University of Southern California Dissertations and Theses

Determining the mechanism and necessity of matrix metalloproteinase 2 (MMP-2) nuclear localization for proficient skeletal muscle differentiation

(USC Thesis Other)

Determining the mechanism and necessity of matrix metalloproteinase 2 (MMP-2) nuclear localization for proficient skeletal muscle differentiation

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2022 Nimisha Mazumdar

Determining the mechanism and necessity of matrix metalloproteinase 2 (MMP-2) nuclear
localization for proficient skeletal muscle
differentiation

by

Nimisha Mazumdar

A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA

In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR MEDICINE)

August 2022
ii
Acknowledgements
First and foremost, I would like to thank the entire Rice lab for their support and
contributions to my academic pursuits. I want to thank Dr. Judd Rice, my principal investigator,
who has been a patient mentor, guiding me through each step of my thesis, and Benjamin Weekley
for helping me navigate the every-day challenges of pursuing research. Without their constant
encouragement to ask challenging questions and innovate novel approaches, I would not have been
able to overcome the struggles I faced. I would also like to thank a former Rice lab member, Vivian
Chen, for laying the groundwork for my project and being a helpful resource.

Next, I want to thank my thesis committee, Drs. Oliver Bell and Woojin An, for overseeing
the progress of my thesis and providing insightful advice. I also want to thank Emily Hsu who
helped me optimize protocols for certain Aims of my thesis.

Lastly, I would like to thank my friends and family for their support through this rigorous
curriculum.

iii
Table of contents

Acknowledgements …ii
List of Tables …iv
List of Figures …iv
List of Abbreviations …v
Abstract …vi
Chapter I: Introduction
Nucleosomes and histones
Histone tails: a site targeted for epigenetic regulation
Matrix Metalloproteinase-2: a novel protease required for H3NT proteolysis
…1
…1
…2
…4
Chapter II: Results
Aim 1: Identifying the minimal MMP-2 protein domain required for nuclear localization
Aim 2: Determining the components of the nuclear MMP-2 multiprotein complex
Aim 2A: Determining components of the native nuclear MMP-2 complex
Aim 2B: Identifying transient interactions facilitating nuclear MMP-2’s recruitment to
specific sites of the genome.
…8
…8
…18
…20
…25

Chapter III: Discussion
Is Fibronectin the minimal domain for MMP-2 nuclear localization?
MMP-2 nuclear interactions facilitating recruitment to sites of H3NT proteolysis
…32
…32
…35
Chapter IV: Materials and Methods
Construction and assembly of truncated constructs
Plasmid extraction
Transient transfection
Viral transduction
Purification of Nuclei
High salt nuclear soluble extraction
Chromatography
HA-Immunoprecipitation
Western analysis
Immunofluorescent staining
miniTurbo-ID and Streptavidin pulldown
…39
…39
…40
…40
…41
…42
…43
…43
…44
…44
…45
…46
References …48

iv
List of Tables

Table 1. List of candidates identified interacting with nuclear MMP-2 via Q XL
anion exchanger column chromatography and HA-immunoprecipitation.
…24
Table 2. List of candidates identified interacting with nuclear MMP-2 via miniTurbo-
ID biotinylation and mass spectrometry.
…31

List of Figures

Figure 1. Map representing specific sites of post-translational modifications on
histone tails.
…2
Figure 2. Structure of MMP-2. …5
Figure 3. Construction of truncated MMP-2 constructs. …9
Figure 4. Subcellular localization of truncated MMP-2 constructs. …11
Figure 5. Immunofluorescent (IF) staining of transiently transfected HEK293 and
U2OS cells.
…13
Figure 6. Subcellular localization of MMP2-3xHA truncated constructs in U2OS and
HEK293 stable cells.
…15
Figure 7. Immunofluorescent (IF) staining of stable U2OS cells at 63x magnification. …16
Figure 8. Experimental flowchart of approaches to identify nuclear MMP2-
interacting proteins.
…19
Figure 9. Enrichment of ProMMP2-3xHA from the Q XL anion exchanger at elution
with 0.6 M of NaCl.
…21
Figure 10. Enrichment of ProMMP2-3xHA following a 100 mM glycine pH 3.0
elution.
…23
Figure 11. Purification of nuclear MMP2-interacting proteins by Bio-ID proximity
labeling.
…27
Figure 12. Isolation of nuclear MMP2-interacting proteins following Streptavidin
bead pulldown.
…29
Figure 13. Volcano plot highlighting statistically significant candidates that were
upregulated in PromMMP2-miniTurbo nuclear extracts.
…30

v
List of abbreviations
MMP Matrix metalloprotease
H3NT Histone H3 N-terminal
PTM Post-translational modification
ECM Extracellular matrix

vi
Abstract

Previously, a canonical extracellular matrix protease, matrix metalloproteinase 2 (MMP-
2), was reported to unexpectedly function in the nucleus to cleave the N-terminal tail of the DNA-
associated histone H3 protein and is required for the activation of myogenic genes during skeletal
muscle differentiation. Although histone H3 N-terminal tail (H3NT) proteolysis within chromatin
was first observed over 60 years ago, the specific genomic sites selectively targeted for H3NT
proteolysis, and the biological significance of this evolutionarily conserved epigenetic
modification remain largely unknown.
In my thesis, I leverage established biochemical approaches to discover the minimal protein
domain required for MMP-2’s nuclear localization as well as to identify nuclear proteins
interacting with MMP-2. I aim to determine how these nuclear interactions facilitate H3NT
proteolysis and affect gene activation. To this end, I devise truncated constructs of the MMP-2
protein for the first aim and study their subcellular localization in different cell lines. For the
second aim, and to study the nuclear interactions recruiting MMP-2 to specific sites of the genome
concurrent with H3NT proteolysis, I adopt a Q XL anion column chromatography approach and a
more modern Bio-ID proximity labeling approach. The results of these two aims highlight nuclear
interactions that facilitate H3NT proteolysis and affect gene activation and most importantly,
ascertain the necessity of a non-nuclear protease for myogenic differentiation.

1
Chapter I
Introduction
Nucleosomes and histones
An organism’s genetic information is encoded by DNA and stored in the nucleus in a highly
condensed form known as chromatin. In the classic dogma, DNA is transcribed into RNA, which
translates to proteins that uniquely folds on itself to serve a plethora of structural and functional
needs of different cell types. To store such vast amounts of DNA, chromatin comprises of smaller
subunits, namely nucleosomes, which are sequences of genomic DNA wrapped around histone
proteins and held together by linker DNA. More specifically, 145-147 bp of DNA wrap around
histones that are made of 2-sets of 4 core proteins in a 1.7 left-handed super helical turn. The core
proteins that make up histones are highly basic proteins that include histone H2A, H2B, H3 and
H4. All these proteins have a N-terminal amino acid tail that protrudes from the nucleosome. H2A
histones have an additional C-terminus amino acid tail that also protrudes from the nucleosome.
Each of these tails ranges from 25-40 amino acids in length and help regulate chromatin structure.
DNA tightly wrapped around the core nucleosome is not actively transcribed and is also protected
from nuclease digestion. On the other hand, the linker DNA that is aimed at holding the structure
of a nucleosome, is rapidly digested (Cutter & Hayes, 2015).
The role of histone N-terminal tails has been widely studied with respect to nucleosome
structure and function. They have been reported to interact with the nuclear environment, other
histones, and linker DNA (Mutskov et al., 1998). Several residues are targeted sites for reversible
post-translational modifications (PTM) such as acetylation, methylation, and phosphorylation
catalyzed by different classes of enzymes. Figure 1 illustrates the residues that get post-
translationally modified. “Writer” enzymes are responsible for the addition of a PTM to a specific
2
residue, “eraser” enzymes are responsible for the removal of a PTM from a specific residue and
“reader” enzymes are responsible for interpreting these PTMs and translating them to a specific
function with respect to transcription (Keppler and Archer, 2008).

Figure 1. Map representing specific sites of post-translational modifications on histone tails
(Keppler and Archer, 2008).
Ac = acetylation, P = phosphorylation, me = methylation, Ub = ubiquitination.

Histone tails: a site targeted for epigenetic regulation
The identification of post-translational modifications on histone tails has led to a need to
investigate the functional significance of these PTMs on transcriptional activity and DNA
accessibility. Modifying certain residues, alters chromatin state and hence increases or decreases
transcriptional activity. For instance, the acetylation of histone H3 on lysine residues at position
9, 14, 18, and 23, by histone acetyltransferases (HATs) increases transcriptional activity. The
observation was further supported when a conserved transcriptional regulator Gen5 was identified
to possess HAT activity (Grant, 2001). Several such well-characterized modifications gave rise to
the term “histone code”, which comprises of PTMs on histone tails that act as a signal for the
recruitment or removal of certain transcriptional regulatory proteins. However, in addition to these
3
reversible PTMs, histone tails have been observed to undergo a more permanent modification,
termed “histone tail proteolysis”, which is the clipping or cleavage of these histone tails at a
specific residue by certain proteases.
First reported in 1959 in calf-thymus histones, histone tail proteolysis is a well-documented
process that is not fully elucidated. (Philips and Johns, 1959). In 1976, a protease associated with
calf-thymus chromatin was found to facilitate a clipping activity that resulted in the formation of
a truncated C-terminal from histone H2A. While the biological significance of the cleaved histone
H2A tail was not definitely identified, the protease facilitating the clipping activity was much later
identified to be a neutrophil-specific protease (Dhaenens et al., 2014). In 1993, another H2A
specific protease was identified in chicken liver and was believed to have tissue-specific
expression. Recently, the proteolysis of histone H3 N-terminal tails, also abbreviated as H3NT
proteolysis, has received wide traction. Observed across a multitude of cellular model systems,
H3NT proteolysis was first identified in Tetrahymena chromatin (Allis et al., 1985). However, it
was the discovery of Cathespin L that fueled investigation into proteases facilitating H3NT
proteolysis. Cathespin L is a protease that predominantly localizes to the lysosome and functions
to facilitate protein digestion in the extracellular matrix. In 2008, Cathespin L was identified to be
the primary protease facilitating histone H3 N-terminal tail clipping during mouse embryonic stem
cell differentiation (Duncan et al, 2008). Subsequently, Cathespin D was identified as the primary
protease facilitating H3NT proteolysis during mammary gland differentiation (Khalkhali-Ellis et
al., 2014). This was a very interesting discovery as it suggested that unlike other post-translational
modifications that adopt one enzyme uniformly to perform the modification, histone tail
proteolysis adopts different proteases to facilitate the cleavage activity, in different differentiation
systems.
4
Matrix Metalloproteinase-2: a novel protease required for H3NT proteolysis
The extracellular matrix (ECM) is an important regulator for tissue development and cell
differentiation. It is constantly degraded and remodeled in a tightly regulated manner to ensure
normal homeostasis (Cox and Erler, 2011). Matrix metalloproteinases (MMPs) are important
multi-domain zinc-dependent endopeptidases part of the family of metalloproteases that interact
with specific components of the ECM, mainly fibronectin and collagen, to facilitate ECM turnover.
MMPs are ubiquitously expressed and dysfunction of all MMPs have been associated with a
plethora of diseases, including cancer metastasis, tissue damage, neurological defects, and
inflammation, among others. First characterized as enzymes facilitating tumor growth, the family
of metalloproteases currently has 28 MMPs in vertebrates, 23 members of which are also present
in humans (Gersh and Catchpole, 1949). This family of proteins can further be divided into (a)
matrilysins, (b) collagenases, (c) gelatinases, (d) transmembrane MT-MMPs, and (e) vitronectin-
like other MMPs (Xie et al., 2017).
Most MMPs have conserved structural features, but distinct enzymatic properties and
cleavage patterns. The general structure comprises of a signal N-terminal peptide required for
secreting the MMP into the extracellular matrix, a ~80 amino-acid pro-domain required for
inactivating MMP and secreting it as a zymogen, a ~160 amino-acid catalytic domain required for
enzymatic activity, and a variable “hinge” region connecting the catalytic domain to the
hemopexin domain, which is a ~200 amino-acid domain also required for enzymatic activity. The
inhibitory pro-domain consists of 3 helices and a highly conserved region, PRCGXPD, known as
“cysteine switch”, which is responsible for interacting with the catalytic Zn
2+
ion to maintain
latency and prevent substrate binding (Laronha and Calderia, 2020). Figure 2 highlights the
different domains and 3-dimensional structure of ProMMP-2 as well as the cysteine switch of the
5
pro-domain. MMP-2, part of the sub-group of gelatinases, also includes another “fibronectin”
domain, which are gelatin-binding repeats similar to a motif found in fibronectin (Xie et al, 2017).

