University of Southern California Dissertations and Theses

Lactate modulates astrocytic and neuronal plasticity

(USC Thesis Other)

Lactate modulates astrocytic and neuronal plasticity

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Lactate modulates astrocytic and neuronal plasticity

by

Adam Jacob Lundquist

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
Neuroscience

August 2021

Copyright 2021 Adam Jacob Lundquist
ii
Acknowledgements
I would like to thank my advisor, Dr. Michael Jakowec, and co-mentor Dr. Giselle
Petzinger for training me. We knew that we could work well together only a few weeks
into my rotation with the lab, and it is that understanding, confidence and commitment to
each other that has made my lab experience the past five years so insightful and fruitful.
You welcomed me into your team, took a chance on my ideas when our backs were up
against the wall, and trusted me to develop and carry these projects for the lab. I have
experienced awesome highs and tragic lows in your company, finding lessons in every
success and every failure, and always felt at home in the hallway of the annex. Through
your guidance and mentorship, I have learned what it means to be a scientist and how it
is about so much more than the data you generate. Thank you for showing me the
humanity of our science, for allowing me to connect with our patients, and to see how
forward thinking we must be to help others. You and your family have provided
innumerable opportunities to advise, develop, and support me, for which I will forever be
indebted.
Before I arrived in Los Angeles to start the Neuroscience Graduate Program, I was
lucky enough to be trained by two incredible female neuroscientists, whose kindness,
patience, and instruction made graduate school possible. I was first exposed to
neuroscience by Dr. Anna Orr during summers at the Gladstone Institute. Anna’s rigor,
dedication, and humanity sparked a fire for neuroscience in me that I did not know
existed. For this and your unabashed love for astrocytes, which I now proudly carry, I
thank you. It was Anna’s training that prepared me for independent work at North Central
College, where Dr. Margaret Gill took me under her wing and gave me a runway to develop
iii
these scientific and experimental skills as a Richter Scholar. Maggie and I started at North
Central the same year – I as a transfer student, her as a new assistant professor – and I
was humbled to learn from and teach for such an incredible scientist. Maggie exposed
me to the idea of graduate school, going so far as to finesse me a pass to the 2015
Society for Neuroscience meeting in Chicago so I could attend the graduate school fair,
a pivotal moment where graduate school became a real possibility. Maggie balanced
building my confidence with critical feedback on the quality of my writing and my science,
ultimately resulting in a mentor I continue to turn to for advice and perspective. For this
and all the laughs, I thank you.
Thank you to all my colleagues, administrators, and mentors at the University of
Southern California. Thanks to Drs. Enrique Cadenas and Daniel Holschneider for
providing support and critical feedback on my doctoral work. Thanks to Dr. Nick Llewellyn
and Nicolaus Jakowec for your dedication to my project and helping our virus dreams
come true – half of the work in this dissertation could not have been completed in such
a timely manner without your commitment and assistance. Thanks to Dawn Burke,
Deanna Solórzano, Ariana Perez, and Morgan Nagatani for the administrative help and
joyous camaraderie along the way. Thanks to Neuroscience Graduate Program
leadership, including Dr. Pat Levitt, Dr. Judith Hirsch, Dr. Jason Zevin, and Dr. Jeannie
Chen, for helping to advocate for a better graduate experience. Thanks to members of
the Neuroscience Graduate Program for trusting me to serve on your behalf as a member
of the Neuroscience Graduate Forum, and to the fellow board members who shared a
common goal of making graduate school the best it can be for our classmates. Thanks
iv
to members of the 2016 cohort for helping me to feel at home amongst our community
of scientists.
A tremendous amount of thanks goes to my lab mates: Dr. Matthew Halliday, for
being so welcoming and whose idea to have a beer together one Tuesday night helped
spark my dissertation ideas; Ilse Flores, for being so quick with a laugh and willing to
serve as emotional support in the lab; and to Erin Donahue, whose energy is unmatched
and helps keep the lab buzzing with action. I also want to thank my undergraduate
mentees for giving me the chance to train you and trusting in our work together. To Jacky
Parizher, Tyler Gallagher, and Susan Kishi, it has been my distinct pleasure to work with
and mentor you, and see you develop into such strong, hardworking, intelligent, and
successful scientists and professionals.
My time at the University of Southern California has been infinitely enriched by the
support and love of an incredible group of friends within the Neuroscience Graduate
Program. Thanks to Edward Catich, whom I met at recruitment in February 2016 and was
my first and best friend the past 5 years; Eddie, you have been like a brother to me, and I
have a ton of love for you and your friendship. Thanks to Adam Mezher, who was quick
to befriend me when we started together in 2016, kept me grounded and always made
me laugh. His tragic death in the spring of 2018 robbed us of more joy together; I miss
you so much, my friend. Thanks to Alicia Quihuis, Zachary Murdock, Eric Hendricks, Phillip
Maire, Soyoung Choi, and Aida Bareghamyan for so many family dinners and hangouts
and happy hours, filled to the brim with laughs and memories. Your love and care have
been fundamental to my growth and made the trials of graduate school not only bearable,
but so much fun.
v
So much thanks and praise are due for Nora Benavidez, whose love, support, and
honesty have catalyzed tremendous personal and professional growth. I feel incredibly
fortunate to have met and known you, to bear witness to your genius, kindness, and
humility first as a friend and now as your partner. Thank you for being my refuge, my
confidant, my advocate, my bridge to a world I didn’t know could exist, a world we’ve
created together. I am so excited for our next steps after this chapter closes. I love you
so much.
Above all, I would not be where I am without the unending love and encouragement
from my parents, Bruce and Denise Lundquist, and my siblings, Amie and Karl McShane,
Emily Lundquist, Aaron and Madisen Lundquist, Caleb Lundquist, and Molly Lundquist.
There is too little space to detail all you have done to help me develop into the scientist,
and the man, that I am today. I love you and thank you for everything, past, present, and
future.

vi
Table of Contents
Acknowledgements ................................................................................................................ ii
List of Tables ......................................................................................................................... vii
List of Figures ...................................................................................................................... viii
Abstract .................................................................................................................................. xi
Introduction ............................................................................................................................ 1
Chapter 2: Exercise induces region-specific remodeling of astrocyte morphology and
reactive astrocyte gene expression patterns in male mice .............................................. 17
Chapter 3: Exogenous L-lactate promotes astrocyte plasticity but is not sufficient for
enhancing striatal synaptogenesis or motor behavior in mice ......................................... 46
Chapter 4: Knockdown of astrocytic monocarboxylate transporter 4 in the motor cortex
leads to loss of dendritic spines and a deficit in motor learning ...................................... 78
Chapter 5: Loss of astrocytic MCT4 modulates dopamine in the striatum to enhance
behavioral performance .................................................................................................... 116
Discussion and Future Directions ..................................................................................... 153
Bibliography ....................................................................................................................... 168

vii
List of Tables

Table 2.1 – Statistical analysis for morphological analyses in Chapter 2 34
Table 3.1 – List of primers used for qRT-PCR analysis in Chapter 3 66
viii
List of Figures

Figure 1.1 – Structural, genetic, and physiological changes underlie short- and long-term
neuroplastic adaptations. 2
Figure 1.2 – Astrocytes regulate and respond to neuronal insults. 3
Figure 1.3 – Astrocyte complexity increases across complex nervous systems. 4
Figure 1.4 – Cell type-specific sequestration of metabolism. 5
Figure 1.5 – Simplified model of how astrocyte-neuron lactate shuttle contributes to
synaptic plasticity and gene expression changes. 8
Figure 1.6 – Astrocyte-neuron lactate shuttle is regulated by and proceeds through
various mechanisms to promote neuroplasticity. 11
Figure 2.1 – Morphological assessment of GFAP-positive astrocytes. 28
Figure 2.2 – Representative GFAP immunohistochemistry of the three regions sampled.
30
Figure 2.3 – Aerobic exercise causes astrocyte morphological changes in a region- and
time-specific manner. 32
Figure 2.4 – Aerobic exercise does not change astrocyte numbers. 35
Figure 2.5 – Sholl analysis of astrocytes reveals distinct arborization patterns across
regions and exercise duration. 38
Figure 2.6 – Aerobic exercise induces reactive astrocyte gene expression. 40
Figure 3.1 – Primary mouse astrocyte cultures are highly pure. 64
Figure 3.2 – L-lactate and 3,5-DHBA administration to primary astrocytes induces
changes in astrocyte plasticity gene expression. 67
ix
Figure 3.3 – L-lactate administration induces changes in astrocyte plasticity gene
expression in mice. 68
Figure 3.4 – L-lactate administration causes morphological remodeling of striatal
astrocytes, but not cortical astrocytes. 70
Figure 3.5 – L-lactate administration does not increase synaptogenesis in striatum or
ectorhinal cortex in mice. 72
Figure 3.6 – L-lactate administration does not improve motor performance on the
accelerating rotarod. 74
Figure 4.1 – In vitro validation of pSico-EGFP-shMCT4 vector. 98
Figure 4.2 – pSico-EGFP-shMCT4 causes loss of MCT4 expression in C8-D1A-CreERT2
astrocytes. 100
Figure 4.3 – In vivo validation of Lenti-pSico-EGFP-shMCT4 to knockdown astrocytic
MCT4. 102
Figure 4.4 – Astrocyte-specific knockdown of MCT4 impairs motor learning without
affecting locomotion or motor coordination. 104
Figure 4.5 – Lentiviral injection and tamoxifen administration do not impact motor
behavior. 105
Figure 4.6 – Loss of astrocytic MCT4 causes cortical dendritic spine loss. 107
Figure 4.7 – Loss of astrocytic MCT4 in M1 does not affect cortical dendrites in M2. 108
Figure 4.8 – Loss of astrocytic MCT4 decreases postsynaptic protein expression. 109
Figure 4.9 – Motor learning correlates with postsynaptic protein expression. 110
Figure 4.10 – Astrocytic MCT4 knockdown causes task-specific decrease in glucose
uptake in motor-related brain regions. 112
x
Figure 5.1 – Specific knockdown of the lactate transporter MCT4 in striatal astrocytes.
135
Figure 5.2 – Astrocytic MCT4 knockdown improves motor performance. 137
Figure 5.3 – Stereotypy and object recognition improve following astrocytic MCT4
knockdown. 138
Figure 5.4 – Astrocytic MCT4 knockdown does not affect expression of striatal synaptic
proteins. 140
Figure 5.5 – MCT4 knockdown enhances presynaptic dopamine markers in the dorsal
striatum. 141
Figure 5.6 – Astrocytic MCT4 knockdown affects dopamine receptor expression and
canonical signaling products in the dorsal striatum. 142
Figure 5.7 – Unilateral knockdown of astrocytic MCT4 increases tyrosine hydroxylase
immunoreactivity in dorsal striatum. 144
Figure 5.8 – Unilateral knockdown of astrocytic MCT4 drives spontaneous rotational
behavior and heightens sensitivity to amphetamine. 146
Figure 5.9 – Astrocytic MCT4 knockdown increases amphetamine-induced cFos
expression. 148
Figure 5.10 – Astrocytic MCT4 knockdown does not impact dendritic spine density in the
dorsal striatum. 149
Figure 6.1 – Proposed mechanism by which striatal astrocytic lactate may modulate
dopamine. 169
xi
Abstract

Cellular energetics underpin the function and dysfunction of all organisms, and the
adaptation of energy systems is fundamental to the evolution and adaptation necessary
for survival. Evolution and adaptation within the central nervous system is perhaps best
characterized by neuroplasticity. The study of neuroplasticity is a theme our laboratory
has been rigorously engaging with for the past two decades, especially in the context of
neurodegenerative diseases such as Parkinson’s disease (Davies et al., 2017; Petzinger
et al., 2013).
Parkinson’s disease is the second most common neurodegenerative disease and
is characterized by a selective loss of midbrain dopaminergic neurons, whose projections
regulate motor and cognitive behaviors by the release of the neuromodulating chemical
dopamine (Petzinger et al., 2015). In the absence of disease-modifying therapeutics,
current therapies aim to limit dopamine metabolism and replace the lost dopaminergic
tone throughout the brain to curb the severity of symptomology; however, these
pharmacological approaches are well-documented to have severe drawbacks in long-
term therapy and do not slow or reverse disease progression or pathological
development (Girasole et al., 2018). In conjunction with pharmacological therapies,
exercise training is a robust adjuvant capable of improving motor and cognitive function
and may slow overall disease progression (Fisher et al., 2004; Petzinger et al., 2007).
Exercise is well-known to affect several elements important to brain function, including
through increasing neurogenesis, cerebral blood flow, and dendritic synaptogenesis
(Black et al., 1990; Toy et al., 2014; van Praag et al., 1999). In addition, exercise can modify
the function of non-neuronal cells, particularly astrocytes which are necessary supporters
xii
of neuronal health and help to regulate various aspects of neuroplasticity (Bernardi et al.,
2013; Lundquist et al., 2019). Importantly, exercise is a whole-body experience requiring
a coupling of peripheral systems (i.e., musculoskeletal, and cardiorespiratory) to central
nervous system function and is executed by a careful coordination and collaboration by
these distinct yet interconnected systems (Rasmussen et al., 2011).
However, the exact mechanisms that govern the benefits of exercise, or how those
molecular and cellular changes may contribute to the potentially disease-modifying
application of exercise in the context of neurodegenerative diseases like Parkinson’s
disease, remains incompletely known. One potential mechanism may be energetic
changes, perhaps best represented by lactate metabolism. Lactate is a common
byproduct of intensive aerobic exercise and is known to be produced by the periphery and
consumed by the central nervous system, in addition to be produced within the brain by
astrocytes during periods of high-intensity neuronal activity, such as during exercise
(Dalsgaard et al., 2004; Matsui et al., 2017; van Hall et al., 2009). Thus, lactate potentially
links peripheral and central activity during exercise, and can be used to study how
astrocytes contribute to neuroplasticity. To address this gap in knowledge, the studies
described in this dissertation represent a consolidated effort to isolate and examine the
contribution of lactate, both peripherally and centrally derived, in driving neuroplasticity.

Introduction

A central tenant of central nervous system function is neuroplasticity; this phenomenon
explains how the central nervous system can undergo various experiences, ranging from
sensory stimulation to systemic degeneration, and change accordingly to maximize
function in a behaviorally relevant and beneficial manner. For almost 100 years, the
central nervous system’s dynamic nature has been seen as necessary for the stability of
the system at large (Cannon, 1932). Underlying this adaptive response are myriad
physical and molecular remodeling, including a push-and-pull between growth and
destruction of signaling infrastructure, enhancement and diminishment of energetic
pathways, and spatial and temporal specificity of cellular responses. Since Santiago
Ramon y Cajal’s seminal histological work that paved the way for the neuron doctrine in
1888, understanding these adaptive responses has been a principal theme for
neuroscience research. The neuroplastic changes most studied include neuronal
morphological changes (the addition and elimination of synaptic dendrites and outgrowth
of axonal growth cones), electrophysiological changes (long-term potentiation or
depression of neural circuits, expression patterns of ionotropic and metabotropic
neurotransmitter receptors or ion channel localization and function), and behavioral
changes (long-term memory and habit formation) (Fig. 1.1).
However, while classical neuroscience research laid the foundation for understanding
neuron-specific changes that trigger overall changes in central nervous system function,
recent work has resulted in a growing consensus that non-neuronal cells are also capable
of undergoing adaptation in response to experience and their own responses inform and
2
direct the neuron-specific plasticity that is most heavily studied and is behaviorally and
clinically relevant.
Figure 1.1 – Structural, genetic, and physiological changes underlie short- and long-term
neuroplastic adaptations. Effective and efficient synaptic plasticity is carried out through
adaptive processes in neurons, on the timescale of seconds to hours or even days. This process
is simply laid out in this figure, where electrochemical activity causes fluxes in ions, setting of
myriad intracellular cascades that ultimately lead to structural remodeling of dendrites (to build
more connections in this particular circuit), greater insertion of ion channels and neurotransmitter
receptors (typically glutamate receptors, to field inputs on new sprouting), and increases in
transcription and translation of intracellular kinases and transcription factors necessary for
fueling and maintaining the heightened neuroplastic state. Adapted from (Lamprecht & LeDoux,
2004).

The central nervous system is a complex entity, comprised of dozens of cell-types all
uniquely positioned to meaningfully contribute to overall systemic function.
Technological and computational advancements, now utilizing and analyzing single-
nucleus RNA sequencing, continues to expand the atlas of cell types across model
organisms, even developing a whole-body mouse atlas (The Tabula Muris Consortium,
2018); however, further insight of this work can be found elsewhere (Zeisel et al., 2018).
3
For the purposes of this dissertation, cell-type specialization within the central nervous
system can broadly be defined in terms of neuronal and non-neuronal cell-types, namely
glia. While neurons are electrically active and are responsible for information flow
throughout the brain via electrochemical signaling, glia are non-electrically active and are
responsible for the support of neuronal function. Glia can be further subdivided into the
macroglia, including astrocytes and oligodendrocytes, and microglia; the former two cell-
types provide energetic, trophic, and structural support for neurons, while the latter is the
brain-resident macrophage responsible for synapse elimination and immune
surveillance.

Figure 1.2 – Astrocytes regulate and respond to neuronal insults. Astrocyte-neuron interactions
in health, reactivity and disease in the central nervous system. Astrocytes help to regulate
synaptic function as part of the tripartite synapse, molecularly and structurally react to local
insults (inflammation, disease, exercise), and form essential cellular barriers to allow for axonal
healing in severe neuronal damage. Adapted from (Khakh & Sofroniew, 2015).

Astrocytic contributions to neuronal synaptic formation, maintenance, and plasticity
are of particular interest, as astrocytes recycle glutamate to buffer against excitotoxicity,
provide neurotrophic factors necessary for neuroplasticity, and coordinate intercellular
4
metabolism of neuronal waste products and supply energetic substrates such as lactate
(Fig. 1.2) (Allen & Lyons, 2018; Magistretti & Allaman, 2018; Santello et al., 2019).
Astrocytes are morphologically complex cells, contacting thousands of individual
synapses with perisynaptic processes, forming the insulated tripartite synapse (Martín et
al., 2015; Verkhratsky & Nedergaard, 2018); it is these perisynaptic processes that
physically remodel to support synaptic plasticity and carry lactate transporters
responsible for supplying lactate to neuronal synapses (Fig. 1.3) (Pellerin et al., 2005).

Figure 1.3 – Astrocyte complexity increases across complex nervous systems. Morphological
distinction between mouse protoplasmic astrocyte (A) and human protoplasmic astrocyte (B),
which are typically found in the neocortex. The massive increase in astrocyte diameter and
volume across evolution is likely due to increased synaptic demands in higher organisms –
human astrocytes may interact with hundreds of thousands, or even millions of synapses.
Adapted from (Verkhratsky & Nedergaard, 2018).

The supply of lactate from astrocytes to neurons represents a viable strategy for
neurons to meet the high energetic demands of synaptic activity (Attwell & Laughlin,
2001; Harris et al., 2012). Neurons derive most of their energy from mitochondrial
5
respiration, which is evident by the mitochondria capacity in synapses due to the high
rate of ATP production afforded by mitochondrial metabolism (Sheng & Cai, 2012). As
glial cells do not require as much ATP for cellular maintenance relative to neurons,
redistribution of energetic resources such as lactate represent a strategy to scale up local
metabolism in response to increased neuronal activity (Fig. 1.4).

Figure 1.4 – Cell type-specific sequestration of metabolism. Energy production is flexible in the
central nervous system, but can be largely divided into pro-glycolytic cells, such as astrocytes,
and pro-mitochondrial respiration cells, such as neurons. This division of metabolic labor –
astrocytes producing lactate for neurons to metabolize – forms the basis of the astrocyte-neuron
lactate shuttle. Cooperative metabolism may play crucial roles in mediating synaptic plasticity
and behavioral adaptation. Adapted from (Magistretti & Allaman, 2018).

6
Canonically, lactate is produced from pyruvate following the glycolytic catabolism of
glucose under conditions of low oxygen saturation, termed anaerobic respiration (Brooks,
2018). Shunting of pyruvate to lactate in low oxygen situations – away from its entry into
the tricarboxylic acid (TCA) cycle and subsequent oxygen-dependent mitochondrial
respiration – ensures the cell can recycle reducing NADH to NAD
+
to allow glycolysis to
produce a small yet continuous flow of ATP. Within this canonical view a cell in distress
under low oxygen saturation, or a mutated cancerous tumor where limited oxygen flow
heightens glycolytic processes, producing and releasing lactate to recycle NADH is an
effective strategy, particularly in the short term until normal oxygen concentration is
restored (Brooks, 2020). Such metabolic flexibility is key to cellular survival and explains
why organisms can successfully survive highly stressful metabolic situations, such as
fleeing from prey (Brooks, 2018).
However, work over the past several decades has demonstrated that this metabolic
flexibility is not limited to stressful periods of low oxygen concentration and that lactate
production is far from a waste product in the grand scheme of organismal metabolism.
Skeletal muscle – responsible for helping our bodies move through space – are an ideal
model to study lactate metabolism in. Muscle cells undergo periods of high intensity
activity, where local oxygen delivery decreases, and lactate metabolism greatly increases
(Ide et al., 2000; Juel & Halestrap, 1999); such shifts in metabolic output are necessary
for these cells to meet ATP production to maintain the fast-acting role of muscle
contraction which allows for reliable and continuous movement. In fact, very recent work
using radioactive
14
C-containing glucose demonstrated that whole-body metabolism is
driven by circulating levels of lactate, not glucose (Hui et al., 2017); thus, peripheral
7
metabolism of lactate is a highly relevant and effective strategy to meet multiorgan
energetic demands. Mechanisms to shuttle lactate between structures, cells or organs
represent an opportunity to maximize whole-organism ATP production by utilizing
glycolytic byproducts from distinct systems. Such shuttles exist at the level of lactate
shuttling from the cytoplasm to the mitochondria within individual cells, as well as at the
level of metabolic cooperation between different organ systems, such as skeletal
muscles and the heart (Brooks, 2018). Within the central nervous system, where energy
consumption represents 20% of total body metabolism, lactate shuttles represent a
valuable method to continuously supply energetic substrates to metabolically demanding
neurons.
Lactate shuttling, particularly between astrocytes and neurons, was first observed and
proposed nearly 30 years ago first in culture systems (Pellerin & Magistretti, 1994), then
further confirmed in vitro and in vivo using chemical manipulations of brain lactate
production (Bittar et al., 1996; Gibbs et al., 2006; Pellerin et al., 1998). The astrocyte-
neuron lactate shuttle posits that in response to increased neuronal activity and
glutamate release, astrocytes recycle glutamate (metabolizing it to glutamine for
shuttling back to neurons) in an ATP-dependent manner. In response to this energetic
demand, astrocytes upregulate glycolytic processes, either through glycogenolysis
(catabolism of glycogen granules within astrocytes) or glucose uptake and produce
pyruvate and ATP (Fig. 1.5) (Alberini et al., 2018).
To recycle necessary glycolytic intermediates, astrocytes convert pyruvate to lactate
which is subsequently shuttled out of astrocytes via monocarboxylate transporter 4
(MCT4) and shuttled into neurons via MCT2, where it can be converted back to pyruvate
8
and sent for mitochondrial respiration to produce the elevated ATP necessary for
synaptic activity (Barros, 2013).

Figure 1.5 – Simplified model of how astrocyte-neuron lactate shuttle contributes to synaptic
plasticity and gene expression changes. Astrocytes, positioned with peri-synaptic and peri-
vascular endfeet, are able to regulate glucose uptake, lactate metabolism, and shuttle to neurons
to fuel neurons during high synaptic activity. This figure also points out the important, and well-
studied, contribution of astrocytic glycogen. Glycogen represents a fast-acting, short term tool
for astrocytes to meet energetic demand before blood flow responds to local brain regions,
supplying more glucose for lactate production. Adapted from (Alberini et al., 2018).

Since the introduction of this cooperative metabolic hypothesis, dozens of papers
from labs across the world have further explored, detailed, and refined the movement of
lactate from astrocytes to neurons. Genetic analysis of metabolic enzymes revealed that
astrocytes are predominately glycolytic, producing high levels of lactate even in normal
oxygen conditions (Bittar et al., 1996). Over time, technological advancements in in vivo
imaging and fluorescent indicators have allowed for the real-time imaging of lactate flow
9
from astrocytes to neurons, visualizing this intrinsic metabolic gradient (Mächler et al.,
2016). Methods to disrupt this energetic connection via lactate, using various RNA
interference targeting astrocytic and neuronal MCTs, and pharmacological approaches
have demonstrated the necessity of lactate shuttling from astrocytes to neurons in
facilitating several behavioral types, particularly focusing on hippocampus-mediated
behaviors (Descalzi et al., 2019; Harris et al., 2019; Murphy-Royal et al., 2020;
Netzahualcoyotzi & Pellerin, 2020; Suzuki et al., 2011; Wang et al., 2017). Additionally,
lactate shuttling has been extended to regulation of other metabolic processes, such as
the recycling of oxidated lipid droplets (Liu et al., 2017) and has been implicated in the
formation of neurotoxic astrocyte subtypes in response to central nervous system injury
(Wheeler et al., 2020). Elsewhere, several groups have connected lactate to intracellular
signaling pathways, facilitated through the Gi/o-protein coupled hydroxycarboxylic acid
receptor 1 (HCAR1); lactate- and HCAR1-mediated signaling has been shown to be
involved in brain angiogenesis, neuronal excitability, and neurotransmitter release (de
Castro Abrantes et al., 2019; Morland et al., 2017; Tang et al., 2014). Beyond astrocytes,
lactate shuttling between oligodendrocytes (or their peripheral nervous system analogs
Schwann cells) and neuronal axons, which the former insulates with fatty myelin sheaths
to enhance electrical signaling for long-range neuronal projections, has been shown to be
important for maintaining axonal structure and function (Lee et al., 2012).
While these discoveries provide evidence for the obligation of lactate shuttling in
maintaining several properties of neuronal function contributing to synaptic plasticity and
behavioral adaptation, evidence also exists disproving the necessity of lactate shuttling
in the brain under certain circumstances (Díaz‐García & Yellen, 2019; Dienel, 2018; Dienel
10
& Cruz, 2016). Part of the discrepancies regarding the reality of the lactate shuttle in
supplying neuronal energy may be due in part to differences in the brain region studied
and experimental manipulations; a vast majority of work characterizing lactate shuttling
has focused on in vivo hippocampal lactate shuttling, while work disproving it has been
focused elsewhere, such as the primary visual cortex or in ex vivo slice cultures (Díaz-
García et al., 2017). Thus, there is a need within the field of neuroscience to understand
how such energetics contribute to neuronal function, neuroplasticity, and behavioral
adaptation, and whether lactate shuttling is universal to diverse brain regions and region-
specific behavioral tasks.
Despite ongoing tensions to resolve the debate surrounding lactate’s utility in the
brain, lactate as a fuel source represents an adaptation to high intensity neuronal activity.
To meet circuit-specific energetic demands in response to challenging stimuli or
experience, the brain is likely capable of switching to diverse metabolic pathways to meet
systemic demands. Exercise represents an energetic challenge to the body and the brain,
which evolutionarily has been necessary for the foraging of food, escape from predators,
and other evolutionarily relevant behaviors. Therefore, the brain should be aligned
energetically with the body, utilizing all resources as thoroughly as possible to produce
immediate success in behaviors, resulting in long-term neuroplastic adaptations that
define classical neuroplasticity and the brain’s true function, to be plastic and undergo
experience-dependent change to practice habits, remember important spatial and
temporal details, and so forth.
Thus, neuroenergetics as a lens through which to study plasticity is a valuable
perspective, providing insights into the fundamental building blocks for cellular function,
11
which are carried out in cell-type specific manners, adapted to evolve, and change in a
specific manner in response to specific stimuli. The work presented in this dissertation
focuses this lens on the idea that lactate acts as a critical mediator of the nervous
system’s own evolution – neuroplasticity – which happens on the timescale of days and
weeks, both at rest and in response to exercise. The chapters presented herein provide
evidence that: one, the brain undergoes region-specific adaptations in astrocytic
structure and function (astrocytic plasticity) in response to exercise that are relevant for
neuronal plasticity; two, peripheral lactate promotes this conserved astrocytic plasticity
response, highlighting a potential body-brain connection in the context of exercise (a high-
lactate state); three, lactate derived from astrocytes is critical in maintaining neuronal
and synaptic function, and its absence results in clear impairments in motor learning; and
four, astrocytic lactate is capable of modulating neurotransmitter activity and likely plays
an important role in aligning and scaling metabolism with synaptic activity (Fig. 1.6).

Figure 1.6 – Astrocyte-neuron lactate shuttle is regulated by and proceeds through various
mechanisms to promote neuroplasticity. Adapted from (Magistretti & Allaman, 2018).
12
Chapter 2 of this dissertation starts by examining how exercise impacts astrocytic
function throughout the adult mouse brain, revealing fundamental differences in genetic
and physical astrocytic response, depending on the brain region studied, that are later
explored in Chapters 4 and 5. The experiments described in Chapter 2 focus on the
structural plasticity of astrocytes, a well-characterized response of astrocytic plasticity
that is necessary for synaptic plasticity to occur in neurons (Bernardinelli et al., 2014).
They additionally examine changes in the expression of astrocyte-specific genes that
may reveal astrocytic function in the context of supporting neuroplasticity in exercise
(Zamanian et al., 2012). This work demonstrated that there is a temporal- and region-
specific response in astrocytic structure, providing clarity to divergent results in previous
experiments studying astrocytic structural change following exercise, and disproved a
widely touted idea that astrocytes multiply in response to exercise, a phenomenon termed
‘gliogenesis’. Additionally, genetic analysis showcased a similarly region-specific
astrocytic response to exercise that highlighted the beneficial response of astrocytes to
exercise to support neuroplasticity.
Chapter 3 of this dissertation uses the findings from Chapter 2 as a framework for
investigating the potential role of peripheral, brain-exogenous lactate in driving the effects
of exercise on astrocytic and neuronal plasticity. Chapter 3 uses in vitro experiments to
focus first on the impact of lactate – as a metabolic substrate and a signaling molecule
– on astrocytic function, then examines whether these astrocyte-specific changes are
recapitulated in vivo when mice receive injections of lactate that mimic blood lactate
levels experienced during previous exercise studies. By combining cell culture and animal
experiments, this work established that exogenous lactate is capable of enacting
13
astrocytic changes that mirror structural and genetic changes seen in previous exercise
studies (including Chapter 2) but does not have any overt effects on neuronal function,
studied through behavioral paradigms and synaptic histological analysis. This work
extended previous efforts to delineate how exercise modifies increased cerebral blood
flow – which was found to be mediated through lactate signaling (Morland et al., 2017)
– and established lactate as a promising exercise mimetic capable of promoting
astrocyte plasticity to effectively prime neurons to undergo robust neuroplasticity in the
event of heightened neuronal activity. Such results presented a mechanistic link tying
peripheral metabolism to central nervous system plasticity and revealed that astrocytic
priming and plasticity – known to contribute to neuroplasticity – may be upstream of
neuronal plasticity.
Chapter 4 of this dissertation turns the focus from peripheral lactate to astrocyte-
derived lactate and its subsequent shuttling between astrocytes and neurons, which
plays an important role in learning and memory in the hippocampus (Harris et al., 2019;
Herrera-López et al., 2020; Suzuki et al., 2011) and has not been explored in the context
of motor behaviors. The focus of studies presented in Chapter 3 was to define how
astrocytic lactate contributes to synaptic plasticity and motor behavior in the motor
cortex and better understand the energetic demands of motor behaviors which are
significantly impaired in Parkinson’s disease. The experiments in Chapter 4 combine
inducible Cre-lox recombination with microinjections of lentivirus containing Cre-
dependent short-hairpin RNA (shRNA) expression plasmids to knockdown a lactate
transporter, monocarboxylate transporter 4 (MCT4) specifically in the mouse motor
cortex. In doing so, one can study the role of astrocytic lactate in a time-, region-, and cell
14
type-specific manner. This experimental framework is supplemented by motor behavior
paradigms, high-resolution histological analysis, western immunoblotting, and task-
specific mapping of neuronal activity using a near-infrared analog of 2-deoxyglucose to
understand how astrocytic lactate shuttling contributes to neuronal activity during a
motor task. Using this multipronged approach, I first developed and validated the shRNA
viral approach in a newly constructed astrocyte cell line in vitro, then applied the viral
manipulation in vivo to understand how astrocytic MCT4 contributes to synaptic
structure, neuronal function, and motor behaviors. Specific loss of astrocytic MCT4 in the
motor cortex led to a dramatic loss of neuronal synapses, particularly in the postsynaptic
dendritic compartment, and this synaptic loss manifested in a specific motor learning
deficit. Furthermore, energetic deficits in the motor cortex were found to propagate
throughout motor-related regions, demonstrating that energy deficits in one node can
affect the activity of other connected regions, which may explain observed motor
behavior shortcomings. Chapter 4 represents a major step forward in understanding how
lactate contributes to neuroplasticity. The results of this set of experiments are the first
demonstration that loss of lactate flow from astrocytes to neurons is harmful to synaptic
maintenance in the cortex and specifically causes motor learning deficits, reaffirming the
probable importance of lactate in fueling and contributing to neuroplasticity.
Chapter 5 of this dissertation extends the approach developed and validated in the
motor cortex in Chapter 4 to its application in the dorsal striatum. The dorsal striatum
receives major dopaminergic innervation from the midbrain, particularly the substantia
nigra which degenerates selectively in Parkinson’s disease, and controls several distinct
motor and cognitive behaviors (Petzinger et al., 2013). Thus, the purpose of these
15
experiments was to understand how lactate contributes to striatal synaptic and
dopaminergic function in the hopes of gaining baseline knowledge for use in the context
of dopamine depletion animal models of Parkinson’s disease. Using the same approach
as described in Chapter 4, temporal-, region- and cell type-specific knockdown of
astrocytic MCT4 is achieved in the dorsal striatum and motor and cognitive behaviors
were used to guide subsequent molecular investigations. Bilateral knockdown of MCT4
in the dorsal striatum resulted in an enhancement of motor and cognitive behaviors and
had no effect on striatal synaptic integrity; in fact, the only impact of MCT4 knockdown
on striatal microcircuitry was evidenced by an increase in presynaptic dopamine,
assessed both by high-performance liquid chromatography and western immunoblotting.
Unilateral knockdown of striatal MCT4 allowed for the application of classical
pharmacological investigations of the dopaminergic system to ascertain how dopamine
may be changing in response to a loss of striatal astrocytic MCT4. Apomorphine
administration, which agonizes dopamine receptors on striatal spiny projection neurons,
had no effect on behavior; alternatively, amphetamine administration, which causes
massive presynaptic dopamine efflux and blocks its reuptake (Björklund & Dunnett,
2019), caused a significant increase in rotational behavior, affirming previous findings at
the level of protein expression, neurochemistry and behaviors modulated by dopamine.
The results of Chapter 5 are groundbreaking, as only one other group has described an
effect of lactate in modulating neurotransmitter levels (Tang et al., 2014) and lactate is
not thought to impact dopamine function. These findings represent another novel
demonstration in which lactate may affect neuroplasticity and could have implications
for how exercise is capable of positively modifying behavioral outcomes in psychiatric
16
disorders with heightened dopamine levels (Gorczynski & Faulkner, 2010; Jouroukhin et
al., 2018).
Taken together, the chapters contained within this dissertation represent a
multimodal approach to understanding how the potential mechanism of lactate
metabolism and signaling contributes to astrocytic and neuronal plasticity. Using in vitro
and in vivo models, combined with cell-type specific genetic manipulations, various
behavioral paradigms, pharmacology and diverse molecular and genetic readouts, the
experiments discussed 1) delineate and clarify region-specific astrocytic responses to
exercise before narrowing the focus to 2) understand how peripheral-derived lactate is
capable of accounting for this variability of astrocytic and neuronal plasticity in exercise,
and 3) how region-specificity further defines the contributions of astrocytic lactate to
neuronal plasticity. Despite our laboratory’s history of working in animal models of
neurodegeneration, all this work was carried out in healthy cells and mice. This was
intentional in the design of these experiments; by using a template of ‘normal’ astrocytes,
I believed I could better leverage and understand the genetic, pharmacological, and
behavioral manipulations necessary to ascertain how lactate contributes to astrocytic
and neuronal plasticity. Indeed, these findings represent new insights for understanding
how cellular metabolism is linked to neuroplasticity, how the body and brain are linked
via lactate, and how lactate plays divergent roles in two adjacent and functionally related
brain regions. The results detailed in this dissertation, thus, serve as a guidepost for
future studies concerned with the neuroenergetics and astrocytic contributions to
neuroplasticity carried out in the context of dopamine-depleted animal models of
Parkinson’s disease.
17
Chapter 2: Exercise induces region-specific remodeling of
astrocyte morphology and reactive astrocyte gene expression
patterns in male mice
Adapted from Lundquist et al., 2019. Journal of Neuroscience Research, 97 (9), 1081-
1094.

Abstract
Astrocytes are essential mediators of many aspects of synaptic transmission and
neuroplasticity. Exercise has been demonstrated to induce neuroplasticity and synaptic
remodeling, such as through mediating neurorehabilitation in animal models of
neurodegeneration. However, the effects of exercise on astrocytic function, and how such
changes may be relevant to neuroplasticity remain unclear. Here, we show that exercise
remodels astrocytes in an exercise- and region-dependent manner as measured by GFAP
and SOX9 immunohistochemistry and morphological analysis in male mice. Additionally,
qRT-PCR analysis of reactive astrocyte gene expression showed an exercise-induced
elevation in brain regions known to be activated by exercise. Taken together, these data
demonstrate that exercise actively modifies astrocyte morphology and drives changes in
astrocyte gene expression and suggest that astrocytes may be a central component to
exercise-induced neuroplasticity and neurorehabilitation.

18
Introduction
Once thought to be limited to a structural and developmental role, glia are now
known to be involved in maintaining brain homeostasis and actively engaging in
experience-dependent neuroplasticity (Khakh & Sofroniew, 2015). Astrocytes are a
diverse class of glia found throughout the mammalian brain. They play a central role in
several important processes including establishing and regulating the blood brain barrier,
modulating blood flow, regulating synaptogenesis and physical remodeling of synapses,
and participating in neuronal bioenergetics by providing key substrates for metabolism
(Abbott et al., 2006; Chung et al., 2013; Eroglu & Barres, 2010; Suzuki et al., 2011). Such
synergism between astrocytes and the neurons they contact reflects the close
relationship that exists between these two cell types in regulating synaptic activity.
Neuroplasticity can simply be defined as changes in synaptic strength and
structure in response to experience (Petzinger et al., 2015). One form of experience that
can initiate neuroplasticity is physical activity or exercise. Studies from our laboratories
have shown that exercise, in the form of intensive treadmill running, can lead to regional
and circuit-specific changes in synaptic connectivity, blood flow and behavior in normal
rodents as well as neurotoxin-induced and genetic models of neurodegenerative
disorders (Petzinger et al., 2013; Wang et al., 2013). Furthermore, others have shown that
exercise leads to changes in astrocyte function, including glutamate-glutamine recycling
(Bernardi et al., 2013), aquaporin-4 expression (Brockett et al., 2015), and hippocampal
proliferation (de Senna et al., 2017). Such findings clearly implicate a role for astrocytes
in mediating exercise-dependent plasticity.
19
The purpose of this study was to build upon our findings on the mechanisms of
neuroplasticity with exercise and to explore the potential role that astrocytes may play.
Interestingly, studies have shown that exercise can cause astrocytic response, termed
astrogliosis, based on the elevation of glial fibrillary acidic protein (GFAP) expression (Li
et al., 2005), while others reporting its suppression (Bernardi et al., 2013). Such
differences in the characterization of the astrocytic response to exercise may be partially
explained by astrocyte heterogeneity (Khakh & Sofroniew, 2015), differences in the type
of exercise intervention, including intensity and type (voluntary vs. forced) (Kinni et al.,
2011), or by strain- or species-dependent difference in exercise ability (Billat et al., 2005).
In this study, we utilized GFAP and SOX9 (SRY-related HMG-box gene 9, an astrocyte-
specific nuclear marker) immunohistochemistry to determine changes in astrocyte
morphology and proliferation in regions of the healthy mouse brain that display
neuroplasticity in response to running on a motorized treadmill. Selection of brain regions
of interest was based on our previous studies mapping exercise-induced changes in
regional cerebral blood flow (rCBF), and synaptogenesis and synaptic plasticity (Kintz et
al., 2013; Wang et al., 2013). Furthermore, we used quantitative RT-PCR to examine
astrocyte-specific transcript expression of genes involved in metabolic responses to
exercise. Together, these studies aimed to determine if exercise leads to changes in
astrocyte morphology, proliferation and gene expression; and to understand if such
changes are associated with region specific responses to neuroplasticity, or simply
reflect generalized and global changes in the brain.

