University of Southern California Dissertations and Theses

Profiling of protein methylation in mammalian cells and methods for deep methylproteomic analysis

(USC Thesis Other)

Profiling of protein methylation in mammalian cells and methods for deep methylproteomic analysis

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Profiling of Protein Methylation in Mammalian Cells
And Methods for Deep Methylproteomic Analysis

by
Nicolas G. Hartel

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)

May 2022

Copyright 2021 Nicolas Hartel

ii
Acknowledgments
I would like to acknowledge the amazing friends and scientists that have taught me so much and
supported my research throughout the journey of my PhD at USC. My mentor and advisor, Dr. Nicholas A.
Graham has always been there for me and introduced me to the world of mass spectrometry proteomics
which I have developed a great passion for that I will keep with me the rest of my life. His patience and
willingness to teach are extraordinary. I also thank my committee members, Dr. Noah Malmstadt and Dr.
Marcelo Coba for their collaborative endeavors and for being a part of my dissertation committee.
Other scientists whose collaborations and advice have been essential to my research include Dr. Jian Qin,
Dr. Jian Xu, Dr. Brent Wilkinson, Dr. Alireza Delfarah, Dr. Dongqing Zheng, Dr. James Joly, Dr. Lu Wang,
Christopher Hughes, Melanie MacMullan, Belinda Garana, Brandon Chew, and Christopher Liu. It has
been one of the greatest honors of my life to have the chance to work with all of them.
I also acknowledge my personal friends, Artur Tarverdyan, Ani Tarverdyan, Ben Zarostky, Ronald
Zarotsky, and Luba Zarostky who have been there for me every time I needed them.
I also thank the people of the Mork Family Department at the Viterbi School of Engineering at
USC, especially Andy Chen, Marcus Walker, and Anthony Tritto for their support and guidance.
I would also like to thank the team at Amgen, especially my supervisors Christina Chen and Brooke
Rock, who selected me for an internship that taught me so much about industry and got me even more
excited about the field of biotechnology and what is possible.
Lastly, I thank my brother Martin Hartel, and my parents Carl and Antonella who have supported
me my whole life with love and affection without which I would not be who I am.
I thank all for their love and support, for without it I would not have made it this far.

iii
Table of Contents
Acknowledgements ……………………………………………………………………………………….ii
List of Tables ……………………………………..…………………………………………………....….v
List of Figures ……………………………..………………………………………………...……….…...vi
Abstract …………………………………………..……………………………………….……….……..vii
1. Chapter 1. Introduction …………………………...……………………………………………….....1
1.1. Biological role of protein methylation ……………………………………………………….....1
1.2. Protein methylation as a post translational modification ………………………………….........2
1.3. LC-MS based proteomics ……………………………...…………………………………….....4
1.3.1. Challenges of methyl peptide analysis ………………………………………...…………7
2. Chapter 2. Deep Methylation Profiling of PRMT1 in 293T cells ……………………...…………..9
2.1. Objective ……………………………………...……………………………………………......9
2.2. Materials and Methods ………………………...…………………………………………….....8
2.3. Results ………………………………………...………………………………………………15
2.3.1. Orthogonal enrichment of methylated peptides cover different subsets of protein
methylome ……………………...………………………………………………………..15
2.3.2. Label free quantitation of methyl peptides shows high reproducibility ………………...18
2.3.3. Quantitative Analysis of MMA peptides from shPRMT1 293T cells ……...…………...20
2.3.4. Characteristic neutral losses enable discrimination of SDMA and ADMA ……...…..…22
2.3.5. Quantitative analysis of DMA peptides from shPRMT1 Cells ………………...…...…..25
2.3.6. Lysine methylation is not affected by PRMT1 knockdown ……………………...……..27
2.3.7. Integrated analysis of methyl-arginine forms reveal novel PRMT1 substrates and
substrate scavenging …………………………………...………………………….…….28
2.4. Discussion and Conclusion ……………………………...……………………………………31
iv
3. Chapter 3. Improved Discrimination of Asymmetric and Symmetric Arginine Dimethylation by
Optimization of the Normalized Collision Energy in Liquid Chromatography−Mass
Spectrometry Proteomics ……………………………………………………………………..…….37
3.1. Objective ……………………………………………...………………………………………37
3.2. Materials and Methods …………………...…………………………………………………...37
3.3. Results ………………………………………...………………………………………………40
3.3.1. Higher NCE improves discrimination of ADMA and SDMA in synthetic peptides …..40
3.3.2. Optimization of NCE using endogenous DMA Peptides ……...……………………….42
3.3.3. Optimized NCE improves assignment of ADMA and SDMA spectra in orthogonal
methyl-peptide enrichment techniques …………………………………...…………….45
3.3.4. Optimized NCE improves confidence of ADMA assignment through higher occurrence
of NL across ion series ……………………………………...………………………….47
3.4. Conclusion …...………………………………………………………………………………….49
4. Chapter 4. Methylproteomics of Mouse CARM1 KO Quadricep Muscle Tissue …………...….51
4.1. Objective …...……………………………………………………………………………………51
4.2. Materials and Methods ……………………………………………………………………….....51
4.3. Results ……………………...……………………………………………………………………55
4.3.1. Arginine methylome profiling in mouse skeletal muscle ..………………………......…55
4.3.2. Monomethyl Analysis of Mouse CARM1 KO in Skeletal Muscle ..…………………...58
4.3.3. Quantitative and Integrative Analysis of CARM1-mediated Protein Arginine
Methylation in Mouse Skeletal Muscle ..……………………………………………….60
4.4. Conclusion ..……………………………………………………………………………………..62
5. Tables …………………………………………………………………………………………….......63
6. References ……………………………………………………………………………………….......65

v
List of Tables
Table 1: Original “Long” gradient for SCX ………………………………………………………....63
Table 2: New proposed “Short” gradient for SCX …………………………………………………..63
Table 3: Summary of spectra identified by Long SCX gradient .……………………………………63
Table 4: Summary of spectra identified by Short SCX gradient …..………………………………...63
Table 5: Summary of spectra identified by SCX & IAP technique in shPRMT1 293T cells .………64
Table 6: Distribution of arginine methylation sites compared to other widespread post-translational
modifications ……..………………………………………………………………………...64

vi
List of Figures
Figure 1.1 Forms of Arginine and Lysine Methylation ………………………………...…………..3
Figure 1.2 Workflow of LCMS based Proteomics ……………….…………………………...……7
Figure 2.1 SCX and IAP enrich methyl peptides and target different subsets of protein
methylome ……….…...……………………………………………………………….17
Figure 2.2 Label free quantitation of methyl peptides is highly reproducible …………………….19
Figure 2.3 Quantitative analysis of MMA peptides from shPRMT1 293T cells …………………21
Figure 2.4 Characteristic neutral loss of methylamine and dimethylamine allows discrimination of
SDMA and ADMA spectra …………………………………………………………...23
Figure 2.5 Quantitative Analysis of DMA peptides from shPRMT1 293T cells …………………26
Figure 2.6 Lysine methylation is unaffected by shPRMT1 ……………………………………….28
Figure 2.7 Integrated analysis of methyl-arginine forms reveals novel PRMT1 substrates and
ADMA substrate scavenging ……………………………..…………………………...30
Figure 3.1 Higher NCE improves the generation of NL ions in synthetic ADMA and SDMA
peptides …………………………………….………………………………………….41
Figure 3.2 Higher NCE improves andromeda scores of synthetic ADMA and SDMA peptides ...42
Figure 3.3 Optimization of NCE for discrimination of ADMA and SDMA in endogenous methyl-
peptides ………………………………………………..………………………………44
Figure 3.4 Increased identification of MMA peptides at optimized NCE ………………………...45
Figure 3.5 Increased Andromeda scores at higher NCE for peptides with and without neutral
loss ………………………………………………………………………………….…46
Figure 3.6 Comparison of standard and optimized NCE for high pH SCX and ADMA
immunoaffinity methyl-peptide purification ……………………..…………………...47
Figure 3.7 Higher NCE improves occurrence of NL across ion series, improving the confidence of
ADMA identification ………………………………………………………………….48
Figure 3.8 Optimized NCE enables ADMA/SDMA assignment of a doubly dimethylated
peptide …………………………………...……………………………………………49
Figure 4.1 Arginine methylproteomic profiling of skeletal muscle ………………………………57
Figure 4.2 Monomethyl arginine analysis of CARM1 mKO muscle ……………………………..59
Figure 4.3 Quantitative and integrated analysis of CARM1-mediated protein arginine
dimethylation in skeletal muscle ……………………………………………………...61

vii
Abstract
Protein methylation has been implicated in many important biological contexts including
signaling, metabolism, and transcriptional control. Despite the importance of this post-translational
modification, the global analysis of protein methylation by mass spectrometry-based proteomics has
not been extensively studied due to the lack of robust, well-characterized techniques for methyl
peptide enrichment. To better investigate protein methylation, we utilize two orthogonal methods for
methyl peptide enrichment: immunoaffinity purification (IAP) and high pH strong cation exchange
(SCX). These techniques are able to enrich separate sets of methyl peptides enhancing coverage of
the available protein ‘methylome’, or group of methylated proteins. Use of these techniques is applied
to investigate the effects of knock-down of PRMT1, the most active enzyme involved in arginine
methylation events in cells. Despite its important biological roles, arginine dimethylation remains an
understudied post-translational modification. Partly, this is because the two forms of arginine
dimethylation, asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA), are
isobaric and therefore indistinguishable by traditional mass spectrometry techniques. Thus, there
exists a need for methods that can differentiate these two modifications. Recently, it has been shown
that the ADMA and SDMA can be distinguished by the characteristic neutral loss (NL) of
dimethylamine and methylamine, respectively. However, the utility of this method is limited because
the vast majority of dimethylarginine peptides do not generate measurable NL ions. We report that
increasing the normalized collision energy (NCE) in a higher-energy collisional dissociation (HCD)
cell increases the generation of the characteristic NL that distinguish ADMA and SDMA. By
analyzing both synthetic and endogenous methyl-peptides, we identify an optimal NCE value that
maximizes NL generation and simultaneously improves methyl-peptide identification. Finally, a
combination of these strategies was used to measure the effects of in vivo knockout of CARM1 on the
methylome of mouse skeletal muscle tissue, showcasing an application of these techniques beyond
proteins collected through traditional cell culture techniques.
1
1. Chapter 1. Introduction
1.1. Biological Role of Protein Methylation
Protein methylation was first identified over 60 years ago by Ambler and Rees in bacterial
flagellar protein (1) and again in eukaryotes by Paik and Kim in 1967 through radioactive
labelling of calf thymus nuclei (2), in which they identified fractions of lysine and arginine whose
elutions were shifted from the unmodified arginine and lysine fractions. It would not be for
another 20 years that new advances would be made in the field of protein methylation as discussed
in the review by Clarke in 1993 in which methylated proteins were involved in the protein damage
response associated with aging (3). As biochemical techniques matured the role of protein
methylation on arginine residues became better understood and was found to be catalyzed by
protein arginine methyltransferases (PRMTs). Work in the 1990’s and early 2000’s revealed
protein methylation was associated with cell cycle proteins and the DNA damage response
through methylation of HNRNPUL1 and MRE11 by PRMT1 allowing them to localize to the site
of DNA damage (4).
Studies have also identified RNA binding proteins as a heavily methylated class of
enzymes in which methylation affects RNA-protein interaction, RNA splicing, localization, and
transcription (5–8). Many proteins involved in the spliceosome are methylated in arginine-
glycine-glycine or RGG domains where these residues are thought to contribute to favorable
stacking interaction with RNA bases (9, 10). Transcription factors are also well known interactors
of PRMTs such as the case of E2F-1 which increased in expression when PRMT5 was depleted,
coinciding with decreased cell growth and apoptosis (11). Knockouts of PRMT1 and PRMT5 are
lethal in mouse and cell models further demonstrating their important functions in growth and
development (12, 13).
2
Overexpression of PRMTs have also been found in certain cancers such as PRMT2 in
breast carcinoma (14), and PRMT6 in lung cancer (15). Certain vulnerabilities have also been
studied such as the dependence on PRMT5 in glioblastoma with passenger deletion of the gene
MTAP, in which knockdown of PRMT5 was lethal to cell viability (16–18). This work has spurred
development of PRMT5 inhibitors which are being screened in clinical trials for anti-cancer
properties (19). PRMT4 also known as CARM1 regulates cancer cell metabolism by methylation
of the glycolytic enzyme PKM2, thereby shifting cell metabolism from oxidative phosphorylation
to aerobic glycolysis (20). Another report on pancreatic ductal adenocarcinoma found that
methylation of metabolic enzyme MDH1 on R248 by PRMT4 interrupted glutamine metabolism,
thus sensitizing cancer cells to oxidative stress and limiting cell division (21).
These examples highlight the emerging biological role of protein methylation and provide
motivation to study this modification in greater detail.

1.2. Protein Methylation as a Post Translational Modification
Post-translational modification of proteins (PTMs) regulate diverse cellular processes by
conferring complexity to proteins in addition to their primary amino acid sequence (22). Well
studied modifications such as phosphorylation of serine, threonine, and tyrosine residues are
known to regulate a host of biological events from metabolism to mitosis (23). Inhibitors of
protein phosphorylation have also been used as anti-cancer therapeutics such as imatinib mesylate,
a tyrosine kinase inhibitor introduced in 1998 for patients with chronic myeloid leukemia (24).
Since then new generations of inhibitors and combination therapies have improved patient
3
outcomes as protein phosphorylation was studied further (25). Recent studies have now identified
protein methylation by protein arginine methyltransferases, as regulators of metabolism, cell
cycle, signal transduction, and transcriptional control (5, 6, 20, 21, 26–30).
Methylation of proteins occurs primarily on the ionic residues lysine and arginine.
Methylation of lysine can occur as a monomethyl, dimethyl, or trimethyl form and arginine
methylation also exists in three forms, a monomethyl arginine (MMA), asymmetric dimethyl
arginine (ADMA), and symmetric dimethyl arginine (SDMA) in which the latter two dimethyl
forms are isobaric in mass but differ in the structural position of the second methyl group either
on the same guanidino nitrogen as for ADMA, or in the case of SDMA on the opposing nitrogen
(Fig. 1.1).

Figure 1.1 Forms of lysine and arginine methylation. Adapted from (28).
Nine PRMTs are known to exist in mammalian genomes which can be classified by the
type of arginine methylation that is catalyzed by each. Type I PRMTs catalyze MMA and ADMA
marks whereas Type II enzymes catalyze MMA and SDMA marks (31). Type I PRMTs include
4
PRMT1, PRMT2, PRMT3, PRMT4, PRMT6, and PRMT8, followed by the Type II enzymes
PRMT5 and PRMT7 although PRMT7 may act as a third class known as Type III catalyzing only
MMA on certain substrates (32). The most active PRMT is PRMT1 which accounts for the
majority of methylation events in mammalian cells, followed by PRMT5 (33). Notably
methylation of arginine does not alter the charge of the residue but does affect steric hindrance and
removes a potential hydrogen bonding site for each methylation moiety (34).

All PRMTs use S-adenosyl methionine (SAM) as a the methyl-donor molecule. SAM is
the second most used enzyme substrate next to ATP (35). They methyl group on SAM is bound to
a sulfur atom which destabilizes the usually inert methylthiol group allowing attack by
nucleophiles such as N, O, and S (36).
It has been observed that most PRMTs exhibit a preference for methylation of RGG motifs
within protein sequences, especially those of RNA binding proteins (9). Proteins harboring RGG
sequences typically contain additional RNA binding domains such as hnRNPK domains and/or
other RNA recognition motifs (37).

1.3. LC-MS Based Proteomics
Similar to how genomics is the study of the entirety of a genome, proteomics is the study of
the entirety of the proteins that are transcribed and translated from that genome (38).
Identification and quantitation of a large number of proteins at the cellular and tissue level
constitutes the field of proteomics. Additional measurable characteristics of interest include
protein expression, interactions with other proteins/biomolecules, subcellular localization, and
5
potential modifications to the protein structure and sequence (39). Technological advances in
high resolution mass spectrometry and liquid chromatography have allowed the robust detection
of nearly all proteins in an organisms proteome (40).
Innovations such as electro-spray ionization (ESI) (41) is an example of “soft” ionization
techniques that has proven quite adept at the in-source ionization of fragile peptides and
proteins. Relatively recent breakthroughs for mass spectrometry mass analyzers include the
development of the orbitrap mass analyzer with mass accuracy measurements in the 2-3ppm
range, m/z window of up to 6000, and dynamic range > 10
3
have been a boon to the field of
proteomics (42). Development of nano-flow ultra high performance liquid chromatography
(UHPLC) instruments using flow rates in the low hundreds of nanoliters per minute have
allowed highly sensitive detection due to the relatively high concentration of biomolecules in
the flow regime (43). This technology coupled with ESI and orbitrap mass spectrometers form
the basis of cutting-edge proteomics research.
The protocol for typical LCMS proteomic experiments is highlighted in Figure 1.2.
Biological samples from either cell culture or tissue extracts are lysed with mass spec friendly
detergents or other reagents, then for a “bottom-up” proteomics experiment, the proteins are
enzymatically digested with a protease to cleave the initial proteins into a mix of shorter
peptides. In order to reduce sample complexity, fractionation can be performed to either
separate groups of peptides based on a chemical property or to enrich certain types of
modifications on peptides. After some cleanup steps to remove contaminating reagents (known
as desalting in colloquial terms) the peptide samples can be dissolved and loaded onto a
UHPLC and separated on an analytical column where the effluent is electro-sprayed into the ion
optics of the mass spectrometer. Inside the mass spectrometers, survey scans of the entire mass
6
spectrum are collected and in the case of a “Data-Dependent Acquisition”, or DDA, the most
abundant ions of usually charge greater than or equal to two are selected for isolation by a
quadrupole. In the case of tandem mass spectrometry, or MS2, selected ions can then be
concentrated and sent to a fragmentation chamber where the analyte is collided with high
energy nitrogen gas in a process known as high-energy collision dissociation (HCD) in order to
fragment the peptide into a series of ions known as “b” and “y” ions which are produced by
cleavage along the peptide backbone. Other types of fragmentation are possible which result in
different types of ions being produced in the fragmentation chamber. From the collision
chamber, the fragment ions are then sent to the orbitrap for mass analysis which now will
consist of a spectrum of only the fragment ions of the initially isolated ion. This MS/MS
spectrum will contain m/z’s that can be compared to an in silico database of digested proteins
from the original organism from which the sample is derived. Matching spectra can be scored
based of closeness of matching or similarity to the predicted spectra from the databased and
then various validation strategies can be used to remove low-scoring spectra, for example by
using a reverse-decoy approach (44). In the case of post-translational modifications, mass
additions or deficits can be considered during the spectra matching process to detect whether
certain chemical groups have been added or taken away from biological processes.
Taken together LCMS based proteomics can provide a window into the world of proteins
that previous generations of scientists could have only dreamed of. It is for this ability that
LCMS proteomics is an indispensable tool for the study of proteins and other biomolecules in
the fields of biochemistry, biology, health, and disease.

