FUNCTIONAL PROTEOMIC ANALYSIS OF THE DOPAMINE

TRANSPORTER INTERACTOME

by

Sarah Michelle Rogstad

B.S., Harvey Mudd College, 2008

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

Pharmacology Program

2013

This thesis for the Doctor of Philosophy degree by

Sarah Michelle Rogstad

has been approved for the

Pharmacology Program

by

J. David Port, Chair

Robert C. Murphy

Tatiana G. Kutateladze

Nancy R. Zahniser

Alexander Sorkin

Christine C. Wu, Advisor

Date 11/18/13

ii Rogstad, Sarah Michelle (Ph.D., Pharmacology)

Functional Proteomic Analysis of the Dopamine Transporter Interactome

Thesis directed by Professor Christine C. Wu.

ABSTRACT

Dopamine transporter (DAT) is a twelve-transmembrane domain integral membrane protein. It functions by clearing the neurotransmitter dopamine (DA) from the synapse in order to terminate synaptic transmission. DAT has been associated with a variety of diseases and disorders including ADHD, Parkinson’s disease, and schizophrenia and is affected by drugs of abuse such as cocaine and amphetamines. This dissertation focused on examination of the DAT interactome under multiple conditions in both cell and mouse models through immunoprecipitation (IP) coupled with mass spectrometry (MS).

The DAT interactome was analyzed using cell models that stably expressed three dually tagged DAT mutants and parental cells. The tags allowed for highly specific IPs using tag-specific antibodies with the parental cell IPs as a negative control. Analysis of these mutants, which had varying expression levels, indicated that IP-MS studies require normalization prior to comparative analysis. Thus, a normalization method was developed for selected reaction monitoring (SRM) studies based on antibody peptide peak areas. This method was implemented with supplementary cell MS studies.

Additionally, knock-in mice expressed DAT with an HA tag to allow for similar IP-MS studies using an HA antibody. Such studies were conducted on untreated striatal homogenates as well as on striatal synaptosomes treated with amphetamine. Antibody based normalization was applied in these studies as well.

iii Both cell and mouse studies identified, through data-dependent acquisition based

MS analyses, a variety of putative DAT associated proteins under both normal conditions and mutated or drug treated conditions. Relative quantitation of these proteins was conducted at the peptide level using SRM analysis for verification and further assessment purposes. The normalization method developed throughout this dissertation is recommended for future IP-MS studies as it accounts for intersample variability and results in more comparable data across experiments.

The form and content of this abstract are approved. I recommend its publication.

Approved: Christine C. Wu

iv

To my husband, Zachary Rogstad,

and my parents, Hugh and Jane Moore

v TABLE OF CONTENTS

CHAPTER Page

I. THE DOPAMINE TRANSPORTER 1

Sturcture and Function 1

Expression and Trafficking 5

Polymorphisms and Modifications 7

Dopaminergic Disorders 9

Drugs of Abuse 10

Known DAT Interactome 15

Summary 18

II. MASS SPECTROMETRY BASED PROTEOMICS 20

Sample Preparation 21

Chromatography 24

Ionization 26

Mass Spectrometers 28

Summary 34

III. ANALYSIS OF THE MEMBRANE TOPOLOGY OF LEUCINE TRANSPORTER 36

Introduction 36

Experimental Procedures 40

Results and Discussion 43

Summary 48

vi IV. RELATIVE NORMALIZATION OF DAT TO IMMUNOGLOBULINS 49

Introduction 49

Experimental Procedures 52

Results and Discussion 59

Summary 65

V. ANALYSIS OF THE DAT INTERACTOME IN PORCINE AORTIC ENDOTHELIAL CELLS 69

Introduction 69

Experimental Procedures 73

Results and Discussion 75

Summary 84

VI. ANALYSIS OF THE DAT INTERACTOME IN MICE WITH AMPHETAMINE AND PMA TREATMENTS 87

Introduction 87

Experimental Procedures 90

Results and Discussion 95

Summary 106

VII. CONCLUSIONS AND FUTURE DIRECTIONS 109

Introduction 109

Summary of Findings 110

Future Directions 114

Concluding Remarks 1145

REFERENCES 117

vii APPENDIX

A. IgG SRM Transition Information 130

B. IgG Normalization Calculations and Statistics 135

C. Mouse IP SRM Transition Information 141

D. Mouse IP Sum Normalization 146

viii LIST OF TABLES

Table Page

1.1 Known DAT Interactome 16

4.1 Western Blot Densitometry 61

4.2 Antibody Profiling Characterization 62

5.1 Most Abundant Proteins Across DAT IP Samples 80

5.2 Proteins with Most Changed Abundances between FL and ΔN IPs 81

5.3 Proteins with Most Changed Abundances between FL and ΔC IPs 82

6.1 Interactome Candidates for SRM Analyses 96

ix LIST OF FIGURES

Figure Page

1.1 Chemical Structure of the Biogenic Amines 2

1.2 Conformations of LeuT 3

1.3 LeuT-DAT Protein Sequence Alignment 4

1.4 DAT Trafficking 6

1.5 Putative DAT Membrane Topology and Post-translational Modifications 8

1.6 Chemical Structures of DAT-interacting Drugs 12

1.7 Effects of Drugs on DAT Trafficking 13

2.1 MS Sample Preparation 22

2.2 Electrospray Ionization 27

2.3 Tandem MS 29

3.1 Membrane Shaving Schematic 37

3.2 Structures of Leu and pL 39

3.3 3H-Leu Uptake Assay 44

3.4 Identified LeuT Peptides 46

3.5 LeuT Representative SRM Chromatography 47

4.1 IP Schematic and Densitometry 50

4.2 Characterization of Mouse IgG 58

4.3 Characterization of Rabbit IgG 59

4.4 Mouse IgG SRM 63

4.5 Rabbit IgG SRM 64

4.6 Relative Normalization of YFP-DAT AUCs to Mouse IgG AUCs 66

x 4.7 Relative Normalization of YFP-DAT AUCs to Rabbit IgG AUCs 67

5.1 YFP-HA-DAT Constructs and Expression 71

5.2 DAT Sequence Coverage 76

5.3 IgG and DAT Peptide Sums 78

5.4 Normalized DAT Peptide Levels 79

5.5 Normalized peptide Abundance Levels of Putative DAT Interactome Members 84

6.1 Peptide Levels of Quantitation 99

6.2 IgG Normalization of IgG Peptides 100

6.3 IgG Normalization on DAT Peptides 102

6.4 DAT Time Courses 103

6.5 DAT Interactor Time Courses 104

xi CHAPTER I

THE DOPAMINE TRANSPORTER

Dopamine transporter (DAT) is a twelve transmembrane domain (TMD) integral membrane protein (IMP)1. As a member of the solute carrier 6 (SLC6) gene family of neurotransmitter transporters, DAT functions by clearing extracellular dopamine (DA), from the synapse in order to terminate its synaptic transmission2,3. Dysfunction of DAT is known to be associated with a variety of diseases and disorders, such as attention deficit hoperactivity disorder (ADHD)4, Parkinson’s disease5, and schizophrenia5. Drugs of abuse such as amphetamine and cocaine also affect DAT. This results in a decrease in

DAT function and a subsequent increase in DA signaling6,7. Thus, DAT is highly significant clinically. Furthermore, increasing our comprehension of DAT function and processing is critical for furthering our understanding of how these disorders and drugs affect people and how to treat them.

This chapter outlines DAT’s structure and function, expression and trafficking, known polymorphisms and modifications, associated disorders, alterations with drug treatments, and known interacting proteins. It also serves to illustrate the complex relationship between these attributes and relate these issues to the following dissertation, which uses mass spectrometric methods to analyze DAT and the DAT interactome.

Structure and Function

The SLC6, or neurotransmitter sodium symporter (NSS) family, is a group of 12

TMD IMPs, including DAT as well as norepinephrine, serotonin, glycine, and GABA

1 transporters3. In addition to their 12 TMDs, intracellular amino and carboxyl termini and glycosylation of the second extracellular loop structurally characterize this family of transporters8. Functionally, these transporters co-transport their specific substrate with two sodium ions and one chloride ion using the pre-existing neuronal sodium gradient for energy9. These substrates are known as the biogenic amines and are highly similar in structure (Figure 1.1). Currently, none of the mammalian NSS family members have a solved crystal structure. This conundrum largely results from the fact that they contain 12

TMDs. These domains lie within the lipid bilayer and largely consist of non-polar residues making them highly hydrophobic and difficult to crystalize. However, the crystal structure of one bacterial homolog of the family has been solved in a variety of different bound states.

NH2 NH2 OH

NH2 HO HO

HO HO N H OH

Dopamine Serotonin Norepinephrine

Figure 1.1 – Chemical Structures of the Biogenic Amines. The biogenic amines transported by the NSS family are dopamine, serotonin, and norepinephrine. All three contain a six-membered aromatic ring with at least one attached hydroxyl group as well as a primary amine group attached to the ring by a carbon chain.

The leucine transporter (LeuT) from Aquifex aeolicus was first crystalized while bound to its substrate leucine and two sodium ions in 20059,10. Yamashita et al found that leucine was bound to LeuT between TMDs 1 and 6. This structure revealed a conformation that resembled a “shallow shot glass”, which did not allow for solvent

2 accessibility to the substrate. This “occluded state” indicated that the transporter likely forms outward and inward facing conformations as well, thus allowing for movement of the substrate through the transporter across the plasma membrane. Since 2005, LeuT has been re-crystallized with multiple additional substrates. Several antidepressants that act as non-competitive inhibitors were found to lock the transporter into its occluded state11,12.

a) LeuT b)

1b 6a 8 1b

6a

10 10

6b c) 1a

3 8 1a 3 6b

1b LeuT 3 6a

10 LeuT

LeuT 8

d) e) 1a 6b

8 Non-Comp 1b 1b 6a 3 Comp 6a

10 10 LeuT

3 8 1a

1a 6b 6b

Figure 1.2 – Conformations of LeuT. The a) outward facing conformation and b) inward facing conformation were most recently solved13. The original solved LeuT conformation c) is known as the substrate occluded conformation10. Two additional states have been solved with d) competitive14 and e) non-competitive inhibitors11,12. These states are known as outward-locked and locked-occluded states, respectively.

Additionally, tryptophan, which acts as a competitive inhibitor, was found to lock the transporter into an open state14. More recently, LeuT mutants allowed for the analysis of crystal structures in outward-open and inward-open states as well13. When combined,

3 these crystal structures depict a structural look at how LeuT transports leucine across the membrane (Figure 1.2).

|----- LeuT 1 MEKKR EHWATRLGLI 15 DAT 1 MSKSKCSVGL MSSVVAPAKE PNAVGPKEVE LILVKEQNGV QLTSSTLTNP RQSPVEAQDR ETWGKKIDFL 70

------TMD 1------| |------TMD 2------| LeuT 16 LAMAGNAVGL GNFLRFPVQA AENGGGAFMI PYIIAFLLVG IPLMWIEWAM GRYGGAQGHG TTPAIFYLLW 85 DAT 71 LSVIGFAVDL ANVWRFPYLC YKNGGGAFLV PYLLFMVIAG MPLFYMELAL GQFNREGAAG ------VW 132

|------TMD 3------| LeuT 86 R-NRFAKILG VFGLWIPLVV AIYYVYIESW TLGFAIKFLV GLVPE------PPPNATDP ------DSI 140 DAT 133 KICPILKGVG FTVILISLYV GFFYNVIIAW ALHYLFSSFT TELPWIHCNN SWNSPNCSDA HPGDSSGDSS 202

|------TMD 4------| |------LeuT 141 ------LRP FKEFLYS--- YIGVPKGDEP ILKP--SLFA YIVFLITMFI NVSILIRGIS KGIERFAKIA 198 DAT 203 GLNDTFGTTP AAEYFERGVL HLHQSHGIDD LGPPRWQLTA CLVLVIVL------LYFSLW KGVKTSGKVV 266

--TMD 5------| |------TMD 6------| LeuT 199 MPTLFILAVF LVIRVFLLET PNGTAADGLN FLWTPDFEKL KDPGVWIAAV GQIFFTLSLG FGAIITYASY 268 DAT 267 WITATMPYVV LTALLLRGVT LPG-AIDGIR AYLSVDFYRL CEASVWIDAA TQVCFSLGVG FGVLIAFSSY 335

|------TMD 7------| LeuT 269 VRKDQDIVLS GLTAATLNEK AEVILGGSIS IPAAVAFFGV ANAVAIAKAG AFNLG--FIT LPAIFSQTAG 336 DAT 336 NKFTNNCYRD AIVTTSINSL TSFSSG--FV VFSFLGYMAQ KHSVPIGDVA KDGPGLIFII YPEAIATLPL 403

|------TMD 8------| |------TMD 9------| |------LeuT 337 GTFLGFLWFF LLFFAGLTSS IAIMQPMIAF LEDELKLSRK HAVLWTAAIV FFSAHLVMFL NKS-----LD 401 DAT 404 SSAWAVVFFI MLLTLGIDSA MGGMESVITG LIDEFQLLHR HRELFTLFIV LATFLLSLFC VTNGGIYVFT 473

-----TMD 10------| |------TMD 11------LeuT 402 EMDFWAGTIG VVFFGLTELI IFFWIFGADK AWEEINRGGI IKVPRIYYYV M-RYITPAFL AVLLVVWARE 470 DAT 474 LLDHFAAGTS ILFGVLIEAI GVAWFYGVGQ FSDDIQQ-MT GQRPSLYWRL CWKLVSPCFL LFVVVVSIVT 542

------| |------TMD 12------| LeuT 471 YIPKIMEETH WTVWITRFYI IGLFLFLTFL VFLAERRRNH ESAGTLVPR 519 DAT 543 FRPPHYGAYI FPDWANALGW VIATSSMAMV PIYAAYKFCS LPGSFREKLA YAIAPEKDRE LVDRGEVRQF 612

LeuT --- DAT 613 TLRHWLKV 620

Figure 1.3 – LeuT-DAT Protein Sequence Alignment. Exact sequence matches are represented in dark blue with white lettering, while similar residues are denoted with light blue highlighting. Transmembrane domains of LeuT are highlighted with yellow labeling. Alignment conducted using protein Basic Local Alignment Search Tool (BLAST) from the NCBI. TMD locations based on Yamashita et al. 200510.

These multiple structural states help to understand how LeuT functions, and by proxy how the rest of the NSS family functions as well. Yet, at only 22% overall sequence similarity to human DAT, LeuT may not seem to be a good representative for

DAT structure (Figure 1.3). However, within TMDs, particularly numbers 1, 2, 6, and 8 that surround the LeuT binding site, the sequence similarity is much higher, approximately 50%15. This indicates a high likelihood that the transport mechanism is

4 maintained between these species16,17 allowing for more confident comparisons. Chapter

III of this dissertation discusses a mass spectrometry based approach for the analysis of the LeuT binding site membrane topology and the numerous associated complications.

Expression and Trafficking

DA neurotransmission is associated with cognition, motor activity, emotion, and mood18. Neurons in which this process occurs are known as dopaminergic neurons and have their cell bodies primarily located in the ventral tegmental area and substantia nigra19. In order to transport DA, DAT is exclusively expressed in dopaminergic neurons, which was determined by its colocalization with tyrosine hydroxylase, the rate-limiting enzyme that catalyzes a main step in the DA synthesis pathway20.

Under normal conditions, after DAT is synthesized in the endoplasmic reticulum it is N-glycosylated in the Gogli apparatus in the somatodendritic compartment of dopaminergic neurons. It is then transported to the plasma membrane before being transported to the axon where it transports DA into the presynaptic neuron from the synaptic cleft (Figure 1.4)6. DAT localization varies throughout the neuron. In cell bodies and dendrites, where DAT cannot internalize

DA, it is localized primarily in early, late, and recycling endosomes in addition to the plasma membrane21. Contrarily, in the axons, where DA in internalized DAT can be found predominantly in recycling endosomes and on the plasma membrane22.

Additionally, DAT can be internalized through multiple pathways. This can either occur constitutively or can be induced through protein kinase C (PKC) activation. As

5 DAT localization is in a dynamic state, DAT is constantly being internalized and recycled. This process was determined to be clathrin and dynamin dependent through small interfering RNA knock down assays, indicating that internalization is a result of clathrin-mediated endocytosis18,23. This process, along with DAT membrane localization, has also been found to be associated with flotillins24,25.

DA DA

PKC

-Ub NEDD4-2

Endocytosis/ Recycling

Degradation

Figure 1.4 – DAT Trafficking. Under regular conditions, DAT internalizes DA from the synapse into the cell while at located on the plasma membrane. It is constitutively internalized in a PKC dependent manner that has also been shown to involve the E3 ubiquitin-protein ligase NEDD4-2. This internalization can lead either to recycling back to the plasma membrane or lysosomal degradation signaled through ubiquitination (Ub).

Tyrosine kinases and PKC have been indicated as regulators of constitutive trafficking19,20,23,26. Furthermore, activation of PKC using phorbol esters induces increased DAT internalization, which suggests involvement of PKC in DAT

6 regulation6,27. PKC activation has also been shown to result in DAT phosphorylation, thus further indicating the involvement of PKC in DAT regulation28,29.

PKC mediated DAT internalization has been shown to result in increased ubiquitination and subsequently accelerated degradation22,30. Mutant DAT constructs with consequently altered trafficking patterns will be analyzed with respect to their relative expression levels as well as their changes in interactome members in Chapter IV.

Polymorphisms and Modifications

Post-translational modifications (PTMs) are critical for protein processing and signaling. The most well studied DAT PTMs include N-glycosylation, ubiquitination, and phosphorylation (Figure 1.5).

In human DAT (hDAT), N-glycosylation occurs at three canonical sites (Asn181,

Asn188, Asn205) within the large second extracellular loop. Mutation and deglycosylation studies have found that glycosylation at these sites is important but not necessary for DAT expression and function. Non-glycosylated DAT underwent increased endocytosis, resulting in lower levels of surface DAT and transported DA with lowered efficiency31.

Three ubiquitination sites (Lys19, Lys27, Lys35) have been discovered in the N- terminal region of hDAT32. These sites are involved in the degradation pathway, such that when they are ubiquitinated there is an increase in endocytosis and subsequently lysosomal degradation of DAT. This pathway was found to be activated by protein kinase

C (PKC)30 and mediated by the E3 ubiquitin ligase Nedd4-233,34.

7 Phosphorylation

Ubiquitination

Glycosylation

1 6

5 4 2 3 DA 8 7 9 10 11 12

HOOC

NH 2

Figure 1.5 – Putative DAT Membrane Topology and Post-translational Modifications. Transmembrane domain positions are based on LeuT crystallographic studies as well as DAT mutational studies. Sites with observed modifications, including phosphorylation, ubiquination, and glycosylation are denoted.

Protein kinase regulation has been shown to regulate DAT activity. More specifically, phosphorylation of DAT has been observed along with activation of PKC 28.

However, only more recently were any of the specific modification sites identified. The locations of these modifications were first restricted to the N-terminal of the protein35 before three specific N-terminal serines (Ser4, Ser7, Ser13) were identified as phosphorylation sites in recombinant rat DAT. One of these sites (Ser7) was confirmed in rat and human DATs as a site for both basal and PKC activated phosphorylation36. An

8 additional phosphorylation site was identified at an N-terminal threonine in rats (Thr53), and was found to be affected by PKC activation37.

One of the less commonly observed putative DAT modifications is the disulfide bond. A disulfide bond has been identified in the second extracellalur loop in drosophila melanogaster DAT38. In human DAT, disulfide bonds have been implicated in the oligomerization of DAT, in TMDs 4 and 639.

The DAT1 gene located on chromosome 5p15.3 encodes DAT. Currently, few polymorphisms have been identified in DAT1 and even fewer have been thoroughly characterized40. The most notable of these polymorphisms is a variable number tandem repeat (VNTR). This 40 base pair repeat is located in the 3’ untranslated region of the gene and was found to be repeated between three and eleven times, where the nine and ten repeat alleles are the most common41. Comparisons of these genotypes have shown differences in DAT availability42. The different alleles have been associated with

ADHD43-45 and cognitive social tendencies46-48. A recent meta-analysis of DAT1 VNTR studies found that that there was no apparent significant association between striatal DAT availability in individuals with different VNTR alleles across a variety of studies49.

However, a separate study then showed that these polymorphisms modulate DAT levels in the midbrain, but that modulation is influenced by both ethnic background and age, which could explain the previous confusion50.

Dopaminergic Disorders

As was mentioned previously, DAT is associated with a variety of diseases and disorders. Individuals suffering from Parkinson’s disease (PD) experience neuronal

9 degeneration as well as the development of Lewy bodies. This degeneration is partially caused by a decrease in midbrain dopamine as well as a related decrease in midbrain dopaminergic neurons and DAT5. This decline in dopaminergic signaling results in the loss of motor function commonly associated with the disease. Glycosylation of DAT has been associated with dopaminergic degeneration in parkinsonian patients as well as in rodent and monkey models51.

Schizophrenia has also been found to be associated with dopaminergic pathways and DAT. However, a recent meta-analysis found that there was not significant evidence to support the claim. Particularly, it was found that striatal DAT expression levels were not significantly affected by either illness duration or antipsychotic medications, and that there was in fact no correlation between the density of striatal dopaminergic terminals and disease progression52. However, DAT involvement in schizophrenia has not been ruled out.

Numerous studies have linked ADHD with dopaminergic pathways and DAT.

This is particularly due to the DAT1 gene polymorphism that was mentioned in the previous section. Furthermore, DAT mutants found in ADHD have been found to be insensitive to amphetamine and PKC activated endocytosis53.

Drugs of Abuse

Drugs of abuse such as amphetamines and cocaine alter DAT regulation and function. Most of these drugs diminish surface DAT levels, though they generally achieve this end through distinctive mechanisms. This largely results in an increase in dopaminergic signaling due to the decrease in DAT’s ability to function and a subsequent

10 increase in extracellular DA levels. Such increased DA signaling often leads to drug reward, reinforcement, and addiction6,7.

Amphetamine is a psychostimulant drug known to increase wakefulness and decrease appetite. It is commonly prescribed as a treatment for ADHD but is also used recreationally. At the molecular level, amphetamine and other amphetamine-like drugs, such as methamphetamine, which are structurally similar to DA, act as substrates for

DAT and thus inhibit DA transport (Figure 1.6). DAT-mediated transport of amphetamine into the cell initiates downstream signaling events. This results in an increase in intracellular calcium levels, which leads to activation of PKC and calcium/calmodulin-dependent kinase II (CaMKII) and inhibition of Akt kinase activity54,55. Through generally undetermined mechanisms, this pathway leads to a slow internalization of DAT resulting in increased levels of DAT in early and recycling endosomes (Figure 1.7a)56,57. Additionally, treatment with amphetamine has been found to produce a DAT-mediated DA efflux, increasing extracellular DA levels even further through CamKII mediated phosphorylation of the DAT N-terminus58. Evidence has also been found of an initial rise in surface DAT that has been attributed to an increase in delivery of DAT to the plasma membrane instead of a reduction in endocytosis59. This discovery demonstrated that DAT trafficking in the presence of amphetamine is biphasic; initially DAT is rapidly trafficked to the cell surface, then it is internalized. Treatment of mouse synaptosomes with amphetamine and the subsequent effects on DAT and the DAT interactome will be assessed in Chapter VI.

11 NH2 N H Amphetamine Methamphetamine

O

O

N O

O Cocaine

N O

Benztropine

Figure 1.6 – Chemical Structures of DAT-interacting Drugs. Amphetamine and its derivatives act as substrates to DAT, as they are structurally similar to DA. Cocaine and some pharmaceuticals, such as benztropine, act as competitive inhibitors by binding DAT in an overlapping binding site to that of DA.

Cocaine, like amphetamine, is believed to increase extracellular DA levels by acting as a competitive inhibitor of DA transport. Cocaine shares some structural components with DA but contains additional ring structures (Figure 1.6). Cocaine binds to DAT at a binding site that overlaps with that of DA such that DAT is no longer able to transport DA into the cell60. Once DA internalization is blocked, DAT is trafficked to the

12 membrane from early and recycling endosomes in a compensatory manner61, resulting in a delayed increase in DA internalization. This delay has been associated with the addictive properties of the drug (Figure 1.7b)62. Recent studies have implicated phosphorylation of DAT at Ser7 in the alteration of cocaine binding affinity36.

Benztropines are a class of anticholinergic drugs marketed mainly for the treatment of Parkinson’s disease. These drugs are structurally similar to cocaine (Figure

1.6) and similarly inhibit DAT-mediated transport of DA, but with less intensive behavioral effects. The binding site of these drugs has been found to overlap with that of

DA within the primary substrate-binding pocket, and is also slightly shifted from that of cocaine. These multiple binding sites indicate the existence of distinct DAT conformations63.

Studies have also shown that DAT is affected by ethanol. Ethanol treatment results in an increase in cell surface DAT and a subsequent increase in DA internalization64. This ethanol potentiation was found to be dependent on particular residues within the second extracellular loop65, which further illustrates the importance of this region of the protein.

Another important chemical used in DAT trafficking studies is phorbol 12- myristate 13-acetate (PMA). This phorbol ester activates PKC, which is known to regulate DAT trafficking pathways. Treatment with PMA results in rapid clathrin- mediated endocytosis of DAT thus increasing extracellular DA levels (Figure 1.7c). This is followed by degradation of the internalized transporters66. The effects of this chemical on DAT levels and the DAT interactome will be analyzed in Chapter VI.

13 a) ? Amph

PKC -Ub NEDD4-2 Ca2+ DA Endocytosis/ ? Recycling

CamKII

Degradation

Akt

b)

DA DA DA

Cocaine DA DA DA DA

PKC

-Ub NEDD4-2

Endocytosis/ Recycling

Degradation

c) DA DA DA DA

PKC -Ub NEDD4-2 PMA Endocytosis/ Recycling

Degradation

Figure 1.7 – Effects of Drugs on DAT Trafficking. a) With the addition of amphetamine to the system, DAT will transport amphetamine into the cell. This initiates signaling pathways and results in increased calcium levels, which activate PKC and CamKII and inhibit Akt. Through pathways that have not yet been fully determined, this results in an increase in endocytosis and/or a decrease in recycling of DAT. Additionally, intracellular amphetamine signaling results in a DA efflux, even further increasing extracellular DA levels. b) Cocaine acts as a competitive inhibitor of DA by binding to DAT. This results initially in an increase in extracellular DA. However, intracellular DAT is then trafficked to the cell surface, which subsequently depletes extracellular DA levels. c) The phorbol ester PMA activates PKC. This results in increased DAT internalization and degradation thereby increasing extracellular DA levels.

14 Overall, these drug-DAT interactions indicate that complicated trafficking pathways are at play in this system. Although much is understood about their mechanisms of action, much is still unknown at this time. Thus, this project analyzes the effects of some of these drugs upon the DAT interactome in order to further the understanding of the effects of drugs on this system. In the future, a better understanding of the effects of these drugs could lead to the creation of altered drug substitutes for the treatment of addicts as well as pharmaceutical compounds for the treatment of dopaminergic diseases, which could act with limited negative side effects.

Known DAT Interactome

A variety of proteins have been identified that interact with DAT (Table 1-1).

These protein-protein interactions have been indicated in various pathways involved in

DAT trafficking and regulation. The majority of DAT’s confirmed interacting proteins were found using yeast two-hybrid (Y2H) assays. Additional interactors have been identified using the split-ubiquitin system, crosslinking assays, fluorescence resonance energy transfer (FRET) microscopy, co-immunoprecipitation, and immunoprecipitation coupled with mass spectrometry.

Various group have discovered DAT interacting proteins using Y2H screening assays. In these assays, transcription factor segments are hybridized to bait and prey proteins such that when bait and prey interact the transcription factor is activated and initiates transcription of a reporter gene67. This method has identified α-synuclein68,focal adhesion protein Hic-569, protein interacting with C kinase 1 (PICK1)70, receptor of protein kinase C 1 (RACK1)71, and syntaxin 1A71. Additionally, synaptogyrin 3 was

15 identified using the split ubiquitin system, a modified version of the Y2H assay, which allows for identification of protein-protein interactions with integral membrane proteins72,73.

DAT interactions have also been identified through co-immunoprecipitation (co-

IP), in which an antibody targeted to the protein of interest is used to pull down the protein and any bound proteins from a sample to be identified through western blot analysis. This method has resulted in the identification of such proteins as the D2 dopamine receptor74, protein phosphatase PP2A75, and PKC-βI and II76. This method has also been used to confirm interactors initially identified through other methods.

Additionally, IP studies have performed in conjunction with mass spectrometry (MS) in order to identify interactors without the prior knowledge necessary to perform a western blot. Such studies have identified a variety of DAT-interacting proteins. The most promising of which, based on confirmatory studies, include dynamin I, voltage gated potassium channel Kv2.1, neurocam, piccolo, and synapsin I77.

Multiple studies have also pointed toward DAT oligomerization. A FRET microscopy study of cells co-expressing DAT tagged with both yellow and cyan fluorescent proteins (YFP/CFP) revealed direct interactions between DAT molecules56.

This study also showed that the FRET signal was strongest within endosomes showing that DAT is efficiently oligomerizing while internalized. Another study used cysteine crosslinking to stabilize DAT interactions. When monitored using SDS-PAGE, molecular weight bands were consistent with oligomerization up to the tetramer state39.

16

Interactor Discovery Method(s) Reference(s) α-synuclein Yeast two-hybrid, Co-IP Lee 2001 (68) BRCA2 IP-MS Maiya 2007 (77) D2 dopamine Co-IP Lee 2007 (74) receptor Dopamine Crosslinking, FRET microscopy Hastrup 2003 (39), transporter Sorkina 2003 (56) Dynamin I IP-MS Maiya 2007 (77) Hic-5 Yeast two-hybrid Carneiro 2002 (69) Kv2.1 IP-MS Maiya 2007 (77) Neurocam IP-MS Maiya 2007 (77) Piccolo (Aczonin) IP-MS Maiya 2007 (77) PICK1 Yeast two-hybrid Torres 2001 (70) PP2A Co-IP Bauman 2000 (75) PKC-βI/II Co-IP Johnson 2005 (76) RACK1 Yeast two-hybrid Lee 2004 (71) SNAP-25 Yeast two-hybrid Torres 2006 (78) Synapsin I IP-MS Maiya 2007 (77) Synaptogyrin 3 Split ubiquitin, Co-IP Egaña 2009 (73) Syntaxin 1A Yeast two-hybrid Lee 2004 (71)

Table 1.1 – Known DAT Interactome. Previously identified DAT interactors are shown along with the method of discovery.

These protein interactors are involved in a variety of actions and have diverse functions. Together they represent the complex dynamic system of the DAT interactome that is far from fully understood. Increasing comprehension of this interactome, especially under multiple biologically relevant conditions, will aid in the understanding of DAT itself, which is critical for better comprehension of the effects of drugs and disorders on dopaminergic systems as a whole. The majority of this dissertation deals with unraveling the complexities of this interactome using different structural mutants, as can be seen in Chapter V, and in drug treated systems, seen in Chapter VI.

17 Summary

This chapter has illustrated that DAT is a functionally and biologically significant protein. It undergoes complex trafficking patterns and can be regulated in numerous ways, through modifications as well as through protein-protein interactions. These patterns become even more complex when drugs such as amphetamine and cocaine are added. Therefore, it is critical to study the association between DAT trafficking patterns, modifications, and protein interactions both with and without drug treatments in order to fully understand the system. Such an understanding will help to elucidate possible treatments for both diseases and drug abuse.

This dissertation aims to unravel parts of this multifaceted DAT interactome through mass spectrometric analyses, an overview of which will be further explained in the following chapter, in order to better understand DAT trafficking and regulation. This analysis begins in Chapter III, which initially aimed to determine the membrane topology of LeuT in order to discern that of DAT. However, the hydrophobicity of these membrane proteins along with several other factors proved this effort insurmountable with the current tools and project timeline. This endeavor revealed characteristics of both

LeuT and DAT that proved critical in further analyses. Difficulties with this initial project led to a more focused analysis of the DAT interactome.

Analysis of the DAT interactome under distinctive conditions could help clarify the dynamic nature of DAT trafficking. Such DAT interactome studies were conducted using an IP-MS approach. Early analysis of structurally altered DAT constructs illustrated the need for a quantitative normalization method for IP-MS studies, which is described in Chapter IV. Previously, there was a basic assumption inherent in this widely

18 used IP-MS methodology that resulted in quantitative discrepancies. This new analytical tool accounts for internal inconsistency and results in a more robust analysis.

The newly developed normalization method was then applied to subsequent interactome studies. First, changes in the DAT interactome were analyzed as a result of structurally different DAT constructs. These structural alterations resulted in distinctive trafficking patterns and associated interacting proteins. This further indicates a relationship between DAT interacting partners and the regulation of its trafficking patterns, which is discussed in detail in Chapter V. Additionally, treatments with both amphetamine and PMA, as shown in Chapter VI, resulted in other changes in interactome protein levels.

This dissertation developed a normalization method for use in IP-MS analyses.

The application of this quantitative method revealed a complex system of interactors, which is transformed in response to both structural alterations and drug treatments. Such changes indicate that DAT trafficking is highly dynamic in response to a wide variety of effectors and this trafficking is largely dependent upon the DAT interactome. Use of this quantitative assay negated some of the inconsistencies inherent in the IP-MS process, and should thus be utilized for future experiments of this type.

19 CHAPTER II

MASS SPECTROMETERY BASED PROTEOMICS

At the time this dissertation was written, in August 2013, over 7500 studies had been conducted on the DAT since 1975, based on articles accessible through PubMed, with over 400 articles published per year for the past decade. However, when a more specific Medical Subject Headings (MeSH) search is conducted for “dopamine transporter” and “mass spectrometry” this number drops substantially to 25 total studies.

Additionally, a MeSH search of “dopamine transporter” and “proteomics” results in only two total studies. This drop is not due to a lack of interest or capability in the mass spectrometry (MS)-proteomics world, as there are well over 200,000 articles published on MS along with over 43,000 studies on proteomics, both with a wide variety of applications. This discrepancy creates a window of opportunity that allows for the application of this powerful, quantitative tool to the analysis of DAT. Multiple MS methods are discussed with regards to the analysis of DAT within this dissertation. This chapter will give on overview of those, and other, proteomics techniques, as well as the surrounding theory.

A standard MS-based proteomics experiment begins with a biological sample of interest for a particular question. This sample is then prepared using a method specific to the question being asked, before proteolytic digestion. The peptides created from this digestion are separated and ionized before reaching the mass spectrometer. Depending on what the final goal of the experiment is, one of several types of mass spectrometer may be used to analyze the sample. Subsequent data analysis and processing methods are

20 determined based on the type of MS experiment employed. Throughout this dissertation, various combinations of sample preparation, MS detection, and data analysis will be employed. Multiple options for these approaches will be detailed throughout this chapter.

Sample Preparation

Prior to proteomic MS analysis, samples can be prepared in a variety of ways.

Samples can originate from practically any protein source including, but not limited to, cell culture, mouse model, and clinical human samples. Whole tissue or cell homogenates can be prepared for analysis as they are. However, more complex samples often need to be further enhanced or enriched before meaningful data can be collected. This can be achieved through a variety of methods that can be used either by themselves or in conjunction with other methods (Figure 2.1).

Cell culture and tissue samples are generally homogenized and/or lysed using mechanical forces such as those from a glass dounce tissue grinder for homogenization and chemical detergents for lysing membranes. These processes allow for increased protein accessibility. In order to look specifically at different components of a cell, samples can be separated using subcellular fractionation. This generally involves high- speed centrifugation of a sample layered onto a sucrose gradient, which separates cellular components based on density79. In studies targeting a particular protein, affinity purification (AP) or immunoprecipitation (IP) can be used to enrich the sample for the protein of interest79-81. AP and IP enrichment methods will be discussed further in

Chapter IV. Samples can be further enriched via sodium dodecyl sulfate polyacrylamide

(SDS-PAGE) gel electrophoresis in order to separate proteins based on molecular weight.

21 Additionally, the use of two-dimensional gel electrophoresis allows for initial separation based on isoelectric point followed by traditional SDS-PAGE separation82. Many of these steps can be combined for use on a single set of samples. Yet, unnecessary enrichment steps may result in loss of sample such that, if possible, it may be beneficial to use only those that are necessary for the particular study83,84.

Initial Sample

Subcellular Fractionation

Immunoprecipitation Antigen Antigen Gel

Electrophoresis A/G A/G Peptide Digest

Figure 2.1 – MS Sample Preparation. Preparation of tissue and cell samples for MS analysis can involve a variety of steps including subcellular fractionation, immunoprecipitation, and both one- and two-dimensional gel electrophoresis prior to digestion.

Membrane topology experiments described in Chapter III examine LeuT from bacterial membrane fractions. While Chapters IV-VI involve either fractionated or unfractionated samples originating from both mice and porcine cells that were then immunoprecipitated to enrich for DAT and DAT associated proteins for further MS analyses. One common issue that arises from the myriad sample preparation methods

22 results from the fact that both target analytes, LeuT and DAT, are integral membrane proteins. Such proteins include regions that are highly hydrophobic due to their location within the membrane, and require additional preparative steps to ensure their proper analysis. These steps will be mentioned throughout this chapter. Although membrane proteins contain regions that are highly hydrophobic, they also contain regions that are hydrophilic, which reside outside of the membrane, and are thus amphipathic overall. As a result, during sample preparation, membrane protein samples must undergo special treatment to ensure that they are fully solubilized and thus accessible by proteases for digestion. Such solubilization can be achieved through the use of MS compatible detergents, organic solvents, and organic acids that are additionally suitable for digestion85.