Figure 2. Structure of MMP-2 (Laronha and Calderia, 2020).
A. Structure of Pro-MMP2, with the pro-domain in orange. B. 3-dimensional structure of Pro-
MMP2.

The activity of MMPs is tightly regulated at various steps which contributes to its wide
range of functions. The regulation can occur at the level of its transcription, secretion and/or
localization, activation, and interaction with inhibitors. Specifically, the regulation of MMP
activation from its pro-form to an enzymatically active form can be achieved through different
ways. The pro-domain can either be enzymatically cleaved by another endopeptidase, or it can be
chemically modified to release Zn
2+
ion from the cysteine switch (Loffek et al., 2011). Certain
MMPs are activated by interactions with MMPs, for instance, MMP-2. MMP-2 interacts with a
tissue-inhibitor of metalloproteinases (TIMP) protein to form a complex with MMP-2, TIMP2 and
6
MMP-14 (or membrane type-1 MMP) tethered to the plasma membrane (Nagase, 1998). This
interaction between MMP-2 and TIMP2 regulates MMP-2 activity and specifies its localization to
the extracellular matrix, where it selectively degrades type IV collagen (among other targets).
However, the canonical notion that MMPs are just drivers of ECM remodeling and only
interact with specific substrates of the ECM, is evolving. A report investing the H3NT-cleaved
regions in mammals using a novel chromatin immunoprecipitation of acetylated chromatin,
identified MMP-9 as the primary H3NT protease required for proficient osteoclastogenesis (Kim
et al., 2016). The implications of this finding include the recruitment and accumulation of a non-
nuclear protease in the nucleus for the activation of lineage-specifying genes. Based on this, the
Rice lab aimed to identify the protease facilitating H3NT proteolysis during myogenic
differentiation. Using C2C12 cells as a model, the accumulation of histone H3 cleaved product
was concurrent with an increase in nuclear MMP-2 activity. The introduction of shRNA specific
to MMP-2 impaired the process of H3NT proteolysis and hence, myogenic differentiation as well.
Furthermore, the reintroduction of MMP-2 in the extracellular matrix alone, where it is canonically
present to degrade type IV collagen, was not sufficient to restore H3NT proteolysis, thereby
suggesting a novel role of MMP-2 in the nucleus (Rice et al., 2021). This finding gives rise to the
pressing question of how a non-nuclear protease localizes to the nucleus for a specific function.
Surmounting evidence indicates that many MMPs, MMP-2 included, enter the nucleus,
and regulate certain events. Some MMPs have shown to localize to the nucleus via endocytosis or
post-translational modifications. Others, such as the sub-group of matrilysins, have shown to
possess a sequence known as a “nuclear localization signal” (NLS), present typically in the
catalytic domain (Xie et al, 2017). The sequence site of this NLS has shown to vary among
different sub-groups of the metalloproteases family. For certain members of the gelatinase sub-
7
group (which includes MMP-2 and MMP-9), it has been shown that a sequence similar to a known
NLS maybe present on the carboxyl -C terminal of full length pro-MMPs. However, despite the
wide range of reports indicating nuclear MMP activity, a specific mechanism detailing its
localization or a clear NLS sequence and/or DNA-binding motifs have not been identified yet.
In my thesis, I address two aims regarding MMP-2’s function in the nucleus. The first aim
is to identify the minimal MMP-2 protein domain/s required for nuclear localization and the second
aim is to identify the nuclear proteins it interacts with to facilitate H3NT proteolysis at specific
sites of the genome. As stated, H3NT proteolysis has been shown to be required for activation of
genes required for myogenic differentiation and hence, I aim to identify how MMP-2 gets recruited
to these sites that facilitate gene activation. The second aim is further divided into two parts to
identify both, the native and transient interactions facilitating MMP-2’s nuclear activity. To answer
the first aim, I devise truncated constructs eliminating one or more domain/s from the full length
MMP-2 sequence and study their subcellular localization by Western blots and immunofluorescent
microscopy. For the second aim, I adopt a column chromatography approach and a Bio-ID
proximity labelling approach for each of the sub-aims. Collectively my results have identified the
Fibronectin domain to be important for nuclear localization and have shortlisted specific
candidates that interact with MMP-2 in the nucleus. Thus, by highlighting the different types of
proteins MMP-2 interacts with and a mechanistic basis for its nuclear localization, the results of
my thesis ascertain the necessity of a non-nuclear protease (such as MMP) for H3NT proteolysis.
8
Chapter II
Results
Aim 1: Identifying the minimal MMP-2 protein domain required for nuclear localization
MMP-2 has 5 canonical domains, each with a unique function with respect to its activity
and/or structure. The 5 domains are the pro domain (Pro), enzymatic domain (Enz), Fibronectin
domain (Fn), collagen-binding domain (Col), and hemopexin domain (Hex). To identify the
mechanism underlying MMP-2’s nuclear localization, a series of truncated MMP-2 constructs
were designed by eliminating one or more domain/s from the N-terminus.
The progressively truncated MMP-2 constructs were designed on Benchling (2021),
amplified via polymerase chain reaction, assembled onto a CP3 lentiviral backbone via Gibson
assembly, transformed into DHF Competent E. Coli (New England Biolabs), from which they
were extracted and transiently transfected into HEK293 cells for analysis; the protocols for these
steps are detailed under “Materials and Methods”. The truncated constructs (Figure 3A) were
verified by Sanger Sequencing at Genewiz Azenta (South Plainfield, NJ) prior to their
transfections. The HEK293 kidney cell line was selected for this approach as it is known to be
easy to transfect. The other components of the CP3 lentiviral backbone included an EF1 promoter,
an IRES sequence for co-expression and a green fluorescent protein (GFP) sequence for
subcellular visualization (Fig 3B).
9

Figure 3. Construction of truncated MMP-2 constructs.
A. Order of truncated constructs with the progressive elimination of one (or more) domains from
the N-terminus. A 3xHA tag was fused to the construct at the C-terminus. B. The sequence of
components on the CP3 lentiviral backbone including the truncated MMP-2 construct, EF1
promoter, 3xHA tag, IRES sequence for co-expression of the truncated construct and GFP and
GFP (green fluorescent protein) for visualization. C. All the components of the CP3 lentiviral
backbone as prepared on Benchling (2021).

10
Previously, another M.S. student in the Rice lab attempted to address this question by
constructing these truncations using a pE2C vector for the Gibson assembly. The desired sequence
was then transferred into a PQXCP retroviral plasmid (Addgene ID:17386) by the means of
Gateway cloning for expression in mammalian cells. The high salt subcellular fractionation
following the transient transfection was analyzed via a Western blot, using the 𝛼 HA antibody.
Prior results by the Rice lab indicated that the antibody for MMP-2 yielded poor results and was
not accurate, thereby requiring another antibody to stain. The analysis of this previous approach
using the 𝛼 HA antibody indicated that the constructs with the Fibronectin (Fn) domain were
present in both the cytosolic and nuclear extracts; the domains following the removal of the Fn
domain were absent from the nuclear extracts (Figure 4A). However, this approach yielded a low
transfection efficiency (as measured by the number of cells expressing GFP relative to the number
of cells plated). The levels of ColMMP2-3xHA and HexMMP2-3xHA protein expression in the
cytosolic fraction was also lesser as compared to the other constructs. Concurrently, a gift from
the Bell lab (University of Southern California, 2020) enabled the direct insertion of the constructs
into a CP3 lentiviral vector backbone. Using a lentiviral vector as the backbone for the Gibson
assembly eliminated the need for Gateway cloning and improved the transfection efficiency.
The western analysis following a subcellular fractionation of the transiently transfected
constructs contradicted the preliminary results obtained by the previous M.S. student. The transient
transfections of these new constructs indicated that all the constructs were present in the nuclear
and cytosolic fractions. However, there were caveats and inconsistencies in plasmid expression
levels, in both fractions. The truncated proteins seemed to be expressed in high levels, resulting in
their localization to the nucleus (Figure. 4B).
11

Figure 4. Subcellular localization of truncated MMP-2 constructs.
Pro = ProMMP2-3xHA, Enz = EnzMMP2-3xHA, Col = ColMMP2-3xHA, Hex = HexMMP2-3xHA, Fn
= FnMMP2-3xHA. A. Western blot following a high salt nuclear extraction of transiently transfected
constructs in HEK293 cells. A rabbit monoclonal HA antibody (Cell Signaling Technologies) was used
for primary antibody staining, followed by Alexa Fluor 680 goat-anti-rabbit secondary antibody (Thermo
Fisher). PR-Set7 was used as a positive control for nuclear staining. B. Western blot following a high salt
nuclear extraction of transiently transfected constructs in HEK293 cells. The constructs were assembled
on a lentiviral plasmid backbone. Same primary and secondary antibody used as (A).