20
Methods
Mice
Male, 8-10 week old C57BL/6J mice (n=28, The Jackson Laboratory, Bar Harbor,
ME) were used for this study and were housed at University of Southern California
vivarium. Female mice were not used, as estradiol (E2) has been shown to decrease GFAP
expression and astrocytic morphology (Rozovsky et al., 2002). All procedures were
approved by the Institutional Animal Care and Use Committee of the University of
Southern California and conducted in accordance with the National Research Council’s
Guide for the Care and Use of Laboratory Animals (Committee for the Update of the Guide
for the Care and Use of Laboratory Animals; National Research Council, 2010). Mice were
housed in groups of 4 mice per cage and maintained on a reverse 12-hour light/dark cycle
(lights off at 0700) with ad libitum access to food and water.
Animal Groups and Treadmill Exercise
Mice were randomly divided into four groups: sedentary (n=8), and exercise for one
(n=8), two (n=8), or four (n=4) weeks. Mice were exercised on motorized treadmills (EXER-
6M, Columbus Instruments, Columbus, OH) for 1 hour/day, five days/week as previously
described (Fisher et al., 2004) with slight modification. Briefly, mice begin in a warm-up
phase, consisting of increasing speeds over the span of 15 minutes, before progressing
to high speed running for 30 minutes and finishing with decreasing speed over the final
15 minutes. Sedentary animals were placed on a stationary treadmill adjacent to the
exercising mice. The maximum speed for intensive running increased over time, reaching
12m/min at the end of one week, and reaching 17.5 m/min at the end of four weeks.
21
Brain Tissue Collection
Brain tissue was collected after the final exercise session at the end of one week
of exercise (including sedentary mice), two weeks of exercise, and four weeks of exercise.
One hour after the final exercise session, half of each group (n = 4 mice) was sedated
with intraperitoneal injections of tribromoethanol (250 mg/kg body weight) and assessed
for lack of toe-pinch response before being transcardially perfused with 50 ml of ice-cold
0.9% saline followed by 100 ml of ice-cold 4% paraformaldehyde in phosphate buffered
saline, pH 7.2 (PFA-PBS). Whole brains were extracted and placed in 4% PFA-PBS at 4°C
for 24 hours, followed by sinking in 20% sucrose. Whole brains were rapidly frozen by
submersion in 2-methylbutane cooled on dry ice and stored at -80°C until use.
Immunohistochemical Analysis of Astrocytic Morphology
Astrocyte morphology was assessed using GFAP-stained sections. Whole brains
were coronally sliced (30µm thickness) on a sliding cryostat (Leica CM1900, Leica
Microsystems, Wetzlar, Germany). Briefly, sections were washed in Tris-buffered saline,
pH 7.2, with 0.2% Triton X-100 (TBST) for 1 hour at room temperature, blocked in 4%
normal goat serum (NGS, S-1000, Vector Labs, Burlingame, CA) in TBST for 2 hours at
room temperature, and incubated overnight at 4°C in primary antibody (2% NGS in TBST)
with rabbit anti-GFAP (1:2000, Agilent Cat# Z0334, RRID: AB_10013382). Sections were
washed with TBST (3x30 min) and incubated in secondary antibody (2% NGS in TBST)
with Alexa 568-conjugated goat anti-rabbit (1:5000, Thermo Fisher Scientific, Cat# A-
11011, RRID: AB_143157) for 90 minutes. Sections were washed with TBST (3x30 min),
mounted onto gelatin-coated slides, and coverslipped (Vectashield Hardset Antifade with
22
DAPI; H-1500, Vector Labs, Burlingame, CA). Confocal images were taken on an IXB-DSU
spinning disk Olympus BX-61 (Olympus America, Melville, NY) and captured with an
ORCA-R2 digital CCD camera (Hamamatsu, Bridgewater, NJ) and MetaMorph Advanced
software (Molecular Devices, San Jose, CA).
Immunohistochemical Analysis of Astrocyte Number
Astrocyte number in regions of interest was assessed using SOX9-stained
sections (Sun et al., 2017). Whole brains were sliced as above, washed with TBST for 1
hour at room temperature, blocked in 4% NGS in TBST for 2 hours at room temperature,
and incubated overnight at 4°C in primary antibody (2% NGS in TBST) with rabbit anti-
SOX9 (1:2000, Millipore Cat# AB5535, RRID: AB_2239761). Sections were washed with
TBST (3x30 min) and incubated in secondary antibody (2% NGS in TBST) with Alexa 568-
conjugated goat anti-rabbit (1:5000, Thermo Fisher Scientific, Cat# A-11011, RRID:
AB_143157) for 90 minutes. Sections were washed with TBST (3x30 min), mounted onto
gelatin-coated slides, and coverslipped (Vectashield Hardset Antifade with DAPI, Vector
Labs).
For SOX9-GFAP co-localization, sections were washed with TBST as before,
blocked in 4% normal donkey serum (NDS, S30, EMD Millipore) in TBST for 2 hours at
room temperature, and incubated overnight at 4°C in primary antibody (2% NDS in TBST)
with rabbit anti-SOX9 (1:2000, Millipore Cat# AB5535, RRID:AB_2239761). Sections were
washed with TBST as before and incubated in secondary antibody (2% NDS in TBST) with
Alexa 568-conjugated donkey anti-rabbit (1:5000, Thermo Fisher Scientific Cat# A10042,
RRID:AB_2534017). Sections were washed as before, and staining protocol was repeated
23
with blocking (4% NDS in TBST) and overnight incubation at 4°C with primary antibody
(2% NDS in TBST) with rabbit anti-GFAP (1:2000, Agilent Cat# Z0334, RRID:
AB_10013382). Sections were washed as before and incubated in secondary antibody
(2% NDS in TBST) with Alexa 488-conjugated donkey anti-rabbit (1:5000, Thermo Fisher
Scientific Cat# A-21206, RRID:AB_2535792). Sections were washed as before, mounted
onto gelatin-coated sildes and coverslipped (Vectashield Hardset Antifade with DAPI,
Vector Labs).
Morphological Analysis of Astrocytic Structure
GFAP-positive astrocytes were morphologically analyzed for corrected total
cellular fluorescence (CTCF, a measure of fluorescence intensity), distal to proximal
length, and number of primary processes. Astrocytes were imaged at 10x magnification
(UPlan Fl, Olympus) and analyzed in three regions of the brain: i) prefrontal cortex (PFC,
Bregma +2.0 to +1.6 mm A.P., ±0.5 mm M.L., -1.5 to -3.0 mm D.V.); ii) striatum (STR, +1.4
to -0.2 mm A.P., ±0.8 to 2.5 mm M.L., -2.5 to -4.5 mm D.V.); and iii) ectorhinal cortex (ETC,
-1.3 to -3.0 mm A.P., ±3.2 to 4.2 mm M.L., -3.0 to -3.5 mm D.V.). Selection of these
anatomical regions was based upon previous findings of exercise-induced changes in
regional blood flow in the PFC and STR but not in the ETC (Guo et al., 2017; Wang et al.,
2013). Astrocytic area was determined by CTCF according to previously published
guidelines (Burgess et al., 2010), with modifications. Intact astrocytes were manually
traced and total area and integrated density determined and measured in Fiji (NIH, ver.
1.52b) (Schindelin et al., 2012). Mean gray values for background adjacent to measured
cells was recorded and CTCF calculated as: integrated density – (average background
24
mean gray value x area of astrocyte). Distance between the most distal and proximal
process was measured and reported in microns. The number of primary processes
emanating from the soma was also counted. Morphological parameters measured by
CTCF were captured in arbitrary units based upon the above calculation. At least three
tissue sections through each anatomical region were analyzed per animal (n=4 mice per
group).
Unbiased Counting of Astrocytes
SOX9-positive nuclei were analyzed using an unbiased counting approach. SOX9-
positive nuclei were imaged at 10x magnification (UPlan Fl, Olympus) with identical
camera settings in the same three regions previously described (PFC, STR, ECT). A grid
of 300 μm x 300 μm squares was overlaid on each image and SOX9-positive nuclei were
manually counted in five nonadjacent squares, averaged across sections, and the percent
change relative to sedentary was calculated. At least three sections through each
anatomical region were analyzed per animal (n=4 mice per group).
Sholl Analysis of Astrocyte Arborization
Astrocytic arborization (a measurement of process ramification and reactivity)
was determined using Sholl analysis (Sholl, 1953). Individual, GFAP-positive astrocytes
were randomly selected from areas of interest (including PFC, STR, and ETC as delineated
above) and imaged using a 60x water-immersion objective (UplanSApo, Olympus) in z-
stacks at 0.5 µm intervals. Images were manually traced using the Simple Neurite Tracer
plug-in for Fiji before segments were max z-projected and measured using the Sholl
Analysis plug-in (Ferreira et al., 2014). Forty concentric circles (with increasing radii in
25
steps of 2 µm) centered around the soma were placed on top of segmented astrocytes
and the number of intersections across each circle recorded. Data were plotted, as the
distance from soma vs. number of intersections, for comparisons including area under
the curve. For representative astrocytes, segments were filled using the Simple Neurite
Tracer plug-in with manually selected thresholds. At least three tissue sections through
each anatomical region were analyzed per animal (n=4 mice per group).
Quantitative RT-PCR for Astrocytic Genes of Interest
To explore astrocytic activation, we examined the pattern of expression of several
genes including Gfap (glial fibrillary acidic protein, Gene ID 14580), Thbs2
(thrombospondin 2, Gene ID 21826), Lif (leukemia inhibitory factor, Gene ID 16878), and
Il6 (interleukin 6, Gene ID 16193). Immediately after the final exercise session, half of
each exercise group (sedentary, one week and two weeks exercise; n = 4 mice per group)
were sacrificed via cervical dislocation and whole brains were extracted. Fresh tissue was
rapidly microdissected in blocks from i) PFC (Bregma +2.0 to +1.4 mm A.P., rostral to
corpus callosum; ± 1 mm M.L. from midline to the corpus callosum, and -1.5 to -3.0 mm
D.V.) (Natalie Kintz et al., 2017); ii) STR (Bregma +1.2 to -0.2mm A.P., including tissue
bordered ventrally by the anterior commissure, dorsally by the corpus callosum, medially
by the lateral ventricle, and ±2.5 mm laterally from the midline) (Kintz et al., 2013); and iii)
ETC (Bregma -2.0 to -3.0 mm A.P., laterally bounded by the edge of the cortex and
extending 1mm medially to the edge of the striatum, and -3.0 to -3.5 mm D.V.) Tissues
were placed in RNAlater Stabilization Solution (QIAGEN, Germantown, MD) and stored at
4°C overnight. RNA was extracted using RNEasy Mini Kit (QIAGEN, Germantown, MD)
26
according to manufacturer’s guidelines for animal tissue, with an additional chloroform
extraction step. Total RNA concentration was measured by absorbance spectroscopy
(BioPhotometer, Eppendorf, Hauppauge, NY). Complimentary DNA (cDNA) was
generated by reverse transcription from 1 µg of sample RNA using QuantiTect Reverse
Transcription Kit (QIAGEN, Germantown, MD) following manufacturer’s guidelines. qRT-
PCR was run with 1 µl cDNA and QuantiTect SYBR Green (QIAGEN, Germantown, MD) on
an Eppendorf Mastercycler Ep Realplex (Eppendorf, Hauppauge, NY) using a program of
15 min at 95°C, followed by 40 cycles of 15 seconds at 94°C, 30 seconds at 55°C, and 30
seconds at 72°C. Following completion of cycling, a melting curve of products was
generated. Data was collected and normalized on Eppendorf Realplex ep software
(Eppendorf, Hauppauge, NY). Standard ΔΔCT analysis (Livak & Schmittgen, 2001) was
used to quantify fold changes in gene expression in exercise groups normalized to
sedentary controls, with Actb serving as a housekeeping gene. The primers used are as
follows (5’3’): Actb forward GGCTGTATTCCCCTCCATCG; Actb reverse
CCAGTTGGTAACAATGCCATG; Gfap forward CGGAGACGCATCACCTCTG; Gfap reverse
AGGGAGTGGAGGAGTCATTCG; Il6 forward TAGTCCTTCCTACCCCAATTTCC; Il6 reverse
TTGGTCCTTAGCCACTCCTTC; Lif forward ATTGTGCCCTTACTGCTGCTG; Lif reverse
GCCAGTTGATTCTTGATCTGGT; Thbs2 forward CTGGGCATAGGGCCAAGAG; Thbs2
reverse GCTTGACAATCCTGTTGAGATCA.
Statistical Analysis
All statistical tests were carried out and graphs made using Prism 8.0 (GraphPad,
San Diego, CA), with statistical significance set at p < 0.05. For morphological and cell
27
counting measurements, area under the curve, and qRT-PCR analysis, one-way ANOVA
with Dunnett’s multiple comparisons was used with the sedentary group serving as a
control. For Sholl analysis, two-way ANOVA with Dunnett’s multiple comparisons was
used with the sedentary group serving as a control.

Figure 2.1 – Morphological assessment of GFAP-positive astrocytes. Prefrontal cortex, striatum,
and ectorhinal cortex (red-shaded regions, with anterior-posterior range given relative to Bregma)
were included in morphological analysis. Examples of the morphological metrics are shown on a
representative astrocyte, including corrected total cellular fluorescence (CTCF), distal-proximal
distance, number of primary processes, and Sholl analysis.

28
Results
The overall analysis approach used in these studies is shown in Figure 2.1. The
upper row indicates the anatomical regions and Bregma range from which brain tissues
were collected including the prefrontal cortex (PFC), striatum (STR), and ectorhinal cortex
(ETC). The middle panel shows a representative Z-stack of a GFAP-positive astrocyte.
The bottom panel outlines the measurement criteria including corrected total cellular
fluorescence (CTCF, arbitrary units) of GFAP, maximal distal-proximal measurement of a
single astrocyte (microns), the Sholl analysis of the astrocytic range of influence (number
of intersections), and the number of primary processes that emerge from the cell body.
A low and high magnification of representative GFAP-positive astrocytic
immunohistochemical staining of sections through the selected regions of interest
corresponding to the PFC, STR, and ETC are shown in Figure 2.2. A summary of
morphological metrics (CTCF, distal-to-proximal distance, and number of primary
processes; mean ± SEM) can be found in Table 2.1.
29
Figure 2.2 – Representative GFAP immunohistochemistry of the three regions sampled. Left
hand column shows low-magnification images of prefrontal cortex, striatum, and ectorhinal
cortex (a, c, and e); right hand column shows corresponding high-magnification images of
boxes area. Scale bars: 200μm.

30
Exercise Increases the GFAP Fluorescence of Astrocytes
In the PFC, exercise resulted in a statistically significant increase in CTCF values
(arbitrary units, au) with exercise (n = 31 to 49 cells, F(3, 165) = 11.27, p < 0.001). CTCF
values statistically significantly increased with one week of exercise (52% increase, p <
0.001) but not with two or four weeks of exercise (p = 0.986 and p = 0.233, respectively)
compared to sedentary controls.
In the STR, CTCF statistically significantly changed with exercise (n=86-146 cells,
F(3, 479) = 38.98, p < 0.001). Specifically, CTCF statistically significantly decreased at one
week of exercise (34% decrease, p < 0.001) and significantly increased at four weeks of
exercise (31% increase, p < 0.001) relative to sedentary animals, and did not significantly
differ with two weeks of exercise (p = 0.495).
In the ETC, CTCF was significantly affected by exercise (n = 25 to 55 cells, F(3, 181)
= 10.03, p < 0.001). Specifically, CTCF increased significantly with one week of exercise
(92% increase, p < 0.001) and four weeks of exercise (64% increase, p = 0.0102) and did
not significantly differ from sedentary animals with two weeks of exercise (p > 0.999)
(Figure 2.3A).

31
Figure 2.3 –Aerobic exercise causes astrocyte morphological changes in a region- and
time-specific manner. (a) Corrected total cellular fluorescence (CTCF) for intact, GFAP-
positive astrocytes following 1, 2, or 4 weeks of exercise across regions of interest and
measured in arbitrary fluorescent units. (b) Distal-proximal distance as measured from
astrocytes quantified previously and measured in microns. (c) Number of primary
processes projecting from astrocytes quantified previously. (d) Representative traces of
prefrontal cortex, striatum, and ectorhinal cortex astrocytes (60x magnification) following
1, 2, or 4 weeks of exercise. Scale bar: 20μm. n = 4 mice per time points; mean ± SD with
all data points in gray. One-way ANOVA with Dunnett’s multiple comparisons. * p < 0.05;
** p < 0.01; *** p < 0.001 relative to sedentary control.

a b c
d
32

Exercise Modifies the Distal-Proximal Distance of Astrocyte Processes
Distal-proximal (D-P) distances in GFAP-positive astrocytes reflected the findings
in the CTCF measurements (Figure 2.3B).
In the PFC, exercise had a statistically significantly effect on D-P measurements
(n = 30 to 52 cells, F(3, 149) = 5.97, p < 0.001). Specifically, one week of exercise
significantly increased D-P distance (23% increase, p < 0.001) but two and four weeks of
exercise had no significant effect on D-P distance (p = 0.531 and p > 0.999, respectively)
compared to sedentary animals.
In the STR, exercise also statistically significantly changed D-P distance (n = 35 to
86 cells, F(3, 334) = 28.28, p < 0.001). D-P distance decreased significantly with one week
of exercise (24% decrease, p < 0.001) and significantly increased with two and four weeks
of exercise (10% increase, p = 0.0132 and 19% increase, p < 0.001, respectively) relative
to sedentary animals.
In the ETC, D-P distance was also statistically significantly increased by exercise
(n = 38 to 55 cells, F(3, 181) = 8.75, p < 0.001). Interestingly, D-P distance significantly
increased with one and two weeks of exercise (13% and 11% increase, respectively; p =
0.023 and p < 0.001, respectively), but was not significantly different from sedentary
animals at four weeks of exercise (p = 0.228).

Exercise Changes the Number of Striatal Astrocyte Primary Processes
The number of primary processes emanating from an identifiable astrocyte soma
were counted in each brain region and exercise duration (Figure 2.3C). In the PFC (n = 30
33
to 52 cells, F(3, 108) = 0.68, p = 0.566) and the ETC (n = 38 to 55 cells, F(3, 139) = 0.11, p =
0.956), there was no exercise effect on the number of primary processes on GFAP-
positive astrocytes.
In the STR, exercise did have a statistically significant effect on the number of
primary processes (n = 21 to 48 cells, F(3, 123) = 13.11, p < 0.001). The number of primary
processes significantly decreased at one week (18% decrease, p = 0.0060) before
showing a trend in increasing at two weeks (11% increase, p = 0.0829) and significantly
increasing at four weeks of exercise (18% increase, p = 0.0012) relative to sedentary
controls.
Table 2.1. Statistical analysis for morphological analyses. Mean ± SEM for all morphological
analyses. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001 relative to sedentary control.
Morphology Duration of exercise Prefrontal cortex Striatum Ectorhinal cortex
CTCF (au)
Sedentary 27230 ± 2025 19163 ± 909.1 26228 ± 2403
One week 41483 ± 4129 *** 12626 ± 361.7 **** 42592 ± 2939 ****
Two weeks 29540 ± 2036 21944 ± 776.4 * 26897 ± 1770
Four weeks 21533 ± 1550 25089 ± 1329 **** 36245 ± 2562 *
Distal-
Proximal
(μm)
Sedentary 90.16 ± 3.82 90.52 ± 2.02 102.5 ± 3.17
One week 111.6 ± 3.87 *** 68.98 ± 3.01 **** 116.3 ± 3.50 ***
Two weeks 97.62 ± 3.51 99.47 ± 2.37 * 114.1 ± 3.27 ****
Four weeks 91.07 ± 3.50 107.7 ± 2.45 **** 94.78 ± 3.30
Primary
processes
Sedentary 4.84 ±0.15 4.69 ± 0.14 4.611 ± 0.14
One week 4.56 ± 0.12 3.86 ± 0.19 ** 4.556 ± 0.14
Two weeks 4.64 ± 0.16 5.21 ± 0.24 4.639 ± 0.14
Four weeks 4.59 ± 0.16 5.55 ± 0.15 ** 4.543 ± 0.13

34
Exercise Does Not Influence Astrocyte Number
To assess the potential that aerobic exercise causes astrocyte proliferation in
regions of interest (Li et al., 2005), we used immunohistochemistry to examine the
astrocyte-specific nuclear protein SOX9 to sample the number of astrocytes (Sun et al.,
2017). In the regions sampled (PFC, STR, ETC), we did not see any statistically significant
change in the number of SOX9-positive astrocytes in exercised animals relative to
sedentary animals (PFC: p = 0.430, n = 1039 nuclei; STR: p = 0.811, n = 1131 nuclei; ETC:
p = 0.758, n = 799 nuclei) (Figure 2.4).

Figure 2.4 – Aerobic exercise does not change astrocyte numbers. (top) Astrocyte counts were
assessed using SOX9 immunostaining and unbiased counting of SOX9-positive nuclei in non-
adjacent 300μm x 300μm squares. Average counts of SOX9-positive nuclei in sedentary animals
(PFC: n = 45.2; STR: n = 45.7; ECT: n = 42.3) were set as 100% and percent relative changes in
exercised animals were compared. (bottom) Representative image showing SOX9 colocalizes
with GFAp in the STR (scale bar = 10μm). n = 4 mice per time point. Data are shown as box-and-
whisker plots, where the upper- and lower-bounds of the box represent the 75
th
and 25
th
percentile,
respectively; the horizontal line represents the median; and the whiskers indicate the highest and
lowest values. One-way ANOVA with Dunnett’s multiple comparisons.
35
Exercise Modulates Astrocyte Complexity
Figure 2.5 summarizes Sholl analysis of astrocyte arborization, process
ramification, and morphological complexity in exercised animals relative to sedentary
animals by counting astrocytic process intersections of overlaid concentric circles
(Ferreira et al., 2014).
In the PFC, exercise showed a statistically significant effect on the number of
intersections (F(3, 1312) = 73.82, p < 0.001). One week of exercise did not significantly
increase the number of intersections (Figure 2.5A) or area under the curve (AUC) analysis
(n = 10 cells, p = 0.213) (Figure 2.5B). In contrast, two weeks of exercise significantly
decreased the number of intersections in distal parts of the cell (> 40μm from soma, p =
0.0084, n = 9 cells) but did not affect AUC analysis (p = 0.840), while four weeks of
exercise significantly decreased both the number of intersections (p < 0.05, n = 9 cells)
as well as significantly decreasing AUC compared to sedentary (499.7 ± 111.9 vs 680 ±
73.34 intersections • μm, p = 0.0298).
In the STR, exercise also showed a statistically significant effect on the number of
intersections (F(3, 1476) = 67.09, p < 0.001). We found a significant exercise effect for AUC
analysis (n = 10 cells per time point, F(3, 36) = 13.46, p < 0.001). One week of exercise
significantly decreased the number of intersections (n = 10 cells, p < 0.05) as well as the
AUC (411.7 ± 97.85 vs 568.4 ± 78.64 intersections • μm, p = 0.0064), while two weeks of
exercise had no effect on the number of intersections (n = 10 cells, p > 0.05) or AUC (p =
0.806). However, four weeks of exercise significantly increased the number of
intersections (n = 10 cells, p < 0.05) but did not significantly increase the AUC (p = 0.676)
(Figure 2.5C-D).
36
In the ETC, our analysis showed a statistically significant effect of exercise on the
number of intersections (F(3, 1189) = 83.48, p < 0.001). We additionally showed a significant
effect of exercise on AUC analysis (n = 8-9 cells per time point, F(3, 29) = 19.94, p < 0.001).
One week of exercise showed a trend for increased intersections in proximal areas of
astrocytes compared to sedentary (<40μm from soma, p = 0.079, n = 8 cells) without any
significant difference in AUC (p = 0.280). Two weeks of exercise also showed a trend for
increased proximal intersections (p = 0.078, n = 9 cells) but had no effect on AUC (p =
0.435). Finally, four weeks of exercise increased intersections significantly both
proximally (< 40μm from soma) and distally (> 40μm from soma) (p < 0.001, n = 8 cells)
as well as increasing the AUC (909.4 ± 166.9 vs 558.8 ± 86.01 intersections • μm, p <
0.001) (Figure 2.5E-F).

37

Figure 2.5 – Sholl analysis of astrocytes reveals distinct arborization patterns across regions
and exercise duration. (a, c, and e) Number of intersections for prefrontal cortex, striatum, and
ectorhinal cortex astrocytes per 2μm step. (b, d, and f) Area-under-the-curve analysis of Sholl plots
for a, c, and e, respectively. N = 4 mice per time point. Data are shown as mean intersections every
2μm (Sholl plot) or box-and-whisker plots, where upper- and lower-bounds of the box represent
the 75
th
and 25
th
percentile, respectively; the horizontal line represents the median; and the
whiskers indicate the highest and lowest values (AUC, intersections x μm). Two-way ANOVA with
Dunnett’s multiple comparisons (Sholl) and one-way ANOVA with Dunnett’s multiple comparisons
(AUC). * p < 0.05; ** p < 0.01; *** p < 0.001 relative to sedentary control.
38

Exercise Induces Differential Gene Expression in Astrocytes
To examine the effect of exercise on astrocytic gene expression, we selected four
genes of interest based upon previous analysis of reactive astrocytes (Zamanian et al.,
2012) for qRT-PCR analysis (Figure 2.6), including Gfap (Glial Fibrillary Acidic Protein, a
marker of reactive astrocytes), Lif (Leukemia Inhibitory Factor, a trophic and
differentiation factor), Thbs2 (thrombospondin, an astrocyte-secreted synaptogenic
factor), and Il6 (Interleukin-6, a pro- and anti-inflammatory chemokine).
In the PFC, exercise resulted in a statistically significant increase in Gfap
expression (F(2, 9) = 1184, p < 0.001), at two weeks of exercise (7.39-fold increase relative
to sedentary control, p < 0.001); an increase in Thbs2 (F(2, 9) = 97.41, p < 0.001) at both one
week (2.16-fold increase, p = 0.0066) and two weeks of exercise (4.1-fold increase, p <
0.001); and an increase in Il6 (F(2, 9) = 29.32, p. < 0.001) at one week (3.23-fold increase, p
< 0.001) but not two weeks of exercise (p = 0.262), without any exercise effect on Lif
expression (F(2, 9) = 0.57, p = 0.582).
In the STR, our analysis showed Gfap expression statistically significantly
increased (F(2, 9) = 6.13, p = 0.0368) at one week (2.98-fold increase, p = 0.0146) and two
weeks of exercise (2.80-fold increase, p = 0.0311); Thbs2 expression increased (F(2, 9) =
6.48, p = 0.018) at one week (2.95-fold increase, p = 0.0211) and two weeks of exercise
(2.89-fold increase, p = 0.0240); Il6 expression showed an exercise effect (F(2, 9) = 86.97,
p < 0.001) but in particular expression did not significantly increase at one week of
exercise (1.52-fold increase, p = 0.145) but was significantly increased at two weeks of
exercise (4.53-fold increase, p < 0.001); and Lif expression increased significantly (F(2, 9)
39
= 6.02, p = 0.0219) at one week (2.06-fold increase, p = 0.0203) and two weeks of exercise
(1.93-fold increase, p = 0.0377).
In the ETC, exercise did not result in any statistically significant change in
expression of the four sampled genes, excluding a statistically significant increase of Il6
expression at two weeks of exercise (9.25-fold increase, p < 0.001).
Figure 2.6 – Aerobic exercise induces reactive astrocyte gene expression. (a) Fold induction of
Gfap with 1 and 2 weeks of exercise. (b) Fold induction of Thbs2 with 1 and 2 weeks of exercise.
(c) Fold induction of Il6 with 1 and 2 weeks of exercise. (d) Fold induction of Lif with 1 and 2
weeks of exercise. N = 4 mice per time point; data are mean + SD. One-way ANOVA with Dunnett’s
multiple comparisons. * p < 0.05; ** p < 0.01; *** p < 0.001 relative to sedentary control.

40
Discussion
Astrocytes comprise 15-20% of all cells in the rodent brain and support
neurotransmission through recycling of neurotransmitters and energetic and trophic
substrate delivery (Bélanger et al., 2011; Khakh & Sofroniew, 2015; Sun et al., 2017).
Astrocytes orchestrate regional cerebral blood flow (rCBF) and synaptic plasticity by
regulating synaptogenesis and modulating the blood brain barrier to promote experience-
dependent neuroplasticity (Abbott et al., 2006; Christopherson et al., 2005; Eroglu &
Barres, 2010). Astrocytes respond to changes in brain activity and homeostasis by
entering a reactive state (Sofroniew & Vinters, 2010). One aspect of reactive astrocytes
is a change in cellular morphology, as demonstrated by increased expression of glial
fibrillary acidic protein (GFAP, an intermediate filament), and cellular hypertrophy
(Wilhelmsson et al., 2006); while the activation of GFAP has been predominantly
characterized in the context of brain injury and disease, increased expression of GFAP
has also been implicated as a surrogate marker of astrocyte reactivity in response to
aerobic exercise (Li et al., 2005; Saur et al., 2014; Sofroniew & Vinters, 2010). Aerobic
exercise is a form of experience-dependent neuroplasticity that changes rCBF in a circuit-
specific manner and rescues cognitive and motor behaviors in animal models of
dopamine dysfunction (Fisher et al., 2004; Kintz et al., 2013; Petzinger et al., 2007; Toy et
al., 2014; Wang et al., 2013). Astrocytes monitor synaptic activity by close contact
between distal processes and neurons, and changes in GFAP expression and astrocyte
morphology are regulated in response to synaptic activity (Bernardinelli et al., 2014).

41
Our results demonstrate that exercise increases the expression of GFAP and
mediates changes in astrocyte morphology. In this study, we focused on specific
neuroanatomical regions of interest (PFC, STR, and ETC) based upon published changes
in rCBF following exercise (Wang et al., 2013). We found that astrocytes within the PFC
increased GFAP expression, increased their distal-to-proximal span, and increased overall
arborization following one week of exercise before returning to baseline morphology even
with the continuation of exercise. This contrasts with findings in the STR that showed
astrocytes initially decreased GFAP expression (and less morphological complexity)
following one week of exercise but increased GFAP expression and morphological
complexity with the continuation of exercise. We found distinct changes in astrocyte
morphology and GFAP expression in the ETC. For example, distal-to-proximal length
increased at one and two weeks of exercise but did not change with four weeks of
exercise as it did in the STR. Alternately, the number of astrocyte primary processes did
not change with exercise in the ETC, similarly to the PFC but in contrast to the STR. While
the PFC and STR showed similar temporal and structural differences in morphology in
response to exercise, the ETC showed both similarities and differences in astrocytic
phenotype that may be reflective of regional astrocyte heterogeneity (Chai et al., 2017).
While some morphological features of astrocytes appear to be region-specific based on
circuit activation, other morphological features may be more global. In a mouse model of
Alzheimer’s disease, increased GFAP expression was found throughout the brain
(Kamphuis et al., 2012) but no morphological analysis was done on these astrocytes to
delineate in greater detail astrocyte reactivity and its functional correlate. Future studies
using GFAP as a surrogate marker of morphology should incorporate more detailed
42
morphological metrics to link GFAP activation and astrocyte function. These differences
in region-specific morphological changes of astrocytes may reflect differences in the role
they play in supporting synaptic integrity and mediating synaptogenesis, as well as
modulating metabolic functions at the blood brain barrier.
Astrocytes can respond to changes in synaptic activity by altering messenger RNA
transcript expression. These astrocyte-specific transcripts are involved in a wide array of
mechanisms, including synaptogenesis, metabolism, and angiogenesis (Abbott et al.,
2006; Christopherson et al., 2005; Stogsdill et al., 2017). To explore the effects of exercise
on astrocyte function and their contribution to synaptogenesis, we analyzed by qRT-PCR
the expression of four genes associated with astrocyte reactivity (Gfap, Thbs2, Lif, and
Il6) (Zamanian et al., 2012). Exercise resulted in the increased expression in Gfap in the
PFC and STR, but not in ETC, suggesting an exercise-mediated association between
activated astrocytes and circuits involved in motor behavior, especially cortico-striatal
circuits. In support of the role of astrocytes in mediating synaptogenesis, we found that
expression of the synaptogenic molecule Thbs2 also increased in the PFC and STR, but
not in ETC, reflecting region-specific activation. Thrombospondins promote the formation
of synapses and previous work from our group has shown an increase in synaptogenesis
following exercise within these same regions (Christopherson et al., 2005; Morel et al.,
2017; Toy et al., 2014). The cytokine LIF (leukemia inhibitor factor) plays a role in neural
development and plasticity. We found increased expression of Lif in the STR following
exercise, further supporting exercise-dependent neuroplasticity (Bauer et al., 2007;
Petzinger et al., 2007). Unlike Gfap, Thbs2, and Lif, the expression of the cytokine IL6
increased across all regions of interest following exercise. IL6 works in both a pro- and
43
anti-inflammatory capacity and its elevated expression may suggest a global response
to exercise (Gruol & Nelson, 1997). The expression of Il6 can be mediated in multiple
ways, including by peripheral lactate derived from muscles during exercise (Andersson et
al., 2005; Erta et al., 2012); however, the precise role of IL6 in the brain following exercise
remains unknown. Previous work has shown an exercise-induced increase in Il6
expression in the hippocampus but not in the cortex or cerebellum of mice; however, such
changes were observed following acute exercise which is in contrast to our chronic
exercise paradigm (Rasmussen et al., 2011). Taken together, changes in gene expression
associated with reactive and synaptogenic astrocyte function occur in a region-specific
manner, while those genes associated with neuropoietic function occur in both a region-
specific and global pattern. Such patterns suggest that exercise differentially activates
astrocytic gene expression and may ultimately reflect differences in astrocytic structure
and function.
Exercise-induced GFAP, morphological, and transcriptional changes of astrocytes
suggest that aerobic exercise can modulate astrocyte function to enhance exercise-
induced neuroplasticity. Our immunohistochemical analysis of the astrocyte-specific
transcription factor SOX9 did not reveal any changes in astrocyte number in our regions
of interest, contrary to previous reports (Li et al., 2005). Our studies support that exercise
increases GFAP expression and morphological changes, but do not show that astrocyte
proliferation is occurring. Adult hippocampal neurogenesis has been shown to be
elevated with exercise but evidence for neurogenesis or gliogenesis in the striatum
following exercise is lacking (van Praag, 2005; van Praag et al., 1999). Thus, future studies
should explore the impact of different exercise paradigms on cellular proliferation in
44
regions activated by motor behavior, such as the striatum. These results suggest that
exercise is not changing the relative astrocyte composition in these regions, but instead
is modifying astrocytic function. Aerobic exercise enhances experience-dependent
neuroplasticity and synaptogenesis, leading to cognitive and motor circuit connectivity
(Fisher et al., 2004; Petzinger et al., 2013). In the context of normal brain function, exercise
can act as a mechanism for brain maintenance, allowing motor and cognitive learning to
occur (Davies et al., 2017). In the context of neurodegenerative diseases, such as
Parkinson’s or Huntington’s disease, where motor and cognitive dysfunction is evident,
skilled exercise can be harnessed to restore dysfunctional circuits, including cortico-
striatal circuitry (Petzinger et al., 2015; Wang et al., 2016). While much of motor learning
and exercise-induced neuroplasticity has focused on the protection and restoration of
neurons, studies from our group are beginning to identify and support an important role
for astrocytes. We hypothesize that astrocytes physically remodel to locally support
neurons as new, exercise-specific circuits are activated. As plasticity begins to refine
such circuits in an activity-dependent manner (i.e., with more exercise), astrocytes
respond to facilitate synaptogenesis and provide neurotrophic support to contribute to
exercise-induced neuroplasticity. In the rat hippocampus, gene expression patterns for
important neuroplasticity markers, such as BDNF, synaptotagmin, and NMDA receptors,
undergo exercise-dependent changes over various lengths of exercise duration,
supporting the notion that chronic exercise is capable of remodeling neuronal function
(Molteni et al., 2002). Increased synaptic activity, which occurs during exercise, is
energetically demanding (Attwell & Laughlin, 2001), and astrocytes can provide energetic
support by supplying neurons with lactate via the astrocyte-neuron lactate shuttle (ANLS)
45
(Pellerin & Magistretti, 1994). The maintenance of the ANLS has been shown to be
important for learning, exhaustive exercise capacity, and memory (Boury-Jamot et al.,
2016; Matsui et al., 2017; Suzuki et al., 2011). Exercise effects on the ANLS and metabolic
cooperation between neurons and astrocytes may be important for exercise-induced
neuroplasticity. Additionally, future studies should seek to understand the connection
between peripheral and central factors, including muscle-derived sources of lactate and
cytokines (E et al., 2013; Fischer, 2006). Lactate is capable of inducing angiogenesis
through the Gi/o-protein coupled receptor HCA1 and may likely play a role in facilitating
region-specific, exercise-induced blood flow changes that map with circuit-specific
neuroplasticity (Morland et al., 2017).
In conclusion, findings from this study, based on changes in astrocytic morphology
and gene expression, support astrocytic roles in neuroplasticity and synaptogenesis as
mediated through exercise. While astrocytes are important in neuroplasticity and
synaptogenesis, the underlying mechanisms that drive such a response remain
incompletely known. Future studies should seek to explain the precise molecular
components that may play a role in driving astrocytic response to exercise, and whether
they are derived from local neurons and other glial cells or peripheral sources.
Additionally, how exercise modifies the blood brain barrier and astrocytic regulation of
blood flow may also be an important avenue for more completely understanding exercise-
induced neuroplasticity and synaptogenesis.

46
Chapter 3: Exogenous L-lactate promotes astrocyte plasticity but
is not sufficient for enhancing striatal synaptogenesis or motor
behavior in mice
Adapted from Lundquist et al., 2021. Journal of Neuroscience Research, 99 (5), 1433-
1447

Abstract
L-Lactate is an energetic and signaling molecule that may be produced through astrocyte-
specific aerobic glycolysis and is elevated in striatal muscle during intensive exercise. L-
Lactate has been shown to promote neurotrophic gene expression through astrocytes
within the hippocampus, however its role in neuroplasticity within the striatum remains
unknown. This study sought to investigate the role of peripheral sources of L-lactate in
promoting astrocyte-specific gene expression and morphology as well as its role in
neuroplasticity within the striatum of healthy animals. Using in vitro primary astrocyte cell
culture, administration of L-lactate increased the expression of the neurotrophic factors
Bdnf, Gdnf, Cntf and the immediate early gene cFos. L-Lactate’s promotion of
neurotrophic factor expression was mediated through the lactate receptor HCAR1 since
application of the HCAR1 agonist 3,5-DHBA also increased expression of Bdnf in primary
astrocytes. Similar to our previous report demonstrating exercise-induced changes in
astrocytic structure within the striatum, L-lactate administration to healthy mice led to
increased astrocyte morphological complexity as well as astrocyte-specific neurotrophic
expression within the striatum. Our study failed to demonstrate an effect of peripheral L-
lactate on synaptogenesis or motor behavior. Insufficient levels and/or inadequate
47
delivery of L-lactate through regional cerebral blood flow within the striatum may account
for the lack of these benefits. Taken together, these novel findings suggest a potential
framework that links peripheral L-lactate production within muscle and intensive exercise
with neuroplasticity of specific brain regions though astrocytic function.

48
Introduction
Over the last two decades, exercise has emerged as an effective tool to enhance
neurogenesis and synaptogenesis in healthy animals (Cotman & Berchtold, 2002), as well
as providing a non-pharmacological therapy for improving motor performance in
individuals with Parkinson’s disease (PD) (Petzinger et al., 2013). While the underlying
molecular mechanisms responsible for the benefits of exercise are not yet fully
elucidated, studies suggest a role for neurotrophic factors (NTFs) such as brain-derived
neurotrophic factor (BDNF) in promoting neuronal morphological changes and behavioral
improvement (Lista & Sorrentino, 2010; Lu, 2003). In addition, studies in our laboratories
utilizing animal models of PD have identified several exercise-induced mechanisms that
underlie neuroplasticity to improve motor behavior including increased regional cerebral
blood flow (rCBF), synaptogenesis, and alterations in dopamine and glutamate
neurotransmission (Kintz et al., 2013; Petzinger et al., 2015; Toy et al., 2014).