Once samples are enriched to the desired degree, they are then ready to be digested. Digestion can be conducted on samples that are in solution or on specific bands that result from either of the electrophoresis techniques. In-gel digestion allows for less complex analytical samples79,86, while in solution digestion allows for a quicker, more comprehensive analysis80,81. In solution samples are generally cleared of contaminants and proteins are concentrated using a procedure such as methanol/chloroform82,87 or acetone/trichloroacetic acid precipitations83,84,88. Proteins are then resolubilized in a MS compatible solvent. Dithiothreitol (DTT) is added to both gel bands and in solution samples in order to reduce disulfide bonds, and iodoacetamide (IAA) is added to alkylate those same bonds. These steps ensure that disulfide bonds are broken and non-reactive.

At this point, proteins are ready to be digested into peptides by one of several proteases. Analysis of digested protein samples through MS is known as bottom-up

23 proteomics, as opposed to top-down proteomics, where whole proteins are analyzed using different fragmentation techniques. The most common protease used in bottom-up proteomics experiments is trypsin, which cleaves the carboxyl side of the peptide bond after both lysine and arginine residues. This results in an average peptide size of 10-12 residues and a corresponding average molecular weight of 800-2000 Daltons89.

Additional proteases can be used in proteomics experiments including LysC and GluC, which cleave at the carboxyl end of lysine and glutamate residues respectively; AspN, which cleaves on the amino side of asparagine residues; chymotrypsin, which cleaves on the carboxyl side of hydrophobic residues; and proteinase K, which is a non-specific protease. Trypsin is used in the vast majority of the experiments described within this dissertation, although the analysis of LeuT in Chapter III uses proteinase K and chymotrypsin as well. After digestion, peptides can be further cleaned using C18 spin columns, pipette tips, or on-line columns, which bind and release peptides allowing for wash steps to dispose of unwanted contaminants. However, as with other additional processing steps, such clean up steps can also result in the loss of sample.

Chromatography

Prior to MS analysis, complex peptide samples are generally separated using chromatographic techniques. Purified peptides may be analyzed via direct infusion, but most samples are too complex to distinguish between separate species without the help of chromatographic separation. Both gas and liquid phase chromatography are often coupled with mass spectrometry. This dissertation, however, uses only liquid chromatography, as this technique is more appropriate for the analysis of larger less volatile compounds.

24 High-performance, or high-pressure, liquid chromatography (HPLC) systems can be placed on-line with mass spectrometers allowing samples to be analyzed directly after separation. Such systems contain multiple buffers, generally water and acetonitrile, that are pumped together through plumbing such that a solvent gradient of increasing organic content can be formed. This gradient flows through a column filled with packing material bound with sample to progressively elute the sample from the column. Depending on the instrumentation, this can occur at flow rates within either the microliter or nanoliter per minute range90.

The most common type of chromatography used for MS analysis is reversed- phase (RP) chromatography. In this case, the column will be packed with a non-polar stationary phase. The most commonly used RP resin for the stationary phase is an octadecyl carbon chain (C18) bonded to silica gel particles. Peptides bind to the C18 material and are removed based on hydrophobicity as the organic gradient passes over the column. Additionally, two (or more) dimensions of peptide separation can be achieved through multidimensional protein identification technology (MudPIT) methods. Such methods combine multiple chromatography resins in sequence. Generally, the peptide mixture is bound first to a strong cation exchange (SCX) resin, which binds peptides based on charge. Pulses of increasing salt concentrations release fractions from the SCX resin onto the RP resin, where peptides are further separated91,92. This process results in multiple distinct chromatograms for a sample and theoretically allows for a deeper look at signals with greater signal-to-noise ratios. These multiple steps of separation make

MudPITs suitable for the analysis of membrane proteins. Within this dissertation,

25 MudPITs are used briefly in the analysis of LeuT membrane topology in Chapter III, while the remaining studies relied solely on RP separations.

Chromatographic difficulties arise from the separation of hydrophobic IMP peptides. Under regular chromatographic conditions, these peptides tend to be underrepresented. However, the application of increased temperature to the column can results in increased recovery of hydrophobic peptides and increased identification of membrane proteins93-95. Additional issues with column performance can also be sample specific. Complex, “sticky” samples can fully clog a column, while impure samples can result in continual slow column degradation. For example, the presence of sepharose beads that have not been fully extracted from an IP sample can result in not only poor spectral results, but they can also have column-fouling effects. Thus, sample clean up, as mentioned in the previous section, is critical to column performance. Additionally, it may be necessary to include forethought in the initial experimental planning stage about column lifespan. For example, if a sample is known to have column-fouling effects, it may be necessary to limit the number of sample injections upon the column.

Ionization

In order to be analyzed by the mass spectrometer, samples must first be ionized.

Thus, the first of three major parts of a mass spectrometer is the ion source. The charging process of ionization allows for the mass spectrometer to measure the mass-to-charge ratio (m/z) of the analyte. The two major modes of sample ionization used for biomolecules are matrix assisted laser desorption ionization (MALDI)96 and electrospray ionization (ESI). However, because ESI is used exclusively in this dissertation, it will be

26 the only mode described in this section. This method is significant enough to the field that in 2002, John Fenn, was awarded the Nobel Prize in Chemistry for the development of the method in the 1980s97.

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + ++ + + + + + + + + + + ++ ++ + + + + + + + + + + + + + + ++ ++ + + + + + + + + + + + + ++ + + + + + + + + + + + + + + ++ + + + + + + + + + ++ ++ ++ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

+ Power - Supply

Figure 2.2 – Electrospray ionization. Sample mixed with solvent is pumped through a capillary with an applied charge. As the sample exits the capillary, it forms a cone shape, known as a Taylor cone, until it splits to form multiply charged droplets. As the droplets evaporate the charges become more concentrated until they split apart into smaller droplets that are eventually vaporized into the MS.

This process converts neutral liquid molecules into gaseous charged particles that can be sprayed into the mass spectrometer for analysis. The initial liquid phase allows for the coupling of HPLC systems directly to mass spectrometers such that separation and analysis can occur simultaneously. Physically, there are several components of an ESI system. These include a capillary tip, through which the sample flows, an opening into the mass spectrometer, and a high voltage power supply, which applies a charge to the system. During ESI, three major steps occur. Firstly, charged droplets are formed at a capillary tip. Then, as the solvent evaporates, these droplets shrink and become more highly charged until, finally, they are transformed into gas-phase ions (Figure 2.2)98.

27 Mass Spectrometers

After the ion source, the two remaining parts of the mass spectrometer are the mass analyzer and the detector. These two entities are intricately connected. This section will discuss the MS techniques used in this dissertation along with which types of mass analyzers and detectors are most useful for such analyses. Methods discussed include data-dependent acquisition (DDA) and selected reaction monitoring (SRM), which both use tandem MS (MS/MS or MS2).

During MS/MS, specific peptide ions are initially selected based on their m/z. The selected ions are then fragmented further, before being again selected for and finally detected, generally by an electron multiplier. Peptide fragmentation occurs in very specific patterns as can be seen in Figure 2.3a.

MS/MS can be performed in two ways: tandem-in-space and tandem-in-time. In tandem-in-space MS/MS experiments, the steps of this process are physically separated, whereas, in tandem-in-time MS/MS experiments, the steps are temporally separated such that the entire process occurs in the same part of the mass spectrometer at distinct times

(Figure 2.3b). Tandem-in-space studies take place in instruments with multiple mass analyzer components such as a triple quadrupole, while tandem-in-time studies occur in trapping instruments such as a quadrupole ion trap99.

As their name implies, triple quadrupole instruments contain three distinct, in-line quadrupole mass analyzers. Quadrupoles consist of four parallel rods centered around an inner axis to which dynamic electric fields are applied. These mass analyzers retain ions within certain m/z windows based on the applied fields, as described in the

28 a) y8 y7 y6 y5 y4 y3 y2 y1

L A Y A I T P E K!

b1 b2 b3 b4 b5 b6 b7 b8

b) MS1 MS2

CID

Figure 2.3 – Tandem MS. a) MS/MS spectrum and associated peptide fragmentation pattern. b) Schematic of MS/MS process. After a full MS1 scan, precursor peptide ions are selected for fragmentation via collision induced dissociation (CID). The resulting product ions are then monitored in an MS/MS scan. This process can occur in different quadrupoles in tandem-in-space instruments, or in the same mass analyzer in tandem-in- time instruments.

29 Mathieu differential equation100. In tandem-in-space MS/MS studies with a triple quadrupole, these m/z windows are used to select for precursor ions in quadrupole 1 (Q1).

Fragmentation occurs in the second quadrupole (q2) through collision induced dissociation (CID), during which a neutral gas such as argon flows into the chamber where the precursor ions will collide with it. The third quadrupole (Q3) then filters the resulting fragment ions, which are subsequently measured by the detector to create a mass spectrum101.

In SRM experiments, triple quadrupoles are used for the targeted analysis of specific precursor and product ions of interest. The areas under the curve (AUCs) of the chromatographic peaks of the product ions are detected by the electron multiplier for quantification. This allows for the use of a quantitative metric to compare relative peptide and protein levels (by proxy) across multiple samples. Additionally, with synthetic stable isotope labeled peptide standards this can be done with absolute quantification102. Over the last several years, Skyline103, a freely available program for the analysis of SRM data has gained widespread popularity. It is used in this dissertation for both SRM experimental design and analysis. Although this method was first developed over 30 years ago for the analysis of small molecules, Nature Methods named this application, known as MS-based targeted proteomics, as the method of the year in 2012104.

Tandem-in-time MS/MS was traditionally performed using three dimensional and linear quadrupole ion traps. Theoretically, this trapping is similar to the quadrupole, however they have different structural configurations. The three dimensional quadrupole ion trap consists of an array of three circular electrodes, which create a parabolic trapping well105. The linear ion trap, the more sensitive of these two instruments106, is more

30 structurally similar to a quadrupole. This mass analyzer also consists of four parallel rods, however the rods are hyperbolic in shape instead of cylindrical as those in a standard quadrupole107. Another, newer, mass analyzer used for similar experiments is the

Orbitrap. Within this mass analyzer ions orbit around an axial electrode and are trapped by a mix of centrifugal and electrical forces108. Recent instrumentation advances that have coupled orbitrap mass analyzers with linear ion traps have allowed for the analysis of more peptides per scan-cycle109. This has resulted in increased identification of less abundant species, and is thus being used with increased frequency throughout the field.

Accordingly, Chapters IV-VI, the data from which were collected after the lab acquired an Orbitrap, take advantage of these qualities and use the instrument exclusively for DDA experiments.

These tandem-in-time MS/MS instruments are often used for profiling experiments. These experiments are used to identify the composition of complex protein samples through DDA. In DDA, the most abundant ions in the initial full MS scan are selected for fragmentation and the resulting product ion m/z values are detected. This creates a spectrum containing individual m/z values for each of the fragment ions from the initial peptide parent ion. This spectrum can be interpreted to identify the amino acid sequence of the specific peptide through a comparison of the m/z values and the known masses of the amino acids. Over the course of a chromatographic gradient, thousands of

MS/MS spectra can be collected in order to identify many protein components within a given sample resulting in a multitude of data. Luckily, computer programs have been developed that can search these data against a protein database of choice. This process

31 first identifies the peptide and then determines what protein each peptide corresponds to for an individual spectrum.

Searching algorithms such as SEQUEST110 and Mascot111 compare observed spectra to theoretical spectra, based on a specified sequence database, resulting in a peptide spectrum match (PSM) with an associated score. Throughout this dissertation,

SEQUEST is used for database searching. Searches are also conducted against randomized decoy databases112 and post-processed using the algorithm Percolator113. This uses semi-supervised machine learning to distinguish between the multiple database identifications resulting in more confident peptide identifications. Additionally, in

Chapters IV-VI, these data is further processed using Hardklör114 and Bullseye115, which work together to deconvolute high resolution MS data.

Profiling MS/MS using DDA traditionally results in a number of spectral counts for each peptide analyzed, indicating the number of times that particular ion triggered an

MS/MS event. The number of spectral counts observed for a particular peptide is often correlated to its abundance. However, this number can be misleading, as less abundant species may not trigger an MS/MS event in some individual runs resulting in semi- random sampling116. In fact, this leads to a generally low analytical reproducibility, such that less than 75% of peptide identifications are consistent between technical replicates117,118. Instrument advancements have allowed for high mass accuracy MS1 spectra of precursor peptides. Retention-time alignment of peptide identification can allow for increased documentation of peptides that may not have triggered an MS/MS event in all samples or replicates. Topograph is a computer program that performs such alignments and subsequently integrates the area under extracted ion chromatograms

32 (XICs) for the MS1 peaks119,120. This results in a highly useful method for relative label- free quantitation of peptides from DDA data, which was used in Chapters V and VI for analysis of this type of experiment.

An additional MS/MS method that is gaining popularity for use with proteins is data independent acquisition (DIA). This method presents some of the best qualities of both of DDA and SRM analyses, as it is both quantitative and allows for analysis of a majority of the peptides within a sample121. However, due to its relatively new application to the proteomics field this approach has not been characterized as fully as the aforementioned analytical methods. For this technique, instead of fragmenting the most abundant precursor ions within a full MS scan, specific small m/z ranges are fragmented systematically. This allows for a quantitative analysis of the entire m/z range without the semi-random sampling issue of DDA. Results of this method include a full library of spectra that can later be mined for answers to hypothesis driven research questions.

Recent advances in high resolution and high mass accuracy instruments have allowed for application of this method to proteomics samples and result in increased protein and proteome sequence coverage122. Due to its more recent development for proteomics applications, this method was not employed within this dissertation; however, its use was investigated during the experimental process, and would be incredibly useful for further analysis of this system.

In both profiling and SRM studies, large amounts of data are produced. The statistical analysis of these data can be cumbersome and it is often difficult to determine the proper course of action. Multiple technical replicates are used for each sample in order to establish the most accurate abundance values, as previously mentioned, however,

33 even this does not account for the variability between runs of DDA experiments.

Statistical analysis of datasets often includes tests such as the Student’s t-test and analysis of variance (ANOVA) to determine whether the abundance of a peptide of interest is significantly different between control and experimental samples. However, in profiling experiments, it can still be difficult to distinguish between identifications of interest and common contaminants even after such statistical tests. Thus, great forethought and careful consideration must go into the statistical analysis of these complex datasets.

Summary

Mass spectrometry based proteomics is a highly impactful field that is still on the rise. The toolbox of various techniques in this field allow for detailed analyses of complex biological systems. Profiling experiments delve into the cohort of proteins involved in a particular system, while targeted analyses can quantify changes in the levels of those proteins due to altered conditions. The expected complexity of the DAT interactome makes MS-analysis the perfect method for exploration of this system.

Additionally, the shortage of completed MS studies on DAT indicates an excellent opportunity for this application.

Multiple sample preparation and MS methodologies are used throughout this dissertation. Chapter III, which focuses on the membrane topology of LeuT, couples membrane fractionation with proteinase K digestion and was analyzed through MudPIT profiling experiments. The remaining three experimental chapters focus on IP samples.

Once such experiments began, it was clear that an IP normalization method was necessary for any future analyses. Thus, in Chapter IV such an analytical method was

34 developed using relative SRM measurements of both DAT and antibody peptides.

Chapters V and VI apply this analytical method to IP experiments quantified using DDA coupled with label-free differential analysis, in both chapters, and SRM analysis, in

Chapter VI. Thus, this dissertation combines profiling and targeted MS analyses of the

DAT interactome in both cell and mouse models under altered conditions to further the understanding of DAT trafficking and processing, enhanced by the aforementioned normalization method.

35 CHAPTER III

ANALYSIS OF THE MEMBRANE TOPOLOGY OF LEUCINE TRANSPORTER

Introduction

Leucine transporter (LeuT) is a bacterial homolog of the NSS family. As was mentioned in Chapter I, LeuT is the only transporter of this family with a solved crystal structure10, a difficult task due to the high hydrophobicity of these transporters resulting from their 12 TMDs. Although the sequence homology between DAT and LeuT is relatively low overall, within certain regions, such as the TMDs involved in substrate binding, the sequence homology is much higher. This structural similarity allows for the use of LeuT as a proxy for better understanding and estimating DAT structure. The aim of this chapter was to develop an assay for analysis of LeuT membrane topology that would later be applicable to DAT. Although, this endeavor was largely unsuccessful, the analytical process of this project imparted a better understanding of the methodological difficulties of working with highly hydrophobic multi-TMD IMPs.

Membrane shaving techniques have been developed for IMPs, which take advantage of protease accessibility. Such methods digest and separate peptides that are located within the membrane from those that are located intra- and extracellularly (Figure

3.1). When coupled with mass spectrometry, this allows for the identification of membrane embedded peptides (MEPs) along with protease accessible peptides (PAPs)123.

The use of these methods along with MS profiling methods has drastically increased identification of IMPs in various membrane samples. Prior to this method, the greatest enrichment for IMPs resulted in approximately 65% of protein identifications accounted

36 for by IMPs. However, with the PAP/MEP methods, this number has been increased to

82% of identified proteins. Consequently, peptides identified with this method largely overlap with previously predicted TMDs124.

High pH/Proteinase K (hppK) or Chymotrypsin

CNBr

PAPs

MEPs

Figure 3.1 – Membrane Shaving Schematic. Membrane proteins are digested using to high pH/Proteinase K (hppK) method, which shave off the protease accessible peptides (PAPs) leaving the membrane embedded peptides (MEPs) in the membrane. The membrane is solubilized and CNBr is used to digest the MEPs

The original membrane shaving method used high pH and proteinase K (hpPK).

In this method, biological samples are exposed to high pH in order to create non- solubilized membrane sheets123. The addition of high pH results in disruptions in the phospholipid bilayer, likely due to the added charge differentials125. Proteinase K (PK), a non-specific protease, is then added to cleave the proteins it can access. The high pH also serves to limit PK activity, such that peptides are formed that are generally a suitable size for MS analysis. Centrifugation can be used to separate out the membranes from the digested PAPs. The membranes are then solubilized using an organic acid, such as formic acid, before the remaining protein segments are digested with cyanogen bromide.

37 Additionally, this process can be completed stepwise, with the high pH applied after initial PK addition, in order to distinguish between extracellular and intracellular

PAPs123. Similarly, physical agitation has been coupled with chymotryptic digestion for a comparable analysis with varied sequence coverage126. This chymotryptic membrane shaving technique is coupled with MS analysis in this chapter for an investigation of

LeuT membrane topology.

LeuT is known to have multiple structural conformations, based on crystallographic studies. These conformation states include a substrate bound state along with outward and inward facing states. Thus, if the membrane topology of LeuT is examined in multiple states, it should reveal structural information about the binding pocket. Additionally, after establishment of this method in LeuT, and comparison to the known structural conformations, such studies can be repeated in DAT to further establish structural homology and understanding.

In LeuT, such studies can be conducted through crosslinking. Several photoreactive amino acids have been synthesized, which covalently bind interacting proteins when exposed to ultraviolet (UV) light. Photo-Leucine (pL) is identical to leucine (Leu), except that pL contains a diazirine ring where Leu has only a methyl group

(Figure 3.2a). When exposed to UV light, the diazirine ring breaks to release nitrogen gas and the remaining carbene binds to nearby proteins. These photoreactive residues were developed for incorporation into proteins for interaction studies127, but can also be used as a LeuT substrate. UV exposure after addition of pL to LeuT crosslinks pL to LeuT, within the binding pocket. This allows for a comparison of the outward facing conformation (without pL) to the substrate-occluded conformation (with pL) when

38 coupled with membrane shaving techniques and MS analysis. However, this method is based on the assumption that LeuT transports pL as it transports Leu. This assumption is tested within this chapter prior to further MS analysis.

Leucine Photo-Leucine

Figure 3.2 – Structures of Leu and pL. pL contains a diazirine ring where Leu has a simple methyl group.

In order to identify peptides within the binding pocket, these crosslinking and membrane shaving methods can be coupled with profiling MS analysis. This analysis can be performed using multidimensional protein identification technology (MudPIT), which was described in Chapter II, which allows for in-depth examination of the sample through several levels of chromatographic separation. Once candidate binding pocket peptides are identified, they can then be relatively quantified using selected reaction monitoring

(SRM) to measure relative differences in peptide accessibility between conformation states.

This chapter aimed to look at conformational changes through this pipeline with the intention of further applying the method to DAT once established. However, complications arose that will be detailed in the results and discussion section of this chapter, which hindered this process. These complications likely arose largely from the

39 hydrophobicity of these transporters. Luckily, lessons learned from these experiments were applicable to future experiments and allowed for further analysis of DAT through alternative analytical avenues.

Experimental Procedures

LeuT Expression and Materials – All solvents were purchased from Fisher

Scientific (Rockford, Il), and all other chemicals were from Fisher Scientific, JT Baker

(Center Valley, PA), or Sigma (St. Louis, MO) unless otherwise noted.

All LeuT used in this chapter was prepared through the transformation of a pET16b vector containing the LeuT sequence (from the lab of Dr. Eric Gouaux, Vollum

Institute of Oregon Health & Science University) into OverExpress C41(DE3) cells

(Lucigen).

Inner Membrane Enrichment – Following transformation and induction, the cells were pelleted before resuspension in 20% sucrose buffer (with 7mM EDTA, 200 U

DNAse 1, and 100 µg/µL lysozyme) for lysis through lysozyme addition (400 µg/ml) and sonication. The lysed cells were layered onto a four-step sucrose gradient, which was then centrifuged for 16 hours at 100,000 x g and 4°C. Fractions were collected and the fraction containing the IM band was used for further analyses128. Fractions were centrifuged for 30 minutes at 20,817 x g and 4°C. Final IM pellets were stored at -80°C until further use.

LeuT Uptake Assays – IM pellets were resuspended to 0.1 mg/mL in uptake buffer (20 mM HEPES-Tris, 100 mM potassium gluconate, 100 mM NaCl, pH 7.4), while pL and Leu were resuspended to 10 mM in uptake buffer. Leu or pL was added to

40 IMs to a final concentration of 100 µM. Samples with Leu, pL, or without any substrate were exposed to UV light. These samples were incubated for 30 minutes at room temperature in the dark, before being crosslinked on ice for 20 minutes using a 245 nm

UV light. This crosslinked the pL sample, while the other two conditions were used as controls. Three additional samples contained no substrate without UV exposure. All six samples were incubated at 37°C in a shaking water bath before addition of 3H-Leu

(specific activity = 115.4 Ci/mmol, Perkin-Elmer Life Sciences, Boston, MA) to a final concentration of 0.1 µM. Cold pL or Leu was added to 100 uM to competition samples, prior to a 3 minute incubation at 37°C, which was then quenched on ice. Samples were then harvested via rapid vacuum filtration through 0.05% polyethyenimine-soaked filter paper (GF/B; Brandel, Gaithersburg, MD). The amount of [3H]-Leu accumulated was determined through liquid scintillation counting. These uptake studies were performed in the lab of Dr. Nancy Zahniser.

Chymotryptic PAPs Preparation – IM pellets were resuspended in 2M Urea/100 mM Tris-HCl to a final concentration of 1 mg/mL. Vesicles were resuspended using an insulin syringe. Dithiothreitol (DTT) was added to a final concentration of 5 mM and samples were incubated at 60°C for 30 minutes to reduce reactive cysteines.

Iodoacetamide (IAA) was then added to a final concentration of 15 mM and samples were incubated for 30 minutes in the dark at room temperature in order to alkylate reduced cysteines. Chymotrypsin was added to a 1:50 enzyme:substrate ratio, and samples were again resuspended using an insulin syringe before overnight incubation at

25°C. An equal volume of aqueous-organic buffer (10% acetonitrile) was added before

MEPs were pelleted out of the sample through ultracentrifugation for 30 minutes and 4°C

41 in a TLA-55 fixed angle rotor at 100,000 x g. The supernatant (PAPs) was collected and adjusted to 5% formic acid prior to MS analysis.

MicroLC-Mass Spectrometry – MudPIT MS runs were performed on an

LTQ linear ion trap mass spectrometer (Thermo Fisher Scientific, Waltham, MA) coupled with an Agilent 1100 binary HPLC and autosampler system for profiling of

LeuT. Samples were loaded off-line onto a fused-silica (Polymicro Tech, Pheonix, AZ) microcapillary column (100µm i.d. with a 5 µm laser-pulled tip) packed with 15 cm of

5µm 125 Å Aqua C18 RP material (Phenomenex, Torrance, CA). Column temperature was maintained at 40°C by a column heater built in-house, as previously described93.

50µg of sample was analyzed with 12-step MudPITs, as previously described124. Mass spectra were acquired through DDA with a single full mass scan followed by 5 MS/MS scans.

Targeted SRM analyses were conducted using a ThermoElectron MicroAS autosampler/Surveyor HPLC pump system and TSQ Quantum Access mass spectrometer

(Waltham, MA) running Xcaliber software. Column composition and temperature were consistent with those used for MudPIT analyses. Peptides for analysis were determined using P3 Predictor to identify chymotryptic peptides within the TMDs of the binding pocket. This resulted in 7 proteotypic peptides. Doubly charged precursor ions and the 6 most abundant product ions (a combination of b and y) were monitored in SRM mode.

Samples were analyzed using a 65-minute linear gradient. Data were acquired with a Q1 and Q3 resolution of 0.7 m/z. Transitions were monitored with a total cycle time of 1 s.

The rf-only q2 collision cell was pressurized with 1 mTorr of argon gas and collision energies of the transitions ranged from 22 to 38 V. LeuT IM preps were analyzed with

42 and without crosslinked pL. Additionally, peptide standards were injected individually and co-injected with sample to verify peaks.

Results and Discussion

This chapter used a combination of techniques for the analysis of LeuT membrane topology with the intention of further applying this newly developed methodology to

DAT membrane topology. Firstly, pL was crosslinked to enriched bacterial IM samples expressing LeuT. The assumption that pL acts similarly to Leu in this system was assessed through uptake assays. LeuT was crosslinked into the substrate-occluded state prior to membrane shaving for both DDA and SRM MS analyses. Issues occurred throughout these analytical steps, most likely resulting from sample complexity, which will be further discussed.

For all experiments in this chapter, LeuT was overexpressed in C43(DE1) modified E. coli cells, which are known for their ability to express membrane proteins at high levels129. However, an early issue was the overexpression of LeuT in IMs.

Preliminary optimization of the IM isolation protocol resulted in only a two-fold enrichment of LeuT while outer membrane (OM) proteins were not found to be de- enriched, as estimated by western blot. Additional optimizations were employed, however, as can be seen below, these difficulties were still encountered in both identification and quantification of LeuT peptides.

In order to lock LeuT into its substrate-occluded conformation, this study took advantage of the Leu analog pL. If pL is in the LeuT binding site at the time of UV exposure, it results in a stabilized conformational change for LeuT into the substrate- occluded position. This allows for the identification of domains exposed in this

43 conformation using the PAPs method. Initial studies quantified the uptake of 3H-leucine

(3H-Leu) by membranes enhanced for LeuT under different conditions with varying levels of pL and leucine with and without UV crosslinking. These tests showed that uptake of 3H-Leu was significantly different between samples with and without substrate, whether that substrate was Leu or pL (Figure 3.3).

3H-Leucine Uptake 2500

2000

1500 * 1000 * ** ** Average CPM Average 500

0 Uptake Leu pL UV Leu UV pL UV Comp Comp

Figure 3.3 – 3H-Leu uptake assay. Uptake and UV samples contained no substrate. Competition (Comp) samples had substrate added directly concurrent with 3H-Leu. UV, Leu UV, and pL UV were crosslinked on ice for 20 min. Uptake was measure through scintillation counting in counts per minute (CPM). All substrate was added to 100µM. Data were analyzed with an unpaired T-tests between either Uptake and Comp or UV and Leu/pL UV samples (*:p<0.05, **:p<0.01).

Samples analyzed in this study began with a LeuT membrane enriched sample.

The “Uptake” sample contained only the LeuT enriched membranes in order to find the maximal 3H-Leu uptake. In competition samples (“Comp”), 100 µM pL or leucine was added immediately prior to the addition of 3H-Leu. In crosslinked samples, 100 µM pL, leucine, or neither was added to the sample before 20 minutes of UV exposure, prior to the performance of the uptake assay. The “UV” sample, which did not contain any substrate controlled for any effects the UV exposure had on the transporter. However, no

44 significant difference was found between the Uptake and UV samples. Significant differences (P<0.05) were found between the Comp and Uptake samples. Similarly, more significant differences (P<0.01) were found between the UV and Uptake samples.

Altogether, these data shows that pL acts comparably to Leu as a substrate for LeuT. This finding allows for further analysis of LeuT membrane topology using pL to lock the transporter into the substrate-occluded conformation.

Once the functionality of pL was established, chymo-PAPs digestion was applied to samples with and without crosslinked pL. Samples were then profiled through MudPIT analysis. Across multiple chymo-PAPs MudPITs, only one LeuT peptide was identified.

While this was a promising peptide, based on its location within the binding pocket this singular peptide was not enough to determine membrane topology. Additional profiling studies used chymotryptic and tryptic digestions of IM and whole cell samples. These studies resulted in the identification of three additional LeuT peptides (Figure 3.4).

However, based on in silico analysis using P3 Predictor, LeuT contains 16 prospective chymotryptic peptides, 7 of which lie within TMDs 1, 3, 6, and 8. These peptides vary in their positions relative to the membrane; some extend intracellularly, some reside completely within the membrane, and others extend extracellularly. With so few possible peptides, it is not surprising that such limited results were achieved. Therefore, chymotrypsin is likely not the optimal protease to cleave the LeuT TMDs, based on the primary amino acid sequence. Proteinase K is not an optimal substitute as it cleaves at any residue such that its peptides cannot be readily predicted for further SRM analyses.

Other proteases such as GluC, PepsinA, LysC, and LysN, could have been used; however, their compatibility with membrane shaving techniques has not been assessed.

45 Figure 3.4 – Identified LeuT Peptides. Locations of identified LeuT peptides (blue boxes) are shown with respect to TMDs (orange highlights). Only the first peptide, located in TMD 1, was found using Chymo-PAPs digestion. Additional peptides were found using alternate digestion methods.

Targeted SRM analyses were conducted on LeuT enriched membrane samples to determine the feasibility of further analysis of the topology and conformational changes of the transporter. It was previously found that the chymotrypsin method of producing

PAPs (chymo-PAP method) resulted in higher sequence coverage than the hppK-PAP method130, thus SRM studies were conducted using samples digested with the chymo-

PAP method. In addition to identifying proteolytic peptides, P3 Predictor identified the expected transitions from each precursor ion to all possible product ions that could be used to create a method appropriate for a triple quadrupole MS. Enriched LeuT samples were digested using the chymo-PAP method before being run with the aforementioned targeted MS methods. Initial studies identified all seven of the chymotryptic binding pocket peptides in the sample, with multiple strong product ions for each precursor ion

(Figure 3.5). This showed that the peptides were protease accessible using the chymo-

46 PAPs method, and allowed for further quantitative analyses through coinjections of synthetic versions of the peptides. Additional studies were conducted with LeuT IM enriched samples and without crosslinked pL. Synthetic peptides were also coinjected for peak verification. Area under the curve (AUC) was calculated for each condition and overall, was found to be lower in crosslinked samples than in non-crosslinked samples.

However, these differences were not found to be statistically significant.

Standard only Sample only Standard & Sample

TIC 887.07 TIC 887.07 TIC 887.07 4.03 E 6 1.44 E 5 5.00 E 5

887.07 ! 211.14 887.07 ! 211.14 887.07 ! 211.14 5.03 E 3 4.87 E 4 2.56 E 3

887.07 ! 324.23 887.07 ! 324.23 3.32 E 3 887.07 ! 324.23 4.63 E 4 9.15 E 3

887.07 ! 423.30 887.07 ! 423.30 3.55 E 3 887.07 ! 423.30 1.96 E 5 2.55 E 4

887.07 ! 522.37 887.07 ! 522.37 887.07 ! 522.37 3.21 E 5 9.54 E 3 4.16 E 4

887.07 ! 564.34 887.07 ! 564.34 887.07 ! 564.34 4.36 E 4 1.97 E 3 3.42 E 3

887.07 ! 593.40 887.07 ! 593.40 887.07 ! 593.40 1.28 E 6 5.23 E 4 1.70 E 5

887.07 ! 706.49 887.07 ! 706.49 887.07 ! 706.49 1.98 E 6 8.07 E 4 2.54 E 5

Figure 3.5 – LeuT Representative SRM Chromatography. Total Ion Chromatogram (TIC) is shown in the top row. Subsequent rows depict precursor ion-product ion transitions. The left column shows a standard only run, the middle column is sample only, and the right column is a coinjection of standard and sample. This figure shows a chymotryptic peptide that resides in TMD 3: IPLVVAIY. Together with the DDA data, these non-significant changes likely indicate a low recovery of LeuT peptides during chromatography. This could be a result of the high hydrophobicity of the peptides due to their location within the membrane. The elevated chromatographic temperature should reduce this ill effect. However, this does not seem to be the case in this particular instance.

47 Summary

This chapter attempted to use pL as a substrate for LeuT in order to lock the transporter into specific conformations for MS analysis of membrane topology. Firstly, pL was tested as a suitable replacement of Leu, as a LeuT substrate within bacterial IM fractions. Based on the uptake studies described in this chapter, pL appears to act comparably to Leu. After the legitimacy of using pL was established. LeuT IM enriched samples were crosslinked with pL and subjected to chymo-PAPs digestion, which effectively shaved peptides from the membrane. This technique, when compared between samples with and without pL, was meant to identify peptides within the binding pocket with altered accessibility upon LeuT conformational changes. However, when coupled with both profiling and targeted MS analyses, this resulted in low peptide recovery rates.

Few LeuT peptides were identified during MudPIT analyses, and, during SRM analyses, peptides that were targeted did not change significantly between conformational states.

This issue is likely a result of the hydrophobicity of the peptides involved. Thus, in future experiments, more care was taken to account for the hydrophobicity and complexity of

IMPs.

48 CHAPTER IV

RELATIVE NORMALIZATION OF DAT TO IMMUNOGLOBULINS1

Introduction

Initial characterization of DAT stably expressed in porcine aortic endothelial cells

(PAE) resulted in unforeseen complications. Three different DAT constructs with different trafficking and localization patterns33,131, along with their interacting partners, were to be analyzed via multiple mass spectrometric methods. These constructs will be further discussed in the subsequent chapter. The PAE-DAT trafficking patterns have already been extensively studied via microscopy in the Sorkin lab, thus creating a useful starting point for the analysis of the DAT interactome in an easily maintainable model system. Upon initial targeted SRM analysis of immunoprecipitates of these cell lines, it was found that the constructs had different expression levels such that it was difficult to compare their binding partners without first being able to accurately compare the abundance of DAT. Thus, it was clear that a novel analytical method was necessary for further analysis within this system.

Affinity purification (AP) constitutes a category of protein enrichment strategies that can be utilized for the analysis of protein-protein interactions and protein complexes80,81. In order to identify the proteins involved in these interactions and complexes, and subsequently to quantify such interactors, AP is frequently coupled with

MS80,81,132,133. One common method of AP is immunoprecipitation (IP). In a typical IP

1 Much of the data included in this chapter was previously reported in the following publication: Rogstad, S. M., Sorkina, T., Sorkin, A. & Wu, C. C. Improved precision of proteomic measurements in immuniprecipitation based purifications using relative quantitation. Anal Chem (2013). doi:10.1021/ac4002222

49 experiment (Figure 4.1a), an antibody is coupled to a solid phase (most often protein A or protein G conjugated sepharose beads), and incubated with the biological sample of interest. The beads are then washed, and the attached proteins are removed through elution or denaturation134. The beads are pelleted out of the solution such that the final sample for analysis contains antibody, the protein of interest, and any associated proteins.

a) Monoclonal b) DAT(Densitometry( Antibody """Co6IP" FL" ΔC" DAT A/G A/G 160,000" (+ Lysate) " 120,000"

Antigen Antigen GFP Elution 80,000"

40,000"

Density((arbitrary(units)(( 0"

10"sec"" 30"sec" 2"min" 30"sec" 2"min" 5"min" Exposure(Time(

Figure 4.1 – IP Schematic and Densitometry. a) In a standard IP experiment, protein A or protein G coupled sepharose beads are combined with an antibody and a sample lysate. The antibody binds to both the beads and its target antigen, which may also affinity enrich a variety of interacting proteins. Proteins are eluted from the beads leaving the antibody in solution with the antigen and interacting proteins. b) Three replicate western blots of full length DAT (FL), N-terminally deleted DAT (ΔN), and C-terminally deleted DAT (ΔC) αGFP IPs were performed, probing with either αDAT or αGFP primary antibody. In the αGFP blots, the ΔN-DAT was more prominent than the other lanes, while in the αDAT blots ΔN-DAT was not detected as the antibody is directed toward the N-terminus of DAT. Densitometry measurements of FL-DAT and ΔC-DAT bands at three different exposure times resulted in dramatic variations in densities.

IP-MS is a strong pipeline that can be utilized for a variety of experiments. One common pitfall of this strategy is that the concentration of beads can vary slightly between samples due to changes in bead slurry distribution resulting in variations in the amount of antibody bound and consequently in the amount of antigen and interacting

50 proteins in the eluate. Additionally, antibody affinity for an antigen could change with mutations to the antigen. This could result in lowered recovery of the mutated antigen, which could be misconstrued as biologically significant. Stable isotope labeling by amino acids in cell culture (SILAC)135 is also often combined with IP and ensures that samples undergo identical preparations. However, if the specific biological conditions of interest could alter the affinity of the antigen to the antibody, the sample loads would no longer be comparable. These issues indicate that a normalization method is needed in IP-MS experiments to control for discrepancies. Specifically, a constant variable between IP experiments could be used to mitigate such variations.