12
Simultaneously, another approach to visualize the subcellular localization of these
constructs was adopted – immunofluorescent (IF) staining and imaging of cells. To this end, U2OS
human sarcoma cells were cultured for the transient transfections and IF microscopy. The two
main reasons to include another cell line, in addition to HEK293 cells, were – first, HEK293 cells
do not endogenously express MMP-2, while U2OS cells do. This would further enable
investigations regarding MMP-2’s nuclear localization and interactions, governing its recruitment
to specific sites of the genome where H3NT proteolysis is observed. The second reason facilitating
the use of U2OS cells was that they are relatively larger in size and hence, it would be easier to
distinguish between the cytosolic and nuclear regions of U2OS cells as compared to HEK293 cells.
Furthermore, different ratios of Lipofectamine 3000 (uL) to DNA (ug) were tested to
identify the concentrations that ensured maximum transfection efficiency and minimal
overexpression of proteins leading to confounding results. The three ratios that were experimented
with were 0.5:1, 1:1 and 1.5:1 (Lipofectamine 3000-to-DNA). The results from the HEK293 IF
experiment showed that not all GFP positive cells were expressing the HA-tagged protein of
interest. This indicated that not all cells that had taken up the plasmid, were expressing the
truncated construct. However, it reiterated that ProMMP2-3xHA and EnzMMP2-3xHA constructs
were localizing to the nucleus (Figure 5A). These findings underscored the need to generate stable
cell lines as opposed to analyzing transient transfections. Lastly, it also indicated that a 1:1 or 1.5:1
ratio of transfection reagent-to-DNA resulted in more cells expressing GFP. To better visualize
the subcellular localization of the constructs, a group of U2OS cells were permeabilized with 0.5%
triton-X prior to fixing with 2% PFA. This helped expel the cytosolic contents while leaving the
nuclear regions intact. The other group of transiently transfected U2OS cells were fixed first in the
13
same manner as HEK293 cells were. The results of comparing these approaches enabled us to
visualize the differences in localization (Figure 5B).

Figure 5. Immunofluorescent (IF) staining of transiently transfected HEK293 and U2OS cells.
A. IF stained HEK293 cells transiently transfected with EnzMMP2-3xHA construct and ProMMP2-3xHA
construct under 20x magnification. Cells were transiently transfected with a 1:1 ratio of DNA to lipid-based
Lipofectamine reagent, followed by fixation with 2% PFA. Rabbit monoclonal HA antibody (Cell Signaling
Technologies) for primary antibody staining and Alexa Fluor 568 goat-anti-rabbit secondary antibody
(Thermo Fisher) was used as the secondary antibody. Cells were counterstained with DAPI prior to
mounting coverslip. Cells were imaged under the HA channel and GFP channel. B. IF stained transiently
transfected U2OS cells with ColMMP2-3xHA and ProMMP2-3xHA under 63x magnification. Cells were
either fixed with 2% PFA and then stained or permeabilized with 0.5% triton-X to release cytosolic contents
(before fixing and staining). IF protocol is the same as (A).

14
However, the main issue with this approach was that U2OS cells are not as easy to
transiently transfect as HEK293 cells are. Hence, the plasmid expression and localization patterns
were inconsistent with the previous results, as even the ColMMP2-3xHA construct was present in
the nucleus. Furthermore, there was high background signaling intensity (indicated by the high
levels of auto-exposure), implying that there was not enough signal to image. With these results,
the approach was again modified to generate stable cell lines in both U2OS and HEK293 cells by
the means of viral transduction.
To this end, HEK293T cells were transfected with the plasmid of interest along with other
vectors required to produce virus, which was then used to generate stable cell lines; the details of
these respective steps are under “Materials and Methods”. A subcellular fractionation of the
transduced cells followed by a Western blot supported the results from the preliminary experiment
i.e., the Fibronectin domain is required for nuclear localization. In the HEK293 stable cells, the
cytosolic fraction from the ProMMP2-3xHA cells were used as a positive control and the wild-
type HEK293 nuclear extracts were used as a negative control. The Pro-, Enz- and FnMMP2-
3xHA constructs were seen in the cytosolic fractions and (in lesser intensity) in the nuclear
fractions. However, the ColMMP-2 was only seen in the cytosolic fraction (Figure 6A).
Furthermore, given that HEK293 cells do not endogenously express MMP-2, none of the truncated
constructs were present in the chromatin fraction. In the case of the stable U2OS cells, the Pro-
and EnzMMP2-3xHA constructs were present in all 3 fractions (Figure 6B). As a positive control
for nuclear presence, ProMMP2 with a known SV40 nuclear localization signal was used.
Interestingly, the Hex construct was not detected in the cytosolic fraction but was present in the
nuclear and chromatin fractions. Similarly, the Col construct was not present in the nuclear or
cytosolic fractions but was faintly present in the chromatin fractions. As elaborated upon further
15
in the “Discussion” section, the localization patterns for the Col- and HexMMP2-3xHA constructs
needs to be investigated further given their small size and hence, relatively unstable nature.

Figure 6. Subcellular localization of MMP2-3xHA truncated constructs in U2OS and
HEK293 stable cells.
Pro = ProMMP2-3xHA, Enz = EnzMMP2-3xHA, Col = ColMMP2-3xHA, Fn = FnMMP2-3xHA,
WT = wild-type, cyto = cytosolic fraction, NE = nuclear extract, Chr = chromatin. A. Western blot
of subcellular fractions from HEK293 stable cells collected following a high salt extraction. A
rabbit monoclonal HA antibody (Cell Signaling Technologies) was used for primary antibody
staining, followed by Alexa Fluor 680 goat-anti-rabbit secondary antibody (Thermo Fisher). B.
Western blot of subcellular fractions from U2OS stable cells collected following a high salt
extraction. Same conditions used as in (A).

16
Furthermore, the immunofluorescent staining of the stable U2OS cells reiterated the
nuclear presence of the Pro- and EnzMMP2-3xHA, while the Col construct was absent (Figure 7).
This observation, with respect to the localization of the ColMMP2-3xHA construct, was
inconsistent with the Western blot from the subcellular fractionation, further supporting the
possibility of protein instability affecting its patterns of localization.

Figure 7. Immunofluorescent (IF) staining of stable U2OS cells at 63x magnification.
Exposure time for the cells that were permeabilized first were kept constant relative to the cells
that were fixed first. Cells were counterstained with DAPI for nuclear staining and imaged under
the HA channel. A. IF of ProMMP2-3xHA constructs that were (left) fixed first and then
permeabilized, and (right) permeabilized first and then fixed. B. IF of EnzMMP2-3xHA construct.
C. IF of ColMMP2-3xHA construct.

Put together, the collective results of the Western analysis and immunofluorescent
approach suggests that the Fibronectin domain is important for MMP-2 nuclear localization as the
Pro- and EnzMMP2-3xHA constructs have consistently been present in the nuclear and cytosolic
extracts, whereas the constructs following the removal of the Fibronectin domain have not
consistently been reported in the nuclear regions.
17
Thus, to ascertain the necessity of the Fibronectin domain for nuclear localization, a 𝛥 Fn
construct was prepared, which included all the domains except the Fibronectin domain. It was
hypothesized that if the Fibronectin domain is necessary for nuclear localization, then the 𝛥 Fn
construct will be present only in the cytosolic fraction and not the nuclear fraction. The construct
has been transformed into and extracted from E. Coli colonies via a miniprep. Following the
sequence verification and subsequent transduction of the construct in U2OS and HEK293 cells, a
Western blot and immunofluorescent staining will be performed.

18
Aim 2: Determining the components of the nuclear MMP-2 multiprotein complex
Nuclear MMP-2 activity results in a relatively small fraction of H3NT proteolysis within
chromatin, similar to other known cell types that display H3NT proteolysis (Azad et al., 2018).
These findings support the selective targeting of MMP-2 protease activity to specific genomic
regions rather than random and unregulated H3NT proteolysis. Based on these observations we
hypothesize that selective MMP-2 genomic targeting and protease activity are directed by
unidentified MMP-2-associated nuclear proteins. The goal of this aim is to identify the components
of the nuclear MMP-2 multiprotein complex and ultimately determine how these interactions
facilitate selective MMP-2 nucleation and H3NT proteolysis within the genome.
Consistent with our findings in the mouse C2C12 myogenesis cell line model, it was
recently reported that MMP-2 also functions as the nuclear H3NT protease in the U2OS human
osteosarcoma cell line (Ali et al., 2021). We decided to leverage the U2OS cell line for these
studies because it is a more homogenous model, compared to the C2C12 myogenesis model, that
could provide less ambiguous results. To confidently identify components of the nuclear MMP-2
multiprotein, we used two different biochemical approaches as illustrated in Figure 8. The first
approach utilizes conventional chromatography to partially purify the MMP-2 complex followed
by immunoprecipitation (IP)-mass spectrometry for protein identification. The second approach is
a more modern proximity-labeling approach that also incorporates IP-mass spectrometry for
protein identification. We reasoned that comparison between these two approaches would reveal
nuclear MMP-2 associated proteins with the highest confidence for further investigation.

19

Figure 8. Experimental flowchart of approaches to identify nuclear MMP2-interacting
proteins.
Left: Nuclear soluble extracts from U2OS human osteosarcoma cell line stably expressing full
length MMP2 fused to an HA epitope tag (ProMMP2-3xHA) were fractionated using Q XL anion-
exchange column chromatography. ProMMP2-3xHA enriched in the 0.6 M NaCl fraction was
immunoprecipitated using HA-magnetic beads. The eluted material was fractionated by SDS-
PAGE and the gel outsourced for protein identification.
Right: U2OS cells stably expressing ProMMP2-miniTurbo or a miniTurbo-NLS negative control
were incubated with biotin to biotinylate proximal proteins. The nuclear soluble extracts were
immunoprecipitated with streptavidin- conjugated magentic beads and washed extensively. On-
bead trypsin digest was performed and the eluted peptides were outsourced for protein
identification.
Bottom: Proteins identified in both approaches will represent bona fide nuclear MMP2- interacting
proteins for further investigation.