Astrocytes may also underlie these exercise-related benefits by impacting
neuronal structure and function as demonstrated by alterations in morphology, the
expression of NTFs, and the support of neuronal metabolism (Bernardinelli et al., 2014;
Magistretti & Allaman, 2018; Suzuki et al., 2011). Due to their close proximity to synapses
and blood capillaries, astrocytes are well positioned to bridge mechanisms coupling
neuronal activity and rCBF. Our recent studies support such a link where we have shown
that exercise in normal mice results in alterations in astrocyte morphology as well as
increases in astrocyte-specific gene expression including mRNA transcripts involved in
lactate metabolism and transport within the striatum (Halliday et al., 2019; Lundquist et
al., 2019). Interestingly, these changes appear to occur in a circuit-specific manner
49
reflecting changes in rCBF with exercise (Wang et al., 2013). One subset of genes altered
with exercise are those involved in L-lactate uptake and metabolism. Not only is L-lactate
important in hippocampal memory formation and long-term potentiation acting through
the astrocyte-neuron lactate shuttle (Suzuki et al., 2011) but it may also serve as a
modulator of NTF expression by enhancing transcription of genes including Bdnf (Coco
et al., 2013). While astrocytes are able to generate L-lactate through aerobic glycolysis,
another source of CNS L-lactate comes from the periphery, a byproduct of striated
muscle metabolism, the result of increased physical activity or exercise (Ide et al., 2000;
van Hall et al., 2009). Such peripheral L-lactate can play an important role in
neuroplasticity through the promotion of NTF expression to improve learning and
memory, increase angiogenesis, and support neuronal metabolism and energy
production (E et al., 2013; El Hayek et al., 2019; Morland et al., 2017). Exercise has also
been shown to activate the G protein-coupled receptor hydroxycarboxylic acid receptor 1
(HCAR1), the major receptor for L-lactate on astrocytes leading to increased expression
of VEGF and promoting angiogenesis (Morland et al., 2017). Taken together, these
findings suggest that peripheral L-lactate produced through muscle activity in exercise
may serve as an important means to promote neuroplasticity by inducing astrocyte-
specific genes (Eroglu & Barres, 2010).

The purpose of this study was to investigate the effects of peripheral L-lactate on
astrocyte-specific gene expression involved in synaptic plasticity or morphology both in
vitro, using astrocyte cell cultures, and in vivo, using healthy adult mice. We used qRT-
PCR to analyze changes in the expression of several astrocytic genes involved in
50
synaptogenesis including the NTFs Bdnf, glial cell line-derived neurotrophic factor (Gdnf),
ciliary neurotrophic factor (Cntf), and the immediate early gene cFos. Furthermore, we
investigated whether physiological levels of L-lactate injected into healthy mice could
recapitulate exercise-enhanced benefits on neuroplasticity as demonstrated by changes
in synaptogenesis, specifically through immunohistochemical staining for synaptophysin
and postsynaptic density protein 95 (PSD95), as well as by improved motor performance.
Findings from these studies may help to reveal the potential underlying mechanisms of
exercise-enhanced neuroplasticity linking L-lactate, especially from the periphery, with
striatal neuronal connectivity, and the central role played by astrocytes.

51
Methods
Animals
No pre-registration was performed for this study. C57BL/6J mice 10 to 14-weeks
of age (The Jackson Laboratory, Bar Harbor, ME) of both sexes were used for all
experiments described and were housed at the University of Southern California vivarium.
Mice were housed in groups of 4 to 5 per cage and maintained on a reverse 12-hour
light/dark cycle (lights off at 0700 hours) with ad libitum access to food and water.
Postnatal day 0-4 (PND0-4) pups from in-house C57BL/6J breeding pairs were used for
the generation of primary astrocyte cultures, which are described in more detail below.
All procedures were approved by the Institutional Animal Care and Use Committee of the
University of Southern California (IACUC protocol #9766, #21044) and conducted in
accordance with the National Research Council’s Guide for the Care and Use of
Laboratory Animals (Committee for the Update of the Guide for the Care and Use of
Laboratory Animals; National Research Council, 2010). No randomization was explicitly
performed to allocate subjects in the study.

Experimental Design and In Vivo Lactate Administration
All experiments were conducted between 0900 and 1400 hours. Mice (n = 32) were
divided into two groups for all experiments (n = 16 per group, equal number of male and
female) and intraperitoneal (I.P.) injected daily for 10 days with either a solution of
sodium L-lactate (Cat# 71718, Millipore-Sigma, St. Louis, MO; 2g/kg body weight,
estimated final concentration 10mM) (Morland et. al, 2017), dissolved in 0.9% saline, or
an equal volume of vehicle (0.9% saline). After each daily injection, mice were placed on
52
motorized horizontal treadmills (EXER-6M, Columbus Instruments, Columbus, OH) to
walk for 10 min/day at a speed of 6m/min. Mice that received vehicle injections were
placed on an adjacent treadmill to walk for 10 min/day at a speed of 6m/min. All mice
were maintained at this low level of activity on the treadmill where L-lactate levels have
been reported to not increase in rodents (Billat et al., 2005; Soya et al., 2007). All mice
were monitored and handled daily, and no mice required veterinary care, died, or had to
be otherwise excluded during the course of experiments.

Rotarod Motor Performance
Motor performance following in vivo L-lactate administration was tested on an
accelerated rotarod using a modification of previously published methods (Rothwell et
al., 2014). Following L-lactate administration as detailed above, a subset of mice (n = 5
per group) were trained on an accelerating rotarod (3cm diameter rod, divided into five
lanes; Ugo Basile, Comerio, Italy). The rotarod accelerated over the course of 300 seconds
from 6 to 60 rpm, and speed at time of fall and latency to fall were automatically recorded
by magnetic trip plates. Mice were acclimated to the rotarod for 90 seconds before the
start of their first trial of each day and were trained for five trials per day for four days,
with a one-minute intertrial interval. A trial ended when the mouse made a complete
backward revolution, fell off, or reached the 300 second threshold. Rotarod training was
conducted by persons blind to the experimental group assignment of the mice.

53
Brain Tissue Collection
Whole brains were collected on the last day of injection and walking. After the
cessation of walking, mice were divided into groups for either fresh tissue collection (n =
6 mice, see below) or transcardial perfusion (n = 5 mice). For transcardial perfusion, one
hour after the final session, mice were heavily sedated with intraperitoneal injections of
tribromoethanol (250 mg/kg body weight) and assessed for lack of toe-pinch response.
The use of tribromoethanol for mouse sedation is standard and approved for use in our
protocol. Mice were transcardially perfused with 50 ml of ice-cold 0.9% saline followed
by 100 ml of ice-cold 4% paraformaldehyde in phosphate buffered saline, pH 7.2 (PFA-
PBS). Brains were extracted, transferred to 4% PFA-PBS at 4°C overnight and then to a
20% sucrose cryoprotection solution. After sinking, brains were rapidly frozen in 2-
methylbutane cooled on dry ice and stored at -80°C until use.
Fresh tissue dissections were carried out immediately following the final session
of injection and walking. Mice from both vehicle and L-lactate groups (n = 6 mice each)
were sacrificed by cervical dislocation, as approved in our protocol and to limit potential
effects of anesthesia on gene expression, and the brains were resected to ice-cold PBS,
pH 7.2. Brains were microdissected to isolate the striatum (STR; Bregma +1.2 to -0.2mm
A.P., including tissue bordered ventrally by the anterior commissure, dorsally by the
corpus callosum, medially by the lateral ventricle, and ±2.5 mm laterally from the midline)
(Halliday et al., 2019; Kintz et al., 2013) and ectorhinal cortex (ETC; Bregma -2.0 to -3.0
mm A.P., laterally bounded by the edge of the cortex and extending 1mm medially to the
edge of the striatum, and -3.0 to -3.5 mm D.V.) (Halliday et al., 2019).

54
Primary Astrocyte Isolation and Culture
Mouse astrocytes were prepared as previously described (Jouroukhin et al., 2018).
Briefly, whole brains were removed from PND0-4 mice to ice-cold Dulbecco’s phosphate
buffered saline (DPBS) and the cortices were freed of meninges, microdissected, and
minced into small chunks. Cortices were digested with 0.25% trypsin-EDTA (Cat.# 25-510,
Genesee Scientific, San Diego, CA) in DPBS at 37°C for 20 minutes with inversion every 5
minutes. Tissue lysate was centrifuged at 300 x g for 5 min, supernatant removed, and
tissue pellet resuspended in astrocyte media (DMEM with 10% fetal bovine serum and 1%
penicillin/streptomycin; Cat.# 25-500, Cat.# 25-550, Cat.# 25-512, Genesee Scientific).
Tissue pellet was triturated with decreasing size pipettes until a single cell suspension
was achieved, which was passed through a 40μm filter to strain large clumps of debris.
The resulting single cell suspension was plated on 75 cm2 tissue-culture treated flasks
and grown in a 37°C incubator with 5% CO2. Loosely attached glial cells were washed
with DPBS and removed by vigorously shaking flasks by hand for 2 minutes before
exchanging DPBS with fresh astrocyte media. Astrocytes were confluent by 7 days in vitro
(DIV7), when cultures were incubated with 0.25% trypsin-EDTA at 37°C and replated on
poly-D-lysine (PDL) coated coverslips or 6- or 12-well tissue-culture treated plates until
future analysis. Astrocytes were confluent and used for experimentation by DIV14±1 day.
Resultant astrocyte cultures were highly pure (on average >96% astrocytes) and purity
was assessed through SOX9, GFAP, and ALDH1L1 immunocytochemistry and
quantitative real-time PCR of cell-type specific genes as described below (Figure 1).

In vitro Administration of Compounds
55
All experimental compounds were prepared in 50 or 100x stock solutions in 0.9%
saline and kept frozen at -20°C until the day of experimentation. Primary astrocytes in 6-
or 12-well plates had media exchanged on the day of experimentation and then stock
solutions were added directly to culture media to their final concentration for 1 hour
(unless otherwise noted). Final concentration for L-lactate was 10mM dissolved in 0.9%
saline, while the final concentration of 3,5-DHBA (3,5-dihydroxybenzoic acid; Cat.#
D1100000) was 2.5mM dissolved in 0.9% saline.

Immunofluorescent Staining
Immunohistochemical Staining: Whole brains (fixed) were sliced in coronal
sections (30 μm thickness) on a sliding freezing microtome (Leica CM1900, Leica
Microsystems, Wetzlar, Germany) as previously described (Lundquist et al., 2019). Briefly,
sections were washed in Tris-buffered saline, pH 7.2, with 0.2% Triton X-100 (TBST),
blocked in 4% normal goat serum (NGS; Vector Laboratories Cat# S-1000,
RRID:AB_2336615) in TBST, and incubated overnight at 4°C in primary antibody (2% NGS
in TBST) with either rabbit anti-GFAP (1:2000, Agilent Cat# Z0334, RRID: AB_10013382),
rabbit anti-synaptophysin (IgG, 1:2000, Abcam Cat# ab32127, RRID:AB_2286949) or
mouse anti-PSD-95 (IgG1, 1:2000, Millipore Cat# AB9708, RRID:AB_2092543). Sections
were washed with TBST and incubated in secondary antibody (2% NGS in TBST) with
Alexa 568-conjugated goat anti-rabbit (1:5000, Thermo Fisher Scientific, Cat# A-11011,
RRID: AB_143157), or Alexa 568-conjugated goat anti-mouse IgG1 (1:5000, Thermo
Fischer Scientific, Cat# A-21124, RRID: AB_2535766). Sections were washed, mounted
onto gelatin-coated slides, and coverslipped (Vectashield Hardset Antifade with DAPI; H-
56
1500, Vector Labs, Burlingame, CA). Confocal images were taken on an IXB-DSU spinning
disk Olympus BX-61 (Olympus America, Melville, NY) and captured with an ORCA-R2
digital CCD camera (Hamamatsu, Bridgewater, NJ) and MetaMorph Advanced software
(Molecular Devices, San Jose, CA).
Immunocytochemical Staining: Primary astrocytes grown on PDL-coated
coverslips were washed twice with ice-cold DPBS before being fixed with ice-cold 4%
PFA-PBS (pH 7.2) for 10 minutes. Astrocytes were washed with TBS, permeabilized with
TBST, and washed again with TBS. Astrocytes were blocked with 10% normal goat serum
(NGS; S-1000, Vector Labs, Burlingame, CA) in TBST before incubating overnight at 4°C
in primary antibody solution (4% NGS in TBST) with rat anti-GFAP (Millipore Cat# 345860-
100UG, RRID:AB_211868), rabbit anti-GFAP (1:2000, Agilent Cat# Z0334, RRID:
AB_10013382), rabbit anti-SOX9 (1:2000, Millipore Cat# AB5535, RRID:AB_2239761) and
mouse anti-ALDH1L1 (1:500, UCDavis/NIH NeuroMab Facility Cat# 75-164,
RRID:AB_10671695). Astrocytes were washed the following day with TBS and incubated
in secondary antibody solution (2% NGS in TBST) with goat anti-mouse 488 (1:5000,
Thermo Fisher Scientific Cat# A-11001, RRID:AB_2534069), goat anti-rabbit 488 (1:5000,
Thermo Fisher Scientific Cat# A-11008, RRID:AB_143165), goat anti-rat 568 (1:5000,
Thermo Fisher Scientific Cat# A-11077, RRID:AB_2534121) or goat anti-rabbit 568
(1:5000, Thermo Fisher Scientific, Cat# A-11011, RRID: AB_143157).

Analyzing Synaptic Puncta
Synaptophysin- and PSD-95-positive puncta were captured in the striatum and
ectorhinal cortex with a 40x objective and identical camera settings, and images were
57
analyzed using the following workflow in Fiji/ImageJ. First the background was
subtracted, then images were manually thresholded at approximately the upper 1% of
signal to eliminate non-specific or overlapping puncta. Finally, the Analyze Particles
plugin for ImageJ (Schindelin et al., 2012) was used to quantify puncta for the entire field
of view of the image. Two or more tissue sections through each anatomical region (Dong,
2008) were analyzed per animal (n = 6 mice per group). Synaptic puncta were analyzed
by persons blind to the experimental group.

SOX9 Colocalization for Culture Purity
Assessment of relative purity of primary astrocytes was performed through
colocalization analysis of SOX9-DAPI double positive nuclei according to published
methods (Manders et al., 1993). Briefly, two to three coverslips from four separate
experiments with astrocytes stained with SOX9 and DAPI were captured with a 20x
objective before using the Colocalization plugin for ImageJ for relative percentage of
green/blue signal colocalization.

Sholl Analysis
Morphological analysis of GFAP-positive astrocytes in the striatum and ectorhinal
cortex were assessed using Sholl analysis as previously described (Lundquist et al., 2019;
Sholl, 1953). Briefly, z-stack images of astrocytes were captured and manually
segmented using the Simple Neurite Tracer plugin for ImageJ (Longair et al., 2011;
Schindelin et al., 2012). Segments were maximally projected to form a composite,
segmented astrocyte, and concentric circles with increasing radii of 2 μm were overlaid
58
and the number of intersections was counted automatically and plotted by distance by
the Sholl plugin for ImageJ (Ferreira et al., 2014). Two or more tissue sections through
each anatomical region were analyzed per animal (n = 6 mice per group). Sholl analysis
of astrocyte morphology was conducted by persons blind to the experimental condition
of the animal.

RNA Isolation and Quantification
Brain tissue: Regions of interest (STR and ETC) were bilaterally microdissected (as
described above) and submerged in an RNA stabilization solution (pH 5.2) at 4°C,
containing in mM: 3.53 ammonium sulfate, 16.66 sodium citrate, and 13.33 EDTA
(ethylenediaminetetraacetic acid). Tissue was transferred to a sterile tube containing 300
μl TRI-reagent (Cat.# 11-330T, Genesee Scientific) and homogenized with a mechanical
pestle before centrifuging at 13,000 x g for 3 minutes. Supernatant was removed to a new
tube where 250 μl of chloroform was added and tubes vigorously shaken twice for 10
seconds followed by 3 minutes of resting on ice and centrifugation at 13,000 x g for 18
minutes at 4°C. The upper, aqueous layer was carefully removed to a new tube, an equal
volume of 100% ethanol was added, and the sample was thoroughly mixed before RNA
purification using the Zymo Direct-zol RNA Miniprep (Cat.# 11-330, Genesee Scientific)
according to the manufacturer’s instructions. RNA was eluted in 35 μl of DNAse/RNAse
free water before spectrophotometric analysis of RNA purity and concentration.
Primary astrocyte cultures: Culture media was removed, and astrocytes were
washed twice with ice-cold DPBS before direct addition of TRI-reagent to each well (250
μl or 500 μl of TRI-reagent for 12- or 6-well plates, respectively). Cells were gently scraped
59
with individual cell scrapers before removing to respective, sterile tubes on ice. 250 μl of
chloroform was added and tubes vigorously shaken twice for 10 seconds followed by 3
minutes of resting on ice and centrifugation at 13,000 x g for 18 minutes at 4°C. The
upper, aqueous layer was carefully removed to a new tube, an equal volume of 100%
ethanol was added, and the sample was thoroughly mixed before RNA purification using
the Zymo Direct-zol RNA Miniprep according to the manufacturer’s instructions. RNA was
eluted in 35 μl of DNAse/RNAse free water before spectrophotometric analysis of RNA
purity and concentration.

Complementary DNA Synthesis and Quantitative RT-PCR
Complementary DNA (cDNA) was synthesized from either 200 ng (astrocyte
cultures) or 400 ng (brain tissue) of isolated RNA using the qPCRBIO cDNA Synthesis Kit
(Cat.# PB30.11-10, PCR Biosystems, Wayne, PA) following manufacturer’s guidelines
before being diluted 1:5 in DNAse/RNAse free water and stored at -20°C. Gene expression
changes were measured with quantitative RT-PCR (qRT-PCR) similarly as previously
described (Halliday et al., 2019; Lundquist et al., 2019). Briefly, qRT-PCR was run with 2 μl
of cDNA and qPCRBIO SyGreen master mix (Cat.# PB20.11-01, PCR Biosystems) on an
Eppendorf Mastercycler Ep Realplex (Eppendorf, Hauppauge, NY) using a program of 15
min at 95°C, followed by 40 cycles of 15 seconds at 94°C, 30 seconds at 55°C, and 30
seconds at 72°C. Data was collected and normalized on Eppendorf Realplex ep software.
Standard delta-CT analysis (Livak & Schmittgen, 2001) was used to quantify fold changes
in gene expression in experimental groups normalized to controls, with Actb serving as a
housekeeping gene. A complete list of primer pairs can be found in Table 3.1.
60

Statistical Analysis
All statistical tests were carried out and graphs made in Prism 8.3 (GraphPad, San
Diego, CA) with statistical significance set at p < 0.05. No sample size calculations were
performed prior to the start of the study but are based on previous publications from our
lab (Halliday et al., 2019; Lundquist et al., 2019). Both male and female mice were used
for this study and sexes were collapsed within groups due to small sample sizes within
each group. Sex differences were not studied in the context of our data. We acknowledge
this as a shortcoming, and all results should be interpreted with this information in mind.
All data included was normally distributed as assessed by Shapiro-Wilk normality testing,
and no data points were excluded from analysis. Simple linear regression of accelerating
rotarod behavior was used to assess differences in initial coordination and learning rate
between groups. Unpaired, two-tailed T-tests were used for all qRT-PCR analyses, total
astrocyte process intersections in Sholl analysis, and total synaptophysin- or PSD95-
positive puncta. Two-way ANOVA with Bonferroni’s multiple comparisons were used for
analysis of Sholl plots. For Sholl plots, distance from soma served as the within-subject
comparison and L-lactate administration served as the between-subject comparison. A
mixed effects model (with trial number and L-lactate administration serving as fixed
effects) with Bonferroni multiple comparisons was used to analyze accelerating rotarod
performance across all trials. Where feasible, p values are listed on the corresponding
figures and all p values are listed in the results section.

61
Figure 3.1 – Primary mouse astrocyte cultures are highly pure. (a,b) Representative
immunocytochemistry of primary astrocyte cultures demonstrate characteristic nuclear and
cytoplasmic markers of astrocytes. (c) Over 96% of all DAPI-stained nuclei colocalized with the
astrocytic nuclear marker SOX9 (n = 10 coverslips from four independent culture experiments).
(d) Cell-specific transcriptional analysis with qRT-PCR verified astrocyte-enriched transcripts with
very little non-astrocyte lineage gene expression. N = 4 samples from four independent culture
experiments; box plots show minimum to maximum spread of values with individual points
overlaid.

62
Results
L-lactate and 3,5-DHBA administration elevates the expression of astrocyte-specific gene
transcripts in primary astrocyte cultures
The effect of L-lactate administration on astrocyte-specific gene transcripts was
performed in primary astrocytes cultures. Figure 3.1 shows the near confluence of
astrocytes based on SOX9 immunohistochemical staining, demonstrating a highly
enriched astrocyte culture. Gene transcript expression was assessed following an acute
(1 hour, 10mM), in vitro exposure of L-lactate (Figure 3.2). Following L-lactate exposure,
there was an increase in the expression of Gfap (a marker of astrocytic reactivity; 1.49-
fold increase, t=6.813, df=8, p < 0.001), Thbs2 (a glycoprotein involved in synaptogenesis;
1.84-fold increase, t=5.729, df=8, p < 0.001), and genes for the neurotrophic factors Gdnf
(glial cell line-derived neurotrophic factor; 2.27-fold increase, t=3.52, df=8, p = 0.007), Bdnf
(brain-derived neurotrophic factor; 1.64-fold increase, t=3.112, df=8, p = 0.014), and Cntf
(ciliary neurotrophic factor; 3.22-fold increase, t=34, df=8, p < 0.001). Additionally, there
was a significant increased expression of the immediate early gene cFos (3.71-fold
increase, t=8.01, df=8, p < 0.001). No significant change in expression was seen for genes
involved in glutamate neurotransmission including glutamine synthetase (Glul; t=1.267,
df=8, p = 0.241) and the glutamate transporter (Slc1a3; t=0.1447, df=8, p = 0.889) (n = 4-
5 independent cultures for all genes, Figure 3.2A).

63
Table 3.1. List of primers used for qRT-PCR experiments.
Gene Name Gene ID Forward Sequence (5’-3’) Reverse Sequence (5’-3’)
Aif1 11629 ATCAACAAGCAATTCCTCGATGA CAGCATTCGCTTCAAGGACATA
Actb 11461 GGCTGTATTCCCCTCCATCG CCAGTTGGTAACAATGCCATG
Bdnf 12064 CTTCCTAGCGGTGTAGGCTG CCTGGTGGAACTTCTTTGGC
Cntf 12803 TCTGTAGCCGCTCTATCTGG GGTACACCATCCACTGAGTCAA
Cx3cr1 13051 GAGTATGACGATTCTGCTGAGG CAGACCGAACGTGAAGACGAG
cFos 14281 CGGGTTTCAACGCCGACTA TTGGCACTAGAGACGGACAGA
Gdnf 14573 TTATGGGATGTCGTGCCTGT AGCCATATTGGCGGCG
Gfap 14580 CGGAGACGCATCACCTCTG AGGGAGTGGAGGAGTCATTCG
Glul 14645 CACCCCTGGTTTGGAATGGA GTAATACGGGCCTTGGGGT
Mbp 17196 GAGACCCTCACAGCGATCCAAG GGAGGTGGTGTTCGAGGTGTC
Nefl 18039 CCGTACTTTTCGACCTCCTACA CTTGTGTGCGGATAGACTTGAG
Slc1a3 20512 ACAATGGTGTGGACAAACGC CGTGGCTGTGATTATTG
Sox10 20665 ACACCTTGGGACACGGTTTTC TAGGTCTTGTTCCTCGGCCAT
Syt1 20979 CTGTCACCACTGTTGCGAC GGCAATGGGATTTTATGCAGTTC
Thbs2 21826 CTGGGCATAGGGCCAAGAG GCTTGACAATCCTGTTGAGATCA

64
Figure 3.2 – L-lactate and 3,5-DHBA administration to primary astrocytes induces changes in
astrocyte plasticity gene expression. (a) Administration of 10mM L-lactate to primary astrocytes
increased gene expression in several genes, as measured by qRT-PCR. (b) Administration of 3,5-
DHBA to primary astrocyte cultures increases gene expression as measured by qRT-PCR. N = 4-
5 samples from five independent culture preparations analyzed with unpaired two-tailed t tests;
box plots show minimum to maximum spread of values with individual points overlaid.

Following exposure to the L-lactate receptor HCAR1 agonist 3,5-DHBA (1 hour,
2.5mM), to primary astrocyte culture, there was a significant increase in the expression
of Gfap (18.93-fold increase, t=3.628, df=8, p = 0.006), Thbs2 (162.9-fold increase,
t=2.709, df=8, p = 0.027), Bdnf (144.5-fold increase, t=3.352, df=8, p = 0.010), and Glul
(7.99-fold increase, t=2.782, df=8, p = 0.024). There was no significant effect of 3,5-DHBA
exposure on the expression of the glutamate transporter (Slc1a3; t=0.644, df=8, p = 0.538)
(n = 4-5 independent cultures for all genes, Figure 3.2B).
L-lactate administration elevates the expression of astrocyte-specific gene transcripts in
normal healthy mice
L-lactate (2g/kg, estimated final concentration of 10mM) (Morland et al., 2017) or
saline was administered by I.P. injection in normal mice (n = 5 mice per group) for 10 days
and astrocyte-specific gene expression was determined by qRT-PCR in the striatum (STR)
and ectorhinal cortex (ETC). These two regions of the brain were selected based on the

65
report of exercise-induced increase (STR) or no change (ETC) in rCBF (Z. Wang et al.,
2013). In the STR, L-lactate significantly increased the expression of genes including Gfap
(3.57-fold increase, t=4.305, df=8, p = 0.003), Thbs2 (6.67-fold increase, t=4.733, df=8, p
= 0.001), and the neurotrophic factor genes Gdnf (15.3-fold increase, t=3.405, df=8, p =
0.009), Bdnf (30.1-fold increase, t=3.99, df=8, p = 0.004) and Cntf (82-fold increase,
t=3.955, df=8, p = 0.004) relative to saline treated animals. Additionally, there was a
significantly increased expression of the immediate early gene cFos (30.1-fold increase,
t=6.002, df=8, p < 0.001). The genes involved in glutamate neurotransmission were also
increased in expression including Glul (glutamine synthetase; 6.44-fold increase, t=3.043,
df=8, p = 0.016) and Slc1a3 (glutamate transporter; 4.38-fold increase, t=2.349, df=8, p =
0.047) relative to saline treated animals (Figure 3.3A). Within the ETC, L-lactate
administration did not result in any significant changes in gene expression compared to
saline administration (Figure 3.3B).

Figure 3.3 – L-lactate administration induces changes in astrocyte plasticity gene expression
in mice. (a) 10 days of 10mM L-lactate administration significantly increased expression of
astrocyte plasticity genes in the striatum of mice, as assessed by qRT-PCR. (b) 10 days of
10mM L-lactate administration did not affect expression of plasticity-related genes in the
ectorhinal cortex of mice, as assessed by qRT-PCR. N = 5 mice per group analyzed with
unpaired two-tailed t tests; box plots show minimum to maximum spread of values with
individual points overlaid.

66
L-lactate administration results in altered astrocyte morphology in normal healthy mice
L-lactate (2g/kg, estimated final concentration of 10mM) or saline was
administered by I.P. injection in normal mice (n = 6 mice per group) for 10 days and Sholl
analysis was used to examine two parameters, including (i) astrocytic arborization and
(ii) the total number of intersections in both the STR and ETC. L-lactate administration
resulted in a significant increase in the arborization of astrocytic processes (n = 17 cells
per group, two-way ANOVA; distance from soma: F(40, 1312) = 93.24, p < 0.001; lactate: F(1,
1312) = 99.23, p < 0.001; interaction: F(40, 1312) = 0.580, p = 0.984) compared to saline
controls (Figure 3.4A). Similarly, the total number of intersections per striatal
astrocyte was also significantly increased following L-lactate administration when
compared to saline controls (n = 17 cells per group, unpaired two-tailed t-test, t=2.882,
df=32, p = 0.007) (Figure 3.4B). In the ETC, L-lactate administration did not have a
significant effect on the overall morphological complexity of astrocytes (n = 12-13 cells
per group, two-way ANOVA; distance from soma: F(40, 942) = 62.48, p < 0.001; L-lactate: F(1,
942) = 1.674, p = 0.196; interaction: F(40, 942) = 0.337, p > 0.999) (Figure 3.4C) or on the total
number of intersections per astrocyte (n = 12-13 cells per group, unpaired two-tailed t-
test, t=0.3768, df=23, p = 0.710) compared to saline controls (Figure 3.4D).
67
Figure 3.4 – L-lactate administration causes morphological remodeling of striatal astrocytes
but not cortical astrocytes in mice. (a) 10 days of 10mM L-lactate administration in vivo
significantly increased the complexity of striatal astrocyte morphology as measured by Sholl
analysis. (b) The total number of intersections per astrocyte, as measured by Sholl analysis, was
significantly increased in striatal astrocytes following L-lactate administration. (c) 10 days of
10mM L-lactate administration in vivo did not affect ectorhinal cortical astrocyte morphology, as
measured by Sholl analysis. (d) The total number of intersections per astrocyte, as measured by
Sholl analysis, did not significantly differ in ectorhinal cortical astrocytes following L-lactate
administration in vivo. N = 6 mice per group analyzed with two-way ANOVA with Bonferroni’s
multiple comparisons (Sholl curves) or unpaired two-tailed t tests (total intersections). Sholl
curves show mean ± standard deviation, while box plots show minimum to maximum spread of
values with individual points overlaid.

68
L-lactate administration does not increase striatal PSD-95 or synaptophysin protein
expression in normal healthy mice
The patterns of expression of PSD-95 and synaptophysin (two proteins involved in
synaptogenesis) (Toy et al., 2014) were determined by counting the number of
immunopositive puncta in the STR and ETC. L-lactate administration to mice did not
result in a significant increase in the amount of synaptophysin positive puncta compared
to saline controls in either the STR or the ETC (n = 15-20 sections per group, STR: unpaired
two-tailed t-test, t=1.345, df=38, p = 0.188; ETC: unpaired two-tailed t-test, t=1.003, df=29,
p = 0.324) (Figure 3.5B-C). Similarly, there was no significant difference in PSD95 positive
puncta in these same mice following L-lactate administration compared to saline
controls in either the STR or the ETC (n = 12 sections per group, STR: unpaired two-tailed
t-test, t=0.5941, df=22, p = 0.559; ETC: unpaired two-tailed t-test, t=1.714, df=22, p = 0.100)
(Figure 3.5D-E).

69
Figure 3.5 – L-lactate administration does not increase synaptogenesis in striatum or ectorhinal
cortex in mice. (a) Representative low magnification images showing where striatum (top) and
ectorhinal cortex (bottom) synaptic puncta were sampled from. (b) Representative
immunohistochemistry showing synaptophysin-positive puncta in the striatum (left) and
ectorhinal cortex (right). (c) Synaptophysin-positive puncta did not significantly differ between
saline- and L-lactate administered mice. (d) Representative immunohistochemistry showing
PSD95-positive puncta in the striatum (left) and ectorhinal cortex (right). (e) PSD95-positive
puncta did not significantly differ between saline- and L-lactate-administered mice. N = 6 mice
per group analyzed with unpaired two-tailed t tests. Box plots show minimum to maximum spread
of values with individual points overlaid.

70

L-lactate administration does not improve motor performance on the accelerating rotarod
in normal healthy mice
The effect of L-lactate administration on motor performance in mice was
determined by the latency-to-fall from the accelerating rotarod. Over the 20-trial period, L-
lactate administration did not significantly alter performance on the rotarod (n = 5 mice
per group, mixed effects model; L-lactate administration: F(1, 8) = 0.300, p = 0.598; trial
number: F(4.97, 39.74) = 4.707, p = 0.002) (Figure 3.6A). Additional linear regression analysis
of terminal speed curves (Figure 3.6B) revealed that L-lactate did not significantly change
initial coordination (Y-intercept; n = 5 mice per group, F = 0.9537, DFd = 197, p = 0.330) or
the learning rate (slope of line; n = 5 mice per group, F = 0.984, DFd = 196, p = 0.322)
between the L-lactate and control groups (Figure 3.6C).
71

Figure 3.6 – L-lactate administration does not improve motor performance on the accelerating
rotarod. (a) Latency to fall (in seconds) for saline- and L-lactate-administered mice did not differ
over 20 trials of learning. (b) Terminal speed (in rpm) for saline- and L-lactate-administered mice
did not differ over 20 trials of learning. (c) Both initial coordination (in rpm) and learning rate
(rpm/trial) did not significantly differ between saline- and L-lactate-administered mice over 20
trials of learning. N = 5 mice per group analyzed with mixed effects model (rotarod performance)
or linear regression (learning rate and initial coordination). Rotarod performance curves show
mean ± standard deviation, while box plots show minimum to maximum spread of values with
individual points overlaid.

72
Discussion
The purpose of this study was to investigate the effects of L-lactate on the
expression of astrocyte-specific genes involved in either synaptic plasticity or astrocytic
morphology in vitro, using astrocyte cell cultures, and in vivo, within the striatum of adult
mice undergoing treadmill walking. L-lactate has been shown to play a role in BDNF
expression within the hippocampus leading to increased neurogenesis and improved
memory and learning in mice (El Hayek et al., 2019). Thus, we further explored whether
the intraperitoneal (IP) administration of L-lactate to treadmill walking mice results in the
elevated expression of neurotrophic factors (NTFs) and proteins involved in synaptic
plasticity within the striatum (Carrard et al., 2018; Margineanu et al., 2018). We found that
the administration of physiologic levels of L-lactate, as well as the HCAR1 agonist 3,5-
DHBA to astrocyte cultures, increased the transcript expression of several NTFs including
brain-derived neurotrophic factor (Bdnf), ciliary neurotrophic factor (Cntf), and glial cell
line-derived neurotrophic factor (Gdnf), as well as the astrocyte-specific transcripts Gfap
and Thbs2. We also found that daily IP injections of 10mM L-lactate for 10 days, was
associated with increased expression of NTF transcripts, including Bdnf, Gdnf, and Cntf
along with increased astrocytic branching and Gfap and Thbs2 expression within the
striatum but not the ETC. The striatal region-specific changes associated with L-lactate
administration may be due to the fact that mice were engaged in striatal related motor
activity (treadmill walking) that may increase delivery of L-lactate to this region and is
consistent with a previous study by Wang et al. reporting increased striatal but not ETC
rCBF in treadmill running mice (Wang et al., 2013). While the role of L-lactate on neuron-
derived gene expression important for synaptic plasticity, and mediated through NMDA
73
receptor dependent and/or independent processes, has been reported (Yang et al., 2014),
the role of L-lactate on astrocytic-derived gene expression involved in synaptic plasticity
has not been fully elucidated. Our findings support the potential role that astrocytes may
play in L-lactate mediated gene expression important for synaptic plasticity and astrocyte
morphology. L-lactate is known to function as both a metabolite and/or a signaling
molecule within the CNS, and our finding that 3,5-DHBA was able to replicate L-lactate
associated increases in expression of Bdnf and Gfap supports the possibility that L-
lactates effects on astrocyte-specific genes important for synaptic plasticity within the
CNS may be in part related to its receptor-mediated signaling effects.

The in vitro and in vivo administration of L-lactate increased the expression of the
immediate early gene cFos, as well as several astrocyte-specific genes including Gfap
and Tbsp2 that are important in astrocyte morphology and synaptic structure and
function (Christopherson et al., 2005; Eroglu et al., 2009; Sofroniew & Vinters, 2010).
These genes also increased in expression after administration of the Hydroxycarboxylic
Acid Receptor 1 (HCAR1) agonist 3,5-DHBA to astrocyte cultures supporting a potential
role for L-lactate in modulating gene expression through receptor-mediated cell signaling.
Gfap serves as a marker of reactive astrogliosis and morphological remodeling
(Sofroniew & Vinters, 2010). Thbs2 (also known as TSP2; a glycoprotein involved in
synaptogenesis) mediates excitatory synaptogenesis through astrocyte-neuron
interactions involving the postsynaptic α2-δ1 receptor (Christopherson et al., 2005; Eroglu
et al., 2009). Consistent with L-lactate’s effect on gene expression, daily in vivo
administration of L-lactate, compared to saline, was associated with changes in astrocyte
74
morphology in mice undergoing treadmill walking. These changes included increased
arborization, as defined by the number of GFAP-positive branching processes, and
increased number of branch intersections, based on Sholl analysis (Lundquist et al., 2019;
Sholl, 1953). Changes in astrocyte morphology were observed in the striatum but not in
the ETC, which was similar to our findings with L-lactate administration and region-
specific changes in NTF, Gfap, and Thps2 gene expression. Future studies will determine
if these striatal specific effects in astrocyte morphology may be regulated though L-
lactate-dependent cell signaling and induction of NTF expression.

We observed that in vivo administration of L-lactate was associated with increased
astrocyte-specific gene expression of GluI and Slc1a3, both involved in glutamate
neurotransmission. Glul (glutamine synthetase), and Slc1a3 (glutamate transporter
EAAT1) regulate excitatory glutamatergic neurotransmission (Perego et al., 2000),
regulate the synaptic occupancy of glutamate (Rothstein et al., 1996), and mediate
neuroplasticity as shown in models of learning and memory (Li et al., 2012). No
significant changes in GluI and Slc1a3 gene expression were observed in vitro.
Differences in glutamatergic gene expression between in vivo and in vitro L-lactate
administration may be due in part to the absence of neurons and neuronal activity in
astrocyte-only cultures. For example, astrocyte-dependent regulation of glutamate
neurotransmission is increased in the presence of astrocyte-neuron co-cultures (Hasel et
al., 2017; Pellerin & Magistretti, 1994). In addition, changes in these genes were observed
in the striatum, a region engaged in motor behavior, but not in the ETC, a region not
engaged in motor behavior during treadmill walking (Wang et al., 2013, 2016). However,
75
in walking mice, these L-lactate-induced changes in astrocyte-specific genes involved in
glutamate neurotransmission were not accompanied by changes in synaptogenesis as
indicated by the lack of change in synaptic protein markers PSD-95 and synaptophysin
expression nor improved motor performance on the rotarod. Future studies will
determine if higher levels of neuronal engagement or activity through high intensity
treadmill running are necessary to induce synaptogenesis in the presence of L-lactate.

In the CNS, L-lactate is highly expressed in astrocytes and may function as either
a signaling molecule or as a metabolite (Barros, 2013; Brooks, 2020). With respect to cell
signaling, L-lactate acts through its Gi/o-protein coupled receptor, Hydroxycarboxylic Acid
Receptor 1 (HCAR1), the primary lactate receptor expressed in astrocytes (Lauritzen et
al., 2014). We found that both L-lactate and the HCAR1 agonist 3,5-DHBA increased the
expression of genes involved in synaptic plasticity including NTFs and the immediate
early gene cFos. 3,5-DHBA-mediated gene expression in astrocytes was significantly
higher than after L-lactate administration, which may be due to differences in receptor
affinity for 3,5-DHBA (IC50 of 1.4mM) compared to L-lactate (IC50 of ~29mM) (Lauritzen
et al., 2014). L-lactate cell signaling through HCAR1 and its effects on NTFs gene
expression as we observed is consistent with reports using other cell lines and has been
suggested to be mediated in part by cFos (Liu et al., 2012; Morland et al., 2015; Shaulian
& Karin, 2002). While HCAR1 is the predominant L-lactate signaling receptor, there are
reports that other G-protein coupled receptors may also be involved in other brain regions
(Ahmed et al., 2009; Tang et al., 2014).