7

Figure 1.2 Typical LCMS based proteomic workflow

1.3.1. Challenges of Methyl Peptide Analysis
One major challenge of analyzing methyl proteins is the low stoichiometry of methylation
with regards to the general unmethylated proteome. Later on, we show that protein methylation
occurs to a similar frequency as other post-translational modifications such as phosphorylation
which require enrichment from non-modified background peptides after protease digestion, and
as such methylated peptides require specific enrichment techniques for their respective analysis.
Methylated peptides identified by mass spectrometry are also prone to higher false
discovery rates than non-modified peptides (45). Therefore, care must be taken when evaluating
results from sequence database approaches of large-scale protein methylation studies. Possible
remedies involve stricter false discovery cutoffs or use of labelled substrates such as
13
CD3
methionine in cell culture where it is metabolically converted to
13
CD3S-adenosyl methionine
8
which is incorporated into proteins, directly labelling the PTM known as the heavy-methyl
SILAC approach (46).
An additional challenge is the isobaric nature of the two forms of arginine dimethylation
which contain the identical mass shifts and therefore pose a challenge for identification by mass
spectrometry. Even immunoaffinity enrichment techniques using antibodies made to recognize
asymmetric or symmetric dimethyl motifs are not guaranteed to enrich the correct form of
dimethylation.

9
2. Chapter 2. Deep methylation profiling of PRMT1 in 293T Cells
2.1. Objective
Protein arginine methyltransferase 1 (PRMT1) is the primary methyltransferase in
mammalian cells accounting for over 90% of endogenous methylation events (33). Therefore, we
sought to investigate the effect of PRMT1 knockdown on human 293T cells expressing a short
hairpin RNA for PRMT1. PRMT1 is a type I PRMT meaning it catalyzes the MMA forms and
ADMA forms of arginine methylation. While the strong activity of PRMT1 is documented, there
is very little data on the exact substrate specificity of PRMTs in general, with studies only able to
identify a small handful of potential targets (4), usually without the high resolution needed to
assign site specificity.
Upregulation of PRMT1 has also been associated with human cancers such as breast cancer
(47, 48), prostate cancer (49), leukemia (50), and bladder cancer (15). A major class of
methylated proteins are histones, which are involved in DNA regulation through a variety of
other PTMs such as histone acetylation and phosphorylation (51). PRMT1 catalyzes an ADMA
mark on histone H4 at position R3 (52) which is positively correlated with increased tumor grade
and recurrence (49). Due to the strong evidence of PRMT1’s role in cancer and it’s lack of
substrate methyl site localization, there is a need to provide further insight into the activity of this
interesting protein.

2.2. Materials and Methods
Cell Culture — LN229 cells and HEK 293T cells expressing short hairpin RNA (shRNA)
against PRMT1 or control were grown in DMEM media (Corning, Corning, NY) supplemented
with 10% FBS (Omega Scientific, Tarzana, CA) and 100 U/ml penicillin/streptomycin (Thermo
10
Scientific, Waltham, MA). Cells were cultured at 37 °C in humidified 5% CO2 atmosphere.
Generation of HEK 293T cells stably expressing shRNA against PRMT1 or control were
previously described (26). shPRMT1 and shControl cells were cultured with 4 ug/ml puromycin
to maintain selection.
Cell Lysate Preparation — Cells were washed with PBS, scraped, and lysed in 50 mM Tris pH
7.5, 8 M urea, 1 mM activated sodium vanadate, 2.5 mM sodium pyrophosphate, 1 mM B-
glycerophosphate, and 100 mM sodium phosphate. Protein concentrations were measured by
bicinchoninic assay. Lysates were sonicated and cleared by high speed centrifugation and then
filtered through 0.22 um filter. Proteins were reduced, alkylated, and quenched with 5 mM
dithiothreitol, 25 mM iodoacetamide, 10 mM dithiothreitol, respectively. Lysates were 4-fold
diluted in 100 mM Tris pH 8.0 and digested with trypsin at a 1:100 ratio and then quenched with
addition of trifluoroacetic acid to pH 2. Peptides were purified using reverse-phase Sep-Pak C18
cartridges (Waters, Milford, MA) and eluted with 30% acetonitrile, 0.1% TFA and then dried by
vacuum. Dried peptides were subjected to high pH strong cation exchange or antibody
immunoaffinity purification.
Immunoblot Analysis — Cells were lysed in modified RIPA buffer (50mM Tris-HCl (pH 7.5),
150 NaCl, 50 mM B-glycerophosphate, 0.5 mM NP-40, 0.25% sodium deoxycholate, 10 mM
sodium pyrophosphate, 30 mM sodium fluoride, 2 mM EDTA, 1 mM activated sodium vanadate,
20 ug/ml aprotinin, 10 ug/ml leupeptin, 1 mM DTT, and 1 mM phenylmethylsulfonyl fluoride).
Whole-cell lysates were resolved by SDSPAGE on 4–15% gradient gels and blotted onto
nitrocellulose membranes (BioRad, Hercules, CA). Membranes were blocked for 1 h in nonfat
milk, and then incubated with primary and secondary antibodies overnight and for 2 h,
respectively. Blots were imaged using the Odyssey Infrared Imaging System (LiCor, Lincoln,
11
NE). Primary antibodies used for Western blot analysis were: mono-methyl arginine (8015, Cell
Signaling Technology, Danvers, MA), asymmetric dimethyl arginine motif (13522, Cell
Signaling), symmetric di-methyl arginine motif (13222, Cell Signaling), PRMT1 (2449, Cell
Signaling), and anti-B-actin (10081–976, Proteintech, Rosemont, IL).
High pH Strong Cation Exchange (SCX) — As described previously (53), in brief, 1 mg of
digested protein was resuspended in loading buffer (60% acetonitrile, 40% BRUB (5 mM
phosphoric acid, 5 mM boric acid, 5 mM acetic acid, pH 2.5) and incubated with high pH SCX
beads (Sepax, Newark, DE) for 30 min, washed with washing buffer (80% acetonitrile, 20%
BRUB, pH 9), and eluted into five fractions using elution buffer 1 (60% acetonitrile, 40%
BRUB, pH 9), elution buffer 2 (60% acetonitrile, 40% BRUB, pH 10), elution buffer 3 (60%
acetonitrile, 40% BRUB, pH 11), elution buffer 4 (30% acetonitrile, 70% BRUB, pH 12), and
elution buffer 5 (100% BRUB, 1 M NaCl, pH 12). Eluates were dried, resuspended in 1%
trifluoroacetic acid and desalted on STAGE tips (54) with 2 mg of HLB material (Waters) loaded
onto 300 ul tip with a C8 plug (Empore, Sigma, St. Louis, MO).
Immunoaffinity Purification (IAP) — Ten milligrams of digested proteins were dissolved in
1x immunoprecipitation buffer (50 mM MOPS, 10 mM Na2HPO4, 50 mM NaCl, pH 7.2, Cell
Signaling). Modified symmetric dimethyl arginine peptides, asymmetric dimethyl arginine
peptides, and monomethyl arginine peptides were immunoprecipitated by addition of 40 ul of
PTMScan Symmetric Di-Methyl Arginine Motif Kit (13563, Cell Signaling), PTMScan
Asymmetric Di-Methyl Arginine Motif Kit (13474, Cell Signaling), and PTMScan Mono-Methyl
Arginine Motif Kit (12235, Cell Signaling), respectively. Modified methyl lysine peptides were
enriched with PTMScan Pan-Methyl Lysine Kit (14809). Lysates were incubated with PTMScan
motif kits for 2 hr at 4 °C on a rotator. Beads were centrifuged and washed two times in 1X
12
immunoprecipitation buffer followed by three washes in water, and modified peptides were
eluted with 2 x 50 ul of 0.15% TFA and desalted on STAGE tips with C18 cores (Empore,
Sigma). Enriched peptides were resuspended in 50 mM ammonium bicarbonate (Sigma) and
subjected to a second digestion with trypsin for 2 h per the manufacturer’s recommendation,
acidified with trifluoroacetic acid to pH 2 and desalted on STAGE tips.
Mass Spectrometric Analysis — All LC-MS experiments were performed on a nanoscale
UHPLC system (EASY-nLC1200, Thermo Scientific) connected to an Q Exactive Plus hybrid
quadrupole-Orbitrap mass spectrometer equipped with a nanoelectrospray source (Thermo
Scientific). Peptides were separated by a reversed-phase analytical column (PepMap RSLC C18,
2 um, 100 Å, 75 um x 25 cm) (Thermo Scientific). For high pH SCX fractions a “Short” gradient
was used where flow rate was set to 300 nl/min at a gradient starting with 0% buffer B (0.1%
FA, 80% acetonitrile) to 29% B in 142 min, then
washed by 90% B in 10 min and held at 90% B for 3. The maximum pressure was set to 1,180
bar and column temperature was constant at 50 °C. For IAP samples a “Slow” gradient was used
where flow rate was set to 300 nl/min at a gradient starting with 0% buffer B to 25% B in 132
min, then washed by 90% B in 10 min. Dried SCX fractions were resuspended in buffer A and
injected as follows, E1: 1.5 ul/60 ul, E2–5: 5 ul/6 ul. IAP samples were resuspended in 7 ul and
6.5 ul was injected. The effluent from the HPLC was directly electrosprayed into the mass
spectrometer. Peptides separated by the column were ionized at 2.0 kV in the positive ion mode.
MS1 survey scans for DDA were acquired at resolution of 70k from 350 to 1,800 m/z, with
maximum injection time of 100 ms and AGC target of 1e6. MS/MS fragmentation of the 10 most
abundant ions were analyzed at a resolution of 17.5k, AGC target 5e4, maximum injection time
13
120 ms for IAP samples, 240 ms for SCX samples, and normalized collision energy 26. Dynamic
exclusion was set to 30 s and ions with charge 1 and >6 were excluded.
Identification and Quantitation of Peptides — MS/MS fragmentation spectra were searched
with Proteome Discoverer SEQUEST (version 2.2, Thermo Scientific) against the in-silico
tryptic digested Uniprot H. sapiens database with all reviewed with isoforms (release Jun 2017,
42,140 entries). The maximum missed cleavage rate was set to 5. Trypsin was set to cleave at R
and K. Dynamic modifications were set to include mono-methylation of arginine or lysine (R/K,
+14.01565), di-methylation of arginine or lysine (R/K, +28.0313), tri-methylation of lysine (K,
+42.04695), oxidation on methionine (M, +15.995 Da, and acetylation on protein N terminus
(+42.011 Da). Fixed modification was set to carbamidomethylation on cysteine residues (C,
+57.021 Da). The maximum parental mass error was set to 10 ppm and the MS/MS mass
tolerance was set to 0.02 Da. Peptides with sequence of six to fifty amino acids were considered.
Methylation site localization was determined by ptm-RS node in Proteome Discoverer, and only
sites with localization probability greater or equal to 75% were considered. The False Discovery
Rate threshold was set strictly to 0.01 using Percolator node validated by q-value. Relative
abundances of parental peptides were calculated by integration of area under-the-curve of the
MS1 peaks using Minora LFQ node in Proteome Discoverer 2.2. The Proteome Discoverer
export peptide groups abundance values were log2 transformed, normalized to the corresponding
samples median values, and significance was determined using a permutation-based FDR
approach in the Perseus environment (55) (release 1.6.2.3) with a q-value FDR of 0.05 and S0
value of 0.5.
Methyl False Discovery Estimation — The “Decoy PSMs” export from Proteome Discoverer
2.2 was filtered for decoy methyl PSMs and the decoy q-values from the Percolator node were
14
extracted and compared with the target methyl PSM q-values. Target methyl PSMs were
removed until a 1% FDR was achieved as described (44).
Neutral Loss Identification in MaxQuant — The modifications SDMA and ADMA were
added to MaxQuant’s library with the added mass of dimethyl on arginine and the corresponding
neutral loss masses of 31.042 for SDMA and 45.058 for ADMA assigned in the “Neutral Loss”
Table in Configuration (56). The missed cleavage rate was set to 5 and all other settings were
kept unchanged. All RAW files were searched with monomethyl(K/R), ADMA, SDMA, and
oxidation of methionine as variable modifications. Carbamidomethylation was kept as a fixed
modification. Neutral losses and their masses were extracted from the msms.txt file. Only target
methyl peptides that passed the 1% Methyl FDR filter were considered for analysis. An
Andromeda cutoff score of 56 was also used to filter spectra to reduce the number of incorrect
assignments. A custom R script was used to remove neutral losses that did not have the
corresponding b/y ion present (e.g. if y6* but not y6 was present, the neutral loss was removed).
A few spectra were confirmed by manual inspection to ensure the accuracy of the Andromeda
search. For identified ADMA/SDMA neutral losses, the Andromeda output was matched to
Proteome Discoverer data by MS2 scan number.
Motif Analysis — Motifs were analyzed by MotifX (57) and MOMO from MEME suite (58) to
detect statistically significant patterns in methylation sequence data. Two sample motif analysis
was performed using Two Sample Logo (59).

15
2.3. Results
2.3.1. Orthogonal Enrichment of Methylated Peptides Cover Different
Subsets of Protein Methylome
We first optimized the LC gradients for high pH SCX and IAP to enhance the detection
of hydrophilic methylated peptides (Tables 1-2). By shortening and lengthening the gradients for
SCX and IAP, respectively, modest improvements were made in instrument time and number of
unique methyl peptides identified (Tables 3-4). Next, we applied our workflow to lysates from
293T cells expressing short hairpin RNA against PRMT1 (shPRMT1) or a negative control
(shControl) (Fig. 2.1A). Across all experiments, we identified 1,720 methylation sites on 778
proteins with a strict 1% methyl-peptide FDR. A summary of the peptide spectral matches
(PSMs) from each technique is provided in Table 5. Each technique enriched the expected type
of arginine methylation: MMA IAP identified primarily MMA peptides, ADMA IAP and SDMA
IAP identified primarily dimethyl arginine (DMA) peptides, and SCX identified primarily DMA
peptides (Fig. 2.1B). PanK and SCX both identified an even distribution of monomethyl lysine
(Kme1), dimethyl lysine (Kme2), and trimethyl lysine (Kme3) PSMs. Similar to previous
reports, we identified roughly 5 times as many methyl-arginine sites as methyl-lysine sites (60).
Notably, the overlap of unique methyl peptides enriched by SCX and IAP was relatively low for
all types of arginine and lysine methylation (Fig. 2.1C), demonstrating that these enrichment
techniques target different methyl arginine peptides. Gene ontology analysis of methyl peptides
identified by SCX and IAP demonstrated that both techniques were highly enriched for RNA
binding proteins (Fig. 2.1D), in agreement with known properties of methyl proteins (5, 7, 61–
65). In addition, IAP enriched proteins related to DNA binding and transcription factor binding,
whereas SCX enriched proteins related to nucleoside-triphosphatase activity and hydrolase
16
activity. Comparison of the number of methyl arginine sites per PSM revealed that IAP generally
enriched singly methylated peptides while SCX enriched multi-methylated peptides, with some
SCX PSMs containing up to four methylation sites (Fig. 2.1E). Taken together, this data suggests
that the usage of both IAP and SCX methods is required to achieve a more complete coverage of
the protein arginine methylome.
17

Figure 2.1 SCX and IAP enrich methyl peptides and target different subsets of protein methylome. A)
Schematic of the methyl enrichment workflow for high pH SCX and IAP (66, 67). 293T cells expressing the
negative control short hairpin RNA shControl or a short hairpin RNA against PRMT1 (shPRMT1) were lysed in 8
M urea and then digested to peptide with trypsin. Next, 1 mg and 10 mg of tryptic peptides were subjected to high
pH SCX or IAP enrichment. For SCX, five fractions eluted at increasing pH were collected. For IAP, the lysates
were sequentially incubated with the indicated IAP antibodies. Samples were then analyzed using a Q Exactive Plus
mass spectrometer. LC-MS data was searched using Proteome Discoverer, and methyl peptides were subjected to a
strict 1% methyl FDR. The number of methyl PSMs for each sample is shown in Table I. B) Number of methyl
PSMs showing the indicated type of arginine methylation for each enrichment technique. Only high confidence
18
spectra passing the 1% methyl FDR were considered. “Mixed” peptides contained a mixture of mono/di methylation
on R and mono/di/tri methylation on K on the same peptide. C) SCX and IAP identify different subsets of the
protein methylome. Overlap of (top) identified MMA peptides comparing SCX and MMA IAP, (middle) identified
DMA peptides comparing SCX and ADMA/SDMA IAPs, and (bottom) identified mono/di/tri-methyl lysine
peptides comparing SCX and PanK IAP. D) Gene ontology of methyl peptides enriched by SCX and IAPs. Unique
non-overlapping gene symbols from SCX and IAP were compared against a human background using GOrilla. All
shown ontologies had an FDR q-value < 0.01 as calculated by GOrilla (68). E) Number of methylation sites per
PSM for each methyl peptide enrichment protocol. All IAP methods primarily enriched single methylated peptides,
whereas SCX identified a larger fraction of di-, tri-, and tetra-methylated peptides.