One such variable is the amount of the antibodies themselves. Immunoglobulin G

(IgG) is the class of antibody that is commonly used for IP experiments. These molecules consist of two γ heavy chains and two light (κ or λ) chains connected by disulfide bridges and noncovalent bonds. Both chains consist of one variable and one constant domain. The sequences of variable regions in both the heavy and light chains differ between specific antibodies in order to form antigen-specific binding sites136,137.

Thus, the constant regions are more reliable target regions than the variable regions for proteomic analyses and could be utilized for the analysis of a wide variety of antibodies.

Using IPs from each of the three dually tagged DAT construct expressing cell lines, IgG levels were used internal standards to which DAT levels could be normalized.

This allowed for the standardization of input levels based on the effectiveness of the IPs.

This concept is often applied to western blots where IgGs can be used as a loading control. However, with the limited dynamic range of westerns with enhanced chemiluminescence (ECL), it can be inaccurate to compare these levels between samples

51 (Figure 4.1b)138,139. Additionally, monitoring IgG levels, as opposed to bait levels, for normalization is critical for samples such as these cell lines where bait levels are not expected to be constant due to modifications and expression differences.

Polyclonal rabbit IgG and monoclonal mouse IgG1 antibodies targeting separate tags on the DAT constructs were used in these experiments as they are the most widely used antibody types for IPs140. IgG levels in IP samples were monitored via selected reaction monitoring (SRM) after selecting for proteotypic peptides for quantitative analysis. Once the IgG peptides of interest were determined, AUCs of DAT peptides were compared with those of IgG peptides for relative normalization before comparing

DAT levels between samples. Normalization of peptides of interest to internal standard peptides has been shown to be a robust method that can determine relative abundance without absolute quantitation141,142.

This study explored the utilization of IgGs in IP-MS studies through a relative normalization and quantitation pipeline. After identification and optimization of proteotypic peptides within IgG constant domains, it was shown that normalization of the

AUC of an antigen peptide to that of an IgG peptide could significantly alter their relative quantitations such that increased variation between IP samples resulted in significant changes between pre- and post-normalized protein levels.

Experimental Procedures

Reagents and Cell Lines – All solvents were purchased from Fisher Scientific

(Rockford, Il), and all other chemicals were from Fisher Scientific, JT Baker (Center

Valley, PA), or Sigma (St. Louis, MO) unless otherwise noted.

52 Antibodies ordered from Abcam (Cambridge, MA) include rabbit-anti-GFP polyclonal ab290, rabbit-anti-β-actin polyclonal ab8227, rabbit-anti-catalase polyclonal ab1877, and mouse-anti-GAPDH monoclonal ab9484 clone number mAbcam9484.

Additional antibodies used were mouse-anti-β-actin monoclonal Sigma A5441 clone AC-

15 and mouse-anti-HA.11 monoclonal Covance (Princeton, NJ) MMS-101P clone

16B12.

HPLC purified peptides (>95% purity) were purchased from Thermo Scientific

(Rockford, Il).

Porcine aortic endothelial (PAE) cells stably expressed DAT constructs that were dually tagged with yellow fluorescent protein (YFP) on the N-terminus and hemagglutinin (HA) within the second extracellular loop. Parental cells were used along with full-length (FL), N-terminally deleted (ΔN), and C-terminally deleted (ΔC) YFP-

HA-DAT expressing cell lines. The FL and ΔN PAE cell lines were previously described33,131. The ΔC cell line was similarly generated, where the 30 C-terminal residues of DAT were removed, and the last two residues of the truncated protein were also mutated in order to retain the PDZ binding domain (LAY àLKV). PAE cells were grown in Ham’s F12 medium containing 10% fetal bovine serum (FBS).

Immunoprecipitation – Cells were grown to 80-90% confluency on 245 mm square plates from Corning (Corning, NY) before being rinsed 2 times with DPBS and harvested using a cell scraper on ice in 2 mL lysis buffer (20 mM Tris-HCl, pH8, 137 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA) with protease inhibitors

(aprotonin, leupeptin and pepstatin A at 1 µg/mL and bestatin at 4 µg/mL final concentrations) added immediately prior to use. Cell pellets were broken up via pipetting.

53 Samples were incubated with rotation for 30 minutes at 4°C and then centrifuged at

20,817 x g for 30 minutes at 4°C to pellet cellular debris. Antibody was added to the supernatant (55µg of rabbit αGFP antibody, or 30µg of mouse αHA antibody) and samples were incubated with rotation overnight at 4°C. Negative control samples were incubated overnight without antibody. Sepharose beads (protein A conjugated for rabbit

αGFP antibody samples, or protein G conjugated for mouse αHA antibody samples) were added at a ratio of 100µL beads to 1 mL of sample, and samples were nutated at 4°C for 4 hours. Samples were rinsed three times with 500µL lysis buffer before elution in 100µL

200mM glycine, pH 2.5 for rabbit αGFP antibody samples. Mouse αHA antibody samples were denatured in 100 µL Laemmli buffer without bromophenol blue by nutating

10 minutes at room temperature and boiling 5 minutes at 100°C. Both sample-types were then centrifuged for 15 minutes at 4°C and 20,817 x g. Supernatants were transferred to clean tubes. 50µg of bovine serum albumin (BSA) was added to IP samples as a carrier protein for further processing steps.

MS Sample Preparation – IPed samples and untreated antibodies were precipitated twice with methanol and chloroform as previously described87. Protein pellets were then resuspended in 0.2% RapiGest (Waters Corporation, Milford, MA) in

50 mM ammonium bicarbonate. Samples were then sonicated using a Microson

Ultrasonic cell disruptor (Misonix, Farmingdale, NY; 6 x 1 sec pulses, power 1.5), before being boiled for 5 minutes at 100°C. Samples were then diluted with 50 mM ammonium bicarbonate to 0.1% RapiGest before being sonicated again. Protein concentrations were determined using the DC protein assay kit (Bio-Rad, Hercules, CA). Samples were diluted to 0.5 or 1 mg/mL. Dithiothreitol (DTT) was added to a final concentration of 5

54 mM and samples were incubated at 60°C for 30 minutes to reduce reactive cysteines.

Iodoacetamide (IAA) was then added to a final concentration of 15 mM and samples were incubated for 30 minutes in the dark at room temperature in order to alkylate reduced cysteines. CaCl2 was added to a final concentration of 1 mM before modified trypsin (Promega, Madison, WI) was added to a 1:50 enzyme:substrate ratio. Samples were incubated overnight at 37°C using a Thermomixer (Eppendorf, Westbury, NY). HCl was added to a final concentration of 200 mM and incubated at 37°C for 45 minutes to hydrolyze the RapiGest. Samples were centrifuged at 20,817 x g and 4°C for 30 minutes and supernatant was collected twice before MS analysis. IP samples using the αHA antibody were cleaned using Pierce® C18 tips (100µL, Thermo Scientific, Rockford, IL), condensed using a CentriVap Centrifugal Concentrator Model 7810014 (Labconco,

Kansas City, MO), and resuspended in 0.1% formic acid to a final concentration of 0.25 mg/mL.

Western Blots – 10 µg of each IP sample was loaded onto a 10% SDS-PAGE gel, and were subjected to electrophoresis. Proteins were transferred to a PVDF membrane via semi-dry transfer. Monoclonal mouse or polyclonal rabbit primary antibodies directed to HA or GFP, respectively, and secondary antibodies from the appropriate animals conjugated with horseradish peroxidase were utilized. Bands were detected via enhanced chemiluminescence (ECL) using the Pierce SuperSignal West Pico Chemiluminescent

Substrate (Rockford, Il) and X-ray films. Films were semi-quantitatively measured using

Image J software.

NanoLC-Mass Spectrometry – Tryptic digests of antibodies were analyzed using a Proxeon EASY-nLC II and an Orbitrap Elite mass spectrometer with Xcaliber software.

55 The LC system was used to load 2 µg of antibody digest onto a fused silica (Polymicro

Tech, Phoenix, AZ) microcapillary column (75 µm i.d., 360 µm o.d. with a 5 µm laser pulled tip) packed with 15 cm of 5 µm 125 Å Aqua C18 reversed-phase (RP) material

(Phenomenex, Torrance, CA). Column temperature was maintained at 25°C using a column heater built in-house as previously described93. The mobile phase buffers used were Buffer A (98% water, 2% acetonitrile, 0.1% formic acid) and Buffer B (100% acetonitrile, 0.1% formic acid). Samples were analyzed with a 75 minute linear gradient at 200 nL/min that started at 100% Buffer A and reached 50% Buffer B by 65 minutes with an additional 2-minute increase to 95% Buffer B that was maintained for the final 8 minutes. A 15-minute wash and a 50-minute, 100 fmol bovine standard (6 Bovine Protein

Digest Equal Molar Mix, Michrom Bioresources, Auburn, CA) gradient were run between each of the five randomized replicate sets of samples. Data-dependent acquisition was used to collect tandem mass spectra at a resolution of 60,000. One high resolution MS profile full scan was followed by 10 MS/MS centroid ion trap scans using dynamic exclusion. Results were searched using SEQUEST110, processed with

Bullseye115, post-processed using Percolator113, and proteins were inferred using

MSDaPl143.

SRM analyses were conducted with IP tryptic digests and were analyzed using a

Proxeon EASY-nLC II and a TSQ Vantage Triple Stage Quadrupole mass spectrometer with Xcaliber software. The LC system was used to load 1 µg αHA IP or 2µg αGFP IP sample onto a microcapillary column, as described above but packed to 30 cm.

Temperature was maintained at 40°C as has been previously demonstrated effective for membrane samples93-95. Samples were analyzed with a previously optimized 75-minute

56 linear gradient at 250 nL/min starting at 100% Buffer A and reaching 30% Buffer B by

65 minutes with an additional 2-minute increase to 95% Buffer B that was maintained for the final 8 minutes. A 15-minute wash and a 50-minute 100 fmol bovine standard gradient were run between each of the three randomized replicate sets of samples.

SRM methods were created using Skyline103 for the analysis of peptides representing DAT, YFP, rabbit IgG (heavy and light chains), and mouse IgG1 (heavy and light chains). The DAT-YFP method analyzed 6 peptides, the rabbit IgG method analyzed 11 peptides, and the mouse IgG method analyzed 10 peptides. For each of these methods, doubly charged parent ions and between 3 and 12 singly charged product y-ions were monitored. Data were acquired with a Q1 and Q3 resolution of 0.7 m/z. Transitions were monitored with a dwell time of 0.020 s or 0.024 s with a total cycle time of 1 s

(Appendix A). The RF-only q2 collision cell was pressurized with 1.5 mTorr of argon gas. SRM data were analyzed using Skyline to measure and calculate AUC, coefficient of variance (CV), and retention time (RT).

The upper and lower limits of detection (LOD) and limits of quantitation (LOQ) were measured for four representative synthetic peptides, ALPAPIEK from the rabbit heavy chain, VTQGTTSVVQSFNR from the rabbit light chain, VNSAAFPAPIEK from the mouse heavy chain, and TSTSPIVK from the mouse light chain. Zero mol, 10 amol,

100 amol, 1 fmol, 10 fmol, 100 fmol, and 1 pmol of each of these peptides were coinjected with both 100 fmol bovine standard and 1 µg whole parental PAE cell lysate tryptic digest; the four representative peptides were analyzed via SRM. The doubly charged parent ion and three to five singly charged product y-ions were monitored with a

57 dwell time of 0.059 s (Appendix A). Otherwise SRM methods were the same as above.

Results were analyzed with Skyline.

Mouse immunoglobulin gamma chain, a) b) constant region c) N C Parental FL Δ Δ " 1 AKTTPPSVYP LAPGSAAQTN SMVTLGCLVK 30" 31 GYFPEPVTVT WNSGSLSSGV HTFPAVLQSD 60 " 61 LYTLSSSVTV PSSPRPSETV TCNVAHPASS 90" 250 91 TKVDKKIVPR DCGCKPCICT VPEVSSVFIF 120 " 121 PPKPKDVLTI TLTPKVTCVV VDISKDDPEV 150" 150 151 QFSWFVDDVE VHTAQTQPRE EQFNSTFRSV 180 " 181 SELPIMHQDW LNGKEFKCRV NSAAFPAPIE 210" 211 KTISKTKGRP KAPQVYTIPP PKEQMAKDKV 240 " 100 241 SLTCMITDFF PEDITVEWQW NGQPAENYKN 270" 75 271 TQPIMNTNGS YFVYSKLNVQ KSNWEAGNTF 300 " 301 TCSVLHEGLH NHHTEKSLSH SPGK 324" " Legend Mouse immunoglobulin kappa light chain HV – Heavy Chain Variable Region HC – Heavy Chain Constant Region " L – Light Chain Variable Region 1 ELVMTQSPLS LSVSLGDQAS ISCRSSQSLV 30" V LC – Light Chain Constant Region 50 31 HTNGNTYLHW YLQKPGLSPK LLIYIVSNRF 60 " – K.DVLTITLTPK.V 61 SGVPDRFSGS GSGTDFTLKI SRVEAEDLGV 90" – R.VNSAAFPAPIEK.T 91 YFCSQSTHVP GTFGGGTKLE IKRADAAPTV 120 " – K.APQVTYIPPPK.E – R.QNGVLNSWTDQDSK.D 121 SIFPPSSEQL TSGGASVVCF LNNFYPKDIN 150" – K.DSTYSMSSTLTLTK.D 151 VKWKIDGSER QNGVLNSWTD QDSKDSTYSM 180 " – K.TSTSPIVK.S 181 SSTLTLTKDE YERHNSYTCE ATHKTSTSPI 210" 211 VKSFNRGEC" "

Figure 4.2 – Characterization of mouse IgG. a) Western blot showing DAT (upper box) and IgGs (lower box) in the four IP samples. The blot was probed with mouse anti-HA primary antibody followed by anti-mouse HRP secondary antibody that recognized both the anti-HA primary antibody attached to the DAT and that which was used in the IP. Upper bands indicate multimerized DAT, center bands represent mature glycosylated DAT, and the lowest bands represent unmodified DAT. b) Peptides are highlighted within the mouse IgG constant regions. c) Peptides are highlighted on a cartoon representation of an antibody molecule.

Normalization – IgG peptide AUCs of three technical replicates were averaged for each peptide within each sample. Individual DAT and YFP peptide AUCs were each divided by the average IgG peptide AUC to create normalized AUCs. Unaltered and normalized DAT/YFP AUCs were all divided by the average FL-YFP-HA-DAT AUC of the specific dataset, such that all numbers ranged between 0 and 1 in order to compare pre- and post-normalization values. T-tests (2-tailed, type 2) were performed in Microsoft

Excel comparing pre- and post-normalized AUC values for ΔN-YFP-HA-DAT and ΔC-

58 YFP-HA-DAT IP samples. In these calculations, four DAT/YFP peptides were normalized to each of the six IgG peptides, in both rabbit and mouse samples.

H H a) b) Rabbit immunoglobulin gamma chain, c) V V N C constant region LV LV Parental FL Δ Δ ! 1 GQPKAPSVFP LAPCCGDTPS STVTLGCLVK 30! LC LC 31 GYLPEPVTVT WNSGTLTNGV RTFPSVRQSS 60 ! 61 GLYSLSSVVS VTSSSQPVTC NVAHPATNTK 90! 91 VDKTVAPSTC SKPTCPPPEL LGGPSVFIFP 120 ! 250 121 PKPKDTLMIS RTPEVTCVVV DVSQDDPEVQ 150! H H 151 FTWYINNEQV RTARPPLREQ QFNSTIRVVS 180 ! C C 130 181 TLPITHQDWL RGKEFKCKVH NKALPAPIEK 210! 211 TISKARGQPL EPKVYTMGPP REELSSRSVS." 240 ! 100 241 LTCMINGFYP SDISVEWEKN GKAEDNYKTT 270! Legend 271 PAVLDSDGSY FLYNKLSVPT SEWQRGDVFT 300 ! HV – Heavy Chain Variable Region 301 CSVMHEALHN HYTQKSISRS PGK 323! H – Heavy Chain Constant Region 70 C LV – Light Chain Variable Region LC – Light Chain Constant Region Rabbit immunoglobulin kappa-b4 chain, – K.DTLMISR.T constant region – R.EQQFNSTIR.V ! – K.ALPAPIEK.T 55 – R.GQPLEPK.V 1 DPVAPTVLIF PPAADQVATG TVTIVCVANK 30 ! – K.LSVPTSEWQR.G 31 YFPDVTVTWE VDGTTQTTGI ENSKTPQNSA 60 ! – K.VTQGTTSVVQSFNR.G 61 DCTYNLSSTL TLTSTQYNSH KEYTCKVTQG 90 ! 91 TTSVVQSFNR GDC 103 ! !

Figure 4.3 – Characterization of rabbit IgG. a) Western blot showing DAT (upper box) and IgGs (lower box) in the four IP samples. The blot was probed with the rabbit αGFP primary antibody followed by an anti-rabbit HRP secondary antibody that recognized the αGFP primary antibody attached to the DAT and that which was leftover from the IP. Upper bands indicate multimerized DAT, center bands represent mature glycosylated DAT, and the lowest bands represent unmodified DAT. b) Peptides are highlighted within the rabbit IgG constant regions. c) Peptides are highlighted on a cartoon representation of an antibody molecule.

Results and Discussion

This study developed a relative normalization and quantitation pipeline for IP-MS samples through SRM analysis of IgG constant domain peptides. Tryptic digests of monoclonal and polyclonal antibodies were profiled using nanoLC-MS/MS to select for the best proteotypic IgG constant domain peptides for subsequent quantitative nanoLC-

SRM assays. IP experiments were then conducted using multiple distinctive DAT constructs as the targeted antigens. Rabbit polyclonal and mouse monoclonal antibodies were utilized for IPs and subsequent analysis, as they are the most widely used types of antibodies in IP experiments140. NanoLC-SRM was performed on these IP samples,

59 targeting both DAT and IgG tryptic peptides. AUCs of IgG peptides were then used for the relative normalization of DAT peptides.

Densitometry is not always an accurate representation of protein abundances in a sample due to the limited dynamic range of ECL, the most common method of western blot detection138,139. Thus, this method does not always produce a precise measurement for loading controls. For example, western blots of samples IPed with αGFP or αHA antibodies included bands corresponding to multiple forms of DAT as well as bands corresponding to IgGs (Figures 4.2a, 4.3a). However, fewer DAT bands are present in the lanes where DAT is less abundant. Particularly in the mouse αHA IP blot, it is difficult to go beyond a qualitative analysis of the samples through western blot (Figure 4.2a).

Replicate western blots of IPs were conducted and densitometry was used for quantification. The densities of ΔC-DAT and FL-DAT changed drastically between different exposure times as well as between reprobing of the same blot with an αGFP primary antibody after an αDAT primary antibody, as the ΔN-DAT cannot be detected by the αDAT antibody and was much more abundant than the other two DAT constructs in this particular study (Figure 4.1a, Table 4.1).

Accurately measuring the levels of IgGs, which are already exposed on the same blots, could help to normalize DAT densitometric values, however these bands are also overexposed on the blots, since these must be highly abundant in order to pull down a substantial amount of DAT (Figure 4.3b,c). Thus, the quantitative analysis of DAT levels based on western blot densitometry was inconclusive in these IP samples due to the limited dynamic range of the method.

60 10 Second Exposure 30 Second Exposure 2 Minute Exposure 1° Sample Avg. Std. Antibody Std. Dev. Avg. Area Std. Dev. Avg. Area Area Dev. FL-DAT αGFP 1554.3 1917.4 7918.0 5946.3 39811.3 23275.3 ΔN-DAT αGFP 90262.8 10821.9 123380.8 4378.0 149806.3 12033.8 ΔC-DAT αGFP 2797.9 2748.3 10487.8 8059.2 39069.9 20465.4 FL-IgG αGFP 27536.5 6530.3 34955.6 4277.3 56987.2 2652.0 ΔN-IgG αGFP 31313.6 2705.6 38245.1 1698.5 54403.5 5013.5 ΔC-IgG αGFP 27609.5 7802.6 35395.0 7538.0 50085.0 4658.6 30 Second Exposure 2 Minute Exposure 5 Minute Exposure FL-DAT αDAT 98144.2 5985.3 119577.4 15001.7 139154.4 10814.9 ΔN-DAT αDAT ------ΔC-DAT αDAT 7209.3 1352.3 21933.3 3235.8 37623.3 1773.9

Table 4.1 – Western Blot Densitometry. Three replicate western blots of FL, ΔN, and ΔC αGFP IPs were performed, probing with both an αDAT primary antibody as well as an αGFP antibody. In the αGFP blots, the ΔN was more prominent than the other lanes, while in the αDAT blots ΔN cannot be detected since the antibody is directed toward the N-terminus of DAT. When the FL and ΔC bands were measured for densitometry using ImageJ, at three different exposure times the differences between the densities vary dramatically. In the αGFP blots, IgGs are detected along with DAT. With different exposure times, the variation within these measurements can be seen.

To determine the representative peptides for IgG constant domains, three rabbit and two mouse antibodies were digested with trypsin and analyzed by nanoLC-MS/MS.

Multiple IgG heavy and light chain peptides were identified across samples and replicates from the same samples (Table 4.2). The peptides identified within this study, as well as additional peptides identified in previous IP-MS/MS studies (data not shown), were selected for targeted analysis of IgG levels in IPed samples.

To select the optimal proxy peptides for quantitative analysis of IgGs, thirteen tryptic rabbit IgG peptides and ten tryptic mouse IgG1 peptides were identified via nanoLC-MS/MS were monitored using nanoLC-SRM. These peptides reside within both the heavy chain and light chain constant regions for both animals, and were monitored in

αGFP IP samples or αHA IP samples from parental, FL-DAT, ΔN-DAT, and ΔC-DAT

PAE cells. Three technical replicates of all eight samples and corresponding bead only

61 control samples were produced. In both sets of IPs, six IgG peptides (Figures 4.2b,c,

4.3b,c) were reproducibly observed between biological samples and technical replicates but not in bead only control samples (Figures 4.4a, 4.5a).

Heavy Chain Constant Region Light Chain Constant Region

% % # # # # Animal Antigen Sequence Sequence Peptides Spectra Peptides Spectra Coverage Coverage Rabbit GFP 51.08 14 59 79.61 6 51

Rabbit Catalase 72.14 56 295 91.26 21 203

Rabbit β-Actin 30.96 8 27 26.21 1 1

Mouse GAPDH 36.01 13 35 52.06 5 9 (IgG2b)

Mouse β-Actin 47.53 11 56 66.66 10 40 (IgG1)

Table 4.2 – Antibody Profiling Characterization. Multiple rabbit and mouse antibodies were digested with trypsin and underwent nanoLC-MS/MS on an Orbitrap Elite spectrometer in order to identify candidate peptides for further SRM analysis.

The AUC of each of these twelve peptides was calculated using Skyline. One-way analysis of variance (ANOVA) was used to compare the AUCs of each of the IgG peptides between IPs from the four cell lines. This analysis showed that with the rabbit antibody, only two of the peptides (GQPLEPK and LSVPTSEWQR) had statistically different (p<0.05) abundances between samples. However, with the mouse antibody, only one of the monitored peptides was not present at statistically different levels between samples (TSTSPIVK). Thus, the rabbit αGFP IP experiment appears to have been more uniformly performed, as there was less variance in IgG peptide abundance between

62 samples. Whereas in the mouse αHA IP experiment, IgG abundances significantly varied

between samples.

a) Parental& Full&Length& ΔN& ΔC& b) c) 1.0E+08' VNSAAFPAPIEK$ TSTSPIVK$ 1.0E+08& 1.0E+08' 1.0E+08' 1.0E+07& 1.0E+06' 1.0E+07' 1.0E+07' *

1.0E+06& $ 1.0E+04' 1.0E+06' 1.0E+06' 1.0E+05& 1.0E+05' 1.0E+05'

AUC$(log$scale)$ 1.0E+02' * 1.0E+04& 1.0E+04' 1.0E+04' AUC$(log$scale)$ AUC$(log$scale)$ 1.0E+03' 1.0E+03' 1.0E+03& 1.0E+00' 1.0E+00' 1.0E+02' 1.0E+04' 1.0E+06' 1.0E+00' 1.0E+02' 1.0E+04' 1.0E+06' AUC*(log*scale) 1.0E+02& Amount$of$Pep5de$(amol)$ Amount$of$Pep5de$(amol)$ 1.0E+01& - In Bovine Matrix - In Lysate Matrix

1.0E+00& K.TSTSPIVK.S$$ K.DVLTITLTPK.V$$ K.APQVYTIPPPK.E$ R.VNSAAFPAPIEK.T$ K.DSTYSMSSTLTLTK.D$ K.DTLMISR.T* K.ALPAPIEK.T* R.GQPLEPK.V*

R.EQQFNSTIR.V* R.QNGVLNSWTDQDSK.D$ K.LSVPTSEWQR.G* Heavy Chain Light Chain K.VTQGTTSVVQSFNR.G* Peptides Peptides Heavy Chain Peptides Light Chain Peptide

Figure 4.4 – Mouse IgG SRM. a) SRM analysis of 6 mouse IgG tryptic peptides within 4 IP samples. AUCs of each peptide were significantly different between samples based on ANOVA testing (p<0.05), except for TSTSPIVK. b) + c) LOD curves of two representative mouse IgG1 peptides.

To determine the dynamic range of detection of the peptides of interest, four

synthetic representative peptides were coinjected at varying concentrations with either

bovine matrix or cell lysate matrix and monitored via SRM. This study determined that

three of the four peptides could be detected at levels between 100 amol and 10 pmol in

bovine matrix, while in cell lysate matrix all four peptides could be detected in the 1

fmol-10 pmol range (Figures 4.4b,c, 4.5b,c).

Antigen levels were also monitored via SRM. Two YFP and two DAT peptides

expressed in all three DAT constructs (Figure 4.6a) were monitored via SRM in

conjunction with IgG peptides in both αGFP and αHA IPs of parental, FL-DAT, ΔN-

DAT, and ΔC-DAT PAE cells. Normalization was performed by dividing the replicate

AUCs of a peptide from the target protein (YFP-DAT) by the average AUC of an IgG

peptide within the same sample. This resulted in post-normalization AUC values for each

63 YFP-DAT peptide. Pre- and post-normalization values were compared between different conditions by dividing all values by the average FL-DAT AUC (pre- or post- normalization, respectively) such that all values were set on a scale from 0-1 (Figure

4.6b-e, 4.7a-d). T-tests were conducted to compare the pre- and post-normalized ΔN-

DAT and ΔC-DAT abundances. These were performed for all four target peptides with all twelve IgG peptides (with both rabbit and mouse antibodies, Appendix B). In rabbit

αGFP IPs the differences between the pre- and post-normalized values were less significant than those from the mouse αHA IPs. Three normalized ΔN abundance values were significantly different (p<0.05) from the corresponding pre-normalized values in the rabbit αGFP IP samples, while none of the ΔC abundance values underwent significant changes with normalization. However, in the mouse αHA IP samples, 10 of the ΔN and

18 of the ΔC abundance values resulted in significant changes after normalization

(p<0.05).

a) Parental' Full'Length' ΔN' ΔC' b) ALPAPIEK$$ c) VTQGTTSVVQSFNR$

1.0E+08' 1.0E+08' 1.0E+08' 1.0E+07' 1.0E+07' 1.0E+07' 1.0E+06' 1.0E+05' 1.0E+06' 1.0E+06'

1.0E+04'$ 1.0E+05' 1.0E+05' 1.0E+03' 1.0E+02' 1.0E+04' 1.0E+04' AUC$(log$scale)$ AUC$(log$scale)$

AUC$(log$scale)$ 1.0E+01' 1.0E+00' 1.0E+03' 1.0E+03' 1.0E+00' 1.0E+02' 1.0E+04' 1.0E+06' 1.0E+00' 1.0E+02' 1.0E+04' 1.0E+06' Amount$of$Pep5de$(amol)$ Amount$of$Pep5de$(amol)$ K.DTLMISR.T$ K.ALPAPIEK.T$ R.GQPLEPK.V$ - In Bovine Matrix - In Lysate Matrix R.EQQFNSTIR.V$ K.LSVPTSEWQR.G$ K.VTQGTTSVVQSFNR.G$

Heavy Chain Light Chain Peptides Peptide

Figure 4.5 – Rabbit IgG SRM. a) SRM analysis was conducted on 6 rabbit IgG tryptic peptides within 4 IP samples. AUCs of each peptide were mostly similar between samples, except for GQPLEPK and LSVPTSEWQR which had significantly different AUCs between samples based on ANOVA testing (p<0.05). b) + c) LOD curves of two representative rabbit IgG peptides.

64 The significant difference between IPs conducted with the different antibodies is not surprising. There was more inter-sample variation in IgG abundances with the mouse antibody samples than there was with rabbit antibody samples (Figures 4.4a, 4.5a). In fact, there is a loose correlation between significant differences in peptide abundance between IP samples via ANOVA analysis and significant changes between pre- and post- normalized abundance values. Peptides that had significantly different abundances between samples when compared with ANOVA were more likely to result in significantly different pre- and post-normalization values (Appendix B). For example, all three of the significant changes after normalization in the rabbit αGFP IP samples occurred with normalization to the LSVPTSEWQR peptide, which was shown to have significantly different levels between samples. Similarly, in the mouse αHA IPs, normalization to the TSTSPIVK peptide, which did not have significantly different levels between samples, only resulted in one significant normalization value, the least of any of the mouse IgG peptides. Thus, more significant changes occurred with normalization when there was a more significant difference in peptide abundance between the compared samples.

Summary

The use of appropriate loading controls is imperative in quantitative and semi- quantitative experiments. For IP-MS, IgGs are an excellent loading control option that has been frequently overlooked. In this study, we propose a pipeline that uses multiple

IgG constant domain peptides for relative normalization and quantitation of the antigen of interest within IP samples using both mouse and rabbit antibodies.

65 a)# YFP DAT

ΔN HA ΔC Legend$ ##### ΔN#–#N0Terminal#Dele:on######–#FEGDTLVNR# ΔC#–#C0Terminal#Dele:on######–#GVTLPGAIDGIR####### ######–#SAMPEGYVQER## #####–#AYLSVFYR###### b)## c)# SAMPEGYVQER$ FEGDTLVNR$

1.5# 1.4# 1.2# 1.0# 1.0# 0.8# **# **# 0.6# ***# 0.5# 0.4# DAT$Abundance$ DAT$Abundance$ 0.2# 0.0# 0.0# FL# ΔN# ΔC# FL# ΔN# ΔC# d)# e)# GVTLPGAIDGIR$ AYLSVDFYR$

1.4# 1.4# 1.2# 1.2# 1.0# 1.0# *# 0.8# 0.8# 0.6# ***# 0.6# 0.4# 0.4# DAT$Abundance$ 0.2# DAT$Abundance$ 0.2# 0.0# 0.0# FL# ΔN# ΔC# FL# ΔN# ΔC# ####0#Average#AUC#######0#Average#Normalized#AUC#

Figure 4.6 – Relative normalization of YFP-DAT AUCs to mouse IgG AUCs. a) Schematic of YFP and DAT peptide locations. b-e) Comparison of pre- and post- normalization DAT abundance using one example mouse IgG heavy chain peptide (VNSAAFPAPIEK) to normalize two YFP (b,c) and two DAT (d,e) peptides. (* p<0.10, **p<0.05, ***p<0.01).

66 a) SAMPEGYVQER$ b) FEGDTLVNR$ 1.40# 1.40#

1.20# 1.20#

1.00# ** 1.00#

0.80# 0.80#

0.60# 0.60#

DAT$Abundance$ 0.40# DAT$Abundance$ 0.40# $(%$of$Full$Length$AUC)$ $(%$of$Full$Length$AUC)$ 0.20# 0.20#

0.00# 0.00# Parental# Full# ΔN# ΔC# Parental# Full# ΔN# ΔC# Length# Length# c) GVTLPGAIDGIR$ d) AYLSVDFYR$ 1.20# 2.00# 1.80# 1.00# 1.60# 0.80# 1.40# 1.20# * 0.60# 1.00# ** 0.80# 0.40# DAT$Abundance$ DAT$Abundance$ 0.60#

$(%$of$Full$Length$AUC)$ 0.20# $(%$of$Full$Length$AUC)$ 0.40# 0.20# 0.00# 0.00# Parental# Full# ΔN# ΔC# Parental# Full# ΔN# ΔC# Length# Length# Average AUC Average Normalized AUC

Figure 4.7 – Normalization of YFP-DAT Peptides to Rabbit IgGs. Comparison of pre- and post-normalization DAT abundance using one example of one rabbit IgG heavy chain peptide (LSVPTSEWQR) to normalize two YFP (a,b) and two DAT (c,d) peptides. (* p<0.05, **p<0.01).

Within the proposed pipeline, after an IP is performed with multiple samples, IgG and antigen peptides are monitored via SRM. The resultant average DAT AUCs are divided by the average IgG AUCs to create a post-normalized AUC value. When the pre- and post-normalized AUCs are compared there can be significant differences between the two values. This indicates that there are varying IgG levels between IP samples. As the changes in normalized values show, this can result in faulty quantitation in current IP experiments, since existing IP protocols do not generally account for such input variations. Such changes can be noticed or possibly monitored via western blot, but due

67 to the limited dynamic range of the method, it is not always possible to compare the IgG levels while also monitoring the levels of antigen. However, when using SRM to compare the IgG levels, it is possible to compare both the IgG levels between samples and the IgG levels to the antigen levels within the same sample, assuming the antigen is concentrated enough for detection.

Thus, this study has shown that relative normalization of target peptides to IgG peptides can result in significant changes in relative peptide abundance using both monoclonal and polyclonal antibodies from mouse and rabbit, respectively. Particularly, if there is a significant difference in IgG peptide AUC between IP conditions, then it is likely that relative normalization will result in a significant change in target protein peptide levels. This normalization pipeline is used in our future IP experiments for not only DAT, but for co-precipitating proteins of interest as well.

68 CHAPTER V

ANALYSIS OF THE DAT INTERACTOME IN PORCINE AORTIC

ENDOTHELIAL CELLS

Introduction

In Chapter IV, the concept of combining immunoprecipitation (IP) and mass spectrometry (MS) was introduced. Inherent issues with this process were discussed and a method for alleviating such issues was presented. In this chapter, that method will be applied not only to DAT levels, but to associated protein levels as well. As was previously mentioned, three DAT constructs were produced for analysis. These constructs were created in the Sorkin lab with a yellow-fluorescent protein (YFP) tag on the N-terminus of the protein along with a hemagglutinin (HA) tag located within the second extracellular loop. The creation of the full-length (FL) dually tagged construct was the first instance of a successful epitope tag insertion within an extracellular loop of the SLC6 family (Figure 5.1). When stably or transiently expressed in both HeLa and porcine aortic endothelial (PAE) cells, the construct was present at levels previously experienced with other DAT constructs. It was also found to undergo trafficking such that it is expressed both on the plasma membrane and internally. Additionally, the specific location of the HA tag within the extracellular loop does not appear to interfere with N- glycosylation, as evidenced by the presence of both glycosylated and non-glycosylated molecular weight bands on western blots. The construct was also shown to be functional

3 in PAE cells, with apparent affinity (KD) and maximal velocity (Vmax) values for H-DA

69 uptake similar to those of unlabeled DAT in the same system33. Thus, this construct appears to be fully functional with regular expression patterns.

After the preliminary study on the FL construct, an additional dually-tagged construct was created, in which the 65 amino acids of the N-terminus were removed (ΔN,

Figure 5.1). A specialized endocytosis assay was created that initially labels DAT in formaldehyde-fixed cells using an HA antibody. Secondary antibody labeled with cyanine 5 (Cy5) fluorescent dye is added to bind DAT on the plasma membrane. The membrane is then permeabilized and an additional secondary antibody labeled with cyanine 3 (Cy3) fluorescent dye is added to bind remaining unlabeled DAT, located inside the cell. These two labels can be distinguished through fluorescence microscopy to determine the amount of both internalized and surface DAT. Applying this method to

PAE and HeLa cells stably expressing either FL- or ΔN-YFP-HA-DAT showed an increased level of internalization in the ΔN-YFP-HA-DAT expressing cells. By adding a recycling inhibitor, monesin, to PAE cells that stably expressed either FL- or ΔN-YFP-

HA-DAT, it was determined that this increase in internalized DAT in the ΔN construct cells was due to an increase in internalization and not in the alternative decrease in recycling131. Further mutational analysis of the N-terminus of DAT showed that residues

60-65 were specifically responsible for the increase in internalization, by using a construct where those six residues were all substituted with alanine residues.

Furthermore, both the ΔN and 60A65 mutants were shown to have negligible [3H]DA uptake levels. Taken together, these data indicate that the N-terminus, specifically residues 60-65, is critical for proper DAT trafficking and functionality. This region has been associated with the stabilization of the outward-facing conformation, and its

70 involvement in endocytosis indicates that there may be a relationship between DAT conformation and trafficking131.