20
● Aim 2A: Determining components of the native nuclear MMP-2 complex
To identify the members of the native nuclear MMP-2 complex, I leveraged a stable U2OS
cell line generated by another graduate student in the Rice Lab, that expresses the full-length mouse
MMP-2 protein fused to a 3x-HA epitope tag at the C-terminus (ProMMP2-HA). The U2OS
ProMMP2-HA and wild type negative control cells were grown to confluence in ten 15 cm culture
dishes. Nuclei were purified from other cellular components, as described in the “Methods”
section. The nuclear soluble extracts were purified and collected from the insoluble chromatin
fraction using moderate salt conditions (400 mM NaCl) and high-speed centrifugation. The
extracts were carefully diluted in a low salt buffer to achieve a final NaCl concentration of ~100
mM. The diluted extracts (input) were applied to an equilibrized 1 ml Q XL column using a syringe
at a rate of <0.5 ml/minute.
As shown in Figure 9, the Western analysis demonstrated that this approach allowed
maximal retention of ProMMP-2 to the column with a substantial amount of non-specific proteins
being removed (flow through). The column was then washed with 10 column volumes of buffers
containing increasing concentrations of NaCl. Western analysis demonstrated that the bulk of
ProMMP2-HA was eluted in the first 2 mls of the 0.6 M buffer and that little of the protein was
retained on column following a 1 M NaCl elution (Figure 9).The 𝛼 HA Western analysis for
assessing the levels of ProMMP2-3xHA in different fractions, was stained by Coomassie to
confirm equal levels of loading.
21

Figure 9. Enrichment of ProMMP2-3xHA from the Q XL anion exchanger at elution with
0.6 M of NaCl.
Top: Nuclear soluble extracts from U2OS ProMMP2-3xHA (P) and wild type negative control
(W) were fractionated on a Q XL anion-exchange column using increasing concentrations of NaCl.
Western analysis demonstrates enrichment of ProMMP2-3xHA in the first 1 ml elution using 0.6
M NaCl. L = ladder, Input = nuclear soluble extracts, FT = flow-through/unbound, Fn= fraction
(n = fraction number). W = wild type U2OS, P = U2OS ProMMP2-3xHA.
Bottom: Western blot was stained with Coomassie to ensure equal loading.

22
To purify the ProMMP2-HA complex, I performed an immunoprecipitation using anti-HA
conjugated magnetic beads. The first 2 ml fraction of the 0.6 M NaCl eluted nuclear extracts from
the Q XL chromatography were slowly diluted to a final NaCl concentration of ~300 mM before
applying to the equilibrated HA-beads. Following an overnight incubation, the unbound proteins
were collected, and the HA-beads were washed extensively in a low stringency buffer. The beads
were then incubated with excessive amounts of an HA peptide to competitively elute ProMMP2-
HA and associated proteins. Although these proteins were not detected by Coomassie staining, the
more sensitive silver staining approach confirmed the elution of ProMMP2-HA and other proteins
(Figure 10). To assess the efficiency of the HA peptide elution, the beads were incubated with 100
mM glycine pH 3.0 and then boiled in SDS-PAGE load buffer to elute any remaining bound
proteins. The results indicate that incubation with the HA peptide was inefficient at eluting
ProMMP2-HA from the beads (Figure 10). Despite the low efficiency, the HA peptide eluted
material was fractionated briefly by SDS-PAGE and a 3 cm gel slice was submitted for in-gel
trypsin digestion and protein identification.
23

Figure 10. Enrichment of ProMMP2-3xHA following a 100 mM glycine pH 3.0 elution.
Left: Western blot of samples from the immunoprecipitation of the 0.6 M NaCl fraction using
HA-magnetic beads followed by elution using excess HA peptide, 100 mM glycine pH 3.0 or
SDS-PAGE load dye.
Right: Silver staining of gel to detect bound proteins.

The results obtained from mass spectrometry were filtered such that candidates with
peptide counts of 1 or less in the experimental groups and peptide counts greater than 0 in the
control group were filtered out. Using the CRAPome database, non-specific candidates that are
commonly identified in several other mass spectrometry experiments were also filtered out. The
cut-off range for the CRAPome specificity was 10% i.e., candidates that were identified in greater
than 71 experiments of 713 mass spectrometry experiments via the CRAPome database were
eliminated. The following table includes the list of candidates identified following the HA-
immunoprecipitation mass-spectrometry analysis (Table 1). This list will be cross-referenced with
the list of candidates identified in Aim 2B.

24
Gene
symbol
WT HA
CRAPome
Specificity (%) Average SC Max SC Unique Total Unique Total
TRRAP 0 0 7 8 1.82 1.6 3
RUNX1 0 0 2 3 0.42 3.3 8
COL1A1 0 0 2 2 2.93 7 46
SEC11A 0 0 2 2 1.12 1.5 3
UACA 0 0 2 2 1.12 2.8 7
MRPS30 0 0 4 5 2.93 1.3 3
SPATS2L 0 0 4 6 4.75 2.3 10
NMNAT1 0 0 2 2 0.70 1.6 2
ISG20L2 0 0 2 3 2.37 1.1 2
MMP2 0 0 1 2 2.51 2.1 6
ARVCF 0 0 4 4 0.42 2.7 5
HELZ 0 0 3 4 6.70 4 31
BCAR1 0 0 2 2 0.56 1.5 2
EIF2B1 0 0 2 2 4.19 1.5 4
EIF4E2 0 0 2 2 7.54 5.9 19
MRPL21 0 0 2 2 6.28 1.3 4
SYF2 0 0 2 2 6.01 2.2 8
C7orf50 0 0 2 2 8.66 2 6
Table 1. List of candidates identified interacting with nuclear MMP-2 via Q XL anion
exchanger column chromatography and HA-immunoprecipitation.

25
● Aim 2B: Identifying transient interactions facilitating nuclear MMP-2’s recruitment
to specific sites of the genome.
As a supplemental and complementary approach to Aim 2A, the proximity labeling
approach was used to identify nuclear MMP-2 interacting proteins (Cho et al., 2020). As illustrated
in Figure 11, the experimental protein is fused to the catalytic site of a biotinylate. The fusion
protein is expressed in cells followed by incubation with the biotin substrate. Those proteins that
are in close proximity to the fusion protein become artificially labeled with biotin. By leveraging
the high affinity nature of biotin-streptavidin, the biotinylated proteins are selectively captured on
streptavidin-conjugated beads. Stringent wash conditions facilitate the removal of non-specific
proteins from the beads. An on-bead trypsin digestion is performed to capture peptide fragments
of the remaining biotinylated proteins, which are then identified by mass spectrometry.
For consistency with Aim 2A, I performed these experiments in the U2OS cell line using
the mouse ProMMP-2 fusion protein. The CP3 lentiviral backbone was once again used with the
sequence for the Turbo enzyme attached to the C-terminus of ProMMP-2. As a negative control
and to measure the normal levels of biotinylation within the nucleus of the U2OS cells, another
construct with sequences only for the miniTurbo-ID enzyme and SV40 (a known NLS) was
prepared (Figure 11B). This construct would localize to the nucleus and biotinylate proximal
proteins in the nucleus, which, when compared to the ProMMP2-Turbo construct, would help
eliminate all non-MMP2 specific interactions. Both these plasmids were transduced into HEK293,
U2OS and C2C12 cells to prepare stable cells, as performed by another member of the Rice lab.
The U2OS ProMMP-2-Turbo and NLS-Turbo cells were grown to confluence in two 15
cm dishes and incubated with 5 mM biotin substrate for increasing periods of time. The cells were
washed extensively in PBS to remove any remaining free biotin and then nuclei were isolated as
26
described under “Methods”. Nuclei were resuspended in RIPA buffer and the nuclear soluble
fraction was purified from the insoluble chromatin fraction by high speed centrifugation. The
subcellular fractionation followed by the western blot of the U2OS stable cells indicated that in as
little as 15 minutes, ProMMP2-3xHA gets biotinylated by the miniTurbo-ID enzyme (Figure 11C).
The levels of biotinylation did not significantly change with time (as noted by the relatively
constant signal from each condition). Another negative control used in this experiment, was a
group of U2OS cells transduced with the ProMMP2-miniTurbo construct, but not treated with
biotin. Based on these results, it was decided that cells would continue to be grown in biotin free
media overnight and would be treated with biotin for an hour prior to the nuclear extraction.
27

Figure 11. Purification of nuclear MMP2-interacting proteins by BioID proximity labeling.
A. ProMMP2 fused to the miniTurbo enzyme should induce biotinylation of ProMMP2-proximal proteins
in the presence of biotin. The biotinylated proteins can then be purified by streptavidin-
immunoprecipitation and the proteins identified by mass spectrometry. B. Schematic of ProMMP2-
miniTurbo and miniTurbo-NLS negative control stably expressed in U2OS cells. C. Top: Western analysis
using a streptavidin-IR800 antibody demonstrates effective biotinylation of nuclear ProMMP2- miniTurbo
(and potential binding partners) following a 30 minute incubation with 100 mM biotin.
Bottom: The presence of the HA-tagged construct was detected using a rabbit monoclonal HA antibody
(Cell Signaling Technologies) which was followed by staining with Alexa Fluor 680 goat-anti-rabbit
secondary antibody (Thermo Fisher).
28
After treating cells with biotin and collecting the nuclear soluble fractions, the next step
was to isolate the biotinylated proteins from the nuclear extracts. To this end, the nuclear soluble
fraction was applied to equilibrated streptavidin-magnetic beads and incubated overnight in the
cold room. The unbound fraction was collected, and the beads were washed extensively in high
stringency buffer to remove any non-specific proteins. The beads were then incubated with trypsin
to proteolyze the biotinylated proteins bound to the streptavidin beads and the resulting soluble
peptide fragments were collected. Three independent biological replicates were performed for the
ProMMP-2-Turbo and NLS-Turbo negative control and the 6 peptide mixtures were submitted for
protein identification by mass spectrometry. The western analysis of the different fractions
collected before and after the Streptavidin bead pulldown, highlight enrichment of ProMMP2-
3xHA in the fraction eluted from the beads (Figure 12). This supports the previous finding that
MMP-2 (and its binding partners) gets biotinylated.
29

Figure 12. Isolation of nuclear MMP2-interacting proteins following Streptavidin bead
pulldown.
Immuno- precipitation of nuclear RIPA extracts using streptavidin magnetic beads followed by
Western analysis with an α-HA antibody (top) or streptavidin-IR800 antibody (bottom). Results
demonstrate the enrichment of biotinylated ProMMP2- miniTurbo after immunoprecipitation as
well as all biotinylated proteins. Inp = nuclear RIPA extract input, UB = unbound/ flow-through,
B= bound/elution with SDS-PAGE load dye, Wn = wash (n denotes the number of washes), Pro =
U2OS ProMMP2-3xHA-miniTurbo, Turbo = U2OS miniTurbo-NLS. The bead-bound proteins
were digested with trypsin and the resulting peptides outsourced for protein identification by mass
spectrometry.