2.3.2. Label Free Quantitation of Methyl Peptides Shows High
Reproducibility
Next, we sought to test the reproducibility of SCX and IAP methyl peptide quantitation in
293T cells. Label-free quantitation (LFQ) values of methyl peptides passing the 1% methyl FDR
were filtered to remove peptides with missing values, log2 transformed, and then normalized by
the sample median (Fig. 2.2A). Across all experiments, we quantified 943 methylation sites on
451 proteins (46% of all identified peptides), of which 262 sites were measured by 2 or more
techniques. Similar to identified methyl peptides (Fig. 2.1C), the overlap between quantified
methyl peptides was low between SCX and IAP (Fig. 2.2B). Scatter plots of LFQ values from
biological replicates showed high correlation for each methyl-peptide enrichment technique
(Pearson correlation coefficients ranging from 0.85 to 0.95, Fig. 2.2C). Together, this data
demonstrates that SCX and IAP both generate reproducible LFQ data for methylated peptides.
19

Figure 2.2 Label free quantitation of methyl peptides is highly reproducible. A) Schematic of the label free
quantitation (LFQ) workflow for methyl peptides. LFQ values were imported to Perseus where peptides with
missing values or methionine oxidation were removed. Values were log2 transformed and median normalized. A
permutation-based t test with an FDR q-value of 0.05 and S0 value of 0.5 was performed to identify significantly
changing methyl peptides. B) SCX and IAP quantify different subsets of the protein methylome. Overlap of (top)
quantified MMA peptides comparing SCX and MMA IAP, (middle) quantified DMA peptides comparing SCX and
ADMA/SDMA IAPs, and (bottom) quantified mono/di/tri-methyl lysine peptides comparing SCX and PanK IAP.
C) Scatter plots of biological replicates demonstrate high reproducibility. The log10 LFQ values between biological
replicates were plotted for each methyl peptide enrichment. Each dot is colored by the local density of points. The
number of peptides and the Pearson correlation coefficient for each comparison is shown.

20
2.3.3. Quantitative Analysis of MMA Peptides from shPRMT1 293T cells
PRMT1 has been reported to account for over 90% of ADMA methylation events in
mammalian cells (33). We therefore investigated how PRMT1 knockdown affected MMA in
293T cells. Knockdown of PRMT1 was confirmed by Western blotting (Fig. 2.3A). We also
observed a general increase in MMA levels in PRMT1 knockdown cells, consistent with other
reports (69). SCX and IAP profiling by LC-MS revealed many changing MMA peptides, with
good agreement for peptides captured by both techniques (bolded sites, Fig. 2.3B). Comparing
shControl and shPRMT1 cells using a permutation-based t-test in Perseus, we found 61
significantly increased and 58 significantly decreased MMA peptides (q < 0.05) in PRMT1
knockdown cells (Fig. 2.3C). Significantly changing MMA sites were enriched for PRMT1
interactors from the EBI database (Fisher’s Exact p-value = 0.044). Of the total 119 significantly
changing methyl peptides, 5 came from SCX enrichment, and 114 from IAP enrichment. Motif
analysis of MMA sites with a log2 fold change greater than 1.5 in either direction compared
against non-significantly changing MMA sites recapitulated the RGG motif common to methyl
arginine sites (Fig. 2.3D). A preference for serine in the -4 and -2 positions and tyrosine in the +3
position was also observed for changing MMA sites. Gene ontology analysis revealed that
significantly changing MMA peptides were enriched for mRNA metabolic process and mRNA
splicing compared to non-changing MMA peptides (Fig. 2.3E). Taken together, this data reveals
that PRMT1 knockdown dramatically reshaped the MMA proteome.
21
Figure 2.3 Quantitative analysis of MMA peptides from shPRMT1 293T cells.
A) Western blotting confirmed reduced PRMT1 expression and increased MMA levels upon PRMT1 knockdown.
293T cells expressing shControl or shPRMT1 were lysed and analyzed by Western blotting with antibodies against
either PRMT1 or MMA. Actin was used as an equal loading control. B) The MMA methylome is substantially
altered by PRMT1 knockdown. Heatmap of peptide level differences for methyl peptides captured by SCX and IAP,
sorted by gene name. Median normalized log2 LFQ values were unit normalized and colored by fold change as
indicated. Methyl peptides measured by both techniques are bolded and showed good quantitative agreement. The
column at right indicates which methyl peptide enrichment protocol identified each peptide (SCX or MMA IAP).
Bold text indicates peptides identified in both SCX and MMA IAP. C) Volcano plot of MMA peptides enriched by
SCX and IAP demonstrating 61 and 58 significantly increased and decreased methyl peptides, respectively, in
PRMT1 knockdown cells. The shape indicates which methyl peptide enrichment protocol identified each peptide.
Filled shapes indicate q-value < 0.05 by permutation t test in Perseus. Red points denote known interactors of
22
PRMT1 according to the EBI Int Act database (70), and significantly changing MMA peptides were enriched for
PRMT1 interactors (p < 0.044 by Fisher’s Exact test). D) Two sample motif analysis of changing MMA sites
recapitulated the known RGG motif of protein arginine methylation. The motif was generated using Two Sample
Logo by comparing MMA peptides with absolute value of log2 fold change > 1.5 against MMA peptides with
absolute value of log2 fold change < 1. A p value of 0.05 was used to generate the motif. E) Gene ontology analysis
of MMA peptides with absolute value of log2 fold change > 1.5 against MMA peptides with absolute value of log2
fold change < 1. D) demonstrated that changing MMA sites were enriched for proteins with mRNA metabolic
process (GO:0016071) and mRNA splicing (GO:0000398). Enriched ontologies were identified using GOrilla and
passed an FDR q-value < 0.25.

2.3.4. Characteristic Neutral Losses Enable Discrimination of SDMA and
ADMA
We next sought to investigate the effect of PRMT1 knockdown on protein arginine
dimethylation (DMA). Distinguishing the isobaric ADMA and SDMA PTMs should be possible
based on characteristic neutral losses from both dimethylarginine forms (56, 71–74). For
ADMA, the neutral loss of dimethyamine causes mass loss of 45.058 Da (Fig. 2.4A), whereas for
SDMA, the neutral loss of monomethylamine causes mass loss of 31.042 Da (Fig. 2.4B). We
automated the search for ADMA and SDMA by adding these neutral loss masses to the
Andromeda search engine in MaxQuant (75).
23

Figure 2.4 Characteristic neutral loss of methylamine and dimethylamine allows discrimination of SDMA
and ADMA spectra respectively. A) Mechanism of neutral loss for ADMA resulting in neutral loss of
dimethylamine (45.058 Da). B) Mechanism of neutral loss for SDMA resulting in neutral loss of monomethylamine
(31.042 Da). C) Analysis of synthetic ADMA and SDMA peptides from Zolg et al. (72) were searched for neutral
losses using the Andromeda search engine. The synthetic peptides contained identical sequences and a single non-C-
terminal ADMA or SDMA. Ambiguous identifications were defined as peptides who had conflicting neutral loss
assignments and whose mean Andromeda scores differed by less than 30. Andromeda successfully identified 54.9%
24
(78/142) of ADMA modifications and 73.1% (106/145) of SDMA modifications. Only one SDMA peptide was
incorrectly identified as ADMA. D) MaxQuant neutral loss search applied to quantified DMA peptides from 293T
cells expressing either shControl or shPRMT1 after enrichment by either SCX or ADMA/SDMA IAPs. Peptides
showing neutral loss from the Andromeda search in MaxQuant were matched to their corresponding peptides in the
LFQ data from Proteome Discoverer 2.2. Matching was performed by matching peptide sequence, retention time,
methyl sites, and sample origin between neutral loss data and LFQ data. E) Annotated spectra of an ADMA peptide
from SCX showing neutral losses of 45.058 for nine y ions. The inset peaks of each neutral loss fragment are shown
with some fragments showing the isotopic envelope typical of peptides. An inset table shows the masses for each y
ion and y ion neutral loss fragment, as well as the differences in mass for each pair. F) Annotated spectra of an
SDMA peptide from SCX showing neutral losses of 31.042 Da for six y ions. The inset peaks of each neutral loss
fragment are shown with some fragments showing the isotopic envelope typical of peptides. An inset table shows
the masses for each y ion and y ion neutral loss fragment, as well as the differences in mass for each pair.

We then tested the accuracy of this approach using a publicly available data set consisting
of synthetic peptides modified with either ADMA or SDMA (72) (Pride ID: PXD009449). The
synthetic peptides contained identical sequences and a single non-C-terminal ADMA or SDMA.
Andromeda successfully identified 54.9% (78/142) of ADMA modifications based on the neutral
loss of dimethylamine (Fig. 2.4C) and 73.1% (106/145) of SDMA modification based on the
neutral loss of monomethylamine, with only 1 incorrect identification (i.e., SDMA identified as
ADMA). We then applied this analysis to our DMA peptides enriched by SCX and IAP and
found between 15-37% of our quantified peptides were annotated as either ADMA or SDMA
(Fig. 2.4D). Manual inspection of identified ADMA and SDMA peptides confirmed the accuracy
of this approach (Fig. 2.4E,F). Although SCX should equally enrich both ADMA and SDMA, a
large majority of neutral loss annotated DMA peptides (86%, 228 of 265 total) were annotated as
ADMA. In addition, we found that the IAP antibodies exhibited a strong preference for their
intended targets, with 87.5% and 78.4% of quantified peptides identified by ADMA and SDMA
IAP identified as ADMA and SDMA, respectively. Thus, the neutral loss of dimethylamine and
monomethylamine can unambiguously discriminate ADMA and SDMA, respectively.

25
2.3.5. Quantitative Analysis of DMA Peptides from shPRMT1 293T Cells
We next investigated how PRMT1 knockdown affected ADMA and SDMA in 293T
cells. Immunoblotting for ADMA and SDMA on shPRMT1 293T lysates revealed a slight
decrease in ADMA methylation and an increase in SDMA methylation in PRMT1 knockdown
cells (Fig. 2.5A). SCX and IAP profiling by LC-MS revealed many significantly changing DMA
peptides, with good quantitative agreement for DMA peptides identified by both SCX and IAP
(bolded sites, Fig. 2.5B). For each methyl enrichment technique (SCX, ADMA IAP, SDMA
IAP), we compared shControl and shPRMT1 cells using a permutation-based t test in Perseus.
We found 2 significantly increased SCX peptides (Fig. 2.5C), 4 significantly increased ADMA
IAP peptides (Fig. 2.5D), and 3 significantly increased SDMA IAP peptides (Fig. 2.5E) (q <
0.05). Two methyl peptides (DHX9 R1249/ R1253/R1265 and HNRNPA3 R246) were
significantly upregulated in both ADMA and SDMA IAP data sets. SCX also enriched peptides
with both MMA and DMA modifications, creating “mixed” methyl peptides, including four
significantly changing peptides (data not shown). Motif analysis of the downregulated ADMA
peptides revealed a preference for arginine in the -2 position and leucine/aspartic acid/asparagine
in the -1 position (Fig. 2.5F). Gene ontology analysis revealed that significantly changing
ADMA peptides were enriched for nitrogen compound transport, organic substance transport,
RNA localization, and protein localization compared with the background (Fig. 2.5G).
26

Figure 2.5 Quantitative Analysis of DMA peptides from shPRMT1 293T cells. A) Immunoblot of ADMA
and SDMA on 293T cells expressing shControl or shPRMT1. 293T cells expressing shControl or shPRMT1 were
lysed and analyzed by Western blotting with antibodies against either ADMA or SDMA. ADMA levels were
decreased, and SDMA levels were increased in PRMT1 knockdown cells. Actin was used as an equal loading
control and is shown in Fig. 3A. B) The DMA methylome is substantially altered by PRMT1 knockdown. Heatmap
of dimethyl peptide levels enriched by SCX, ADMA IAP, or SDMA IAP, sorted by gene name. Median normalized
log2 LFQ values were unit normalized and colored by fold change as indicated. The columns at right indicate
27
observed neutral losses (ADMA or SDMA) and which methyl peptide enrichment protocol identified each peptide.
Bold text indicates peptides that were identified by multiple enrichment protocols. C–E, Volcano plots of DMA
peptides enriched by SCX (C), ADMA IAP (D), and SDMA IAP (E) demonstrating significantly increased DMA
peptides in PRMT1 knockdown cells. The type of neutral loss, if observed, is indicated by shape, and the filled in
points indicate q-value < 0.05 by permutation t test in Perseus. F) Two-sample motif analysis of downregulated
dimethyl peptides showing ADMA neutral loss compared with unchanging background dimethyl peptides. All
downregulated ADMA peptides from all experiments were combined for the foreground with log2 fold change <-1
and a p value < 0.1. All unchanging DMA peptides from all experiments were used as the background with a log2
fold change between 0.5 and -0.5 with a p value > 0.20. G) Gene ontology analysis revealed that proteins with
significantly changing ADMA sites were enriched for nitrogen compound transport, organic substance transport,
establishment of RNA localization, and protein localization. ADMA peptides with absolute value log2 fold change >
1 and Student’s t test p value < 0.1 were compared against ADMA peptides with absolute value log2 fold change <
0.5 and Student’s t test p value < 0.2. Enriched ontologies were identified in GOrilla and passed an FDR q-value <
0.25.

2.3.6. Lysine Methylation Is Not Affected by PRMT1 Knockdown
In addition to arginine methylation, both SCX and IAP can enrich peptides containing
methyl lysine. To test whether PRMT1 depletion affected protein lysine methylation, 293T cells
expressing shPRMT1 or the shControl were subjected to both SCX and Pan-methyl-K IAP
followed by mass spectrometry. Label-free quantitation of mono-, di-, and tri-methyl lysine by
IAP showed only 1 site significantly changed for SCX, HMGN2 K40, (Fig. 2.6A) and no
significant changes in protein lysine methylation for PanK IAP (Fig. 2.6B). Together, these data
demonstrate that PRMT1 depletion affects methylation of arginine but not lysine residues.

28
Figure 2.6 Lysine methylation is unaffected by shPRMT1 A) Volcano plot of quantified methyl lysine
peptides enriched by SCX on 293T cells expressing shControl and shPRMT1. LFQ values were log2 transformed,
median normalized, and subjected to a Perseus permutation-based t-test to asses significance with parameters q <
0.05 and S0 = 0.5. Filled symbols represent a q-value < 0.05. Only one site, HMGN2 K40, was found to
significantly change upon PRMT1 knockdown. B) Similar to A) but for PanK IAP enriched methyl lysine peptides.
No significant changes were observed for PanK IAP upon knockdown of PRMT1.

2.3.7. Integrated Analysis of Methyl-arginine Forms Reveals Novel PRMT1
Substrates and Substrate Scavenging
Because PRMT1 catalyzes both MMA and ADMA, PRMT1 substrates may exhibit both
downregulated MMA and ADMA levels in PRMT1 knockdown cells (Fig. 2.7A). However,
because other PRMT1s can also catalyze MMA, it is possible that PRMT1 targets will exhibit
downregulated ADMA but upregulated MMA levels. We therefore reasoned that integrating
29
results from MMA and DMA would enable a more comprehensive view of the PRMT1
methylome. We identified 17 methylation sites on 11 proteins that exhibited downregulated
DMA levels and confirmed ADMA neutral loss (Fig. 2.7B). Several of these methylation sites
exhibited increased MMA levels concomitant with decreased ADMA levels (e.g. EWSR1 R460).
In addition, we found 12 methylation sites on 9 proteins with upregulated DMA levels and
confirmed ADMA neutral loss (Fig. 2.7C), consistent with scavenging by other Type 1 PRMTs
in the absence of PRMT1 activity. Next, because arginine dimethylation is more likely to result
in missed tryptic cleavages (66), we reasoned that we might identify PRMT1 substrates by
examination of ADMA peptides with and without missed tryptic cleavages. Using this approach,
we identified SON R996 as a high confidence PRMT1 target. In our SCX data, SON R996 was
present in three forms: unmethylated with tryptic cleavage at R996 (i.e. LAPRPLMLASR, R996
in bold underline), monomethylated with missed cleavage at R996 (i.e. LAPRPLMLASRR), and
ADMA-modified with missed cleavage at R996 (i.e. LAPRPLMLASRR). The levels of the
unmethylated, MMA, and ADMA peptides were increased (log2 fold change +1.09), decreased
(log2 fold change -1.24), and decreased (log2 fold change -1.50), respectively, in PRMT1
knockdown cells (Fig. 2.7D). In total, we found 4 examples where identification of peptides with
and without missed cleavages enabled deeper understanding of methylation dynamics, although
SON R996 was the only putative PRMT1 target. Finally, we identified one DMA site,
HNRNPA1 R206, that was increased in abundance in SCX, ADMA IAP, and SDMA IAP data
(Fig. 2.7E) and that exhibited both ADMA and SDMA neutral losses. For this methyl peptide,
ADMA neutral losses were more frequent in shControl cells, and SDMA neutral losses were
more prevalent in shPRMT1 cells (Fig. 2.7F).