FL-YFP-HA-DAT ΔN-YFP-HA-DAT ΔC-YFP-HA-DAT HA HA HA

YFP LKV

YFP YFP

Figure 5.1 – YFP-HA-DAT Constructs and Expression. Three dually-tagged DAT constructs were expressed in PAE cells. All three had a YFP tag on the N-terminus and an HA tag within the second extracellular loop. Constructs included full length (FL), N- terminally deleted (ΔN), and C-terminally deleted versions (ΔC). As is shown by YFP fluorescence microscopy, the FL construct is primarily expressed on the plasma membrane, the ΔN construct is found on the plasma membrane with increased expression in endosomes, while the ΔC construct is found largely in the ER. These microscope images are from the Sorkin lab, published131 and unpublished data. Insets represent high magnification of indicated region for FL and ΔN, and a representative ER fluorescence pattern for ΔC144. Scale bars, 10 µm.

Another dually tagged mutant DAT construct was created that lacks the C- terminus (ΔC, Figure 5.1). This construct is truncated by 27 residues, where the final

LAY residues are substituted to LKV in order to retain the PDZ binding domain, to which the protein interacting with C kinase-1 (PICK1) has been shown to bind. This unpublished mutant construct is primarily expressed in the endoplasmic reticulum (ER), indicating a trafficking defect created by the loss of the C-terminus. These three

71 constructs (FL, ΔN, and ΔC) allow for the analysis the DAT interactome in three distinctive DAT trafficking systems.

Throughout the remainder of this dissertation, the DAT interactome will be characterized in multiple systems through the coupling of IP, which was described in the previous chapter, to MS, which was thoroughly detailed in Chapter II. The dual YFP and

HA tags present in the aforementioned DAT constructs allow for IPs using either YFP or

HA antibodies. Additionally, a parental cell line was used as a negative control to limit false-positive interactors.

In standard MS/MS profiling studies, DDA analysis can often have misleading results. Spectral count analysis, as its name implies, counts the number of spectra that identify particular peptides in order to compare peptide and protein abundances.

However, the semi-random sampling involved in this process can skew this sort of analysis such that less abundant peptides often go unidentified in some samples due to inter-run inconsistencies in MS/MS triggering events116. Recent instrument and programming advancements have allowed for more accurate comparisons between DDA peptide IDs based on chromatographic alignment of peptide identifications. Aligned, high mass accuracy MS1 peaks can then be quantified through integration of the areas under the curve for extracted ion chromatograms (XICs) in the program Topograph119,120. This process, known as label-free differential analysis, results in improved reproducibility because peptides that were identified in one sample that may not have triggered an

MS/MS event in another sample can now be quantified based on MS1 peak intensity in

DDA data.

72 This chapter employed IP-MS/MS to analyze changes in the DAT interactome between three different dually tagged DAT constructs expressed in PAE cells. As the distinct DAT mutants all undergo different trafficking patterns, it was expected that there would be corresponding differences in their interacting proteins. These differences were relatively quantified based on DDA data through chromatographic alignment. Further quantitative analysis was conducted through application of the normalization process described in Chapter IV.

Experimental Procedures

Reagents and Cell Lines – All solvents were purchased from Fisher Scientific

(Rockford, Il), and all other chemicals were from Fisher Scientific, JT Baker (Center

Valley, PA), or Sigma (St. Louis, MO) unless otherwise noted. The rabbit anti-GFP primary polyclonal ab290 antibody was ordered from Abcam (Cambridge, MA). Parental

PAE cells along with those stably expressing FL, ΔN, and ΔC YFP-HA-DAT constructs were grown as described in the previous chapter.

Immunoprecipitation – Two plates of each of the four cell types were grown to

80-90% confluency on 245 mm square plates from Corning (Corning, NY) before being rinsed 2 times with DPBS and harvested using a cell scraper on ice in 2 mL DPBS. Plates from each cell type were pooled together before centrifugation for 5 minutes at 1000 x g and 4°C. Supernatant was removed and cell pellets were frozen at -80°C for approximately one month before being thawed on ice for IP. Pellets were resuspended in

4mL lysis buffer (described in Chapter IV) with Halt Protease and Phosphatase Inhibitor

Cocktail (Thermo Scientific, Rockford, IL). IPs were then performed as described in

73 Chapter IV using the 22 µg of anti-GFP primary antibody per sample with all other amounts of reagent doubled to account for the increase in starting material.

MS Sample Preparation – Protein precipitation and digestion of IP samples were performed as described in Chapter IV. Digests were cleaned using Pierce® C18 Spin

Columns (Thermo Scientific, Rockford, IL), condensed using a CentriVap Centrifugal

Concentrator Model 7810014 (Labconco, Kansas City, MO), and resuspended in 0.1% formic acid to a final concentration of 0.25 mg/mL.

NanoLC-Mass Spectrometry – Tryptic digests of IP samples were analyzed similarly to samples in Chapter IV, using a Proxeon EASY-nLC II and an Orbitrap Elite mass spectrometer with Xcaliber software. The major difference for this analysis was the maintenance of column temperature at 40°C instead of 25°C to aid in the separation of hydrophobic peptides from membrane proteins94,95,124. Additionally, the analytical gradient used with these samples differed slightly. Samples were analyzed with a previously optimized 70 minute gradient at 250 nL/min that started at 100% Buffer A, ramped up to 5% Buffer B by 5 minutes and reached 30% Buffer B by 50 minutes with an additional 2 minute increase to 80% Buffer B that maintained for 5 minutes before a 2 minute decrease back to 100% Buffer A which was maintained for the final 11 minutes.

A 15 minute wash and a 50 minute 100 fmol bovine standard (6 Bovine Protein Digest

Equal Molar Mix, Michrom Bioresources, Auburn, CA) gradient were run in between each of the four randomized replicate sets of samples. Data-dependent acquisition was used to collect tandem mass spectra at a resolution of 60,000. One FTMS profile full scan was followed by 10 MS/MS centroid ion trap scans using dynamic exclusion.

74 Data Analysis – Profiling results were deconvoluted using Hardklör114 and

Bullseye115, searched using SEQUEST110 against a homemade human and porcine concatenated database. This combinatorial database was necessary because the porcine database alone was not well enough curated. High sequence homology between human and porcine sequences allowed for increased identifications. Data were post-processed using Percolator113. For initial analyses, data were managed and peptides were inferred using MSDaPl143. For later analyses, Peptide identifications were aligned and MS1 XIC peak areas were calculated using Topograph119,120. Peak areas were normalized within each dataset relative to the sum of the total peptide MS1 XIC areas across all injections.

All MS1 XIC were normalized to immunoglobulin levels based on the total sum of the 187 IgG peptides identified across the three technical replicates of each sample.

Student’s t-tests were conducted between parental and DAT samples to identify differentially abundant peptides. Resulting p-values were converted to q-values using the

R software program TkQVALUE145. Differential peptides were considered significant with q-values < 0.05. Additionally, peptide abundances ratios were compared between samples. Average protein abundance ratios were further examined using Ingenuity

Pathway Analysis (IPA, Ingenuity®Systems, www.ingenuity.com).

Results and Discussion

In this chapter, cells expressing multiple epitope tagged DAT constructs were

IPed using antibodies directed to the tags. These IP samples were then analyzed through

DDA profiling. Comparisons were conducted between samples through label-free differential analysis coupled with normalization to immunoglobulin levels. Peptide levels

75 were compared between sample via ratio analysis to determine the most abundant and most altered protein interactors between samples.

Dually tagged DAT constructs, containing an N-terminal YFP tag along with an

HA tag in the second extracellular loop, were stably expressed in PAE cells. Samples of all three cell lines (FL, ΔN, and ΔC) along with the parental cell line were enriched for

DAT through IP using an antibody directed toward GFP, which also detects YFP. After tryptic digestion, these four IP samples were analyzed through top ten MS/MS analysis using an Orbitrap Elite mass spectrometer, with three technical replicates of each sample.

Figure 5.2 – DAT Sequence Coverage. Variability in sequence coverage between samples is shown here as a result of spectral count analysis. Vertical black lines indicate the locations of truncations in ΔN and ΔC DAT.

For initial data analysis, peaks were deconvoluted with Hardklör114 and

Bullseye115, data were searched using SEQUEST110 against a concatenated human and porcine database, which allowed for increased identification as the porcine database alone

76 is not well curated. Prior to further analyses the data were then post-processed with

Percolator113. Peptides were then inferred using the Mass Spectrometry Data Platform

(MSDaPl)143. Spectral count analysis of these data showed that there was a much higher sequence coverage of DAT in the FL IP, at 23.1%, than in either ΔN or ΔC IPs, at 8.7% and 14.0%, respectively (Figure 5.2). Overall these sequence coverage numbers seem rather low, however, when such percentages are recalculated in respect to location in the membrane, the peptide identifications account for 45.9%, 15.7%, and 23.44% of the total non-TMD sequences. These levels are a vast improvement, and are reminiscent of the findings of Chapter III, which discussed the issues with the identification of highly hydrophobic membrane peptides. a) 1.05E+11 Sum of IgG Peptides b) Sum of Tryptic DAT Peptides 1E+11 8E+09 9.5E+10 6E+09 9E+10 Abundance 4E+09 8.5E+10 2E+09 8E+10 Normalized Abundance Abundance Normalized Par FL ∆N ∆C Par FL ∆N ∆C 0 Par FL ∆N ∆C Pre-normalization Post-normalization Figure 5.3 – IgG and DAT peptide sums. a) IgG abundances are shown as a sum of all IgG and IgK constant region peptides before and after normalization. Prior to normalization, ANOVA showed significant differences (p<0.05) between samples; while after normalization these values were no longer significantly different. b) Summed DAT abundance showed that DAT levels are higher in FL compared to both ΔN and ΔC IPs.

However, there is still a large discrepancy between the DAT peptide levels across these samples. From this method of data analysis, it is impossible to say whether the additional peptides identified only in the FL sample are due to an increase in protease accessibility to the particular regions of interest. Alternatively, these discrepancies could

77 simply be a result of lesser abundance in the terminally deleted constructs, or a combination of the two reasons. In order to better examine this question, a more quantitative approach is necessary. Such an approach is possible with the use of label-free differential analysis to achieve more reliable quantitation coupled with the immunoglobulin normalization method, discussed in Chapter IV, to account for inter- sample deviations.

Topograph was used to chromatographically align percolated data. All resulting data were then normalized by sample based on the total sum of IgG peptides. Using the sum of multiple peptides to create a normalization factor limits the influence of any one peptide and allows for a more general representation of IgG abundance. Prior to normalization, the IgG sums across the samples were significantly different based on

ANOVA (p<0.05). However, after analysis, ANOVA showed no significant differences between these values (Figure 5.3a). The sums of tryptic DAT peptides across samples indicate that the FL sample indeed has an increased DAT abundance when compared to the truncated mutants (Figure 5.3b). This corresponds to western blot analysis as seen in

Chapter IV (Figure 4.3). However, this analysis also revealed that although not all of the aforementioned DAT peptides triggered MS/MS events in all samples there was still evidence of their presence. Correlated MS1 peaks for each of these peptides were present in all replicates of each of the DAT samples and replicates, except for those peptides within the deleted regions of the ΔN and ΔC samples (Figure 5.4). This difference in

DAT peptide identification and sequence coverage instills additional confidence in the differential analysis method over more traditional spectral counting methods.

78 K.CSVGLMSSVVAPAK.EK.CSVGLMSSVVAPAK. K.EQNGVQLTSSTLTNPR.QK.EQNGVQLTSSTLTNP R.QSPVEAQDR.E

1.6E+08 8.0E+08 8.0E+06

1.2E+08 6.0E+08 6.0E+06

8.0E+07 4.0E+08 4.0E+06

4.0E+07 2.0E+08 2.0E+06

Normalized Abundance Abundance Normalized 0.0E+00 0.0E+00 0.0E+00 Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C

R.EGAAGVWK.I R.GVLHLHQSHGIDDLGPPR.WR.GVLHLHQSHGIDDLG R.GVTLPGAIDGIR.A

8.0E+07 2.5E+08 5.0E+08

2.0E+08 4.0E+08 6.0E+07 1.5E+08 3.0E+08 4.0E+07 1.0E+08 2.0E+08 2.0E+07 5.0E+07 1.0E+08 Normalized Abundance Abundance Normalized 0.0E+00 0.0E+00 0.0E+00 Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C

R.AYLSVDFYR.L K.FTNNCYR.D K.HSVPIGDVAK.D

6.0E+08 1.2E+07 3.0E+08

4.8E+08 2.4E+08 9.0E+06 3.6E+08 1.8E+08 6.0E+06 2.4E+08 1.2E+08 3.0E+06 1.2E+08 6.0E+07

Normalized Abundance Abundance Normalized 0.0E+00 0.0E+00 0.0E+00 Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C

R.PSLYWR.L K.FCSLPGSFR.E K.LAYAIAPEK.D

1.6E+08 5.0E+08 6.0E+07 4.0E+08 1.2E+08 4.5E+07 3.0E+08 8.0E+07 3.0E+07 2.0E+08 4.0E+07 1.5E+07 1.0E+08

Normalized Abundance Abundance Normalized 0.0E+00 0.0E+00 0.0E+00 Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C Figure 5.4 – Normalized DAT Peptide Levels. Tryptic DAT peptide abundances are depicted graphically across samples. The graphs are listed from most N-terminal (upper left) to most C-terminal (lower right).

79 FL ∆N ∆C Accession # Protein Description Avg Par Par Par 285.3 IPI00219765.4 SLC6A3 Sodium-dependent dopamine transporter 690.10 134.63 31.45 9 178.7 YFP YFP 399.96 104.90 31.42 6 ANKRD13A Ankyrin repeat domain-containing 149.2 IPI00217831.4 350.13 91.65 6.01 protein 13A 6 IPI00015148.3 RAP1B Ras-related protein Rap-1b 52.89 55.68 37.93 48.83 IPI00005737.1 SURF4 Isoform 1 of Surfeit locus protein 4 69.90 21.36 6.92 32.73 VKORC1L1 Vitamin K epoxide reductase IPI00166079.4 64.65 18.47 8.20 30.44 complex subunit 1-like protein 1 IPI00022143.3 FAM62A Isoform 1 of Extended synaptotagmin-1 56.49 23.25 7.04 28.93 3 beta-hydroxysteroid dehydrogenase/Delta 5-->4- Q9N119 30.35 6.75 14.33 17.14 isomerase OS UBC;UBB;RPS27A ubiquitin and ribosomal IPI00179330.6 35.23 12.57 2.05 16.62 protein S27a precursor A1XQS3 Mitochondrial NDUFA4 OS 28.24 13.78 5.02 15.68 KHDRBS1 Isoform 1 of KH domain-containing, IPI00008575.3 RNA-binding, signal transduction-associated 17.82 12.52 16.02 15.45 protein 1 PTPLAD1 Protein tyrosine phosphatase-like IPI00008998.3 28.32 13.04 4.61 15.33 protein PTPLAD1 IPI00010154.3 GDI1 Rab GDP dissociation inhibitor alpha 12.25 11.51 11.03 11.60 C7E3P9 Tripartite motif-containing protein 21 OS 12.00 12.41 10.18 11.53 SLC25A3 Isoform A of Phosphate carrier protein, IPI00022202.3 21.99 7.21 2.79 10.66 mitochondrial Q6VY03 Intercellular adhesion molecule-2 OS 11.65 12.24 5.65 9.85 Q95JG9 Carnitine palmitoyltransferase I OS 14.50 9.52 3.33 9.12 IPI00013271.1 DERL1 Derlin-1 15.39 7.70 3.64 8.91 IPI00219622.3 PSMA2 Proteasome subunit alpha type-2 9.08 8.52 8.71 8.77 C8C420 Solute carrier family 25 member 3 OS 15.53 7.47 2.73 8.58 C6EQ34 Endothelial protein C receptor OS 12.78 5.03 5.90 7.91 IPI00009236.5 CAV1 Isoform Alpha of Caveolin-1 15.67 6.25 1.27 7.73 SRGAP2 SLIT-ROBO Rho GTPase-activating IPI00479125.3 16.18 4.83 1.64 7.55 protein 2 IPI00003833.3 MTCH2 Mitochondrial carrier homolog 2 12.03 5.79 3.16 6.99 B0LY43 Monocarboxylic acid transporter 1 (Fragment) OS 15.09 3.90 1.01 6.67 Q2QLE2 Caveolin-2 OS 13.71 5.06 1.13 6.63 P00336 L-lactate dehydrogenase B chain OS 8.09 6.04 5.42 6.51 IPI00001159.10 GCN1L1 Translational activator GCN1 6.20 5.33 6.50 6.01 GNB2 Guanine nucleotide-binding protein IPI00003348.3 9.26 5.48 1.73 5.49 G(I)/G(S)/G(T) subunit beta-2 Immunglobulin heavy chain variable region Q6B6Z2 5.67 5.78 4.91 5.45 (Fragment) OS Dolichyl-diphosphooligosaccharide--protein Q9GL01 6.85 5.49 2.94 5.09 glycosyltransferase subunit 2 OS Q95242 Platelet endothelial cell adhesion molecule OS 6.96 5.54 2.58 5.03 Table 5.1 – Most abundant proteins across DAT IP samples. Protein accession numbers and descriptions are shown along with protein average DAT IP ratios to corresponding parental values. The final column shows the average of these values across samples.

80 Accession # Protein Description ∆N/FL IPI00024417.1 HIP1R Huntingtin-interacting protein 1-related protein 2.83 IPI00418471.6 VIM Vimentin 2.59 Q95N04 Dihydrolipoamide acetyltransferase OS 2.28 IPI00296099.6 THBS1 Thrombospondin-1 2.26 B7U3Y5 Lectin galactoside-binding soluble 9 protein OS 2.24 PDHA2 Pyruvate dehydrogenase E1 component subunit alpha, testis-specific IPI00024087.3 1.81 form, mitochondrial IPI00844578.1 DHX9 ATP-dependent RNA helicase A 1.80 IPI00288941.1 NCOA5 Nuclear receptor coactivator 5 1.52 KRT75 cDNA FLJ60809, highly similar to Homo sapiens cytokeratin type II IPI00005859.4 1.51 (K6HF), mRNA IPI00180956.6 Putative uncharacterized protein ENSP00000382160 1.40

Accession # Protein Description FL/∆N IPI00219765.4 SLC6A3 Sodium-dependent dopamine transporter 5.13 Q9N119 3 beta-hydroxysteroid dehydrogenase/Delta 5-->4-isomerase OS 4.50 B0LY43 Monocarboxylic acid transporter 1 (Fragment) OS 3.87 IPI00217831.4 ANKRD13A Ankyrin repeat domain-containing protein 13A 3.82 YFP YFP 3.81 IPI00166079.4 VKORC1L1 Vitamin K epoxide reductase complex subunit 1-like protein 1 3.50 IPI00479125.3 SRGAP2 SLIT-ROBO Rho GTPase-activating protein 2 3.35 IPI00005737.1 SURF4 Isoform 1 of Surfeit locus protein 4 3.27 IPI00022202.3 SLC25A3 Isoform A of Phosphate carrier protein, mitochondrial 3.05 IPI00291467.7 SLC25A6 ADP/ATP translocase 3 2.94 Table 5.2 – Proteins with most changed abundances between FL and ΔN IPs. The top ten most increased and decreased proteins are shown along with the average fold change values for the protein as a whole.

After analysis of DAT peptide abundances, further investigation of associated protein abundances was conducted through t-test analyses and ratio comparisons. First,

DAT IP sample Topograph peptide abundances were compared to Par values for the same peptides via t-test. From the p-values, q-values were calculated using the R program

TkQVALUE145. Peptides with significant q-values (p<0.05) were used for continued analyses. DAT IP peptide abundances were further compared to Par values through ratios. These ratios were then used to calculate the most abundant proteins across the three sample types by averaging peptide abundance for all peptides within a protein.

81 Proteins that had only one peptide identification were excluded from further analyses.

Table 5.1 lists the proteins with at least five-fold average increase across all samples to show the most abundant putative interactors.

Additional protein level analyses were conducted in order to identify DAT interacting protein variation between DAT samples. To do this, peptide ratios of peptides previously found to be significantly different from the parental IP were calculated between truncation mutant samples and the FL sample. Comparisons with significant t- tests (p<0.05) were further analyzed. Ratios were again averaged per protein, ignoring any proteins that had only one peptide identification. The ten most increased and decreased proteins in ΔN and ΔC IPs as compared to FL can be seen in Tables 5.2 and

5.3, respectively.

Protein abundance levels were examined with Ingenuity Pathway Analysis (IPA) software. When the average protein levels across samples were analyzed, it was found that upregulated networks containing DAT were largely involved in connective tissue disorders, cellular assembly and organization, and molecular transport. Average protein ratios comparing FL to ΔN IPs identified changes in networks involved in carbohydrate metabolism, neurological disease, and cellular assembly and organization. Whereas, ΔC comparisons revealed network changes involved in cellular assembly and organization, lipid metabolism, psychological disorders. The common factor in these network change comparisons is cellular assembly and organization.

82 Accession # Protein Description ∆C/FL Q95N04 Dihydrolipoamide acetyltransferase OS 2.36 PDHA2 Pyruvate dehydrogenase E1 component subunit alpha, testis-specific IPI00024087.3 2.19 form, mitochondrial IPI00020984.2 CANX cDNA FLJ55574, highly similar to Calnexin 1.78 IPI00033494.3 MYL12B Myosin regulatory light chain MRLC2 1.65 IPI00413922.7 MYL6B 11 kDa protein 1.50 IPI00027834.3 HNRNPL Heterogeneous nuclear ribonucleoprotein L 1.50 IPI00018140.3 SYNCRIP Isoform 1 of Heterogeneous nuclear ribonucleoprotein Q 1.41 IPI00171903.2 HNRNPM Isoform 1 of Heterogeneous nuclear ribonucleoprotein M 1.30 IPI00033907.1 ANAPC1 Anaphase-promoting complex subunit 1 1.24 IPI00216348.1 DYNC1I2 Isoform 2C of Cytoplasmic dynein 1 intermediate chain 2 1.17

Accession # Protein Description FL/∆C IPI00219765.4 SLC6A3 Sodium-dependent dopamine transporter 21.94 IPI00179330.6 UBC;UBB;RPS27A ubiquitin and ribosomal protein S27a precursor 17.17 YFP YFP 12.73 IPI00009236.5 CAV1 Isoform Alpha of Caveolin-1 12.38 IPI00005737.1 SURF4 Isoform 1 of Surfeit locus protein 4 10.11 IPI00479125.3 SRGAP2 SLIT-ROBO Rho GTPase-activating protein 2 9.87 IPI00022143.3 FAM62A Isoform 1 of Extended synaptotagmin-1 8.02 IPI00166079.4 VKORC1L1 Vitamin K epoxide reductase complex subunit 1-like protein 1 7.89 IPI00022202.3 SLC25A3 Isoform A of Phosphate carrier protein, mitochondrial 7.89 IPI00008998.3 PTPLAD1 Protein tyrosine phosphatase-like protein PTPLAD1 6.14 Table 5.3 – Proteins with most changed abundances between FL and ΔC IPs. The top ten most increased and decreased proteins are shown along with the average fold change values for the protein as a whole.

Protein level analysis is useful in identifying the most abundant proteins and associated abundance changes within a system, especially when multiple peptides are included per protein. However, more subtle changes at the peptide level are likely to go unnoticed when only focusing on the protein as a whole. Thus, it is also necessary to look at the abundances of the individual peptides of these candidate proteins. Two examples of this are shown in Figure 5.5. One of the most increased proteins in the ΔN IP sample is vimentin, an intermediate filament protein. This protein increased in the ΔN sample at both the protein level and at the peptide level. All five peptides associated with this

83 protein display the same relative pattern across peptides, with the ΔN sample being most abundant and significantly higher than the other sample abundances (Figure 5.5a). This increase could be related to the increased intracellular location of DAT in the ΔN cells.

The second example shown is calnexin, an ER chaperone protein. This protein was one of the most increased proteins overall in the ΔC IP sample, when compared with the FL. Unlike vimentin, this protein has a more variable abundance pattern between peptides. Abundance in the ΔC sample is significantly increased (p<0.05) over FL for four of the five tryptic peptides shown in Figure 5.5b (not significant in

K.TPYTIMFGPDK.C). However, the ΔN shows a significant increase over FL in one peptide (K.TPELNLDQFHDK.T) but a significant decrease in two peptides

(K.APVPTGEVYFADSFDR.G and K.TPYTIMFGPDK.C). Such discrepancies could be the result of post-translational modifications, the presence of which would interfere with

MS identification of the peptide. The general increase in calnexin in the ΔC IP sample makes sense as the ΔC DAT variant remains mostly in the ER. This is thought to be due to the loss of its C-terminus hindering further protein processing. Thus, increased interaction with an ER chaperone, which also appears to interact with properly processed

DAT, correlates with the expected trafficking patterns of the construct. This correlation could help to further explain the stunted trafficking of this mutant construct.

Summary

The main focus of this chapter was to implement the immunoglobulin normalization technique onto a large dataset. This was conducted upon IPs directed

84 toward three different DAT constructs stably expressed in PAE cells in order to identify both consistent and changing members of the DAT interactome in this expression system.

Application of the normalization method was based on the total sum of IgG and

IgK constant domain peptides. This allowed for normalization factors that were less dependent upon individual peptide abundance differences and were thus a better representation of the immunoglobulin levels as a whole. As DAT levels are significantly different between samples, this normalization introduces an additional level of assurance in the dataset and inter-sample comparisons. a) R.TYSLGSALRPSTSR.S K.ILLAELEQLK.G R.KVESLQEEIAFLKK.L R.QVQSLTCEVDALK.G R.ISLPLPNFSSLNLR.E

3500000 2500000 1600000 1200000 3000000 3000000 1400000 2000000 1000000 2500000 2500000 1200000 800000 2000000 1000000 2000000 1500000 800000 600000 1500000 1500000 1000000 600000 400000 1000000 1000000 400000 500000 200000 500000 Normalized Abundance Abundance Normalized 500000 200000 0 0 0 0 0 Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C b) K.APVPTGEVYFADSFDR.GK.APVPTGEVYFAD R.GTLSGWILSK.A K.TPELNLDQFHDK.TPELNLDQFHDK.T K.TPYTIMFGPDKK.TPYTIMFGPDK.C K.LHFIFR.H

7000000 10000000 20000000 7000000 4500000 9000000 18000000 4000000 6000000 6000000 8000000 16000000 3500000 5000000 7000000 14000000 5000000 3000000 6000000 4000000 12000000 4000000 2500000 5000000 10000000 2000000 3000000 4000000 8000000 3000000 1500000 2000000 3000000 6000000 2000000 2000000 4000000 1000000 1000000 1000000 Normalized Abundance Abundance Normalized 1000000 2000000 500000 0 0 0 0 0 Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C Par FL ∆N ∆C Figure 5.5 – Normalized peptide abundance levels of putative DAT interactome members. a) Vimentin peptide abundance levels are consistent across both samples and peptides. b) Calnexin abundance levels are more variable between peptides and samples.

After normalization, ratios between DAT IP samples and the parental control IPs were calculated in order to identify putative DAT interactome members. Peptide ratios were averaged for individual proteins to provide a general identifier of protein abundance. Table 5.1 shows the most abundant proteins identified in this manner. Further

85 analysis using IPA showed that these consistent interactome members are largely involved in connective tissue disorders, cellular assembly and organization, and molecular transport. Additionally, protein levels were compared between truncation mutants and the FL sample. This identified several proteins that are either increased or decreased in ΔN and ΔC samples relative to FL levels. However, because DAT levels are also decreased in ΔN and ΔC samples when compared to FL, these decreased protein levels may not be indicative of a relative change as compared to DAT. Yet, these proteins still appear to be reliably present as they tend to be significantly altered from parental values in all samples. Furthermore, for similar comparative reasons, the proteins that are increased in the truncation mutant IPs when compared to FL appear to be even more significantly altered. Specifically, having an increased amount of interacting protein with less DAT as bait indicates an even larger increase in the protein at hand. These increased proteins are also pertinent to the given biological pathways and associated changes specific to the mutant constructs.

This chapter showed the successful implementation of the immunoglobulin normalization pipeline onto a large DDA dataset based on relative label-free differential analysis. Further experiments described in Chapter VI will continue to develop this normalization technique with both additional DDA data as well as targeted SRM data.

86 CHAPTER VI

ANALYSIS OF THE DAT INTERACTOME IN MICE WITH AMPHETAMINE

AND PMA TREATMENTS

Introduction

In the previous chapter, the DAT interactome was explored using multiple DAT constructs in PAE cells through IP-MS/MS with the aid of the IgG normalization method.

This chapter continues experimentation with the application of the normalization method.

However, in this chapter additional sample types will be analyzed using multiple analytical MS techniques. Particularly, drug treatments are implemented with synaptosomal preparations from an HA-DAT mouse model. These samples are then analyzed using both DDA and SRM analyses.

Phorbol esters, such as phorbol 12-myristate 13-acetate (PMA), activate PKC.

This results in rapid clathrin-mediated endocytosis of DAT followed by degradation of the internalized transporters thus increasing extracellular DA levels66. Amphetamine acts through a similar but distinct pathway. Amphetamine is similar enough to DA structurally that it acts as a substrate and is transported into the cell by DAT. This initiates downstream signaling events resulting in an increase in intracellular calcium levels, which activates both PKC and calcium/calmodulin-dependent protein kinase II

(CaMKII), and also inhibits Akt kinase activity54,55. This process has been found to involve several proteins, including syntaxin 1A146 and flotillin-124. Through currently unknown mechanisms, this results in both an increase in intracellular DAT levels, due to either an increase in internalization, a decrease in recycling, or a combination of the two,

87 and a subsequent increase in extracellular DA levels. Additionally, intracellular amphetamine signaling results in a DA efflux along with a secondary increase in trafficking of DAT to the plasma membrane, which when combined increase extracellular

DA levels even further58,59,76. Furthermore, levels of surface DAT have been found to rapidly increase with amphetamine treatments through unknown pathways59. Uncovering the proteins involved in these complex pathways could help to reveal more specific mechanistic details and lead to a more comprehensive understanding of the system as a whole.

An epitope-tagged DAT knock-in mouse model has been developed wherein a hemagluttinin (HA) tag was inserted into the second extracellular loop of DAT147. In this model, HA-DAT expression levels, distribution patterns, and DA uptake kinetics were shown to not be significantly altered from those of wild-type (WT) DAT. Combining this mouse model with an IP-MS pipeline allows for the analysis of the DAT interactome in vivo with a negative control using WT mice. Additionally, this strategy allows for the implementation of additional drug treatments with condition specific WT controls.

For this proteomics pipeline, profiling DDA studies were used to identify potential DAT interactome members from drug-treated synaptosomal IPs. As traditional

DDA methods provide less reliable quantitation, due to their semi-random sampling, a more quantitative method for peptide and protein abundance analyses is necessary.

Identified candidates, and the changes in their levels between samples, can then be relatively quantified through targeted selected reaction monitoring (SRM). SRM analyses are highly selective and sensitive with a broad dynamic range allowing for reliable quantitation.

88 For both DDA and SRM IP-MS data, IgG normalization was applied in order to account for any intersample variability. In both types of experiment, it is important that this normalization is a result of multiple antibody peptides to better represent overall levels of the IgG. In the same vein, multiple peptides should also be quantitated for DAT and any interactome members as well. This will help elucidate changes in abundance at the peptide level more reliably. Due to peptide level alterations such as modifications and domain effects, individual peptide levels are not always indicative of protein levels as a whole, thus multiple peptides should be used for quantification of any protein.

Unfortunately, this is not always the case in proteomics studies, and single peptides are often misinterpreted as representative for the entire protein.

This chapter investigates the use of IgG-based IP-MS normalization with several applications using both WT and HA-DAT mice. Firstly, the DAT interactome in these mice is explored through DDA analysis of IPs from striatal dissections. The interactome is then further examined through DDA profiling of striatal synaptosome IPs with and without the addition of amphetamine or PMA. The combination of these two profiling studies culminated in a list of potential DAT interactome candidates for further targeting through SRM for relative quantitation with IgG normalization. This was performed over treatment time courses with both of the aforementioned drugs in striatal synaptosomes in order to identify changes in the interactome as a result of the drug treatments. This analytical pipeline takes advantage of the IgG normalization method for both identification and quantitation purposes.

89 Experimental Procedures

Reagents and Mice - Solvents were purchased from Fisher Scientific (Rockford,

Il); other chemicals were from Fisher Scientific, JT Baker (Center Valley, PA), or Sigma

(St. Louis, MO) unless noted. Mouse-anti-HA.11 monoclonal antibody was purchased from Covance (Princeton, NJ) MMS-101P clone 16B12.

HA-DAT knock-in mice were previously described 147. WT and HA-DAT mice were bred and housed in accordance with the U.S. National Institutes of Health (NIH) guidelines (NIH Publication No. 80-23, revised 1996) as approved by the Institutional

Animal Care and Use Committee at the University of Pittsburgh.

Striatal IPs – Striata were dissected from 3 WT and 3 HA-DAT mice and homogenized individually in 500 µL lysis buffer (20 mM Tris-HCl, pH 7.4, 137 mM

NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA) with Halt Protease and

Phosphatase Inhibitor Cocktail (Thermo Scientific, Rockford, IL). Samples were nutated at 4°C for 10 minutes to ensure lysis and centrifuged for 15 minutes at 4°C and 100,000 x g. Supernatants were added to 200 µL protein G conjugated sepharose bead slurry (50%, equilibrated three times in lysis buffer) and nutated for 2 hours at 4°C to pre-clear.

Samples were centrifuged 10 minutes at 4°C and 1,000 x g to pellet beads. Protein concentrations of supernatants were determined using the DC protein assay kit (Bio-Rad,

Hercules, CA). 700 µg of each sample was diluted to 1.4 µg/µL final concentration before addition of HA.11 antibody at a 1:30 antibody:protein ratio and overnight nutation at

4°C. Negative control bead only samples were incubated overnight without antibody.

Equilibrated protein G conjugated sepharose beads (200 µL 50% slurry) were added, and samples were nutated at 4°C for 4 hours. Beads were rinsed three times with 500µL lysis

90 buffer and frozen at -80°C until elution. Beads were thawed on ice and 200 µL Laemmli buffer without bromophenol blue was added to samples after the rinses, samples were nutated at room temperature for 10 minutes, then boiled at 100°C for 5 minutes. Both types of samples were then centrifuged 15 minutes at 4°C and 20,817 x g and the supernatants were transferred to clean eppindorf tubes.

Synaptosomal Drug Treatment IPs – Striata were dissected from 6 WT and 6 HA-

DAT mice and homogenized with a dounce homogenizer in 0.32M sucrose buffer (5 mM

HEPES, pH 7.4). Homogenates were pooled for each type of mouse and centrifuged for

10 minutes at 1000 x g and 4°C. Supernatants were split into thirds and centrifuged again for 20 minutes at 12,500 x g to prepare synaptosomes as previously described26,148.

Synaptosomal pellets were resuspened in Krebs Ringer-HEPES (KRH) buffer (120 mM

NaCl, 4.7 mM KCl, 2.2 mM CaCl2, 1.2 mM MgSO4, 1.2 mM KH2PO4, 20 mM HEPES,

0.1% bovine serum albumin (BSA), 0.18% glucose). PMA was added to 1 µM, amphetamine was added to 20 µM, or an equivalent volume of DMSO was added to each sample and incubated at 37°C for 30 minutes. Samples were centrifuged at 12,500 x g and 4°C for 20 minutes, and pellets were resuspended in triton-glycine HEPES (TGH) lysis buffer (1% Triton-X 100, 10% glycerol, 25 mM HEPES, 1 mM EDTA, 1 mM

EGTA, 1 mM NaOV, 1 mM NaF, 1 mM PMSF) and nutated for 20 minutes at 4°C before a 10 minute centrifugation at 4°C and 20,817 x g. The supernatants were then pre-cleared and IPed as described above.

Synaptosomal Drug Treatment Time-course IPs – Synaptosomes were prepared similarly for the drug treatment time-course samples as in the initial drug treatment synaptosomal IPs. However, pooled samples were split for six different conditions.

91 Synaptosomes in KRH buffer were treated with equivalent volumes of DMSO for 15 minutes, 1 µM PMA for 5 or 15 minutes, water for 30 minutes, or 20 µM amphetamine for 15 or 30 minutes. These times and concentrations were established based on the literature 26,148,149.

MS Sample Preparation – For all three sets of IPs, 50µg of bovine serum albumin

(BSA) was added as a carrier protein before further processing steps. Samples were precipitated with methanol and chloroform as previously described87. Protein pellets were resuspended in 0.2% RapiGest (Waters Corporation, Milford, MA) in 50 mM ammonium bicarbonate, sonicated using a Microson Ultrasonic cell disruptor (Misonix, Farmingdale,

NY; 6 x 1 sec pulses, power 1.5), and boiled for 5 minutes at 100°C. Samples were diluted with 50 mM ammonium bicarbonate to 0.1% RapiGest before being sonicated again. Protein concentrations were determined using the DC protein assay kit (Bio-Rad,

Hercules, CA). Total protein of 30 µg was aliquoted and diluted to 0.5 mg/mL.

Dithiothreitol (DTT) was added to 5 mM final concentration and samples were incubated at 60°C for 30 minutes to reduce disulfide bonds. Iodoacetamide (IAA) was added to 15 mM final concentration and samples were incubated for 30 minutes in the dark at room temperature in order to alkylate reduced cysteines. CaCl2 was added to a final concentration of 1 mM before modified trypsin (Promega, Madison, WI) was added to a

1:50 enzyme:substrate ratio. Samples were incubated overnight at 37°C using a

Thermomixer (Eppendorf, Westbury, NY). HCl was added to a final concentration of 200 mM and incubated at 37°C for 45 minutes in a Thermomixer to hydrolyze the RapiGest.