76
In neurons, L-lactate is known to play an important role in the expression of genes
involved in synaptic plasticity acting through both cell signaling and metabolic pathways.
L-lactate’s transport into and between cells of the CNS is mediated by several different
transporters including monocarboxylate transporters (MCT1 and 4 on astrocytes; MCT2
on neurons) (Bélanger et al., 2011), pannexin and connexin hemichannels (Karagiannis et
al., 2016), and an unidentified potassium sensitive, activity-dependent channel (Sotelo-
Hitschfeld et al., 2015). Once inside the cell the conversion of L-lactate to pyruvate
increases NADH levels leading to the regulation of NMDA receptors, influx of calcium,
and induction of a spectrum of genes important in synaptic plasticity including protein
kinases, immediate early genes like cFos and Arc, and Bdnf (Herrera-López et al., 2020;
Izumi et al., 1997; Margineanu et al., 2018; Yang et al., 2014). L-lactate can also work
through non-NMDA pathways, which may involve the conversion of L-lactate to pyruvate
that than enters the TCA cycle to undergo oxidative metabolism to form ATP. This
process has been demonstrated to induce a number of genes and proteins involved in
synaptic plasticity including protein kinases like mitogen-activate protein kinase (MAPK)
and transcription factors such as cAMP- responsive element-binding protein (CREB)
(Hasel et al., 2017). In addition, L-lactate may work through epigenetic mechanisms
including its activity as a weak deacetylase inhibitor targeting histone deacetylation
(HDAC) and resulting in increased transcription of a wide spectrum of transcripts
(Latham et al., 2012). The role of L-lactate and HDAC activity is complex since L-lactate
has also been shown to activate the NAD-dependent deacetylase Sirtuin1 (SIRT1), which
increases the levels of peroxisome proliferator-activated receptor-gamma coactivator
(PGC-1alpha), a member of a large family of transcription coactivators that promotes the
77
expression of secreted myokine precursors including irisin, also known as fibronectin
type III domain-containing protein 5 (FNDC5), which in turn can mediate Bdnf expression
(El Hayek et al., 2019).
Limitations to consider in our studies include the need to further elucidate HCAR1-
dependent cell signaling mechanisms that may regulate NTF expression in astrocytes
and including the role of other G-protein coupled receptors. Studies were carried out in
healthy adult mice and we do not yet know if animal models of disease may respond
differently to L-lactate administration.
Our results demonstrate that the administration of L-lactate and an HCAR1 agonist
can induce astrocyte-specific transcripts involved in synaptic plasticity including NTFs
(Bdnf, Gdnf, Cntf), astrocyte morphology (Gfap, Tbsp2), and glutamate neurotransmission
(GluI, Slc1a3). In addition, our findings support the role of L-lactate in modulating
astrocyte morphology in striatum of adult mice undergoing treadmill walking. Future
studies will explore the role of L-lactate as a cell signaling molecule and its role in
metabolism in meditating the expression of astrocyte-specific genes involved in synaptic
plasticity.

78
Chapter 4: Knockdown of astrocytic monocarboxylate transporter
4 in the motor cortex leads to loss of dendritic spines and a
deficit in motor learning
Adapted from Lundquist et al., 2021. Progress in Neurobiology, in preparation.

Abstract
Astrocytes support synaptic plasticity and learning through a number of
mechanisms, including by providing energetic substrates such as L-lactate. This
intercellular coupling is called the astrocyte-neuron lactate shuttle (ANLS), wherein L-
lactate is shuttled out of astrocytes via monocarboxylate transporter (MCT) 4 and taken
up by neurons via MCT2. The ANLS has been shown to contribute to hippocampus-
dependent processes. Despite these advances, the importance of the ANLS in other
behaviors, such as motor learning, remains relatively unexplored. Motor learning is
sequenced through a well-described orchestration of cortical and subcortical activity
leading to circuit-specific synaptogenesis. While much is known about the structural and
molecular aspects of motor learning, to date there is little evidence to detail how
intercellular metabolic coupling may be relevant to motor learning. In this study, we
investigated the importance of the ANLS in motor learning through knockdown of
astrocytic MCT4 expression using a cell-type specific lentiviral vector approach. Here, we
show that a loss of astrocytic MCT4 causes diminished dendritic spine density and size,
and such diminished synaptogenesis was mediated by decreased expression of Arc. The
observed deficits in synaptic structure following MCT4 knockdown are reflected in a
79
decrease in 2-deoxyglucose uptake during an accelerating rotarod task, in which MCT4
knockdown mice exhibited impaired motor performance and learning. Collectively, this
study demonstrates an important role for astrocytic MCT4 in cortical synaptic
homeostasis and motor learning acquisition.

80
Introduction
Astrocytic contributions to neuronal health, and by extension neuroplasticity, have
historically been viewed as composed of non-active and supportive measures, such as
ion buffering and neurotransmitter recycling. Recent work, however, has challenged this
neuron-centric dogma; astrocytic activity, either through controlling synaptogenesis,
releasing energy substrates, or regulating blood flow, is now known to play a direct and
central role in neuroplastic processes, such as learning (Allen & Lyons, 2018; Verkhratsky
& Nedergaard, 2018). Astrocytic influence on learning has been best described in the
hippocampus and related circuits that process memory formation (Adamsky et al., 2018),
but astrocytes have also been shown to be important for directing and controlling
behaviors elsewhere, including the striatum and cortex (Chai et al., 2017; Khakh &
Sofroniew, 2015; Perea et al., 2014). Astrocytic coordination and shuttling of energy
substrates is vital to maintaining neuronal health (Liu et al., 2017; Polyzos et al., 2019); in
the case of L-lactate, its production by astrocytes and consumption by neurons has
repeatedly been shown to be important for hippocampal-dependent learning in mice, rats,
and chicks (Gibbs et al., 2006; Netzahualcoyotzi & Pellerin, 2020; Suzuki et al., 2011).
Astrocytic shuttling of L-lactate to neurons – termed the astrocyte-neuron lactate shuttle
(ANLS) – is a well-described phenomenon and extends beyond the hippocampus, where
it plays a role in drug-seeking behaviors (Boury-Jamot et al., 2016), stress-induced deficits
in LTP (Murphy-Royal et al., 2020) and decision making in chronic pain (Wang et al., 2017).

The ANLS is not yet known to apply universally across different brain structures
and circuits, and evidence exists in certain experimental approaches where the ANLS is
81
not necessary or relevant to neuronal function (Díaz‐García & Yellen, 2019; Dienel & Cruz,
2016). Thus, whether the ANLS contributes to motor learning – a well-studied group of
behaviors that depend upon both cortical and subcortical circuit activity for acquisition
and retention respectively (Kawai et al., 2015; Peters et al., 2017; Sathyamurthy et al.,
2020; Yin et al., 2009) – is not known. Deficits in motor learning and weakening of motor-
related circuits are hallmarks of neurodegenerative diseases, such as Parkinson’s and
Huntington’s (Petzinger et al., 2013), and these diseases have known risk factors related
to cellular metabolism (Abou-Sleiman et al., 2006; Bose & Beal, 2016; Polyzos et al.,
2019). Therefore, it is important to understand the underlying metabolic mechanisms that
dictate motor learning and, by extension, may be important for restoration of motor
circuits in disease, such as through aerobic exercise (Davies et al., 2017; Petzinger et al.,
2015).
Considering these lines of evidence, we hypothesized that the ANLS was important
for motor learning. To study the ANLS in a motor learning task, we used a cell-type
specific, Cre-recombinase mediated lentiviral approach to selectively knockdown the
expression of monocarboxylate transporter 4 (MCT4), which shuttles L-lactate out of
astrocytes to neuronal synapses. Here, we combine in vitro and in vivo approaches to
validate astrocyte-specific knockdown of MCT4 and show that loss of astrocytic MCT4
leads to a loss of dendritic spines in mouse primary motor cortex. We further
demonstrate that dendritic spine loss is mediated in part by a decrease in Arc expression
and leads to a task-specific decrease in neuronal activity measured via fluorescent 2-DG
uptake. Finally, astrocytic MCT4 knockdown diminished motor learning and performance
on the accelerating rotarod. Together, these results provide novel evidence for astrocytic
82
MCT4 as important to cortical dendritic spine maintenance and motor learning
acquisition and extends the ANLS as a possible model of cellular metabolic coordination
in motor circuits.

83
Methods
Mice
Heterozygous Aldh1l1-CreERT2
+/-
(termed Cre
(+ve)
mice) mice expressing Cre
recombinase selectively in astrocytes (Srinivasan et al., 2016) and Aldh1l1-CreERT2
-/-

(Cre
(-ve)
mice) littermates were generated from in-house breeding pairs of Aldh1l1-
CreERT2
+/-
and Aldh1l1-CreERT2
-/-
mice at the University of Southern California. All
animals were group housed, up to five animals per cage, and had ad libitum food and
water access on an inverse dark-light cycle (lights off/on 0700/1900 hours). CreERT2
gene expression was assessed by PCR with primers towards CreERT2 or the wildtype
sequence insert according to previously described methods (Srinivasan et al., 2016).
Daily intraperitoneal (i.p) injections of tamoxifen (Sigma, St. Louis, MO; 20 mg/ml,
dissolved in corn oil; 75 mg tamoxifen per kg body weight) were administered for five
days to induce MCT4 shRNA gene expression (vector construction described below) in
Cre
(+ve)
mice. Cre
(-ve)
mice also received daily injections of tamoxifen as control. All
experimental procedures in animals were approved by the Institutional Animal Care and
Use Committee at the University of Southern California (Protocol No. 21044) and carried
out in compliance with the National Institutes of Health Guide for the Care and Use of
Laboratory Animals, 8
th
Edition, 2011.

Plasmids and Vector Construction
MSCV (Murine Stem Cell Virus)-CreERT2 puro plasmid was a gift from Tyler Jacks
(Addgene, Watertown, MA; plasmid # 22776; http://n2t.net/addgene:22776; RRID:
Addgene_22776). pCL-Eco was a gift from Inder Verma (Addgene plasmid # 12371;
84
http://n2t.net/addgene:12371; RRID: Addgene_12371). pCMV-VSV-G was a gift from Bob
Weinberg (Addgene plasmid #8454; http://n2t.net/addgene:8454; RRID: Addgene_8454).
pCMV-deltaR8.2 was a gift from Didier Trono (Addgene plasmid # 12263;
http://n2t.net/addgene:12263; RRID: Addgene_12263). The pSico-EGFP plasmid was a
gift from Tyler Jacks (Addgene plasmid #11578; http://n2t.net/addgene:11578; RRID:
Addgene_11578). pSico contains loxP segments, flanking an insert site for pre-designed
short hairpin RNA (shRNA) sequences, allowing for Cre-mediated recombination. To
target the mouse MCT4 gene (Slc16a3), shRNA sequences specific to Slc16a3 were
designed using the pSicoligomaker software (Ventura Laboratory,
https://venturalaboratory.com/home/downloads/) and synthesized as single stranded
cDNA oligonucleotides with Hpa1 and Xho1 overhangs (IDT Technologies, Coralville, IA).
The shRNA MCT4 top and bottom cDNA sequences were annealed together and cloned
into the pSico-EGFP vector plasmid using restriction enzymes Hpa1 and Xho1. For MSCV-
CreERT2 retroviral vector, MSCV-CreERT2, pCL-Eco, and pCMV-VSV-G plasmids were
transfected into HEK-293T cells using calcium phosphate as previously described
(Llewellyn et al., 2019). To construct pSico-EGFP-shMCT4 lentivirus vector (termed Lenti-
pSico-EGFP-shMCT4), pSico-EGFP-shMCT4, pCMV delta R8.2, and pCMV-VSV-G plasmids
were all transfected into HEK-293T cells. After 2 days, the supernatant containing the
vector was collected and passed through a 0.45um filter. In addition, 100X concentrated
vector was made by ultracentrifugation (100,000 xg for 2 hours) through a 20% sucrose
cushion, resuspended in 1:1 PBS (phosphate buffered saline): FBS (fetal bovine serum)
and stored at -80
o
C until use.
85

Astrocyte Cell Line Expressing CreERT2
The C8-D1A astrocyte cell line (CRL-2541, ATCC, Manassas, VA) was purchased
and maintained in DMEM-10 (DMEM with 10% fetal bovine serum and 1%
penicillin/streptomycin: Genesee Scientific, San Diego, CA). C8-D1A cells were modified
to express an inducible CreERT2 protein (known further as C8-D1A-CreERT2 astrocytes)
by transfecting C8-D1A cells with MSCV-CreERT2 retroviral vector. CreERT2 expression
was constitutively selected for using puromycin (2 ug/mL, Sigma) and maintained in
DMEM-10 media.

Immunocytochemistry of Astrocyte Cell Cultures
C8-D1A-CreERT2 astrocytes were seeded on poly-D-lysine coated glass coverslips
(1x10
5
cells/coverslip) and transduced lentivirus pSico-EGFP-shMCT4 (1x10
10
genome
copies/ml). After 5 days, cells were washed and fixed with ice-cold 4% PFA-PBS and
stained with rabbit anti-SOX9 (astrocytic nuclear marker; 1:2000, Millipore, Burlington,
MA; Cat# AB5535, RRID: AB_2239761) and Alexa 568-conjugated secondary antibody
(1:5000, Thermo Fisher Scientific, Waltham, MA; Cat# A11011, RRID: AB_143157) before
mounting on microscope slides (Vectashield Hardset Antifade with DAPI; Vector
Laboratories, Burlingame, CA). Coverslip images were obtained on an IXB-DSU spinning
disk Olympus BX-61 microscope (Olympus America, Center Valley, PA) equipped with an
ORCA-R2 digital CCD camera (Hamamatsu Photonics, Bridgewater, NJ) and MetaMorph
Advanced software (Molecular Devices, San Jose, CA).

86
Quantitative real-time PCR
Following MCT4 shRNA expression in C8-D1A-CreERT2 astrocytes, total RNA was
extracted (Zymo Research, Irvine, CA) and 100 ng of RNA reverse transcribed to cDNA
(PCR Biosystems, Wayne, PA) as previously described (Lundquist et al., 2019).
Quantitative RT-PCR was performed on an Eppendorf Mastercycler ep Realplex as
previously described (Lundquist et al., 2019) using the following primer pairs: Slc16a1
forward: TGTTAGTCGGAGCCTTCATTT; reverse: CACTGGTCGTTGCACTGAATA; Slc16a3
forward: TCACGGGTTTCTCCTACGC; reverse: GCCAAAGCGGTTCACACAC; Actb forward:
GGCTGTATTCCCCTCCATCG; reverse: CCAGTTGGTAACAATGCCATG.

Flow Cytometry
Flow cytometry was used to assess EGFP expression in C8-D1A or C8-D1A-
CreERT2 astrocytes transfected with Lenti-pScio-EGFP-shMCT4. Cells were treated with
0.25% Trypsin (Genesee Scientific), fixed in 4% paraformaldehyde, and separated on a
FACS Canto Flow Cytometer (BD Biosciences, San Jose, CA). Quantitative analysis of cell
numbers was performed with FloJo software (BD Biosciences).

Calcium Imaging
C8-D1A-CreERT2 astrocytes were seeded in 96 well plates and immediately
infected with Lenti-pSico-EGFP-shMCT4 (1x10
12
viral particles/ml). 1uM 4-
hydroxytamoxifen (4-OHT; Sigma) was added to select wells 5 days prior to imaging to
allow for adequate Cre-mediated recombination and expression of MCT4 shRNA;
remaining wells did not receive 4-OHT treatment as control. On imaging day, Fluo-4AM
87
(50ug; Biotium, Fremont, CA; Cat# 50018) was dissolved in 50ul Pluronic F-127 (Biotium,
Cat# 59004) and further diluted with 10mL HBSS with 20 mM HEPES to make Fluo-4
solution. Astrocyte culture medium was removed, cells were washed with ice-cold HBSS,
and 200 ul of Fluo-4 solution was added to each well and incubated at 37°C for 45
minutes. Fluo-4 solution was removed, replaced with fresh HBSS + 20mM HEPES, and
the plate returned to 37°C for an additional 20 minutes to allow for complete cleavage of
Fluo-4. Calcium imaging was performed on a Synergy H1 Hybrid Multimode Reader
(BioTek, Winooski, VT) with 494nm excitation and 516nm emission filters. Baseline
fluorescence was recorded every 30 seconds for 3.5 minutes, and then 100uM ATP
(Sigma) was injected and mixed in all wells, and fluorescence recorded every 30 seconds
for another 9 minutes for a total of 26 measurements (including t =0). Fluorescence
intensity in each well was normalized to the average baseline fluorescence and plotted
across the imaging session as deltaF/F.

Stereotactic Surgery for Vector Delivery to Motor Cortex
Mice were anesthetized with 3% isoflurane and maintained at 1.5% isoflurane
throughout the surgery. Cre
+
and Cre
-
littermates were placed in a stereotactic frame
(Stoelting, Wood Dale, IL) and their skulls were exposed with a single scalpel cut. Bore
holes were drilled in the skull over the primary motor cortex using stereotactic
coordinates and Lenti-pSico-EGFP-shMCT4 (1.0ul, 1x10
12
genome copies/mL) was
injected bilaterally using a 26-gauge syringe (701N, Hamilton, Reno, NV) at a rate of 0.20
ul/min. Lentivirus was targeted to the primary motor cortex (anterior-posterior [AP]: +1.5
mm; medial-lateral [ML]: ±1.8 mm; dorsal-ventral [DV]: -1.5 mm from Bregma on the
88
surface of the skull). The incision was closed with a single stainless-steel surgical staple
(Stoelting) and mice allowed to recover on heating pads with Medigel (containing 1mg
carprofen, Clear H2O, Westbrook, ME) refreshed daily for 3 days. Mice were used 2 weeks
after stereotactic surgery to allow for adequate viral expression.

Immunohistochemistry for Transfection of MCT4 shRNA Vector
Mice were transcardially perfused with ice-cold saline and 4% PFA-PBS. Whole
brains were postfixed with 4% PFA-PBS before being transferred to 30% sucrose until they
sank. Brains were frozen in dry ice-cooled 2-methylbutane and sliced on a freezing
cryostat in the coronal orientation at 50um before being stored in cryoprotectant solution
at -20ºC. Sections were processed for immunohistochemistry as previously described
(Lundquist et al., 2019). Briefly, sections were washed and permeabilized in TBS with
0.05% Tween-20 (TBST), nonspecific staining was blocked with blocking solution (TBST
and 4% normal donkey or normal goat serum) and incubated overnight at 4ºC with SOX9
antibody (1:2000, Millipore, AB5535, RRID: AB_2239761). Sections were washed and
incubated with Alexa 568-conjugated secondary antibody (1:5000, Thermo Fisher
Scientific, A11011, RRID: AB_143157) then coverslipped (Vectashield Hardset Antifade
with DAPI; Vector Laboratories). All confocal images were taken on an IXB-DSU spinning
disk Olympus BX-61 (Olympus America) and captured with an ORCA-R2 digital CCD
camera (Hamamatsu) and MetaMorph Advanced software (Molecular Devices). Mouse
brain sections were imaged for co-localization of SOX9 (an astrocyte-specific nuclear
marker) and GFP (expressed by the MCT4 shRNA lentivirus) in the motor cortex to assess
relative transduction of astrocytes by stereotactic lentivirus injection. Primary motor
89
cortex was located using anatomical landmarks and images taken in both the RFP and
GFP channels throughout the primary motor cortex. Images were acquired across at least
three sections per mouse. Colocalization of RFP and GFP signals were analyzed using
ImageJ (Schindelin et al., 2012), as previously detailed (Lundquist et al., 2019).

Western Immunoblotting
Mice were rapidly euthanized by cervical dislocation and whole brains quickly
removed and the primary motor cortex (1.0 mm anterior to Bregma; dorsal to the corpus
callosum spanning 1.0 mm to 2.5 mm from midline) rapidly dissected, snap frozen on dry
ice, and stored at -80
o
C. Cortical tissue samples were lysed in RIPA buffer with protease
and phosphatase inhibitors (MSSAFE, Sigma), centrifuged at 16,000 x g and the soluble
fraction collected for subsequent protein analysis. Total protein content was determined
by BCA analysis (ThermoFisher) and 20 ug of protein resolved on a 10% Tris-Glycine gel
by electrophoresis (BioRad, Hercules, CA). Total protein was transferred to nitrocellulose
membranes (BioRad), blocked in OneBlock buffer (Cat# 20-314, Genesee Scientific) and
probed with the following antibodies overnight at 4ºC: mouse anti-PSD95 (1:2000,
Millipore, Cat# MAB1596, RRID: AB_2092365), mouse anti-synaptophysin (1:2000,
Abcam, Cambridge, MA; Cat# ab8049, RRID: AB_2198854), rabbit anti-MCT1 (1:1000,
Thermo Fisher Scientific, Cat# PA5-72957, RRID: AB_2718811) rabbit anti-MCT4 (1:1000,
Novus Biologicals, Littleton, CO; Cat# NBP1-81251, RRID: AB_11033184), rabbit anti-Arc
(H-300, 1:500, Santa Cruz Biotechnology, Dallas, TX; Cat# sc-15325, RRID: AB_634092),
and mouse anti-beta actin (1:5000, LI-COR, Lincoln, NE; Cat# 926-42212, RRID:
AB_2756372). Membranes were washed, incubated with corresponding goat anti-mouse
90
or goat anti-rabbit conjugated near-infrared secondary fluorescent antibodies (1:5000, LI-
COR, Cat# 926-32211, RRID: AB_621843; and Cat# 926-68070, RRID: AB_10956588), and
scanned on an Odyssey imaging system (LI-COR). Relative protein expression was
quantified by optical density and normalized to beta-actin as loading control.

Golgi-Cox Staining for Dendritic Spine Density
Mice were rapidly euthanized by cervical dislocation and decapitation and whole
brains removed. Brains were used for Golgi-Cox staining (FD NeuroTechnologies,
Columbia, MD) following the manufacturer’s protocol as previously described (Toy et al.,
2014). Briefly, brains were impregnated for two weeks, slowly frozen on dry-ice chilled 2-
methylbutane, sectioned in the coronal plane in 100um thickness, and processed
according to the manufacturer’s protocol. Pyramidal neurons residing in layer 5 of the
primary motor cortex were identified, and 15 to 20 cells were imaged per mouse using an
Olympus BX50 (Olympus America) and captured with a KAPELLA digital CCD camera
(Jenoptik, Jena, Germany). Golgi impregnated cells were imaged across at least three
sections per mouse, and multiple spans of dendrites were counted and analyzed per cell
using FIJI and BIOQUANT software as previously described (Toy et al., 2014). Total
dendritic counts were averaged across all cells imaged per mouse.

Near-Infrared 2-Deoxyglucose Mapping
Following the final trial on the accelerating rotarod, mice were injected
intraperitoneally with 10nmol of near-infrared-conjugated 2-deoxyglucose (2DG-IR; LI-
COR, Cat# 926-08946). Mice were placed on the rotarod at a slow, constant speed (8 rpm)
91
for 30 minutes before being euthanized, perfused, and brains removed and processed for
immunohistochemistry. Four slices containing structures of interest (motor cortex [AP:
+1.3 from Bregma], dorsal striatum [AP: +0.5], motor nuclei of thalamus [AP: -1.3], and
dorsal hippocampus [AP: -1.8]) were selected per mouse; all efforts were made to match
Bregma levels across individual mice and groups. Slices were washed in PBS, mounted
on gelatin-subbed microscope slides, and slides scanned using identical settings on an
Odyssey Near-Infrared imaging system (LI-COR) before analyzing with FIJI. First, overall
fluorescent intensity of matching slices in both groups were analyzed in FIJI; no
differences were observed between mice or groups, indicating equal administration and
uptake of the tracer. 2DG-IR positive signals were then thresholded to exclude any non-
specific background, aligned to a standard mouse atlas (Dong, 2008), and total optical
density of 2DG-IR puncta within specific anatomical regions was quantified and averaged
across sections for each mouse.

Motor Behavior Testing
The impact of MCT4 knockdown in M1 was assessed with several tests of motor
behavior including activity in the open field, reversal climbing on the pole test, and latency
to fall from the rotarod. All behavioral assessments took place during the first half of the
dark cycle (0900-1200 hours). Mice were brought to the behavioral suite and allowed to
acclimate to the room for 30 minutes before any behavior testing.
Open Field Test was used to assess total locomotion behavior. Behavioral testing was
conducted twice, once before tamoxifen administration, and once prior to motor learning
on the accelerating rotarod. Mice were placed in an open field chamber (30cm x 30cm x
92
30cm white plywood box) for five minutes and total movement was recorded by a digital
video camera at 30 frames per second. Locomotion was analyzed using the EzTrack
pipeline (Pennington et al., 2019) and distance was binned in one-minute intervals. Motor
behavior was expressed as total distance traveled in meters.
Pole Test was used to assess the ability to descend a wooden pole and test speed and
dexterity. Behavioral testing with the pole test was conducted twice, once before
tamoxifen administration, and once prior to motor learning on the accelerating rotarod.
Briefly, mice were placed at the top of a 50 cm wooden pole in an empty cage filled with
normal bedding, and the time to descend to the base of the pole is recorded. The trial was
repeated five times, with a one-minute inter-trial interval.
Accelerating Rotarod was used to assess motor behavior and learning of the task.
Following tamoxifen administration to activate MCT4 shRNA expression in Cre
+
mice, all
mice were tested for motor learning and coordination on the accelerating rotarod with
slight modifications as previously described (Rothwell et al., 2014). Briefly, mice were
oriented to the stationary rod for three minutes before their first trial. Mice then
completed two trials per day for four days, with a 5-minute intertrial during which mice
were returned to their home cage. The rotarod (3 cm diameter rod, divided into five lanes;
Ugo Basile, Comerio, Italy) accelerated over the course of 300 seconds from 4 to 40 rpm,
and speed at time of fall and latency to fall automatically recorded by magnetic trip
plates. A trial ended when the mouse made a complete backward revolution, fell off, or
reached the 300 second threshold. Learning curves were fit with linear regression
analysis and slope defined as the learning rate and the Y-intercept of the fitted regression
93
line defined as the initial coordination, a measure of baseline motor coordination on the
accelerating rotarod (Rothwell et al., 2014).

Statistical Analysis
Sample sizes were calculated based on our previously published work and all efforts were
made to minimize the number of mice used. For glucose mapping, histology, and
dendritic spine analysis, we used three male mice (n = 3) per group. For western blotting,
we used five male mice (n = 5) per group. For behavioral experiments, we used eight male
mice (n = 8) per group. In vitro experiments were conducted with at least three
independent biological replicates from two independent culture preparations. All data
was tested for normality using D’Agostino & Pearson test. If both groups (control and
MCT4 shRNA) passed normality tests, an unpaired or paired t-test was used; this was the
case with almost all the data analyzed in this paper. In the case of peak calcium
fluorescence and dendritic spine analysis, these data were not normally distributed and
were analyzed using a Kolmogorov-Smirnov test. To compare overall performance on
behavioral tasks (open field, pole test and accelerating rotarod), a 2-way ANOVA with
Sidak’s multiple comparisons was used. All statistical analyses were conducted using
Prism (version 8.2, GraphPad) with significance denoted as p < 0.05.

94
Results
In Vitro Validation of MCT4 shRNA Knockdown Vector
To selectively knockdown MCT4 expression in astrocytes, a Cre-inducible pSico
expression plasmid carrying a short hairpin RNA (shRNA) specific to the mouse transcript
Slc16a3 (MCT4) (Ventura et al., 2004) was constructed (Fig. 4.1a). The C8-D1A astrocyte
cell line was transfected with plasmid MSC::CreERT2 puro (Kumar et al., 2009) to express
a tamoxifen-inducible Cre recombinase in a cell line termed C8-D1A-CreERT2 astrocytes.
Both C8-D1A and C8-D1A-CreERT2 astrocytes were stably transfected with pSico-EGFP-
shMCT4 as demonstrated by expression of GFP using flow cytometry at 2-, 4-, and 7-days
post-transfection (Fig. 4.1b) Following induction of Cre-recombinase in C8-D1A-CreERT2
astrocytes by application of the tamoxifen metabolite 4-hydroxytamoxifen (4-OHT) there
was a statistically significant reduction in GFP expression at 2 and 4 days (day 4, unpaired
two-tailed t-test, t=4.099, df=4, p = 0.015; day 7, unpaired two-tailed t-test, t=22.67, df=4,
p < 0.001) (Fig. 1c). There was no change in GFP expression in C8-D1A cells transfected
with pSico-GFP-shMCT4 following 4-OHT administration indicating cell-specific Cre-
mediated expression. These data are quantitated in (Fig. 4.1c).

95
Figure 4.1 – In vitro validation of pSico-EGFP-shMCT4 vector. a. Schematic highlighting pSico-
EGFP plasmid carrying Cre-dependent MCT4 shRNA, with MCT4 shRNA sequence detailed in
inset box. b. Representative flow cytometry plots of C8-D1A-CreERT2 astrocytic cell line
transduced with pSico (C8-D1A-CreERT2::pSico-EGFP-shMCT4) treated with saline or 4OHT (left),
and quantification of EGFP expression following 1μM 4OHT treatment (right). c. Representative
flow cytometry plots of C8-D1A astrocytic cell line transduced with pSico (C8-D1A::pSico-EGFP-
shMCT4) treated with saline or 4OHT, and quantification of EGFP expression in C8-D1A::pSico-
EGFP-shMCT4 transduced astrocytes following 1μM 4-OHT treatment. Data are mean ± s.e.m.; *
p < 0.05, *** p < 0.001, unpaired two-sided t-test (n = 3 experiments per group).

To evaluate the efficacy of vector-mediated knockdown MCT4 transcript, pSico-
EGFP-shMCT4 was packaged into lentivirus particles and transfected into C8-D1A-
CreERT2 astrocytes. Following 7 days of 4-OHT administration, MCT4 transcript and
96
protein expression were evaluated by qRT-PCR and western blotting, respectively. pSico-
shMCT4 decreased MCT4 transcript expression by 68 ± 7% (unpaired two-tailed t-test,
t=3.102, df=6, p = 0.021) which resulted in an over 70% decrease in protein expression
(Fig. 4.2a). Knockdown by pSico-shMCT4 was specific for MCT4 mRNA and did not alter
expression of MCT1 (Slc16a1) transcript expression (unpaired two-tailed t-test, t=0.101,
df=6, p = 0.923) (Fig. 4.2a). To verify that MCT4 knockdown was not cytotoxic, cell viability
was evaluated using real-time imaging with the calcium indicator Fluo-4 in the presence
of 100 μM ATP to activate astrocyte specific P2Y1 and P2X7 receptor channels (Agulhon
et al., 2008). Both non-transfected and pSico-shMCT4 transfected C8-D1A-CreERT2
astrocytes responded similarly to ATP exposure, with transfected cells (Kolmogorov-
Smirnov test, D=0.563, p < 0.001) (Fig. 4.2b).
97

Figure 4.2 – pSico-EGFP-shMCT4 causes loss of MCT4 expression in C8-D1A-CreERT2
astrocytes. a. Loss of MCT4 (Slc16a3) mRNA and protein expression following activation of
shRNA expression. Data are mean ± s.e.m; ns, p > 0.05, * p < 0.05, unpaired two-sided t-test (n =
3 replicates per group). Bands shown are representative from the same experiment. b. MCT4
knockdown does not affect ATP-mediated calcium response in astrocytes (measured with Fluo-
4). Data are mean ± s.e.m. Some error bars are too small to be displayed (n = 32 wells across 3
experiments per group).

98
In Vivo Assessment of Vector Targeting on MCT Expression in M1 Cortex
Lenti-pSico-EGFP-shMCT4 particles were injected into the primary motor cortex
(M1) of mice expressing an astrocyte-specific tamoxifen-inducible Cre recombinase
(Cre
(+)ve
, termed MCT4 knockdown or shRNA; or Cre
(-)ve
mice, termed control) (Srinivasan
et al., 2016) (Fig. 4.3a). Lentiviral transduction based on EGFP fluorescence was limited
to cells in M1, without spreading into adjacent secondary motor cortex (M2) (Fig. 4.3b).
Within M1, approximately 42 ± 12% of transduced cells were positive for the astrocyte
specific marker Sox9 (Fig. 4.3b). To promote Cre recombinase and shMCT4 expression,
both MCT4 knockdown and control mice were administered tamoxifen for five days and
ten days after the last injection, molecular and behavioral effects of MCT4 knockdown
were determined. MCT4 protein expression was significantly decreased in M1 (unpaired
t-test, t=4.600, df=6, p = 0.004) following Cre induction. There was no change in MCT1
expression (unpaired t-test, t=0.792, df=6, p = 0.459) demonstrating the specificity of the
shRNA targeting (Fig. 4.3c).
99

Figure 4.3 – In vivo validation of Lenti-pSico-EGFP-shMCT4 to knockdown astrocytic MCT4. a.
Experimental schematic and timeline for in vivo application of Lenti-pSico-EGFP-shMCT4. b.
Example micrograph showing localization of lentivirus to primary motor cortex (M1), and Lenti-
pSico-EGFP-shMCT4 (LV) localizes with astrocytic nuclear marker SOX9 in M1, with ~42% of all
transfected cells being astrocytes. c. Activation of MCT4 shRNA decreases expression of MCT4
protein but not MCT1 protein in M1. Data are mean ± s.e.m. ns, p > 0.05; ** p < 0.01, unpaired two-
sided t-test (n = 5 mice per group). Bands shown are representative lanes.

100
MCT4 Knockdown Impacts Motor Behavior Learning
Motor performance following knockdown of MCT4 in M1 cortical astrocytes was
determined with several tests including the open field, descent from the pole test, and
latency to fall on the accelerating rotarod. Both MCT4 knockdown and control mice did
not show any difference in overall distance traveled in the open field (unpaired t-test,
t=1.064, df=14, p = 0.305) (Fig. 4.4a) or time to descend a 50cm pole (unpaired t-test,
t=1.134, df=14, p = 0.276) (Fig. 4.4b). Loss of astrocytic MCT4 in the motor cortex
decreased the overall performance on the accelerating rotarod (two-way ANOVA, MCT4
shRNA: F(1,112) = 8.644, p = 0.004) (Fig. 4.4c), highlighted by a 43% reduction in learning
rate (simple linear regression, F(1, 124) = 4.604, p = 0.034) (Fig. 4.4d). Additionally, we did
not observe any difference between groups following stereotactic surgery or tamoxifen
administration on locomotion in the open field (Fig. 4.5a-d) or performance on the pole
test (Fig. 4.5e-g).
101
Figure 4.4 – Astrocyte-specific knockdown of MCT4 impairs motor learning without affecting
locomotion or motor coordination. a. MCT4 knockdown in motor cortex does not affect overall
locomotion in the open field. b. MCT4 knockdown in motor cortex does not affect time to descend
on the pole test. c. MCT4 knockdown in motor cortex decreases overall performance on the
accelerating rotarod, assessed by terminal speed. d. MCT4 knockdown in motor cortex
diminished learning rate (calculated as the slope of the performance curve) on the accelerating
rotarod. Data are mean ± s.e.m., ** p < 0.01 (n = 8 mice per group; repeated measures 2-way
ANOVA with Bonferroni correction) (c) or boxplots showing quantiles (25, 50, 75%) with central
line marking the median and plus denoting the mean, ns, p > 0.05, * p < 0.05 (n = 8 mice per group;
unpaired two-sided t-test) (a, b, d).

102

Figure 4.5 – Lentiviral injection and tamoxifen administration do not impact motor behavior. a.
Locomotion as measured in the open field does not differ within or between groups before
MCT4 shRNA induction. b. Total distance traveled in the open field is not impacted due to
surgery, pre-tamoxifen induction of MCT4 shRNA (n = 8 mice per group, unpaired two-sided t-
test). c. Learning rate in the accelerating rotarod is not correlated with total locomotion in the
open field in control mice (simple linear regression). d. Learning rate in the accelerating rotarod
is not correlated with total locomotion in the open field in MCT4 shRNA mice (simple linear
regression). e. Coordination as measured in the pole test does not differ within or between
groups before MCT4 shRNA induction. f. Best time to descend in the pole test is not impacted
due to surgery, pre-tamoxifen induction of MCT4 shRNA. g. Lentiviral surgery and tamoxifen
induction of MCT4 shRNA do not affect average first-last trial times to descend in the pole test.

103
MCT4 knockdown Reduces Dendritic Spine Density in M1
Astrocytic MCT4 is important for learning and memory and the associated
learning-induced increase in dendritic spine formation in the hippocampus
(Netzahualcoyotzi & Pellerin, 2020; Suzuki et al., 2011; Vezzoli et al., 2020). To determine
the effect of MCT4 knockdown on dendritic spine density on M1 neurons Golgi-Cox
impregnation was used to label neurons following tamoxifen-induced MCT4 knockdown
(Fig. 4.6a). Knockdown of astrocytic MCT4 significantly decreased both dendritic spine
density (Kolmogorov-Smirnov test, D=0.351, p < 0.001) and dendritic spine width
(Kolmogorov-Smirnov test, D=0.143, p < 0.001) compared to control mice (Fig. 4.6b).
There was no evidence of cell loss based on cortical thickness of M1 assessed with Nissl
staining (unpaired t-test, t=0.111, df=36, p = 0.912) (Fig. 4.6c) in agreement with
astrocytic MCT4 knockdown in the hippocampus where no cell loss was observed
(Netzahualcoyotzi & Pellerin, 2020). In addition, there was no evidence of non-specific
dendritic spine loss due to surgery or M1-based MCT4 knockdown, as dendritic spine
analysis in adjacent M2 secondary motor cortex showed no effect of M1-MCT4
knockdown on dendritic spine density or width (Fig. 4.7).
104

Fig. 4.6 – Loss of astrocytic MCT4 causes cortical dendritic spine loss. a. Representative layer
5 pyramidal neuron in primary motor cortex of control (top) and MCT4 shRNA (bottom) mouse
filled via Golgi-Cox impregnation, with example dendritic branches highlighted. b. MCT4
knockdown causes decrease in dendritic spine density (top) and dendritic spine width (bottom).
Data are histogram of the frequency distribution of spine density (top) and spine head width
(bottom), with boxplot insets, showing quantiles (25, 50, 75%) with central line marking the
median and plus denoting the mean; *** p < 0.001 (n = 3 mice per group; Kolmogorov-Smirnov
test). c. Representative images of Nissl-stained motor cortex from control and MCT4 shRNA
mice; dashed lines represent approximate anatomical borders for the primary motor cortex.
MCT4 knockdown in primary motor cortex does not change cortical thickness, reflecting no
change in neuronal density. Data are mean ± s.e.m. (n = 3 mice per group).

105

Figure 4.7 – Loss of astrocytic MCT4 in M1 does not affect dendrites in M2. a. Representative
layer 5 pyramidal neuron in primary motor cortex of control (top) and MCT4 shRNA (bottom)
mouse filled via Golgi-Cox impregnation, with example dendritic branches highlighted. b. MCT4
knockdown in M1 does not affect dendritic spine density (top) or dendritic spine width (bottom)
in M2. Data are histogram of the frequency distribution of spine density (top) and spine head
width (bottom), with boxplot insets, showing quantiles (25, 50, 75%) with central line marking the
median and plus denoting the mean; ns p > 0.05 (n = 3 mice per group; Kolmogorov-Smirnov test).

MCT4 Knockdown Reduces Synapse-Specific Protein Marker Expression
Previous reports suggested that loss of MCT4 in the hippocampus reduces
dendritic spine formation, possibly through decreased expression of the activity-
regulated cytoskeleton-associated protein Arc (Suzuki et al., 2011). Western
immunoblotting was used to determine the patterns of expression of Arc protein as well
as synaptophysin and post-synaptic protein of 95 kDa (PSD95), proteins localized at the
pre-synaptic and post-synaptic termini, respectively, in M1 cortex of MCT4 knockdown
106
mice compared to control mice. Synaptophysin protein expression was unchanged
(unpaired t-test, t=1.741, df=8, p = 0.120) (Fig. 4.8a); however, PSD95 and Arc protein
expression were significantly decreased following MCT4 knockdown compared to
control mice (PSD95: unpaired t-test, t=2.474, df=8, p = 0.039; Arc: unpaired t-test,
t=2.549, df=8, p = 0.034) (Fig. 4.8b, c). Interestingly, in line with previous reports (Hosp et
al., 2013), learning rate on the accelerating rotarod significantly correlated with both Arc
protein expression and PSD95 protein expression in the primary motor cortex (Fig. 4.9).
Figure 4.8 – Loss of astrocytic MCT4 decreases postsynaptic protein expression. a. MCT4
knockdown does not significantly decrease expression of pre-synaptic protein synaptophysin. b.
MCT4 knockdown decreases expression of postsynaptic protein PSD95. c. MCT4 knockdown
decreases expression of Arc (activity-regulated cytoskeletal-associated protein). Representative
bands from western blots of primary motor cortex from control (left) and MCT4 shRNA (right)
mice. Data are mean ± s.e.m. ns, p > 0.05; *, p < 0.05 (n = 5 mice per group; unpaired, two-sided t-
test). All western blot quantifications are normalized relative to each lane’s β-actin intensity, then
normalized such that the mean of the control group is equal to 1.