30
Figure 2.7 Integrated analysis of methyl-arginine forms reveals novel PRMT1 substrates and ADMA
substrate scavenging. A) Schematic depicting the expected trends in MMA and ADMA levels for methylation sites
targeted by PRMT1 for both MMA and ADMA methylation (top) and ADMA but not MMA methylation (bottom).
B) Integrated analysis of MMA and ADMA levels revealed novel PRMT1 substrates. Log2 fold change of
methylation levels for shPRMT1 cells compared with shControl cells are shown for different methyl peptide
enrichment protocols: DMA peptides identified by SCX, SDMA IAP, ADMA IAP, MMA peptides identified by
31
SCX, and MMA IAP. Peptides were selected based on decreased ADMA levels in one or more experiments, and the
presence of MMA data for the same methylation site. † denotes methyl peptides with confirmed ADMA neutral loss.
Bold outline indicates methyl peptides with FDR q-value < 0.1 by permutation t test in Perseus. C) Integrated
analysis of MMA and ADMA levels revealed substrate scavenging in the absence of PRMT1 activity. Log2 fold
change of methylation levels for shPRMT1 cells compared with shControl cells are shown for different methyl
peptide enrichment protocols: DMA peptides identified by SCX, SDMA IAP, ADMA IAP, MMA peptides
identified by SCX, and MMA IAP. Peptides were selected based on increased ADMA levels in one or more
experiments, and the presence of MMA data for the same methylation site. † denotes methyl peptides with
confirmed ADMA neutral loss. Bold outline indicates methyl peptides with FDR q-value < 0.1 by permutation t test
in Perseus. D) Identification of SON R996 as a PRMT1 substrate. A peptide with neutral loss confirmed ADMA
R989 was upregulated upon PRMT1 knockdown. A peptide with neutral loss confirmed ADMA R989 and MMA
R996 was downregulated significantly in the mixed SCX data set. A peptide with DMA R989, neutral loss
confirmed ADMA R996, and a missed cleavage at R996 was downregulated upon PRMT1 knockdown. This
suggests that PRMT1 knockdown reduced ADMA R996 and MMA R996, allowing tryptic cleavage, thereby
resulting in increased levels of the fully cleaved tryptic peptide with ADMA R989. E–F, HNRNPA1 R206 exists in
both ADMA and SDMA modified form and may switch from ADMA to SDMA upon PRMT1 knockdown. E) The
methyl peptide SGSGNFGGGRGGGFGGNDNFGR (DMA site underlined and italicized) was upregulated in SCX,
ADMA IAP, and SDMA IAP. The log2 fold change of shPRMT1 cells compared with shControl cells is shown,
normalized to shControl. F) Analysis of neutral loss ions demonstrated that ADMA neutral losses were primarily
identified in shControl cells, whereas SDMA neutral losses were primarily identified in shPRMT1 cells. (left) Each
bar represents a PSM with the y axis representing the number of neutral losses observed for that PSM. There were
no identified neutral losses in either ADMA IAP or SDMA IAP in the shControl cells. (right) The percentage of
ADMA and SDMA neutral losses observed for shControl and shPRMT1 cells.

2.4. Discussion and Conclusion
Despite its relevance for signal transduction, metabolism, transcription, and other cellular
phenotypes, protein arginine and lysine methylation remain understudied. This is partly because
of the inherent difficulty of enriching a small, neutral PTM and partly because methyl peptide
enrichment strategies have been less comprehensively studied than other PTMs. In this study, we
compared two methyl peptide enrichment techniques: high pH SCX and IAP (Fig. 2.1A). We
found that the two techniques were mostly orthogonal for both methyl peptide identification and
LFQ quantitation, demonstrating that comprehensive measurement of the protein methylome
requires multiple methyl peptide enrichment strategies. Although SCX and IAP enrich different
methyl peptides, both SCX and IAP peptides were enriched for the GO annotation RNA binding,
consistent with the known function of protein methylation (5, 7, 61–65) (Fig. 2.1D). One
explanation for the low overlap between SCX and IAP is the tendency of DMA to result in
32
missed tryptic cleavages (66). Because SCX enriches highly positively charged peptides, SCX
preferentially identifies multi-methylated peptides with missed cleavages and therefore more
positive charge (Fig. 2.1E). In contrast, because MMA is readily cleaved by trypsin, MMA IAP
enriches significantly more MMA peptides than SCX because these peptides are less likely to
have missed cleavages. Using these orthogonal methyl peptide enrichment techniques, we
investigated how PRMT1 knockdown remodeled the protein methylome and found significant
changes to 127 methylarginine sites (q < 0.05) on 78 proteins (Figs. 2.3, 2.5). We observed that
PRMT1 knockdown significantly affected only one lysine methylation site (Fig. 2.6), although
our data support previous observations that lysine methylation is much less abundant in vivo than
arginine methylation (66, 67). Of the significantly changing arginine methylation sites, the large
majority were MMA (119 of 127, or 93.7%), with most significantly changing MMA sites
identified by IAP rather than SCX (114 of 119, or 95.8%). Because PRMT1 catalyzes MMA
modifications, the 58 significantly downregulated MMA sites we observed in PRMT1
knockdown cells may represent PRMT1 MMA targets (Fig. 2.3C). Conversely, because
accumulation of MMA can result from inhibition of PRMT1-mediated ADMA modification, the
61 significantly increased MMA sites we observed in PRMT1 knockdown cells may represent
PRMT1 ADMA sites. Consistent with this hypothesis, significantly changing MMA sites were
enriched for known PRMT1 interactors (Fisher’s Exact p value < 0.044). In addition,
identification of the RGG motif from significantly changing MMA sites (Fig. 2.3D) further
confirmed that PRMT1 targets GAR (glycine arginine rich) motifs (9). Taken together, these
results demonstrate the PRMT1-mediated regulation of MMA in 293T cells.
Recent reports have indicated that the neutral losses of dimethylamine and methylamine
can discriminate between ADMA and SDMA, respectively (56, 71–74). Here, we have extended
33
those findings by reanalyzing data from synthetic peptides with either ADMA or SDMA
modifications (72). Although 25–45% of synthetic peptides did not generate identifiable neutral
losses (Fig. 2.4C), the accuracy of ADMA and SDMA identification was very high for spectra
with identified neutral losses: 78/78 and 106/107 peptides for ADMA and SDMA, respectively.
We subsequently applied this approach to both SCX and IAP methyl peptide enrichment
strategies. In addition to confirming the general specificity of the ADMA and SDMA IAP
antibodies, we found that SCX, a technique which should not be biased toward either ADMA or
SDMA peptides, identified 228 ADMA but only 37 SDMA peptides (Fig. 2.4D). This result
indicates ADMA may be more prevalent in mammalian cells than SDMA (6:1 ratio), consistent
with observations in mouse embryonic fibroblasts that SDMA is present in proteins at 10-fold
lower concentration than ADMA (69). Analysis of arginine dimethylation also revealed
considerable changes upon knockdown of PRMT1. In PRMT1 knockdown cells, we observed
decreased ADMA levels by immunoblot, consistent with loss of PRMT1’s type I
methyltransferase activity (Fig. 2.5A). Additionally, we observed increased SDMA levels by
immunoblot in PRMT1 knockdown cells, consistent with substrate scavenging by type II PRMTs
in the absence of PRMT1 activity. In our LC-MS data, we identified several candidates for type
II PRMT scavenging including HNRNPA1 206, FBL R45;R49, and SERBP1 R177; R181 (Fig.
2.5B). Interestingly, HNRNPA1 showed evidence of methyl switching from ADMA to SDMA.
First, levels of HNRNPA1 R206 were increased in PRMT1 knockdown cells in SCX, ADMA
IAP, and SDMA IAP data sets (Fig. 2.7E). Second, analysis of the neutral losses in each data set
revealed PRMT1 knockdown increased the percentage of SDMA neutral losses (0% in shControl
and 80% in shPRMT1 cells) and decreased the percentage of ADMA neutral losses (100% in
shControl and 20% in shPRMT1 cells) (Fig. 2.7F). Taken together, this data indicates that
34
HNRNPA1 R206 can exist as either ADMA and SDMA and suggests that PRMT1 knockdown
results in a switch from ADMA to SDMA. This finding is particularly interesting because
switching between methyl forms can affect protein function (6) and PRMT5-mediated
methylation of HNRNPA1 regulates translation mRNAs containing internal ribosome entry sites
(62). Thus, PRMT1 and the interplay between ADMA and SDMA modifications may regulate
HNRNPA1 activity.
Further, we identified high confidence PRMT1 ADMA targets by integrating different
forms of arginine methylation (e.g. MMA, SDMA, ADMA) (Fig. 2.7A). For example, PRMT1
knockdown resulted in decreased ADMA levels at EWSR1 R460 (log2 fold change -1.91, q-
value 0.14 in SCX DMA with ADMA neutral loss), suggesting that R460 may be a PRMT1
ADMA target. However, by considering that PRMT1 knockdown also significantly increased the
levels of EWSR1 R460 MMA (log2 fold change 2.91, q-value 0.09 in SCX MMA and log2 fold
change 5.61, q-value < 0.004 in MMA IAP), our confidence that EWSR1 R460 is a PRMT1
ADMA site is increased. Using this approach, we identified 17 high confidence PRMT1
substrates (Fig. 2.7B). In addition, we identified one additional high confidence PRMT1 target
by comparing peptides with and without missed cleavages. We identified three peptides with and
without missed tryptic cleavages from the protein SON that contained R996 either without
methylation or with MMA or ADMA modifications (Fig. 2.7D). Because PRMT1 resulted in
increased levels of the unmethylated peptide and decreased levels of both the MMA and ADMA
peptides, this data suggests that SON R996 is both a target of PRMT1 MMA and ADMA
modification. A similar integrative analysis yielded ADMA sites that we predict are scavenged
by other type I PRMTs in the absence of PRMT1 (Fig. 2.7C). These peptides all contained
ADMA neutral loss and showed increased ADMA and MMA levels in PRMT1 knockdown cells.
35
However, for these peptides, we cannot exclude the possibility that increased methylation is
driven by increases in total protein abundance. Taken together, our data demonstrate the value of
integrative analysis to elucidate the complex dynamics of arginine methylation. Taken together,
our data highlight that PRMT1 is likely to regulate both RNA:protein interactions and the protein
subcellular localization. Although methylated proteins are generally enriched for RNA binding
proteins (Fig. 2.1D), both significantly changing MMA sites and significantly decreased ADMA
sites in PRMT1 knockdown cells were further enriched for RNA-related processes (Figs. 2.3E
and 2.5G). In addition, of the 12 proteins on which we identified 18 high confidence PRMT1
targets, 10 are known RNA binding proteins (i.e. all proteins in Figs. 2.7B and 2.7D except
MAP3K20 and WDR70). In addition, arginine methylation can affect protein subcellular
localization, including the nucleo-cytoplasmic shuttling of RNA binding proteins (8, 76). In our
data, PRMT1 knockdown decreased MMA of DHX9 R1160, a residue that regulates the nuclear
localization of DHX9 (77). Interestingly, SDMA levels of DHX9 on the neighboring residues
R1249/R1253/R1265 were increased in PRMT1 knockdown cells, suggesting that DHX9
becomes more accessible to Type II PRMTs when localized to the cytoplasm.
In summary, our results confirm that PRMT1 regulates a substantial amount of
arginine methylation in mammalian cells. The fact that over 90% of significantly changing
methyl arginine sites are not known interactors of PRMT1 demonstrates the need for continued
comprehensive analysis of PRMTs and their substrates. This is especially relevant considering
the growing body of evidence that dysregulation of arginine methylation may contribute to
diseases including cancer (31). Our findings validate the utility of using high pH SCX and IAP
for enrichment of methyl peptides and to enhance coverage and quantitation of the methylome.
The dynamic interplay between different methylation marks highlights the need for further
36
development of methods to quantify site occupancy across all methylation forms, as has been
done for simpler PTMs including phosphorylation (58). Toward this end, improved methods to
distinguish ADMA and SDMA through fragmentation patterns will be valuable. Finally, given
our demonstration that high pH SCX and IAP are largely orthogonal, the continued incorporation
of fractionation techniques (24) and alternative methyl-peptide enrichment strategies (29–31)
will enable deeper analysis of the protein methylome.

37
3. Chapter 3. Improved Discrimination of Asymmetric and
Symmetric Arginine Dimethylation by Optimization of the
Normalized Collision Energy in Liquid Chromatography−Mass
Spectrometry Proteomics
3.1. Objective
Despite its important biological roles, arginine dimethylation remains an understudied
post-translational modification. Partly, this is because the two forms of arginine dimethylation,
asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA), are isobaric
and therefore indistinguishable by traditional mass spectrometry techniques. Thus, there exists a
need for methods that can differentiate these two modifications. Recently, it has been shown that
the ADMA and SDMA can be distinguished by the characteristic neutral loss (NL) of
dimethylamine and methylamine, respectively(56, 71–74). However, the utility of this method is
limited because the vast majority of dimethylarginine peptides do not generate measurable NL
ions. The objective of this study is to provide new mass spectrometry methods utilizing common
fragmentation modes such as normalized collision energy (NCE)(78, 79) to enhance the
generation of neutral losses of peptides containing dimethylated arginine residues.
3.2. Materials and Methods
Cell Culture — HEK 293T cells were grown in Dulbecco’s modified Eagle’s medium (Corning)
supplemented with 10% fetal bovine serum (Omega Scientific) and 100 U/mL
penicillin/streptomycin (Thermo Scientific). Cells were cultured at 37 °C in a humidified 5%
CO2 atmosphere.
Cell Lysate Preparation — Cells were prepared as described in Chapter 2.
38
High pH Strong Cation Exchange (SCX) — High pH SCX enrichment was performed as
described in Chapter 2, with the only difference being that the fifth SCX fraction was not
collected.
Immunoaffinity Purification (IAP) — IAPs were performed as described in Chapter 2, but
only the PTMScan asymmetric dimethyl arginine motif kit was used (13474, Cell Signaling).
Mass Spectrometric Analysis — All liquid chromatography (LC)−MS experiments were
performed on a nanoscale UHPLC system (EASY-nLC1200, Thermo Scientific) connected to a
Q Exactive Plus hybrid quadrupole-Orbitrap mass spectrometer equipped with a
nanoelectrospray source (Thermo Scientific). Peptides were separated by a reversed-phase
analytical column (PepMap RSLC C18, 2 μm, 100 Å, 75 μm × 25 cm) (Thermo Scientific). For
high pH SCX fractions, the flow rate was set to 300 nL/min at a gradient starting with 0% buffer
B (0.1% FA, 80% acetonitrile) to 29% B in 142 min, then washed by 90% B in 10 min, and held
at 90% B for 3 min. The maximum pressure was set to 1180 bar, and the column temperature
was constant at 50 °C. For immunoaffinity purification (IAP) samples, the flow rate was set to
300 nL/min at a gradient starting with 0% buffer B to 25% B in 132 min and then washed by
90% B in 10 min. Dried SCX fractions were resuspended in buffer A and injected as follows: E1:
1.5 μL/60 μL; E2-4: 5μL/6 μL. IAP samples were resuspended in 7 μL, and 6.5 μL was injected.
The effluent from the high-performance liquid chromatograph was directly electrosprayed into
the mass spectrometer. Peptides separated by the column were ionized at 2.0 kV in the positive
ion mode. MS1 survey scans for data-dependent acquisition were acquired at a resolution of 70k
from 350 to 1800 m/z, with a maximum injection time of 100 ms and an automatic gain control
(AGC) target of 1 × 106. MS/MS fragmentation of the 10 most abundant ions were analyzed at a
resolution of 17.5k, AGC target of 5 × 104, and a maximum injection time of 120 ms for IAP
39
samples and 240 ms for SCX samples, and normalized collision energies (NCE) of 26, 30, 32,
34, 38, 42, 46, and 50 were tested. Dynamic exclusion was set to 30 s and ions with charge 1
and >6 were excluded. The MS proteomics data have been deposited to the ProteomeXchange
Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository with
the dataset identifier PXD017193.
Identification of Peptides — MS/MS fragmentation spectra were searched with Proteome
Discoverer SEQUEST (version 2.2, Thermo Scientific) against the in-silico tryptic digested
Uniprot H. sapiens database with all reviewed with isoforms (release Jun 2017, 42,140 entries).
The maximum missed cleavage rate was set to 5. Trypsin was set to cleave at R and K. Dynamic
modifications were set to include mono-methylation of arginine or lysine (R/K, +14.01565), di-
methylation of arginine or lysine (R/K, +28.0313), tri-methylation of lysine (K, +42.04695),
oxidation on methionine (M, +15.995 Da, and acetylation on protein N terminus (+42.011 Da).
Fixed modification was set to carbamidomethylation on cysteine residues (C, +57.021 Da). The
maximum parental mass error was set to 10 ppm and the MS/MS mass tolerance was set to 0.02
Da. Peptides with sequence of six to fifty amino acids were considered. Methylation site
localization was determined by ptm-RS node in Proteome Discoverer, and only sites with
localization probability greater or equal to 75% were considered. The False Discovery Rate
threshold was set strictly to 0.01 using Percolator node validated by q-value.
Methyl False Discovery Estimation — Methyl false discovery estimation was performed as
described in Chapter 2.
Neutral Loss Identification in MaxQuant — Neutral loss identification was performed as
described in Chapter 2.