Samples were centrifuged at 20,817 x g at 4°C for 30 minutes and supernatant was collected twice. Supernatants were cleaned using Pierce® C18 Spin Columns (8mg,

92 Thermo Scientific, Rockford, IL), condensed using a CentriVap Centrifugal Concentrator

Model 7810014 (Labconco, Kansas City, MO), and resuspended in 0.1% formic acid to a

0.25 or 0.5 mg/mL final concentration.

NanoLC-Mass Spectromtery – Tryptic striatal and synaptosomal drug treatment

IP digests were analyzed using a Proxeon EASY-nLC II and an Orbitrap Elite mass spectrometer with Xcaliber software. 2 µg of each sample were loaded onto a fused silica

(Polymicro Tech, Phoenix, AZ) microcapillary column (75 µm i.d., 360 µm o.d. with a 5

µm laser pulled tip) packed with 30 cm of 5 µm 125 Å Aqua C18 reversed-phase (RP) material (Phenomenex, Torrance, CA) maintained at 40°C, using a column heater built in-house as previously described93. The mobile phase buffers used were Buffer A (98% water, 2% acetonitrile, 0.1% formic acid) and Buffer B (100% acetonitrile, 0.1% formic acid). Samples were analyzed with a previously optimized 70-minute gradient at 250 nL/min that started at 100% Buffer A ramped up to 5% Buffer B by 5 minutes and reached 30% Buffer B by 50 minutes with an additional 2 minute increase to 80% Buffer

B and maintained for 5 minutes before a 2 minute decrease back to 100% Buffer A that was maintained for the final 11 minutes. A 15-minute wash and a 50-minute 100 fmol bovine standard (6 Bovine Protein Digest Equal Molar Mix, Michrom Bioresources,

Auburn, CA) gradient were run in between each of the four randomized replicate sets of samples. Data-dependent acquisition was used to collect tandem mass spectra at a resolution of 60,000. One FTMS profile full scan was followed by 10 MS/MS centroid ion trap scans using dynamic exclusion.

Selected reaction monitoring (SRM) methods were created using Skyline103 for the analysis of peptides representing DAT and putative binding partners identified in the

93 MS/MS studies. These methods were refined through preliminary SRM studies on additional striatal, synaptosomal, and drug treated synaptosomal IPs along with empirical refinement150. Data were acquired with a Q1 and Q3 resolution of 0.7 m/z. Transitions were monitored with a total cycle time of 1 s. The RF-only q2 collision cell was pressurized with 1.5 mTorr of argon gas. Monitored transitions and associated collision energies can be found in Appendix C.

SRM analyses were conducted with synaptosomal drug treatment time-course IP digests using a Proxeon EASY-nLC II and a TSQ Vantage Triple Stage Quadrupole mass spectrometer with Xcaliber software. The LC system was used to load 1 µg digest onto a microcapillary column, as described above. Temperature was maintained at 40°C.

Samples were analyzed using a 75 minute gradient at 250 nL/min that started at 100%

Buffer A ramped up to 5% Buffer B by 5 minutes and reached 30% Buffer B by 50 minutes with an additional 2 minute increase to 87% Buffer B that was maintained for 5 minutes before a 2 minute decrease back to 100% Buffer A that was maintained for the final 16 minutes. Four technical replicate injections of all 12 samples were conducted in randomized order, and a 60-minute 100 fmol bovine standard gradient was run after every six injections for quality control purposes.

The upper and lower limits of detection (LOD) and limits of quantitation (LOQ) were measured for two representative synthetic peptides from each protein. Zero mol, 10 amol, 100 amol, 1 fmol, 10 fmol, 100 fmol, and 1 pmol of each peptide were coinjected separately with 1 µg digested striatal homogenate, 100 fmol bovine standard, and a mix of 50 fmol bovine standard with 0.5 µg digested αHA antibody. These peptides were analyzed via SRM using the methods mentioned above. Results were analyzed with

94 Skyline. Additional confirmatory studies were conducted by coinjecting 100 fmol of synthetic peptide with each of the drug treated synaptosomal IP samples in order to verify peaks using the same SRM method as above.

MS Data Processing – Profiling results were deconvoluted using Hardklor114 and

Bullseye115, searched using SEQUEST110, post-processed using Percolator113. Peptide identifications were aligned and MS1 XIC peak areas were calculated using

Topograph119,120. Peak areas were normalized within each dataset relative to the sum of the total peptide MS1 XIC areas across all injections. SRM data were analyzed using

Skyline103 to measure and calculate peptide area under the curve (AUC), coefficient of variance (CV), and retention time (RT). For both profiling and SRM results, all peptide areas were normalized to immunoglobulin levels based on the sum of 6 constant domain peptides described previously151 (Appendix D). Student’s t-tests were used to compare between HA-DAT and WT samples, and ANOVAs were used to compare intersample variability. For both statistical tests, a significance cut-off of p < 0.05 was implemented.

Furthermore, p-values from DDA data were used to calculate q-values using the R program TkQVALUE145. Ingenuity Pathway Analysis (IPA, Ingenuity®Systems, www.ingenuity.com) was used to aid in data interpretation.

Results and Discussion

This study used an HA-tagged knock-in mouse model for IP-MS analysis of the

DAT interactome in both striatal homogenate and striatal synaptosome samples. Mouse striata were homogenized in sucrose prior to synaptosomal preparation through centrifugation. Synaptosomes were treated with either PMA or amphetamine to analyze

95 changes in the interactome with drug treatments, or with DMSO or water as a negative control, respectively. IPs were conducted on synaptosomes as well as untreated striatal homogenates with an antibody specific for the HA-tag, such that WT mice could be used as additional negative controls. These samples were initially analyzed through DDA analysis. Peptide and protein targets were selected from these data for further targeted

SRM analysis on a triple quadrupole mass spectrometer using synaptosomal samples that had been subjected to a time-course of drug treatments. IP data were then normalized to antibody levels using IgG peptides to account for intersample variability.

In order to determine the basic components of the DAT interactome in this unique mouse model, tryptic digests of striatal homogenate IPs from both the WT and HA-DAT mice were profiled using a top ten DDA analysis on a Thermo ScientificTM Orbitrap Elite with four randomized technical replicates and three biological replicates.

The resulting data were then deconvoluted using Hardklör114 and Bullseye115 before being searched and post-processed with Sequest110 and Percolator113. Label-free differential analysis was performed using Topograph119,120, which allowed for chromatographic peak area based comparisons between samples, including that of low abundance peptides that may not have triggered an MS/MS event in all replicates and samples. Peptide abundances were normalized based on the sum of the six mouse IgG peptides analyzed in Chapter IV. Although a normalization based on more peptides may be optimal in DDA experiments, these six peptides were used for consistency with the future SRM experiments, as the addition of too many peptides into a single SRM method is impractical, due to scan speed and scheduling constraints. T-tests were conducted between HA-DAT and WT sample abundances, and q-values were calculated from the

96 HA:WT Ratio Protein Peptide Striatum DMSO PMA Amph EVELILVK 9.80 10.30 12.82 FCSLPGSFR 85.43 14.52 10.02 6.95 DAT – Dopamine LAYAITPEK 488.14 10.14 8.15 8.21 Transporter AYLSVDFYR 288.23 13.63 10.45 9.02 EGAAGVWK 35.39 6.27 14.43 22.65 CSVGPMSSVVAPAK 10.52 3.83 0.52 2.96 Atp1b1 – Na+/K+ - YNPNVLPVQCTGK 2.58 2.73 3.65 1.86 transporting ATPase VAPPGLTQIPQIQK 1.77 2.48 2.72 2.33 Camk2g – CaM FYFENLLSK 2.39 2.06 1.63 kinase II gamma GAILTTMLVSR 1.90 1.87 1.59 EVCFACVDGK 3.73 2.03 1.51 Cltc – Clathrin heavy chain LLLPWLEAR 2.51 2.99 1.78 VSQPIEGHAASFAQFK 1.32 2.41 2.42 DNFTLIPEGTNGTEER 1.60 2.39 2.08 Dpysl2 – GIQEEMEALVK 1.52 2.15 2.73 1.96 Dihydropyrimidinase -related protein 2 IVLEDGTLHVTEGSGR 2.64 5.21 1.36 ISVGSDADLVIWDPDSVK 1.85 4.73 1.87 YYLDSLDR 2.46 2.74 1.36 Gnao1 – Guanine LWGDSGIQECFNR 4.65 2.23 1.36 nucleotide binding

protein alpha MEDTEPFSAELLSAMMR 2.04 2.83 3.06 1.68 IGAGDYQPTEQDILR 1.87 2.33 0.85 1.55 LDAQVQELHAK 1.96 1.62 0.94 NSLQEQQEEEEEAR 7.57 1.14 2.55 1.12 Myh10 – Myosin X ADMEDLMSSK 1.25 0.65 1.97 LEVNMQAMK 2.47 2.95 3.44 AEEDSMEDPYELK 1.24 1.07 0.97 ISLPETGLAPTPSSQTK 1.93 2.47 1.49 Pclo - Piccolo ALGGELAAIPSSPQPTPK 1.79 2.36 1.76 FAEELEWER 1.54 2.07 1.10 DNALLAQLIQDK 1.38 2.46 2.14 1.62 Stxbp1 – Syntaxin HYQGTVDK 0.76 2.67 1.63 binding protein 1 VLVVDQLSMR 1.40 1.81 2.52 1.43 EPLPSLEAVYLITPSEK 2.41 2.45 1.24 Syngr1 – AGGAFDPYTLVR 2.88 3.59 2.76 2.69 Synaptogyrin 1 DNPLNEGTDAAR 6.49 1.63 2.11 Table 6.1 – Interactome Candidates for SRM Analyses. Peptides and proteins are listed with their normalized HA-DAT to WT ratios from discovery experiments. These peptides were further analyzed via SRM. Green highlighting indicates significant differences (p < 0.05).

97 resultant p-values using the R program TkQVALUE145. Peptides with significant q- values (q < 0.05) were used for further comparative analysis. There were 162 peptides, representing 64 proteins, which were identified to be differentially abundant between the two samples. Based on analysis using IPA, these identifications indicated involvement in cancer, molecular transport, and neurological disease networks. These networks are highly critical and show that this complex DAT interactome is both diverse and necessary for proper functionality.

The general components of the DAT interactome with and without drug treatments were also examined. WT and HA-DAT mouse striatal synaptosomes were treated with PMA, amphetamine, or DMSO prior to IP and tryptic digestion, and samples underwent the same MS/MS pipeline as the aforementioned striatal homogenates. This resulted in over 1000 proteins that had significant changes between WT and HA-DAT samples with greater than 2-fold change in both DMSO and PMA treated samples and over 400 proteins in amphetamine samples. This increase in total identifications, when compared to the striatal homogenate samples, could likely be a result of the multiple enrichment strategies involved along with a probable increase in false positives. It was thus necessary to investigate some of these candidates further in order to verify their associations with DAT and the DAT interactome.

Based on the results of these two discovery datasets, a list of proteins of interest was selected for further SRM analyses. This selection was based upon several factors.

First, proteins were selected that had robust signals from multiple peptides. Second, proteins of interest were selected that are known to interact with DAT, are involved in trafficking or signaling, or are specifically neuronal or synaptosomal. The identified

98 a) IgG Heavy Chain: VNSAAFPAPIEK

1.E+08 Area

1.E+06

1.E+04

1.E+02 1.E+00 1.E+02 1.E+04 1.E+06 Amount of Standard (amol) b) DAT: DAT: AYLSVDFYR LAYAITPEK

1.E+08 1.E+08 1.E+06 1.E+06 1.E+04 1.E+04 Area Area 1.E+02 1.E+02 1.E+00 1.E+00 1.E+00 1.E+02 1.E+04 1.E+06 1.E+00 1.E+02 1.E+04 1.E+06 Amount of Standard (amol) Amount of Standard (amol) c) Stxbp1: Stxbp1: DNALLAQLIQDK EPLPSLEAVYLITPSEK

1.E+08 1.E+08 1.E+06 1.E+06 1.E+04 1.E+04 Area Area 1.E+02 1.E+02 1.E+00 1.E+00 1.E+00 1.E+02 1.E+04 1.E+06 1.E+00 1.E+02 1.E+04 1.E+06 Amount of Standard (amol) Amount of Standard (amol)

Figure 6.1 – Peptide Levels of Quantitation. Increasing amounts of synthetic peptides representative of a) IgGs, b) DAT, c) and putative interactome members were coinjected with striatum, bovine, and mixed bovine/antibody matrices. All axis are based on log scales.

99 peptides from these proteins underwent preliminary SRM analyses in untreated synaptosomal IPs to further constrain the list for analysis within a single SRM method

(Table 6.1). Once this refined list was compiled, LOQ studies were conducted using synthetic peptides for at least two peptides per protein of interest (Figure 6.1). These studies were performed in three different background matrices: digested striatal homogenate, bovine standard, and a mix of bovine standard with digested αHA antibody.

Synthetic peptides were generally detectable in all matrices in the 10 fmol to 1 pmol range, while some peptides were detectable down to 100 amol. a) DSTYSMSSTLTLTK No TSTSPIVK VNSAAFPAPIEK Sum Normalization Normalization Normalization Normalization

4.0E+07

3.0E+07

2.0E+07

Abundance 1.0E+07

0.0E+00 HA1 HA2 HA3 HA4 HA6 HA1 HA2 HA3 HA4 HA6 HA1 HA2 HA3 HA4 HA6 HA5 HA5 HA5 HA1 HA2 HA3 HA4 HA6 WT1 WT2 WT3 WT4 WT6 WT1 WT2 WT3 WT4 WT6 WT1 WT2 WT3 WT4 WT6 HA5 WT5 WT5 WT5 WT1 WT2 WT3 WT4 WT6 WT5 p = 0.056 p = 0.0013 p = 0.0086 p = 0.66 b) APQVYTIPPPK! No TSTSPIVK VNSAAFPAPIEK Sum Normalization Normalization Normalization Normalization

8.0E+07

6.0E+07

4.0E+07

Abundance 2.0E+07

0.0E+00 HA1 HA2 HA3 HA4 HA6 HA1 HA2 HA3 HA4 HA6 HA1 HA2 HA3 HA4 HA6 HA1 HA2 HA3 HA4 HA6 HA5 HA5 HA5 HA5 WT1 WT2 WT3 WT4 WT6 WT1 WT2 WT3 WT4 WT6 WT1 WT2 WT3 WT4 WT6 WT1 WT2 WT3 WT4 WT6 WT5 WT5 WT5 WT5 p = 0.031 p = 0.0031 p = 0.093 p = 0.24

Figure 6.2 – IgG normalization of IgG peptides. Three normalization methods were implemented on IgG peptides. This is shown here with a) DSTYSMSSTLTLTK from the heavy chain and b) APQVYTIPPPK from the light chain. Two of these methods used single peptides while the third used the sum of all six monitored IgG peptides. ANOVA p values are listed underneath individual graphs. Numbers 1-6 indicate sample treatment where 1-3 represent the points on the PMA time course and 4-6 represent those on the Amph time course.

100 Time-course drug treatments were performed on WT and HA-DAT synaptosomes with PMA and amphetamine including a zero time-point using DMSO or water, respectively. Additionally, the PMA time-course included 5- and 15-minute treatments, while the amphetamine time-course consisted of 15- and 30-minute treatments. These treatment times were determined through the literature to be either partially or fully effective for each of the individual drugs26,148,149. These samples underwent SRM analysis targeting peptides from DAT, mouse IgG/K, and the aforementioned putative DAT interactome members. Peptide areas under the curve (AUCs) were calculated using

Skyline103 and then normalized to IgG levels in order to account for intersample variability. Additionally, synthetic peptides were coinjected with samples in separate experiments in order to verify peptide peaks.

Three different methods for IgG normalization were taken into consideration, based on the 6 previously described mouse IgG peptides, all of which were monitored in the

SRM method151. The six analyzed IgG peptides analyzed included three peptides from the constant region of the heavy chain and three from the constant region of the light chain. With three technical replicates, average CVs across all 12 samples for these IgG peptides ranged from 10.0% to 28.8%, indicating that some of these peptides are being more stably monitored and are thus more suitable on their own for normalization than others due to this signal stability. The two peptides with the lowest CVs

(VNSAAFPAPIEK and TSTSPIVK) were chosen for the individual peptide normalization approach, because they had the most stable signals. Additionally, a sum normalization approach was conducted which used normalization factors based on the sum of all six monitored IgG peptides. In all normalization methods, relative

101 normalization factors were created based on the most abundant sample, and all abundance values were then multiplied by this sample specific factor. These normalization methods were evaluated in IgG peptides (Figure 6.2) and in DAT peptides

(Figure 6.3) using ANOVAs to determine intersample variations (p < 0.05). As IgG levels should be equivalent across samples, normalization of IgG peptides should result in less variation between samples based on this metric. a) AYLSVDFYR No TSTSPIVK VNSAAFPAPIEK Sum Normalization Normalization Normalization Normalization

2.2E+06

1.7E+06

1.1E+06

5.5E+05 Abundance

0.0E+00 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 p = 0.60 p = 0.034 p = 0.59 p = 0.74 b) LAYAITPEK No TSTSPIVK VNSAAFPAPIEK Sum Normalization Normalization Normalization Normalization

3.2E+06

2.4E+06

1.6E+06

Abundance 8.0E+05

0.0E+00 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 p = 0.023 p = 1.9E-5 p = 0.19 p = 0.056

Figure 6.3 – IgG normalization of DAT peptides. Three normalization methods were implemented on DAT peptides. This is shown here with a) AYLSVDFYR and b) LAYAITPEK. Two of these methods used single peptides while the third used the sum of all six monitored IgG peptides. ANOVA p-values are listed underneath individual graphs. Numbers 1-6 indicate sample treatment where 1-3 represent the points on the PMA time course and 4-6 represent those on the Amph time course.

102 ANOVAs of IgG peptides pre- and post-normalization showed that normalization lowered variability with both the peptide with the lowest CV value (VNSAAFPAPIEK) and the sum normalization. For both of these methods before normalization, four of the six peptides have ANOVA p-values less than 0.05, indicating variation, whereas after normalization with this peptide, only two peptides were still at varying levels across samples. Contrarily, when normalization was conducted using TSTSPIVK, five of the six peptides had variable levels between samples (Figure 6.2). When DAT peptides were normalized using these three methods, again both VNSAAFPAPIEK and sum normalization decreased intersample variability in HA-DAT samples, while TSTSPIVK normalization actually increased intersample variability. The VNSAAFPAPIEK and sum normalization techniques had very similar results. In further analyses, only the sum normalization method is used; it is based on a combination of multiple peptide signals, and is therefore more representative of the protein as a whole.

After implementation of the sum normalization method, drug treatment time courses were visualized with bar graphs and intersample comparisons were conducted using

ANOVAs. Of the DAT peptides, only EGAAGVWK, which is located in the first intracellular loop, has variable levels at the middle time point with both treatments.

Additionally, LAYAITPEK, which is located in the C-terminal tail of DAT, was found to decrease in the final time point of the PMA treatment time course (Figure 6.4).

103 a) PMA

EVELILVK EGAAGVWK AYLSVDFYR FCSLPGSFR LAYAITPEK

4.0E+06 1.6E+05 1.2E+06 6.0E+05 1.6E+06

3.0E+06 1.2E+05 9.0E+05 4.5E+05 1.2E+06

2.0E+06 8.0E+04 6.0E+05 3.0E+05 8.0E+05

1.0E+06 4.0E+04 3.0E+05 1.5E+05 4.0E+05

Normalized Abundance Abundance Normalized 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0 5 15 0 5 15 0 5 15 0 5 15 0 5 15 Time Time Time Time Time b) Amph EVELILVK EGAAGVWK AYLSVDFYR FCSLPGSFR LAYAITPEK

5.0E+06 2.0E+05 1.2E+06 4.0E+05 1.6E+06

4.0E+06 1.6E+05 9.0E+05 3.0E+05 1.2E+06 3.0E+06 1.2E+05 6.0E+05 2.0E+05 8.0E+05 2.0E+06 8.0E+04 3.0E+05 1.0E+05 4.0E+05 1.0E+06 4.0E+04

Normalized Abundance Abundance Normalized 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0 15 30 0 15 30 0 15 30 0 15 30 0 15 30 Time Time Time Time Time

Figure 6.4 – DAT time courses. Normalized DAT abundances are shown over a) PMA and b) amphetamine time-courses. While most peptides remain stable across the time courses, EGAAGVWK varies significantly at the middle time point with both treatments, and LAYAITPEK decreases significantly by the end of the PMA time course based on both ANOVA and Student’s t-tests (p < 0.05). Blue columns indicate WT values while red columns indicate HA-DAT values.

Normalized time courses of putative interactome members did not show significant variability across time points (Figure 6.5). Aside from DAT, of the 11 peptides from the 4 proteins that were significantly increased (t-test, p < 0.05) from WT controls, only one peptide in each of the treatment types was significantly altered over the time course based on ANOVA (p < 0.05). Both of these peptides were from the guanine nucleotide-binding protein G(o) subunit alpha (Gnao1). With PMA treatment,

LWGDSGIQECFNR was increased at the intermediate time point, whereas with amphetamine treatment, YYLDSLDR was increased at the final time point. These changes could be due to peptide modifications. For example, if the serine residue in either

104 peptide were phosphorylated, it would interfere with detection of that peptide and could alter the observed abundance. As a G protein, Gnao1 is involved as a modulator in transmembrane signaling systems, thus this association and changes could indicate changing signaling pathways associated with the drug treatments. a) PMA ATP1b1 - ATP1b1 - Gnao1 - Gnao1 - Gnao1 – VAPPGLTQIPQIQK YNPNVLPVQCTGK IGAGDYQPTEQDILR LWGDSGIQECFNR YYLDSLDR

60000 120000 120000 45000 120000 40000 50000 100000 100000 100000 35000 40000 80000 80000 30000 80000 25000 30000 60000 60000 60000 20000 20000 40000 40000 15000 40000 10000 10000 20000 20000 20000 5000

Normalized Abundance Abundance Normalized 0 0 0 0 0 0 5 15 0 5 15 0 5 15 0 5 15 0 5 15 Time Time Time Time Time b) Amph

ATP1b1 - ATP1b1 - Gnao1 - Gnao1 - Gnao1 – VAPPGLTQIPQIQK YNPNVLPVQCTGK IGAGDYQPTEQDILR LWGDSGIQECFNR YYLDSLDR

70000 160000 120000 30000 140000 60000 140000 100000 25000 120000 50000 120000 100000 100000 80000 20000 40000 80000 80000 60000 15000 30000 60000 60000 40000 10000 20000 40000 40000 10000 20000 20000 5000 20000

Normalized Abundance Abundance Normalized 0 0 0 0 0 0 15 30 0 15 30 0 15 30 0 15 30 0 15 30 Time Time Time Time Time

Figure 6.5 – DAT interactor time courses. Normalized DAT interactor abundances are shown over a) PMA and b) amphetamine time-courses. The normalized abundances of representative peptides are shown from two of the putative interactome proteins. Both peptides from ATP1b1 remain stable over the time courses as measure by intersample ANOVAs (p < 0.05). One Gnao1 peptide (LWGDSGIQECFNR) shows variable levels across the PMA time course. Another Gnao1 peptide (YYLDSLDR) is increasing across the amphetamine time course as shown by both ANOVA and t-test (p < 0.05). Blue columns indicate WT values while red columns indicate HA-DAT values.

However, the rest of the peptides remained stably above WT levels without changing over the time courses. This is demonstrated in Figure 6.5 with

Sodium/potassium-transporting ATPase subunit beta-1 (Atp1b1). Atp1b1 is a plasma membrane protein involved in cell adhesion and polarity that regulates the number of

105 sodium pumps on the plasma membrane152. An interaction between DAT and ATP1b1 could indicate cooperation in maintaining cellular Na+ levels.

Additional proteins in which this steady association occurred include

Dihydropyrimidinase-related protein 2 (Dpysl2), a protein involved in neuronal development and polarity, and syntaxin binding protein 1 (Stxbp1), a synaptic vesicle protein with a titular protein known to interact with DAT. Dpysl2 is a neuronal protein that has been associated with schizophrenia153, depression154, and Alzheimer’s disease155.

As DAT has also been associated with both schizophrenia and depression, it could be highly informative to further analyze this association in order to increase comprehension of these complex disorders. Furthermore, the interaction between syntaxin 1A and DAT has been shown to promote the amphetamine-induced DA efflux146, such that we would expect this interaction to change with and without amphetamine treatment.

These steady protein and peptide levels indicate that many of the proteins identified in this study, although significantly different between the HA-DAT and WT samples, are not changing with the administered drug treatments. The knowledge that these proteins remain unchanged with these drug treatments show that they are more stable DAT interactors and could be important for more basic DAT trafficking patterns.

Additional protein interactors could be responsible for the alterations in DAT trafficking and should be further analyzed.

Summary

Within this chapter, the IgG normalization method was implemented on several sets of IP data using two different MS analytical techniques. These IPs were all

106 conducted using epitope-tagged HA-DAT knock-in mice with and without drug treatments.

Initial experiments explored the DAT proteome in the HA-DAT knock-in mice through comparative label-free differential DDA analysis. This was done in both striatal homogenates and in striatal synaptosomes treated with PMA, amphetamine, or DMSO

(as a negative control). Resulting peptide abundances were normalized based on the sum of the six previously characterized mouse IgG peptides. These profiling studies revealed many potential DAT interactome members. The list of which was then narrowed down based on previously known DAT interactome members, roles in trafficking and signaling pathways, and specific localization to synaptosomes for further targeted analyses. The remaining proteins constitute a pool of potential candidates for later studies.

Targeted SRM analyses were conducted on the focused list of candidates using

IPs of synaptosomes treated with PMA or amphetamine over a time course, beginning with no treatment and including both partially and fully effective time points for both treatments. Multiple normalization techniques were performed on these data and it was verified that the sum normalization was the most applicable method. After sum normalization, these studies showed that multiple peptides from four proteins were significantly increased in the HA-DAT samples compared to the WT, aside from DAT.

However, only two of these peptides changed significantly over the drug treatment time courses. Both of these peptides were from the g-protein Gnao1. This could indicate that modification of these peptides is occurring with drug treatments, a result that would have an important impact on future drug treatment studies. Additional protein candidates,

107 including Atp1b1, Dpysl2, and Stxbp1, remained steady across the time courses, but could still be interesting for further studies based on their biological relevance.

This chapter demonstrated use of the IgG normalization method upon three large datasets in order to identify and quantify members of the DAT interactome in epitope tagged HA-DAT mice. Multiple variations of this technique were implemented, and the summed method was chosen for further usage. Additionally, this method aided in the identification of several novel DAT interactome members, though only one of these was found to have increased abundance levels with these drug treatments. These results validate the need for a normalization method such as this one in similar, future studies.

108 CHAPTER VII

CONCLUSIONS AND FUTURE DIRECTIONS

Introduction

This dissertation focused on a proteomic analysis of the dopamine transporter

(DAT) and the proteins that interact with it under various conditions. DAT clears the synapse of dopamine in order to terminate its transmission and thus proper functionality of DAT is critical in dopaminergic signaling pathways. DAT has been found to undergo complex trafficking patterns that are affected by multiple external stimuli such as drugs of abuse.

As DAT and the rest of the mammalian members of the neurotransmitter sodium symporter (NSS) family do not yet have solved crystal structures, their bacterial homolog leucine transporter (LeuT) currently serves as a structural proxy. This allows for assumptions to be made regarding the effects of structural conformations based on further modeling and mutational studies. Thus, mutational analyses of DAT have been associated with these structural conformations and the resulting trafficking patterns131.

The DAT interactome has been found to be a complex, dynamic entity. A variety of proteins interact with DAT, several of which have been shown to interact directly.

Previous studies have indicated that additional interactors likely effect DAT trafficking and processing patterns78. However, few studies have thus far analyzed DAT interacting proteins through MS based proteomic analyses, neglecting a powerful technique that can help unravel the complex trafficking patterns for future applications. This dissertation used several IP-MS approaches to study the DAT interactome under multiple conditions

109 and during this process a new analytical tool was developed for future IP-MS studies.

Use of this novel tool will help ensure that further MS studies of IPed proteins is both more reliable and applicable to a broad spectrum of biological research.

Summary of Findings

Chapter III – The membrane topology of the bacterial DAT homolog LeuT was examined through membrane shaving techniques. As LeuT is the only member of the

NSS family with a solved crystal structure, it was the optimal starting point for MS-based structural analysis. This analysis was meant to define the membrane accessible regions of the binding pocket in different LeuT conformations. It was determined that a photo- activatable leucine analog (pL) acts similarly to leucine as a substrate for LeuT through

3H-Leu uptake assays. This analog was intended to lock LeuT into its substrate occluded conformation when crosslinked via UV exposure, thus crosslinking pL to LeuT before membrane shaving and MS analyses. However, complications ensued at this point in the project resulting in low LeuT sequence coverage such that differentiation between peptide levels with and without pL was impossible. Although this project was unsuccessful, it was an interesting and novel approach that unveiled several difficulties and subsequent resolutions inherent in the proteomic analysis of these highly complex and hydrophobic proteins. The resulting processing techniques for analyzing such complex proteins were applied throughout the remainder of the research within this dissertation to ensure higher quality results.

110 Chapter IV – A normalization method was developed for IP-MS studies based on the abundance levels of antibody peptides. Particularly, antibodies were chosen as the basis for normalization because antibody abundances should be constant across conditions, whereas bait levels may be altered with mutations and should not be present in control samples. This normalization to antibody levels ensures equal sample loading for further analyses. This study was conducted using IPs of three different DAT constructs with varying levels of expression in PAE cells. IPs were targeted toward two different tags on the DAT constructs using both rabbit and mouse antibodies, which are the most widely used types of antibodies. Specifically, peptides in both the heavy and light chain constant domains of these IgGs were targeted for SRM analysis. This allowed for relative quantitation of these peptides within the different IP samples. Both IgG and

DAT peptide abundances were normalized to each of the different IgG peptides for analysis on the changes in relative abundance. Significant changes were found between pre- and post-normalized values. Additionally, it was found that intersample variability was decreased with this normalization resulting in more reliable relative quantitation between samples.

Chapter V – The IgG normalization method was applied to a complex dataset to analyze the DAT interactome in cells. Three dually tagged DAT constructs with distinctive trafficking and expression patterns were stably expressed in PAE cells. These constructs included an N-terminal YFP tag as well as an HA tag in the second extracellular loop, and consisted of a full-length (FL) version as well as N-terminally

(ΔN) and C-terminally (ΔC) deleted versions. Using antibodies targeted to the epitope tags, these constructs were IPed from the cells in order to analyze the changes in the DAT

111 interactome associated with these differing trafficking patterns. Normalization was conducted on label-free differential DDA data using a normalization factor based on the sum of all identified heavy and light chain constant domain peptides. This resulted in a decrease in the variability of IgG abundances between samples, while DAT levels remained distinctly different as has been seen previously through western blot analyses.

This analysis identified several proteins that increased when compared to parental controls across all three DAT constructs. These comparative analyses were conducted at both the peptide and protein level in order to account for differences due to modifications and domain effects. Additionally, several proteins were shown to increase in the ΔN and

ΔC IPs when compared to the FL. These putative interactome members included previously known interactors as well as additional novel proteins that are of future interest. One such protein is calnexin, which is an ER chaperone protein. This protein is found in increased abundance in the ΔC IPs when compared to the FL, which can be correlated to the ΔC-DAT localization in the ER. Targeted research based on this finding would lead to a more robust understanding of DAT trafficking out of the ER.

Chapter VI – The DAT interactome was further examined using a knock-in mouse model expressing DAT tagged with an HA epitope within the second extracellular loop.

Striatal homogenates and striatal synaptosomes treated with PMA, amphetamine, or

DMSO were analyzed through IPs coupled with DDA profiling. Normalization of these samples was implemented using a summed normalization factor based on the six mouse

IgG peptides that were characterized in Chapter IV. These studies identified putative

DAT interactors based on robust signals from multiple peptides with differential ratio comparisons between HA-DAT and WT samples. A list of peptides and corresponding

112 proteins was composed based on these DDA data for further targeted SRM analyses, peptide verification, and relative quantitation. Proteins were chosen based on previously known DAT interactions, involvement in trafficking and signaling pathways, and synaptosomal expression patterns. This generated a list of additional putative interactors that could be further analyzed in additional studies.

For targeted analyses, striatal synaptosomes were treated with PMA or amphetamine over time courses, which represented zero, partial, and full effectiveness of the drug treatments prior to IP and SRM. Resulting IgG and DAT peptide abundances were normalized based on three different normalization factors for comparison. These methods included normalization based on either of the two IgG peptides with the lowest variability as well as normalization based on the sum of the six targeted IgG peptides.

The summed normalization method was chosen for further analyses due to its better representation of the protein as a whole. The additional targeted peptides were normalized using the sum method. This revealed four proteins that were increased in the

HA-DAT synaptosomal IPs when compared to WT. However, only one protein displayed variability across the time courses, and this protein did not change uniformly across the time course with different peptides. Instead, one peptide was increased across the amphetamine time course while another varied at the intermediate time point of the PMA time course. This variability indicates the possibility of modifications or domain effects acting specifically upon certain regions of the protein. Verification of such modifications or domain effects could help further elucidate the mechanisms involved in these pathways.

113 Overall, these data presented in this dissertation showed that the implementation of the IgG normalization method accounts for the intersample variability inherent in the

IP-MS pipeline. Application of this method, for both DDA and SRM analyses, is most effective when using multiple IgG peptides to provide a more accurate representation of the protein as a whole, as well as multiple peptides of all other analyzed proteins to account for peptide level changes. This normalization method will produce a decrease in intersample variability in future IP-MS studies resulting in more exact comparative analyses and greater confidence in protein-protein interaction identifications.

Future Directions

Biological confirmation of the identified DAT interactors should help to clarify the nature of such interactions. Such studies should also show whether the interactions are direct or part of a protein complex. Direct interactions can be identified through multiple methods, including fluorescence resonance energy transfer (FRET) microscopy and yeast two-hybrid assays. Additionally, RNA interference studies could help to understand the downstream effects of these interactions. Confirmation of these interactions will help elucidate the mechanisms involved in these altered trafficking patterns. Furthermore, additional bioinformatics approaches could be employed in order to delve further into the intricacies of the profiling results.

Supplementary MS studies could also be used to examine and confirm interactors orthogonally. For example, data independent acquisition (DIA) could be used to collect large amounts of quantitative high resolution MS/MS data. DIA was considered for analysis within this dissertation, however time limitations prohibited from such

114 investigations. DIA data can later be mined for additional interactors as well as for post- translational modifications. Such analysis will allow for a non-biased representation of sample components and, unlike SRM data, will allow for subsequent analyses of the same dataset if additional putative interactors and modifications come to light. Overall application of this method would be beneficial to this project as it is both highly quantitative and hypothesis driven, thus representing the best attributes of both of the MS methods presented in this dissertation.

Concluding Remarks

Currently, many IP-MS pipelines tend to overlook several inherent issues with the technique. Primarily, studies often assume that IP samples with equal concentrations are equivalent when loaded onto the mass spectrometer such that samples are compared on a one-to-one basis. However, this is not always the case and this assumption can skew data.

This fact is frequently taken into account with western blot analysis of IPs in the form of loading controls, yet this approach is less quantitative due to limited dynamic range. Prior to this dissertation no standardized method had been created to account for this issue in

MS applications. Therefore, this dissertation created and implemented a novel and crucial normalization method to ensure equivalent intersample comparative quantitation in IP-

MS studies.

The normalization technique described within this dissertation utilizes immunoglobulin constant domain peptides from the antibody used for the IP.

Normalization is based on IgG levels, as opposed to bait levels, as bait levels may be altered with mutations and under different IP conditions. Additionally, characterization of

115 both mouse and rabbit IgGs was completed, the two most common animals used for IPs, so that the method can be applied across the field.

Another common issue with IP-MS studies is the use of individual peptides for quantification of a protein as a whole. This can easily misrepresent protein abundance, as some peptides may be more or less abundant under different conditions as a result of modifications or domain accessibility changes. Thus, it is necessary to monitor multiple peptides in all analyzed proteins. This practice is applicable to both the IgGs used in normalization and to the bait and the proteins associated with it.

Within this dissertation, IgG normalization with multiple peptides was applied to several IP-MS datasets in order to analyze changes in DAT trafficking patterns. The application of this method resulted in decreased intersample variability and also identified several putative DAT interactome members in both cells and mice.

Accordingly, future IP-MS studies should normalize data to IgG levels based on multiple peptides from the antibody. This type of normalization method is of critical importance for further IP-MS studies in order for findings to be replicable and applicable. It is the recommendation of this dissertation that such normalization methods be implemented immediately into all current and future experiments of this sort.

116 REFERENCES

1. Nelson, N. N. The family of Na+/Cl- neurotransmitter transporters. J Neurochem 71, 1785–1803 (1998).

2. Iversen, L. L. Role of transmitter uptake mechanisms in synaptic neurotransmission. Br J Pharmacol 41, 571–591 (1971).

3. Amara, S. G. & Kuhar, M. J. Neurotransmitter transporters: recent progress. Annu. Rev. Neurosci. 16, 73–93 (1993).

4. Cook, E. H. et al. Association of attention-deficit disorder and the dopamine transporter gene. American Journal of Human Genetics 56, 993–998 (1995).

5. Seeman, P. & Niznik, H. B. Dopamine receptors and transporters in Parkinson's disease and schizophrenia. FASEB J 4, 2737–2744 (1990).