30
The results of the mass spectrometry were analyzed in the same manner as explained under
Aim 2A. Comparative analysis of the mass spectrometry data between ProMMP-2-Turbo and the
NLS-Turbo negative control samples was performed using the SimpliFi software package (ProtiFi,
2022). To identify the most likely ProMMP-2 interacting proteins, a stringent statistical cutoff of
p < 0.005 and at least a 4-fold difference between the samples was applied (Figure 13).

Figure 13. Volcano plot highlighting statistically significant candidates that were
upregulated in PromMMP2-miniTurbo nuclear extracts.

31

Gene
symbol P-value Fold change
CRAPome
Specificity % Average SC Max SC
CCT8 0.00001111 57.28 62.99 36.7 267
RPAP3 0.0002066 48.08 15.92 4.6 13
HELZ 0.0005629 14.75 6.70 4 31
PIK3C2A 0.0006974 8.681 1.12 1.3 2
MMP2 0.0009745 239.5 2.51 2.1 6
TNRC6B 0.001137 6.342 22.63 19.4 97
MYH9 0.002802 5.938 62.57 51.8 667
RAI 0.005379 5.3 0.00 0 0
MAGED1 0.005972 4.321 12.57 2 19
VCL 0.006059 4.326 21.79 4.1 28
TAGLN2 0.00814 4.528 57.26 9.9 63
AHNAK2 0.009648 4.368 9.64 33.1 378
AHKHD1 0.0105 7.22 0.00 0 0
STAU 0.01209 4.53 0.00 0 0
CNOT1 0.01399 5.288 18.16 3.6 37
CCT6 0.01803 4.886 0.00 0 0
XRN1 0.01889 5.575 14.66 16.3 49
CCT2 0.02208 12.29 59.78 19.3 245
PIH1D1 0.03871 7.705 11.17 1.8 6
Table 2. List of candidates identified interacting with nuclear MMP-2 via miniTurbo-ID
biotinylation and mass spectrometry.
32
Chapter III
Discussion
MMP-2 is an extracellular matrix protein, previously shown to be required for histone H3
N-terminal proteolysis and myogenic gene activation. Here, I aimed to identify a mechanism
underlying MMP-2 nuclear localization and nuclear interactions facilitating MMP-2’s recruitment
to specific sites of the genome concurrent with H3NT proteolysis. For the first aim, the Fibronectin
(Fn) domain has been identified to play a role in MMP-2 nuclear localization. The constructs
following the removal of the Fibronectin domain were inconsistently present in the nucleus. For
the second aim, a Q XL anion exchanger column chromatography approach as well as a Bio-ID
proximity labeling approach was adopted to identify proteins interacting with MMP-2 in the
nucleus. Devising and optimizing protocols for two independent approaches ensured a higher
confidence in the list of candidates shortlisted. Ultimately, these approaches expand the
investigation into MMP-2’s nuclear activity and the functional significance of H3NT proteolysis
with respect to cell differentiation and development.

Is Fibronectin the minimal domain for MMP-2 nuclear localization?
For Aim 1, truncated constructs were devised to assess the minimal domain required for
MMP-2 nuclear localization. The preliminary approach required an additional step of Gateway
cloning as an empty pE2C vector backbone was used for the Gibson assembly. The replacement
of the pE2C vector with a CP3 lentiviral plasmid improved the transfection efficiency and helped
optimize the protocol. However, the transient transfection of the truncated constructs still
demonstrated caveats in the plasmid expression levels, as seen in Figure 4B. Thus, switching to
producing stable cells as opposed to performing transient transfections ensured that differences in
33
expression levels were on account of the presence (or absence) of a particular domain, and not
because of the experimental setup. To this end, the Western analysis and immunofluorescent
staining and imaging of the stable constructs demonstrated that the ProMMP-2 construct was
localizing to the nucleus in both HEK293 and U2OS cells (Figures 6 and 7). This was interesting
as HEK293 cells do not endogenously express MMP-2 or undergo histone tail proteolysis (Rice et
al., 2021). Thus, seeing ProMMP-2 localize to the nucleus, indicates that the machinery for MMP-
2 nuclear localization is conserved within MMP-2 and is not cell-specific.
Furthermore, the truncated constructs following the removal of the Fibronectin domain
were not consistently present in the nucleus (Figure 7). The western analysis contradicted the
immunofluorescent microscopy for the ColMMP2-3xHA stable construct. While the western blot
did not detect the Col construct in the nuclear extracts, there was faint signaling seen via the
immunofluorescent staining. The inconsistent subcellular expression for the Col- and HexMMP2-
3xHA constructs could be on account of unstable and non-functional proteins being produced since
most of the domains were eliminated from these constructs. To address this issue, it was initially
proposed to use SUMO tags to prevent these constructs from being prematurely degraded. By
preventing the truncated proteins from being targeted for destruction, these tags would ensure for
accurate subcellular localization. However, since these results could also indicate that the
Fibronectin domain is required for nuclear localization and the removal of this domain prevents
consistent nuclear localization, an alternate solution was adopted to address this issue. A ∆Fn
construct was prepared eliminating only the Fibronectin domain, while keeping the other domains
in order. While this construct is yet to be sequence verified and stably transduced into U2OS and
HEK293 cells, the results of this experiment would ascertain a mechanism underlying MMP-2
nuclear localization, by demonstrating how the elimination of the Fn domain prevents MMP-2
34
from localizing to the nucleus. A subsequent experiment reintroducing the Fibronectin domain
back into the ∆Fn construct will positively determine the necessity of this domain for nuclear
localization, while reiterating its sufficiency.
The ∆Fn construct will be analyzed via the same biochemical techniques as the previous
constructs were, having previously optimized the protocols for the transfection, transduction,
extraction, and fluorescent microscopy. The controls for the fluorescent microscopy were cells that
were fixed before permeabilization, and cells that were permeabilized before fixation. Initially,
these controls were imaged under auto-exposure, thereby maximizing the signal intensity for the
different constructs. However, the main drawback is that it significantly increases the background
staining intensity, indicating the presence of constructs in the nucleus, even if that is not the case.
To avoid such confounding results, the exposure time for all the cells permeabilized first was kept
constant relative to the exposure time of the cells that were fixed first. This ensured that every
construct’s nuclear signaling intensity was relative to their cytoplasmic signaling intensity. By
adjusting the exposure time and ensuring uniformity across the controls, the results from the
immunofluorescent microscopy of the ∆Fn construct will answer the proposed research question.
Ultimately, the goal is to be able to reproduce this experiment in C2C12 mouse satellite
cells, to understand the physiological significance of MMP-2 nuclear localization with respect to
H3NT proteolysis and myogenic gene activation. As stated, the presence of MMP-2 in the nucleus
of HEK293 cells indicates that there is an intrinsic way for MMP-2 to localize to the nucleus, even
in cells, where it is not normally expressed. This supports the proposed mechanism with respect to
the role of the Fibronectin domain. However, other reports have highlighted alternate mechanisms,
given that there has been a growing consensus that other MMPs, in addition to MMP-2, have novel
roles in the nucleus. A report highlighting the presence of catalytically active MMP-2 in the
35
nucleus of endothelial and neuronal cells also highlighted the nuclear localization of TIMP-
1 (Sinha et al., 2014). While this report did not identify a mechanistic basis for MMP-2 nuclear
localization, it suggested the possibility of MMP-2 co-localizing to the nucleus with the TIMP-1
protein. This finding is further supported by the fact that MMP-2 is activated from its pro-form by
associating with MT1-MMP and TIMP-2, another member of the TIMP family of proteins. One
experiment to test this hypothesis would be to prevent TIMP-1 from localizing to the nucleus or
knocking down the protein and then studying MMP-2’s subcellular localization to further establish
a correlation between the two.
Alternatively, other studies have identified putative nuclear localization signals in MMPs
using bioinformatic softwares such as PSORT (Frolova et al., 2020). This tool identified a nuclear
localization signal like sequence (PKWRKTH) in the catalytic domain of MMP-2 based on a
previous report identifying MMP-2 in the nucleus of cardiomyocytes (Kwan et al., 2004). Other
members of the Rice lab have utilized such bioinformatic tools to identify NLS-like sequences in
MMP-2, but have not been successful in identifying a sequence with a low specificity cut-off. As
the specificity cut-off for NLS-like sequence increases, the less accurate, a sequence is likely to
be. Future studies could be aimed at identifying whether different MMPs adopt different ways to
localize to the nucleus via these putative NLS-like sequences, or if the Fibronectin domain,
identified in this study, plays a role in tandem with TIMP proteins to ultimately localize MMP-2
to the nucleus.