40
3.3. Results
3.3.1. Higher NCE Improves Discrimination of ADMA and SDMA in
Synthetic Peptides
Fragmentation of DMA-containing peptides can result in NL of dimethylamine and
monomethylamine from ADMA and SDMA, respectively (Fig. 3.1A). These NLs generate mass
shifts of 31.042 Da for SDMA peptides and 45.058 Da for ADMA peptides. To assess the effect
of NCE on NL generation, we analyzed a set of 140 synthetic peptides containing either ADMA
or SDMA that had been fragmented at 25, 30, or 35 NCE in the HCD cell of an Orbitrap Fusion
Lumos spectrometer (PRIDE ID: PXD009449)(72). Although the total number of DMA PSMs
was not significantly changed by increasing NCE, the number of PSMs with either ADMA or
SDMA NL strongly increased at higher NCE (Fig. 3.1B). In addition, the total number of unique
ADMA and SDMA peptides with NL substantially increased with higher NCE (Fig. 3.1C).
Notably, synthetic SDMA peptides generated more NL than synthetic ADMA peptides. This is
potentially because there are two bonds in SDMA whose fragmentation produces the NL but
only one such bond in ADMA. For these synthetic peptides, even NCE 35 did not appear to have
reached a maximum in terms of NL generation.
41

Figure 3.1 Higher NCE improves the generation of NL ions in synthetic ADMA and SDMA peptides. (A)
Mechanism of NL of dimethylamine and monomethylamine from ADMA and SDMA, respectively. (B) Total
number of DMA PSMs from a dataset of synthetic ADMA and SDMA peptides. The proportion of PSMs displaying
either an ADMA or SDMA NL in the HCD cell of an Orbitrap Fusion Lumos is shown for each NCE. (C) Total
number of unique peptides with assignable ADMA or SDMA NL for each NCE. (D) Violin plot of the number of
NL ions present in DMA PSMs for each NCE. Each point represents a PSM with a corresponding number of NL
ions observed in that spectra.

In addition, higher NCE values significantly increased the number of NL ions observed per
spectra for both synthetic ADMA and SDMA peptides (Fig. 3.1D). Comparison of the
Andromeda scores for peptides identified at all three NCEs revealed significantly higher scores
for spectra collected at higher NCE (Fig. 3.2). Increased Andromeda scores at higher NCE likely
arise because Andromeda considers the number of NLs in calculating the score for each PSM.
Taken together, this data supports that increasing NCE improves the ability to distinguish the
42
isobaric isomers ADMA and SDMA by generating the characteristic NL ions without sacrificing
peptide-sequencing capabilities.

Figure 3.2 Higher NCE improves andromeda scores of synthetic ADMA and SDMA peptides. Boxplots of
the andromeda scores of synthetic dimethyl peptides fragmented at each NCE energy. The difference in scores at 25
and 30 were significant by Student’s paired t-test with p < 0.038 for ADMA and p < 0.006 for SDMA. MaxQuant
was configured to consider neutral losses as part of its scoring algorithm, which occur more frequently at higher
NCEs.

3.3.2. Optimization of NCE Using Endogenous DMA Peptides
Next, we sought to investigate the effect of higher NCE on NL generation on endogenous
DMA peptides enriched from whole cell lysates. We thus enriched methyl-peptides from 293T
whole cell lysates using SCX chromatography at high pH(66, 80). This antibody-free approach
enriches methyl-peptides based on the increased positive charge of peptides with missed trypsin
cleavages at methylarginine. Analyzing the third SCX elution fraction at NCE values ranging
from 26 to 42 on a Q Exactive Plus mass spectrometer (Fig. 3.3A), we found that the number of
DMA PSMs was substantially increased at NCE values of 30 and 34 compared to the standard
value of NCE 26 (Fig. 3.3B). In addition, the number of PSMs with ADMA or SDMA NL
greatly increased as the NCE was increased from 26 to 30 and 34. However, at NCEs greater
30 25
160
120
80
40
Andromeda Score
HCD NCE Energy
25 35 30 35
p < 0.038
p < 0.006
Synth ADMA Synth SDMA A
43
than 34, both the number of identified DMA peptides and the number of PSMs with ADMA or
SDMA NL were decreased.
In addition, the number of unique DMA peptides showing ADMA NL increased as the
NCE was increased from the standard value of 26 (Fig 3.3C). The number of unique DMA
peptides with SDMA NL was also slightly increased at NCE 30 compared to the standard NCE
value of 26 (Fig. 3.3D). Notably, the number of peptides with SDMA NL was ∼8 times lower
than the number with ADMA NL, even though synthetic SDMA peptides generated more NL
than synthetic ADMA peptides (Fig. 3.1C). This likely reflects the lower abundance of SDMA
compared to ADMA in vivo (69). As with synthetic DMA peptides, the number of NL ions per
spectra increased up to a maximum at 34 NCE and began diminishing past 34 NCE (Fig. 3.3E).
Last, we estimated the methyl-FDR for each NCE. The percolator q-value cutoff required to
reach a 1% methyl FDR increased as the NCE increased up to 42 NCE, indicating that less
methyl decoy spectra were identified at higher NCEs. This increase in the q-value cutoff also led
to a modest increase in the number of MMA peptides identified (Fig. 3.4). Thus, higher NCE not
only improves the chances of NL observation but also reduces the methylpeptide FDR, thereby
increasing the total number of identified DMA peptides.
44

Figure 3.3 Optimization of NCE for discrimination of ADMA and SDMA in endogenous methyl-peptides. (A)
Schematic of high pH SCX enrichment of methyl-peptides from 293T cells and analysis of elution fraction three by
LC−MS at increasing NCE. (B) Total number of DMA PSMs from fraction three of SCX. PSMs with assignable
ADMA or SDMA NLs are shown. (C) Total number of unique ADMA peptides from fraction three of SCX
identified at each NCE. (D) Total number of unique SDMA peptides from fraction three of SCX identified at each
NCE. (E) Violin plot of the number of NL ions present in DMA PSMs at each NCE level from fraction three of
SCX. Each point represents a PSM with a corresponding number of NL ions within that spectra.

We note that the optimum NCE value for the identification and NL generation of
endogenous methyl-peptides in this single SCX elution fraction was ∼32. Above NCE 32, the
collision energy is likely over-fragmenting the peptides, leaving too few ions for peptide
identification. In contrast, with synthetic DMA peptides, the number of peptides with observed
NL had not saturated even at NCE 35 (Fig. 3.1B). This discrepancy between endogenous and
45
synthetic peptides may occur because of the different MS platforms used (Orbitrap Fusion
Lumos v. Q Exactive Plus) or because of the lower abundance and higher sample complexity of
endogenous methyl-peptides. It is important to note that collision energies in this study are from
a QE Plus and therefore instrument specific. Optimized collision energies will need to be
empirically established for other models of mass spectrometers with different collision cells.
Regardless, the fact that previous methyl-proteomics studies used an NCE of 26 (66, 80)
supports that these studies missed identification and assignment of ADMA/SDMA on the basis
of NL for many DMA peptides.
Figure 3.4 Increased identification of MMA peptides at optimized NCE. A) Total MMA peptide spectrum
matches (PSMs) from all SCX fractions fragmented at 26 and 32 NCE. Mean +/- standard deviation of two
biological replicates is shown. Methyl peptides were removed until a 1% methyl-FDR was achieved for each
fragmentation energy. B) Unique MMA peptides from all SCX fractions fragmented at 26 and 32 NCE. Only
peptides with MMA sites and without DMA sites were counted in panels A & B.

3.3.3. Optimized NCE Improves Assignment of ADMA and SDMA Spectra in
Orthogonal Methyl-Peptide Enrichment Techniques
We next tested whether the optimized NCE value of 32 could improve methyl-peptide
identification and ADMA/SDMA assignment in two orthogonal methyl-peptide enrichment
strategies, high pH SCX and ADMA IAP. We purified methyl-peptides from 293T cell lysates
using both strategies and then analyzed samples at both standard NCE 26 and optimized NCE 32
46
on a Q Exactive Plus (Fig. 3.6A). For methyl-peptides purified by high pH SCX, the number of
unique DMA peptides identified at NCE 32 increased by ∼20% relative to NCE 26.
Additionally, the number of unique DMA peptides with assignable NL increased ∼70% at NCE
32 compared to NCE 26 (Fig. 3.6B). This increase was primarily driven by identification of
ADMA NL. Notably, the Andromeda score and delta scores for methylpeptides with NL was
increased at higher NCE (Fig. 3.5). Methyl localization values were unchanged by NCE (data not
shown).

Figure 3.5 Increased Andromeda scores at higher NCE for peptides with and without neutral loss. Boxplot of
Andromeda scores of dimethyl arginine PSMs fragmented at the specified NCE. Points are colored by whether the
spectrum contained neutral losses of ADMA/SDMA (red) or not (blue). Data from SCX fraction three with two
biological replicates for each NCE.

For methyl-peptides purified by ADMA IAP, increasing the NCE from 26 to 32 did not
significantly increase the total number of unique DMA peptides identified. However, increasing
the NCE did result in a ∼350% increase of DMA peptides with assignable ADMA NL (Fig.
3.6C). Only one SDMA NL was observed in these ADMA IAP samples. Notably, there are still
∼25 and ∼50% of peptides in SCX and IAP, respectively, which were not annotated with NL.
47
Thus, there exists further need to optimize methods for discrimination of ADMA and SDMA by
MS. Regardless, increasing the NCE from 26 to 32 resulted in increases of ∼125% and ∼17%
for NL annotation and methyl-peptide identification. Similar results for methyl-peptide
identifications and NL at higher NCE were obtained when using alternate search engines
MSAmanda and MSFragger (data not shown), although only MSAmanda was able to consider
ADMA/SDMA NL in the search parameters.
Figure 3.6 Comparison of standard and optimized NCE for high pH SCX and ADMA immunoaffinity methyl-
peptide purification. (A) 293T cells were lysed, digested, and subjected to high pH SCX enrichment and ADMA
immunoaffinity enrichment. Enriched SCX fractions and ADMA IAP elution were run at a standard NCE of 26 and
an “optimum” NCE of 32. (B) Number of unique DMA peptides identified across all high pH SCX fractions at 26
and 32 NCE. Peptides with assignable NL are shown. (C) Number of unique DMA peptides identified in the ADMA
IAP at 26 and 32 NCE. Peptides with assignable NL are shown. One SDMA NL peptide was identified in the
ADMA IAP experiment, but this peptide was not included in the chart for clarity.

3.3.4. Optimized NCE Improves Confidence of ADMA Assignment through
Higher Occurrence of NL across Ion Series
DMA spectra that contain only a single NL may decrease the reliability of ADMA and
SDMA assignment by NL. In addition, false identification of a single NL ion may be increased
because NL ions are typically much less abundant than b and y ions in the HCD spectra. We thus
48
compared the MS2 spectra of the same DMA peptide collected at 26 and 32 NCE (TAF15 R206,
peptide GPMTGSSGGDRGGFK where R indicates the DMA-modified residue). At 26 NCE,
GPMTGSSGGDRGGFK generated only one identifiable NL (y11 +) (Fig. 3.7). However, at 32
NCE, the same peptide generated seven identifiable NL ions, all on y ions. As such, the
likelihood of correct ADMA assignment for this peptide is highly increased. Furthermore, the
generation of multiple NL ions can improve the localization of ADMA/SDMA on peptides that
contain multiple dimethyl arginine residues (Fig. 3.8). This is especially beneficial for methyl-
peptides enriched by high pH SCX, which frequently contain missed cleavages and multiple
methylation sites within the same peptide (80).
Figure 3.7 Higher NCE improves occurrence of NL across ion series, improving the confidence of ADMA
identification. Two MS2 spectra are shown for the same peptide fragmented at 26 NCE (above) and 32 NCE
(below). The b and y ions are highlighted in red and blue, respectively, and the NL ions are in black. The inset table
shows the difference of the mass-to-charge ratios of the y+ ions and their corresponding NL ions which is equal to
the mass loss of dimethylamine (45.058 Da). The displayed spectra have similar intensities, reducing the effect from
higher ion loading in the collision chamber.
49

Figure 3.8 Optimized NCE enables ADMA/SDMA assignment of a doubly dimethylated peptide. A)
Mechanism of double neutral loss where a peptide harboring both ADMA and SDMA can simultaneously exhibit
neutral loss of the dimethylamine from ADMA and monomethylamine from SDMA. The resulting deficit mass of
the ion is 76.1 Da which, if observed indicates a “mixed” peptide that must include both modifications. B) A real
peptide with two dimethyl sites was fragmented at 26 and 32 NCE. At 26 NCE, only single neutral loss ions were
observed (b6, b9, and b13), and we were unable to assign either dimethyl site since both sites could generate a loss
of 31 Da. At NCE 32, the same peptide showed a double neutral loss occurring on b4 and b15 with a loss of 76 Da.
This information combined with the neutral loss of b2 (45 Da) allowed assignment of R32 as ADMA and R34 as
SDMA.

3.4. Conclusion
We have demonstrated that using higher NCE during HCD fragmentation improves the
assignment of NL for DMA peptides, thereby enabling improved discrimination of the isobaric
modifications ADMA and SDMA. In addition, increased NCE improves the total number of
DMA peptide identifications in the two methyl-peptide enrichment strategies, high pH SCX and
S R G R G F Q F V S S S L P D I C Y R
S R G R G F Q F V S S S L P D I C Y R
S(DMA)??
S(DMA)??
CNBP (P62633-2)
R32 and R34
from SCX Fraction 2
ADMA
SDMA
b6* =
31Da
b9* =
31Da
b13* =
31Da
b2* =
45Da
b4* =
76Da
b15* =
76Da
26 NCE
One or both
sites is SDMA
from single
neutral loss
Collision
Energy
32 NCE
R32 ADMA
R34 SDMA
from double
neutral loss
at higher NCE
O
O
NH
HN
N NH2
H 3 C H 3 C
O
O
HN
HN
HN NH
H 3 C CH 3
O
O
NH
ADMA SDMA
R R G
O
O
HN
HN
HN NH
H 3 C
N
H
CH 3
Dimethylamine
neutral loss
(-45.058 Da)
H 3 C
CH 3
O
O
N
H
HN
N NH2
H 3 C H 3 C
H 2 N
CH 3
Monomethylamine
neutral loss
(-31.042 Da)
O
O
NH
mass(y
x
mass(y
x
*) = (M - 76.1)
HCD Fragmentation
*) = (M - 45.058 - 31.042)
Both ADMA and SDMA Modifications Present
A
B
50
IAP. Given the emerging evidence that ADMA and SDMA differentially regulate biological
functions, this study demonstrates a simple method to improve proteomic studies of protein
arginine dimethylation by MS (81).