6. Zahniser, N. R. & Sorkin, A. Trafficking of dopamine transporters in psychostimulant actions. Semin Cell Dev Biol 20, 411–417 (2009).

7. Kahlig, K. M. & Galli, A. Regulation of dopamine transporter function and plasma membrane expression by dopamine, amphetamine, and cocaine. Eur J Pharmacol 479, 153–158 (2003).

8. Torres, G. E., Gainetdinov, R. R. & Caron, M. G. Plasma membrane monoamine transporters: structure, regulation and function. Nat Rev Neurosci 4, 13–25 (2003).

9. Amara, S. G. & Sonders, M. S. Neurotransmitter transporters as molecular targets for addictive drugs. Drug Alcohol Depend 51, 87–96 (1998).

10. Yamashita, A., Singh, S. K., Kawate, T., Jin, Y. & Gouaux, E. Crystal structure of a bacterial homologue of Na+/Cl--dependent neurotransmitter transporters. Nature 437, 215–223 (2005).

11. Zhou, Z. et al. LeuT-desipramine structure reveals how antidepressants block neurotransmitter reuptake. Science 317, 1390–1393 (2007).

12. Singh, S. K., Yamashita, A. & Gouaux, E. Antidepressant binding site in a bacterial homologue of neurotransmitter transporters. Nature 448, 952–956 (2007).

13. Krishnamurthy, H. & Gouaux, E. X-ray structures of LeuT in substrate-free outward-open and apo inward-open states. Nature 481, 469–474 (2012).

117 14. Singh, S. K., Piscitelli, C. L., Yamashita, A. & Gouaux, E. A competitive inhibitor traps LeuT in an open-to-out conformation. Science 322, 1655–1661 (2008).

15. Stockner, T. et al. Mutational Analysis of the High-Affinity Zinc Binding Site Validates a Refined Human Dopamine Transporter Homology Model. 9, e1002909 (2013).

16. Rost, B. Twilight zone of protein sequence alignments. Protein Eng. Des. Sel. 12, 85–94 (1999).

17. Beuming, T., Shi, L., Javitch, J. A. & Weinstein, H. A comprehensive structure- based alignment of prokaryotic and eukaryotic neurotransmitter/Na+ symporters (NSS) aids in the use of the LeuT structure to probe NSS structure and function. Mol Pharmacol 70, 1630–1642 (2006).

18. Carlsson, A. Perspectives on the discovery of central monoaminergic neurotransmission. Annu. Rev. Neurosci. 10, 19–40 (1987).

19. Nicholls, J. G., Martin, A. R., Wallace, B. G. & Fuchs, P. A. From Neuron to Brain. 284–286 (Sinauer Associates, Inc).

20. Hoffman, B. J., Hansson, S. R., Mezey, E. & Palkovits, M. Localization and dynamic regulation of biogenic amine transporters in the mammalian central nervous system. Front Neuroendocrinol 19, 187–231 (1998).

21. Nirenberg, M. J., Vaughan, R. A., Uhl, G. R., Kuhar, M. J. & Pickel, V. M. The dopamine transporter is localized to dendritic and axonal plasma membranes of nigrostriatal dopaminergic neurons. jneurosci.org

22. Rao, A., Simmons, D. & Sorkin, A. Differential subcellular distribution of endosomal compartments and the dopamine transporter in dopaminergic neurons. Mol. Cell. Neurosci. 46, 148–158 (2011).

23. Sorkina, T., Hoover, B. R., Zahniser, N. R. & Sorkin, A. Constitutive and protein kinase C-induced internalization of the dopamine transporter is mediated by a clathrin-dependent mechanism. Traffic 6, 157–170 (2005).

24. Cremona, M. L. et al. Flotillin-1 is essential for PKC-triggered endocytosis and membrane microdomain localization of DAT. Nat Neurosci 14, 469–477 (2011).

25. Sorkina, T., Caltagarone, J. & Sorkin, A. Flotillins Regulate Membrane Mobility of the Dopamine Transporter but Are Not Required for Its Protein Kinase C Dependent Endocytosis. Traffic 14, 709–724 (2013).

118 26. Hoover, B. R., Everett, C. V., Sorkin, A. & Zahniser, N. R. Rapid regulation of dopamine transporters by tyrosine kinases in rat neuronal preparations. J Neurochem 101, 1258–1271 (2007).

27. Loder, M. K. & Melikian, H. E. The dopamine transporter constitutively internalizes and recycles in a protein kinase C-regulated manner in stably transfected PC12 cell lines. J Biol Chem 278, 22168–22174 (2003).

28. Vaughan, R. A. Protein Kinase C-mediated Phosphorylation and Functional Regulation of Dopamine Transporters in Striatal Synaptosomes. Journal of Biological Chemistry 272, 15541–15546 (1997).

29. Foster, J. D., Adkins, S. D., Lever, J. R. & Vaughan, R. A. Phorbol ester induced trafficking-independent regulation and enhanced phosphorylation of the dopamine transporter associated with membrane rafts and cholesterol. J Neurochem 105, 1683–1699 (2008).

30. Miranda, M., Wu, C. C., Sorkina, T., Korstjens, D. R. & Sorkin, A. Enhanced ubiquitylation and accelerated degradation of the dopamine transporter mediated by protein kinase C. J Biol Chem 280, 35617–35624 (2005).

31. Li, L.-B. et al. The role of N-glycosylation in function and surface trafficking of the human dopamine transporter. J Biol Chem 279, 21012–21020 (2004).

32. Miranda, M., Dionne, K. R., Sorkina, T. & Sorkin, A. Three ubiquitin conjugation sites in the amino terminus of the dopamine transporter mediate protein kinase C-dependent endocytosis of the transporter. Mol Biol Cell 18, 313–323 (2007).

33. Sorkina, T. et al. RNA interference screen reveals an essential role of Nedd4-2 in dopamine transporter ubiquitination and endocytosis. Journal of Neuroscience 26, 8195–8205 (2006).

34. Vina-Vilaseca, A. & Sorkin, A. Lysine 63-linked polyubiquitination of the dopamine transporter requires WW3 and WW4 domains of Nedd4-2 and UBE2D ubiquitin-conjugating enzymes. J Biol Chem 285, 7645–7656 (2010).

35. Gorentla, B. K., Moritz, A. E., Foster, J. D. & Vaughan, R. A. Proline-Directed Phosphorylation of the Dopamine Transporter N-Terminal Domain †. Biochemistry 48, 1067–1076 (2009).

36. Moritz, A. E. et al. Phosphorylation of dopamine transporter serine 7 modulates cocaine analog binding. J Biol Chem 288, 20–32 (2013).

37. Foster, J. D. et al. Dopamine transporter phosphorylation site threonine 53 regulates substrate reuptake and amphetamine-stimulated efflux.

119 38. Chen, R. et al. Direct evidence that two cysteines in the dopamine transporter form a disulfide bond. Mol Cell Biochem 298, 41–48 (2006).

39. Hastrup, H., Sen, N. & Javitch, J. A. The Human Dopamine Transporter Forms a Tetramer in the Plasma Membrane: Cross-linking of a cysteine in the fourth transmembrane segment is sensitive to cocaine analogs. Journal of Biological Chemistry 278, 45045–45048 (2003).

40. Varga, G. G., Szekely, A. A. & Sasvari-Szekely, M. M. Candidate gene studies of dopaminergic and serotonergic polymorphisms. Neuropsychopharmacol Hung 13, 93–101 (2011).

41. Vandenbergh, D. J. et al. Human dopamine transporter gene (DAT1) maps to chromosome 5p15.3 and displays a VNTR. Genomics 14, 1104–1106 (1992).

42. Heinz, A. Genotype Influences In Vivo Dopamine Transporter Availability in Human Striatum. Neuropsychopharmacology 22, 133–139 (2000).

43. Cook, E. H. et al. Association of attention-deficit disorder and the dopamine transporter gene. American Journal of Human Genetics 56, 993–998 (1995).

44. El-Tarras, A. E. et al. Association study between the dopamine-related candidate gene polymorphisms and ADHD among Saudi Arabia population via PCR technique. Mol Biol Rep 39, 11081–11086 (2012).

45. Volkow, N. D. et al. Evaluating Dopamine Reward Pathway in ADHDClinical Implications. JAMA 302, 1084–1091 (2009).

46. Enter, D., Colzato, L. S. & Roelofs, K. Dopamine transporter polymorphisms affect social approach-avoidance tendencies. Br J Pharmacol 11, 671–676 (2012).

47. Mata, R., Hau, R., Papassotiropoulos, A. & Hertwig, R. DAT1 Polymorphism Is Associated with Risk Taking in the Balloon Analogue Risk Task (BART). PLoS ONE 7, e39135 (2012).

48. Newman, D. P., O'Connell, R. G., Nathan, P. J. & Bellgrove, M. A. Dopamine transporter genotype predicts attentional asymmetry in healthy adults. Eur J Pharmacol 50, 2823–2829 (2012).

49. Costa, A., Riedel, M., Müller, U., Möller, H.-J. & Ettinger, U. Relationship between SLC6A3 genotype and striatal dopamine transporter availability: A meta-analysis of human single photon emission computed tomography studies. Br J Pharmacol 65, 998–1005 (2011).

120 50. Shumay, E., Chen, J., Fowler, J. S. & Volkow, N. D. Genotype and Ancestry Modulate Brain's DAT Availability in Healthy Humans. PLoS ONE 6, e22754 (2011).

51. Afonso-Oramas, D. et al. Dopamine transporter glycosylation correlates with the vulnerability of midbrain dopaminergic cells in Parkinson's disease. Neurobiol. Dis. 36, 494–508 (2009).

52. Fusar-Poli, P. & Meyer-Lindenberg, A. Striatal Presynaptic Dopamine in Schizophrenia, Part I: Meta-Analysis of Dopamine Active Transporter (DAT) Density. schizophreniabulletin.oxfordjournals.org

53. Sakrikar, D. et al. Attention Deficit/Hyperactivity Disorder-Derived Coding Variation in the Dopamine Transporter Disrupts Microdomain Targeting and Trafficking Regulation. Journal of Neuroscience 32, 5385–5397 (2012).

54. Wei, Y. et al. Dopamine Transporter Activity Mediates Amphetamine-Induced Inhibition of Akt through a Ca2+/Calmodulin-Dependent Kinase II-Dependent Mechanism. Mol Pharmacol 71, 835–842 (2006).

55. Garcia, B. G. et al. Akt Is Essential for Insulin Modulation of Amphetamine- Induced Human Dopamine Transporter Cell-Surface Redistribution. Mol Pharmacol 68, 102–109 (2005).

56. Sorkina, T., Doolen, S., Galperin, E., Zahniser, N. R. & Sorkin, A. Oligomerization of dopamine transporters visualized in living cells by fluorescence resonance energy transfer microscopy. J Biol Chem 278, 28274– 28283 (2003).

57. Kahlig, K. M., Javitch, J. A. & Galli, A. Amphetamine regulation of dopamine transport. Combined measurements of transporter currents and transporter imaging support the endocytosis of an active carrier. J Biol Chem 279, 8966– 8975 (2004).

58. Kahlig, K. M. et al. Amphetamine induces dopamine efflux through a dopamine transporter channel. Proc Natl Acad Sci USA 102, 3495–3500 (2005).

59. Johnson, L. A., Furman, C. A., Zhang, M., Guptaroy, B. & Gnegy, M. E. Rapid delivery of the dopamine transporter to the plasmalemmal membrane upon amphetamine stimulation. Neuropharmacology 49, 750–758 (2005).

60. Beuming, T. et al. The binding sites for cocaine and dopamine in the dopamine transporter overlap. Nat Neurosci 11, 780–789 (2008).

121 61. Little, K. Y., Elmer, L. W., Zhong, H., Scheys, J. O. & Zhang, L. Cocaine induction of dopamine transporter trafficking to the plasma membrane. Mol Pharmacol 61, 436–445 (2002).

62. Daws, L. C. et al. Cocaine increases dopamine uptake and cell surface expression of dopamine transporters. Biochem Biophys Res Commun 290, 1545–1550 (2002).

63. Bisgaard, H. et al. The binding sites for benztropines and dopamine in the dopamine transporter overlap. Eur J Pharmacol 60, 182–190 (2011).

64. Mayfield, R. D. & Zahniser, N. R. Dopamine D2 receptor regulation of the dopamine transporter expressed in Xenopus laevis oocytes is voltage- independent. Mol Pharmacol 59, 113–121 (2001).

65. Maiya, R., Buck, K. J., Harris, R. A. & Mayfield, R. D. Ethanol-sensitive sites on the human dopamine transporter. J Biol Chem 277, 30724–30729 (2002).

66. Daniels, G. & Amara, S. Regulated trafficking of the human dopamine transporter. Journal of Biological Chemistry 274, 35794–35801 (1999).

67. Young, K. H. Yeast two-hybrid: so many interactions, (in) so little time... Biol. Reprod. 58, 302–311 (1998).

68. Lee, F. J. S., Liu, F., Pristupa, Z. B. & Niznik, H. B. Direct binding and functional coupling of alpha-synuclein to the dopamine transporters accelerate dopamine-induced apoptosis. The FASEB Journal 15, 916–926 (2001).

69. Carneiro, A. M. et al. The Multiple LIM Domain-Containing Adaptor Protein Hic-5 Synaptically Colocalizes and Interacts with the Dopamine Transporter. J Neurosci 22, 7045–7054 (2002).

70. Torres, G. E. et al. Functional Interaction between Monoamine Plasma Membrane Transporters and the Synaptic PDZ Domain–Containing Protein PICK1. Eur J Pharmacol 30, 121–134 (2001).

71. Lee, K.-H., Kim, M.-Y., Kim, D.-H. & Lee, Y.-S. Syntaxin 1A and Receptor for Activated C Kinase Interact with the N-Terminal Region of Human Dopamine Transporter. Mol Neurobiol 29, 1405–1409 (2004).

72. Grefen, C., Lalonde, S. & Obrdlik, P. Split-Ubiquitin System for Identifying Protein-Protein Interactions in Membrane and Full-Length Proteins. (John Wiley & Sons, Inc., 2001). doi:10.1002/0471142301.ns0527s41

122 73. Egaña, L. A. et al. Physical and functional interaction between the dopamine transporter and the synaptic vesicle protein synaptogyrin-3. J Neurosci 29, 4592– 4604 (2009).

74. Lee, F. J. S. et al. Dopamine transporter cell surface localization facilitated by a direct interaction with the dopamine D2 receptor. EMBO J. 26, 2127–2136 (2007).

75. Bauman, A. L. A. et al. Cocaine and antidepressant-sensitive biogenic amine transporters exist in regulated complexes with protein phosphatase 2A. J Neurosci 20, 7571–7578 (2000).

76. Johnson, L. A., Guptaroy, B., Lund, D., Shamban, S. & Gnegy, M. E. Regulation of Amphetamine-stimulated Dopamine Efflux by Protein Kinase C. Journal of Biological Chemistry 280, 10914–10919 (2005).

77. Maiya, R., Ponomarev, I., Linse, K. D., Harris, R. A. & Mayfield, R. D. Defining the dopamine transporter proteome by convergent biochemical and in silico analyses. Genes Brain Behav. 6, 97–106 (2007).

78. Torres, G. E. The dopamine transporter proteome. J Neurochem 97 Suppl 1, 3– 10 (2006).

79. Cox, B. & Emili, A. Tissue subcellular fractionation and protein extraction for use in mass-spectrometry-based proteomics. Nat Protoc 1, 1872–1878 (2006).

80. Gingras, A.-C., Gstaiger, M., Raught, B. & Aebersold, R. Analysis of protein complexes using mass spectrometry. Nat Rev Mol Cell Biol 8, 645–654 (2007).

81. Köcher, T. & Superti-Furga, G. Mass spectrometry–based functional proteomics: from molecular machines to protein networks. Nat Methods 4, 807–815 (2007).

82. O'Farrell, P. H. High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250, 4007–4021 (1975).

83. Wang, H. et al. Development and Evaluation of a Micro- and Nanoscale Proteomic Sample Preparation Method. J. Proteome Res. 4, 2397–2403 (2005).

84. Addona, T. A. et al. Multi-site assessment of the precision and reproducibility of multiple reaction monitoring–based measurements of proteins in plasma. Nat Biotechnol 27, 633–641 (2009).

85. Wu, C. C. & Yates, J. R. The application of mass spectrometry to membrane proteomics. Nat Biotechnol 21, 262–267 (2003).

123 86. Shevchenko, A., Tomas, H., Havli sbreve, J., Olsen, J. V. & Mann, M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat Protoc 1, 2856–2860 (2007).

87. Wessel, D. & Flügge, U. I. A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal Biochem 138, 141– 143 (1984).

88. Fic, E., Kedracka-Krok, S., Jankowska, U., Pirog, A. & Dziedzicka-Wasylewska, M. Comparison of protein precipitation methods for various rat brain structures prior to proteomic analysis. Electrophoresis 31, 3573–3579 (2010).

89. López-Ferrer, D. et al. Sample treatment for protein identification by mass spectrometry-based techniques. TrAC Trends in Analytical Chemistry 25, 996– 1005 (2006).

90. Mitulovic, G. G. & Mechtler, K. K. HPLC techniques for proteomics analysis--a short overview of latest developments. Brief Funct Genomic Proteomic 5, 249–260 (2006).

91. Link, A. J. et al. Direct analysis of protein complexes using mass spectrometry. Nat Biotechnol 17, 676–682 (1999).

92. Washburn, M. P., Wolters, D. & Yates, J. R. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat Biotechnol 19, 242–247 (2001).

93. Blackler, A. R., Speers, A. E. & Wu, C. C. Chromatographic benefits of elevated temperature for the proteomic analysis of membrane proteins. Proteomics 8, 3956–3964 (2008).

94. Speers, A. E., Blackler, A. R. & Wu, C. C. Shotgun analysis of integral membrane proteins facilitated by elevated temperature. Anal Chem 79, 4613– 4620 (2007).

95. Farias, S. E., Kline, K. G., Klepacki, J. & Wu, C. C. Quantitative improvements in peptide recovery at elevated chromatographic temperatures from microcapillary liquid chromatography-mass spectrometry analyses of brain using selected reaction monitoring. Anal Chem 82, 3435–3440 (2010).

96. Karas, M. & Krüger, R. Ion Formation in MALDI: The Cluster Ionization Mechanism. Chem. Rev. 103, 427–440 (2003).

97. Fenn, J. B. Electrospray Wings for Molecular Elephants (Nobel Lecture). Angew. Chem. Int. Ed. 42, 3871–3894 (2003).

124 98. Kebarle, P. & Verkerk, U. H. Electrospray: From ions in solution to ions in the gas phase, what we know now. Mass Spectrom Rev 28, 898–917 (2009).

99. Johnson, J. V., Yost, R. A., Kelley, P. E. & Bradford, D. C. Tandem-in-space and tandem-in-time mass spectrometry: triple quadrupoles and quadrupole ion traps. Anal Chem 62, 2162–2172 (1990).

100. Mathieu, É. Mémoire sur le mouvement vibratoire d'une membrane de forme elliptique. Journal de mathématiques pures et appliquées 137–203 (1868).

101. de Hoffmann, E. Tandem mass spectrometry: a primer. Journal of Mass Spectrometry 31, 129–137 (1996).

102. Gerber, S. A. Absolute quantification of proteins and phosphoproteins from cell lysates by tandem MS. Proc Natl Acad Sci USA 100, 6940–6945 (2003).

103. Maclean, B. et al. Skyline: an open source document editor for creating and analyzing targeted proteomics experiments. Bioinformatics 26, 966 (2010).

104. Doerr, A. Mass spectrometry–based targeted proteomics. Nat Methods 10, 23–23 (2012).

105. March, R. E. An introduction to quadrupole ion trap mass spectrometry. Journal of Mass Spectrometry 32, 351–369 (1997).

106. Blackler, A. R., Klammer, A. A., Maccoss, M. J. & Wu, C. C. Quantitative comparison of proteomic data quality between a 2D and 3D quadrupole ion trap. Anal Chem 78, 1337–1344 (2006).

107. Douglas, D. J., Frank, A. J. & Mao, D. Linear ion traps in mass spectrometry. Mass Spectrom Rev 24, 1–29 (2004).

108. Hu, Q. et al. The Orbitrap: a new mass spectrometer. J Mass Spectrom 40, 430–443 (2005).

109. Makarov, A. et al. Performance Evaluation of a Hybrid Linear Ion Trap/Orbitrap Mass Spectrometer. Anal Chem 78, 2113–2120 (2006).

110. Eng, J. K., McCormack, A. L. & Yates, J. R. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. Journal of the American Society for Mass Spectrometry 5, 976–989 (1994).

111. Perkins, D. N., Pappin, D. J., Creasy, D. M. & Cottrell, J. S. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20, 3551–3567 (1999).

125 112. Käll, L., Storey, J. D., Maccoss, M. J. & Noble, W. S. Assigning Significance to Peptides Identified by Tandem Mass Spectrometry Using Decoy Databases. J. Proteome Res. 7, 29–34 (2008).

113. Käll, L., Canterbury, J. D., Weston, J., Noble, W. S. & Maccoss, M. J. Semi- supervised learning for peptide identification from shotgun proteomics datasets. Nat Methods 4, 923–925 (2007).

114. Hoopmann, M. R., Finney, G. L. & Maccoss, M. J. High-speed data reduction, feature detection, and MS/MS spectrum quality assessment of shotgun proteomics data sets using high-resolution mass spectrometry. Anal Chem 79, 5620–5632 (2007).

115. Hsieh, E. J., Hoopmann, M. R., Maclean, B. & Maccoss, M. J. Comparison of Database Search Strategies for High Precursor Mass Accuracy MS/MS Data. J. Proteome Res. 9, 1138–1143 (2010).

116. Liu, H., Sadygov, R. G. & Yates, J. R. A Model for Random Sampling and Estimation of Relative Protein Abundance in Shotgun Proteomics. Anal Chem 76, 4193–4201 (2004).

117. Nilsson, T. et al. Mass spectrometry in high-throughput proteomics: ready for the big time. Nat Methods 7, 681–685 (2010).

118. Tabb, D. L. et al. Repeatability and Reproducibility in Proteomic Identifications by Liquid Chromatography−Tandem Mass Spectrometry. J. Proteome Res. 9, 761–776 (2010).

119. Hsieh, E. J. et al. Topograph, a Software Platform for Precursor Enrichment Corrected Global Protein Turnover Measurements. Mol Cell Proteomics 11, 1468–1474 (2012).

120. Bateman, N. W. et al. Maximizing peptide identification events in proteomic workflows utilizing data-dependent acquisition. mcponline.org

121. Gillet, L. C. et al. Targeted Data Extraction of the MS/MS Spectra Generated by Data-independent Acquisition: A New Concept for Consistent and Accurate Proteome Analysis. Mol Cell Proteomics 11, O111.016717–O111.016717 (2012).

122. Blackburn, K., Mbeunkui, F., Mitra, S. K., Mentzel, T. & Goshe, M. B. Improving protein and proteome coverage through data-independent multiplexed peptide fragmentation. J. Proteome Res. 9, 3621–3637 (2010).

126 123. Wu, C. C., Maccoss, M. J., Howell, K. E. & Yates, J. R. A method for the comprehensive proteomic analysis of membrane proteins. Nat Biotechnol 21, 532–538 (2003).

124. Blackler, A. R., Speers, A. E., Ladinsky, M. S. & Wu, C. C. A shotgun proteomic method for the identification of membrane-embedded proteins and peptides. J. Proteome Res. 7, 3028–3034 (2008).

125. Howell, K. E. & Palade, G. E. Hepatic Golgi fractions resolved into membrane and content subfractions. J Cell Biol 92, 822–832 (1982).

126. Moore, S. M. & Wu, C. C. in link.springer.com 155–164 (Humana Press, 2012). doi:10.1007/978-1-62703-023-6_9

127. Suchanek, M., Radzikowska, A. & Thiele, C. Photo-leucine and photo- methionine allow identification of protein-protein interactions in living cells. Nat Methods 2, 261–267 (2005).

128. Hill, R. A. & Sillence, M. N. Improved membrane isolation in the purification of beta 2-adrenoceptors from transgenic Escherichia coli. Protein Expr Purif 10, 162–167 (1997).

129. Wagner, S. et al. Tuning Escherichia coli for membrane protein overexpression. Proceedings of the National Academy of Sciences 105, 14371–14376 (2008).

130. Blackler, A. R. Shotgun Proteomic Methods for Integral Membrane Proteins: Applications to the Leucine and Dopamine Transporters. Dissertation, University of Colorado, Anschutz Medical Campus. 2008.

131. Sorkina, T., Richards, T. L., Rao, A., Zahniser, N. R. & Sorkin, A. Negative regulation of dopamine transporter endocytosis by membrane-proximal N- terminal residues. Journal of Neuroscience 29, 1361–1374 (2009).

132. Vermeulen, M., Hubner, N. C. & Mann, M. High confidence determination of specific protein–protein interactions using quantitative mass spectrometry. Eur J Pharmacol 19, 331–337 (2008).

133. Collins, M. O. & Choudhary, J. S. Mapping multiprotein complexes by affinity purification and mass spectrometry. Eur J Pharmacol 19, 324–330 (2008).

134. Bonifacino, J. S. & Dell'Angelica, E. C. Immunoprecipitation. Curr Protoc Cell Biol Chapter 7, Unit 7.2 (2001).

135. Ong, S.-E. et al. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell Proteomics 1, 376–386 (2002).

127 136. Harlow, E. & Lane, D. Antibodies: A Laboratory Manual. (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1988).

137. Schroeder, H. W. & Cavacini, L. Structure and function of immunoglobulins. J. Allergy Clin. Immunol. 125, S41–52 (2010).

138. Harper, D. R. & Murphy, G. Nonuniform variation in band pattern with luminol/horseradish peroxidase western blotting. Anal Biochem 192, 59–63 (1991).

139. Schutz-Geschwender, A., Zhang, Y., Holt, T., McDermitt, D. & Olive, D. M. Quantitative, two-color Western blot detection with infrared fluorescence. LI- COR Biosciences (2004).

140. Markham, K., Bai, Y. & Schmitt-Ulms, G. Co-immunoprecipitations revisited: an update on experimental concepts and their implementation for sensitive interactome investigations of endogenous proteins. Anal Bioanal Chem 389, 461–473 (2007).

141. Agger, S. A., Marney, L. C. & Hoofnagle, A. N. Simultaneous Quantification of Apolipoprotein A-I and Apolipoprotein B by Liquid-Chromatography-Multiple- Reaction-Monitoring Mass Spectrometry. Clinical Chemistry 56, 1804–1813 (2010).

142. Sherrod, S. D. et al. Label-Free Quantitation of Protein Modifications by Pseudo Selected Reaction Monitoring with Internal Reference Peptides. J. Proteome Res. 11, 3467–3479 (2012).

143. Sharma, V., Eng, J. K., Maccoss, M. J. & Riffle, M. MSDaPl (Mass Spectrometry Data Platform): A Proteomics Data Management Platform. Mol Cell Proteomics – (2012). doi:10.1074/mcp.O111.015149

144. Lodish, H. et al. Molecular Cell Biology. (W. H. Freeman and Company, 2008).

145. Storey, J. D. & Tibshirani, R. Statistical significance for genomewide studies. Proc Natl Acad Sci USA 100, 9440–9445 (2003).

146. Binda, F. et al. Syntaxin 1A interaction with the dopamine transporter promotes amphetamine-induced dopamine efflux. Mol Pharmacol 74, 1101–1108 (2008).

147. Rao, A. A., Richards, T. L. T., Simmons, D. D., Zahniser, N. R. N. & Sorkin, A. A. Epitope-tagged dopamine transporter knock-in mice reveal rapid endocytic trafficking and filopodia targeting of the transporter in dopaminergic axons. FASEB J 26, 1921–1933 (2012).

128 148. Richards, T. L. & Zahniser, N. R. Rapid substrate-induced down-regulation in function and surface localization of dopamine transporters: rat dorsal striatum versus nucleus accumbens. J Neurochem 108, 1575–1584 (2009).

149. German, C. L., Hanson, G. R. & Fleckenstein, A. E. Amphetamine and methamphetamine reduce striatal dopamine transporter function without concurrent dopamine transporter relocalization. Br J Pharmacol 123, 288–297 (2012).

150. Bereman, M. S., Maclean, B., Tomazela, D. M., Liebler, D. C. & Maccoss, M. J. The development of selected reaction monitoring methods for targeted proteomics via empirical refinement. Proteomics 12, 1134–1141 (2012).

151. Rogstad, S. M., Sorkina, T., Sorkin, A. & Wu, C. C. Improved precision of proteomic measurements in immuniprecipitation based purifications using relative quantitation. Anal Chem (2013). doi:10.1021/ac4002222

152. Bab-Dinitz, E. et al. A C-Terminal Lobe of the β Subunit of Na,K-ATPase and H,K-ATPase Resembles Cell Adhesion Molecules. Biochemistry 48, 8684–8691 (2009).

153. Arai, M. & Itokawa, M. A hard road in psychiatric genetics: schizophrenia and DPYSL2. J Hum Genet 55, 397–399 (2010).

154. Yang, Y. et al. Proteomics reveals energy and glutathione metabolic dysregulation in the prefrontal cortex of a rat model of depression. Neuroscience 247, 191–200 (2013).

155. Takano, M. et al. Proteomic analysis of the brain tissues from a transgenic mouse model of amyloid β oligomers. Neurochemistry international 61, 347–355 (2012).

129 APPENDIX A

IGG SRM TRANSITION INFORMATION

Table A1. Rabbit IgG SRM Method Transitions and Collision Energies.

Collision Peptide Protein Parent Ion Product y-ion Energy Rabbit IgG GYLPEPVTVTWNSGTLTNGVR Heavy Chain 1131.081674 1404.723003 36.8 Rabbit IgG GYLPEPVTVTWNSGTLTNGVR Heavy Chain 1131.081674 1305.654589 36.8 Rabbit IgG GYLPEPVTVTWNSGTLTNGVR Heavy Chain 1131.081674 1204.606911 36.8 Rabbit IgG DTLMISR Heavy Chain 418.220745 720.407271 15.5 Rabbit IgG DTLMISR Heavy Chain 418.220745 619.359593 15.5 Rabbit IgG DTLMISR Heavy Chain 418.220745 506.275529 15.5 Rabbit IgG DTLMISR Heavy Chain 418.220745 375.235044 15.5 Rabbit IgG TARPPLR Heavy Chain 405.750861 638.409655 15.1 Rabbit IgG TARPPLR Heavy Chain 405.750861 482.308544 15.1 Rabbit IgG TARPPLR Heavy Chain 405.750861 385.25578 15.1 Rabbit IgG TARPPLR Heavy Chain 405.750861 288.203016 15.1 Rabbit IgG EQQFNSTIR Heavy Chain 561.780544 865.452642 19.8 Rabbit IgG EQQFNSTIR Heavy Chain 561.780544 737.394064 19.8 Rabbit IgG EQQFNSTIR Heavy Chain 561.780544 590.32565 19.8 Rabbit IgG EQQFNSTIR Heavy Chain 561.780544 476.282723 19.8 Rabbit IgG VVSTLPIAHQDWLR Heavy Chain 817.954289 1135.600703 27.4 Rabbit IgG VVSTLPIAHQDWLR Heavy Chain 817.954289 1038.547939 27.4 Rabbit IgG VVSTLPIAHQDWLR Heavy Chain 817.954289 925.463875 27.4 Rabbit IgG VVSTLPIAHQDWLR Heavy Chain 817.954289 854.426761 27.4 Rabbit IgG ALPAPIEK Heavy Chain 419.755278 654.382102 15.5 Rabbit IgG ALPAPIEK Heavy Chain 419.755278 557.329339 15.5 Rabbit IgG ALPAPIEK Heavy Chain 419.755278 486.292225 15.5 Rabbit IgG GQPLEPK Heavy Chain 384.716153 711.403566 14.4

130 Rabbit IgG GQPLEPK Heavy Chain 384.716153 583.344989 14.4 Rabbit IgG GQPLEPK Heavy Chain 384.716153 486.292225 14.4 Rabbit IgG GQPLEPK Heavy Chain 384.716153 244.165568 14.4 Rabbit IgG EELSSR Heavy Chain 360.679767 591.309666 13.7 Rabbit IgG EELSSR Heavy Chain 360.679767 462.267073 13.7 Rabbit IgG EELSSR Heavy Chain 360.679767 349.183009 13.7 Rabbit IgG EELSSR Heavy Chain 360.679767 262.15098 13.7 Rabbit IgG AEDNYK Heavy Chain 370.166493 668.288596 14 Rabbit IgG AEDNYK Heavy Chain 370.166493 539.246003 14 Rabbit IgG AEDNYK Heavy Chain 370.166493 424.21906 14 Rabbit IgG AEDNYK Heavy Chain 370.166493 310.176132 14 Rabbit IgG LSVPTSEWQR Heavy Chain 601.811844 903.431906 21 Rabbit IgG LSVPTSEWQR Heavy Chain 601.811844 806.379142 21 Rabbit IgG LSVPTSEWQR Heavy Chain 601.811844 705.331464 21 Ig-Kappa- VTQGTTSVVQSFNR chain 762.394261 936.489755 25.8 Ig-Kappa- VTQGTTSVVQSFNR chain 762.394261 849.457727 25.8 Ig-Kappa- VTQGTTSVVQSFNR chain 762.394261 750.389313 25.8 Ig-Kappa- VTQGTTSVVQSFNR chain 762.394261 651.320899 25.8

131 Table A2. Mouse IgG1 SRM Method Transitions and Collision Energies. Colli sion Peptide Protein Parent Ion Product y-ion Ener gy DCGCKPCICTVPEVSSVFIFPPK PK Mouse Ig gamma-1 1461.704676 1474.830428 46.8 DCGCKPCICTVPEVSSVFIFPPK PK Mouse Ig gamma-1 1461.704676 566.366058 46.8 DCGCKPCICTVPEVSSVFIFPPK PK Mouse Ig gamma-1 1461.704676 469.313295 46.8 DVLTITLTPK Mouse Ig gamma-1 550.831714 985.629209 19.4 DVLTITLTPK Mouse Ig gamma-1 550.831714 886.560795 19.4 DVLTITLTPK Mouse Ig gamma-1 550.831714 773.476731 19.4 DVLTITLTPK Mouse Ig gamma-1 550.831714 672.429053 19.4 DVLTITLTPK Mouse Ig gamma-1 550.831714 559.344989 19.4 VTCVVVDISK Mouse Ig gamma-1 560.307549 1020.539408 19.7 VTCVVVDISK Mouse Ig gamma-1 560.307549 919.49173 19.7 VTCVVVDISK Mouse Ig gamma-1 560.307549 759.461081 19.7 VTCVVVDISK Mouse Ig gamma-1 560.307549 660.392667 19.7 VTCVVVDISK Mouse Ig gamma-1 560.307549 561.324253 19.7 SVSELPIMHQDWLNGK Mouse Ig gamma-1 927.46436 1451.746381 30.7 SVSELPIMHQDWLNGK Mouse Ig gamma-1 927.46436 1338.662317 30.7 SVSELPIMHQDWLNGK Mouse Ig gamma-1 927.46436 1241.609553 30.7 SVSELPIMHQDWLNGK Mouse Ig gamma-1 927.46436 1128.525489 30.7 SVSELPIMHQDWLNGK Mouse Ig gamma-1 927.46436 997.485004 30.7 VNSAAFPAPIEK Mouse Ig gamma-1 622.337695 1030.556772 21.6 VNSAAFPAPIEK Mouse Ig gamma-1 622.337695 943.524744 21.6 VNSAAFPAPIEK Mouse Ig gamma-1 622.337695 872.48763 21.6 VNSAAFPAPIEK Mouse Ig gamma-1 622.337695 801.450516 21.6 VNSAAFPAPIEK Mouse Ig gamma-1 622.337695 654.382102 21.6 VNSAAFPAPIEK Mouse Ig gamma-1 622.337695 486.292225 21.6 APQVYTIPPPK Mouse Ig gamma-1 605.845156 1139.645922 21.1 APQVYTIPPPK Mouse Ig gamma-1 605.845156 1042.593158 21.1 APQVYTIPPPK Mouse Ig gamma-1 605.845156 914.53458 21.1 APQVYTIPPPK Mouse Ig gamma-1 605.845156 815.466166 21.1 APQVYTIPPPK Mouse Ig gamma-1 605.845156 652.402838 21.1 APQVYTIPPPK Mouse Ig gamma-1 605.845156 438.271095 21.1 APQVYTIPPPK Mouse Ig gamma-1 605.845156 341.218332 21.1 QNGVLNSWTDQDSK kappa light chain 796.370983 1292.611721 26.8 QNGVLNSWTDQDSK kappa light chain 796.370983 1193.543307 26.8 QNGVLNSWTDQDSK kappa light chain 796.370983 1080.459243 26.8 QNGVLNSWTDQDSK kappa light chain 796.370983 966.416316 26.8 QNGVLNSWTDQDSK kappa light chain 796.370983 879.384287 26.8 DSTYSMSSTLTLTK kappa light chain 767.868896 1332.671544 25.9 DSTYSMSSTLTLTK kappa light chain 767.868896 1231.623866 25.9 DSTYSMSSTLTLTK kappa light chain 767.868896 1068.560537 25.9 DSTYSMSSTLTLTK kappa light chain 767.868896 981.528509 25.9 DSTYSMSSTLTLTK kappa light chain 767.868896 850.488024 25.9 HNSYTCEATHK kappa light chain 674.29095 1210.515713 23.1 HNSYTCEATHK kappa light chain 674.29095 1096.472785 23.1 HNSYTCEATHK kappa light chain 674.29095 1009.440757 23.1 HNSYTCEATHK kappa light chain 674.29095 846.377428 23.1 HNSYTCEATHK kappa light chain 674.29095 745.32975 23.1

132 TSTSPIVK kappa light chain 416.742368 731.429781 15.4 TSTSPIVK kappa light chain 416.742368 644.397753 15.4 TSTSPIVK kappa light chain 416.742368 543.350074 15.4 TSTSPIVK kappa light chain 416.742368 456.318046 15.4

Table A3. DAT-YFP SRM Method Transitions and Collision Energies.