MMP-2 nuclear interactions facilitating recruitment to sites of H3NT proteolysis
For the second aim, it was hypothesized that selective MMP-2 genomic targeting and
protease activity are directed by unidentified MMP-2-associated nuclear proteins and hence, to
36
identify these proteins with a greater confidence, two independent approaches were adopted. In
Aim 2A, the Q XL anion exchanger column chromatography approach eluted the MMP-2 nuclear
complex following a wash with 0.6M NaCl, from which the MMP-2 nuclear complex was isolated
via a HA-immunoprecipitation. The results from the immunoprecipitation experiment were sent
to mass spectrometry and filtered such that only candidates with positive peptide counts in the
experimental groups and low CRAPome specificity were shortlisted.
The results of Aim 2A indicated interesting nuclear interactions for MMP-2. For instance,
one of the proteins identified with a very high peptide count in the experimental group, was the
TRRAP protein, which is the transformation/transcription associated domain protein. It has shown
to be a key component of the HAT complexes, recruiting histone acetyltransferases to several sites
of the genome for DNA repair, replication and transcription (Murr et al., 2007). Identifying
TRRAP in the subsequent replicate experiments would suggest that TRRAP is a protein that
functions to recruit various known histone tail modifying enzymes, including MMP-2, to specific
sites of the genome. However, while the unique peptide counts for TRRAP were significantly
above the average peptide counts and was identified in less than 2% of the experiments listed on
CRAPome, this protein was identified after only a single replicate of the experiment (i.e., n=1).
Future experiments need to include more independent replicates to increase the validity of this
finding. Furthermore, the protocols for this Aim are yet to be optimized as there is little enrichment
following the HA-immunoprecipitation using the excess HA peptides and a significant number of
proteins were still bound to the beads (which were eluted only after being boiled in SDS-PAGE
load dye for 10 minutes or using 100 mM glycine pH 3.0). The reduced efficiency of the HA-IP
elution poses an issue as it eliminates other unknown candidates that could potentially interact
with nuclear MMP-2.
37
Aim 2B adopted a Bio-ID proximal labeling approach. Using the miniTurbo-ID enzyme,
proteins proximal to the C-terminus of ProMMP2-3xHA were labeled in the presence of biotin,
isolated following a streptavidin bead pulldown and identified via mass spectrometry. A promising
candidate identified in Aim 2B (that was also the only candidate overlapping with the results of
Aim 2A) was the HELZ protein. While little is known about HELZ, it is a probable helicase with
a zinc finger binding domain. The putative RNA helicase, part of the Upf-1 superfamily I class of
helicases, was more recently shown to be important for and promote the initiation of translation
(Hasgall et al., 2011). Additionally, the presence of the zinc finger binding domain supports the
hypothesis that HELZ could potentially recruit MMP-2 to specific sites of the genome for histone
tail proteolysis. It also supports preliminary results obtained by another member of the Rice lab
which highlights a strong association or binding between MMP-2 and RNA polymerase-II at
transcription start sites of different genes. Other candidates identified from Aim 2B with high fold-
change increases were members of the CCT family and RPAP3. However, following the
CRAPome analysis, some members of the CCT family (such as CCT8, CCT2 and CCT6) were
believed to be less genuine interactions, given their high specificity or presence in other
experiments. The other promising candidate, RPAP3, is a component required for the activation
of multi-molecular complexes such as RNA polymerase II and complexes of the PIKK family,
among others (Rodríguez & Llorca, 2020). Our results indicate a very high fold-change increase
with a low p-value, collected from 3 independent replicates, suggesting that this could be another
genuine interaction with nuclear MMP-2 or atleast, a protein in close proximity to nuclear MMP-
2.
However, the main drawback of the candidates identified from Aim 2B is that they are
restricted by the radius in which the Turbo-ID enzyme can biotinylate proteins. All the candidates
38
identified presently, are proximal to the C-terminus of MMP-2 and may not necessarily reflect the
interactions taking place with MMP-2’s catalytic domain, closer to the N terminal. A follow-up
set of experiments could address this issue by having the sequence for the Turbo-ID on the N-
terminus of ProMMP-2 and identifying differences (if any). Furthermore, the peptide
concentration of the samples sent for mass spectrometry was less than desired, similar to the issue
with low enrichment following the HA-IP elution. This suggests that the efficiency of the trypsin
digest needs to be optimized and increased.
Subsequent research questions could investigate whether the mechanism and/or domains
proposed in this study are conserved for all MMPs (or non-nuclear H3NT proteases) that localize
to the nucleus or differ based on their respective nuclear functions and activities. Ultimately, once
these results have been corroborated through replicate experiments, the data would indicate a
potential mechanism by which MMP-2 facilitates H3NT proteolysis. More importantly, it
definitively ascertains the necessity of a non-nuclear protease for myogenic gene activation.

39
Chapter IV
Materials and Methods
Construction and assembly of truncated constructs
The sequence and primers for each truncation was modified from the full-length mMMP-
2 (NM_008610.3) on Benchling (2021). The truncated sequences were amplified via polymerase
chain reaction (PCR) using an empty pE2C donor vector (Addgene ID:17462) as a template, which
consisted of a 3xHA tag. The amplified products fused with the 3xHA tag from the pE2C donor
vector were size- and sequence verified prior to its use. Each construct was run on a 1.5% agarose
gel and was subsequently extracted for assembly via Gibson assembly (Zymoclean Gel DNA
Recovery Kit, Zymo Research). Forward and reverse sequence primers were designed and
procured from Integrated DNA Technologies.
The Gibson assembly required a CP3 lentiviral plasmid backbone which was a gift from
the Bell lab (University of Southern California, 2020). Prior to the assembly, the lentiviral plasmid
was digested with BamHI-HF and EcoRI (New England Biolabs) for 3 hours at 37°C and then
treated with alkaline phosphatase for 30 minutes at 37°C. The plasmid was run on a 1% agarose
gel and extracted in the same manner as other constructs. The lentiviral plasmid backbone already
included sequences for an EF1 promoter, IRES element, green fluorescent protein (GFP),
ampicillin and puromycin resistant genes. GFP was used as a measure of assessing transfection
efficiency and its co-expression with the truncated constructs was ensured by the IRES element.
The ampicillin and puromycin resistant genes allowed for bacterial selection and preparation of
stable cell lines.
The different components were assembled in a specific order based on their homology arms
and were incubated with a Gibson assembly master mix at 50°C for 1 hour. The product was
40
transformed into DH5 𝛼 Competent E. Coli (Thermo Fisher), which was streaked onto agar plates
with ampicillin and incubated at 37°C overnight. Alternatively, the plasmids were also transformed
into NEB Stable Competent E. Coli (New England Biolabs) under the same conditions.

Plasmid extraction
DH5 𝛼 Competent E. Coli colonies that survived and grew on the ampicillin-treated agar
plates were picked, cultured in LB growth medium with ampicillin (1000:1 dilution), and
subsequently mini-prepped to extract the plasmid DNA (Zymogen ZR Plasmid Miniprep-kit,
Classic). The extracted plasmids were sequenced via Genewiz Azenta (South Plainfield, NJ) and
following the confirmation of the correct desired sequence, a larger-scale midi-prep was performed
(Invitrogen HiPure Plasmid Midiprep-kit). To increase the extraction efficiency, the bacterial
colonies picked for the large-scale grow out were incubated in 200 mLs of LB growth media at
37°C overnight. The midi-prepped plasmid DNA was used to transfect or transduce the cells.

Transient transfections
HEK293 (ATCC CRL-1573) cells were plated in 6-well tissue culture plates at a density
of 1.5 cells per well. Each well had 2 mLs of Dulbecco’s Modified Eagle’s Medium (DMEM,
Corning), supplemented with 10% fetal bovine serum, amino acids glutamine and non-essential
amino acids. The plates were left overnight in an incubator at 37°C with 5% CO2. The following
day, the cells looked about 70-80% confluent.
Concurrently, a lipid-based reagent, 3.75 uL Lipofectamine 3000 (Invitrogen), along with
10 uL of p3000 (Invitrogen) was added to 5 uL of 500 ng plasmid DNA in 250 uL of Opti-MEM
(Thermo Fisher). The solution was left to incubate at room temperature for 30 minutes before
41
being added to the 2 mLs of cells in growth medium. The cells were left in the incubator again
overnight at 37°C with 5% CO2. The transfection efficiency was measured by the number of GFP
positive cells relative to the number of cells plated.

Viral Transduction
HEK293T cells (ATCC CRL-11268) cells were plated in 6-well tissue culture plates at a
density of 4.0 cells per well and grown in 2 mLs of regular growth media overnight at 37°C with
5% CO2. The following day, the cells looked about 70-80% confluent.
3 hours prior to transfection, the growth media was replaced by 1.75 mLs of serum free media
(without FBS). Concurrently, the DNA-Lipofectamine 3000 cocktail was prepared and left to
incubate for ~45 minutes prior to the transfection. The ratio of DNA:psPax2:psMD.2 is
2ug:1.5ug:0.5ug (Addgene). psPax2 and psMD.2 are second generation lentiviral plasmids
required to produce virus. 20 uLs of p3000 and Lipofectamine 3000 reagent were added to this
cocktail along with 500 uL of Opti-MEM. These were mixed well prior to their incubation. 250
uL of the transfection cocktail was added per well. 4 hours post-transfection of HEK293T cells
with the DNA-Lipofectamine 3000 cocktail, 500 uL of 50% FBS media was added to each well,
bringing the final concentration of the media to 10% FBS.
C2C12 cells were plated in 6 well plates at a density of 4.0 cells per well and grown in 2
mLs of regular growth media overnight. The following day, these cells were 30-40% confluent. 1-
2 days following the transfection (based on the transfection efficiency), the virus was collected by
spinning the cells in 15 mL conical tubes for 5 minutes. 8 ug/mL of polybrene was added to each
mix of virus. The media from the C2C12 cells was aspirated and replaced with media containing
42
virus and polybrene. These cells were spun on a rotor at a temperature of 30 and speed of 1100
RCF for 1hour. Subsequently, they were left in the incubator overnight at 37°C with 5% CO2.
The following morning, the media was changed and replaced with fresh growth media.
Using the fluorescent microscope, the transduction efficiency was measured. The cells were then
treated with puromycin or FACS sorted by another member of the Rice lab. The same protocol
was used for U2OS and HEK293 stable cells.

Purification of nuclei
Cells were initially grown in regular growth media overnight at 37°C with 5% CO2. Prior
to purification, the cells were washed three times with ice cold PBS in the cold room, scraped,
collected in 15 ml conical tubes and pelleted by centrifugation at 500 x g for 5 minutes. The
supernatants were decanted, and the cell pellet was resuspended in 1 ml ice cold Low Salt Buffer
(LSB: 20 mM Hepes-KOH pH 7.8, 25% glycerol, 1.5 mM MgCl2, 2 mM EDTA pH 8, 1 mM DTT
and protease inhibitors (1 mM PMSF, 2 ug/ml aprotinin, pepstatin and leupeptin)) and transferred
to chilled 1.5 ml microfuge tubes. Cells were centrifuged 500 x g for 5 minutes, resuspended in
another 1 ml LSB and incubated on ice for >15 minutes to induce cell swelling. To lyse the cells,
110 ul of 10% NP-40 detergent (1% final) was added followed by gentle pipetting (10x) and passed
through a 25-gauge syringe (10x) to liberate nuclei. The nuclei were isolated by centrifugation at
500 x g for 5 minutes and any remaining cytosolic contaminants were removed by washing the
nuclei 2x in LSB.