51
4. Chapter 4. CARM1 Substrate Identification in Mouse Quadricep
Muscle
*Note – This Chapter is the result of a collaborative effort with Dr. Vladimir Ljubicic and
Tiffany VanLieshout of the Department of Kinesiology of McMaster University, Canada.
4.1. Objective
The methyltransferase CARM1 (aka PRMT4) has received attention as an important enzyme for
regulating transcription/ RNA processing (82), cancer development (83), and more recently in
the maintenance of skeletal muscle cells (84). The objective of this work is to investigate the
effect of CARM1 skeletal muscle KO (85) on methylated proteins of mouse skeletal muscle
tissue. Using optimized methyl-peptide analysis procedures developed in previous chapters, here
we seek to characterize putative substrates of CARM1 which deposits the MMA and ADMA
mark on substrate proteins. Mice made deficient in CARM1 using a Cre/loxP system revealed
that the absence of CARM1 mitigated denervation-induced atrophy in the extensor digitorum
longus muscle after 7 days of disuse (85). Identifying differentially methylated substrates of
CARM1 in skeletal muscle could shed light on the exact proteins involved in skeletal muscle
response to atrophy.
4.2. Materials and Methods
Whole Cell Lysate Preparation — The muscle tissues were lysed by probe sonication in 8 M
urea, 50 mM Tris-HCl pH 7.5, 1 mM activated sodium vanadate, 2.5 mM sodium pyrophosphate,
1 mM β-glycerophosphate and 100 mM sodium phosphate. Insoluble cell debris were filtered by
0.22 µm syringe filter. Protein concentration was measured with a BCA assay (Pierce, PI23227).
Lysates were reduced with 5 mM DTT, alkylated with 25 mM iodoacetamide, quenched with 10
52
mM DTT, and acidified to pH 2 with 5% trifluoracetic acid. Proteins were then digested to
peptides using a 1:100 trypsin-to-lysate ratio by weight. Tryptic peptides were desalted by
reverse phase C18 StageTips and eluted with 30% acetonitrile. One milligram of peptide was
saved for high pH SCX enrichment and five milligrams were saved for immunoaffinity
enrichment. The eluents were vacuumed dried, and 750 ng/injection was submitted to LC-MS.
High pH Strong Cation Exchange (SCX) — High pH SCX enrichment was performed as
described in Chapter 2, with the only difference being that the fifth SCX fraction was not
collected.
Immunoaffinity Purification (IAP) — Five milligrams of digested proteins were dissolved in
1x immunoprecipitation buffer (50 mM MOPS, 10 mM Na2HPO4, 50 mM NaCl, pH 7.2, Cell
Signaling). Modified symmetric dimethyl arginine peptides, asymmetric dimethyl arginine
peptides, and monomethyl arginine peptides were immunoprecipitated by addition of 40 ul of
PTMScan Symmetric Di-Methyl Arginine Motif Kit (13563, Cell Signaling), PTMScan
Asymmetric Di-Methyl Arginine Motif Kit (13474, Cell Signaling), and PTMScan Mono-Methyl
Arginine Motif Kit (12235, Cell Signaling), respectively. Lysates were incubated with PTMScan
motif kits for 2 hr at 4 °C on a rotator. Beads were centrifuged and washed two times in 1X
immunoprecipitation buffer followed by three washes in water, and modified peptides were
eluted with 2 x 50 ul of 0.15% TFA and desalted on STAGE tips with C18 cores (Empore,
Sigma). Enriched peptides were resuspended in 50 mM ammonium bicarbonate (Sigma) and
subjected to a second digestion with trypsin for 2 h per the manufacturer’s recommendation,
acidified with trifluoroacetic acid to pH 2 and desalted on STAGE tips.
Mass Spectrometric Analysis — All LC-MS experiments were performed on a nanoscale
UHPLC system (EASY-nLC1200, Thermo Scientific) connected to an Q Exactive Plus hybrid
53
quadrupole-Orbitrap mass spectrometer equipped with a nanoelectrospray source (Thermo
Scientific). Peptides were separated by a reversed-phase analytical column (PepMap RSLC C18,
2 μm, 100 Å, 75 μm × 25 cm) (Thermo Scientific). For whole cell lysates flow rate was set to
250 nl/min at a gradient from 3% buffer B (0.1% formic acid, 80% acetonitrile) to 38% B in 110
min, followed by a 10-min washing step to 85% B. For high pH SCX the flow rate was set to
250 nl/min at a gradient starting with 0% buffer B to 29% B in 142 min, then washed by 90% B
in 10 min, and held at 90% B for 3. The maximum pressure was set to 500 bar and column
temperature was constant at 55 °C. For IAP samples the flow rate was set to 250 nl/min at a
gradient starting with 0% buffer B to 25% B in 132 min, then washed by 90% B in 10 min. Dried
SCX fractions were resuspended in buffer A and injected as follows, E1: 1.5 μl/60 μl, E2–4: 5
μl/6 μl. IAP samples were resuspended in 7 μl and 6.5 μl was injected. The effluent from the
HPLC was directly electrosprayed into the mass spectrometer. Peptides separated by the column
were ionized at 2.0 kV in the positive ion mode. MS1 survey scans for DDA were acquired at
resolution of 70k from 350 to 1,800 m/z, with maximum injection time of 100 ms and AGC
target of 1e6. MS/MS fragmentation of the 10 most abundant ions were analyzed at a resolution
of 17.5k, AGC target 5e4, maximum injection time 120 ms for IAP samples, 240 ms for SCX
samples, 65 ms for whole cell lysate samples, and normalized collision energy 32. For whole cell
lysates the normalized collision energy was set to 26. Dynamic exclusion was set to 30 s and ions
with charge 1 and >6 were excluded.
Identification and Quantification of Peptides — MS/MS fragmentation spectra were searched
with Proteome Discoverer SEQUEST (version 2.2, Thermo Scientific) against the in-silico
tryptic digested Uniprot Mus Musculus database with all reviewed with isoforms (release Jun
2017). The maximum missed cleavage rate was set to 4. Trypsin was set to cleave at R and K.
54
Dynamic modifications were set to include mono-methylation of arginine or lysine (R/K,
+14.01565), di-methylation of arginine or lysine (R/K, +28.0313), tri-methylation of lysine (K,
+42.04695), oxidation on methionine (M, +15.995 Da, and acetylation on protein N terminus
(+42.011 Da). Fixed modification was set to carbamidomethylation on cysteine residues (C,
+57.021 Da). The maximum parental mass error was set to 10 ppm and the MS/MS mass
tolerance was set to 0.02 Da. Peptides with sequence of six to fifty amino acids were considered.
Methylation site localization was determined by ptm-RS node in Proteome Discoverer, and only
sites with localization probability greater or equal to 75% were considered. The False Discovery
Rate threshold was set strictly to 0.01 using Percolator node validated by q-value. Relative
abundances of parental peptides were calculated by integration of area-under-the-curve of the
MS1 peaks using Minora LFQ node in Proteome Discoverer 2.2. The Proteome Discoverer
export peptide groups abundance values were log2 transformed, normalized to the corresponding
samples median values, and significance was determined using a t-test permutation-based
approach in the Perseus environment (release 1.6.2.3) to generate p-values, followed by a
Benjamini-Hochberg correction in R to generate FDR-adjusted p-values. For FDR-adjusted p
values < 0.05 peptide abundance values were unit normalized and imported to Morpheus (86) for
display on heatmaps. When possible protein abundance values from the whole cell lysate
experiments were log2 transformed, median normalized and were subtracted from the
corresponding peptide normalized values for each respective sample to correct for changes in
total protein abundance.
Methyl False Discovery Estimation — Methyl false discovery estimation was performed as
described in Chapter 2.
55
Neutral Loss Identification in MaxQuant — Neutral loss identification was performed as
described in Chapter 2.
Motif Analysis — Two sample motif analysis was performed using Two Sample Logo (59).
Gene Ontology — Gene ontologies were performed using PantherDB (87) with a background
mouse proteome for biological process, molecular function, and cellular component ontologies.

4.3. Results
4.3.1. Arginine Methylome Profiling of Mouse Skeletal Muscle
To understand the role of CARM1 in skeletal muscle arginine methylation, we first sought
to define the arginine methylome in this tissue. We employed a previously optimized workflow
(80) to enrich methyl peptides from mouse quadriceps (QUAD) muscles using complementary
immunoaffinity purification (IAP) and high pH strong cation exchange (SCX) techniques followed
by liquid chromatography-mass spectrometry (Fig. 4.1A). Analysis of muscles from WT mice
identified over 1,150 methylation sites on 313 proteins with a strict 1% methyl-peptide FDR. GO
analyses revealed that methylarginine peptides were enriched for RNA binding, mRNA processing,
muscle system processes, and actin filament binding (Fig. 4.1B). Analysis of IAP and SCX data
revealed that each method enhanced the expected type of arginine methylation (Fig. 4.1C). In
addition, there was little overlap of unique methyl peptides refined by IAP and SCX (Fig. 4.1D),
highlighting the importance of employing both enrichment techniques. IAP and SCX generally
enriched singly methylated peptides, with some IAP-enriched peptides containing up to 4
methylation sites (Fig. 4.1E).
We also observed that 5.6% of arginine residues from arginine methylated proteins were
methylated. In comparison, 6.9%, 7.2%, and 12.7% of serine, threonine, and tyrosine residues,
56
respectively, are phosphorylated (88), 10.3% of lysine residues are ubiquitinated (89), and 6.9%
of arginine residues are ADP-ribosylated (90) on proteins that bear those marks in skeletal
muscle (Table 6). Neutral loss analysis of WT dimethylarginine peptides showed an approximate
6:1 ratio of ADMA to SDMA modifications with a small fraction of peptides bearing both
modifications within the same amino acid sequence (Fig. 4.1F). This ratio of ADMA:SDMA is
consistent with previous reports of 6:1 (80) and 10:1 (69). Collectively, these data suggest that
arginine methylation in skeletal muscle is observed with a similar prevalence to that of the much
more comprehensively studied serine and threonine phosphorylation, as well as lysine
methylation.

57
Figure 4.1 Arginine methylproteomic profiling of skeletal muscle. (A) Overview of methylarginine
proteomic workflow. Mouse quadriceps muscles from WT animals were subject to high pH strong cation exchange
(SCX) and immuno-affinity purification (IAP) to enrich for peptides containing symmetric dimethylarginine
(SDMA), asymmetric dimethylarginine (ADMA), and monomethylarginine (MMA). Samples were then analyzed
using liquid chromatography-mass spectrometry (LC-MS/MS). Identified spectra were corrected to a 1% methyl
false discovery rate (FDR) and annotated for neutral loss to discriminate ADMA from SDMA sites. n = 6. (B) Gene
ontology (GO) of protein accessions from methylarginine peptides compared to the mouse background proteome,
separated by enrichment method (sorted by Biological Process and Molecular Function). Ontologies passed an FDR
threshold < 0.25. (C) Peptide spectral matches (PSMs) of each methyl type are shown for each enrichment
experiment. High confidence spectra passing the 1% methyl FDR were considered. Arginine methylation type is
58
organized into dimethylarginine (DMA), MMA, or mixed. Mixed peptides contained a mixture of mono/di
methylation on arginine (R) and mono/di/trimethylation on lysine (K) on the same peptide. Values shown are the
sum of identified PSMs from n = 6 samples. (D) Overlap of identified methylarginine peptides from each
enrichment method. (E) Number of PSMs with n methyl sites per spectrum. (F) Pie chart of dimethylarginine
peptides showing either ADMA, SDMA, or mixed neutral loss from WT samples.

4.3.2. CARM1-mediated Arginine Monomethylation in Skeletal Muscle
Following this we explored how the removal of CARM1 in skeletal muscle impacts the
arginine methylome in this tissue. In the TA muscle we found that global MMA marks were
reduced (-10%; p < 0.05) in mKO muscle relative to WT (Fig. 4.2A,B), however the specific
unmethylated form of SmB was similar between genotypes. CARM1 mKO had no impact on
PRMT1, -5, and -6 content in skeletal muscle. In contrast, PRMT7 expression was significantly
upregulated (+45%) in muscles from mKO mice compared to their WT littermates. We then
quantified the specific impact of CARM1 mKO on the arginine methylome in QUAD muscles of
WT and CARM1 mKO animals using our comprehensive LC-MS proteomics approach. We
found 16 MMA peptides that were significantly increased and 20 MMA peptides that were
significantly decreased in mKO muscles (FDR-corrected p-value < 0.05) (Fig. 4.2C,D). When
possible, we normalized the abundance of methylarginine peptides to total protein levels.
Notably, we were unable to normalize any upregulated MMA peptides to total protein levels, so
we cannot exclude that changes in total protein abundance caused the upregulation in MMA.
Motif analysis uncovered a [P/A]RFVT sequence that was enriched in MMA sites decreased in
CARM1 mKO muscles, whereas the MMA sites that were unchanged with CARM1 mKO were
enhanced for the RGG motif commonly found in methylarginine peptides (Fig. 4.2E). GO
analysis revealed that MMA peptides that were significantly decreased in mKO animals were
enriched for several muscle terms, including sarcomere organization, muscle cell development,
myofibril assembly, sarcomere localization, and Z-disc localization (Fig. 4.2F).
59
Figure 4.2 Monomethyl arginine analysis of CARM1 mKO muscle. (A) Representative Western blots for
monomethylarginine (MMA), unmethylated form of SmB (SmBme0), PRMT1, PRMT5, PRMT6, and PRMT7 in
the TA muscles from WT and mKO mice. The ponceau stain demonstrates equal loading while approximate
molecular weights (MWs) are to the right of each image. (B) Graphical summary of MMA, SmBme0, PRMT1,
PRMT5, PRMT6, and PRMT7 protein levels in mKO relative to WT mice. Bars indicate group means, whiskers
represent SEMs, points illustrate individual results, and the dotted line denotes 1 (or WT level). Statistical analysis
was completed using a student’s t-test. *, p < 0.05 vs. WT. n = 29 – 37. (C) Heatmap of MMA peptide level
differences from methyl sites passing an FDR-corrected p-value < 0.05. Raw label-free quantitation (LFQ) values
were log2 transformed, median normalized and sites were normalized to total protein level when possible. T-tests
were performed in Perseus and sites were unit normalized for display on the heatmap. The enrichment technique for
each site is shown in a color column on the right. n = 6 for each genotype. (D) Volcano plot of MMA sites
comparing CARM1 mKO to WT. Sites enriched by IAP and SCX are represented by triangles and squares,
60
respectively. Sites with an FDR-corrected q-value < 0.05 are filled in black. Sites that are normalized to the protein
level are outlined in red. (E) Motif enrichment analysis of MMA sites downregulated in CARM1 mKO muscle.
Downregulated MMA sites (log2 fold change [log2FC] < 0 and q < 0.05) were compared against un-changing MMA
sites (FDR-corrected p-value > 0.05) using TwoSampleLogo. Enriched amino acids appear in the upper row and de-
enriched amino acids appear in the lower row. The changes to the motifs as the p-value cutoff is increased is shown.
(F) Gene ontology (GO) analysis of depleted MMA protein accessions compared to mouse proteome. Displayed
ontologies pass an FDR threshold of FDR-corrected p-value < 0.25. BP = biological process, CC = Cellular
Component.

4.3.3. Quantitative and Integrative Analysis of CARM1-mediated Protein
Arginine Methylation in Skeletal Muscle
Next, we examined protein dimethylarginine (DMA) including both ADMA and SDMA
marks. ADMA levels of bona fide CARM1 targets (91, 92) BRG1-associated factor 155
(BAF155) and polyadenylate-binding protein 1 (PABP1), as well as marked CARM1 substrates
were reduced by 40-65% in muscles from mKO mice relative to WT (Fig. 4.3A,B). Global
myocellular ADMA and SDMA content, which are largely indicative of PRMT1 and PRMT5
methyltransferase activities, respectively, were similar between WT and mKO animals. Next,
using LC-MS proteomics, we identified numerous peptides with ADMA and SDMA
modifications that were differentially expressed between muscles from WT and mKO mice (Fig.
4.3C,D). As with MMA, we normalized the abundance of DMA peptides to total protein levels
when possible. Specifically, 9 DMA peptides were significantly increased, and 10 were
significantly decreased in CARM1 mKO muscle relative to WT (FDR-corrected p-value < 0.05)
(Fig. 4.3C). Analysis of DMA sites revealed that a [P/A]RYPLP motif was common to
downregulated dimethylarginine peptides, while motifs containing an RGG sequence did not
display altered methylarginine status in mKO tissues (Fig. 4.3E). We identified several putative
CARM1 substrate sites, including synaptopodin 2-like protein (synpo2l, also known as
myopodin) R466, R476, R953, and R955, as well as titin R296, R304, R318, R1405, and R1414
(Fig. 4.3F). Interestingly, we also observed a potential switch from ADMA to SDMA marks on
61
isoform 3 of RNA binding fox-1 homolog 1 (Rbfox1) in mKO skeletal muscle (Fig. 4.3G).
Taken together, these data demonstrate that removal of CARM1 significantly remodels the
protein arginine methylome in mouse skeletal muscle.
Figure 4.3 Quantitative and integrated analysis of CARM1-mediated protein arginine dimethylation
in skeletal muscle. (A) Representative Western blots of BAF155 arginine asymmetric dimethylation
(BAF155me2a), total BAF155, arginine asymmetric dimethylation of PABP1 (PABP1me2a), total
PABP1, arginine methylated CARM1 substrates, ADMA at glycine and arginine-rich motifs (ADMA-
GAR), and SDMA in TA muscles from WT and mKO mice. The ponceau stain demonstrates equal
loading while approximate MWs (kda) are to the right of each image. (B) Graphical summary of
BAF155me2a, BAF155, PABP1me2a, PABP1, CARM1 substrates, ADMA, and SDMA protein levels in
62
the TA muscles of mKO mice relative to WT animals. Bars indicate group means, whiskers represent
SEMs, points illustrate individual results, and the dotted line denotes 1 (or WT level). Statistical analysis
was completed using a student’s t-test. *, p < 0.05 vs. WT. n = 29 – 37. (C) Volcano plot of
dimethylarginine sites showing neutral loss, log2 fold change, and whether the site was normalized to the
protein level. Different shapes denote ADMA, SDMA, or dimethylarginine (DMA). Sites with an FDR-
corrected p-value < 0.05 are filled in black. Sites that are normalized to the protein level are outlined in
red. (D) Heatmap of significantly altered (FDR-corrected p-value < 0.05) protein normalized
dimethylation sites and non-protein normalized dimethylation sites. Raw LFQ values were normalized to
their respective protein level if available. The prefixes [d] and [m] denote the type of methylation, di- or
mono- respectively, and neutral losses are indicated with an asterisk and either [ad] or [sd] for asymmetric
or symmetric neutral loss, respectively. The enrichment technique for each site is shown in a color
column on the right. n = 6 for each genotype. (E) Two sample motif plot with foreground of
dimethylation sites with log2FC < -0.5 and p < 0.1 compared to background motifs that are unchanging
|log2FC| < 0.2. (F) Integrated analysis of significant (FDR-corrected p-value < 0.05) putative substrates of
CARM1 showing decrease in mono and dimethylation in CARM1 mKO. Neutral loss is denoted with an
asterisk. Mixed peptides contained a mixture of mono/di methylation on arginine (R) sites. (G) Potential
ADMA to SDMA methyl switching of isoform 3 of Rbfox1 sites identified in ADMA and SDMA IAPs.