Collision Peptide Protein Parent Ion Product y-ion Energy EQNGVQLTSSTLTNPR HA-DAT 872.944846 1487.781246 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 1373.738319 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 1316.716855 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 1217.648441 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 1089.589864 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 976.505799 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 875.458121 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 788.426093 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 701.394064 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 600.346386 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 487.262322 29.1 EQNGVQLTSSTLTNPR HA-DAT 872.944846 386.214643 29.1 GVTLPGAIDGIR HA-DAT 584.837862 1012.578571 20.5 GVTLPGAIDGIR HA-DAT 584.837862 911.530892 20.5 GVTLPGAIDGIR HA-DAT 584.837862 798.446828 20.5 GVTLPGAIDGIR HA-DAT 584.837862 701.394064 20.5 GVTLPGAIDGIR HA-DAT 584.837862 644.3726 20.5 GVTLPGAIDGIR HA-DAT 584.837862 573.335487 20.5 GVTLPGAIDGIR HA-DAT 584.837862 460.251423 20.5 GVTLPGAIDGIR HA-DAT 584.837862 345.22448 20.5 AYLSVDFYR HA-DAT 567.284931 899.462144 19.9 AYLSVDFYR HA-DAT 567.284931 786.37808 19.9 AYLSVDFYR HA-DAT 567.284931 699.346051 19.9 AYLSVDFYR HA-DAT 567.284931 600.277637 19.9 AYLSVDFYR HA-DAT 567.284931 485.250694 19.9 LAYAIAPEK HA-DAT 488.279117 791.429781 17.6 LAYAIAPEK HA-DAT 488.279117 628.366452 17.6 LAYAIAPEK HA-DAT 488.279117 557.329339 17.6 LAYAIAPEK HA-DAT 488.279117 444.245275 17.6 LAYAIAPEK HA-DAT 488.279117 373.208161 17.6 SAMPEGYVQER YFP 633.792794 1108.50917 21.9 SAMPEGYVQER YFP 633.792794 977.468686 21.9 SAMPEGYVQER YFP 633.792794 880.415922 21.9 SAMPEGYVQER YFP 633.792794 751.373329 21.9 SAMPEGYVQER YFP 633.792794 694.351865 21.9 SAMPEGYVQER YFP 633.792794 531.288536 21.9 SAMPEGYVQER YFP 633.792794 432.220122 21.9 FEGDTLVNR YFP 525.764363 774.410442 18.7 FEGDTLVNR YFP 525.764363 717.388979 18.7 FEGDTLVNR YFP 525.764363 602.362036 18.7 FEGDTLVNR YFP 525.764363 501.314357 18.7 FEGDTLVNR YFP 525.764363 388.230293 18.7

133 Table A4. Synthetic IgG Peptide SRM Method Transitions and Collision Energies.

Collision Peptide Protein Parent Ion Product y-ion Energy

Mouse Ig gamma-1 VNSAAFPAPIEK 622.337695 1030.556772 21.6

Mouse Ig gamma-1 VNSAAFPAPIEK 622.337695 943.524744 21.6

Mouse Ig gamma-1 VNSAAFPAPIEK 622.337695 872.48763 21.6

Mouse Ig gamma-1 VNSAAFPAPIEK 622.337695 801.450516 21.6

Mouse Ig gamma-1 VNSAAFPAPIEK 622.337695 654.382102 21.6

Mouse Ig gamma-1 VNSAAFPAPIEK 622.337695 486.292225 21.6

TSTSPIVK kappa light chain 416.742368 731.429781 15.4

TSTSPIVK kappa light chain 416.742368 644.397753 15.4

TSTSPIVK kappa light chain 416.742368 543.350074 15.4

TSTSPIVK kappa light chain 416.742368 456.318046 15.4 Rabbit IgG Heavy ALPAPIEK Chain 419.755278 654.382102 15.5 Rabbit IgG Heavy ALPAPIEK Chain 419.755278 557.329339 15.5 Rabbit IgG Heavy ALPAPIEK Chain 419.755278 486.292225 15.5

VTQGTTSVVQSFNR Ig-Kappa-chain 762.394261 936.489755 25.8

VTQGTTSVVQSFNR Ig-Kappa-chain 762.394261 849.457727 25.8

VTQGTTSVVQSFNR Ig-Kappa-chain 762.394261 750.389313 25.8

VTQGTTSVVQSFNR Ig-Kappa-chain 762.394261 651.320899 25.8

134 APPENDIX B

IGG NORMALIZATION CALCULATIONS AND STATISTICS

Table B1. Pre- and post-normalized area comparison of ΔN-DAT in αGFP IP. Original AUC values were compared to normalized areas by normalizing all values to their corresponding FL-DAT area value (pre- or post-normalization).

ΔN-DAT Average Average YFP-DAT Peptide IgG Peptide Pre- Post- S.D S.D. Normalized Normalized Area Area DTLMISR GVTLPGAIDGIR 0.2398 0.0293 0.1913 0.0234 DTLMISR AYLSVDFYR 0.6395 0.1261 0.5101 0.1006 DTLMISR SAMPEGYVQER 0.7180 0.0859 0.5727 0.0685 DTLMISR FEGDTLVNR 0.6594 0.2810 0.5259 0.2242 EQQFNSTIR GVTLPGAIDGIR 0.2398 0.0293 0.2189 0.0268 EQQFNSTIR AYLSVDFYR 0.6395 0.1261 0.5837 0.1151 EQQFNSTIR SAMPEGYVQER 0.7180 0.0859 0.6554 0.0784 EQQFNSTIR FEGDTLVNR 0.6594 0.2810 0.6018 0.2565 ALPAPIEK GVTLPGAIDGIR 0.2398 0.0293 0.1984 0.0243 ALPAPIEK AYLSVDFYR 0.6395 0.1261 0.5292 0.1044 ALPAPIEK SAMPEGYVQER 0.7180 0.0859 0.5942 0.0711 ALPAPIEK FEGDTLVNR 0.6594 0.2810 0.5457 0.2326 GQPLEPK GVTLPGAIDGIR 0.2398 0.0293 0.1892 0.0231 GQPLEPK AYLSVDFYR 0.6395 0.1261 0.5045 0.0995 GQPLEPK SAMPEGYVQER 0.7180 0.0859 0.5664 0.0678 GQPLEPK FEGDTLVNR 0.6594 0.2810 0.5202 0.2217 LSVPTSEWQR GVTLPGAIDGIR 0.2398 0.0293 0.1359 0.0166 LSVPTSEWQR AYLSVDFYR 0.6395 0.1261 0.3624 0.0715 LSVPTSEWQR SAMPEGYVQER 0.7180 0.0859 0.4069 0.0487 LSVPTSEWQR FEGDTLVNR 0.6594 0.2810 0.3737 0.1593 VTQGTTSVVQSFN GVTLPGAIDGIR R 0.2398 0.0293 0.2584 0.0316 VTQGTTSVVQSFN AYLSVDFYR R 0.6395 0.1261 0.6891 0.1359 VTQGTTSVVQSFN SAMPEGYVQER R 0.7180 0.0859 0.7737 0.0926 VTQGTTSVVQSFN FEGDTLVNR R 0.6594 0.2810 0.7105 0.3028

135 Table B2. Pre- and post-normalized area comparison of ΔC-DAT in αGFP IP. Original AUC values were compared to normalized areas by normalizing all values to their corresponding FL-DAT area value (pre- or post-normalization).

ΔC-DAT Average Average YFP-DAT Peptide IgG Peptide Pre- Post- S.D S.D. Normalized Normalized Area Area DTLMISR GVTLPGAIDGIR 0.1146 0.0229 0.1134 0.0227 DTLMISR AYLSVDFYR 0.1484 0.0306 0.1469 0.0303 DTLMISR SAMPEGYVQER 0.0969 0.0813 0.0958 0.0805 DTLMISR FEGDTLVNR 0.1335 0.0573 0.1321 0.0567 EQQFNSTIR GVTLPGAIDGIR 0.1146 0.0229 0.0964 0.0193 EQQFNSTIR AYLSVDFYR 0.1484 0.0306 0.1248 0.0258 EQQFNSTIR SAMPEGYVQER 0.0969 0.0813 0.0815 0.0684 EQQFNSTIR FEGDTLVNR 0.1335 0.0573 0.1123 0.0482 ALPAPIEK GVTLPGAIDGIR 0.1146 0.0229 0.1065 0.0213 ALPAPIEK AYLSVDFYR 0.1484 0.0306 0.1380 0.0285 ALPAPIEK SAMPEGYVQER 0.0969 0.0813 0.0900 0.0756 ALPAPIEK FEGDTLVNR 0.1335 0.0573 0.1241 0.0533 GQPLEPK GVTLPGAIDGIR 0.1146 0.0229 0.1248 0.0250 GQPLEPK AYLSVDFYR 0.1484 0.0306 0.1617 0.0334 GQPLEPK SAMPEGYVQER 0.0969 0.0813 0.1055 0.0886 GQPLEPK FEGDTLVNR 0.1335 0.0573 0.1455 0.0624 LSVPTSEWQR GVTLPGAIDGIR 0.1146 0.0229 0.0921 0.0184 LSVPTSEWQR AYLSVDFYR 0.1484 0.0306 0.1193 0.0246 LSVPTSEWQR SAMPEGYVQER 0.0969 0.0813 0.0779 0.0654 LSVPTSEWQR FEGDTLVNR 0.1335 0.0573 0.1073 0.0461 VTQGTTSVVQSFN GVTLPGAIDGIR R 0.1146 0.0229 0.1387 0.0278 VTQGTTSVVQSFN AYLSVDFYR R 0.1484 0.0306 0.1797 0.0371 VTQGTTSVVQSFN SAMPEGYVQER R 0.0969 0.0813 0.1172 0.0984 VTQGTTSVVQSFN FEGDTLVNR R 0.1335 0.0573 0.1616 0.0694

136 Table B3. Pre- and post-normalized area comparison of ΔN-DAT in αHA IP. Original AUC values were compared to normalized areas by normalizing all values to their corresponding FL-DAT area value (pre- or post-normalization).

ΔN Average Average YFP-DAT Peptide IgG Peptide Pre- Post- S.D S.D. Normalized Normalized Area Area DVLTITLTPK GVTLPGAIDGIR 0.5426 0.0667 0.4568 0.0562 DVLTITLTPK AYLSVDFYR 0.3153 0.0982 0.2654 0.0827 DVLTITLTPK SAMPEGYVQER 0.2892 0.1874 0.2435 0.1578 DVLTITLTPK FEGDTLVNR 0.4413 0.0447 0.3715 0.0376 VNSAAFPAPIEK GVTLPGAIDGIR 0.5426 0.0667 0.4221 0.0519 VNSAAFPAPIEK AYLSVDFYR 0.3153 0.0982 0.2452 0.0764 VNSAAFPAPIEK SAMPEGYVQER 0.2892 0.1874 0.2250 0.1458 VNSAAFPAPIEK FEGDTLVNR 0.4413 0.0447 0.3432 0.0348 APQVYTIPPPK GVTLPGAIDGIR 0.5426 0.0667 0.2038 0.0251 APQVYTIPPPK AYLSVDFYR 0.3153 0.0982 0.1184 0.0369 APQVYTIPPPK SAMPEGYVQER 0.2892 0.1874 0.1086 0.0704 APQVYTIPPPK FEGDTLVNR 0.4413 0.0447 0.1657 0.0168 QNGVLNSWTDQDS GVTLPGAIDGIR K 0.5426 0.0667 0.1909 0.0235 QNGVLNSWTDQDS AYLSVDFYR K 0.3153 0.0982 0.1109 0.0346 QNGVLNSWTDQDS SAMPEGYVQER K 0.2892 0.1874 0.1018 0.0659 QNGVLNSWTDQDS FEGDTLVNR K 0.4413 0.0447 0.1553 0.0157 DSTYSMSSTLTLTK GVTLPGAIDGIR 0.5426 0.0667 0.1583 0.0195 DSTYSMSSTLTLTK AYLSVDFYR 0.3153 0.0982 0.0920 0.0286 DSTYSMSSTLTLTK SAMPEGYVQER 0.2892 0.1874 0.0844 0.0547 DSTYSMSSTLTLTK FEGDTLVNR 0.4413 0.0447 0.1287 0.0130 TSTSPIVK GVTLPGAIDGIR 0.5426 0.0667 0.5530 0.0680 TSTSPIVK AYLSVDFYR 0.3153 0.0982 0.3213 0.1001 TSTSPIVK SAMPEGYVQER 0.2892 0.1874 0.2947 0.1910 TSTSPIVK FEGDTLVNR 0.4413 0.0447 0.4497 0.0456

137 Table B4. Pre- and post-normalized area comparison of ΔC-DAT in αHA IP. Original AUC values were compared to normalized areas by normalizing all values to their corresponding FL-DAT area value (pre- or post-normalization).

ΔC Average Average YFP-DAT Peptide IgG Peptide Pre- Post- S.D S.D. Normalize Normalized d Area Area DVLTITLTPK GVTLPGAIDGIR 0.2955 0.0512 0.1397 0.0242 DVLTITLTPK AYLSVDFYR 0.1481 0.0662 0.0700 0.0313 DVLTITLTPK SAMPEGYVQER 0.2644 0.0776 0.1250 0.0367 DVLTITLTPK FEGDTLVNR 0.2242 0.0149 0.1060 0.0070 VNSAAFPAPIEK GVTLPGAIDGIR 0.2955 0.0512 0.1407 0.0244 VNSAAFPAPIEK AYLSVDFYR 0.1481 0.0662 0.0706 0.0315 VNSAAFPAPIEK SAMPEGYVQER 0.2644 0.0776 0.1259 0.0369 VNSAAFPAPIEK FEGDTLVNR 0.2242 0.0149 0.1068 0.0071 APQVYTIPPPK GVTLPGAIDGIR 0.2955 0.0512 0.0785 0.0136 APQVYTIPPPK AYLSVDFYR 0.1481 0.0662 0.0393 0.0176 APQVYTIPPPK SAMPEGYVQER 0.2644 0.0776 0.0702 0.0206 APQVYTIPPPK FEGDTLVNR 0.2242 0.0149 0.0595 0.0040 QNGVLNSWTDQDS GVTLPGAIDGIR K 0.2955 0.0512 0.0681 0.0118 QNGVLNSWTDQDS AYLSVDFYR K 0.1481 0.0662 0.0341 0.0153 QNGVLNSWTDQDS SAMPEGYVQER K 0.2644 0.0776 0.0609 0.0179 QNGVLNSWTDQDS FEGDTLVNR K 0.2242 0.0149 0.0517 0.0034 DSTYSMSSTLTLTK GVTLPGAIDGIR 0.2955 0.0512 0.0458 0.0079 DSTYSMSSTLTLTK AYLSVDFYR 0.1481 0.0662 0.0230 0.0103 DSTYSMSSTLTLTK SAMPEGYVQER 0.2644 0.0776 0.0410 0.0120 DSTYSMSSTLTLTK FEGDTLVNR 0.2242 0.0149 0.0347 0.0023 TSTSPIVK GVTLPGAIDGIR 0.2955 0.0512 0.2168 0.0376 TSTSPIVK AYLSVDFYR 0.1481 0.0662 0.1087 0.0485 TSTSPIVK SAMPEGYVQER 0.2644 0.0776 0.1940 0.0569 TSTSPIVK FEGDTLVNR 0.2242 0.0149 0.1645 0.0109

138 Table B5. Statistical Analysis of IgG Normalization in Rabbit IgG IPs. ANOVA p- values were calculated using Microsoft Excel on the AUCs of each rabbit IgG peptide between the four cell line IP samples. The pre- and post-normalization AUC values for YFP-DAT peptides in ΔN and ΔC IP samples were compared via T-test.

ΔC ΔN Normaliz- YFP-DAT ANOVA IgG Peptide Normalization ation Peptide p-value T-test p-value T-test p- value DTLMISR GVTLPGAIDGIR 0.59201 0.0886 0.9523 DTLMISR AYLSVDFYR 0.59201 0.2371 0.9537 DTLMISR SAMPEGYVQER 0.59201 0.0838 0.9886 DTLMISR FEGDTLVNR 0.59201 0.5552 0.9777 EQQFNSTIR GVTLPGAIDGIR 0.44476 0.4131 0.3523 EQQFNSTIR AYLSVDFYR 0.44476 0.6017 0.3656 EQQFNSTIR SAMPEGYVQER 0.44476 0.4038 0.8144 EQQFNSTIR FEGDTLVNR 0.44476 0.8064 0.6495 ALPAPIEK GVTLPGAIDGIR 0.67344 0.1331 0.6783 ALPAPIEK AYLSVDFYR 0.67344 0.3082 0.6874 ALPAPIEK SAMPEGYVQER 0.67344 0.1268 0.9203 ALPAPIEK FEGDTLVNR 0.67344 0.6179 0.8452 GQPLEPK GVTLPGAIDGIR 0.01567 0.0787 0.628 GQPLEPK AYLSVDFYR 0.01567 0.2192 0.6382 GQPLEPK SAMPEGYVQER 0.01567 0.0744 0.9066 GQPLEPK FEGDTLVNR 0.01567 0.5375 0.819 LSVPTSEWQR GVTLPGAIDGIR 0.02686 0.0059 0.2569 LSVPTSEWQR AYLSVDFYR 0.02686 0.0297 0.2694 LSVPTSEWQR SAMPEGYVQER 0.02686 0.0055 0.7685 LSVPTSEWQR FEGDTLVNR 0.02686 0.2003 0.5711 VTQGTTSVVQSFNR GVTLPGAIDGIR 0.74126 0.4969 0.3105 VTQGTTSVVQSFNR AYLSVDFYR 0.74126 0.6676 0.3236 VTQGTTSVVQSFNR SAMPEGYVQER 0.74126 0.488 0.7958 VTQGTTSVVQSFNR FEGDTLVNR 0.74126 0.8409 0.6171

139 Table B6. Statistical Analysis of IgG Normalization in Mouse IgG IPs. ANOVA p- values were calculated using Microsoft Excel on the AUCs of each mouse IgG peptide between the four cell line IP samples. The pre- and post-normalization AUC values for YFP-DAT peptides in ΔN and ΔC IP samples were compared via T-test.

ΔN ΔC ANOVA IgG Peptide YFP-DAT Peptide Normalization Normalization p-value T-test p-value T-test p-value DVLTITLTPK GVTLPGAIDGIR 0.00522 0.1636 0.0089 DVLTITLTPK AYLSVDFYR 0.00522 0.5382 0.1384 DVLTITLTPK SAMPEGYVQER 0.00522 0.7627 0.0481 DVLTITLTPK FEGDTLVNR 0.00522 0.1076 0.0002 VNSAAFPAPIEK GVTLPGAIDGIR 0.00354 0.0689 0.0091 VNSAAFPAPIEK AYLSVDFYR 0.00354 0.3848 0.1407 VNSAAFPAPIEK SAMPEGYVQER 0.00354 0.6636 0.0492 VNSAAFPAPIEK FEGDTLVNR 0.00354 0.04 0.0002 APQVYTIPPPK GVTLPGAIDGIR 0.0035 0.0012 0.0021 APQVYTIPPPK AYLSVDFYR 0.0035 0.0314 0.0513 APQVYTIPPPK SAMPEGYVQER 0.0035 0.1932 0.0138 APQVYTIPPPK FEGDTLVNR 0.0035 0.0006 0.0001 QNGVLNSWTDQ DSK GVTLPGAIDGIR 0.01253 0.001 0.0017 QNGVLNSWTDQ DSK AYLSVDFYR 0.01253 0.0273 0.0438 QNGVLNSWTDQ DSK SAMPEGYVQER 0.01253 0.1775 0.0114 QNGVLNSWTDQ DSK FEGDTLVNR 0.01253 0.0005 <0.0001 DSTYSMSSTLTLT K GVTLPGAIDGIR 0.04947 0.0007 0.0011 DSTYSMSSTLTLT K AYLSVDFYR 0.04947 0.0194 0.0318 DSTYSMSSTLTLT K SAMPEGYVQER 0.04947 0.1432 0.0079 DSTYSMSSTLTLT K FEGDTLVNR 0.04947 0.0003 <0.0001 TSTSPIVK GVTLPGAIDGIR 0.42949 0.8599 0.0983 TSTSPIVK AYLSVDFYR 0.42949 0.9444 0.4518 TSTSPIVK SAMPEGYVQER 0.42949 0.9732 0.2735 TSTSPIVK FEGDTLVNR 0.42949 0.8307 0.005

140 APPENDIX C

MOUSE IP SRM TRANSITION INFORMATION

Table C1. DAT, DAT Interactome, and IgG Transitions and Collision Energies Product Collision Start End Peptide Protein Parent Ion y-ion Energy Time Time EVELILVK DAT 471.797143 714.476003 17.1 31.09 39.09 EVELILVK DAT 471.797143 585.43341 17.1 31.09 39.09 EVELILVK DAT 471.797143 472.349346 17.1 31.09 39.09 EVELILVK DAT 471.797143 359.265282 17.1 31.09 39.09 EVELILVK DAT 471.797143 246.181218 17.1 31.09 39.09 FC[+57.0]SLPGSFR DAT 535.758026 763.409714 19.0 24.78 32.78 FC[+57.0]SLPGSFR DAT 535.758026 676.377686 19.0 24.78 32.78 FC[+57.0]SLPGSFR DAT 535.758026 563.293622 19.0 24.78 32.78 FC[+57.0]SLPGSFR DAT 535.758026 466.240858 19.0 24.78 32.78 FC[+57.0]SLPGSFR DAT 535.758026 322.187366 19.0 24.78 32.78 LAYAITPEK DAT 503.2844 821.440346 18.0 20.39 28.39 LAYAITPEK DAT 503.2844 658.377017 18.0 20.39 28.39 LAYAITPEK DAT 503.2844 587.339903 18.0 20.39 28.39 LAYAITPEK DAT 503.2844 474.255839 18.0 20.39 28.39 LAYAITPEK DAT 503.2844 373.208161 18.0 20.39 28.39 AYLSVDFYR DAT 567.284931 899.462144 19.9 31.94 39.94 AYLSVDFYR DAT 567.284931 786.37808 19.9 31.94 39.94 AYLSVDFYR DAT 567.284931 699.346051 19.9 31.94 39.94 AYLSVDFYR DAT 567.284931 600.277637 19.9 31.94 39.94 AYLSVDFYR DAT 567.284931 485.250694 19.9 31.94 39.94 EGAAGVWK DAT 409.213777 631.356222 15.2 15.24 23.24 EGAAGVWK DAT 409.213777 560.319108 15.2 15.24 23.24 EGAAGVWK DAT 409.213777 489.281994 15.2 15.24 23.24 EGAAGVWK DAT 409.213777 432.260531 15.2 15.24 23.24 EGAAGVWK DAT 409.213777 333.192117 15.2 15.24 23.24 C[+57.0]SVGPMSSVVAPAK DAT 695.34688 1142.623806 23.8 31.08 39.08 C[+57.0]SVGPMSSVVAPAK DAT 695.34688 1043.555392 23.8 31.08 39.08 C[+57.0]SVGPMSSVVAPAK DAT 695.34688 986.533929 23.8 31.08 39.08 C[+57.0]SVGPMSSVVAPAK DAT 695.34688 386.239795 23.8 31.08 39.08 C[+57.0]SVGPMSSVVAPAK DAT 695.34688 315.202681 23.8 31.08 39.08 YNPNVLPVQC[+57.0]TGK Atp1b1 745.377026 1212.640519 25.3 22.93 30.93 YNPNVLPVQC[+57.0]TGK Atp1b1 745.377026 1001.544828 25.3 22.93 30.93 YNPNVLPVQC[+57.0]TGK Atp1b1 745.377026 902.476414 25.3 22.93 30.93 YNPNVLPVQC[+57.0]TGK Atp1b1 745.377026 789.39235 25.3 22.93 30.93 YNPNVLPVQC[+57.0]TGK Atp1b1 745.377026 593.271172 25.3 22.93 30.93 VAPPGLTQIPQIQK Atp1b1 745.440483 1222.715398 25.3 30.34 38.34 VAPPGLTQIPQIQK Atp1b1 745.440483 955.557107 25.3 30.34 38.34 VAPPGLTQIPQIQK Atp1b1 745.440483 854.509428 25.3 30.34 38.34 VAPPGLTQIPQIQK Atp1b1 745.440483 726.450851 25.3 30.34 38.34 VAPPGLTQIPQIQK Atp1b1 745.440483 613.366787 25.3 30.34 38.34 FYFENLLSK Camk2g 580.802957 1013.530223 20.3 42.17 50.17

141 FYFENLLSK Camk2g 580.802957 850.466895 20.3 42.17 50.17 FYFENLLSK Camk2g 580.802957 703.398481 20.3 42.17 50.17 FYFENLLSK Camk2g 580.802957 347.228896 20.3 42.17 50.17 FYFENLLSK Camk2g 580.802957 234.144832 20.3 42.17 50.17 GAILTTMLVSR Camk2g 581.336641 807.4393 20.3 43.61 51.61 GAILTTMLVSR Camk2g 581.336641 706.391621 20.3 43.61 51.61 GAILTTMLVSR Camk2g 581.336641 605.343943 20.3 43.61 51.61 GAILTTMLVSR Camk2g 581.336641 474.303458 20.3 43.61 51.61 GAILTTMLVSR Camk2g 581.336641 361.219394 20.3 43.61 51.61 EVC[+57.0]FAC[+57.0]VDGK Cltc 592.757366 796.365801 20.7 28.10 36.10 EVC[+57.0]FAC[+57.0]VDGK Cltc 592.757366 649.297387 20.7 28.10 36.10 EVC[+57.0]FAC[+57.0]VDGK Cltc 592.757366 578.260273 20.7 28.10 36.10 EVC[+57.0]FAC[+57.0]VDGK Cltc 592.757366 418.229624 20.7 28.10 36.10 EVC[+57.0]FAC[+57.0]VDGK Cltc 592.757366 319.161211 20.7 28.10 36.10 LLLPWLEAR Cltc 555.837134 884.498864 19.6 36.50 44.50 LLLPWLEAR Cltc 555.837134 771.4148 19.6 36.50 44.50 LLLPWLEAR Cltc 555.837134 674.362036 19.6 36.50 44.50 LLLPWLEAR Cltc 555.837134 488.282723 19.6 36.50 44.50 LLLPWLEAR Cltc 555.837134 375.198659 19.6 36.50 44.50 VSQPIEGHAASFAQFK Cltc 858.938836 1402.711376 28.7 42.87 50.87 VSQPIEGHAASFAQFK Cltc 858.938836 1192.574548 28.7 42.87 50.87 VSQPIEGHAASFAQFK Cltc 858.938836 1063.531955 28.7 42.87 50.87 VSQPIEGHAASFAQFK Cltc 858.938836 798.414465 28.7 42.87 50.87 VSQPIEGHAASFAQFK Cltc 858.938836 727.377351 28.7 42.87 50.87 DNFTLIPEGTNGTEER Dpysl2 896.921037 1202.564771 29.8 31.27 39.27 DNFTLIPEGTNGTEER Dpysl2 896.921037 1089.480707 29.8 31.27 39.27 DNFTLIPEGTNGTEER Dpysl2 896.921037 863.38535 29.8 31.27 39.27 DNFTLIPEGTNGTEER Dpysl2 896.921037 705.316208 29.8 31.27 39.27 DNFTLIPEGTNGTEER Dpysl2 896.921037 591.27328 29.8 31.27 39.27 GIQEEMEALVK Dpysl2 623.82102 948.47066 21.6 35.80 43.80 GIQEEMEALVK Dpysl2 623.82102 819.428066 21.6 35.80 43.80 GIQEEMEALVK Dpysl2 623.82102 690.385473 21.6 35.80 43.80 GIQEEMEALVK Dpysl2 623.82102 559.344989 21.6 35.80 43.80 GIQEEMEALVK Dpysl2 623.82102 430.302396 21.6 35.80 43.80 IVLEDGTLHVTEGSGR Dpysl2 841.939033 1228.591654 28.2 20.71 28.71 IVLEDGTLHVTEGSGR Dpysl2 841.939033 1113.564711 28.2 20.71 28.71 IVLEDGTLHVTEGSGR Dpysl2 841.939033 705.352593 28.2 20.71 28.71 IVLEDGTLHVTEGSGR Dpysl2 841.939033 606.284179 28.2 20.71 28.71 IVLEDGTLHVTEGSGR Dpysl2 841.939033 376.193908 28.2 20.71 28.71 ILDLGITGPEGH Dpysl2 611.327327 995.47925 21.2 30.30 38.30 ILDLGITGPEGH Dpysl2 611.327327 767.368243 21.2 30.30 38.30 ILDLGITGPEGH Dpysl2 611.327327 710.34678 21.2 30.30 38.30 ILDLGITGPEGH Dpysl2 611.327327 597.262716 21.2 30.30 38.30 ILDLGITGPEGH Dpysl2 611.327327 496.215037 21.2 30.30 38.30 ISVGSDADLVIWDPDSVK Dpysl2 958.486013 1171.635751 31.7 40.09 48.09 ISVGSDADLVIWDPDSVK Dpysl2 958.486013 1058.551687 31.7 40.09 48.09 ISVGSDADLVIWDPDSVK Dpysl2 958.486013 846.399209 31.7 40.09 48.09 ISVGSDADLVIWDPDSVK Dpysl2 958.486013 660.319896 31.7 40.09 48.09

142 ISVGSDADLVIWDPDSVK Dpysl2 958.486013 545.292953 31.7 40.09 48.09 AVVYSNTIQSIIAIIR Gnai1 881.017086 1328.789626 29.3 34.12 42.12 AVVYSNTIQSIIAIIR Gnai1 881.017086 913.582928 29.3 34.12 42.12 AVVYSNTIQSIIAIIR Gnai1 881.017086 785.52435 29.3 34.12 42.12 AVVYSNTIQSIIAIIR Gnai1 881.017086 472.324194 29.3 34.12 42.12 AVVYSNTIQSIIAIIR Gnai1 881.017086 401.28708 29.3 34.12 42.12 YYLDSLDR Gnao1 522.753464 881.436323 18.6 24.45 32.45 YYLDSLDR Gnao1 522.753464 718.372994 18.6 24.45 32.45 YYLDSLDR Gnao1 522.753464 605.28893 18.6 24.45 32.45 YYLDSLDR Gnao1 522.753464 490.261987 18.6 24.45 32.45 YYLDSLDR Gnao1 522.753464 403.229959 18.6 24.45 32.45 LWGDSGIQEC[+57.0]FNR Gnao1 791.359364 1282.548075 26.6 30.55 38.55 LWGDSGIQEC[+57.0]FNR Gnao1 791.359364 1110.499669 26.6 30.55 38.55 LWGDSGIQEC[+57.0]FNR Gnao1 791.359364 853.362112 26.6 30.55 38.55 LWGDSGIQEC[+57.0]FNR Gnao1 791.359364 725.303535 26.6 30.55 38.55 LWGDSGIQEC[+57.0]FNR Gnao1 791.359364 596.260942 26.6 30.55 38.55 MEDTEPFSAELLSAMMR Gnao1 979.438836 1352.670105 32.3 39.61 47.61 MEDTEPFSAELLSAMMR Gnao1 979.438836 1108.548927 32.3 39.61 47.61 MEDTEPFSAELLSAMMR Gnao1 979.438836 950.479784 32.3 39.61 47.61 MEDTEPFSAELLSAMMR Gnao1 979.438836 708.353127 32.3 39.61 47.61 MEDTEPFSAELLSAMMR Gnao1 979.438836 595.269063 32.3 39.61 47.61 IGAGDYQPTEQDILR Gnao1 838.417933 1262.637542 28.1 28.00 36.00 IGAGDYQPTEQDILR Gnao1 838.417933 1099.574213 28.1 28.00 36.00 IGAGDYQPTEQDILR Gnao1 838.417933 971.515636 28.1 28.00 36.00 IGAGDYQPTEQDILR Gnao1 838.417933 773.415194 28.1 28.00 36.00 IGAGDYQPTEQDILR Gnao1 838.417933 644.3726 28.1 28.00 36.00 LDAQVQELHAK Myh10 626.338226 824.462478 21.7 43.67 51.67 LDAQVQELHAK Myh10 626.338226 725.394064 21.7 43.67 51.67 LDAQVQELHAK Myh10 626.338226 597.335487 21.7 43.67 51.67 LDAQVQELHAK Myh10 626.338226 468.292893 21.7 43.67 51.67 LDAQVQELHAK Myh10 626.338226 355.208829 21.7 43.67 51.67 NSLQEQQEEEEEAR Myh10 859.876826 1276.528779 28.7 42.87 50.87 NSLQEQQEEEEEAR Myh10 859.876826 1147.486186 28.7 42.87 50.87 NSLQEQQEEEEEAR Myh10 859.876826 1019.427609 28.7 42.87 50.87 NSLQEQQEEEEEAR Myh10 859.876826 891.369031 28.7 42.87 50.87 NSLQEQQEEEEEAR Myh10 859.876826 762.326438 28.7 42.87 50.87 ADMEDLMSSK Myh10 563.741381 940.41143 19.8 20.52 28.52 ADMEDLMSSK Myh10 563.741381 809.370945 19.8 20.52 28.52 ADMEDLMSSK Myh10 563.741381 680.328352 19.8 20.52 28.52 ADMEDLMSSK Myh10 563.741381 452.217345 19.8 20.52 28.52 ADMEDLMSSK Myh10 563.741381 321.176861 19.8 20.52 28.52 LEVNMQAMK Myh10 532.267369 821.400806 18.9 24.46 32.46 LEVNMQAMK Myh10 532.267369 722.332392 18.9 24.46 32.46 LEVNMQAMK Myh10 532.267369 608.289464 18.9 24.46 32.46 LEVNMQAMK Myh10 532.267369 477.24898 18.9 24.46 32.46 LEVNMQAMK Myh10 532.267369 349.190402 18.9 24.46 32.46 AEEDSMEDPYELK Pclo 778.327061 1111.497603 26.3 20.90 28.90 AEEDSMEDPYELK Pclo 778.327061 1024.465574 26.3 20.90 28.90

143 AEEDSMEDPYELK Pclo 778.327061 893.42509 26.3 20.90 28.90 AEEDSMEDPYELK Pclo 778.327061 764.382496 26.3 20.90 28.90 AEEDSMEDPYELK Pclo 778.327061 649.355553 26.3 20.90 28.90 ISLPETGLAPTPSSQTK Pclo 863.964716 1187.626643 28.8 43.75 51.75 ISLPETGLAPTPSSQTK Pclo 863.964716 1086.578964 28.8 43.75 51.75 ISLPETGLAPTPSSQTK Pclo 863.964716 916.473437 28.8 43.75 51.75 ISLPETGLAPTPSSQTK Pclo 863.964716 845.436323 28.8 43.75 51.75 ISLPETGLAPTPSSQTK Pclo 863.964716 647.335881 28.8 43.75 51.75 ALGGELAAIPSSPQPTPK Pclo 867.475251 1193.652464 28.9 42.87 50.87 ALGGELAAIPSSPQPTPK Pclo 867.475251 1122.61535 28.9 42.87 50.87 ALGGELAAIPSSPQPTPK Pclo 867.475251 1051.578236 28.9 42.87 50.87 ALGGELAAIPSSPQPTPK Pclo 867.475251 938.494172 28.9 42.87 50.87 ALGGELAAIPSSPQPTPK Pclo 867.475251 667.377351 28.9 42.87 50.87 FAEELEWER Pclo 604.782753 990.452701 21.0 19.95 27.95 FAEELEWER Pclo 604.782753 861.410108 21.0 19.95 27.95 FAEELEWER Pclo 604.782753 732.367515 21.0 19.95 27.95 FAEELEWER Pclo 604.782753 619.283451 21.0 19.95 27.95 FAEELEWER Pclo 604.782753 490.240858 21.0 19.95 27.95 QALNEIETR Stx1b 537.28292 761.378808 19.0 24.48 32.48 QALNEIETR Stx1b 537.28292 647.335881 19.0 24.48 32.48 QALNEIETR Stx1b 537.28292 518.293287 19.0 24.48 32.48 QALNEIETR Stx1b 537.28292 405.209223 19.0 24.48 32.48 QALNEIETR Stx1b 537.28292 276.16663 19.0 24.48 32.48 DNALLAQLIQDK Stxbp1 671.372266 928.546208 23.0 39.60 47.60 DNALLAQLIQDK Stxbp1 671.372266 815.462144 23.0 39.60 47.60 DNALLAQLIQDK Stxbp1 671.372266 744.42503 23.0 39.60 47.60 DNALLAQLIQDK Stxbp1 671.372266 503.282388 23.0 39.60 47.60 DNALLAQLIQDK Stxbp1 671.372266 390.198324 23.0 39.60 47.60 HYQGTVDK Stxbp1 474.232698 810.399209 17.1 27.50 35.50 HYQGTVDK Stxbp1 474.232698 647.335881 17.1 27.50 35.50 HYQGTVDK Stxbp1 474.232698 519.277303 17.1 27.50 35.50 HYQGTVDK Stxbp1 474.232698 462.255839 17.1 27.50 35.50 HYQGTVDK Stxbp1 474.232698 262.139747 17.1 27.50 35.50 VLVVDQLSMR Stxbp1 580.328816 947.497877 20.3 30.84 38.84 VLVVDQLSMR Stxbp1 580.328816 848.429463 20.3 30.84 38.84 VLVVDQLSMR Stxbp1 580.328816 749.361049 20.3 30.84 38.84 VLVVDQLSMR Stxbp1 580.328816 634.334106 20.3 30.84 38.84 VLVVDQLSMR Stxbp1 580.328816 393.191465 20.3 30.84 38.84 EPLPSLEAVYLITPSEK Stxbp1 943.511499 1249.667445 31.2 42.66 50.66 EPLPSLEAVYLITPSEK Stxbp1 943.511499 1120.624852 31.2 42.66 50.66 EPLPSLEAVYLITPSEK Stxbp1 943.511499 950.519324 31.2 42.66 50.66 EPLPSLEAVYLITPSEK Stxbp1 943.511499 561.287868 31.2 42.66 50.66 EPLPSLEAVYLITPSEK Stxbp1 943.511499 460.240189 31.2 42.66 50.66 AGGAFDPYTLVR Syngr1 633.827494 1010.530558 21.9 28.54 36.54 AGGAFDPYTLVR Syngr1 633.827494 863.462144 21.9 28.54 36.54 AGGAFDPYTLVR Syngr1 633.827494 748.435201 21.9 28.54 36.54 AGGAFDPYTLVR Syngr1 633.827494 488.319108 21.9 28.54 36.54 AGGAFDPYTLVR Syngr1 633.827494 387.27143 21.9 28.54 36.54

144 DNPLNEGTDAAR Syngr1 636.79438 1043.511613 22.0 32.83 40.83 DNPLNEGTDAAR Syngr1 636.79438 833.374785 22.0 32.83 40.83 DNPLNEGTDAAR Syngr1 636.79438 719.331858 22.0 32.83 40.83 DNPLNEGTDAAR Syngr1 636.79438 590.289265 22.0 32.83 40.83 DNPLNEGTDAAR Syngr1 636.79438 432.220122 22.0 32.83 40.83 DVLTITLTPK IgG HC 550.831714 886.560795 19.4 33.96 41.96 DVLTITLTPK IgG HC 550.831714 773.476731 19.4 33.96 41.96 DVLTITLTPK IgG HC 550.831714 672.429053 19.4 33.96 41.96 DVLTITLTPK IgG HC 550.831714 559.344989 19.4 33.96 41.96 VNSAAFPAPIEK IgG HC 622.337695 1030.556772 21.6 23.39 31.39 VNSAAFPAPIEK IgG HC 622.337695 872.48763 21.6 23.39 31.39 VNSAAFPAPIEK IgG HC 622.337695 801.450516 21.6 23.39 31.39 VNSAAFPAPIEK IgG HC 622.337695 654.382102 21.6 23.39 31.39 APQVYTIPPPK IgG HC 605.845156 914.53458 21.1 21.17 29.17 APQVYTIPPPK IgG HC 605.845156 815.466166 21.1 21.17 29.17 APQVYTIPPPK IgG HC 605.845156 438.271095 21.1 21.17 29.17 APQVYTIPPPK IgG HC 605.845156 341.218332 21.1 21.17 29.17 QNGVLNSWTDQDSK IgK LC 796.370983 1193.543307 26.8 23.62 31.62 QNGVLNSWTDQDSK IgK LC 796.370983 1080.459243 26.8 23.62 31.62 QNGVLNSWTDQDSK IgK LC 796.370983 966.416316 26.8 23.62 31.62 QNGVLNSWTDQDSK IgK LC 796.370983 879.384287 26.8 23.62 31.62 DSTYSMSSTLTLTK IgK LC 767.868896 1231.623866 25.9 29.47 37.47 DSTYSMSSTLTLTK IgK LC 767.868896 1068.560537 25.9 29.47 37.47 DSTYSMSSTLTLTK IgK LC 767.868896 981.528509 25.9 29.47 37.47 DSTYSMSSTLTLTK IgK LC 767.868896 850.488024 25.9 29.47 37.47 TSTSPIVK IgK LC 416.742368 731.429781 15.4 11.20 19.20 TSTSPIVK IgK LC 416.742368 644.397753 15.4 11.20 19.20 TSTSPIVK IgK LC 416.742368 543.350074 15.4 11.20 19.20 TSTSPIVK IgK LC 416.742368 456.318046 15.4 11.20 19.20

145 APPENDIX D

MOUSE IP SUM NORMALIZATION

Table D1. Pre- and Post- Normalized Areas. This extensive table displays the pre- and post-normalization areas and associated standard deviations for targeted DAT, putative DAT interactors, and IgG peptides in all 12 time course samples. Numbers 1-6 indicate sample treatment where 1-3 represent the points on the PMA time course and 4-6 represent those on the Amph time course.