107
Figure 4.9 – Motor learning correlates with postsynaptic protein expression. a. Learning rate on
the accelerating rotarod significantly correlates with PSD95 protein expression (simple linear
regression). b. Learning rate on the accelerating rotarod significantly correlates with Arc protein
expression (simple linear regression). c. Relationship between average dendritic spine density
(left) and spine width (right) with learning rate between control and MCT4 shRNA groups. Data
are mean ± s.e.m. on both the X-axis and Y-axis.

108
MCT4 Knockdown Reduces 2-Deoxygluocse Uptake
The effect of MCT4 knockdown in astrocytes on neuronal metabolism was
assessed using the near-infrared fluorescent 2-deoxyglucose (2DG-IR) tracer, which has
previously been shown to be taken-up by mouse cortical neurons in response to activity
(Lundgaard et al., 2015). Mice were administered 10 nmol of 2DG-IR i.p. while walking on
the rotarod for 30 minutes, and the degree of 2D-IR in the brain determined in 3 brain
regions involved in motor behavior including the primary motor cortex (Fig. 4.10a, ‘i’),
dorsal striatum (Fig. 4.10a, ‘ii’), and motor nuclei of the thalamus (focusing on
ventroanterior and ventroposterior nuclei) (Fig. 4.10a, ‘iii’), as well as a region of the brain
not involved in motor behavior, the dorsal hippocampus (Fig. 4.10a, ‘iv’). All three motor
areas showed a decrease in 2DG-IR fluorescence in MCT4 knockdown mice relative to
control mice (motor cortex: unpaired t-test, t=4.994, df=29, p < 0.001; dorsal striatum:
unpaired t-test, t=2.498, df=32, p = 0.018; motor thalamus: unpaired t-test, t=2.161, df=33,
p = 0.038) (Fig. 4.10b). The dorsal hippocampus showed no significant difference in 2DG-
IR uptake (unpaired t-test, t=1.614, df=33, p = 0.116).

109
Figure 4.10 – Astrocytic MCT4 knockdown causes task-specific decrease in glucose uptake in
motor-related brain regions. a. Representative images of near-infrared analog of 2-deoxyglucose
(2DG-IR), pseudocolored to detail gradient of fluorescent uptake and select regions of interest
including motor cortex (i); dorsal striatum (ii); motor thalamus (ventroanterior and ventroposterior
nuclei) (iii); and dorsal hippocampus (iv). All efforts were made to match high-quality,
representative images between groups. Scale bars are 2mm in hemi-sections and 1mm in
cropped regions of interest. b. MCT4 knockdown in the motor cortex causes decreases in 2DG-
IR uptake in motor related regions (motor cortex, dorsal striatum, and motor thalamus) but not in
dorsal hippocampus. Data are boxplots, showing quantiles (25, 50, 75%) with central line marking
the median and plus denoting the mean, and normalized such that the mean fo the control group
is equal to 100%. ns, p > 0.05; *, p < 0.05; ***, p < 0.001 (n = 3 mice per group; unpaired two-tailed
t-test).
110
Discussion
The purpose of our study was to investigate the role of astrocytic MCT4 in motor
learning and synaptic structure in the mouse primary motor cortex. Brain-specific MCTs,
including astrocytic MCT4, have been repeatedly shown to critically contribute to synaptic
plasticity and long-term memory formation in the hippocampus (Suzuki et al., 2011;
Vezzoli et al., 2020). Thus, we explored whether astrocytic MCT4 played a similarly
important role in coordinating motor behavior and synaptic structural remodeling. Using
a temporal-, regional- and cell-type-specific approach in vivo, we knocked down astrocytic
MCT4 expression in mouse motor cortex. We found that a loss of astrocytic MCT4
expression selectively impaired motor learning and performance on the accelerating
rotarod, a deficit that corresponded with a loss of postsynaptic dendritic spines and a
parallel decrease in PSD95 and Arc protein expression in the motor cortex. In addition,
we observed diminished neuronal activity in motor related regions using task-specific 2-
deoxyglucose mapping following knockdown of astrocytic MCT4, further providing
possible mechanisms underlying reduced motor learning. Our findings provide the first
evidence, to our knowledge, of a role for astrocytic MCT4 in mediating cortical synaptic
homeostasis and motor learning. Brain-specific MCTs have been suggested to be
important in supplying lactate as an energetic substrate to support neuronal plasticity
and behavioral adaptations, and our finding that MCT4 knockdown impaired 2-
deoxyglucose uptake supports the possibility that astrocytic MCT4 contributes to
neuronal energy consumption and activity.

111
The observed deficit in overall rotarod performance following astrocytic MCT4
knockdown emerged with only eight trials (two trials/day for four days); this early period
of rotarod training has been shown to capture motor skill acquisition (Sathyamurthy et
al., 2020), an aspect of motor learning that is known to be controlled by the motor cortex
(Kawai et al., 2015). Using linear regression to calculate the slopes of the performance
curves (Rothwell et al., 2014), we were able to confirm the performance deficit was due
to impaired learning, confirming a specific deficit within motor cortex where astrocytic
MCT4 expression was diminished. This motor learning shortfall is novel, as previous
studies have predominately focused on the role of brain-specific MCTs in hippocampus-
dependent behavior and extends the possibility that brain-specific MCTs – and astrocytic
MCT4 – may contribute to a variety of normal behaviors mediated by distinct brain
regions. Importantly, we did not see any evidence of gross motor impairment following
astrocytic MCT4 knockdown using open field (to measure locomotion) and pole test (to
measure coordination). These results suggest that any deficits in performance and motor
learning in the accelerating rotarod paradigm are not linked to motor capacity but are
specific to an inability to effectively learn the motor task following astrocytic MCT4
knockdown. Future studies are necessary to understand if astrocytic MCT4 contributes
to other well-studied motor learning paradigms, such as the skilled reaching task (Xu et
al., 2009), and if it plays a similar role in motor-related regions that contribute to different
aspects of motor behavior, such as motor skill habituation mediated by the dorsal
striatum (Yin et al., 2009).

112
Using Golgi-Cox impregnation, we discovered that knockdown of astrocytic MCT4
in the motor cortex caused a diminishment of dendritic spine density and a decrease in
individual spine head width, and such synaptic structural changes were observed in a
short timeframe (within ~10 days of initial tamoxifen administration), prior to initiation of
any behavioral tasks. Recent results demonstrated that astrocytic MCT4 coordinates
learning-induced synaptic remodeling in the hippocampus in a L-lactate dependent
manner (Vezzoli et al., 2020). Beyond these learning-dependent structural changes, our
work extends the role of astrocytic MCT4 to include synaptic maintenance, suggesting
that astrocytic MCT4 is important in the homeostasis of existing dendritic spines. We
confirmed the specific loss of dendritic spines by examining PSD95 expression in motor
cortex protein lysates; additionally, the decrease in PSD95 was similarly matched by a
decrease in Arc protein expression, a critical mediator of dendritic spine remodeling
particularly in the motor cortex (Hosp et al., 2013). These findings mirror previously
reports of brain-specific MCTs knockdown negatively impacting the expression of PSD95
and Arc in the hippocampus (Suzuki et al., 2011). Notably, the loss of neuronal dendritic
spines was restricted to the primary motor cortex around the site of Lenti-pSico-EGFP-
shMCT4 activity (Fig. 4.7) and did not affect overall cortical thickness, suggestive that
loss of astrocytic MCT4 is not neurotoxic and confirming recent results of a lack of
cytotoxicity following astrocytic MCT4 knockdown in the hippocampus
(Netzahualcoyotzi & Pellerin, 2020).

Synaptic maintenance is an energetically demanding event (Attwell & Laughlin,
2001; Harris et al., 2012) and relative energy consumption can be estimated and
113
visualized using fluorescent probes, including near-infrared 2-deoxyglucose (2DG-IR) that
are retained and unable to be further metabolized by cells. Thorough validation of 2DG-IR
as a probe for investigating brain energetics has been previously completed, where it was
shown that 2DG-IR uptake in the brain is specific to neurons and occurs in an activity-
dependent manner (Lundgaard et al., 2015). We took advantage of the properties of this
specific 2DG analogue to investigate retention of the 2DG-IR tracer as a surrogate marker
of neuronal activity following astrocytic MCT4 knockdown in the motor cortex.
Intraperitoneal injection of 2DG-IR, followed by continuous engagement with the rotarod
motor task, allowed us to visualize the regional uptake of 2DG-IR reflective of task-
specific neuronal activity. Although our knockdown of astrocytic MCT4 was restricted to
the motor cortex, we found that this led to a clear diminishment of 2DG-IR uptake in motor
cortex as well as other motor-related structures during the rotarod task, including the
dorsal striatum and motor nuclei of the thalamus. It is important to acknowledge that
uptake of this specific 2DG-IR probe is likely not reflective of typical fast glucose
transport, as it has been suggested that 2DG-IR complexes and is internalized with the
glucose transporter, but still reflects energy consumption and neuronal activity (Kovar et
al., 2009; Lundgaard et al., 2015). This is particularly true as recent findings reported no
effect of MCT2 or MCT4 knockdown on glucose transporter expression
(Netzahualcoyotzi & Pellerin, 2020) that may affect 2DG-IR uptake during the rotarod task.

Taken together, our results demonstrate that astrocytic MCT4 contributes to
maintenance of synaptic structure and task-specific neuronal activity, both of which may
contribute to observed motor learning deficits on the accelerating rotarod. Without
114
electrophysiology, we cannot explicitly draw a line between our observed deficits in
synaptic structure, 2DG uptake and motor learning with a deficit in learning-induced long-
term potentiation, a hallmark of experience-dependent neuronal plasticity. Future
experiments, such as those utilizing in vivo electrophysiology and/or imaging of cortical
activity using neuronal calcium indicators during motor learning, would further detail the
role of astrocytic MCT4 as an important element of motor learning. Additionally, while the
motor cortex is necessary for the acquisition of a motor learning task (Kawai et al., 2015),
various other subcortical or cerebellar nuclei are known to play important and distinct
roles in various phases of the motor learning routine (Dayan & Cohen, 2011; Doyon et al.,
2009; Sathyamurthy et al., 2020; Yin et al., 2009). It will be important, therefore, to
investigate the relevance of astrocytic MCT4 to these other subcortical or cerebellar
structures. Considering that astrocytic MCT4 is known to shuttle lactate as part of the
astrocyte-neuron lactate shuttle, the deficits in synaptic structure, 2DG uptake and
behavioral adaptation may be the result of impaired energetic supply of lactate to motor
cortex neurons. The impact of brain energetics on healthy aging and neurodegenerative
diseases is becoming increasingly apparent in both animal models of disease and human
neuroimaging studies (Bak & Walls, 2018; Goyal et al., 2017; Polyzos et al., 2019). By
understanding the possibly relevant energy systems in circuits affected by disease –
such as the motor cortex-to-dorsal striatum corticostriatal circuit governing motor
learning and habit formation affected in Parkinson’s disease (Petzinger et al., 2013; Smith
& Graybiel, 2013) – therapeutic development can better target particular intra- and
intercellular processes to provide ample energy to maintain synaptic structure and
function. Thus, we suggest that astrocytic MCT4 is important for maintaining basal
115
synaptic structure and neuronal activity, and its loss leads to specific deficits in motor
learning, further highlighting the importance of astrocytic support of neuronal structure
and function.

116
Chapter 5: Loss of astrocytic MCT4 modulates dopamine in the
striatum to enhance behavioral performance
Adapted from Lundquist et al., 2021. In preparation.

Abstract
L-lactate is a metabolic byproduct of aerobic glycolysis in astrocytes and plays important
energetic and signaling roles to mediate neuroplasticity and behavioral adaptation.
Astrocytic lactate release, via MCT4, is tied to excitatory neurotransmission through the
astrocyte-neuron lactate shuttle (ANLS) to meet synaptic energy demands. Alternatively,
astrocytic lactate may signal through neuronal lactate receptors, resulting in distinct
changes in neuronal activity. Recently, we showed that astrocytic MCT4 knockdown in
the motor cortex diminished dendritic spines, ultimately leading to decreased motor
learning. To build on this, we sought to understand how astrocytic MCT4 is involved in
striatum-specific synaptic function and behavior. Here, astrocyte specific Cre-mediated
knockdown of MCT4 in the dorsal striatum caused a clear increase in object recognition,
stereotypy, and motor performance. Interestingly, MCT4 knockdown mice also exhibit
heightened amphetamine-induced rotational behavior, accompanied by increased
presynaptic dopamine, amphetamine-induced cFos activity, and DARPP-32
phosphorylation. These behavioral and synaptic responses do not appear to be
compensatory in nature, as cortico- and thalamostriatal glutamatergic projections and
striatal synaptic proteins were unchanged following astrocytic MCT4 knockdown. Taken
together, we propose that astrocytic MCT4 exports lactate to act as a negative feedback
117
mechanism on nigrostriatal dopaminergic activity, wherein its loss results in a
hyperdopaminergic state, leading to enhanced behavioral outputs. These findings
broaden the scope of astrocytic lactate function, acting not only as an energy source but
also as a synaptic modulator.

118
Introduction
Astrocytes support the formation, maintenance and modulation of neuronal
synapses through the release of a variety factors, including thrombospondins, serine, and
ATP (Allen & Lyons, 2018; Bowser, 2004; Hamilton & Attwell, 2010; Panatier et al., 2006),
and are capable of greatly influencing neuronal activity and plasticity (Adamsky et al.,
2018; Gourine et al., 2010; Kol et al., 2020). Beyond canonical astrocyte-to-neuron signals
regulating synaptic plasticity, astrocytes contribute to neuroplasticity through
intercellular metabolic shuttling including lipid droplets and L-lactate (Liu et al., 2017;
Magistretti & Allaman, 2018; Margineanu et al., 2018). Astrocytic shuttling of L-lactate, in
response to neuronal activity, is a well-established phenomenon in vitro and in vivo to
supply energy to synapses via monocarboxylate transporters (MCT2 on neurons and
MCT4 on astrocytes); this shuttling is known to affect synaptic structure and function,
manifesting in behavioral changes (Descalzi et al., 2019; Murphy-Royal et al., 2020; Suzuki
et al., 2011; Vezzoli et al., 2020). The astrocyte-to-neuron lactate shuttle (ANLS) has been
repeatedly demonstrated to play an important role in the hippocampus and related long-
term memory tasks. Recently, we extended the role of the ANLS to the cortex by blocking
astrocytic shuttling of lactate through genetic downregulation of MCT4, leading to a loss
of cortical dendrites and impairment in motor learning (see Chapter 4).

Beyond L-lactate’s ability to contribute energetically to neuronal synapses, L-
lactate is also capable of modifying brain vasculature and neuronal function through the
Gi/o-protein coupled receptor HCAR1, for which L-lactate is an agonist (Lauritzen et al.,
2014; Morland et al., 2017). Indeed, L-lactate acting at pre-synaptic HCAR1 in cortical
119
neurons was shown to dampen firing rate, suggesting that L-lactate release plays more
diverse roles at the synapse than purely being a metabolic fuel (de Castro Abrantes et al.,
2019). Interestingly, astrocytic L-lactate release has also been reported to excite neurons
in the locus coeruleus leading to norepinephrine release, acting through a yet identified
L-lactate receptor independent of metabolic consumption or HCAR1 signaling (Tang et
al., 2014).

Astrocytic function is known to vary drastically by region, (Chai et al., 2017; Martín
et al., 2015; Morel et al., 2017) and this likely contributes to the diversity of roles that L-
lactate plays throughout the brain. Thus, we sought to explore how astrocytic L-lactate
shuttling affected synaptic structure, function and behavior in the dorsal striatum, a highly
innervated nucleus whose output governs a variety of motor and cognitive behaviors
(Graybiel, 2008; Graybiel & Grafton, 2015). We utilize unilateral and bilateral lentiviral
knockdown of astrocytic MCT4 to impair astrocyte-to-neuron lactate shuttling in the
context of striatum-mediated behaviors. This approach reveals an unexpected
enrichment of presynaptic dopamine following MCT4 knockdown, as shown through
heightened motor performance, amplified sensitivity to amphetamine and an increase in
presynaptic TH and VMAT2. The MCT4 knockdown-mediated increase in dopaminergic
activity appears to signal primarily through the dopamine D1-receptor, as MCT4
knockdown mice showed exaggerated stereotypical behavior and increased DARPP-32
phosphorylation, cFos expression, and NMDA receptor expression in the striatum. These
findings suggest a novel role for astrocytic lactate as a regulator of dopamine signaling
120
in the striatum and provide further evidence that lactate can potently impact the
production and release of neuromodulators in the brain.

121
Methods
Mice
Heterozygous Aldh1l1-CreERT2
+/-
(Cre
+
) (Srinivasan et al., 2016) and Aldh1l1-
CreERT2
-/-
(Cre
-
) littermates were generated from in-house breeding pairs of Aldh1l1-
CreERT2
+/-
and Aldh1l1-CreERT2
-/-
mice at the University of Southern California. All
animals were group housed, up to five animals per cage, and had ad libitum food and
water access on an inverse dark-light cycle (lights off at 0700, lights on at 1900). CreERT2
expression was assessed by traditional PCR with primers against CreERT2 and the
wildtype insert according to previously described methods (Srinivasan et al., 2016). All
experimental procedures in animals were approved by the Institutional Animal Care and
Use Committee at the University of Southern California (Protocol No. 21044) and carried
out in compliance with the National Institutes of Health guidelines for the care and use
of laboratory animals.

Plasmids and Vector Construction
pCMV-VSV-G was a gift from Bob Weinberg (Addgene plasmid #8454;
http://n2t.net/addgene:8454; RRID: Addgene_8454). pCMV delta R8.2 was a gift from
Didier Trono (Addgene plasmid # 12263; http://n2t.net/addgene:12263; RRID:
Addgene_12263). The pSico plasmid was a gift from Tyler Jacks (Addgene plasmid
#11578; http://n2t.net/addgene:11578; RRID: Addgene_11578). Cloning of short hairpin
RNA (shRNA) segments specific to mouse MCT4 (Slc16a3) into pSico was previously
described (see Chapter 4). For pSico Slc16a3shRNA (pSico-MCT4 shRNA) lentiviral
vector, pSico-Slc16a3shRNA, pCMV delta R8.2 and pCMV-VSV-G plasmids were
122
transfected into HEK-293T cells, collected, and concentrated as previously described
(see Chapter 4) (Llewellyn et al., 2019).
Stereotactic surgery
Mice were anesthetized with 3% isoflurane and maintained at 1.5% isoflurane
throughout the surgery. Cre
+
and Cre
-
littermates were placed in a stereotactic frame
(Stoelting) and their skulls were exposed with a single scalpel cut. Small holes were
drilled in the skull over the striatum using stereotactic coordinates and MCT4 shRNA
lentivirus (1.0ul, 1x10
12
genome copies/mL) was injected using a 26-gauge syringe
(701N, Hamilton, Reno, NV) at a rate of 0.15ul/min. Lentivirus was targeted to the dorsal
striatum at a rostral level (anterior-posterior [AP]: +1.3 mm; medial-lateral [ML]: ±1.4 mm;
dorsal-ventral [DV]: -3.0 mm from surface of the skull) and at a caudal level (AP: +0.0 mm;
ML: ±2.0 mm; DV: -3.1 mm from surface of the skull). In one group, mice were injected
bilaterally, receiving a total of four lentiviral injections (two rostral, two caudal); in a
second group, mice were injected unilaterally in the right hemisphere, receiving a total of
two lentiviral injections (one rostral, one caudal). The incision was closed with a single
stainless-steel surgical staple (Stoelting) and mice were allowed to recover on heating
pads with Medigel (containing 1mg carprofen, Clear H2O) refreshed daily for 3 days. Mice
were not used for further experimentation until ≥ 2 weeks after stereotactic surgeries to
allow for adequate viral expression and recovery from surgery.

Pharmacological treatments
Daily intraperitoneal injections of tamoxifen (Sigma; 20mg/ml, dissolved in corn
oil; 75 mg tamoxifen per kg body weight) were given for five days to induce MCT4 shRNA
123
expression in Cre
+
(MCT4 shRNA) mice. Cre
-
(control) mice also received daily injections
of tamoxifen to ensure there were no effects of drug administration alone on behavioral
or molecular outcomes. Apomorphine (Sigma, Cat. #A4393) was dissolved 5mg/ml in
0.9% saline with 0.1% ascorbic acid and administered subcutaneous at 2mg/kg.
Amphetamine (Sigma, Cat. #A5880) was dissolved 1mg/ml in 0.9% saline and
administered intraperitoneally at 5mg/kg.

Behavioral Testing
The effect of MCT4 knockdown in the dorsal striatum was assessed using motor
and cognitive behavioral tests, as detailed below. All testing and analysis were performed
during the first half of the dark cycle (0900-1200). Mice were brought to the behavioral
suite and allowed to acclimate for at least 30 minutes before any behavioral paradigms
were initiated. All behavioral arenas and apparatuses were thoroughly cleaned with 70%
ethanol and allowed to dry between uses.

Open Field Test was used to assess spontaneous locomotive behavior. Open field testing
was conducted once, prior to any other behavioral tests. Mice were placed in an open
field chamber (30cm x 30cm x 30cm white plywood box) for ten minutes and total
movement was recorded by a digital video camera at 30 frames per second. Locomotion
was analyzed using the EzTrack pipeline (Pennington et al., 2019), and distance was
binned in one-minute intervals.
Pole Test was used to as a test of motor speed and dexterity. Behavioral testing with the
pole test was conducted once, after the open field test. Briefly, mice were placed at the
124
top of a 50cm wooden pole in an empty cage filled with normal bedding, and their time to
descend to the base of the pole is recorded. The trial was repeated five times, with a one-
minute inter-trial interval, before the mice were returned to their home cage.
Self-grooming behavior, a form of stereotypy (Kalueff et al., 2016), was assessed once.
Mice were placed in a clear, 1-liter glass beaker out of sight of other mice and videotaped
for 10 min before being returned to their respective home cages. The duration of a
grooming bout was counted if the grooming sequence initiated lasted at least 0.5
seconds, and its total duration was recorded so long as the chain of grooming sequences
was not broken (Berridge et al., 1987; Kalueff et al., 2016).
Novel object recognition (NOR) was used to assess working memory. The NOR protocol
spanned three days, including habituation (day one), training (day two), and testing (day
three) (Leger et al., 2013). Briefly, mice were habituated to the testing arena (30cm x
30cm x 30cm white plywood box) for 10 minutes on the first day. 24 hours later mice
were returned to the same arena containing two identical objects (either inverted 100mL
glass beaker or blue Lego structure) and allowed to explore the arena and objects for 10
minutes. 24 hours later mice were returned to the same arena, now containing a novel
object in place of a familiar object (Lego in place of beaker, and vice versa), and allowed
to explore the arena and objects for 10 minutes. Mice were videotaped on the training
and testing days and exploratory behavior of the objects were scored manually using
previously described methods (Leger et al., 2013). Briefly, mouse exploratory behavior
was recorded if the mouse sniffed or touched the object while looking at it but was not
counted if the animal climbed on top of the object. A discrimination index score for each
mouse during the test was calculated using exploratory durations for each object (Tnovel
125
– Tfamiliar / Ttotal); an index score of 0 represents equal time spent exploring familiar and
novel objects, whereas an index score of 1 represents all time spent exploring the novel
object. Familiar and novel object sequence and position in the testing arena were
counterbalanced between animals. Additionally, a minimal time of exploration in the
training phase (20 seconds over 10 minutes) was used in analyzing NOR behavior to
minimize inter-subject variability and more accurately measure object discrimination;
across both groups, only one mouse did not reach this criterion.
Accelerating rotarod was used to assess motor learning and retention. Following
baseline motor testing, mice were tested for motor learning and coordination was
previously described (Rothwell et al., 2014). Briefly, mice were oriented to the stationary
rod for three minutes before their first trial. Mice then completed three trials per day for
four days, with a five-minute intertrial interval during which mice were returned to their
home cage. For the first two days, the rotarod (3cm diameter rod, divided into five lanes;
Ugo Basile, Comerio, Italy) accelerated over the course of 300 seconds from 4 to 40 rpm,
and for the last two days, the rotarod accelerated over the course of 300 seconds from 8
to 80 rpm; speed at time of fall and latency to fall were automatically recorded by
magnetic trip plates. Five days after the last trial, motor retention and recall were tested
across three trials, with a five-minute intertrial interval, using the 4 to 40 rpm acceleration
paradigm. A trial ended when the mouse made a complete backward revolution, fell off,
or reached the 300 second threshold. Learning curves were fit with linear regression
analysis, where the slope is defined as the learning rate and the Y-intercept of the fitted
regression line is defined as the initial coordination (Lundquist et al., 2021; Rothwell et al.,
2014).
126
Rotational behavior was assessed in the unilateral knockdown of astrocytic MCT4 as a
measure of spontaneous and drug-induced turning, a measure of possible nigrostriatal
dopamine imbalance (Björklund & Dunnett, 2019; Robinson & Becker, 1983). Spontaneous
rotational behavior was assessed every four days for 10 minutes by placing mice in a 1-
liter glass beaker, out of sight of other mice. For apomorphine or amphetamine-induced
rotations, mice were administered drug solutions subcutaneously or intraperitoneally,
respectively, then placed in a 30cm x 30cm x 30cm white plywood arena for 40 minutes.
All rotational behavior was videotaped and manually scored. A complete rotation was
defined as four consecutive 90º full body turns in the same direction, and the
directionality of turning (ipsilateral or contralateral) was in relation to the side of the
lentiviral knockdown of MCT4 (right hemisphere of the brain).
Immunohistochemistry
Mice were heavily anesthetized with Avertin (tribromoethanol, 250mg/kg) before
being transcardially perfused with ice cold 0.9% saline, followed by 4% PFA. Whole brains
were postfixed in 4% PFA at 4ºC overnight before being transferred to 30% sucrose at
4ºC for two days. Brains were frozen in dry ice-cooled isopentane and sliced on a freezing
cryostat in the coronal orientation at 50µm before being stored in cryoprotectant solution
at -20ºC. Sections were processed for immunohistochemistry as previously described
(Lundquist et al., 2019). Briefly, sections were washed and permeabilized in TBS with
0.05% Tween-20 (TBST), nonspecific staining was blocked with blocking solution (TBST
and 4% normal donkey or normal goat serum) and incubated overnight at 4ºC with SOX9
antibody (1:2000, Millipore, AB5535, RRID: AB_2239761), TH antibody (1:2000, Millipore,
MAB318, RRID: AB_2201528) or cFos antibody (1:1000, Cell Signaling Technology, Cat.
127
#2250, RRID: AB_2247211). Sections were washed and incubated with Alexa 568-
conjugated secondary antibody (1:5000, Thermo Fisher Scientific, A11011, RRID:
AB_143157) before being coverslipped (Vectashield Hardset Antifade with DAPI; Vector
Laboratories). All confocal images were taken on an IXB-DSU spinning disk Olympus BX-
61 (Olympus America) and captured with an ORCA-R2 digital CCD camera (Hamamatsu)
and MetaMorph Advanced software (Molecular Devices).
cFos Quantification
cFos
+
cells were quantified in two areas, guided by previous literature on
amphetamine-induced cFos expression in rodents (Badiani et al., 1998): dorsal striatum
and cortex dorsal to the striatum, including primary motor and somatosensory cortex
areas. A grid of six 500 µm x 500 µm squares was overlaid on images using FIJI
(Schindelin et al., 2012) and a random number generator (1 through 6) was used to
randomly select three squares to sample from per section. Three to four sections were
analyzed per mouse in each area.
Western Blotting
Mice were rapidly euthanized by cervical dislocation and whole brains were
removed before the dorsal striatum was rapidly dissected, snap frozen on dry ice and
stored at -80
o
C. Cortical tissue samples were lysed in RIPA buffer with protease and
phosphatase inhibitor (MSSAFE, Sigma), centrifuged at 16,000 x g and the soluble
fraction was collected for subsequent protein analysis. Total protein content was
determined by BCA analysis and 20ug of protein was resolved on a 10% Tris-Glycine gel
by electrophoresis (BioRad). Total protein was transferred to nitrocellulose membranes
(BioRad), blocked (OneBlock, Genesee Scientific, Cat. #20-314) and probed with the
128
following antibodies overnight at 4ºC: rabbit anti-MCT1 (1:1000, Thermo Fisher Scientific,
Cat# PA5-72957, RRID: AB_2718811), rabbit anti-MCT4 (1:1000, Novus Biologics, Cat#
NBP1-81251, RRID: AB_11033184), mouse anti-synaptophysin (1:2000, Abcam, Cat#
ab8049, RRID: AB_2198854), mouse anti-PSD95 (1:2000, Millipore, Cat# MAB1596, RRID:
AB_2092365), rabbit anti-VGLUT1 (1:250, Thermo Fisher Scientific, Cat# 48-2400, RRID:
AB_2533843), rabbit anti-VGLUT2 (1:250, Thermo Fisher Scientific, Cat# 42-7800, RRID:
AB_2533537), rabbit anti-GAP43 (1:3000, Thermo Fisher Scientific, Cat# PA5-34943,
RRID: AB_2552292), rabbit anti-GAD65 (1:1000, Millipore, Cat# ABN101), goat anti-ChAT
(1:1000, Millipore, Cat# AB144P, RRID: AB_2079751), rabbit anti-DARPP-32 (1:500, Santa
Cruz Biotechnology, Cat# sc-11365, RRID: AB_639000), mouse anti-TH (1:1000, Millipore,
Cat# MAB318, RRID: AB_2201528), rabbit anti-VMAT2 (1:1000, Thermo Fisher Scientific,
Cat# PA5-102661, RRID: AB_2852058), rabbit anti-phospho-DARPP-32 (Thr34) (1:1000,
Cell Signaling Technology, Cat# 12438, RRID: AB_2797914), rat anti-DAT (1:1000,
Millipore, Cat# MAB369, RRID: AB_2190413), rabbit anti-dopamine D1 receptor (1:500,
Millipore, Cat# ABN20, RRID: AB_10561918), rabbit anti-dopamine D2 receptor (1:500,
Millipore, Cat# AB5084P, RRID: AB_2094980), rabbit anti-NMDAR1 (1:300, Millipore, Cat#
AB1516, RRID: AB_2314955) and mouse anti-beta actin (1:5000, LI-COR, Cat# 926-42212,
RRID: AB_2756372). Membranes were washed, incubated with corresponding donkey
anti-goat, goat anti-rat, goat anti-mouse or goat anti-rabbit conjugated near-infrared
secondary fluorescent antibodies (1:5000, LI-COR, Cat# 926-32214, RRID: AB_621846;
Cat# 926-68076, RRID: AB_10956590; Cat# 926-32211, RRID: AB_621843; and Cat# 926-
68070, RRID: AB_10956588), and then scanned on an Odyssey imaging system (LI-COR).
129
Relative protein expression was quantified by optical density and normalized to beta-actin
loading controls using FIJI.

Golgi-Cox Staining
Mice were rapidly euthanized by cervical dislocation and decapitation and whole
brains were extracted. Brains were used for Golgi-Cox staining (FD NeuroTechnologies,
Cat# PK401) following the manufacturer’s protocol as previously described (see Chapter
4) (Toy et al., 2014). Briefly, brains were impregnated for two weeks, slowly frozen with
dry ice-chilled isopentane, sectioned in the coronal plane in 100µm sections, and
processed according to the manufacturer’s protocol. Medium spiny neurons in the dorsal
striatum were identified and 20 cells were imaged per mouse using an Olympus BX50
(Olympus America) and captured with a KAPELLA digital CCD camera (Jenoptik). Cells
were imaged across at least three sections per mouse, and multiple spans of dendrites
were counted and analyzed per cell using FIJI. Total dendritic counts were averaged
across all cells imaged per mouse.

Statistical Analysis
Sample sizes were calculated based on our previously published work and all
efforts were made to minimize the number of mice used. We used five mice per group (n
= 5) for bilateral dorsal striatum MCT4 knockdown experiments, including NOR,
stereotypy, rotarod, open field, pole test, and western blotting. We used six mice per group
(n = 6) for unilateral dorsal striatum MCT4 knockdown experiments, including open field
and rotational behaviors; these six mice were split evenly for histology (n = 3 mice per
130
group) and dendritic spine analysis (n = 3 mice per group). All data was tested for
normality using D’Agostino & Pearson test. If both groups (control and MCT4 shRNA)
passed normality tests, an unpaired t-test was used. If not, a Kolmogorov-Smirnov test
was used instead. Simple linear regression was used to assess TH and VMAT2
expression correlation. To compare overall performance on accelerating rotarod, a 2-way
ANOVA with Sidak’s multiple comparisons was used. Spontaneous rotational behavior
data for each group was analyzed with a 2-way repeated measures ANOVA with Sidak’s
multiple comparisons, then fit with a least squares regression; comparisons between a
simple linear or second-order polynomial was made using Akaike’s Information Criteria.
All statistical analyses were conducted using Prism (version 9.1, GraphPad) with
significance denoted as p < 0.05.

131
Results
Astrocytic MCT4 knockdown improves rotarod performance
The Cre-inducible pSico expression plasmid was used to carry a short hairpin RNA
(shRNA) specific to mouse MCT4 (Slc16a3), packaged into a lentivirus and delivered to
the dorsal striatum of astrocyte-specific Aldh1l1-CreERT2
+
(Cre
+
) or Aldh1l1-CreERT2
-

(Cre
-
) mice (Fig. 5.1a). Lentiviral expression was identified in striatal SOX9
+
astrocytes
(Fig. 5.1b) and resulted in specific loss of MCT4 protein expression (unpaired two-tailed
t-test, t=2.848, df=8, p = 0.022) without any change in MCT1 expression (unpaired two-
tailed t-test, t=0.907, df=8, p = 0.391) (Fig. 5.1c).

132
Figure 5.1 – Specific knockdown of the lactate transporter MCT4 in striatal astrocytes. a.
Experimental approach for assessing astrocytic MCT4 (left) and representative micrograph
showing injection site for lentivirus (LV) in dorsal striatum. Scale bar = 1mm. b. Representative
micrograph of SOX9
+
astrocytes transfected with LV. Scale bar = 200μm. c. MCT4 knockdown
specifically affects MCT4 protein expression, but not related MCT1 expression in the dorsal
striatum. Data are mean ± s.e.m. ns, p > 0.05; * p < 0.05 (n = 5 mice per group, unpaired two-sided
t-test). Bands shown are select representative lanes.

Next, we assessed whether MCT4 knockdown in the dorsal striatum affected basic
locomotion and coordination using the open field and pole test, respectively. MCT4
133
knockdown did not impact locomotion (unpaired two-tailed t-test, t=0.344, df=8, p =
0.740) or coordination (unpaired two-tailed t-test, t=1.208, df=8, p = 0.262) (Fig. 5.2a, b).
Our previous work in motor cortex revealed a deficit in motor performance on the
accelerating rotarod (see Chapter 4), and the dorsal striatum contributes to acquisition
of the rotarod task (Yin et al., 2009). Thus, we tested whether MCT4 knockdown in the
dorsal striatum negatively affected rotarod performance. Interestingly, knockdown mice
outperformed control mice in early trials, with most knockdown mice reaching the
threshold of 300 seconds (Fig. 5.2c). In later trials, increasing the speed of rotation to
better delineate aspects of motor performance did not reveal task-specific repetitive
behavior as others have shown in mouse models of autism (Rothwell et al., 2014). Overall,
motor performance was significantly higher in MCT4 mice compared to control mice
(unpaired two-tailed t test, t=2.825, df=8, p = 0.022) (Fig. 5.2d).
Striatal MCT4 knockdown increases stereotypical and working memory behaviors
To better understand how loss of astrocytic lactate shuttling may manifest in
behaviors mediated by the dorsal striatum, we observed mouse self-grooming. Self-
grooming is a measure of stereotypical behavior governed in part by dopamine acting on
D1-receptor specific striatal spiny projection neurons, as part of the direct pathway of the
basal ganglia (Frederick et al., 2015; Kalueff et al., 2016). MCT4 knockdown mice did not
carry out more grooming bouts (unpaired two-tailed t-test, t=0.293, df=8, p = 0.777), but
grooming chains were continued for significantly longer than controls (unpaired two-
tailed t-test, t=2.824, df=8, p = 0.022) (Fig. 5.3a). Following this, we next used the novel
object recognition task, a measure of working memory, to investigate a cognitive task
mediated by dopamine innervation of the dorsal striatum (Darvas & Palmiter, 2009).
134
Figure 5.2 – Astrocytic MCT4 knockdown improves motor performance. a. MCT4 knockdown in
the dorsal striatum does not affect coordination on the pole test. b. MCT4 knockdown does not
affect locomotion in the open field. Data are mean ± s.e.m or boxplots showing quantiles (25, 50,
75%) with central line marking the median and plus denoting the mean. ns, p > 0.05 (n = 5 mice
per group, unpaired two-sided t-test) (a, b). c. MCT4 knockdown enhances early learning and
retention in the accelerating rotarod. Grey shading indicates the faster, 8-80rpm acceleration
paradigm. d. MCT4 knockdown increases overall performance (area under the curve). Data are
mean ± s.e.m., * p < 0.05 (n = 5 mice per group, unpaired two-sided t-test) (c, d).

MCT4 knockdown revealed a significant improvement in the discrimination of a novel
object relative to controls (unpaired two-tailed t-test, t=3.363, df=8, p = 0.001) (Fig. 5.3b).
These results, along with enhanced motor performance on the accelerating rotarod,
suggest that astrocytic lactate loss enhances performance on tasks that involve
dopamine signaling in the dorsal striatum.

135

Figure 5.3 – Stereotypy and object recognition improve following astrocytic MCT4 knockdown.
a. Self-grooming behavior was assessed over a 10-minute window by measuring total grooming
sessions and duration of grooming (see Methods). b. Novel object recognition paradigm (left)
revealed that object discrimination was improved in MCT4 knockdown mice (right). Data are
boxplots showing quantiles (25, 50, 75%) with central line marking the median and plus denoting
the mean. ns, p > 0.05; *, p < 0.05; **, p < 0.01 (n = 5 mice per group, unpaired two-sided t-test).

Loss of astrocytic MCT4 does not affect expression of representative striatal proteins
Maintenance of dopamine-mediated behaviors has been observed using
developmental interference of the nigrostriatal dopamine system (Golden et al., 2013),
and even mild neurotoxic lesions to the dopaminergic system results in presynaptic
dopamine compensation (Perez et al., 2008). Thus, we used western blotting to first
assess the expression of diverse striatal synaptic proteins, including spiny projection
neurons, interneurons, and excitatory inputs from cortex and thalamus in order to
136
determine if MCT4 knockdown in the dorsal striatum resulted in striatal-specific damage
that may account for behavioral differences. No significant difference was found in the
expression of the ubiquitous presynaptic markers synaptophysin (unpaired two-tailed t-
test, t=1.047, df=8, p = 0.326) or GAP43 (unpaired two-tailed t-test, t=1.475, df=8, p =
0.178); additionally, there was no change in the expression of the postsynaptic protein
PSD95 (unpaired two-tailed t-test, t=.661, df=8, p = 0.527) (Fig, 5.4a, b, e). Expression of
presynaptic glutamatergic markers VGLUT1 (corticostriatal terminals) and VGLUT2
(thalamostriatal terminals) were also unchanged with MCT4 knockdown (VGLUT1:
unpaired two-tailed t-test, t=0.192, df=8, p = .852; VGLUT2: unpaired two-tailed t-test,
t=0.600, df=8, p = 0.565) (Fig. 5.4c, d). The striatal neuronal population is composed
principally of spiny projection neurons (expressing DARPP-32) and cholinergic (ChAT
+
)
and GABAergic (GAD65
+
) interneurons (Kreitzer, 2009). ChAT and GAD65 expression
were not affected following MCT4 knockdown (ChAT: unpaired two-tailed t-test, t=1.351,
df=8, p = 0.213; GAD65: unpaired two-tailed t-test, t=0.312, df=8, p = 0.763) (Fig. 5.4f, g).
Similarly, DARPP-32 expression was unchanged as well (unpaired two-tailed t-test,
t=0.978, df=8, p = 0.356) (Fig. 5.4h). In totality, MCT4 knockdown does not appear to
affect the neuronal composition of the dorsal striatum that may explain behavioral
changes.