4.4. Conclusion
The data presented in this chapter provides the first look into the protein methylome of
mouse skeletal muscle to date identifying 1150 methylation sites on 313 proteins. Potential
substrates and alterations to methyl proteins in the absence of CARM1 are also identified and
evidence of methyl scavenging and methyl-switching from ADMA to SDMA. This work also
serves as an example that methyl enrichment and neutral loss identification strategies can be
applied to real animal tissue with in vivo enzyme knockouts.

63
Tables

Tables 1 & 2 Original “Long” gradient and a new proposed “Short” gradient for SCX. Wang et al. did not collect
spectra for the first 20 min during sample loading, but because we found that methyl PSMs were eluting during this
sample loading phase, we did collect spectra during this phase of the LC gradient.

Tables 3 & 4 Summary of spectra identified by each gradient. Each of the five SCX fractions are shown
individually. The numbers of methyl and nonmethyl PSMs were used to calculate the percent enrichment for each
fraction. The number of MMA, Kme1, DMA, Kme2, PSMs, and mixed methyl PSMs are shown. Mixed PSMs
contained a mixture of methyl marks on the same peptide (e.g., MMA and DMA). The percolator q-value cutoff used
to estimate the methyl FDR is also shown for each technique.

64

Table 5 Summary of spectra identified by each technique. For SCX each of the five fractions are shown
along with each IAP. The number of methyl and nonmethyl PSMs were used to calculate the percent enrichment for
each fraction and IAP. The number of MMA, Kme1, DMA, Kme2, Kme3, and mixed PSMs are shown. Mixed PSMs
contained a mixture of methyl marks on the same peptide (e.g., MMA and DMA). The percolator q-value cutoff used
to estimate the methyl FDR is also shown for each technique.

PTM
PTM occurrence in
mouse skeletal
muscle dataset
PTM occurrence in
mouse (Phosphosite
database)
Amino acid
occurrence in
mouse
Lys ubiquitination 10.3% (43) 14.5% (16) 5.65%
Ser phosphorylation 6.9% (28) 11.9% (16) 8.41%
Thr phosphorylation 7.2% (28) 8.8% (16) 5.35%
Tyr phosphorylation 12.7% (28) 13.9% (16) 2.69%
Arg methylation 5.6% (present study) 6.2% (16) 5.63%
Lys methylation 6.8% (present study) 6.2% (16) 5.65%
Arg ADP-ribosylation 9.2% (27) - 5.63%
Lys acetylation - 10.1% (16) 5.65%

Table 6: Distribution of arginine methylation sites compared to other widespread posttranslational modifications
ADP, adenosine diphosphate; Arg, arginine; Lys, lysine; PTM, post translational modification; Ser, serine; Thr,
threonine; Tyr, tyrosine.

65
References

1. Ambler, R. P., and Rees, M. W. (1959) ɛ-N-Methyl-lysine in Bacterial Flagellar Protein.
Nature. 184, 56–57
2. Paik, W. K., and Kim, S. (1967) Enzymatic methylation of protein fractions from calf
thymus nuclei. Biochem. Biophys. Res. Commun. 29, 14–20
3. Clarke, S. (1993) Protein methylation. Curr. Opin. Cell Biol. 5, 977–983
4. Bedford, M. T., and Richard, S. (2005) Arginine Methylation: An Emerging Regulatorof
Protein Function. Mol. Cell. 18, 263–272
5. Sylvestersen, K. B., Horn, H., Jungmichel, S., Jensen, L. J., and Nielsen, M. L. (2014)
Proteomic Analysis of Arginine Methylation Sites in Human Cells Reveals Dynamic
Regulation During Transcriptional Arrest. Mol. Cell. Proteomics. 13, 2072–2088
6. Zheng, S., Moehlenbrink, J., Lu, Y.-C., Zalmas, L.-P., Sagum, C. A., Carr, S., McGouran, J.
F., Alexander, L., Fedorov, O., Munro, S., Kessler, B., Bedford, M. T., Yu, Q., and La
Thangue, N. B. (2013) Arginine methylation-dependent reader-writer interplay governs
growth control by E2F-1. Mol. Cell. 52, 37–51
7. Liu, Q., and Dreyfuss, G. (1995) In vivo and in vitro arginine methylation of RNA-binding
proteins. Mol. Cell. Biol. 15, 2800–2808
8. Smith, W. A., Schurter, B. T., Wong-Staal, F., and David, M. (2004) Arginine Methylation
of RNA Helicase A Determines Its Subcellular Localization. J. Biol. Chem. 279, 22795–
22798
9. Thandapani, P., O’Connor, T. R., Bailey, T. L., and Richard, S. (2013) Defining the
RGG/RG Motif. Mol. Cell. 50, 613–623
66
10. Wooderchak, W. L., Zang, T., Zhou, Z. S., Acuña, M., Tahara, S. M., and Hevel, J. M.
(2008) Substrate Profiling of PRMT1 Reveals Amino Acid Sequences That Extend Beyond
the “RGG” Paradigm. Biochemistry. 47, 9456–9466
11. Cho, E.-C., Zheng, S., Munro, S., Liu, G., Carr, S. M., Moehlenbrink, J., Lu, Y.-C.,
Stimson, L., Khan, O., Konietzny, R., McGouran, J., Coutts, A. S., Kessler, B., Kerr, D. J.,
and Thangue, N. B. L. (2012) Arginine methylation controls growth regulation by E2F-1.
EMBO J. 31, 1785–1797
12. Yu, Z., Chen, T., Hébert, J., Li, E., and Richard, S. (2009) A Mouse PRMT1 Null Allele
Defines an Essential Role for Arginine Methylation in Genome Maintenance and Cell
Proliferation. Mol. Cell. Biol. 29, 2982–2996
13. Tee, W.-W., Pardo, M., Theunissen, T. W., Yu, L., Choudhary, J. S., Hajkova, P., and
Surani, M. A. (2010) Prmt5 is essential for early mouse development and acts in the
cytoplasm to maintain ES cell pluripotency. Genes Dev. 24, 2772–2777
14. Zhong, J., Cao, R.-X., Zu, X.-Y., Hong, T., Yang, J., Liu, L., Xiao, X.-H., Ding, W.-J.,
Zhao, Q., Liu, J.-H., and Wen, G.-B. (2012) Identification and characterization of novel
spliced variants of PRMT2 in breast carcinoma. FEBS J. 279, 316–335
15. Yoshimatsu, M., Toyokawa, G., Hayami, S., Unoki, M., Tsunoda, T., Field, H. I., Kelly, J.
D., Neal, D. E., Maehara, Y., Ponder, B. A. J., Nakamura, Y., and Hamamoto, R. (2011)
Dysregulation of PRMT1 and PRMT6, Type I arginine methyltransferases, is involved in
various types of human cancers. Int. J. Cancer. 128, 562–573
16. Mavrakis, K. J., McDonald, E. R., Schlabach, M. R., Billy, E., Hoffman, G. R., deWeck,
A., Ruddy, D. A., Venkatesan, K., Yu, J., McAllister, G., Stump, M., deBeaumont, R., Ho,
S., Yue, Y., Liu, Y., Yan-Neale, Y., Yang, G., Lin, F., Yin, H., Gao, H., Kipp, D. R., Zhao,
67
S., McNamara, J. T., Sprague, E. R., Zheng, B., Lin, Y., Cho, Y. S., Gu, J., Crawford, K.,
Ciccone, D., Vitari, A. C., Lai, A., Capka, V., Hurov, K., Porter, J. A., Tallarico, J.,
Mickanin, C., Lees, E., Pagliarini, R., Keen, N., Schmelzle, T., Hofmann, F., Stegmeier, F.,
and Sellers, W. R. (2016) Disordered methionine metabolism in MTAP/CDKN2A-deleted
cancers leads to dependence on PRMT5. Science. 351, 1208–1213
17. Kryukov, G. V., Wilson, F. H., Ruth, J. R., Paulk, J., Tsherniak, A., Marlow, S. E.,
Vazquez, F., Weir, B. A., Fitzgerald, M. E., Tanaka, M., Bielski, C. M., Scott, J. M.,
Dennis, C., Cowley, G. S., Boehm, J. S., Root, D. E., Golub, T. R., Clish, C. B., Bradner, J.
E., Hahn, W. C., and Garraway, L. A. (2016) MTAP deletion confers enhanced dependency
on the PRMT5 arginine methyltransferase in cancer cells. Science. 351, 1214–1218
18. Marjon, K., Cameron, M. J., Quang, P., Clasquin, M. F., Mandley, E., Kunii, K., McVay,
M., Choe, S., Kernytsky, A., Gross, S., Konteatis, Z., Murtie, J., Blake, M. L., Travins, J.,
Dorsch, M., Biller, S. A., and Marks, K. M. (2016) MTAP Deletions in Cancer Create
Vulnerability to Targeting of the MAT2A/PRMT5/RIOK1 Axis. Cell Rep. 15, 574–587
19. Zhu, K., Song, J.-L., Tao, H.-R., Cheng, Z.-Q., Jiang, C.-S., and Zhang, H. (2018)
Discovery of new potent protein arginine methyltransferase 5 (PRMT5) inhibitors by
assembly of key pharmacophores from known inhibitors. Bioorg. Med. Chem. Lett. 28,
3693–3699
20. Liu, F., Ma, F., Wang, Y., Hao, L., Zeng, H., Jia, C., Wang, Y., Liu, P., Ong, I. M., Li, B.,
Chen, G., Jiang, J., Gong, S., Li, L., and Xu, W. (2017) PKM2 methylation by CARM1
activates aerobic glycolysis to promote tumorigenesis. Nat. Cell Biol. 19, 1358–1370
21. Wang, Y.-P., Zhou, W., Wang, J., Huang, X., Zuo, Y., Wang, T.-S., Gao, X., Xu, Y.-Y.,
Zou, S.-W., Liu, Y.-B., Cheng, J.-K., and Lei, Q.-Y. (2016) Arginine Methylation of
68
MDH1 by CARM1 Inhibits Glutamine Metabolism and Suppresses Pancreatic Cancer. Mol.
Cell. 64, 673–687
22. Beltrao, P., Albanèse, V., Kenner, L. R., Swaney, D. L., Burlingame, A., Villén, J., Lim,
W. A., Fraser, J. S., Frydman, J., and Krogan, N. J. (2012) Systematic Functional
Prioritization of Protein Post-translational Modifications. Cell. 150, 413–425
23. Derouiche, A., Cousin, C., and Mijakovic, I. (2012) Protein phosphorylation from the
perspective of systems biology. Curr. Opin. Biotechnol. 23, 585–590
24. Hughes, T., Deininger, M., Hochhaus, A., Branford, S., Radich, J., Kaeda, J., Baccarani,
M., Cortes, J., Cross, N. C. P., Druker, B. J., Gabert, J., Grimwade, D., Hehlmann, R.,
Kamel-Reid, S., Lipton, J. H., Longtine, J., Martinelli, G., Saglio, G., Soverini, S., Stock,
W., and Goldman, J. M. (2006) Monitoring CML patients responding to treatment with
tyrosine kinase inhibitors: review and recommendations for harmonizing current
methodology for detecting BCR-ABL transcripts and kinase domain mutations and for
expressing results. Blood. 108, 28–37
25. Gross, S., Rahal, R., Stransky, N., Lengauer, C., and Hoeflich, K. P. (2015) Targeting
cancer with kinase inhibitors. J. Clin. Invest. 125, 1780–1789
26. Xu, J., Wang, A. H., Oses-Prieto, J., Makhijani, K., Katsuno, Y., Pei, M., Yan, L., Zheng,
Y. G., Burlingame, A., Brückner, K., and Derynck, R. (2013) Arginine methylation initiates
BMP-induced Smad signaling. Mol. Cell. 51, 5–19
27. Hsu, J.-M., Chen, C.-T., Chou, C.-K., Kuo, H.-P., Li, L.-Y., Lin, C.-Y., Lee, H.-J., Wang,
Y.-N., Liu, M., Liao, H.-W., Shi, B., Lai, C.-C., Bedford, M. T., Tsai, C.-H., and Hung, M.-
C. (2011) Crosstalk between Arg 1175 methylation and Tyr 1173 phosphorylation
negatively modulates EGFR-mediated ERK activation. Nat. Cell Biol. 13, 174–181
69
28. Biggar, K. K., and Li, S. S.-C. (2015) Non-histone protein methylation as a regulator of
cellular signalling and function. Nat. Rev. Mol. Cell Biol. 16, 5–17
29. Scoumanne, A., Zhang, J., and Chen, X. (2009) PRMT5 is required for cell-cycle
progression and p53 tumor suppressor function. Nucleic Acids Res. 37, 4965–4976
30. Krapivinsky, G., Krapivinsky, L., Renthal, N. E., Santa-Cruz, A., Manasian, Y., and
Clapham, D. E. (2017) Histone phosphorylation by TRPM6’s cleaved kinase attenuates
adjacent arginine methylation to regulate gene expression. Proc. Natl. Acad. Sci. 114,
E7092–E7100
31. Yang, Y., and Bedford, M. T. (2013) Protein arginine methyltransferases and cancer. Nat.
Rev. Cancer. 13, 37–50
32. Bedford, M. T., and Clarke, S. G. (2009) Protein Arginine Methylation in Mammals: Who,
What, and Why. Mol. Cell. 33, 1–13
33. Tang, J., Frankel, A., Cook, R. J., Kim, S., Paik, W. K., Williams, K. R., Clarke, S., and
Herschman, H. R. (2000) PRMT1 Is the Predominant Type I Protein Arginine
Methyltransferase in Mammalian Cells. J. Biol. Chem. 275, 7723–7730
34. Tripsianes, K., Madl, T., Machyna, M., Fessas, D., Englbrecht, C., Fischer, U., Neugebauer,
K. M., and Sattler, M. (2011) Structural basis for dimethylarginine recognition by the Tudor
domains of human SMN and SPF30 proteins. Nat. Struct. Mol. Biol. 18, 1414–1420
35. Schubert, H. L., Blumenthal, R. M., and Cheng, X. (2003) Many paths to methyltransfer: a
chronicle of convergence. Trends Biochem. Sci. 28, 329–335
36. Cheng, X., and Roberts, R. J. (2001) AdoMet-dependent methylation, DNA
methyltransferases and base flipping. Nucleic Acids Res. 29, 3784–3795
70
37. McBride, A. E., and Silver, P. A. (2001) State of the Arg: Protein Methylation at Arginine
Comes of Age. Cell. 106, 5–8
38. Mallick, P., and Kuster, B. (2010) Proteomics: a pragmatic perspective. Nat. Biotechnol. 28,
695
39. Han, X., Aslanian, A., and Yates, J. R. (2008) Mass spectrometry for proteomics. Curr.
Opin. Chem. Biol. 12, 483–490
40. Elias, J. E., Haas, W., Faherty, B. K., and Gygi, S. P. (2005) Comparative evaluation of
mass spectrometry platforms used in large-scale proteomics investigations. Nat. Methods. 2,
667–675
41. Fenn, J. B., Mann, M., Meng, C. K., Wong, S. F., and Whitehouse, C. M. (1989)
Electrospray Ionization for Mass Spectrometry of Large Biomolecules. Science.
10.1126/science.2675315
42. Proteomics by Mass Spectrometry: Approaches, Advances, and Applications | Annual
Review of Biomedical Engineering [online] https://www-annualreviews-
org.libproxy2.usc.edu/doi/10.1146/annurev-bioeng-061008-124934 (Accessed November 9,
2021)
43. Cutillas, P. R. (2005) Principles of Nanoflow Liquid Chromatography and Applications to
Proteomics. Curr. Nanosci. 1, 65–71
44. Elias, J. E., and Gygi, S. P. (2010) Target-Decoy Search Strategy for Mass Spectrometry-
Based Proteomics. Methods Mol. Biol. Clifton NJ. 604, 55–71
45. Hart-Smith, G., Yagoub, D., Tay, A. P., Pickford, R., and Wilkins, M. R. (2016) Large
Scale Mass Spectrometry-based Identifications of Enzyme-mediated Protein Methylation
Are Subject to High False Discovery Rates. Mol. Cell. Proteomics. 15, 989–1006
71
46. Ong, S.-E., Mittler, G., and Mann, M. (2004) Identifying and quantifying in vivo
methylation sites by heavy methyl SILAC. Nat. Methods. 1, 119–126
47. Goulet, I., Gauvin, G., Boisvenue, S., and Côté, J. (2007) Alternative Splicing Yields
Protein Arginine Methyltransferase 1 Isoforms with Distinct Activity, Substrate Specificity,
and Subcellular Localization *. J. Biol. Chem. 282, 33009–33021
48. Romancer, M. L., Treilleux, I., Leconte, N., Robin-Lespinasse, Y., Sentis, S., Bouchekioua-
Bouzaghou, K., Goddard, S., Gobert-Gosse, S., and Corbo, L. (2008) Regulation of
Estrogen Rapid Signaling through Arginine Methylation by PRMT1. Mol. Cell. 31, 212–
221
49. Seligson, D. B., Horvath, S., Shi, T., Yu, H., Tze, S., Grunstein, M., and Kurdistani, S. K.
(2005) Global histone modification patterns predict risk of prostate cancer recurrence.
Nature. 435, 1262–1266
50. Cheung, N., Chan, L. C., Thompson, A., Cleary, M. L., and So, C. W. E. (2007) Protein
arginine-methyltransferase-dependent oncogenesis. Nat. Cell Biol. 9, 1208–1215
51. Rothbart, S. B., and Strahl, B. D. (2014) Interpreting the language of histone and DNA
modifications. Biochim. Biophys. Acta BBA - Gene Regul. Mech. 1839, 627–643
52. Wang, H., Huang, Z.-Q., Xia, L., Feng, Q., Erdjument-Bromage, H., Strahl, B. D., Briggs,
S. D., Allis, C. D., Wong, J., Tempst, P., and Zhang, Y. (2001) Methylation of Histone H4
at Arginine 3 Facilitating Transcriptional Activation by Nuclear Hormone Receptor.
Science. 293, 853–857
53. Wang, Q., Wang, K., and Ye, M. (2017) Strategies for large-scale analysis of non-histone
protein methylation by LC-MS/MS. Analyst. 142, 3536–3548
72
54. Rappsilber, J., Mann, M., and Ishihama, Y. (2007) Protocol for micro-purification,
enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat.
Protoc. 2, 1896–1906
55. Tyanova, S., and Cox, J. (2018) Perseus: A Bioinformatics Platform for Integrative
Analysis of Proteomics Data in Cancer Research. in Cancer Systems Biology: Methods and
Protocols (von Stechow, L. ed), pp. 133–148, Methods in Molecular Biology, Springer
New York, New York, NY, 10.1007/978-1-4939-7493-1_7
56. Musiani, D., Bok, J., Massignani, E., Wu, L., Tabaglio, T., Ippolito, M. R., Cuomo, A.,
Ozbek, U., Zorgati, H., Ghoshdastider, U., Robinson, R. C., Guccione, E., and Bonaldi, T.
(2019) Proteomics profiling of arginine methylation defines PRMT5 substrate specificity.
Sci Signal. 12, eaat8388
57. Schwartz, D., Chou, M. F., and Church, G. M. (2009) Predicting Protein Post-translational
Modifications Using Meta-analysis of Proteome Scale Data Sets. Mol. Cell. Proteomics
MCP. 8, 365–379
58. Cheng, A., Grant, C., Bailey, T. L., and Noble, W. (2017) MoMo: Discovery of post-
translational modification motifs. bioRxiv. 10.1101/153882
59. Vacic, V., Iakoucheva, L. M., and Radivojac, P. (2006) Two Sample Logo: a graphical
representation of the differences between two sets of sequence alignments. Bioinformatics.
22, 1536–1537
60. Murn, J., and Shi, Y. (2017) The winding path of protein methylation research: milestones
and new frontiers. Nat. Rev. Mol. Cell Biol. 18, 517–527
73
61. Larsen, S. C., Sylvestersen, K. B., Mund, A., Lyon, D., Mullari, M., Madsen, M. V., Daniel,
J. A., Jensen, L. J., and Nielsen, M. L. (2016) Proteome-wide analysis of arginine
monomethylation reveals widespread occurrence in human cells. Sci Signal. 9, rs9–rs9
62. Gao, G., Dhar, S., and Bedford, M. T. (2017) PRMT5 regulates IRES-dependent translation
via methylation of hnRNP A1. Nucleic Acids Res. 45, 4359–4369
63. Shen, E. C., Henry, M. F., Weiss, V. H., Valentini, S. R., Silver, P. A., and Lee, M. S.
(1998) Arginine methylation facilitates the nuclear export of hnRNP proteins. Genes Dev.
12, 679–691
64. Gurunathan, G., Yu, Z., Coulombe, Y., Masson, J.-Y., and Richard, S. (2015) Arginine
methylation of hnRNPUL1 regulates interaction with NBS1 and recruitment to sites of
DNA damage. Sci. Rep. 5, 10475
65. Kzhyshkowska, J., Schütt, H., Liss, M., Kremmer, E., Stauber, R., Wolf, H., and Dobner, T.
(2001) Heterogeneous nuclear ribonucleoprotein E1B-AP5 is methylated in its Arg-Gly-
Gly (RGG) box and interacts with human arginine methyltransferase HRMT1L1. Biochem.
J. 358, 305–314
66. Wang, K., Dong, M., Mao, J., Wang, Y., Jin, Y., Ye, M., and Zou, H. (2016) Antibody-Free
Approach for the Global Analysis of Protein Methylation. Anal. Chem. 88, 11319–11327
67. Guo, A., Gu, H., Zhou, J., Mulhern, D., Wang, Y., Lee, K. A., Yang, V., Aguiar, M.,
Kornhauser, J., Jia, X., Ren, J., Beausoleil, S. A., Silva, J. C., Vemulapalli, V., Bedford, M.
T., and Comb, M. J. (2014) Immunoaffinity Enrichment and Mass Spectrometry Analysis
of Protein Methylation. Mol. Cell. Proteomics. 13, 372–387
74
68. Eden, E., Navon, R., Steinfeld, I., Lipson, D., and Yakhini, Z. (2009) GOrilla: a tool for
discovery and visualization of enriched GO terms in ranked gene lists. BMC
Bioinformatics. 10, 48
69. Dhar, S., Vemulapalli, V., Patananan, A. N., Huang, G. L., Lorenzo, A. D., Richard, S.,
Comb, M. J., Guo, A., Clarke, S. G., and Bedford, M. T. (2013) Loss of the major Type I
arginine methyltransferase PRMT1 causes substrate scavenging by other PRMTs. Sci. Rep.
3, 1–6
70. Kerrien, S., Alam-Faruque, Y., Aranda, B., Bancarz, I., Bridge, A., Derow, C., Dimmer, E.,
Feuermann, M., Friedrichsen, A., Huntley, R., Kohler, C., Khadake, J., Leroy, C., Liban,
A., Lieftink, C., Montecchi-Palazzi, L., Orchard, S., Risse, J., Robbe, K., Roechert, B.,
Thorneycroft, D., Zhang, Y., Apweiler, R., and Hermjakob, H. (2007) IntAct—open source
resource for molecular interaction data. Nucleic Acids Res. 35, D561–D565
71. Nabity, M. B., Lees, G. E., Boggess, M. M., Yerramilli, M., Obare, E., Yerramilli, M.,
Rakitin, A., Aguiar, J., and Relford, R. (2015) Symmetric Dimethylarginine Assay
Validation, Stability, and Evaluation as a Marker for the Early Detection of Chronic Kidney
Disease in Dogs. J. Vet. Intern. Med. 29, 1036–1044
72. Zolg, D. P., Wilhelm, M., Schmidt, T., Médard, G., Zerweck, J., Knaute, T., Wenschuh, H.,
Reimer, U., Schnatbaum, K., and Kuster, B. (2018) ProteomeTools: Systematic
Characterization of 21 Post-translational Protein Modifications by Liquid Chromatography
Tandem Mass Spectrometry (LC-MS/MS) Using Synthetic Peptides. Mol. Cell. Proteomics.
17, 1850–1863
75
73. Brame, C. J., Moran, M. F., and McBroom-Cerajewski, L. D. B. (2004) A mass
spectrometry based method for distinguishing between symmetrically and asymmetrically
dimethylated arginine residues. Rapid Commun. Mass Spectrom. 18, 877–881
74. Rappsilber, J., Friesen, W. J., Paushkin, S., Dreyfuss, G., and Mann, M. (2003) Detection of
arginine dimethylated peptides by parallel precursor ion scanning mass spectrometry in
positive ion mode. Anal. Chem. 75, 3107–3114
75. Cox, J., Neuhauser, N., Michalski, A., Scheltema, R. A., Olsen, J. V., and Mann, M. (2011)
Andromeda: A Peptide Search Engine Integrated into the MaxQuant Environment. J.
Proteome Res. 10, 1794–1805
76. Boisvert, F.-M., Hendzel, M. J., Masson, J.-Y., and Richard, S. (2005) Methylation of
MRE11 Regulates its Nuclear Compartmentalization. Cell Cycle. 4, 981–989
77. Aratani, S., Oishi, T., Fujita, H., Nakazawa, M., Fujii, R., Imamoto, N., Yoneda, Y.,
Fukamizu, A., and Nakajima, T. (2006) The nuclear import of RNA helicase A is mediated
by importin-alpha3. Biochem. Biophys. Res. Commun. 340, 125–133
78. Huang, T.-Y., and McLuckey, S. A. (2010) Gas-phase chemistry of multiply charged
bioions in analytical mass spectrometry. Annu. Rev. Anal. Chem. Palo Alto Calif. 3, 365–
385
79. Elviri, L. (2012) 7 ETD and ECD Mass Spectrometry Fragmentation for the
Characterization of Protein Post Translational Modifications
80. Hartel, N. G., Chew, B., Qin, J., Xu, J., and Graham, N. A. (2019) Deep Protein
Methylation Profiling by Combined Chemical and Immunoaffinity Approaches Reveals
Novel PRMT1 Targets. Mol. Cell. Proteomics. 18, 2149–2164
76
81. Hartel, N. G., Liu, C. Z., and Graham, N. A. (2020) Improved Discrimination of
Asymmetric and Symmetric Arginine Dimethylation by Optimization of the Normalized
Collision Energy in Liquid Chromatography–Mass Spectrometry Proteomics. J. Proteome
Res. 10.1021/acs.jproteome.0c00116
82. Cheng, D., Côté, J., Shaaban, S., and Bedford, M. T. (2007) The Arginine
Methyltransferase CARM1 Regulates the Coupling of Transcription and mRNA
Processing. Mol. Cell. 25, 71–83
83. Kim, Y.-R., Lee, B. K., Park, R.-Y., Nguyen, N. T. X., Bae, J. A., Kwon, D. D., and Jung,
C. (2010) Differential CARM1 expression in prostate and colorectal cancers. BMC Cancer.
10, 197
84. Wang, S.-C. M., Dowhan, D. H., Eriksson, N. A., and Muscat, G. E. O. (2012)
CARM1/PRMT4 is necessary for the glycogen gene expression programme in skeletal
muscle cells. Biochem. J. 444, 323–331
85. Stouth, D. W., vanLieshout, T., and Ljubicic, V. (2020) Investigating the Function of Co-
activator-associated Arginine Methyltransferase 1 (CARM1) During Denervation-induced
Skeletal Muscle Atrophy. FASEB J. 34, 1–1
86. Morpheus [online] https://software.broadinstitute.org/morpheus/ (Accessed November 7,
2021)
87. Mi, H., Ebert, D., Muruganujan, A., Mills, C., Albou, L.-P., Mushayamaha, T., and
Thomas, P. D. (2021) PANTHER version 16: a revised family classification, tree-based
classification tool, enhancer regions and extensive API. Nucleic Acids Res. 49, D394–D403
88. Lin, K.-H., Wilson, G. M., Blanco, R., Steinert, N. D., Zhu, W. G., Coon, J. J., and
Hornberger, T. A. (2021) A deep analysis of the proteomic and phosphoproteomic
77
alterations that occur in skeletal muscle after the onset of immobilization. J. Physiol. 599,
2887–2906
89. Wagner, S. A., Beli, P., Weinert, B. T., Schölz, C., Kelstrup, C. D., Young, C., Nielsen, M.
L., Olsen, J. V., Brakebusch, C., and Choudhary, C. (2012) Proteomic Analyses Reveal
Divergent Ubiquitylation Site Patterns in Murine Tissues. Mol. Cell. Proteomics MCP. 11,
1578–1585
90. Leutert, M., Menzel, S., Braren, R., Rissiek, B., Hopp, A.-K., Nowak, K., Bisceglie, L.,
Gehrig, P., Li, H., Zolkiewska, A., Koch-Nolte, F., and Hottiger, M. O. (2018) Proteomic
Characterization of the Heart and Skeletal Muscle Reveals Widespread Arginine ADP-
Ribosylation by the ARTC1 Ectoenzyme. Cell Rep. 24, 1916-1929.e5
91. Greenblatt, S. M., Man, N., Hamard, P.-J., Asai, T., Karl, D., Martinez, C., Bilbao, D.,
Stathias, V., Jermakowicz, A. M., Duffort, S., Tadi, M., Blumenthal, E., Newman, S., Vu,
L., Xu, Y., Liu, F., Schurer, S. C., McCabe, M. T., Kruger, R. G., Xu, M., Yang, F.-C.,
Tenen, D. G., Watts, J., Vega, F., and Nimer, S. D. (2018) CARM1 Is Essential for Myeloid
Leukemogenesis but Dispensable for Normal Hematopoiesis. Cancer Cell. 33, 1111-
1127.e5
92. Drew, A. E., Moradei, O., Jacques, S. L., Rioux, N., Boriack-Sjodin, A. P., Allain, C.,
Scott, M. P., Jin, L., Raimondi, A., Handler, J. L., Ott, H. M., Kruger, R. G., McCabe, M.
T., Sneeringer, C., Riera, T., Shapiro, G., Waters, N. J., Mitchell, L. H., Duncan, K. W.,
Moyer, M. P., Copeland, R. A., Smith, J., Chesworth, R., and Ribich, S. A. (2017)
Identification of a CARM1 Inhibitor with Potent In Vitro and In Vivo Activity in
Preclinical Models of Multiple Myeloma. Sci. Rep. 7, 17993