Average Average Pre- Post- Sample Protein Peptide Normalization SD Normali SD Area zation Area HA1 Atp1b1 VAPPGLTQIPQIQK 31463.33 9237.31 41833.25 12281.81 HA2 Atp1b1 VAPPGLTQIPQIQK 41596.33 9993.02 43006.74 10331.86 HA3 Atp1b1 VAPPGLTQIPQIQK 37343.33 6828.94 37343.33 6828.94 HA4 Atp1b1 VAPPGLTQIPQIQK 48991.00 1957.29 49694.76 1985.40 HA5 Atp1b1 VAPPGLTQIPQIQK 43031.00 11273.59 44434.18 11641.21 HA6 Atp1b1 VAPPGLTQIPQIQK 44538.33 11090.45 47689.91 11875.22 WT1 Atp1b1 VAPPGLTQIPQIQK 14258.33 5918.30 18813.21 7808.93 WT2 Atp1b1 VAPPGLTQIPQIQK 11050.33 1723.06 15936.82 2485.00 WT3 Atp1b1 VAPPGLTQIPQIQK 12339.00 3053.50 18525.21 4584.38 WT4 Atp1b1 VAPPGLTQIPQIQK 14647.33 5033.85 18009.93 6189.48 WT5 Atp1b1 VAPPGLTQIPQIQK 15708.67 2678.83 17491.48 2982.86 WT6 Atp1b1 VAPPGLTQIPQIQK 12859.00 5259.20 20051.84 8200.99 HA1 Atp1b1 YNPNVLPVQCTGK 63755.33 13385.08 84768.29 17796.63 HA2 Atp1b1 YNPNVLPVQCTGK 88998.00 13012.66 92015.66 13453.88 HA3 Atp1b1 YNPNVLPVQCTGK 86563.33 14458.02 86563.33 14458.02 HA4 Atp1b1 YNPNVLPVQCTGK 100954.33 17479.48 102404.55 17730.57 HA5 Atp1b1 YNPNVLPVQCTGK 87102.00 21915.91 89942.26 22630.56 HA6 Atp1b1 YNPNVLPVQCTGK 111144.67 29352.11 119009.38 31429.10 WT1 Atp1b1 YNPNVLPVQCTGK 23011.67 3886.75 30362.82 5128.38 WT2 Atp1b1 YNPNVLPVQCTGK 21773.33 3630.85 31401.56 5236.42 WT3 Atp1b1 YNPNVLPVQCTGK 24183.33 5655.02 36307.76 8490.19 WT4 Atp1b1 YNPNVLPVQCTGK 36874.00 3191.75 45339.19 3924.48 WT5 Atp1b1 YNPNVLPVQCTGK 33955.67 4596.95 37809.37 5118.67 WT6 Atp1b1 YNPNVLPVQCTGK 22913.33 4006.75 35730.18 6247.98 HA1 Cltc EVCFACVDGK 114769.33 14153.63 152595.87 18818.49 HA2 Cltc EVCFACVDGK 135100.67 8576.59 139681.54 8867.40 HA3 Cltc EVCFACVDGK 140915.67 20602.22 140915.67 20602.22 HA4 Cltc EVCFACVDGK 148578.67 5719.04 150713.01 5801.20 HA5 Cltc EVCFACVDGK 148174.33 48155.97 153006.07 49726.26

146 HA6 Cltc EVCFACVDGK 105946.67 35171.88 113443.56 37660.68 WT1 Cltc EVCFACVDGK 93689.33 12219.25 123618.71 16122.73 WT2 Cltc EVCFACVDGK 94268.67 9289.39 135954.51 13397.19 WT3 Cltc EVCFACVDGK 89103.00 6163.96 133775.20 9254.29 WT4 Cltc EVCFACVDGK 131681.33 24945.06 161911.50 30671.72 WT5 Cltc EVCFACVDGK 110457.33 6205.28 122993.39 6909.54 WT6 Cltc EVCFACVDGK 117577.33 20014.51 183345.63 31209.87 116204.8 HA1 DAT AYLSVDFYR 660226.33 1 877828.66 154504.46 249222.0 HA2 DAT AYLSVDFYR 759398.00 2 785146.98 257672.41 HA3 DAT AYLSVDFYR 859035.33 73705.63 859035.33 73705.63 136652.5 HA4 DAT AYLSVDFYR 775434.67 8 786573.85 138615.61 121266.0 HA5 DAT AYLSVDFYR 680777.00 6 702976.10 125220.36 200851.2 HA6 DAT AYLSVDFYR 838412.67 0 897739.62 215063.64 WT1 DAT AYLSVDFYR 1375.00 551.23 1814.25 727.32 WT2 DAT AYLSVDFYR 2554.67 628.99 3684.35 907.13 WT3 DAT AYLSVDFYR 2320.67 1024.29 3484.14 1537.82 WT4 DAT AYLSVDFYR 2551.00 1349.50 3136.63 1659.31 WT5 DAT AYLSVDFYR 3075.67 388.98 3424.73 433.13 WT6 DAT AYLSVDFYR 2455.67 528.58 3829.27 824.24 HA1 DAT EGAAGVWK 79933.00 9062.41 106277.91 12049.27 HA2 DAT EGAAGVWK 66561.00 7320.41 68817.89 7568.62 HA3 DAT EGAAGVWK 125109.33 13505.77 125109.33 13505.77 HA4 DAT EGAAGVWK 59873.67 8915.18 60733.76 9043.25 HA5 DAT EGAAGVWK 162393.00 20379.03 167688.39 21043.56 HA6 DAT EGAAGVWK 84717.33 7520.63 90712.02 8052.79 WT1 DAT EGAAGVWK 767.33 612.31 1012.46 807.91 WT2 DAT EGAAGVWK 1114.00 332.81 1606.61 479.98 WT3 DAT EGAAGVWK 1875.33 434.08 2815.54 651.70 WT4 DAT EGAAGVWK 1550.33 773.28 1906.24 950.80 WT5 DAT EGAAGVWK 664.67 180.31 740.10 200.77 WT6 DAT EGAAGVWK 672.67 358.01 1048.93 558.27 699847.2 HA1 DAT EVELILVK 2151853.33 6 2861077.22 930508.14 966320.1 HA2 DAT EVELILVK 2355814.00 7 2435692.80 999085.27 518458.6 HA3 DAT EVELILVK 2559836.33 8 2559836.33 518458.68 625839.9 HA4 DAT EVELILVK 2504106.00 9 2540077.68 634830.23 974257.4 1006026.5 HA5 DAT EVELILVK 2140696.33 8 2210501.19 4 1157386. 1239284.2 HA6 DAT EVELILVK 2437414.00 42 2609887.96 9 WT1 DAT EVELILVK 24452.67 26546.74 32264.15 35027.18 WT2 DAT EVELILVK 9135.33 3960.61 13175.00 5712.00

147 WT3 DAT EVELILVK 10766.00 5462.22 16163.58 8200.73 WT4 DAT EVELILVK 7522.33 588.36 9249.24 723.43 WT5 DAT EVELILVK 11810.67 2386.77 13151.09 2657.65 WT6 DAT EVELILVK 8367.00 3141.11 13047.18 4898.12 HA1 DAT FCSLPGSFR 177639.33 45811.60 236187.03 60910.52 111168.7 HA2 DAT FCSLPGSFR 413543.00 3 427565.04 114938.14 HA3 DAT FCSLPGSFR 402405.00 72631.86 402405.00 72631.86 HA4 DAT FCSLPGSFR 249924.33 58902.81 253514.52 59748.96 HA5 DAT FCSLPGSFR 211958.67 40219.36 218870.32 41530.85 HA6 DAT FCSLPGSFR 273289.67 76654.96 292627.93 82079.14 WT1 DAT FCSLPGSFR 2292.33 1836.11 3024.63 2422.66 WT2 DAT FCSLPGSFR 5752.00 4246.45 8295.55 6124.24 WT3 DAT FCSLPGSFR 5034.67 773.32 7558.82 1161.03 WT4 DAT FCSLPGSFR 4795.00 1711.44 5895.79 2104.33 WT5 DAT FCSLPGSFR 4087.67 5243.22 4551.59 5838.29 WT6 DAT FCSLPGSFR 6264.33 4106.43 9768.36 6403.41 HA1 DAT LAYAITPEK 957489.67 66678.60 1273066.26 88655.03 128287.6 HA2 DAT LAYAITPEK 1314840.00 2 1359422.40 132637.48 HA3 DAT LAYAITPEK 1096101.67 59096.02 1096101.67 59096.02 HA4 DAT LAYAITPEK 1061824.00 44679.03 1077077.19 45320.85 166601.7 HA5 DAT LAYAITPEK 1158058.33 9 1195820.85 172034.43 130821.7 HA6 DAT LAYAITPEK 1223444.67 4 1310016.89 140078.82 WT1 DAT LAYAITPEK 3654.33 869.27 4821.72 1146.96 WT2 DAT LAYAITPEK 2745.67 1505.88 3959.81 2171.79 WT3 DAT LAYAITPEK 3497.33 567.83 5250.74 852.52 WT4 DAT LAYAITPEK 5894.00 2370.22 7247.09 2914.35 WT5 DAT LAYAITPEK 3597.67 989.35 4005.97 1101.64 WT6 DAT LAYAITPEK 1985.33 1025.27 3095.85 1598.77 HA1 Dpysl2 DNFTLIPEGTNGTEER 40287.33 2472.86 53565.53 3287.88 HA2 Dpysl2 DNFTLIPEGTNGTEER 57359.67 10877.32 59304.57 11246.13 HA3 Dpysl2 DNFTLIPEGTNGTEER 49013.33 9736.12 49013.33 9736.12 HA4 Dpysl2 DNFTLIPEGTNGTEER 58293.00 5575.50 59130.38 5655.60 HA5 Dpysl2 DNFTLIPEGTNGTEER 53574.33 11212.08 55321.31 11577.69 HA6 Dpysl2 DNFTLIPEGTNGTEER 64659.00 12698.07 69234.34 13596.60 WT1 Dpysl2 DNFTLIPEGTNGTEER 22206.33 2525.28 29300.22 3331.98 WT2 Dpysl2 DNFTLIPEGTNGTEER 19106.33 4984.50 27555.20 7188.65 WT3 Dpysl2 DNFTLIPEGTNGTEER 25771.33 2366.23 38691.91 3552.56 WT4 Dpysl2 DNFTLIPEGTNGTEER 25944.33 4063.37 31900.39 4996.19 WT5 Dpysl2 DNFTLIPEGTNGTEER 25301.00 4160.54 28172.47 4632.73 WT6 Dpysl2 DNFTLIPEGTNGTEER 25112.00 7538.55 39158.70 11755.33

148 HA1 Dpysl2 GIQEEMEALVK 28067.33 3171.51 37317.97 4216.80 HA2 Dpysl2 GIQEEMEALVK 26799.33 5130.33 27708.02 5304.29 HA3 Dpysl2 GIQEEMEALVK 37433.67 8730.98 37433.67 8730.98 HA4 Dpysl2 GIQEEMEALVK 36947.33 753.96 37478.08 764.79 HA5 Dpysl2 GIQEEMEALVK 33471.00 7184.62 34562.44 7418.90 HA6 Dpysl2 GIQEEMEALVK 34950.33 9636.30 37423.46 10318.17 WT1 Dpysl2 GIQEEMEALVK 9102.67 3501.40 12010.54 4619.93 WT2 Dpysl2 GIQEEMEALVK 11711.67 5449.15 16890.60 7858.78 WT3 Dpysl2 GIQEEMEALVK 10702.00 6723.00 16067.50 10093.60 WT4 Dpysl2 GIQEEMEALVK 15745.67 4433.58 19360.41 5451.40 WT5 Dpysl2 GIQEEMEALVK 18364.00 3102.33 20448.17 3454.42 WT6 Dpysl2 GIQEEMEALVK 19526.00 1285.82 30448.10 2005.06 HA1 Dpysl2 ISVGSDADLVIWDPDSVK 20878.00 5646.34 27759.13 7507.30 HA2 Dpysl2 ISVGSDADLVIWDPDSVK 30537.67 2252.40 31573.11 2328.77 HA3 Dpysl2 ISVGSDADLVIWDPDSVK 34025.67 8469.16 34025.67 8469.16 HA4 Dpysl2 ISVGSDADLVIWDPDSVK 42662.00 8128.88 43274.84 8245.65 HA5 Dpysl2 ISVGSDADLVIWDPDSVK 30717.67 5355.89 31719.32 5530.54 HA6 Dpysl2 ISVGSDADLVIWDPDSVK 32979.00 4599.58 35312.63 4925.05 WT1 Dpysl2 ISVGSDADLVIWDPDSVK 11178.33 1984.72 14749.29 2618.74 WT2 Dpysl2 ISVGSDADLVIWDPDSVK 12304.33 2804.88 17745.34 4045.21 WT3 Dpysl2 ISVGSDADLVIWDPDSVK 10788.33 1662.44 16197.11 2495.92 WT4 Dpysl2 ISVGSDADLVIWDPDSVK 14369.00 1146.27 17667.70 1409.42 WT5 Dpysl2 ISVGSDADLVIWDPDSVK 10992.67 1755.81 12240.25 1955.08 WT6 Dpysl2 ISVGSDADLVIWDPDSVK 18721.33 2180.85 29193.34 3400.73 HA1 Dpysl2 IVLEDGTLHVTEGSGR 7208.33 1274.97 9584.11 1695.19 HA2 Dpysl2 IVLEDGTLHVTEGSGR 7983.00 1527.44 8253.68 1579.23 HA3 Dpysl2 IVLEDGTLHVTEGSGR 7299.67 1603.26 7299.67 1603.26 HA4 Dpysl2 IVLEDGTLHVTEGSGR 9228.33 1112.72 9360.90 1128.70 HA5 Dpysl2 IVLEDGTLHVTEGSGR 10349.00 1395.95 10686.47 1441.47 HA6 Dpysl2 IVLEDGTLHVTEGSGR 8920.33 1573.99 9551.55 1685.37 WT1 Dpysl2 IVLEDGTLHVTEGSGR 4402.00 676.79 5808.23 892.99 WT2 Dpysl2 IVLEDGTLHVTEGSGR 3034.00 1520.24 4375.64 2192.50 WT3 Dpysl2 IVLEDGTLHVTEGSGR 4101.00 794.87 6157.06 1193.38 WT4 Dpysl2 IVLEDGTLHVTEGSGR 5762.67 794.21 7085.61 976.53 WT5 Dpysl2 IVLEDGTLHVTEGSGR 5675.67 259.25 6319.81 288.67 WT6 Dpysl2 IVLEDGTLHVTEGSGR 4841.00 386.43 7548.87 602.59 HA1 Gnao1 IGAGDYQPTEQDILR 62879.00 12828.60 83603.13 17056.74 HA2 Gnao1 IGAGDYQPTEQDILR 75632.00 16370.84 78196.46 16925.92 HA3 Gnao1 IGAGDYQPTEQDILR 76417.33 5082.05 76417.33 5082.05 HA4 Gnao1 IGAGDYQPTEQDILR 85123.67 12200.08 86346.47 12375.34 HA5 Gnao1 IGAGDYQPTEQDILR 80478.00 17383.87 83102.27 17950.73

149 HA6 Gnao1 IGAGDYQPTEQDILR 81819.67 21306.53 87609.31 22814.20 WT1 Gnao1 IGAGDYQPTEQDILR 19151.00 2979.81 25268.85 3931.72 WT2 Gnao1 IGAGDYQPTEQDILR 21929.00 6561.65 31626.06 9463.22 WT3 Gnao1 IGAGDYQPTEQDILR 19884.00 3552.05 29852.93 5332.88 WT4 Gnao1 IGAGDYQPTEQDILR 28594.33 4816.90 35158.75 5922.72 WT5 Gnao1 IGAGDYQPTEQDILR 23127.00 2771.89 25751.74 3086.48 WT6 Gnao1 IGAGDYQPTEQDILR 38729.67 5334.70 60393.57 8318.72 HA1 Gnao1 LWGDSGIQECFNR 9109.33 2021.24 12111.66 2687.42 HA2 Gnao1 LWGDSGIQECFNR 30785.00 9266.58 31828.83 9580.78 HA3 Gnao1 LWGDSGIQECFNR 19033.00 3135.05 19033.00 3135.05 HA4 Gnao1 LWGDSGIQECFNR 16076.33 1950.74 16307.27 1978.76 HA5 Gnao1 LWGDSGIQECFNR 13912.33 3447.58 14365.99 3560.00 HA6 Gnao1 LWGDSGIQECFNR 18007.00 4453.08 19281.19 4768.19 WT1 Gnao1 LWGDSGIQECFNR 2409.33 1037.71 3179.00 1369.21 WT2 Gnao1 LWGDSGIQECFNR 3767.67 1426.22 5433.74 2056.90 WT3 Gnao1 LWGDSGIQECFNR 1609.33 357.92 2416.18 537.36 WT4 Gnao1 LWGDSGIQECFNR 7185.33 675.84 8834.87 831.00 WT5 Gnao1 LWGDSGIQECFNR 6325.00 1804.83 7042.84 2009.66 WT6 Gnao1 LWGDSGIQECFNR 4123.33 3070.81 6429.77 4788.50 HA1 Gnao1 YYLDSLDR 69641.00 3493.52 92593.80 4644.93 HA2 Gnao1 YYLDSLDR 80188.00 15011.21 82906.94 15520.20 HA3 Gnao1 YYLDSLDR 88067.33 11723.62 88067.33 11723.62 HA4 Gnao1 YYLDSLDR 90108.67 10617.52 91403.08 10770.04 HA5 Gnao1 YYLDSLDR 89249.00 10393.75 92159.27 10732.68 HA6 Gnao1 YYLDSLDR 105479.33 7352.09 112943.16 7872.33 WT1 Gnao1 YYLDSLDR 22488.00 996.22 29671.87 1314.46 WT2 Gnao1 YYLDSLDR 22284.67 3318.79 32139.00 4786.37 WT3 Gnao1 YYLDSLDR 24425.67 2248.19 36671.59 3375.33 WT4 Gnao1 YYLDSLDR 30011.67 6019.11 36901.46 7400.92 WT5 Gnao1 YYLDSLDR 29559.33 1517.04 32914.09 1689.21 WT6 Gnao1 YYLDSLDR 31440.67 3953.30 49027.38 6164.62 6647212. 21468851.0 8838049.8 HA1 Ig APQVYTIPPPK 16147001.67 07 2 4 4988877. 18800741.9 5158035.6 HA2 Ig APQVYTIPPPK 18184169.67 35 8 7 6200932. 21434490.6 6200932.7 HA3 Ig APQVYTIPPPK 21434490.67 76 7 6 9099028. 24405037.1 9229736.5 HA4 Ig APQVYTIPPPK 24059421.67 27 9 4 8385034. 24971157.4 8658457.6 HA5 Ig APQVYTIPPPK 24182599.67 45 8 8 2904016. 11761104.0 3109507.7 HA6 Ig APQVYTIPPPK 10983873.67 54 5 8 4303335. 18410811.4 5678050.2 WT1 Ig APQVYTIPPPK 13953362.67 25 2 1 2373449. 14277172.9 3422995.2 WT2 Ig APQVYTIPPPK 9899561.67 75 9 6 WT3 Ig APQVYTIPPPK 15096865.33 3656860. 22665748.2 5490244.3

150 50 3 3 4091523. 18538306.8 5030816.6 WT4 Ig APQVYTIPPPK 15077057.33 16 3 9 4374696. 14383506.9 4871190.4 WT5 Ig APQVYTIPPPK 12917473.00 07 3 6 1047563. 18724619.7 1633531.2 WT6 Ig APQVYTIPPPK 12007871.67 79 2 5 1303069 80433464.6 17325449. HA1 Ig DVLTITLTPK 60495053.33 5.57 2 50 9035780. 76588887.2 9342157.5 HA2 Ig DVLTITLTPK 74077146.67 49 7 2 1507409 80536658.6 15074096. HA3 Ig DVLTITLTPK 80536658.67 6.01 7 01 1013458 75303785.6 10280167. HA4 Ig DVLTITLTPK 74237360.00 3.82 3 93 1622721 79970239.8 16756362. HA5 Ig DVLTITLTPK 77444880.00 7.42 0 31 2885937 86739118.6 30901491. HA6 Ig DVLTITLTPK 81006980.00 2.21 1 56 1686480 81735069.7 22252326. WT1 Ig DVLTITLTPK 61946160.00 6.83 1 24 1029185 91082503.3 14842934. WT2 Ig DVLTITLTPK 63155140.00 1.62 7 55 1497017 75261520.9 22475534. WT3 Ig DVLTITLTPK 50129077.33 0.56 0 44 1414608 78238956.0 17393616. WT4 Ig DVLTITLTPK 63631120.00 9.43 8 04 7887173. 82556401.7 8782306.5 WT5 Ig DVLTITLTPK 74141869.33 03 8 6 1387608 96021217.0 21637842. WT6 Ig DVLTITLTPK 61577242.67 6.24 9 67 5121910. 61086924.6 6810028.0 HA1 Ig VNSAAFPAPIEK 45944269.33 50 1 0 1076287 69864760.8 11127813. HA2 Ig VNSAAFPAPIEK 67573538.67 5.96 4 78 3979391. 61357494.6 3979391.6 HA3 Ig VNSAAFPAPIEK 61357494.67 69 7 9 8621585. 61983772.3 8745434.9 HA4 Ig VNSAAFPAPIEK 61105980.00 14 0 0 2460283. 60174932.8 2540509.1 HA5 Ig VNSAAFPAPIEK 58274684.00 10 5 9 6730476. 63688141.7 7206732.2 HA6 Ig VNSAAFPAPIEK 59479322.67 70 3 7 8224392. 64577366.0 10851702. WT1 Ig VNSAAFPAPIEK 48942514.67 92 2 97 2340207. 59424480.8 3375052.7 WT2 Ig VNSAAFPAPIEK 41203977.33 20 0 6 4607422. 65547626.5 6917375.0 WT3 Ig VNSAAFPAPIEK 43658990.67 52 1 0 3574123. 66878307.1 4394636.6 WT4 Ig VNSAAFPAPIEK 54391594.67 02 2 7 3586522. 67750612.1 3993564.7 WT5 Ig VNSAAFPAPIEK 60845154.67 08 8 8 3382211. 46496452.3 5274092.7 WT6 Ig VNSAAFPAPIEK 29817611.33 74 8 3 1466667. 15502661.4 1950063.2 HA1 kappa DSTYSMSSTLTLTK 11659753.00 93 4 9 1661653. 13233248.4 1717995.1 HA2 kappa DSTYSMSSTLTLTK 12799262.67 29 4 2 2474495. 14894908.6 2474495.7 HA3 kappa DSTYSMSSTLTLTK 14894908.67 76 7 6 2335719. 16537889.4 2369272.7 HA4 kappa DSTYSMSSTLTLTK 16303685.67 97 2 8 3843593. 13826085.0 3968926.9 HA5 kappa DSTYSMSSTLTLTK 13389474.67 17 8 0 4190097. 15786302.1 4486593.3 HA6 kappa DSTYSMSSTLTLTK 14743067.33 65 9 5

151 3463816. 13950705.8 4570343.9 WT1 kappa DSTYSMSSTLTLTK 10573095.00 17 5 2 2403211. 13900426.9 3465917.5 WT2 kappa DSTYSMSSTLTLTK 9638332.00 40 0 9 1456200. 13114953.5 2186272.6 WT3 kappa DSTYSMSSTLTLTK 8735413.67 06 0 6 2386537. 13495464.3 2934416.5 WT4 kappa DSTYSMSSTLTLTK 10975753.67 58 1 0 1418636. 13633744.6 1579640.4 WT5 kappa DSTYSMSSTLTLTK 12244130.00 18 7 9 2768502. 18489133.6 4317097.6 WT6 kappa DSTYSMSSTLTLTK 11856857.33 39 9 4 530769.2 HA1 kappa QNGVLNSWTDQDSK 3885883.33 1 5166621.77 705704.10 824012.8 HA2 kappa QNGVLNSWTDQDSK 5311551.00 5 5491650.25 851952.72 1030312. 1030312.7 HA3 kappa QNGVLNSWTDQDSK 5719575.33 73 5719575.33 3 768735.4 HA4 kappa QNGVLNSWTDQDSK 5707333.33 6 5789319.63 779778.40 529727.5 HA5 kappa QNGVLNSWTDQDSK 5032179.00 1 5196270.71 547001.12 1366411. 1463100.6 HA6 kappa QNGVLNSWTDQDSK 5701775.00 92 6105238.56 6 823497.0 1086565.9 WT1 kappa QNGVLNSWTDQDSK 4014361.00 1 5296762.16 1 107863.6 WT2 kappa QNGVLNSWTDQDSK 3674559.00 9 5299458.33 155561.29 484712.7 WT3 kappa QNGVLNSWTDQDSK 4818170.33 6 7233782.20 727725.73 1397168. 1717917.8 WT4 kappa QNGVLNSWTDQDSK 5178368.33 87 6367169.60 0 331577.5 WT5 kappa QNGVLNSWTDQDSK 5098518.67 0 5677161.36 369209.00 111218.5 WT6 kappa QNGVLNSWTDQDSK 2715212.67 1 4233999.67 173429.93 HA1 kappa TSTSPIVK 626413.00 61886.42 832870.87 82283.41 112644.0 HA2 kappa TSTSPIVK 495311.00 5 512105.56 116463.48 HA3 kappa TSTSPIVK 548266.33 42942.37 548266.33 42942.37 HA4 kappa TSTSPIVK 464911.67 58048.63 471590.16 58882.50 HA5 kappa TSTSPIVK 341570.33 60561.73 352708.42 62536.56 HA6 kappa TSTSPIVK 384296.00 57924.12 411489.19 62022.89 WT1 kappa TSTSPIVK 394617.00 39758.47 520678.73 52459.45 WT2 kappa TSTSPIVK 351789.67 51520.26 507351.95 74302.65 WT3 kappa TSTSPIVK 444773.67 14534.57 667762.99 21821.55 139755.7 WT4 kappa TSTSPIVK 791488.00 6 973190.40 171839.58 WT5 kappa TSTSPIVK 440028.00 70775.35 489967.80 78807.81 WT6 kappa TSTSPIVK 337299.33 34975.38 525971.79 54539.28 HA1 Pclo AEEDSMEDPYELK 1466.00 280.74 1949.18 373.26 HA2 Pclo AEEDSMEDPYELK 2178.67 917.07 2252.54 948.16 HA3 Pclo AEEDSMEDPYELK 1441.33 360.59 1441.33 360.59 HA4 Pclo AEEDSMEDPYELK 967.67 154.30 981.57 156.52 HA5 Pclo AEEDSMEDPYELK 1584.67 835.18 1636.34 862.41 HA6 Pclo AEEDSMEDPYELK 1174.33 814.51 1257.43 872.15 WT1 Pclo AEEDSMEDPYELK 1455.00 320.35 1919.80 422.68

152 WT2 Pclo AEEDSMEDPYELK 1283.00 448.12 1850.35 646.28 WT3 Pclo AEEDSMEDPYELK 1572.00 395.81 2360.13 594.25 WT4 Pclo AEEDSMEDPYELK 1731.67 374.89 2129.21 460.95 WT5 Pclo AEEDSMEDPYELK 1991.00 258.09 2216.96 287.38 WT6 Pclo AEEDSMEDPYELK 973.67 361.92 1518.30 564.36 HA1 Pclo ALGGELAAIPSSPQPTPK 26813.33 4330.45 35650.67 5757.71 HA2 Pclo ALGGELAAIPSSPQPTPK 31518.33 8242.52 32587.03 8522.00 HA3 Pclo ALGGELAAIPSSPQPTPK 35651.67 4987.19 35651.67 4987.19 HA4 Pclo ALGGELAAIPSSPQPTPK 31698.33 5492.11 32153.68 5571.00 HA5 Pclo ALGGELAAIPSSPQPTPK 30485.33 457.22 31479.41 472.13 HA6 Pclo ALGGELAAIPSSPQPTPK 38909.67 5662.75 41662.96 6063.46 WT1 Pclo ALGGELAAIPSSPQPTPK 30878.33 3071.73 40742.52 4053.00 WT2 Pclo ALGGELAAIPSSPQPTPK 27420.67 3478.96 39546.16 5017.36 WT3 Pclo ALGGELAAIPSSPQPTPK 28356.33 430.68 42572.91 646.60 WT4 Pclo ALGGELAAIPSSPQPTPK 22878.33 6900.17 28130.53 8484.25 WT5 Pclo ALGGELAAIPSSPQPTPK 24958.67 8774.53 27791.28 9770.37 WT6 Pclo ALGGELAAIPSSPQPTPK 28457.33 3332.63 44375.29 5196.78 HA1 Pclo ISLPETGLAPTPSSQTK 315653.00 59155.76 419688.27 78652.76 HA2 Pclo ISLPETGLAPTPSSQTK 270403.33 2132.90 279571.92 2205.22 HA3 Pclo ISLPETGLAPTPSSQTK 337552.33 12535.95 337552.33 12535.95 HA4 Pclo ISLPETGLAPTPSSQTK 344984.33 47295.66 349940.06 47975.07 HA5 Pclo ISLPETGLAPTPSSQTK 280705.33 57070.63 289858.71 58931.62 HA6 Pclo ISLPETGLAPTPSSQTK 295212.33 85964.45 316101.87 92047.39 WT1 Pclo ISLPETGLAPTPSSQTK 322431.33 34344.78 425433.11 45316.33 WT2 Pclo ISLPETGLAPTPSSQTK 307371.00 32600.62 443291.24 47016.70 WT3 Pclo ISLPETGLAPTPSSQTK 281232.67 46318.69 422229.96 69540.78 WT4 Pclo ISLPETGLAPTPSSQTK 294165.00 34196.35 361696.64 42046.83 WT5 Pclo ISLPETGLAPTPSSQTK 267019.00 39623.15 297323.60 44120.08 WT6 Pclo ISLPETGLAPTPSSQTK 299049.33 12324.64 466326.19 19218.57 HA1 Stxbp1 DNALLAQLIQDK 40503.33 974.65 53852.72 1295.88 HA2 Stxbp1 DNALLAQLIQDK 47984.33 8207.85 49611.34 8486.15 HA3 Stxbp1 DNALLAQLIQDK 51396.00 5954.17 51396.00 5954.17 HA4 Stxbp1 DNALLAQLIQDK 67812.00 12814.74 68786.12 12998.82 HA5 Stxbp1 DNALLAQLIQDK 51389.33 22844.50 53065.06 23589.43 HA6 Stxbp1 DNALLAQLIQDK 62206.00 16634.49 66607.76 17811.56 WT1 Stxbp1 DNALLAQLIQDK 16913.33 1153.36 22316.35 1521.80 WT2 Stxbp1 DNALLAQLIQDK 21929.67 2710.39 31627.02 3908.93 WT3 Stxbp1 DNALLAQLIQDK 20603.00 372.73 30932.41 559.61 WT4 Stxbp1 DNALLAQLIQDK 23141.33 5107.25 28453.90 6279.72 WT5 Stxbp1 DNALLAQLIQDK 29131.67 3414.89 32437.89 3802.45 WT6 Stxbp1 DNALLAQLIQDK 31173.67 4579.31 48611.03 7140.80

153 HA1 Stxbp1 EPLPSLEAVYLITPSEK 17841.00 3095.92 23721.17 4116.29 HA2 Stxbp1 EPLPSLEAVYLITPSEK 21779.67 4262.56 22518.15 4407.09 HA3 Stxbp1 EPLPSLEAVYLITPSEK 24375.67 2052.82 24375.67 2052.82 HA4 Stxbp1 EPLPSLEAVYLITPSEK 27303.00 3057.18 27695.21 3101.10 HA5 Stxbp1 EPLPSLEAVYLITPSEK 24152.67 6152.63 24940.25 6353.26 HA6 Stxbp1 EPLPSLEAVYLITPSEK 25114.67 3725.53 26891.81 3989.15 WT1 Stxbp1 EPLPSLEAVYLITPSEK 11365.00 3205.34 14995.59 4229.30 WT2 Stxbp1 EPLPSLEAVYLITPSEK 11084.33 3495.17 15985.85 5040.74 WT3 Stxbp1 EPLPSLEAVYLITPSEK 9544.00 2208.55 14328.93 3315.82 WT4 Stxbp1 EPLPSLEAVYLITPSEK 11539.00 3310.64 14188.02 4070.67 WT5 Stxbp1 EPLPSLEAVYLITPSEK 10531.67 2182.06 11726.93 2429.70 WT6 Stxbp1 EPLPSLEAVYLITPSEK 17364.67 2253.41 27077.80 3513.88

154