137

Figure 5.4 – Astrocytic MCT4 knockdown does not affect expression of striatal synaptic
proteins. MCT4 knockdown does not affect the expression of presynaptic synaptophysin (a),
postsynaptic PSD95 (b), presynaptic corticostriatal VGLUT1 (c), presynaptic thalamostriatal
VGLUT2 (d), presynaptic GAP43 (e), interneuron marker GAD65 (f), interneuron marker ChAT (g),
or spiny projection neuron marker DARPP32 (h). All data are mean ± s.e.m. ns, p > 0.05 (n = 5
mice per group, unpaired two-sided t-test). Bands shown are select representative bands.

Presynaptic striatal dopamine increased following MCT4 knockdown
Dopaminergic inputs to dorsal striatum, largely originating from the substantia
nigra (Palmiter, 2008), express tyrosine hydroxylase, the rate-limiting enzyme in
dopamine synthesis. Tyrosine hydroxylase expression was increased in dorsal striatum
following MCT4 knockdown (unpaired two-tailed t-test, t=2.952, df=8, p = 0.018) (Fig.
138
5.5a). Similarly, the vesicular monoamine transporter VMAT2 – responsible for vesicular
uptake of dopamine – was increased following MCT4 knockdown (unpaired two-tailed t-
test, t=3.104, df=8, p = 0.015) (Fig. 5.5b). Together, this expression pattern suggests
MCT4 knockdown mediated increase in presynaptic dopamine synthesis and packaging
in the nigrostriatal dopamine terminals; this is further corroborated by a strong correlation
in TH and VMAT2 expression (linear regression, R
2
=0.644, p = 0.005) (Fig. 5.5c).
Figure 5.5 – MCT4 knockdown enhances presynaptic dopamine markers in the dorsal striatum.
a. MCT4 knockdown increases expression of tyrosine hydroxylase (TH). b. MCT4 knockdown
increases vesicular monoamine transporter 2 (VMAT2) expression. c. TH and VMAT2 expression
correlate in the dorsal striatum (simple linear regression with 95% confidence interval). d. MCT4
knockdown does not affect dopamine transporter (DAT) expression. Data are mean ± s.e.m. (a,
b, d), ns, p > 0.05; *, p < 0.05 (n = 5 mice per group, unpaired two-sided t-test). Bands shown are
select representative bands.

139
Dopamine transporter, responsible for clearing dopamine to modulate synaptic
occupancy and presynaptic dopamine synthesis (Giros et al., 1996), is unaffected by
MCT4 knockdown (unpaired two-tailed t-test, t=0.566, df=8, p = 0.587) (Fig. 5.5d).
Figure 5.6 – Astrocytic MCT4 knockdown affects dopamine receptor expression and canonical
signaling products in the dorsal striatum. a. MCT4 knockdown does not affect expression of
dopamine D1 receptors (D1R) in the dorsal striatum. b. MCT4 knockdown decreases dopamine
D2 receptor (D2R) expression in the dorsal striatum. c. Striatal D1R-to-D2R ration is increased
following MCT4 knockdown. d. MCT4 knockdown increases phosphorylation of DARPP-32 at
Thr34. e. Striatal NMDA receptor subunit NR1 expression is increased following MCT4
knockdown. Data are mean ± s.e.m. ns, p > 0.05; *, p < 0.05 (n = 5 mice per group, unpaired two-
sided t-test). Bands shown are select representative bands.

140
Synaptic dopamine receptor ratio changes following MCT4 knockdown to promote striatal
plasticity
Synaptic dopamine receptors on striatal spiny projection neurons gate action
selection, reinforcement, and striatal plasticity that underlies behavioral adaptation
(Anzalone et al., 2012; Kravitz et al., 2010). In dorsal striatum, no effect of MCT4
knockdown was seen on dopamine D1 receptor expression (unpaired two-tailed t-test,
t=0.659, df=8, p = 0.528) (Fig. 5.6a), which is principally involved in promoting movement
and learning in striatum (Hallett et al., 2006; Kravitz et al., 2010). Conversely, dopamine
D2 receptor expression was decreased in the dorsal striatum (unpaired two-tailed t-test,
t=3.162, df=8, p = 0.013) (Fig. 5.6b), which serves to inhibit movement initiation. This
imbalance of D1-D2 leads to an increase in dopamine receptor ratios in the dorsal
striatum (unpaired two-tailed t-test, t=2.309, df=8, p = 0.049) (Fig. 5.6c). Dopamine acts
through D1- or D2-like dopamine receptors to differentially affect intracellular signaling
pathways in the striatum. Following dopamine D1 receptor activation, the striatal protein
DARPP-32 is phosphorylated at the threonine 34 residue, regulating downstream gene
expression (Bateup et al., 2008; Gerfen et al., 1990). In the dorsal striatum, DARPP-32
phosphorylation at Thr34 was significantly increased with MCT4 knockdown (unpaired
two-tailed t-test, t=2.856, df=8, p = 0.021) (Fig. 5.6d). Thus, astrocytic MCT4 knockdown
appears to affect presynaptic dopamine production and release, which in turn increases
D1R-specific activation, without impacting dopamine recycling via DAT. The likely
enhanced activity on D1 dopamine receptors – previously demonstrated by increased
DARPP-32 phosphorylation – is also seen in an increase in NMDA receptor subunit NR1
141
expression (unpaired two-tailed t-test, t=3.011, df=8, p = 0.017) (Fig. 5.6e), known to be
tied to dopamine D1 receptor activity (Hallett et al., 2006).

Figure 5.7 – Unilateral knockdown of astrocytic MCT4 increases tyrosine hydroxylase (TH)
immunoreactivity in dorsal striatum. Representative micrograph of TH staining in a unilateral
MCT4 knockdown, with colored syringe denoting injection site (left) and quantification of
immunoreactivity (right). Data are boxplots showing quantiles (25, 50, 75%) with central line
marking the median and plus denoting the mean. ***, p < 0.001 (n = 3 mice per group, Mann-
Whitney U test).

Unilateral MCT4 knockdown model reveals increased sensitivity to amphetamine
To further delineate the impact of astrocyte lactate shuttling on striatal dopamine,
MCT4 knockdown was localized unilaterally to the right hemisphere and these mice were
used to assess rotational behavior, a behavior tied to interhemispheric dopamine
imbalance (Björklund & Dunnett, 2019) (Fig. 5.7). As before in our bilateral knockdown of
MCT4, no gross motor deficits were detected using the open field in the unilateral model
(Fig. 5.8b). Spontaneous rotations were assessed every four days for 10 minutes. While
MCT4 knockdown did not have a significant impact on the directionality of rotations (2-
way repeated measures ANOVA, shRNA effect: F(1,10) = 4.546, p = 0.059), mouse rotational
behavior was rather dynamic, reflected by a time-by-shRNA effect on spontaneous
142
rotational behavior (2-way repeated measures ANOVA, session x shRNA effect: F(5,50) =
4.436, p = 0.002) (Fig. 5.8c). Mice had similar total rotations across most sessions (2-
way repeated measures ANOVA with Sidak’s multiple comparisons test, session 1: p =
0.035; all other session, p > 0.05) (Fig. 5.8d)
We next used two different pharmacological approaches to demarcate whether
MCT4 knockdown had a larger effect on presynaptic dopamine release or postsynaptic
dopamine receptor sensitivity. Apomorphine was administered to assess dopamine
receptor sensitivity; overall, there was no difference in total rotations between MCT4
knockdown and control groups (left, total rotation time course, 2-way repeated measures
ANOVA, shRNA effect, F(1,10) = 0.107, p = 0.750; right, total summed rotations, unpaired
two-tailed t-test, t=0.848, p = 0.416) (Fig. 5.8e). Next, amphetamine was administered to
evaluate presynaptic dopamine release; amphetamine significantly increased total
rotations in the MCT4 knockdown group (left, total rotation time course, 2-way repeated
measures ANOVA, shRNA effect, F(1,10) = 5.466, p = 0.042; right, total summed rotations,
unpaired two-tailed t-test, t=2.331, df=10, p = 0.042), suggestive of higher presynaptic
dopamine release (Fig. 5.8f).

143

Figure 5.8 – Unilateral knockdown of striatal astrocytic MCT4 drives spontaneous rotational
behavior and heightens sensitivity to amphetamine. a. Experimental approach and timeline for
rotational behavior in unilateral MCT4 knockdown in dorsal striatum. b. Unilateral MCT4
knockdown does not affect overall locomotion in the open field. Data are mean ± s.e.m. (left) and
boxplot showing quantiles (25, 50, 75%) with central line marking the median and plus denoting
the mean. ns, p > 0.05 (n = 6 mice per group, unpaired two-sided t-test). c. Unilateral MCT4
knockdown in dorsal striatum causes dynamic change in spontaneous rotations. Data are mean
± s.e.m. 2-way repeated measures ANOVA (n = 6 mice per group). Curves are best-fit regressions
for each group and corresponding 95% confidence intervals, assessed using Akaike’s Information
Criteria (see Methods). d. Total spontaneous rotations per group across all sessions. Data are
mean ± s.e.m. ns, p > 0.05; *, p < 0.05 (n = 6 mice per group, unpaired two-sided t-test). e.
Apomorphine administration does not increase rotational behavior. Data are mean ± s.e.m. (left)
and boxplot showing quantiles (25, 50, 75%) with central line marking the median and plus
denoting the mean (right). ns, p > 0.05, 2-way repeated measures ANOVA (left); ns, p > 0.05,
unpaired two-sided t-test (right) (n = 6 mice per group). f. Amphetamine administration increases
rotational behavior. Data are mean ± s.e.m. (left) and boxplot showing quantiles (25, 50, 75%) with
central line marking the median and plus denoting the mean (right). *, p < 0.05, 2-way repeated
measures ANOVA, shRNA effect (left); *, p < 0.05, unpaired two-sided t-test (right) (n = 6 mice per
group).
144
MCT4 knockdown increases amphetamine-induced cFos expression
Amphetamine administration induces cortical and striatal cFos expression
(Badiani et al., 1998) due to heightened dopaminergic tone (Berretta et al., 1992).
Following amphetamine administration (5mg/kg intraperitoneally), cFos
+
cell counts in
the dorsal striatum were higher in MCT4 knockdown mice (Fig. 5.9a) relative to control
mice (unpaired two-tailed t-test, t=7.288, df=26, p < 0.001). A similar result was found in
the cortex dorsal to the sampled striatal area (unpaired two-tailed t-test, t=7.136, df=24,
p < 0.001) (Fig. 5.9b); this cortical area, containing primary motor and primary
somatosensory cortices, was previously shown using in situ hybridization studies to be a
focal site of amphetamine-induced c-fos mRNA expression (Badiani et al., 1998).

145
Figure 5.9 – Astrocytic MCT4 knockdown increases amphetamine-induced cFos expression. a.
Striatal cFos
+
cell density increases in MCT4 knockdown mice following amphetamine
administration. c. Cortical cFos
+
cell density increases in MCT4 knockdown mice following
amphetamine administration. Data are boxplot showing quantiles (25, 50, 75%) with central line
marking the median and plus denoting the mean (right). ***, p < 0.001 (n = 3 mice per group,
unpaired two-sided t-test).

146
MCT4 knockdown does not affect striatal dendritic spine density
Changes in dopaminergic and glutamatergic activity in the striatum is well-known
to affect dendritic spine density of striatal spiny projection neurons (Toy et al., 2014;
Yagishita et al., 2014). However, MCT4 knockdown did not affect spine density
(Kolmogorov-Smirnov test, D=0.147, p = 0.100, Fig. 5.10) in the dorsal striatum relative to
control mice, further corroborating earlier work in bilateral striatal MCT4 knockdown mice
showing no effect of MCT4 knockdown on synaptic proteins.
Figure 5.10 – Astrocytic MCT4 knockdown does not impact dendritic spine density in the
dorsal striatum. a. Representative Golgi-impregnated spiny projection neuron with example
dendritic branches from control and MCT4 shRNA groups. Scale bar for branches is 5μm. b.
Striatal dendritic spine density is unaffected following striatal MCT4 knockdown. Data are
histogram of the frequency distribution of spine density and boxplot showing quantiles (25, 50,
75%) with central line marking the median and plus denoting the mean (inset). ns, p > 0.05 (n = 3
mice per group, Kolmogorov-Smirnov test).

147
Discussion
Astrocytic release of lactate in the central nervous system is well-documented
(Karagiannis et al., 2016; Mächler et al., 2016) and MCT4-mediated release of lactate
maintains the astrocyte-neuron lactate shuttle (ANLS) previously demonstrated to be
important in synaptic plasticity and behavior (Netzahualcoyotzi & Pellerin, 2020; Suzuki
et al., 2011). Whether the ANLS applies universally throughout the brain, or there exist
regional differences in intercellular metabolic coupling strategies such as the ANLS, is
not known. The dorsal striatum is a site of diverse and massive integration from
excitatory glutamatergic input, from across cortex and thalamus, and dopaminergic
inputs from the substantia nigra (Foster et al., 2020; Hintiryan et al., 2016; Hunnicutt et
al., 2016), and its output governs a myriad of motor and cognitive behaviors (Graybiel &
Grafton, 2015). The purpose of this study was to explore the importance of astrocytic
lactate shuttling, via MCT4, on synaptic structure and behavior in the dorsal striatum
using astrocyte specific, inducible Cre recombinase to knockdown MCT4 expression.

We previously developed and validated the pSico plasmid and lentiviral packaging
for shRNA-mediated knockdown of astrocytic MCT4 (see Chapter 4) in the motor cortex;
this approach worked similarly well in the dorsal striatum and was specific to MCT4.
Bilateral dorsal striatum knockdown of MCT4 caused a clear enhancement of behavioral
outputs across distinct tasks that are regulated by the dorsal striatum, including the
accelerating rotarod, object discrimination, and stereotypical self-grooming behavior.
Previous work investigating repetitive motor behaviors showed a similar improvement in
rotarod performance and stereotypical behavior that also occurred with hyperactivity in
148
the open field (Rothwell et al., 2014). Importantly, we did not observe any hyperactivity in
the open field, suggesting that changes in motor or stereotypical behavior following
MCT4 knockdown did not result from a lack of inhibition in the dorsal striatum.

The importance of astrocytic lactate shuttling in maintaining synaptic structure
and function has been noted in both the hippocampus and the cortex, which underlies
behavioral deficits associated with a loss of the ANLS (see Chapter 4), (Suzuki et al.,
2011; Vezzoli et al., 2020). Intriguingly, we did not observe any clear loss of various striatal
cell-type specific markers of synapses in the dorsal striatum following MCT4 knockdown.
This included spiny projection neurons and interneurons (DARPP-32 and ChAT/GAD65,
respectively), glutamatergic inputs from cortex (VGLUT1) and thalamus (VGLUT2), and
more universal pre- and postsynaptic markers GAP43, synaptophysin and PSD-95. A lack
of a clear synaptic or cellular lesion is curious, considering that interfering with the ANLS
has been shown to be detrimental to synapses; however, this may be explained by the
localization of MCT2, responsible for uptake of lactate into neurons. In cortex, MCT2
colocalizes with PSD95 (Pierre et al., 2002), and it was further shown that MCT2
colocalizes post-synaptically with AMPA receptors in the hippocampus (Bergersen et al.,
2005); however, in the striatum MCT2 and PSD95 do not colocalize (Pierre et al., 2002).
Physical segregation of MCT2 and the postsynaptic density suggest that neuronal uptake
of lactate in the striatum is not as important for maintaining synaptic energetics (Attwell
& Laughlin, 2001; J. J. Harris et al., 2012), and thus interruptions in lactate shuttling may
not as prominently affect synaptic survival.

149
Dopamine release in the dorsal striatum, and its distinct effect on D1R- and D2R-
expressing striatal spiny projection neuron activity, has been studied extensively and
underlies both the adaptive and maladaptive behavioral outcomes of the basal ganglia,
from normal habit formation to addiction to neurodegenerative disease (Berke, 2018;
Graybiel, 2000). The behavioral enhancement observed in our bilateral dorsal striatum
MCT4 knockdown model were mirrored by increased presynaptic dopamine synthesis –
measured through expression of TH and VMAT2 proteins – and heightened
phosphorylation of striatal DARPP-32, indicative of increased D1R activity. Interestingly,
we also observed a shifting of dopamine receptor balance as D2R expression decreased
following MCT4 knockdown, without affecting D1R expression, leading to a greater D1R-
to-D2R ratio. D1R activity regulates NMDA receptor expression in striatal synapses that
underlie neuroplasticity (Hallett et al., 2006); indeed, we found that expression of NMDA
receptor subunit NR1, common to all functional NMDA receptors, was increased in our
bilateral knockdown of MCT4, further suggesting a boost in D1R activity in the striatum.

The observed enhancement in behavior (motor performance, stereotypy, object
recognition) has been previously linked to dopamine D1 receptor activity (Darvas &
Palmiter, 2009; Frederick et al., 2015; Hotte et al., 2006; Rothwell et al., 2014; Taylor et al.,
2010). Our previous results did not demonstrate any gross change in pre- or postsynaptic
proteins, suggesting that synaptic and neuronal integrity was maintained without
astrocytic MCT4. In addition, presynaptic dopamine synthesis release seemed to be
enhanced and disproportionally acting on dopamine D1 receptors, largely indicating that
changes in behavior may indeed be linked to changes in the nigrostriatal dopamine
150
system. Using a unilateral approach to introduce possible disequilibrium in striatal
dopamine between hemispheres, leading to rotational behaviors – an approach in
unilateral models of dopamine depletion for studying Parkinson’s disease (Björklund &
Dunnett, 2019; Girasole et al., 2018) – allowed us to better understand whether changes
in dopamine release or dopamine receptor sensitivity were driving the observed
behavioral phenotypes. Administration of apomorphine, which agonizes dopamine
receptors, did not affect rotational behaviors, but amphetamine administration, which
causes release of presynaptic dopamine, did provoke a clear increase in rotational
behavior. Amphetamine also caused a significant increase in cFos expression in both
dorsal striatum and cortex in MCT4 knockdown mice, a known byproduct of enhanced
dopaminergic activity (Badiani et al., 1998; Ferguson et al., 2004). Thus, using uni- and
bilateral MCT4 knockdown models, various behaviors, pharmacology and molecular
biology approaches, our results indicate that astrocytic lactate shuttling – via MCT4 –
regulates presynaptic dopamine in the striatum.

The decrease in dopamine D2 receptor expression is puzzling, as western blots of
striatal proteins did not show any effect of MCT4 knockdown on synaptic protein
expression and Golgi impregnation did not show a decrease in dendritic spine density in
spiny projection neurons following MCT4 knockdown. Previous work in
hyperdopaminergic mice that lack the dopamine transporter showed a 50% reduction in
D2 autoreceptor expression in nigrostriatal neurons, suggesting that heightened
dopamine states can negatively regulate dopamine D2 receptor expression (Giros et al.,
1996). Whether the decrease in D2 receptor expression in our study is composed of pre-
151
or postsynaptic is difficult to say, as the postsynaptic dopamine D2 receptor and the
presynaptic dopamine D2 autoreceptor forms differ by just 29 amino acids and
commercial antibodies recognize both isoforms. However, we believe that MCT4
knockdown likely affects the expression of both isoforms, as the processes that either
isoform are known to negatively regulate (presynaptic TH activity and postsynaptic
DARPP-32 phosphorylation) were elevated in MCT4 knockdown mice (Lindgren et al.,
2003).

The mechanism by which a loss of astrocytic lactate release increases presynaptic
dopamine production is not known at this time. It is unlikely that increases in dopamine
are associated with a compensatory mechanism, as seen in moderate nigrostriatal
lesions in nonhuman primates (Perez et al., 2008), as striatal synaptic protein expression
and dendritic spine density were both unchanged in our knockdown model, a
characteristic response in moderate dopaminergic lesions (Meshul et al., 2000). One
possible mechanism may be that astrocytic lactate signaling through presynaptic HCAR1
acts as a negative feedback mechanism on nigrostriatal dopamine terminals in the
striatum, similar to how dopamine negatively regulates its own release through the
dopamine transporter or the D2 autoreceptor (Ford, 2014). HCAR1 and D2 autoreceptors
are Gi/o protein coupled receptors that canonically dampen cAMP activity, CREB
phosphorylation and subsequent expression of CREB-regulated transcripts, including TH
(Kumer & Vrana, 1996). Future work investigating such mechanisms will benefit from the
application of fast-scan cyclic voltammetry to understand dopamine release in the
context of MCT4 knockdown, as well as genetic or pharmacological manipulation of
152
presynaptic HCAR1 in nigrostriatal terminals to assess lactate receptor-mediated
signaling in controlling dopamine production.

Astrocytes have been shown to regulate striatal neuron activity and striatal
dopamine release (Martín et al., 2015; Roberts et al., 2020). Additionally, astrocytic lactate
contributes to norepinephrine release by noradrenergic neurons in the locus coeruleus
(Tang et al., 2014) and lactate can regulate neuronal firing rate via presynaptic HCAR1
activity (de Castro Abrantes et al., 2019). Thus, proof already exists for astrocytic
capability to modulate neuronal activity and neurotransmitter production and release in
diverse subcortical structures. Our results support the ability of astrocytic lactate to
modify neuroplasticity independent of known mechanisms, including energetic support
of synapses (see Chapter 4), and expands the scope of astrocytic lactate’s effects in the
brain to now include dopaminergic signaling. Taken together, we show that MCT4
knockdown results in an enhancement of presynaptic dopamine synthesis and release,
which in turn acts primarily through dopamine D1 receptors to drive molecular changes
in striatal spiny projection neurons and modulate behavioral activity.

153
Discussion and Future Directions
The findings presented in this dissertation represent an effort to understand the
effect of exercise on astrocytic plasticity, characterize a potential molecular mechanism
for lactate in promoting exercise-induced plasticity, and further explore how lactate
contributes to synaptic structure, signaling and behavioral adaptations. Chapter 2
demonstrates that in response to exercise, astrocytes undergo robust experience-
dependent neuroplasticity, just as neurons do. This response appeared to be toward
priming the central nervous system for region- and temporal-specific changes to promote
neuroplasticity and synaptogenesis but lacked a possible mechanism. Chapter 3
attempts to bridge this gap in knowledge by which exercise can modify astrocytic and
neuronal structure and function to promote plasticity, homing in on peripheral lactate
metabolism and its impact on the central nervous system. A combination of in vitro and
in vivo experiments determined that exogenous lactate acts upon astrocytes to stimulate
a pro-synaptogenic response, including a robust increase in neurotrophic gene
expression, but fails to impact motor behavior or synaptogenesis outright. This work
highlighted the potential that lactate plays in coordinating brain-body responses to
exercise, but the lack of neuronal changes suggested that brain-centric lactate sources
may play a more direct role in synaptic plasticity. Chapters 4 and 5 use parallel
approaches to ascertain the necessity of astrocyte-derived lactate, shuttled via its
transporter MCT4, in normal synaptic function in two distinct, but functionally related
brain regions. The results characterize a divergence in lactate-mediated responses
between the brain structures: in the motor cortex, astrocyte-derived lactate is important
in the maintenance of dendrites and normal motor learning; in the dorsal striatum,
154
astrocyte-derived lactate has no effect on dendrites and acts primarily through changing
pre-synaptic dopamine synthesis and activity to enhance dopamine-dependent
behaviors.
The combination of these experimental designs and technical approaches reveals
two important takeaways regarding lactate metabolism and astrocytic and neuronal
plasticity. First, lactate is a powerful modulator of astrocytic and synaptic structure and
function, and the source of the lactate likely determines how it is utilized by and acts upon
cells in the central nervous system.
The work highlighted in Chapter 3 demonstrated that exogenous lactate can drive
a robust genetic response in astrocytes. The application of 3,5-DHBA, an agonist of the
lactate receptor HCAR1, resulted in an even greater genetic response in vitro, suggesting
that lactate could work through both metabolic and intracellular signaling pathways. It is
this divergence in lactate’s function that may very well explain the differences seen in the
stark contrast shown between Chapters 4 and 5. The loss of astrocytic lactate shuttling
in the motor cortex led to a significant decrease in dendritic spine density, with the
synaptic effects playing out primarily post-synaptically possibly through affecting actin
mobilization. However, loss of astrocytic lactate shuttling in the dorsal striatum did not
affect dendritic spine density at all and instead drove the largest effect on presynaptic
dopamine. The specificity of post-synaptic effects following astrocytic MCT4 knockdown
in the motor cortex may very well be explained by the post-synaptic localization of the
neuron-specific transporter of lactate, MCT2, responsible for neuronal lactate uptake
(Pierre et al., 2002) in the rodent cortex; MCT2 has also been shown to colocalize with
glutamatergic receptors in the post-synaptic density in the hippocampus (Bergersen et
155
al., 2005). In contrast, MCT2 does not colocalize to the post-synaptic density in the rodent
striatum, suggesting a distinct capacity for neuronal lactate uptake outside of fueling
highly energetic synaptic activity in the striatum (Lauritzen et al., 2014; Pierre et al., 2002).
Indeed, the work presented in Chapters 4 and 5 support the notion presented by earlier
histological work that neuronal lactate uptake differs between distinct brain regions and
suggests that lactate’s utility is not limited strictly to energetic support.
The second takeaway is that regional heterogeneity of astrocytes and neurons
determines their response to experience-dependent plasticity and changes in
metabolism and is an increasingly important factor to consider in experimental
manipulations studying the interplay between these two cell-types. This second point is
particularly clear in Chapter 2, which underscored how region-specific astrocytic changes
in morphology and gene expression are in response to exercise. In recent years, there has
been a groundswell of research demonstrating the molecular, morphological, and
functional heterogeneity in astrocytes, both between discrete brain structures and across
layers within the same cortical region (Chai et al., 2017; Lanjakornsiripan et al., 2018).This
diversity of astrocytes is important in informing neuronal function as well; in vitro
experiments comparing co-culturing astrocytes and neurons from cortex and striatum
(i.e., cortical astrocyte-neuron cocultures, or cortical astrocytes cultured with striatal
neurons) demonstrated that region specificity in glial support of neurons is important in
normal structural and synaptic development (Morel et al., 2017). Furthermore, very new
work showed that astrocytes and neurons from the same brain region have conserved,
region-specific transcriptional signatures, suggesting that neuronal and astrocytic
maturation informs spatially restricted gene signatures, and by extension region-specific
156
function (Herrero-Navarro et al., 2021). Thus, it is not surprising that exercise, as a form
of experience-dependent neuroplasticity, drives region-specific astrocytic responses.
These astrocytic responses are likely in line with region-specific neuronal changes,
particularly in brain regions directly engaged in the behavior mediating neuroplasticity (in
this case, aerobic treadmill exercise).
The region specificity of astrocytes, and how they contribute to neuronal function,
is abundantly clear in both Chapters 4 and 5, where genetic manipulation of astrocytic
lactate shuttling led to disparate behavioral and molecular outcomes. Astrocytic lactate
transporters are present and expressed on astrocytes across brain regions throughout
postnatal development (Clarke et al., 2018), suggesting that a loss of astrocytic MCT4 in
one brain region would not have an outsized impact on synaptic function and behavior
compared to a different brain region. Thus, it is more likely that differences in the impact
of astrocytic MCT4 knockdown on synaptic structure and function are due to region-
specific differences in how astrocytes and neurons cooperate. This astrocyte-neuron
coordination very likely extends to metabolism: how a particular region contributes to
manifesting a discrete, coordinated behavioral output will determine how much energy it
needs to sustain synaptic activity, which will further determine the metabolic strategies
to meet this synaptic energetic demand. The region-specific adaptations in neuronal and
astrocytic plasticity in response to experience-dependent plasticity, such as exercise, is
dependent upon metabolic flexibility and strategies, and cellular plasticity likely extends
to metabolic plasticity. Indeed, early work not presented in this dissertation showed at
the transcriptional level that components of the astrocyte-neuron lactate shuttle – a
strategy to meet synaptic energy demands (Magistretti & Allaman, 2018) – were
157
heightened in an exercise-dependent manner in exercise-relevant brain regions (Halliday
et al., 2019). In fact, it is that early work that informed the experiments presented within
this dissertation, suggesting that astrocytic lactate shuttling may underlie the positive
synaptogenic and behavioral effects of exercise in animal models of disease, particularly
in animal models of Parkinson’s disease.
While the work presented in this dissertation represents a solid step forward in our
understanding of how astrocyte-neuron metabolic cooperation may underlie
neuroplasticity and behavioral adaptation, there are limitations found in each chapter that
deserve extra space to discuss. Whether the result of shifting laboratory obligations,
changing of hands in the laboratory, or the COVID-19 pandemic, there are further
experiments that could have been performed to better understand and extend the results
presented previously. This section is designed not to belittle the collective work produced
by trainees and collaborators, but to promote reflection and offer suggestions and
commentary to future work carried out in our laboratory and in conjunction with
collaborators, in the hopes that we collectively continue to push toward a stronger truth.
Chapter 2 represented a first pass at understanding how exercise drives astrocytic
plasticity and was intended to act as a birds-eye view by focusing on histological and
morphological changes. While this approach offered a number of novel insights and was
buoyed by the laborious and longitudinal fashion of astrocyte analysis, it lacked a higher
resolution that is now evidently absent and sorely needed to better contextualize the
astrocyte-specific changes observed. The changes in astrocyte gene expression were
generalized from previous work in highly pure astrocytes and would have strongly
benefited from astrocyte isolation to assess cell type-specific transcriptional changes. At
158
the time in the laboratory (2017 and early 2018), we had not yet conceived of and
generated the transgenic mouse lines to express fluorescent tdTomato under the control
of the astrocyte promoter Aldh1l1 – a Cre-driver line essential in Chapters 4 and 5 – and
this was not on the metaphorical radar. RiboTag technologies – which allows for cell type-
specific hemagglutinin tagging and subsequent immunoprecipitation of actively
transcribing ribosomes (Nectow et al., 2017; Yu et al., 2020) – could have been delivered
by astrocyte-specific adeno-associated viruses via stereotaxic surgery to provide region-
specific gene expression in astrocytes. Looking back, seeing as the path forward led
toward metabolic contributions to synaptic plasticity, it would have been additionally
beneficial to examine metabolic changes within astrocytes – either via two-color
fluorescent immunohistochemistry or previously stated RiboTag approaches – to provide
additional context to work previously led by Dr. Matthew Halliday (Halliday et al., 2019).
Lastly, in hindsight it feels obvious that astrocytic changes should be tracked in
relation to synaptic changes, as this morphological, functional, and metabolic link is
continuously explored in the literature and begs to be studied in tandem. Two
technological advancements not currently being utilized by our laboratory would provide
a high degree of functional and structural information regarding neuronal and synaptic
changes in response to experience such as exercise. First, in vivo two-photon imaging in
conjunction with sparsely labeled fluorescent neurons (achievable by the Thy1-YFP ‘H-
line’ mouse model) would provide lifetime analysis of cortical dendrites (Xu et al., 2009)
in response to exercise, which could be aligned with astrocytic structural or functional
changes. Second, in vivo calcium imaging using MiniScopes (Ghosh et al., 2011)
represents real-time functional data of neuronal activation during exercise, which may
159
provide novel insights to how region-specific neuronal ensembles engaged during
exercise change or respond to exercise duration. The utility of existing approaches, such
as western immunoblotting and immunohistochemistry for synaptic protein changes, or
Golgi impregnation for microscopic analysis of dendritic spines, also present viable
options to understand synaptic dynamics in relation to astrocytic changes but represent
snapshots of structural or functional changes compared to the dynamic and longitudinal
approaches outlined above.
Chapter 3 tackled a possible mechanism for the observed astrocyte changes in
exercise by linking peripheral metabolism of lactate to astrocyte-specific transcriptional
and structural adaptations. At the time (fall 2018 – fall 2019), the goal of this project was
to understand how much peripheral lactate contributed to synaptic and astrocytic
changes; this was a pertinent topic raised by myself after the writing and submission of
the funding proposal that ultimately supported Chapters 4 and 5 of this dissertation. How
would we be able to understand how a loss of astrocytic MCT4 – proposed to minimize
astrocytic lactate – impacted exercise-induced behavioral and synaptic recovery in a
mouse model of Parkinson’s if we knew that exercise caused massive peripheral rises in
lactate that was consumed by the brain (Quistorff et al., 2008), but didn’t understand how
that lactate might affect synaptic and astrocytic function? While this approach was
insightful and well-informed based upon existing literature, there are now obvious
shortcomings to the approach.
First, the administration of 3,5-DHBA in vivo to mirror in vitro experimentations
would have greatly enhanced the project and helped to delineate the signaling and/or
metabolic contributions of lactate to the genetic, morphological, and behavioral changes
160
studied. In the same vein of 3,5-DHBA and its activity at the lactate receptor HCAR1, a
viral approach to knockdown astrocytic HCAR1 in vivo existed at the time; I had requested
a lentiviral stock prepared by our collaborators in the Cannon laboratory designed to
knockdown HCAR1, but never applied it in the context of the in vivo experiments
presented in Chapter 3. This approach could have also been explored in vitro as well but
would have either required either the Cre-expressing astrocytic cell line we were still
developing at the time (utilized in Chapter 4) or the use of limited Cre-expressing mouse
pups to generate astrocytic cultures. This would have also provided a great deal of
understanding regarding lactate’s effects in modulating astrocytic function in promotion
of synaptic plasticity. Global MCT1 knockouts are embryonically lethal (Lee et al., 2012),
thus blocking brain uptake of lactate is not possible and limits how much one can
understand how peripheral lactate contributes metabolically to observed astrocytic and
synaptic plasticity. In light of the interesting effects that HCAR1 causes in regulating
presynaptic neuronal activity (de Castro Abrantes et al., 2019) and the dependence of
synapses on lactate to function (Murphy-Royal et al., 2020), lactate’s effect on neuronal
function – either metabolically or signaling – could also be explored using neuron-
specific genetic manipulation of MCT2 or HCAR1 to round out our understanding of
peripheral lactate’s effects on astrocytic and synaptic plasticity. There remains an
intriguing, oft repeated and lingering question since the data in Chapter 3 were generated:
could peripheral lactate, as an incomplete mimetic of exercise, be administered in
conjunction with low-intensity exercise to recapitulate high-intensity exercise, particularly
in animal models of Parkinson’s disease? This may possibly have direct application of
tailoring exercise paradigms for Parkinson’s disease patients with severe, late-stage
161
dopaminergic degeneration, where high-intensity exercise may be motorically
challenging or impossible and the benefits of exercise limited.
Chapter 4 signifies a convergence of new methodologies and experimental
approaches and was a major step forward in our understanding of MCT4, and by
extension astrocytic lactate, and how it contributes to motor learning. Initial in vitro and
in vivo validation of the models and behavioral studies for this project were completed in
late 2019 and showed great promise for studying astrocytic MCT4 in motor behaviors;
this, in turn, prompted a slew of new cohorts reared and surgically prepared for analysis
in early 2020, experimental plans that abruptly ended with the onset of the COVID-19
pandemic and the shuttering of the laboratory for several months. While I attempted to
reclaim experimental losses upon our laboratory’s reopening – like so many researchers
across the globe – there were inevitably experiments that had to be put on the back burner
or completely abandoned in favor of completing more pressing questions. These
experiments will be discussed here, as I believe that the experimental approach of
probing astrocytic MCT4 is rich with insightful deviations. Not all these experiments are
feasible in our laboratory at the moment, but choice collaborations with core facilities or
other laboratories in the future may present opportunities to explore these questions.
First, the lack of a lactate rescue experiments (administering lactate, either via
cannula infusion into the motor cortex or intraperitoneal injection) makes the job of the
corresponding scientist a trickier one in convincing my reader that the lack of astrocytic
lactate is directly responsible for synaptic structural, behavioral, and neuronal activity
deficits observed in Chapter 4. While astrocytes technically have the capacity to uptake
other monocarboxylates from the periphery and subsequently shuttle them out to the
162
perisynaptic space, experimental evidence in normal, physiologically relevant conditions
does not point to a notable role for non-lactate monocarboxylates (such as pyruvate and
ketone bodies) in being MCT4-shuttle substrates (Mächler et al., 2016). It is, of course,
experimentally burdensome to properly discriminate between functionally distinct
substrates commonly transported by the same shuttle; however, a lactate rescue in our
model would have provided a clearer picture as to whether the lack of lactate was truly
the culprit in our astrocytic MCT4 knockdown mice, as others before us have done
(Descalzi et al., 2019; Suzuki et al., 2011).
Second, the introduction of a different motor learning task – such as the skilled
reaching task – would have strengthened the conclusion that astrocytic MCT4 in the
motor cortex contributes to motor learning. This model has been used elegantly to
demonstrate a variety of synaptic changes in response to motor learning (Xu et al., 2009),
often in conjunction with chronic in vivo two-photon imaging of cortical dendrites. This
approach could be used to study how a loss of astrocytic MCT4 affects dendritic spine
dynamics and, in theory, could be combined with Förster resonance energy transfer
(FRET) fluorescent probes designed to track lactate dynamics (Mächler et al., 2016). Such
an approach would allow for real-time imaging of lactate changes in astrocytes with
stable lifetime imaging of dendritic spines, providing possible insights of energy
metabolism during distinct phases of motor learning, how that metabolism changes in
line with dendritic spine changes, and how MCT4 knockdown affects distinct aspects of
dendritic spine remodeling.
Third, dendritic spine pruning is carried out by glial cells, particularly microglia
(Hong et al., 2016). There exists a careful balance of ‘eat me’ and ‘don’t eat me’ signals
163
expressed by neuronal synapses and interpreted by microglia, principally carried out
through the classical complement system (Schafer et al., 2012), that is important in
normal developmental sculpting of neuronal circuits. In instances of aberrant dendritic
spines, such as in the instance of astrocytic MCT4 knockdown, are normal synaptic
recycling mechanisms ramped up as possible energetic supply dwindles? A histological
survey of microglia engulfment of postsynaptic proteins would provide a first pass
understanding of dendritic spine dynamics in the absence of astrocytic MCT4, and a
deeper understanding of how energy metabolism and normal dendritic maintenance
interact via microglia could be carried out using high resolution single cell RNA
sequencing or spatial transcriptomic approaches. Such work might unveil unprecedented
connections between astrocytic metabolism, microglial activity, and circuit refinement
that would suggest that neuroenergetics is an arena in which all cells participate and that
manipulations of neuroenergetics cannot be generalized to a single specific process.
While there was no obvious evidence of microglial activation via western blotting or
immunohistochemistry assessed by expression of Iba1 (which is classically heightened
in activated microglia), enhanced microglial pruning is better characterized by levels of
the complement receptor 3 (Schafer et al., 2012).
Chapter 5 presents a puzzling and fascinating divergence from the findings in
Chapter 4, despite utilizing identical approaches in two functionally related regions. How
might astrocytic MCT4 – and its presumed shuttling of lactate to synapses – cause
enhancement of presynaptic dopamine in the striatum without affecting synaptic
integrity? Certain shortcomings discussed for Chapter 4, such as the lack of a lactate
rescue, are common to Chapter 5 as well. However, because the overall takeaway from
164
Chapter 5 was distinct from Chapter 4, new approaches should be used in the future to
deepen our understanding and shore up our conclusions.
First, the absence of fast-scan cyclic voltammetry employed to measure dopamine
metabolism and evoked release in the striatum is a weakness. The evidence collected
and presented in favor of enhanced presynaptic dopamine – including dopamine-
mediated behaviors, immunohistochemistry, and western blotting of tyrosine hydroxylase
and VMAT2, and pharmacology to ascertain pre- vs. post-synaptic dopamine activity –
builds a strong case for a loss of astrocytic MCT4 promoting more presynaptic
dopamine. However, electrophysiological data would solidify these conclusions. Data
exists for lactate enhancing dopamine in the striatum, as measured by fast-scan cyclic
voltammetry, but this is in acute manipulation of lactate metabolism and uptake in ex-
vivo striatal slices (Msackyi et al., 2020, bioRxiv); it seems likely that more chronic
changes in lactate availability and metabolism (such as in the genetic knockdown of
astrocytic MCT4 in this model) could also affect dopamine in opposing ways to existing
experimental data. Indeed, longitudinal behavioral data studying spontaneous rotation in
unilateral astrocytic MCT4 knockdown in the striatum showed a dynamic change in
rotational direction (Fig. 5.8) suggesting an active rebalancing of dopamine in response
to a loss of astrocytic lactate. Thus, utilizing fast-scan cyclic voltammetry at multiple
timepoints would provide a complete picture of how dopamine responds longitudinally to
changes in lactate availability.
This also serves as a reminder to the relative novelty of lactate in modulating
dopamine, as currently little data exists to direct a curious experimentalist. In light of this
novelty, possible mechanisms driving the observed dopaminergic changes are plentiful
165
and difficult to eliminate based on present literature. One possibility is that astrocytic
lactate release, which is classically considered to heightened in response to synaptic
activity, acts as a negative regulator of presynaptic activity (Fig. 6.1). Thus, in the case of
decreased astrocytic MCT4, typical perisynaptic lactate release is attenuated and fails to
act as a presynaptic brake. This is supported by the fact that the lactate receptor HCAR1
is expressed presynaptically and regulates neuronal excitability; in vitro, HCAR1 knockout
neurons are overactive (de Castro Abrantes et al., 2019). HCAR1 is a Gi/o-protein coupled
receptor (GPCR), whose signaling pathway negatively regulates the expression of many
CREB-regulated genes, including tyrosine hydroxylase and VMAT2 (Kumer & Vrana,
1996). In this scenario, impaired lactate inhibition of HCAR1 could lead to heightened
tyrosine hydroxylase and VMAT2, resulting in increased presynaptic dopamine. Increased
synaptic dopamine tone is known to decrease dopamine D2 receptor expression (Giros
et al., 1996) and is implicated in enhanced motor and stereotyped behaviors, all of which
were seen in astrocytic MCT4 knockdown mice. If this scenario were the mechanism by
which astrocytic shuttling of lactate via MCT4 regulated presynaptic dopamine, a simple
chronic lactate or 3,5-DHBA rescue experiment could help to prove or disprove this
hypothesis. Furthermore, more advanced and novel techniques such as axon terminal-
specific chemogenetic approaches to artificially stimulate inhibitory Gi/o-DREADDs
(designer receptors exclusively activated by designer drugs) (Roth, 2016) or axon-
terminal specific optogenetic manipulations (Copits et al., 2021) could be leveraged to
study this possible mechanism, including in a cell- or pathway-specific manner by
combining viral DREADD/opsin delivery with intersectional mouse genetics. Whether the
change in presynaptic dopamine following astrocytic MCT4 knockdown was specific to
166
the nigrostriatal dopamine system – or is broadly applicable to other dopaminergic
systems (such as dopaminergic axons arising from the ventral tegmental area and
projecting to the nucleus accumbens) – remains to be known.
Figure 6.1 – Proposed mechanism by which striatal astrocytic lactate may modulate dopamine.
In the normal striatum (left), lactate flows via perisynaptic MCT4 to act on presynaptic HCAR1,
which negatively regulates cAMP and CREB transcription, acting as a break on presynaptic
nigrostriatal dopamine synthesis and activity. In the MCT4 knockdown striatum (right), lactate
flow is blocked; diminished lactate tone decreases HCAR1-mediated inhibition of cAMP, leading
to unregulated CREB activity and increase in target genes, including TH and VMAT2, that
ultimately manifest in exaggerated striatal behaviors.