Abstract (if available)

Abstract Protein methylation has been implicated in many important biological contexts including signaling, metabolism, and transcriptional control. Despite the importance of this post-translational modification, the global analysis of protein methylation by mass spectrometry-based proteomics has not been extensively studied due to the lack of robust, well-characterized techniques for methyl peptide enrichment. To better investigate protein methylation, we utilize two orthogonal methods for methyl peptide enrichment: immunoaffinity purification (IAP) and high pH strong cation exchange (SCX). These techniques are able to enrich separate sets of methyl peptides enhancing coverage of the available protein ‘methylome’, or group of methylated proteins. Use of these techniques is applied to investigate the effects of knock-down of PRMT1, the most active enzyme involved in arginine methylation events in cells. Despite its important biological roles, arginine dimethylation remains an understudied post-translational modification. Partly, this is because the two forms of arginine dimethylation, asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA), are isobaric and therefore indistinguishable by traditional mass spectrometry techniques. Thus, there exists a need for methods that can differentiate these two modifications. Recently, it has been shown that the ADMA and SDMA can be distinguished by the characteristic neutral loss (NL) of dimethylamine and methylamine, respectively. However, the utility of this method is limited because the vast majority of dimethylarginine peptides do not generate measurable NL ions. We report that increasing the normalized collision energy (NCE) in a higher-energy collisional dissociation (HCD) cell increases the generation of the characteristic NL that distinguish ADMA and SDMA. By analyzing both synthetic and endogenous methyl-peptides, we identify an optimal NCE value that maximizes NL generation and simultaneously improves methyl-peptide identification. Finally, a combination of these strategies was used to measure the effects of in vivo knockout of CARM1 on the methylome of mouse skeletal muscle tissue, showcasing an application of these techniques beyond proteins collected through traditional cell culture techniques.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Metabolomic and proteomic approaches to understanding senescence and aging in mammary epithelial cells

PDF

Phosphorylation kinetics of T-cell receptor signaling domain in vitro

PDF

Generation and characterization of peptide theranostics by mRNA display

PDF

Phospho-proteomic analysis of immune cell activation

PDF

Probing the effects of transmembrane domains on the continuum mechanics of lipid bilayers

PDF

Engineering chimeric antigen receptor-directed immune cells for enhanced antitumor efficacy in solid tumors

PDF

Identification and characterization of PR-Set7 and histone H4 lysine 20 methylation-associated proteins

PDF

High-throughput nanoparticle fabrication and nano-biomembrane interactions

PDF

Molecular dynamics studies of protein aggregation in unbounded and confined media

PDF

Protein arginine methyltransferases in murine skull development

PDF

The kinetic study of engineered MBD domain interactions with methylated DNA: insight into binding of methylated DNA by MBD2b

PDF

Engineering immunotoxin and viral vectors for cancer therapy

PDF

Understanding the roles of posttranslational modifications in aggregation using synthetic proteins

PDF

Hybrid lipid-based nanostructures

PDF

Characterization, process analysis, and recycling of a benzoxazine-epoxy resin for structural composites

PDF

Simultaneous monomer deposition and polymerization at low substrate temperatures for the formation of porous polymer membranes

PDF

The selective targeting of Ras: a tale of kinetics, mRNA display, and vaccines

PDF

Comparative analysis of DNA methylation in mammals

PDF

Vapor phase deposition of dense and porous polymer coatings and membranes for increased sustainability and practical applications

PDF

The effect of plasma membrane microenvironment on GPCR activity

Asset Metadata

Creator Hartel, Nicolas G. (author)

Core Title Profiling of protein methylation in mammalian cells and methods for deep methylproteomic analysis

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Degree Conferral Date 2021-12

Publication Date 12/15/2023

Defense Date 11/16/2021

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

Tag ADMA,arginine,asymmetric dimethylarginine,asymmetric dimethylation,fragmentation,HCD,LCMS,mass spectrometry,methylation,methyltransferase,MMA,monomethyl arginine,MS,OAI-PMH Harvest,orbitrap,protein,proteomics,PTM,SDMA,symmetric dimethylarginine

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Graham, Nicholas A. (committee chair), Coba, Marcelo (committee member), Malmstadt, Noah (committee member)

Creator Email nhartel@usc.edu,nickorama3@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC18652765

Unique identifier UC18652765

Legacy Identifier etd-HartelNico-10310

Document Type Dissertation

Format application/pdf (imt)

Rights Hartel, Nicolas G.

Type texts

Source 20211221-wayne-usctheses-batch-905-nissen (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