43
High salt nuclear soluble extraction
The pelleted nuclei volume (PNV) was determined by resuspending in 500 ul of LSB and
measuring the difference in volume by pipetting. Nuclei were pelleted and then gently resuspended
in 1 vol of LSB containing 400 mM NaCl (200 mM NaCl final). While vortexing the nuclei on
setting 6, ¼ vol of High Salt Buffer (HSB: 1.2 M NaCl, 20 mM Hepes-KOH pH 7.8, 25% glycerol,
1.5 mM MgCl2, 2 mM EDTA pH 8, 1 mM DTT and protease inhibitors (1 mM PMSF, 2 ug/ml
aprotinin, pepstatin and leupeptin)) was added dropwise to achieve a final NaCl concentration of
~400 mM. The nuclei were rotated in the cold room for >30 minutes to liberate nuclear soluble
components from chromatin. The nuclear soluble component was isolated from the chromatin
following high speed centrifugation at 21,000 x g for 20 minutes. The extracts were used
immediately or stored at -40 ℃ for future studies.

Chromatography
For aim 2A, the nuclear soluble extracts were vortexed on setting 6 while adding dropwise
3 vol LSB to achieve a final NaCl concentration of ~100 mM. The solution was centrifuged at
21,000 x g for 20 minutes to pellet any precipitates before applying the clarified material to an
equilibrized 1 ml Q XL anion-exchange column using a 10 ml syringe at a rate of <0.5 ml/minute.
The flow through containing the unbound proteins was collected for analysis. The column was
sequentially washed with 10 ml of LSB buffers containing increasing concentrations of NaCl (0.2,
0.6 and 1 M) and 2 ml fractions were collected (5 tubes/wash).

44
HA-Immunoprecipitation
The first 2 ml 0.6 M eluted fraction from the chromatography was diluted in 1 vol of LSB
to achieve a final NaCl concentration of 300 mM. The solution was applied to 40 ul of equilibrized
anti-HA conjugated magnetic beads (CST) and rotated overnight at 4C. The unbound proteins were
collected for analysis and the beads were washed 5x in low stringency IP buffer (50 mM Hepes-
KOH pH 7.8, 300 mM NaCl, 1.5 mM MgCl2, 10 mM KCl, 0.2% Triton X-100, 10% glycerol).
The beads were then incubated in 50 ul HA elution buffer (50 mM Hepes-KOH pH 7.8, 100 mM
NaCl, 1.5 mM MgCl2, 0.05% Triton X-100, 10% glycerol) containing 2 mg/ml HA peptide for 30
minutes shaking at 1000 rpm. This first elution fraction was collected, a second elution was
performed, and the samples were combined and stored at -40℃. To determine the efficiency of the
HA peptide elution, the beads were incubated with 100 ul 100 mM glycine pH 2.3 for 30 minutes
at 1000 rpm. This material was collected, and the beads were then incubated with 100 ul SDS-
PAGE load dye, boiled for 10 minutes and collected.

Western Analysis
Samples were prepared per the experiment’s requirements and 6x SDS-PAGE loading dye
was added in a 1:6 ratio (dye-to-sample). The samples in SDS-PAGE loading dye were boiled at
95°C for 10 minutes to ensure complete denaturation of proteins. A 10% or 14% agarose gel was
prepared using 40% bis/acrylamide, which was run at 180 mV for 45-60 minutes (based on the
amount of separation needed for the experiment). A PageRuler protein ladder (Invitrogen) was
also added to the first lane of the gel. A 0.2 um or 0.4 um nitrocellulose membrane was used for
the transfer (depending on the nature of the blot), which was run at 55 mAmps for 90 minutes. A
0.2 um nitrocellulose membrane was preferred for blotting 14% gels (which shows greater
45
separation of smaller proteins, such as histones) and a 0.4 um nitrocellulose membrane was used
for blotting 10% gels (for cytosolic or nuclear extracts). The membrane was blocked in 4% milk
in TBS for 30 minutes, followed by incubation with the primary antibody solution (prepared in
1.5% milk in TBS). For the blocking step, the cells were placed on a rocker at room temperature,
but for the primary body incubation, the cells were left on the benchtop. Following the primary
body incubation, the blot was washed with TBS-T for 5 minutes. After 3 such washes in TBS-T,
the membrane was treated (in the dark) with the secondary antibody solution. Alexa Fluor 680
goat anti-rabbit (Invitrogen) was used for this step and was prepared in 1.5% milk in TBS-T at a
dilution of 1:10k. For samples treated with and extracted from biotinylated media (for Aim 2B),
another secondary antibody treatment was performed after the Alexa Fluor 680 antibody treatment.
The cells were washed a total of 6 washes in TBS-T for 5 minutes each time. They were then
blocked once again, in 3% BSA in TBS for 30 minutes. An IR-800 Streptavidin antibody (Li-Cor)
was used at a concentration of 1:2500 in blocking serum (i.e. 3% BSA in TBS-T) and added to the
cells for 45 minutes-1 hour. The blot was washed in TBS-T 3 times and deionized water, prior to
its imaging on the Odyssey Clx software.

Immunofluorescent staining
U2OS cells were plated in 6-well plates at a density of 2.0 cells per well. Prior to adding
the cells, each well had a microscope glass cover placed inside it for the cells to grow on. When
cells were roughly 80% confluent, they were fixed with 2 mLs of 2% PFA in phosphate buffered
saline (PBS, Corning) for 5-7 minutes at room temperature. Following 3 washes with PBS, they
were permeabilized with 2 mLs of 0.5% triton-X in PBS for the same time and temperature. Again,
following 3 washes with PBS for five minutes each time, the cells were blocked with 2 mLs of 2%
46
goat serum in PBS (blocking serum) for 30-45 minutes at room temperature. Cells that were used
to visualize nuclear localization of different constructs were permeabilized first and then fixed
(using the same reagents as above).
A primary antibody solution was prepared by diluting the antibody in blocking serum. 125
uL of primary antibody solution was added to each well for 2 hours and the plate was left on the
benchtop at room temperature. Following the primary antibody incubation, the wells were washed
3 times, for 5 minutes each time, with blocking serum. A secondary antibody solution using Alexa
Fluor 568 goat-anti-rabbit (Invitrogen) was prepared in a 1:1000 dilution in blocking serum. 1 mL
of this was added to each well and kept aside in the dark for an hour at room temperature.
Following the secondary antibody incubation, each well was washed 3 times, for 5 minutes each
time, with 2 mLs of PBS and then deionized water.
To affix the microscope glass covers onto a cover slide (VWR Scientific), a fine point
tweezer was used to lift the glass covers and place them on the cover slide. Prior to affixing the
glass covers, each cover slide had one drop of DAPI solution added onto it. Once the glass covers
were mounted on a slide, they were left to dry for 2 hours before imaging at 63x on a fluorescent
microscope. The exposure time between the experimental samples was kept constant.

miniTurbo-ID and Streptavidin bead pulldown
The U2OS ProMMP-2-Turbo and NLS-Turbo negative control cells were grown to
confluence in two 15 cm dishes. Biotin (50 uM final) was added to the media and incubated for 1
hour. Cells were washed 3x with ice cold PBS in the cold room to remove any free biotin, cells
were scraped, collected in 15 ml conical tubes and pelleted by centrifugation at 500 x g for 5
minutes. Nuclei were then isolated as described above and the nuclei were resuspended completely
47
in 500 ul of RIPA buffer (50 mM Hepes-KOH pH 7.9, 450 mM NaCl, 1 mM EDTA, 1% Triton
X-100, 0.5% NaDoc, 0.1% SDS) by gentle pipetting. The mixture was incubated on ice for >10
minutes and then clarified by centrifugation at 13,000 x g for 10 minutes. The nuclear soluble
extracts were transferred to a tube containing 25 ul of equilibrized streptavidin magnetic beads and
then rotated overnight in the cold room. The flow through was collected and the beads were washed
sequentially 2x with RIPA buffer, 1x with 1 M KCl, 1x with 0.1 M Na2CO3, 1x with freshly
prepared 2 M urea in 10 mM Tris-HCl pH 8.0 and then 2x PBS. To prepare for the on-bead trypsin
digestion, the beads were washed 4x with 200 ul 50 mM ammonium bicarbonate shaking at 500
rpm for 20 minutes each wash. The mass spec-grade trypsin (Promega) was diluted 1:100 in 50
mM ammonium bicarbonate and 50 ul was applied to the beads. The beads were incubated
overnight shaking at 500 rpm in the cold room. The trypin-digested solution was collected, the
beads were incubated in 50 ul of 50 mM ammonium bicarbonate to capture any residual peptides
and then combined (100 ul total). The samples were lyophilized and sent for protein identification
by mass spectrometry.

48
References
Ali, M. A., Garcia‐Vilas, J. A., Cromwell, C. R., Hubbard, B. P., Hendzel, M. J., Schulz, R. (2021).
Matrix metalloproteinase‐2 mediates ribosomal RNA transcription by cleaving nucleolar
histones. The FEBS Journal, 288(23), 6736–6751. https://doi.org/10.1111/febs.16061

Allis, C. D., Chicoine, L. G., Richman, R., & Schulman, I. G. (1985). Deposition-related histone
acetylation in micronuclei of conjugating tetrahymena. Proceedings of the National Academy
of Sciences, 82(23), 8048–8052. https://doi.org/10.1073/pnas.82.23.8048

Azad, G. K., Swagatika, S., Kumawat, M., Kumawat, R., & Tomar, R. S. (2018). Modifying
chromatin by histone tail clipping. Journal of Molecular Biology, 430(18), 3051–3067.
https://doi.org/10.1016/j.jmb.2018.07.013

Cho, K. F., Branon, T. C., Udeshi, N. D., Myers, S. A., Carr, S. A., Ting, A. Y. (2020). Proximity
labeling in mammalian cells with turboid and split-turboid. Nature Protocols, 15(12), 3971–
3999. https://doi.org/10.1038/s41596-020-0399-0

Cox, T. R., & Erler, J. T. (2011). Remodeling and homeostasis of the extracellular matrix:
Implications for fibrotic diseases and cancer. Disease Models & Mechanisms, 4(2), 165–178.
https://doi.org/10.1242/dmm.004077

Cutter, A. R., & Hayes, J. J. (2015). A brief review of nucleosome structure. FEBS Letters,
589(20PartA), 2914–2922. https://doi.org/10.1016/j.febslet.2015.05.016