Looking forward, the data presented in Chapter 5 may also have applications in
atypical dopaminergic states, including hyperdopaminergic tone encountered in
psychiatric disorders such as schizophrenia or hypodopaminergic tone highlighted by
neurodegenerative diseases such as Parkinson’s disease. How might lactate shuttling go
awry in these situations? Little focus has been directed toward astrocyte-neuron
metabolism in psychiatric or degenerative contexts; recent and future data utilizing high-
167
resolution transcriptional analysis of human post-mortem brain tissue, along with
increasingly reliable animal models of neurological disorders, will provide the opportunity
to examine cell type-specific metabolic cooperation and how it contributes to abnormal
dopaminergic states. In the context of previous work from our laboratory, it is tempting
to consider whether exercise-induced lactate contributes to the regulation of
dopaminergic activity and dopamine receptor expression in dopamine-depleted mouse
models of Parkinson’s disease. In conjunction with established effects of exercise, and
lactate, on establishing a pro-synaptogenic environment through enhancement of
neurotrophic support (Lundquist et al., 2021), lactate may also work to curb inappropriate
synaptic changes, working in a multivariate manner through metabolic, neurotrophic, and
neuromodulatory mechanisms to promote experience-dependent synaptic plasticity
necessary for behavioral recovery.
Taken together, the findings presented in this dissertation represent novel insights for
understanding how cellular metabolism is linked to neuroplasticity, especially focused on
the divergent roles lactate plays across the brain. Carried out in the context of healthy
animals, this work can help inform future studies investigating how cellular energetics
contribute to plasticity. In the context of our own laboratory’s work studying the effects
of exercise on neuroplasticity in dopamine-depleted animal models of Parkinson’s, this
work reinforces lactate as a potentially pivotal molecule in coordinating and driving the
structural and functional changes previously documented and suggests modulation of
lactate as a possible future therapeutic target.