Dhaenens, M., Glibert, P., Lambrecht, S., Vossaert, L., Van Steendam, K., Elewaut, D., & Deforce,
D. (2014). Neutrophil elastase in the capacity of the “H2A-specific protease.” The
International Journal of Biochemistry & Cell Biology, 51, 39–44.
https://doi.org/10.1016/j.biocel.2014.03.017

49
Duncan, E. M., Muratore-Schroeder, T. L., Cook, R. G., Garcia, B. A., Shabanowitz, J., Hunt, D.
F., & Allis, C. D. (2008). Cathepsin L proteolytically processes histone H3 during mouse
embryonic stem cell differentiation. Cell, 135(2), 284–294.
https://doi.org/10.1016/j.cell.2008.09.055

Frolova, A. S., Petushkova, A. I., Makarov, V. A., Soond, S. M., & Zamyatnin, A. A. (2020).
Unravelling the network of nuclear matrix metalloproteinases for targeted drug design.
Biology, 9(12), 480. https://doi.org/10.3390/biology9120480

Gersh, I., & Catchpole, H. R. (1949). The organization of ground substance and basement
membrane and its significance in tissue injury, disease and growth. American Journal of
Anatomy, 85(3), 457–521. https://doi.org/10.1002/aja.1000850304

Grant, P. A. (2001). A tale of histone modifications. Genome Biology, 2(4).
https://doi.org/10.1186/gb-2001-2-4-reviews0003

Hasgall, P. A., Hoogewijs, D., Faza, M. B., Panse, V. G., Wenger, R. H., & Camenisch, G. (2011).
The putative RNA helicase HELZ promotes cell proliferation, translation initiation and
ribosomal protein S6 phosphorylation. PLoS ONE, 6(7).
https://doi.org/10.1371/journal.pone.0022107

Keppler, B. R., & Archer, T. K. (2008). Chromatin-modifying enzymes as therapeutic targets –
part 1. Expert Opinion on Therapeutic Targets, 12(10), 1301–1312.
https://doi.org/10.1517/14728222.12.10.1301

Khalkhali-Ellis, Z., Goossens, W., Margaryan, N. V., & Hendrix, M. J. (2014). Cleavage of histone
3 by cathepsin D in the involuting mammary gland. PLoS ONE, 9(7).
https://doi.org/10.1371/journal.pone.0103230

Kim, K., Punj, V., Kim, J.-M., Lee, S., Ulmer, T. S., Lu, W., Rice, J. C., & An, W. (2016). MMP-
9 facilitates selective proteolysis of the histone H3 tail at genes necessary for proficient
osteoclastogenesis. Genes & Development, 30(2), 208–219.
https://doi.org/10.1101/gad.268714.115
50
Kwan, J. A., Schulze, C. J., Wang, W., Leon, H., Sariahmetoglu, M., Sung, M., Sawicka, J., Sims,
D. E., Sawicki, G., & Schulz, R. (2004). Matrix metalloproteinase‐2 (MMP‐2) is present in
the nucleus of cardiac myocytes and is capable of cleaving poly (adp‐ribose) polymerase
(PARP) in vitro. The FASEB Journal, 18(6), 690–692. https://doi.org/10.1096/fj.02-1202fje

Laronha, H., & Caldeira, J. (2020). Structure and function of human matrix metalloproteinases.
Cells, 9(5), 1076. https://doi.org/10.3390/cells9051076

Loffek, S., Schilling, O., & Franzke, C.-W. (2011). Biological role of matrix metalloproteinases:
A critical balance. European Respiratory Journal, 38(1), 191–208.
https://doi.org/10.1183/09031936.00146510

Murr, R., Vaissière, T., Sawan, C., Shukla, V., & Herceg, Z. (2007). Orchestration of chromatin-
based processes: Mind the TRRAP. Oncogene, 26(37), 5358–5372.
https://doi.org/10.1038/sj.onc.1210605

Mutskov, V., Gerber, D., Angelov, D., Ausio, J., Workman, J., & Dimitrov, S. (1998). Persistent
interactions of core histone tails with nucleosomal DNA following acetylation and
transcription factor binding. Molecular and Cellular Biology, 18(11), 6293–6304.
https://doi.org/10.1128/mcb.18.11.6293
Nagase, H. (1998). Cell surface activation of progelatinase a (PROMMP-2) and cell migration.
Cell Research, 8(3), 179–186. https://doi.org/10.1038/cr.1998.18

Phillips, D. M., & Johns, E. W. (1959). A study of the proteinase content and the chromatography
of thymus histones. Biochemical Journal, 72(3), 538–544. https://doi.org/10.1042/bj0720538

Rice, J. C., Weekley, B. H., Kanholm, T., Chen, Z., Lee, S., Fernandez, D. J., Abrahamson, R.,
Castaldi, A., Borok, Z., Dynlacht, B. D., & An, W. (2021). MMP-2 is a novel histone H3 N-
terminal protease necessary for myogenic gene activation. Epigenetics & Chromatin, 14(1).
https://doi.org/10.1186/s13072-021-00398-4

51
Rodríguez, C. F., & Llorca, O. (2020). RPAP3 C-terminal domain: A conserved domain for the
Assembly of R2TP co-chaperone complexes. Cells, 9(5), 1139.
https://doi.org/10.3390/cells9051139

Sinha, S. K., Asotra, K., Uzui, H., Nagwani, S., Mishra, V., & Rajavashisth, T. B. (2014). Nuclear
localization of catalytically active MMP-2 in endothelial cells and neurons. American journal
of translational research, 6(2), 155–162.

Xie, Y., Mustafa, A., Yerzhan, A., Merzhakupova, D., Yerlan, P., N Orakov, A., Wang, X., Huang,
Y., & Miao, L. (2017). Nuclear matrix metalloproteinases: Functions resemble the evolution
from the intracellular to the extracellular compartment. Cell Death Discovery, 3(1).
https://doi.org/10.1038/cddiscovery.2017.36

Yi, S.-J., & Kim, K. (2018). Histone tail cleavage as a novel epigenetic regulatory mechanism for
gene expression. BMB Reports, 51(5), 211–218.
https://doi.org/10.5483/bmbrep.2018.51.5.053

Abstract (if available)

Abstract Previously, a canonical extracellular matrix protease, matrix metalloproteinase 2 (MMP- 2), was reported to unexpectedly function in the nucleus to cleave the N-terminal tail of the DNA- associated histone H3 protein and is required for the activation of myogenic genes during skeletal muscle differentiation. Although histone H3 N-terminal tail (H3NT) proteolysis within chromatin was first observed over 60 years ago, the specific genomic sites selectively targeted for H3NT proteolysis, and the biological significance of this evolutionarily conserved epigenetic modification remain largely unknown.
In my thesis, I leverage established biochemical approaches to discover the minimal protein domain required for MMP-2’s nuclear localization as well as to identify nuclear proteins interacting with MMP-2. I aim to determine how these nuclear interactions facilitate H3NT proteolysis and affect gene activation. To this end, I devise truncated constructs of the MMP-2 protein for the first aim and study their subcellular localization in different cell lines. For the second aim, and to study the nuclear interactions recruiting MMP-2 to specific sites of the genome concurrent with H3NT proteolysis, I adopt a Q XL anion column chromatography approach and a more modern Bio-ID proximity labeling approach. The results of these two aims highlight nuclear interactions that facilitate H3NT proteolysis and affect gene activation and most importantly, ascertain the necessity of a non-nuclear protease for myogenic differentiation.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Determining the necessity of nuclear matrix metalloproteinase-2 (MMP-2) activity for proficient skeletal muscle differentiation

PDF

Developing new tools to discover and investigate histone H3 proteases

PDF

The role of histone H4 lysine 20 monomethylation in gene expression and differentiation

PDF

Multiple functions of the PR-Set7 histone methyltransferase: from transcription to the cell cycle

PDF

Impacts of post-translational modifications on interactions between G9a and its N-terminus binding partners

PDF

The mechanism of recruitment of Tip60 to ER target genes

PDF

Studies of the biological relevance of Histone H4 Lysine 20 monomethylation: discovery of its role in the cell cycle and localization within the human genome

PDF

Mixed lineage leukemia proteins (MLLs), their effect as coregulators on target gene expression and global histone methylation

PDF

A novel screening approach to identify regulators of polycomb-dependent gene silencing

PDF

Molecular mechanism of the recruitment of SWI/SNF chromatin remodeling complex and histone acetyltransferase to estrogen-responsive promoters

PDF

Regulation of Aurora kinase B and its effect on phosphorylation of G9a/GLP

PDF

Do the ZFX and ZFY transcription factors have redundant or unique functions?

PDF

Epigenetic plasticity of cultured female human embryonic stem cells and regulation of gene expression and chromatin by PR-SET7 mediated H4K20me1

PDF

Characteristics of hydrogen peroxide inducible clone-5 and its potential role as a nuclear receptor coactivator

PDF

The effect of estradiol on TAZ in stromal cell line ST2 in osteoblast differentiation and ER+breast cancer cell

PDF

The mechanisms of somatic cell reprogramming

PDF

The role of Rpd3 and Sir2 in regulation of replication initiation in budding yeast: Rpd3 acts directly on single-copy origins while Sir2 works through ribosomal DNA origins

PDF

Elucidating the functional role of CHD7 associated nuclear PDH complex and other associated proteins on neural crest development

PDF

The essential role of histone H2A deubiquitinase MYSM1 in natural killer cell maturation and HSC homeostasis

PDF

TNNI3K expression affects nucleation of cardiomyocytes and skeletal myoblasts

Asset Metadata

Creator Mazumdar, Nimisha (author)

Core Title Determining the mechanism and necessity of matrix metalloproteinase 2 (MMP-2) nuclear localization for proficient skeletal muscle differentiation

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Medicine

Degree Conferral Date 2022-08

Publication Date 01/19/2023

Defense Date 05/04/2022

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

Tag anion exchange column chromatography,bio-ID proximity labeling,histone tail,nuclear localization,OAI-PMH Harvest,post translational modifications

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Rice, Judd (committee chair), An, Woojin (committee member), Bell, Oliver (committee member)

Creator Email nimisha.mazumdar@gmail.com,nmazumda@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111373212

Unique identifier UC111373212

Legacy Identifier etd-MazumdarNi-10857

Document Type Thesis

Format application/pdf (imt)

Rights Mazumdar, Nimisha

Type texts

Source 20220719-usctheses-batch-955 (batch), 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