168
Bibliography
Abbott, N. J., Rönnbäck, L., & Hansson, E. (2006). Astrocyte–endothelial interactions at
the blood–brain barrier. Nature Reviews Neuroscience, 7(1), 41–53.
https://doi.org/10.1038/nrn1824
Abou-Sleiman, P. M., Muqit, M. M. K., & Wood, N. W. (2006). Expanding insights of
mitochondrial dysfunction in Parkinson’s disease. Nature Reviews Neuroscience,
7(3), 207–219. https://doi.org/10.1038/nrn1868
Adamsky, A., Kol, A., Kreisel, T., Doron, A., Ozeri-Engelhard, N., Melcer, T., Refaeli, R.,
Horn, H., Regev, L., Groysman, M., London, M., & Goshen, I. (2018). Astrocytic
Activation Generates De Novo Neuronal Potentiation and Memory Enhancement.
Cell, 174(1), 59-71.e14. https://doi.org/10.1016/j.cell.2018.05.002
Agulhon, C., Petravicz, J., McMullen, A. B., Sweger, E. J., Minton, S. K., Taves, S. R.,
Casper, K. B., Fiacco, T. A., & McCarthy, K. D. (2008). What Is the Role of Astrocyte
Calcium in Neurophysiology? In Neuron.
https://doi.org/10.1016/j.neuron.2008.09.004
Ahmed, K., Tunaru, S., & Offermanns, S. (2009). GPR109A, GPR109B and GPR81, a
family of hydroxy-carboxylic acid receptors. Trends in Pharmacological Sciences,
30(11), 557–562. https://doi.org/10.1016/j.tips.2009.09.001
Alberini, C. M., Cruz, E., Descalzi, G., Bessières, B., & Gao, V. (2018). Astrocyte glycogen
and lactate: New insights into learning and memory mechanisms. Glia, 66(6),
1244–1262. https://doi.org/10.1002/glia.23250
Allen, N. J., & Lyons, D. A. (2018). Glia as architects of central nervous system formation
and function. Science, 362(6411), 181–185.
169
https://doi.org/10.1126/science.aat0473
Andersson, A. K., Rönnbäck, L., & Hansson, E. (2005). Lactate induces tumour necrosis
factor-α, interleukin-6 and interleukin-1β release in microglial- and astroglial-
enriched primary cultures. Journal of Neurochemistry, 93(5), 1327–1333.
https://doi.org/10.1111/j.1471-4159.2005.03132.x
Anzalone, A., Lizardi-Ortiz, J. E., Ramos, M., De Mei, C., Hopf, F. W., Iaccarino, C., Halbout,
B., Jacobsen, J., Kinoshita, C., Welter, M., Caron, M. G., Bonci, A., Sulzer, D., &
Borrelli, E. (2012). Dual Control of Dopamine Synthesis and Release by Presynaptic
and Postsynaptic Dopamine D2 Receptors. Journal of Neuroscience, 32(26), 9023–
9034. https://doi.org/10.1523/JNEUROSCI.0918-12.2012
Attwell, D., & Laughlin, S. B. (2001). An Energy Budget for Signaling in the Grey Matter of
the Brain. Journal of Cerebral Blood Flow & Metabolism, 21(10), 1133–1145.
https://doi.org/10.1097/00004647-200110000-00001
Badiani, A., Oates, M. M., Day, H. E. W., Watson, S. J., Akil, H., & Robinson, T. E. (1998).
Amphetamine-Induced Behavior, Dopamine Release, and c- fos mRNA Expression:
Modulation by Environmental Novelty. The Journal of Neuroscience, 18(24), 10579–
10593. https://doi.org/10.1523/JNEUROSCI.18-24-10579.1998
Bak, L. K., & Walls, A. B. (2018). Astrocytic glycogen metabolism in the healthy and
diseased brain. Journal of Biological Chemistry, 293(19), 7108–7116.
https://doi.org/10.1074/jbc.R117.803239
Barros, L. F. (2013). Metabolic signaling by lactate in the brain. In Trends in
Neurosciences. https://doi.org/10.1016/j.tins.2013.04.002
Bateup, H. S., Svenningsson, P., Kuroiwa, M., Gong, S., Nishi, A., Heintz, N., & Greengard,
170
P. (2008). Cell type–specific regulation of DARPP-32 phosphorylation by
psychostimulant and antipsychotic drugs. Nature Neuroscience, 11(8), 932–939.
https://doi.org/10.1038/nn.2153
Bauer, S., Kerr, B. J., & Patterson, P. H. (2007). The neuropoietic cytokine family in
development, plasticity, disease and injury. In Nature Reviews Neuroscience.
https://doi.org/10.1038/nrn2054
Bélanger, M., Allaman, I., & Magistretti, P. J. (2011). Brain energy metabolism: Focus on
Astrocyte-neuron metabolic cooperation. In Cell Metabolism (Vol. 14, Issue 6, pp.
724–738). https://doi.org/10.1016/j.cmet.2011.08.016
Bergersen, L. H., Magistretti, P. J., & Pellerin, L. (2005). Selective Postsynaptic Co-
localization of MCT2 with AMPA Receptor GluR2/3 Subunits at Excitatory Synapses
Exhibiting AMPA Receptor Trafficking. Cerebral Cortex, 15(4), 361–370.
https://doi.org/10.1093/cercor/bhh138
Berke, J. D. (2018). What does dopamine mean? Nature Neuroscience, 21(6), 787–793.
https://doi.org/10.1038/s41593-018-0152-y
Bernardi, C., Tramontina, A. C., Nardin, P., Biasibetti, R., Costa, A. P., Vizueti, A. F.,
Batassini, C., Tortorelli, L. S., Wartchow, K. M., Dutra, M. F., Bobermin, L.,
Sesterheim, P., Quincozes-Santos, A., de Souza, J., & Gonçalves, C. A. (2013).
Treadmill Exercise Induces Hippocampal Astroglial Alterations in Rats. Neural
Plasticity, 2013, 1–10. https://doi.org/10.1155/2013/709732
Bernardinelli, Y., Randall, J., Janett, E., Nikonenko, I., König, S., Jones, E. V., Flores, C. E.,
Murai, K. K., Bochet, C. G., Holtmaat, A., & Muller, D. (2014). Activity-dependent
structural plasticity of perisynaptic astrocytic domains promotes excitatory
171
synapse stability. Current Biology, 24(15), 1679–1688.
https://doi.org/10.1016/j.cub.2014.06.025
Berretta, S., Robertson, H. A., & Graybiel, A. M. (1992). Dopamine and glutamate
agonists stimulate neuron-specific expression of Fos-like protein in the striatum.
Journal of Neurophysiology, 68(3), 767–777.
https://doi.org/10.1152/jn.1992.68.3.767
Berridge, K. C., Fentress, J. C., & Parr, H. (1987). Natural syntax rules control action
sequence of rats. Behavioural Brain Research, 23(1), 59–68.
https://doi.org/10.1016/0166-4328(87)90242-7
Billat, V. L., Mouisel, E., Roblot, N., & Melki, J. (2005). Inter- and intrastrain variation in
mouse critical running speed. Journal of Applied Physiology, 98(4), 1258–1263.
https://doi.org/10.1152/japplphysiol.00991.2004
Bittar, P. G., Charnay, Y., Pellerin, L., Bouras, C., & Magistretti, P. J. (1996). Selective
distribution of lactate dehydrogenase isoenzymes in neurons and astrocytes of
human brain. Journal of Cerebral Blood Flow and Metabolism, 16(6), 1079–1089.
https://doi.org/10.1097/00004647-199611000-00001
Björklund, A., & Dunnett, S. B. (2019). The Amphetamine Induced Rotation Test: A Re-
Assessment of Its Use as a Tool to Monitor Motor Impairment and Functional
Recovery in Rodent Models of Parkinson’s Disease. Journal of Parkinson’s Disease,
9(1), 17–29. https://doi.org/10.3233/JPD-181525
Black, J. E., Isaacs, K. R., Anderson, B. J., Alcantara, A. A., & Greenough, W. T. (1990).
Learning causes synaptogenesis, whereas motor activity causes angiogenesis, in
cerebellar cortex of adult rats. Proceedings of the National Academy of Sciences,
172
87(14), 5568–5572. https://doi.org/10.1073/pnas.87.14.5568
Bose, A., & Beal, M. F. (2016). Mitochondrial dysfunction in Parkinson’s disease. Journal
of Neurochemistry, 139, 216–231. https://doi.org/10.1111/jnc.13731
Boury-Jamot, B., Carrard, A., Martin, J. L., Halfon, O., Magistretti, P. J., & Boutrel, B.
(2016). Disrupting astrocyte–neuron lactate transfer persistently reduces
conditioned responses to cocaine. Molecular Psychiatry, 21(8), 1070–1076.
https://doi.org/10.1038/mp.2015.157
Bowser, D. N. (2004). ATP Excites Interneurons and Astrocytes to Increase Synaptic
Inhibition in Neuronal Networks. Journal of Neuroscience, 24(39), 8606–8620.
https://doi.org/10.1523/JNEUROSCI.2660-04.2004
Brockett, A. T., LaMarca, E. A., & Gould, E. (2015). Physical exercise enhances cognitive
flexibility as well as astrocytic and synaptic markers in the medial prefrontal cortex.
PLoS ONE, 10(5). https://doi.org/10.1371/journal.pone.0124859
Brooks, G. A. (2018). The Science and Translation of Lactate Shuttle Theory. Cell
Metabolism, 27(4). https://doi.org/10.1016/j.cmet.2018.03.008
Brooks, G. A. (2020). Lactate as a fulcrum of metabolism. Redox Biology, 35(January),
101454. https://doi.org/10.1016/j.redox.2020.101454
Burgess, A., Vigneron, S., Brioudes, E., Labbé, J.-C., Lorca, T., & Castro, A. (2010). Loss of
human Greatwall results in G2 arrest and multiple mitotic defects due to
deregulation of the cyclin B-Cdc2/PP2A balance. Proceedings of the National
Academy of Sciences, 107(28), 12564–12569.
https://doi.org/10.1073/pnas.0914191107
Carrard, A., Elsayed, M., Margineanu, M., Boury-Jamot, B., Fragnière, L., Meylan, E. M.,
173
Petit, J.-M., Fiumelli, H., Magistretti, P. J., & Martin, J.-L. (2018). Peripheral
administration of lactate produces antidepressant-like effects. Molecular
Psychiatry, 23(2), 392–399. https://doi.org/10.1038/mp.2016.179
Chai, H., Diaz-Castro, B., Shigetomi, E., Monte, E., Octeau, J. C., Yu, X., Cohn, W.,
Rajendran, P. S., Vondriska, T. M., Whitelegge, J. P., Coppola, G., & Khakh, B. S.
(2017). Neural Circuit-Specialized Astrocytes: Transcriptomic, Proteomic,
Morphological, and Functional Evidence. Neuron, 95(3), 531-549.e9.
https://doi.org/10.1016/j.neuron.2017.06.029
Christopherson, K. S., Ullian, E. M., Stokes, C. C. A., Mullowney, C. E., Hell, J. W., Agah, A.,
Lawler, J., Mosher, D. F., Bornstein, P., & Barres, B. A. (2005). Thrombospondins are
astrocyte-secreted proteins that promote CNS synaptogenesis. Cell, 120(3), 421–
433. https://doi.org/10.1016/j.cell.2004.12.020
Chung, W.-S., Clarke, L. E., Wang, G. X., Stafford, B. K., Sher, A., Chakraborty, C., Joung, J.,
Foo, L. C., Thompson, A., Chen, C., Smith, S. J., & Barres, B. A. (2013). Astrocytes
mediate synapse elimination through MEGF10 and MERTK pathways. Nature,
504(7480), 394–400. https://doi.org/10.1038/nature12776
Clarke, L. E., Liddelow, S. A., Chakraborty, C., Münch, A. E., Heiman, M., & Barres, B. A.
(2018). Normal aging induces A1-like astrocyte reactivity. Proceedings of the
National Academy of Sciences, 115(8), E1896–E1905.
https://doi.org/10.1073/pnas.1800165115
Coco, M., Caggia, S., Musumeci, G., Perciavalle, V., Graziano, A. C. E., Pannuzzo, G., &
Cardile, V. (2013). Sodium L-lactate differently affects brain-derived neurothrophic
factor, inducible nitric oxide synthase, and heat shock protein 70 kDa production in
174
human astrocytes and SH-SY5Y cultures. Journal of Neuroscience Research, 91(2),
313–320. https://doi.org/10.1002/jnr.23154
Copits, B. A., Gowrishankar, R., O’Neill, P. R., Li, J.-N., Girven, K. S., Yoo, J. J., Meshik, X.,
Parker, K. E., Spangler, S. M., Elerding, A. J., Brown, B. J., Shirley, S. E., Ma, K. K. L.,
Vasquez, A. M., Stander, M. C., Kalyanaraman, V., Vogt, S. K., Samineni, V. K.,
Patriarchi, T., … Bruchas, M. R. (2021). A photoswitchable GPCR-based opsin for
presynaptic inhibition. Neuron. https://doi.org/10.1016/j.neuron.2021.04.026
Cotman, C. W., & Berchtold, N. C. (2002). Exercise: A behavioral intervention to enhance
brain health and plasticity. In Trends in Neurosciences.
https://doi.org/10.1016/S0166-2236(02)02143-4
Dalsgaard, M. K., Quistorff, B., Danielsen, E. R., Selmer, C., Vogelsang, T., & Secher, N. H.
(2004). A reduced cerebral metabolic ratio in exercise reflects metabolism and not
accumulation of lactate within the human brain. The Journal of Physiology, 554(2),
571–578. https://doi.org/10.1113/jphysiol.2003.055053
Darvas, M., & Palmiter, R. D. (2009). Restriction of dopamine signaling to the
dorsolateral striatum is sufficient for many cognitive behaviors. Proceedings of the
National Academy of Sciences, 106(34), 14664–14669.
https://doi.org/10.1073/pnas.0907299106
Davies, J. M. S., Cillard, J., Friguet, B., Cadenas, E., Cadet, J., Cayce, R., Fishmann, A.,
Liao, D., Bulteau, A. L., Derbré, F., Rébillard, A., Burstein, S., Hirsch, E., Kloner, R. A.,
Jakowec, M., Petzinger, G., Sauce, D., Sennlaub, F., Limon, I., … Davies, K. J. A.
(2017). The Oxygen Paradox, the French Paradox, and age-related diseases.
GeroScience, 39(5–6), 499–550. https://doi.org/10.1007/s11357-017-0002-y
175
Dayan, E., & Cohen, L. G. (2011). Neuroplasticity Subserving Motor Skill Learning.
Neuron, 72(3), 443–454. https://doi.org/10.1016/j.neuron.2011.10.008
de Castro Abrantes, H., Briquet, M., Schmuziger, C., Restivo, L., Puyal, J., Rosenberg, N.,
Rocher, A.-B., Offermanns, S., & Chatton, J.-Y. (2019). The Lactate Receptor HCAR1
Modulates Neuronal Network Activity through the Activation of G α and G βγ
Subunits. The Journal of Neuroscience, 39(23), 4422–4433.
https://doi.org/10.1523/JNEUROSCI.2092-18.2019
de Senna, P. N., Bagatini, P. B., Galland, F., Bobermin, L., do Nascimento, P. S., Nardin, P.,
Tramontina, A. C., Gonçalves, C. A., Achaval, M., & Xavier, L. L. (2017). Physical
exercise reverses spatial memory deficit and induces hippocampal astrocyte
plasticity in diabetic rats. Brain Research, 1655, 242–251.
https://doi.org/10.1016/j.brainres.2016.10.024
Descalzi, G., Gao, V., Steinman, M. Q., Suzuki, A., & Alberini, C. M. (2019). Lactate from
astrocytes fuels learning-induced mRNA translation in excitatory and inhibitory
neurons. Communications Biology, 2(1), 247. https://doi.org/10.1038/s42003-019-
0495-2
Díaz-García, C. M., Mongeon, R., Lahmann, C., Koveal, D., Zucker, H., & Yellen, G. (2017).
Neuronal Stimulation Triggers Neuronal Glycolysis and Not Lactate Uptake. Cell
Metabolism, 26(2). https://doi.org/10.1016/j.cmet.2017.06.021
Díaz‐García, C. M., & Yellen, G. (2019). Neurons rely on glucose rather than astrocytic
lactate during stimulation. Journal of Neuroscience Research, 97(8), 883–889.
https://doi.org/10.1002/jnr.24374
Dienel, G. A. (2018). Brain Glucose Metabolism: Integration of Energetics with Function.
176
Physiological Reviews, 99(1), 949–1045.
https://doi.org/10.1152/physrev.00062.2017
Dienel, G. A., & Cruz, N. F. (2016). Aerobic glycolysis during brain activation: adrenergic
regulation and influence of norepinephrine on astrocytic metabolism. Journal of
Neurochemistry, 138(1), 14–52. https://doi.org/10.1111/jnc.13630
Dong, H. W. (2008). The Allen reference atlas: A digital color brain atlas of the C57Bl/6J
male mouse. In The Allen reference atlas: A digital color brain atlas of the C57Bl/6J
male mouse. (pp. ix, 366–ix, 366). John Wiley & Sons Inc.
Doyon, J., Bellec, P., Amsel, R., Penhune, V., Monchi, O., Carrier, J., Lehéricy, S., & Benali,
H. (2009). Contributions of the basal ganglia and functionally related brain
structures to motor learning. Behavioural Brain Research, 199(1), 61–75.
https://doi.org/10.1016/j.bbr.2008.11.012
E, L., Lu, J., Selfridge, J. E., Burns, J. M., & Swerdlow, R. H. (2013). Lactate administration
reproduces specific brain and liver exercise-related changes. Journal of
Neurochemistry, 127(1), 91–100. https://doi.org/10.1111/jnc.12394
El Hayek, L., Khalifeh, M., Zibara, V., Abi Assaad, R., Emmanuel, N., Karnib, N., El-
Ghandour, R., Nasrallah, P., Bilen, M., Ibrahim, P., Younes, J., Abou Haidar, E., Barmo,
N., Jabre, V., Stephan, J. S., & Sleiman, S. F. (2019). Lactate mediates the effects of
exercise on learning and memory through SIRT1-dependent activation of
hippocampal brain-derived neurotrophic factor (BDNF). The Journal of
Neuroscience, 1661–18. https://doi.org/10.1523/JNEUROSCI.1661-18.2019
Eroglu, Ç., Allen, N. J., Susman, M. W., O’Rourke, N. A., Park, C. Y., Özkan, E., Chakraborty,
C., Mulinyawe, S. B., Annis, D. S., Huberman, A. D., Green, E. M., Lawler, J.,
177
Dolmetsch, R., Garcia, K. C., Smith, S. J., Luo, Z. D., Rosenthal, A., Mosher, D. F., &
Barres, B. A. (2009). Gabapentin Receptor α2δ-1 Is a Neuronal Thrombospondin
Receptor Responsible for Excitatory CNS Synaptogenesis. Cell.
https://doi.org/10.1016/j.cell.2009.09.025
Eroglu, C., & Barres, B. A. (2010). Regulation of synaptic connectivity by glia. Nature,
468(7321), 223–231. https://doi.org/10.1038/nature09612
Erta, M., Quintana, A., & Hidalgo, J. (2012). Interleukin-6, a major cytokine in the central
nervous system. In International Journal of Biological Sciences (Vol. 8, Issue 9, pp.
1254–1266). https://doi.org/10.7150/ijbs.4679
Ferguson, S. M., Norton, C. S., Watson, S. J., Akil, H., & Robinson, T. E. (2004).
Amphetamine-evoked c-fos mRNA expression in the caudate-putamen: the effects
of DA and NMDA receptor antagonists vary as a function of neuronal phenotype
and environmental context. Journal of Neurochemistry, 86(1), 33–44.
https://doi.org/10.1046/j.1471-4159.2003.01815.x
Ferreira, T. A., Blackman, A. V., Oyrer, J., Jayabal, S., Chung, A. J., Watt, A. J., Sjöström,
P. J., & Van Meyel, D. J. (2014). Neuronal morphometry directly from bitmap
images. In Nature Methods (Vol. 11, Issue 10, pp. 982–984).
https://doi.org/10.1038/nmeth.3125
Fischer, C. P. (2006). Interleukin-6 in acute exercise and training: What is the biological
relevance? In Exercise Immunology Review (Vol. 12, pp. 6–33).
https://doi.org/10.1016/j.metabol.2015.10.028
Fisher, B. E., Petzinger, G. M., Nixon, K., Hogg, E., Bremmer, S., Meshul, C. K., & Jakowec,
M. W. (2004). Exercise-induced behavioral recovery and neuroplasticity in the 1-
178
methyl-4-phenyl-1,2,3,6-tetrahydropyridine-lesioned mouse basal ganglia. Journal of
Neuroscience Research, 77(3), 378–390. https://doi.org/10.1002/jnr.20162
Ford, C. P. (2014). The role of D2-autoreceptors in regulating dopamine neuron activity
and transmission. Neuroscience, 282, 13–22.
https://doi.org/10.1016/j.neuroscience.2014.01.025
Foster, N. N., Korobkova, L., Garcia, L., Gao, L., Becerra, M., Sherafat, Y., Peng, B., Li, X.,
Choi, J. H., Gou, L., Zingg, B., Azam, S., Lo, D., Khanjani, N., Zhang, B., Stanis, J.,
Bowman, I., Cotter, K., Cao, C., … Dong, H. (2020). The mouse cortico-basal ganglia-
thalamic network. In bioRxiv. https://doi.org/10.1101/2020.10.06.326876
Frederick, A. L., Yano, H., Trifilieff, P., Vishwasrao, H. D., Biezonski, D., Mészáros, J.,
Urizar, E., Sibley, D. R., Kellendonk, C., Sonntag, K. C., Graham, D. L., Colbran, R. J.,
Stanwood, G. D., & Javitch, J. A. (2015). Evidence against dopamine D1/D2 receptor
heteromers. Molecular Psychiatry, 20(11), 1373–1385.
https://doi.org/10.1038/mp.2014.166
Gerfen, C., Engber, T., Mahan, L., Susel, Z., Chase, T., Monsma, F., & Sibley, D. (1990). D1
and D2 dopamine receptor-regulated gene expression of striatonigral and
striatopallidal neurons. Science, 250(4986), 1429–1432.
https://doi.org/10.1126/science.2147780
Ghosh, K. K., Burns, L. D., Cocker, E. D., Nimmerjahn, A., Ziv, Y., Gamal, A. El, & Schnitzer,
M. J. (2011). Miniaturized integration of a fluorescence microscope. Nature
Methods, 8(10). https://doi.org/10.1038/nmeth.1694
Gibbs, M. E., Anderson, D. G., & Hertz, L. (2006). Inhibition of glycogenolysis in
astrocytes interrupts memory consolidation in young chickens. GLIA, 54(3), 214–
179
222. https://doi.org/10.1002/glia.20377
Girasole, A. E., Lum, M. Y., Nathaniel, D., Bair-Marshall, C. J., Guenthner, C. J., Luo, L.,
Kreitzer, A. C., & Nelson, A. B. (2018). A Subpopulation of Striatal Neurons Mediates
Levodopa-Induced Dyskinesia. Neuron, 97(4), 787-795.e6.
https://doi.org/10.1016/j.neuron.2018.01.017
Giros, B., Jaber, M., Jones, S. R., Wightman, R. M., & Caron, M. G. (1996).
Hyperlocomotion and indifference to cocaine and amphetamine in mice lacking the
dopamine transporter. Nature, 379(6566), 606–612.
https://doi.org/10.1038/379606a0
Golden, J. P., DeMaro, J. A., Knoten, A., Hoshi, M., Pehek, E., Johnson, E. M., Gereau, R.
W., & Jain, S. (2013). Dopamine-Dependent Compensation Maintains Motor
Behavior in Mice with Developmental Ablation of Dopaminergic Neurons. The
Journal of Neuroscience, 33(43), 17095–17107.
https://doi.org/10.1523/JNEUROSCI.0890-13.2013
Gorczynski, P., & Faulkner, G. (2010). Exercise therapy for schizophrenia. Cochrane
Database of Systematic Reviews.
https://doi.org/10.1002/14651858.CD004412.pub2
Gourine, A. V., Kasymov, V., Marina, N., Tang, F., Figueiredo, M. F., Lane, S.,
Teschemacher, A. G., Spyer, K. M., Deisseroth, K., & Kasparov, S. (2010). Astrocytes
Control Breathing Through pH-Dependent Release of ATP. Science, 329(5991),
571–575. https://doi.org/10.1126/science.1190721
Goyal, M. S., Vlassenko, A. G., Blazey, T. M., Su, Y., Couture, L. E., Durbin, T. J., Bateman,
R. J., Benzinger, T. L. S., Morris, J. C., & Raichle, M. E. (2017). Loss of Brain Aerobic
180
Glycolysis in Normal Human Aging. Cell Metabolism, 26(2), 353-360.e3.
https://doi.org/10.1016/j.cmet.2017.07.010
Graybiel, A. M. (2000). The basal ganglia. Current Biology, 10(14), R509–R511.
https://doi.org/10.1016/S0960-9822(00)00593-5
Graybiel, A. M. (2008). Habits, Rituals, and the Evaluative Brain. Annual Review of
Neuroscience, 31(1), 359–387.
https://doi.org/10.1146/annurev.neuro.29.051605.112851
Graybiel, A. M., & Grafton, S. T. (2015). The Striatum: Where Skills and Habits Meet. Cold
Spring Harbor Perspectives in Biology, 7(8), a021691.
https://doi.org/10.1101/cshperspect.a021691
Gruol, D. L., & Nelson, T. E. (1997). Physiological and pathological roles of interleukin-6
in the central nervous system. Molecular Neurobiology, 15(3), 307–339.
https://doi.org/10.1007/BF02740665
Guo, Y., Wang, Z., Prathap, S., & Holschneider, D. P. (2017). Recruitment of prefrontal-
striatal circuit in response to skilled motor challenge. NeuroReport, 28(18), 1187–
1194. https://doi.org/10.1097/WNR.0000000000000881
Hallett, P. J., Spoelgen, R., Hyman, B. T., Standaert, D. G., & Dunah, A. W. (2006).
Dopamine D1 activation potentiates striatal NMDA receptors by tyrosine
phosphorylation-dependent subunit trafficking. The Journal of Neuroscience,
26(17), 4690–4700. https://doi.org/10.1523/JNEUROSCI.0792-06.2006
Halliday, M. R., Abeydeera, D., Lundquist, A. J., Petzinger, G. M., & Jakowec, M. W.
(2019). Intensive treadmill exercise increases expression of hypoxia-inducible
factor 1α and its downstream transcript targets. NeuroReport.
181
https://doi.org/10.1097/wnr.0000000000001239
Hamilton, N. B., & Attwell, D. (2010). Do astrocytes really exocytose neurotransmitters?
Nature Reviews Neuroscience, 11(4), 227–238. https://doi.org/10.1038/nrn2803
Harris, J. J., Jolivet, R., & Attwell, D. (2012). Synaptic Energy Use and Supply. Neuron,
75(5), 762–777. https://doi.org/10.1016/j.neuron.2012.08.019
Harris, R. A., Lone, A., Lim, H., Martinez, F., Frame, A. K., Scholl, T. J., & Cumming, R. C.
(2019). Aerobic Glycolysis Is Required for Spatial Memory Acquisition But Not
Memory Retrieval in Mice. Eneuro, 6(1), ENEURO.0389-18.2019.
https://doi.org/10.1523/ENEURO.0389-18.2019
Hasel, P., Dando, O., Jiwaji, Z., Baxter, P., Todd, A. C., Heron, S., Márkus, N. M., McQueen,
J., Hampton, D. W., Torvell, M., Tiwari, S. S., McKay, S., Eraso-Pichot, A., Zorzano, A.,
Masgrau, R., Galea, E., Chandran, S., Wyllie, D. J. A., Simpson, T. I., & Hardingham, G.
E. (2017). Neurons and neuronal activity control gene expression in astrocytes to
regulate their development and metabolism. Nature Communications, 8(1), 15132.
https://doi.org/10.1038/ncomms15132
Herrera-López, G., Griego, E., & Galván, E. J. (2020). Lactate induces synapse-specific
potentiation on CA3 pyramidal cells of rat hippocampus. PLoS ONE, 15(11
November), 1–22. https://doi.org/10.1371/journal.pone.0242309
Herrero-Navarro, Á., Puche-Aroca, L., Moreno-Juan, V., Sempere-Ferràndez, A., Espinosa,
A., Susín, R., Torres-Masjoan, L., Leyva-Díaz, E., Karow, M., Figueres-Oñate, M.,
López-Mascaraque, L., López-Atalaya, J. P., Berninger, B., & López-Bendito, G.
(2021). Astrocytes and neurons share region-specific transcriptional signatures
that confer regional identity to neuronal reprogramming. Science Advances, 7(15).
182
https://doi.org/10.1126/sciadv.abe8978
Hintiryan, H., Foster, N. N., Bowman, I., Bay, M., Song, M. Y., Gou, L., Yamashita, S.,
Bienkowski, M. S., Zingg, B., Zhu, M., Yang, X. W., Shih, J. C., Toga, A. W., & Dong, H.-
W. (2016). The mouse cortico-striatal projectome. Nature Neuroscience, 19(8),
1100–1114. https://doi.org/10.1038/nn.4332
Hong, S., Dissing-Olesen, L., & Stevens, B. (2016). New insights on the role of microglia
in synaptic pruning in health and disease. Current Opinion in Neurobiology, 36.
https://doi.org/10.1016/j.conb.2015.12.004
Hosp, J. A., Mann, S., Wegenast-Braun, B. M., Calhoun, M. E., & Luft, A. R. (2013). Region
and task-specific activation of Arc in primary motor cortex of rats following motor
skill learning. Neuroscience, 250.
https://doi.org/10.1016/j.neuroscience.2013.06.060
Hotte, M., Thuault, S., Lachaise, F., Dineley, K. T., Hemmings, H. C., Nairn, A. C., & Jay, T.
M. (2006). D1 receptor modulation of memory retrieval performance is associated
with changes in pCREB and pDARPP-32 in rat prefrontal cortex. Behavioural Brain
Research, 171(1), 127–133. https://doi.org/10.1016/j.bbr.2006.03.026
Hui, S., Ghergurovich, J. M., Morscher, R. J., Jang, C., Teng, X., Lu, W., Esparza, L. A.,
Reya, T., Zhan, L., Yanxiang Guo, J., White, E., & Rabinowitz, J. D. (2017). Glucose
feeds the TCA cycle via circulating lactate. Nature.
https://doi.org/10.1038/nature24057
Hunnicutt, B. J., Jongbloets, B. C., Birdsong, W. T., Gertz, K. J., Zhong, H., & Mao, T.
(2016). A comprehensive excitatory input map of the striatum reveals novel
functional organization. ELife, 5. https://doi.org/10.7554/eLife.19103
183
Ide, K., Schmalbruch, I. K., Quistorff, B., Horn, A., & Secher, N. H. (2000). Lactate, glucose
and O 2 uptake in human brain during recovery from maximal exercise. The Journal
of Physiology, 522(1), 159–164. https://doi.org/10.1111/j.1469-7793.2000.t01-2-
00159.xm
Izumi, Y., Benz, A. M., Katsuki, H., & Zorumski, C. F. (1997). Endogenous
monocarboxylates sustain hippocampal synaptic function and morphological
integrity during energy deprivation. Journal of Neuroscience, 17(24), 9448–9457.
https://doi.org/10.1523/jneurosci.17-24-09448.1997
Jouroukhin, Y., Kageyama, Y., Misheneva, V., Shevelkin, A., Andrabi, S., Prandovszky, E.,
Yolken, R. H., Dawson, V. L., Dawson, T. M., Aja, S., Sesaki, H., & Pletnikov, M. V.
(2018). DISC1 regulates lactate metabolism in astrocytes: implications for
psychiatric disorders. Translational Psychiatry, 8(1), 76.
https://doi.org/10.1038/s41398-018-0123-9
Juel, C., & Halestrap, A. P. (1999). Lactate transport in skeletal muscle - role and
regulation of the monocarboxylate transporter. In The Journal of physiology: Vol.
517 ( Pt 3 (pp. 633–642). https://doi.org/10.1177/088307389901400801
Kalueff, A. V., Stewart, A. M., Song, C., Berridge, K. C., Graybiel, A. M., & Fentress, J. C.
(2016). Neurobiology of rodent self-grooming and its value for translational
neuroscience. Nature Reviews Neuroscience, 17(1), 45–59.
https://doi.org/10.1038/nrn.2015.8
Kamphuis, W., Mamber, C., Moeton, M., Kooijman, L., Sluijs, J. A., Jansen, A. H. P.,
Verveer, M., de Groot, L. R., Smith, V. D., Rangarajan, S., Rodríguez, J. J., Orre, M., &
Hol, E. M. (2012). GFAP isoforms in adult mouse brain with a focus on neurogenic
184
astrocytes and reactive astrogliosis in mouse models of Alzheimer disease. PLoS
ONE. https://doi.org/10.1371/journal.pone.0042823
Karagiannis, A., Sylantyev, S., Hadjihambi, A., Hosford, P. S., Kasparov, S., & Gourine, A.
V. (2016). Hemichannel-mediated release of lactate. Journal of Cerebral Blood Flow
and Metabolism, 36(7), 1202–1211. https://doi.org/10.1177/0271678X15611912
Kawai, R., Markman, T., Poddar, R., Ko, R., Fantana, A. L., Dhawale, A. K., Kampff, A. R., &
Ölveczky, B. P. (2015). Motor Cortex Is Required for Learning but Not for Executing
a Motor Skill. Neuron, 86(3), 800–812.
https://doi.org/10.1016/j.neuron.2015.03.024
Khakh, B. S., & Sofroniew, M. V. (2015). Diversity of astrocyte functions and phenotypes
in neural circuits. Nature Neuroscience, 18(7), 942–952.
https://doi.org/10.1038/nn.4043
Kinni, H., Guo, M., Ding, J. Y., Konakondla, S., Dornbos, D., Tran, R., Guthikonda, M., &
Ding, Y. (2011). Cerebral metabolism after forced or voluntary physical exercise.
Brain Research, 1388, 48–55. https://doi.org/10.1016/j.brainres.2011.02.076
Kintz, N., Petzinger, G. M., Akopian, G., Ptasnik, S., Williams, C., Jakowec, M. W., & Walsh,
J. P. (2013). Exercise modifies α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic
acid receptor expression in striatopallidal neurons in the 1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine-lesioned mouse. Journal of Neuroscience Research, 91(11),
1492–1507. https://doi.org/10.1002/jnr.23260
Kintz, Natalie, Petzinger, G. M., & Jakowec, M. W. (2017). Treadmill exercise modifies
dopamine receptor expression in the prefrontal cortex of the 1-methyl-4-phenyl-
1,2,3,6-tetrahydropyridine-lesioned mouse model of Parkinson’s disease.
185
NeuroReport, 28(15), 987–995. https://doi.org/10.1097/WNR.0000000000000865
Kol, A., Adamsky, A., Groysman, M., Kreisel, T., London, M., & Goshen, I. (2020).
Astrocytes contribute to remote memory formation by modulating hippocampal–
cortical communication during learning. Nature Neuroscience, 23(10), 1229–1239.
https://doi.org/10.1038/s41593-020-0679-6
Kovar, J. L., Volcheck, W., Sevick-Muraca, E., Simpson, M. A., & Olive, D. M. (2009).
Characterization and performance of a near-infrared 2-deoxyglucose optical
imaging agent for mouse cancer models. Analytical Biochemistry, 384(2), 254–262.
https://doi.org/10.1016/j.ab.2008.09.050
Kravitz, A. V., Freeze, B. S., Parker, P. R. L., Kay, K., Thwin, M. T., Deisseroth, K., &
Kreitzer, A. C. (2010). Regulation of parkinsonian motor behaviours by optogenetic
control of basal ganglia circuitry. Nature, 466(7306), 622–626.
https://doi.org/10.1038/nature09159
Kreitzer, A. C. (2009). Physiology and Pharmacology of Striatal Neurons. Annual Review
of Neuroscience, 32(1), 127–147.
https://doi.org/10.1146/annurev.neuro.051508.135422
Kumar, M. S., Pester, R. E., Chen, C. Y., Lane, K., Chin, C., Lu, J., Kirsch, D. G., Golub, T. R.,
& Jacks, T. (2009). Dicer1 functions as a haploinsufficient tumor suppressor. Genes
& Development, 23(23), 2700–2704. https://doi.org/10.1101/gad.1848209
Kumer, S. C., & Vrana, K. E. (1996). Intricate regulation of tyrosine hydroxylase activity
and gene expression. Journal of Neurochemistry, 67(2), 443–462.
https://doi.org/10.1046/j.1471-4159.1996.67020443.x
Lamprecht, R., & LeDoux, J. (2004). Structural plasticity and memory. Nature Reviews
186
Neuroscience, 5(1). https://doi.org/10.1038/nrn1301
Lanjakornsiripan, D., Pior, B.-J., Kawaguchi, D., Furutachi, S., Tahara, T., Katsuyama, Y.,
Suzuki, Y., Fukazawa, Y., & Gotoh, Y. (2018). Layer-specific morphological and
molecular differences in neocortical astrocytes and their dependence on neuronal
layers. Nature Communications, 9(1), 1623. https://doi.org/10.1038/s41467-018-
03940-3
Latham, T., Mackay, L., Sproul, D., Karim, M., Culley, J., Harrison, D. J., Hayward, L.,
Langridge-Smith, P., Gilbert, N., & Ramsahoye, B. H. (2012). Lactate, a product of
glycolytic metabolism, inhibits histone deacetylase activity and promotes changes
in gene expression. Nucleic Acids Research, 40(11), 4794–4803.
https://doi.org/10.1093/nar/gks066
Lauritzen, K. H., Morland, C., Puchades, M., Holm-Hansen, S., Hagelin, E. M., Lauritzen, F.,
Attramadal, H., Storm-Mathisen, J., Gjedde, A., & Bergersen, L. H. (2014). Lactate
Receptor Sites Link Neurotransmission, Neurovascular Coupling, and Brain Energy
Metabolism. Cerebral Cortex, 24(10), 2784–2795.
https://doi.org/10.1093/cercor/bht136
Lee, Y., Morrison, B. M., Li, Y., Lengacher, S., Farah, M. H., Hoffman, P. N., Liu, Y.,
Tsingalia, A., Jin, L., Zhang, P.-W., Pellerin, L., Magistretti, P. J., & Rothstein, J. D.
(2012). Oligodendroglia metabolically support axons and contribute to
neurodegeneration. Nature, 487(7408), 443–448.
https://doi.org/10.1038/nature11314
Leger, M., Quiedeville, A., Bouet, V., Haelewyn, B., Boulouard, M., Schumann-Bard, P., &
Freret, T. (2013). Object recognition test in mice. Nature Protocols, 8(12), 2531–
187
2537. https://doi.org/10.1038/nprot.2013.155
Li, J., Ding, Y. H., Rafols, J. A., Lai, Q., McAllister, J. P., & Ding, Y. (2005). Increased
astrocyte proliferation in rats after running exercise. Neuroscience Letters, 386(3),
160–164. https://doi.org/10.1016/j.neulet.2005.06.009
Li, Y.-K., Wang, F., Wang, W., Luo, Y., Wu, P.-F., Xiao, J.-L., Hu, Z.-L., Jin, Y., Hu, G., & Chen,
J.-G. (2012). Aquaporin-4 Deficiency Impairs Synaptic Plasticity and Associative
Fear Memory in the Lateral Amygdala: Involvement of Downregulation of Glutamate
Transporter-1 Expression. Neuropsychopharmacology, 37(8), 1867–1878.
https://doi.org/10.1038/npp.2012.34
Lindgren, N., Usiello, A., Goiny, M., Haycock, J., Erbs, E., Greengard, P., Hokfelt, T.,
Borrelli, E., & Fisone, G. (2003). Distinct roles of dopamine D2L and D2S receptor
isoforms in the regulation of protein phosphorylation at presynaptic and
postsynaptic sites. Proceedings of the National Academy of Sciences, 100(7),
4305–4309. https://doi.org/10.1073/pnas.0730708100
Lista, I., & Sorrentino, G. (2010). Biological Mechanisms of Physical Activity in
Preventing Cognitive Decline. Cellular and Molecular Neurobiology, 30(4), 493–503.
https://doi.org/10.1007/s10571-009-9488-x
Liu, C., Kuei, C., Zhu, J., Yu, J., Zhang, L., Shih, A., Mirzadegan, T., Shelton, J., Sutton, S.,
Connelly, M. A., Lee, G., Carruthers, N., Wu, J., & Lovenberg, T. W. (2012). 3,5-
Dihydroxybenzoic Acid, a Specific Agonist for Hydroxycarboxylic Acid 1, Inhibits
Lipolysis in Adipocytes. Journal of Pharmacology and Experimental Therapeutics.
https://doi.org/10.1124/jpet.112.192799
Liu, L., MacKenzie, K. R., Putluri, N., Maletić-Savatić, M., & Bellen, H. J. (2017). The Glia-
188
Neuron Lactate Shuttle and Elevated ROS Promote Lipid Synthesis in Neurons and
Lipid Droplet Accumulation in Glia via APOE/D. Cell Metabolism, 26(5), 719-737.e6.
https://doi.org/10.1016/j.cmet.2017.08.024
Livak, K. J., & Schmittgen, T. D. (2001). Analysis of Relative Gene Expression Data Using
Real-Time Quantitative PCR and the 2−ΔΔCT Method. Methods, 25(4), 402–408.
https://doi.org/10.1006/meth.2001.1262
Llewellyn, G. N., Seclén, E., Wietgrefe, S., Liu, S., Chateau, M., Pei, H., Perkey, K., Marsden,
M. D., Hinkley, S. J., Paschon, D. E., Holmes, M. C., Zack, J. A., Louie, S. G., Haase, A.
T., & Cannon, P. M. (2019). Humanized Mouse Model of HIV-1 Latency with
Enrichment of Latent Virus in PD-1 + and TIGIT + CD4 T Cells. Journal of Virology,
93(10). https://doi.org/10.1128/JVI.02086-18
Longair, M. H., Baker, D. A., & Armstrong, J. D. (2011). Simple neurite tracer: Open
source software for reconstruction, visualization and analysis of neuronal
processes. Bioinformatics. https://doi.org/10.1093/bioinformatics/btr390
Lu, B. (2003). BDNF and Activity-Dependent Synaptic Modulation. Learning & Memory,
10(2), 86–98. https://doi.org/10.1101/lm.54603
Lundgaard, I., Li, B., Xie, L., Kang, H., Sanggaard, S., Haswell, J. D. R., Sun, W., Goldman,
S., Blekot, S., Nielsen, M., Takano, T., Deane, R., & Nedergaard, M. (2015). Direct
neuronal glucose uptake heralds activity-dependent increases in cerebral
metabolism. Nature Communications, 6(1), 6807.
https://doi.org/10.1038/ncomms7807
Lundquist, A. J., Gallagher, T. J., Petzinger, G. M., & Jakowec, M. W. (2021). Exogenous
l ‐lactate promotes astrocyte plasticity but is not sufficient for
189
enhancing striatal synaptogenesis or motor behavior in mice. Journal of
Neuroscience Research, 99(5). https://doi.org/10.1002/jnr.24804
Lundquist, A. J., Parizher, J., Petzinger, G. M., & Jakowec, M. W. (2019). Exercise
induces region‐specific remodeling of astrocyte morphology and reactive astrocyte
gene expression patterns in male mice. Journal of Neuroscience Research, 97(9),
1081–1094. https://doi.org/10.1002/jnr.24430
Mächler, P., Wyss, M. T., Elsayed, M., Stobart, J., Gutierrez, R., Von Faber-Castell, A.,
Kaelin, V., Zuend, M., San Martín, A., Romero-Gómez, I., Baeza-Lehnert, F.,
Lengacher, S., Schneider, B. L., Aebischer, P., Magistretti, P. J., Barros, L. F., &
Weber, B. (2016). In Vivo Evidence for a Lactate Gradient from Astrocytes to
Neurons. Cell Metabolism, 23(1), 94–102.
https://doi.org/10.1016/j.cmet.2015.10.010
Magistretti, P. J., & Allaman, I. (2018). Lactate in the brain: from metabolic end-product
to signalling molecule. Nature Reviews Neuroscience, 19(4), 235–249.
https://doi.org/10.1038/nrn.2018.19
Manders, E. M. M., Verbeek, F. J., & Aten, J. A. (1993). Measurement of co‐localization
of objects in dual‐colour confocal images. Journal of Microscopy.
https://doi.org/10.1111/j.1365-2818.1993.tb03313.x
Margineanu, M. B., Mahmood, H., Fiumelli, H., & Magistretti, P. J. (2018). L-Lactate
Regulates the Expression of Synaptic Plasticity and Neuroprotection Genes in
Cortical Neurons: A Transcriptome Analysis. Frontiers in Molecular Neuroscience,
11. https://doi.org/10.3389/fnmol.2018.00375
Martín, R., Bajo-Grañeras, R., Moratalla, R., Perea, G., & Araque, A. (2015). Circuit-specific
190
signaling in astrocyte-neuron networks in basal ganglia pathways. Science,
349(6249), 730–734. https://doi.org/10.1126/science.aaa7945
Matsui, T., Omuro, H., Liu, Y.-F., Soya, M., Shima, T., McEwen, B. S., & Soya, H. (2017).
Astrocytic glycogen-derived lactate fuels the brain during exhaustive exercise to
maintain endurance capacity. Proceedings of the National Academy of Sciences,
114(24), 6358–6363. https://doi.org/10.1073/pnas.1702739114
Meshul, C. K., Cogen, J. P., Cheng, H.-W., Moore, C., Krentz, L., & McNeill, T. H. (2000).
Alterations in Rat Striatal Glutamate Synapses Following a Lesion of the Cortico-
and/or Nigrostriatal Pathway. Experimental Neurology, 165(1), 191–206.
https://doi.org/10.1006/exnr.2000.7467
Molteni, R., Ying, Z., & Gómez-Pinilla, F. (2002). Differential effects of acute and chronic
exercise on plasticity-related genes in the rat hippocampus revealed by microarray.
European Journal of Neuroscience. https://doi.org/10.1046/j.1460-
9568.2002.02158.x
Morel, L., Chiang, M. S. R., Higashimori, H., Shoneye, T., Iyer, L. K., Yelick, J., Tai, A., &
Yang, Y. (2017). Molecular and Functional Properties of Regional Astrocytes in the
Adult Brain. The Journal of Neuroscience, 37(36), 8706–8717.
https://doi.org/10.1523/JNEUROSCI.3956-16.2017
Morland, C., Andersson, K. A., Haugen, Ø. P., Hadzic, A., Kleppa, L., Gille, A., Rinholm, J.
E., Palibrk, V., Diget, E. H., Kennedy, L. H., Stølen, T., Hennestad, E., Moldestad, O.,
Cai, Y., Puchades, M., Offermanns, S., Vervaeke, K., Bjørås, M., Wisløff, U., …
Bergersen, L. H. (2017). Exercise induces cerebral VEGF and angiogenesis via the
lactate receptor HCAR1. Nature Communications, 8(1), 15557.
191
https://doi.org/10.1038/ncomms15557
Morland, C., Lauritzen, K. H., Puchades, M., Holm-Hansen, S., Andersson, K., Gjedde, A.,
Attramadal, H., Storm-Mathisen, J., & Bergersen, L. H. (2015). The lactate receptor,
G-protein-coupled receptor 81/hydroxycarboxylic acid receptor 1: Expression and
action in brain. Journal of Neuroscience Research, 93(7), 1045–1055.
https://doi.org/10.1002/jnr.23593
Murphy-Royal, C., Johnston, A. D., Boyce, A. K. J., Diaz-Castro, B., Institoris, A., Peringod,
G., Zhang, O., Stout, R. F., Spray, D. C., Thompson, R. J., Khakh, B. S., Bains, J. S., &
Gordon, G. R. (2020). Stress gates an astrocytic energy reservoir to impair synaptic
plasticity. Nature Communications, 11(1), 2014. https://doi.org/10.1038/s41467-
020-15778-9
Nectow, A. R., Moya, M. V., Ekstrand, M. I., Mousa, A., McGuire, K. L., Sferrazza, C. E.,
Field, B. C., Rabinowitz, G. S., Sawicka, K., Liang, Y., Friedman, J. M., Heintz, N., &
Schmidt, E. F. (2017). Rapid Molecular Profiling of Defined Cell Types Using Viral
TRAP. Cell Reports, 19(3). https://doi.org/10.1016/j.celrep.2017.03.048
Netzahualcoyotzi, C., & Pellerin, L. (2020). Neuronal and astroglial monocarboxylate
transporters play key but distinct roles in hippocampus-dependent learning and
memory formation. Progress in Neurobiology, 194, 101888.
https://doi.org/10.1016/j.pneurobio.2020.101888
Palmiter, R. D. (2008). Dopamine Signaling in the Dorsal Striatum Is Essential for
Motivated Behaviors. Annals of the New York Academy of Sciences, 1129(1), 35–46.
https://doi.org/10.1196/annals.1417.003
Panatier, A., Theodosis, D. T., Mothet, J.-P., Touquet, B., Pollegioni, L., Poulain, D. A., &
192
Oliet, S. H. R. (2006). Glia-Derived d-Serine Controls NMDA Receptor Activity and
Synaptic Memory. Cell, 125(4), 775–784. https://doi.org/10.1016/j.cell.2006.02.051
Pellerin, L., & Magistretti, P. J. (1994). Glutamate uptake into astrocytes stimulates
aerobic glycolysis: a mechanism coupling neuronal activity to glucose utilization.
Proceedings of the National Academy of Sciences, 91(22), 10625–10629.
https://doi.org/10.1073/pnas.91.22.10625
Pellerin, Luc, Halestrap, A. P., & Pierre, K. (2005). Cellular and subcellular distribution of
monocarboxylate transporters in cultured brain cells and in the adult brain. Journal
of Neuroscience Research, 79(1–2), 55–64. https://doi.org/10.1002/jnr.20307
Pellerin, Luc, Pellegri, G., Bittar, P. G., Charnay, Y., Bouras, C., Martin, J. L., Stella, N., &
Magistretti, P. J. (1998). Evidence supporting the existence of an activity-dependent
astrocyte-neuron lactate shuttle. Developmental Neuroscience.
https://doi.org/10.1159/000017324
Pennington, Z. T., Dong, Z., Feng, Y., Vetere, L. M., Page-Harley, L., Shuman, T., & Cai, D.
J. (2019). ezTrack: An open-source video analysis pipeline for the investigation of
animal behavior. Scientific Reports, 9(1), 19979. https://doi.org/10.1038/s41598-
019-56408-9
Perea, G., Yang, A., Boyden, E. S., & Sur, M. (2014). Optogenetic astrocyte activation
modulates response selectivity of visual cortex neurons in vivo. Nature
Communications, 5(1), 3262. https://doi.org/10.1038/ncomms4262
Perego, C., Vanoni, C., Bossi, M., Massari, S., Basudev, H., Longhi, R., & Pietrini, G. (2000).
The GLT-1 and GLAST glutamate transporters are expressed on morphologically
distinct astrocytes and regulated by neuronal activity in primary hippocampal
193
cocultures. Journal of Neurochemistry. https://doi.org/10.1046/j.1471-
4159.2000.0751076.x
Perez, X. A., Parameswaran, N., Huang, L. Z., O’Leary, K. T., & Quik, M. (2008). Pre-
synaptic dopaminergic compensation after moderate nigrostriatal damage in non-
human primates. Journal of Neurochemistry, 105(5), 1861–1872.
https://doi.org/10.1111/j.1471-4159.2008.05268.x
Peters, A. J., Liu, H., & Komiyama, T. (2017). Learning in the Rodent Motor Cortex.
Annual Review of Neuroscience, 40(1), 77–97. https://doi.org/10.1146/annurev-
neuro-072116-031407
Petzinger, G. M., Walsh, J. P., Akopian, G., Hogg, E., Abernathy, A., Arevalo, P., Turnquist,
P., Vuckovic, M., Fisher, B. E., Togasaki, D. M., & Jakowec, M. W. (2007). Effects of
Treadmill Exercise on Dopaminergic Transmission in the 1-Methyl-4-Phenyl-1,2,3,6-
Tetrahydropyridine-Lesioned Mouse Model of Basal Ganglia Injury. Journal of
Neuroscience, 27(20), 5291–5300. https://doi.org/10.1523/JNEUROSCI.1069-
07.2007
Petzinger, G.M., Holschneider, D. P., Fisher, B. E., McEwen, S., Kintz, N., Halliday, M., Toy,
W., Walsh, J. W., Beeler, J., & Jakowec, M. W. (2015). The Effects of Exercise on
Dopamine Neurotransmission in Parkinson’s Disease: Targeting Neuroplasticity to
Modulate Basal Ganglia Circuitry. Brain Plasticity, 1(1), 29–39.
https://doi.org/10.3233/BPL-150021
Petzinger, Giselle M., Fisher, B. E., McEwen, S., Beeler, J. A., Walsh, J. P., & Jakowec, M.
W. (2013). Exercise-enhanced neuroplasticity targeting motor and cognitive
circuitry in Parkinson’s disease. The Lancet Neurology, 12(7), 716–726.
194
https://doi.org/10.1016/S1474-4422(13)70123-6
Pierre, K., Magistretti, P. J., & Pellerin, L. (2002). MCT2 is a Major Neuronal
Monocarboxylate Transporter in the Adult Mouse Brain. Journal of Cerebral Blood
Flow & Metabolism, 22(5), 586–595. https://doi.org/10.1097/00004647-
200205000-00010
Polyzos, A. A., Lee, D. Y., Datta, R., Hauser, M., Budworth, H., Holt, A., Mihalik, S.,
Goldschmidt, P., Frankel, K., Trego, K., Bennett, M. J., Vockley, J., Xu, K., Gratton, E.,
& McMurray, C. T. (2019). Metabolic Reprogramming in Astrocytes Distinguishes
Region-Specific Neuronal Susceptibility in Huntington Mice. Cell Metabolism, 29(6),
1258-1273.e11. https://doi.org/10.1016/j.cmet.2019.03.004
Quistorff, B., Secher, N. H., & Van Lieshout, J. J. (2008). Lactate fuels the human brain
during exercise. The FASEB Journal. https://doi.org/10.1096/fj.08-106104
Rasmussen, P., Vedel, J. C., Olesen, J., Adser, H., Pedersen, M. V., Hart, E., Secher, N. H.,
& Pilegaard, H. (2011). In humans IL-6 is released from the brain during and after
exercise and paralleled by enhanced IL-6 mRNA expression in the hippocampus of
mice. Acta Physiologica. https://doi.org/10.1111/j.1748-1716.2010.02223.x
Rasmussen, Peter, Wyss, M. T., & Lundby, C. (2011). Cerebral glucose and lactate
consumption during cerebral activation by physical activity in humans. The FASEB
Journal : Official Publication of the Federation of American Societies for
Experimental Biology, 25(9), 2865–2873. https://doi.org/10.1096/fj.11-183822
Roberts, B. M., Doig, N. M., Brimblecombe, K. R., Lopes, E. F., Siddorn, R. E., Threlfell, S.,
Connor-Robson, N., Bengoa-Vergniory, N., Pasternack, N., Wade-Martins, R., Magill,
P. J., & Cragg, S. J. (2020). GABA uptake transporters support dopamine release in
195
dorsal striatum with maladaptive downregulation in a parkinsonism model. Nature
Communications, 11(1), 4958. https://doi.org/10.1038/s41467-020-18247-5
Robinson, T. E., & Becker, J. B. (1983). The rotational behavior model: asymmetry in the
effects of unilateral 6-OHDA lesions of the substantia nigra in rats. Brain Research,
264(1), 127–131. https://doi.org/10.1016/0006-8993(83)91129-0
Roth, B. L. (2016). DREADDs for Neuroscientists. Neuron, 89(4).
https://doi.org/10.1016/j.neuron.2016.01.040
Rothstein, J. D., Dykes-Hoberg, M., Pardo, C. A., Bristol, L. A., Jin, L., Kuncl, R. W., Kanai,
Y., Hediger, M. A., Wang, Y., Schielke, J. P., & Welty, D. F. (1996). Knockout of
glutamate transporters reveals a major role for astroglial transport in excitotoxicity
and clearance of glutamate. Neuron. https://doi.org/10.1016/S0896-
6273(00)80086-0
Rothwell, P. E., Fuccillo, M. V., Maxeiner, S., Hayton, S. J., Gokce, O., Lim, B. K., Fowler, S.
C., Malenka, R. C., & Südhof, T. C. (2014). Autism-Associated Neuroligin-3
Mutations Commonly Impair Striatal Circuits to Boost Repetitive Behaviors. Cell,
158(1), 198–212. https://doi.org/10.1016/j.cell.2014.04.045
Rozovsky, I., Wei, M., Stone, D. J., Zanjani, H., Anderson, C. P., Morgan, T. E., & Finch, C.
E. (2002). Estradiol (E2) Enhances Neurite Outgrowth by Repressing Glial Fibrillary
Acidic Protein Expression and Reorganizing Laminin. Endocrinology, 143(2), 636–
646. https://doi.org/10.1210/endo.143.2.8615
Santello, M., Toni, N., & Volterra, A. (2019). Astrocyte function from information
processing to cognition and cognitive impairment. Nature Neuroscience, 22(2),
154–166. https://doi.org/10.1038/s41593-018-0325-8
196
Sathyamurthy, A., Barik, A., Dobrott, C. I., Matson, K. J. E., Stoica, S., Pursley, R., Chesler,
A. T., & Levine, A. J. (2020). Cerebellospinal Neurons Regulate Motor Performance
and Motor Learning. Cell Reports, 31(6), 107595.
https://doi.org/10.1016/j.celrep.2020.107595
Saur, L., Baptista, P. P. A., De Senna, P. N., Paim, M. F., Nascimento, P. Do, Ilha, J.,
Bagatini, P. B., Achaval, M., & Xavier, L. L. (2014). Physical exercise increases GFAP
expression and induces morphological changes in hippocampal astrocytes. Brain
Structure and Function, 219(1), 293–302. https://doi.org/10.1007/s00429-012-
0500-8
Schafer, D. P., Lehrman, E. K., Kautzman, A. G., Koyama, R., Mardinly, A. R., Yamasaki, R.,
Ransohoff, R. M., Greenberg, M. E., Barres, B. A., & Stevens, B. (2012). Microglia
Sculpt Postnatal Neural Circuits in an Activity and Complement-Dependent Manner.
Neuron, 74(4). https://doi.org/10.1016/j.neuron.2012.03.026
Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T.,
Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., Tinevez, J.-Y., White, D. J.,
Hartenstein, V., Eliceiri, K., Tomancak, P., & Cardona, A. (2012). Fiji: an open-source
platform for biological-image analysis. Nature Methods, 9(7), 676–682.
https://doi.org/10.1038/nmeth.2019
Shaulian, E., & Karin, M. (2002). AP-1 as a regulator of cell life and death. Nature Cell
Biology, 4(5), E131–E136. https://doi.org/10.1038/ncb0502-e131
Sheng, Z.-H., & Cai, Q. (2012). Mitochondrial transport in neurons: impact on synaptic
homeostasis and neurodegeneration. Nature Reviews Neuroscience, 13(2), 77–93.
https://doi.org/10.1038/nrn3156
197
Sholl, D. A. (1953). Dendritic organization in the neurons of the visual and motor
cortices of the cat. Journal of Anatomy, 87, 387–406.
Smith, K. S., & Graybiel, A. M. (2013). A Dual Operator View of Habitual Behavior
Reflecting Cortical and Striatal Dynamics. Neuron, 79(2), 361–374.
https://doi.org/10.1016/j.neuron.2013.05.038
Sofroniew, M. V., & Vinters, H. V. (2010). Astrocytes: biology and pathology. Acta
Neuropathologica, 119(1), 7–35. https://doi.org/10.1007/s00401-009-0619-8
Sotelo-Hitschfeld, T., Niemeyer, M. I., Mächler, P., Ruminot, I., Lerchundi, R., Wyss, M. T.,
Stobart, J., Fernández-Moncada, I., Valdebenito, R., Garrido-Gerter, P., Contreras-
Baeza, Y., Schneider, B. L., Aebischer, P., Lengacher, S., Martín, A. S., Le Douce, J.,
Bonvento, G., Magistretti, P. J., Sepúlveda, F. V., … Felipe Barros, L. (2015). Channel-
mediated lactate release by K+ -stimulated astrocytes. Journal of Neuroscience,
35(10), 4168–4178. https://doi.org/10.1523/JNEUROSCI.5036-14.2015
Soya, H., Nakamura, T., Deocaris, C. C., Kimpara, A., Iimura, M., Fujikawa, T., Chang, H.,
McEwen, B. S., & Nishijima, T. (2007). BDNF induction with mild exercise in the rat
hippocampus. Biochemical and Biophysical Research Communications.
https://doi.org/10.1016/j.bbrc.2007.04.173
Srinivasan, R., Lu, T. Y., Chai, H., Xu, J., Huang, B. S., Golshani, P., Coppola, G., & Khakh,
B. S. (2016). New Transgenic Mouse Lines for Selectively Targeting Astrocytes and
Studying Calcium Signals in Astrocyte Processes In Situ and In Vivo. Neuron, 92(6),
1181–1195. https://doi.org/10.1016/j.neuron.2016.11.030
Stogsdill, J. A., Ramirez, J., Liu, D., Kim, Y. H., Baldwin, K. T., Enustun, E., Ejikeme, T., Ji,
R. R., & Eroglu, C. (2017). Astrocytic neuroligins control astrocyte morphogenesis
198
and synaptogenesis. Nature, 551(7679), 192–197.
https://doi.org/10.1038/nature24638
Sun, W., Cornwell, A., Li, J., Peng, S., Osorio, M. J., Aalling, N., Wang, S., Benraiss, A., Lou,
N., Goldman, S. A., & Nedergaard, M. (2017). SOX9 Is an Astrocyte-Specific Nuclear
Marker in the Adult Brain Outside the Neurogenic Regions. The Journal of
Neuroscience, 37(17), 4493–4507. https://doi.org/10.1523/JNEUROSCI.3199-
16.2017
Suzuki, A., Stern, S. A., Bozdagi, O., Huntley, G. W., Walker, R. H., Magistretti, P. J., &
Alberini, C. M. (2011). Astrocyte-neuron lactate transport is required for long-term
memory formation. Cell, 144(5), 810–823.
https://doi.org/10.1016/j.cell.2011.02.018
Tang, F., Lane, S., Korsak, A., Paton, J. F. R., Gourine, A. V., Kasparov, S., &
Teschemacher, A. G. (2014). Lactate-mediated glia-neuronal signalling in the
mammalian brain. Nature Communications, 5(1), 3284.
https://doi.org/10.1038/ncomms4284
Taylor, J. L., Rajbhandari, A. K., Berridge, K. C., & Aldridge, J. W. (2010). Dopamine
receptor modulation of repetitive grooming actions in the rat: Potential relevance
for Tourette syndrome. Brain Research, 1322, 92–101.
https://doi.org/10.1016/j.brainres.2010.01.052
The Tabula Muris Consortium. (2018). Single-cell transcriptomics of 20 mouse organs
creates a Tabula Muris. Nature, 562(7727). https://doi.org/10.1038/s41586-018-
0590-4
Toy, W. A., Petzinger, G. M., Leyshon, B. J., Akopian, G. K., Walsh, J. P., Hoffman, M. V.,
199
Vučković, M. G., & Jakowec, M. W. (2014). Treadmill exercise reverses dendritic
spine loss in direct and indirect striatal medium spiny neurons in the 1-methyl-4-
phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model of Parkinson’s disease.
Neurobiology of Disease, 63, 201–209. https://doi.org/10.1016/j.nbd.2013.11.017
van Hall, G., Strømstad, M., Rasmussen, P., Jans, O., Zaar, M., Gam, C., Quistorff, B.,
Secher, N. H., & Nielsen, H. B. (2009). Blood lactate is an important energy source
for the human brain. Journal of Cerebral Blood Flow and Metabolism : Official
Journal of the International Society of Cerebral Blood Flow and Metabolism, 29(6),
1121–1129. https://doi.org/10.1038/jcbfm.2009.35
van Praag, H. (2005). Exercise Enhances Learning and Hippocampal Neurogenesis in
Aged Mice. Journal of Neuroscience. https://doi.org/10.1523/JNEUROSCI.1731-
05.2005
van Praag, Henriette, Kempermann, G., & Gage, F. H. (1999). Running increases cell
proliferation and neurogenesis in the adult mouse dentate gyrus. Nature
Neuroscience, 2(3), 266–270. https://doi.org/10.1038/6368
Ventura, A., Meissner, A., Dillon, C. P., McManus, M., Sharp, P. A., Van Parijs, L.,
Jaenisch, R., & Jacks, T. (2004). Cre-lox-regulated conditional RNA interference
from transgenes. Proceedings of the National Academy of Sciences, 101(28),
10380–10385. https://doi.org/10.1073/pnas.0403954101
Verkhratsky, A., & Nedergaard, M. (2018). Physiology of Astroglia. Physiological
Reviews, 98(1), 239–389. https://doi.org/10.1152/physrev.00042.2016
Vezzoli, E., Calì, C., De Roo, M., Ponzoni, L., Sogne, E., Gagnon, N., Francolini, M., Braida,
D., Sala, M., Muller, D., Falqui, A., & Magistretti, P. J. (2020). Ultrastructural Evidence
200
for a Role of Astrocytes and Glycogen-Derived Lactate in Learning-Dependent
Synaptic Stabilization. Cerebral Cortex, 30(4), 2114–2127.
https://doi.org/10.1093/cercor/bhz226
Walter B. Cannon. (1932). The wisdom of the body (2nd ed.). Norton and Co.
Wang, J., Tu, J., Cao, B., Mu, L., Yang, X., Cong, M., Ramkrishnan, A. S., Chan, R. H. M.,
Wang, L., & Li, Y. (2017). Astrocytic l -Lactate Signaling Facilitates Amygdala-
Anterior Cingulate Cortex Synchrony and Decision Making in Rats. Cell Reports,
21(9), 2407–2418. https://doi.org/10.1016/j.celrep.2017.11.012
Wang, Z., Myers, K. G., Guo, Y., Ocampo, M. A., Pang, R. D., Jakowec, M. W., &
Holschneider, D. P. (2013). Functional Reorganization of Motor and Limbic Circuits
after Exercise Training in a Rat Model of Bilateral Parkinsonism. PLoS ONE, 8(11),
e80058. https://doi.org/10.1371/journal.pone.0080058
Wang, Z., Stefanko, D. P., Guo, Y., Toy, W. A., Petzinger, G. M., Jakowec, M. W., &
Holschneider, D. P. (2016). Evidence of functional brain reorganization on the basis
of blood flow changes in the CAG140 knock-in mouse model of Huntington’s
disease. NeuroReport, 27(9), 632–639.
https://doi.org/10.1097/WNR.0000000000000587
Wheeler, M. A., Clark, I. C., Tjon, E. C., Li, Z., Zandee, S. E. J., Couturier, C. P., Watson, B.
R., Scalisi, G., Alkwai, S., Rothhammer, V., Rotem, A., Heyman, J. A., Thaploo, S.,
Sanmarco, L. M., Ragoussis, J., Weitz, D. A., Petrecca, K., Moffitt, J. R., Becher, B., …
Quintana, F. J. (2020). MAFG-driven astrocytes promote CNS inflammation. Nature,
578(7796). https://doi.org/10.1038/s41586-020-1999-0
Wilhelmsson, U., Bushong, E. A., Price, D. L., Smarr, B. L., Phung, V., Terada, M., Ellisman,
201
M. H., & Pekny, M. (2006). Redefining the concept of reactive astrocytes as cells
that remain within their unique domains upon reaction to injury. Proceedings of the
National Academy of Sciences, 103(46), 17513–17518.
https://doi.org/10.1073/pnas.0602841103
Xu, T., Yu, X., Perlik, A. J., Tobin, W. F., Zweig, J. A., Tennant, K., Jones, T., & Zuo, Y.
(2009). Rapid formation and selective stabilization of synapses for enduring motor
memories. Nature, 462(7275), 915–919. https://doi.org/10.1038/nature08389
Yagishita, S., Hayashi-Takagi, A., Ellis-Davies, G. C. R., Urakubo, H., Ishii, S., & Kasai, H.
(2014). A critical time window for dopamine actions on the structural plasticity of
dendritic spines. Science, 345(6204), 1616–1620.
https://doi.org/10.1126/science.1255514
Yang, J., Ruchti, E., Petit, J.-M., Jourdain, P., Grenningloh, G., Allaman, I., & Magistretti, P.
J. (2014). Lactate promotes plasticity gene expression by potentiating NMDA
signaling in neurons. Proceedings of the National Academy of Sciences, 111(33),
12228–12233. https://doi.org/10.1073/pnas.1322912111
Yin, H. H., Mulcare, S. P., Hilário, M. R. F., Clouse, E., Holloway, T., Davis, M. I., Hansson,
A. C., Lovinger, D. M., & Costa, R. M. (2009). Dynamic reorganization of striatal
circuits during the acquisition and consolidation of a skill. Nature Neuroscience,
12(3), 333–341. https://doi.org/10.1038/nn.2261
Yu, X., Nagai, J., & Khakh, B. S. (2020). Improved tools to study astrocytes. Nature
Reviews Neuroscience, 21(3). https://doi.org/10.1038/s41583-020-0264-8
Zamanian, J. L., Xu, L., Foo, L. C., Nouri, N., Zhou, L., Giffard, R. G., & Barres, B. A. (2012).
Genomic Analysis of Reactive Astrogliosis. Journal of Neuroscience, 32(18), 6391–
202
6410. https://doi.org/10.1523/JNEUROSCI.6221-11.2012
Zeisel, A., Hochgerner, H., Lönnerberg, P., Johnsson, A., Memic, F., van der Zwan, J.,
Häring, M., Braun, E., Borm, L. E., La Manno, G., Codeluppi, S., Furlan, A., Lee, K.,
Skene, N., Harris, K. D., Hjerling-Leffler, J., Arenas, E., Ernfors, P., Marklund, U., &
Linnarsson, S. (2018). Molecular Architecture of the Mouse Nervous System. Cell,
174(4). https://doi.org/10.1016/j.cell.2018.06.021

Abstract (if available)

Abstract Cellular energetics underpin the function and dysfunction of all organisms, and the adaptation of energy systems is fundamental to the evolution and adaptation necessary for survival. Evolution and adaptation within the central nervous system is perhaps best characterized by neuroplasticity. The study of neuroplasticity is a theme our laboratory has been rigorously engaging with for the past two decades, especially in the context of neurodegenerative diseases such as Parkinson’s disease (Davies et al., 2017; Petzinger et al., 2013). ? Parkinson’s disease is the second most common neurodegenerative disease and is characterized by a selective loss of midbrain dopaminergic neurons, whose projections regulate motor and cognitive behaviors by the release of the neuromodulating chemical dopamine (Petzinger et al., 2015). In the absence of disease-modifying therapeutics, current therapies aim to limit dopamine metabolism and replace the lost dopaminergic tone throughout the brain to curb the severity of symptomology; however, these pharmacological approaches are well-documented to have severe drawbacks in long-term therapy and do not slow or reverse disease progression or pathological development (Girasole et al., 2018). In conjunction with pharmacological therapies, exercise training is a robust adjuvant capable of improving motor and cognitive function and may slow overall disease progression (Fisher et al., 2004; Petzinger et al., 2007). Exercise is well-known to affect several elements important to brain function, including through increasing neurogenesis, cerebral blood flow, and dendritic synaptogenesis (Black et al., 1990; Toy et al., 2014; van Praag et al., 1999). In addition, exercise can modify the function of non-neuronal cells, particularly astrocytes which are necessary supporters of neuronal health and help to regulate various aspects of neuroplasticity (Bernardi et al., 2013; Lundquist et al., 2019). Importantly, exercise is a whole-body experience requiring a coupling of peripheral systems (i.e., musculoskeletal, and cardiorespiratory) to central nervous system function and is executed by a careful coordination and collaboration by these distinct yet interconnected systems (Rasmussen et al., 2011). ? However, the exact mechanisms that govern the benefits of exercise, or how those molecular and cellular changes may contribute to the potentially disease-modifying application of exercise in the context of neurodegenerative diseases like Parkinson’s disease, remains incompletely known. One potential mechanism may be energetic changes, perhaps best represented by lactate metabolism. Lactate is a common byproduct of intensive aerobic exercise and is known to be produced by the periphery and consumed by the central nervous system, in addition to be produced within the brain by astrocytes during periods of high-intensity neuronal activity, such as during exercise (Dalsgaard et al., 2004; Matsui et al., 2017; van Hall et al., 2009). Thus, lactate potentially links peripheral and central activity during exercise, and can be used to study how astrocytes contribute to neuroplasticity. To address this gap in knowledge, the studies described in this dissertation represent a consolidated effort to isolate and examine the contribution of lactate, both peripherally and centrally derived, in driving neuroplasticity.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Experience-dependent neuroplasticity of the dorsal striatum and prefrontal cortex in the MPTP-lesioned mouse model of Parkinson’s disease

PDF

The effect of treadmill running on dendritic spine density in two neurodegenerative disorders: 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model of Parkinson’s disease and CAG₁₄₀ kn...

PDF

Long-term treadmill exercise delays the onset of motor dysfunction, cognitive impairments, and mood disturbances in the CAG 140 mouse model of Huntington's disease via restoration of dopamine neu...

PDF

Physical activity optimizes circuit-specific cellular metabolism in neuroplasticity: a role for hypoxia-inducible factor-1 and its downstream targets

PDF

Exploration and expansion of novel therapeutic strategies targeting brain metabolism in neurodegenerative diseases

PDF

The role of glial cells in spinal muscular atrophy

PDF

Genotype and sex effects on anxiety behavior and functional activation in corticolimbic circuits in serotonin transporter knockout mice (5-HITT KO)

PDF

The contribution of periascular spaces to Parkinson's disease pathophysiology

PDF

Computational investigation of glutamatergic synaptic dynamics: role of ionotropic receptor distribution and astrocytic modulation of neuronal spike timing

PDF

Energy metabolism and inflammation in brain aging: significance of age-dependent astrocyte metabolic-redox profile

PDF

Evaluating the effects of dihydromyricetin on dopamine mediated behaviors using a Parkinson's disease model

PDF

Gene-environment interactions in neurodevelopment

PDF

Longitudinal assessment of neural stem-cell aging

Asset Metadata

Creator Lundquist, Adam Jacob (author)

Core Title Lactate modulates astrocytic and neuronal plasticity

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Neuroscience

Degree Conferral Date 2021-08

Publication Date 07/22/2021

Defense Date 06/04/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag astrocyte,dendritic spines,Dopamine,Exercise,lactate,Learning and Instruction,monocarboxylate transporter,motor behavior,neuroplasticity,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Holschneider, Daniel (committee chair), Cadenas, Enrique (committee member), Jakowec, Michael (committee member), Petzinger, Giselle (committee member)

Creator Email alundqui@usc.edu,lundquistaj@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15616763

Unique identifier UC15616763

Legacy Identifier etd-LundquistA-9823

Document Type Dissertation

Format application/pdf (imt)

Rights Lundquist, Adam Jacob

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu