190323111.Pdf
Characterization of polymers of nucleotide biosynthetic enzymes
By
Sajitha Anthony
April, 2017
A dissertation presented to the faculty of Drexel University College of Medicine in partial fulfillment for the requirements for the degree of Doctor of Philosophy in Molecular and Cellular Biology and Genetics
i
ii
ACKNOWLEDGEMENTS
First and foremost, I would like to thank my mentor Dr. Jeffrey
Peterson for all of his support and guidance throughout the five years that I have been in his lab. He has truly inspired me with his tremendous enthusiasm for science and his constant encouragement. He has changed, for the better, the way I approach and see science. I could not have asked for a better mentor.
Secondly, I would like to thank my committee members for all of their great ideas and their never-ending support. I thank them for answering all of my questions so patiently and always listening to what I had to say. After every meeting I’ve had with them, I always felt encouraged and confident.
I would also like to thank my collaborators over at the University of
Washington—Justin Kollman, Anika Burrell, and Matthew Johnson. It has been such a pleasure working with them. The journey of discovery that we took together has been very exciting. This was my first collaboration, and they have made it such a memorable and positive experience. iii
I would like to thank my lab members, in particular Alex for being an awesome cubemate, colleague, and friend. He has been the sounding board for so many of my project’s ideas over the years and has given such knowledgeable input. I would also like to thank Jackie for being a good colleague and friend. You have advised me on various things and shared so much of your knowledge with me. Also, I thank both of them for sitting in on so many of my practice talks!
I would like to thank my mother, who has always supported me throughout my life. She has always emphasized the importance of education and hard work. She has pushed me to achieve my goals and has always advised me to never settle. Without her undying support and encouragement throughout the years, I would not be where I am today.
And last, but certainly not least, I would like to thank my dear husband Christian. It has been a busy year with us having to adjust to his recent move here, our marriage and my thesis writing all at once. Through it all, he has stood by me always ready to offer encouragement and help. His technological knowhow has been a tremendous help to me. He has guided me through formatting issues (and numerous recoveries!) of my thesis. I am truly grateful for him.
Here’s to the future! iv
TABLE OF CONTENTS
LIST OF TABLES ...... xi
LIST OF ILLUSTRATIONS ...... xii
ABSTRACT ...... xv
Chapter 1: Introduction ...... 1
Nucleotide biosynthesizing enzymes form macromolecular structures
within cells ...... 2
...... 3
The many roles of nucleotides in cellular metabolism ...... 3
The importance of nucleotides in dividing and quiescent cells ...... 3
Nucleotides as intermediates in cellular communication ...... 5
The role of nucleotides in the nervous system ...... 6
Nucleotide sugars as activators of metabolites for lipid synthesis ...... 6
Nucleotides as cofactors in reactions (S-Adenosyl-L-methionine) ...... 7
Nucleotides as electron donors ...... 8 v
Nucleotides as energy carriers (ATP, GTP) ...... 8
The process of purine synthesis and catabolism in the cell ...... 9
De novo nucleotide synthesis ...... 9
The salvage pathway ...... 11
Regulation of synthesis ...... 13
Catabolism of purine nucleotides ...... 14
Purine nucleotide salvage deficiencies and neurological pathologies ...... 15
Purine nucleotide salvage deficiencies and mitochondrial depletion
syndromes ...... 17
Deficiencies of the purine de novo synthesis pathway which result in
immunodeficiency ...... 17
Pyrimidine synthesis and catabolism ...... 19
De novo pyrimidine synthesis ...... 19
The pyrimidine salvage pathway ...... 20
Regulation of pyrimidine synthesis ...... 21
Catabolism of pyrimidine nucleotides ...... 22
Pyrimidine nucleotide salvage deficiencies ...... 23
De novo pyrimidine nucleotide synthesis deficiencies ...... 24
De novo pyrimidine nucleotide synthesis and immunodeficiency ...... 25
Inosine monophosphate (IMPDH) and de novo nucleotide synthesis ...... 26 vi
The IMPDH reaction ...... 26
The structure of IMPDH ...... 27
The role of the Bateman domain and nucleotides in the regulation of
IMPDH activity ...... 29
Activation of IMPDH by monovalent cations ...... 31
Isoforms of IMPDH...... 32
Moonlighting functions of IMPDH ...... 33
Therapeutic targeting of IMPDH activity ...... 35
MPA ...... 35
Ribavirin ...... 37
Other inhibitors of IMPDH ...... 38
IMPDH as filamentous structures ...... 39
Physical characteristics of IMPDH filaments ...... 39
History of the discovery of IMPDH filaments ...... 40
Regulation of IMPDH filament assembly ...... 43
The effect of IMPDH filament assembly on catalytic activity ...... 48
IMPDH can co-assemble with CTPS into filaments ...... 48
CTPS in pyrimidine synthesis ...... 50
Structural characteristic of CTPS ...... 51 vii
Isoforms and expression ...... 52
The regulation of CTPS activity ...... 54
Allosteric regulation ...... 54
Regulation of CTP by post translational modifications ...... 55
The targeting of CTPS activity as a therapeutic ...... 56
Cyclopentenylcytosine Triphosphate (CPEC) and c3Urd ...... 57
Gemcitabine ...... 58
Azaserine ...... 60
DON ...... 62
CTPS as filamentous structures ...... 66
The discovery of CTPS filaments ...... 66
Regulation of CTPS filament assembly ...... 68
The biological role of CTPS filament assembly ...... 77
Chapter 2: Identification of CTPS1 associating proteins within the filaments
...... 84
Introduction ...... 85
Materials and methods ...... 89
Cell culture and transfection ...... 89
Plasmid ...... 89 viii
Transfection and cell culture ...... 91
Immunofluorescence ...... 91
Immunoblotting ...... 92
Capture of biotinylated proteins ...... 93
Silver Stain ...... 93
Identification of proteins using mass spectrometry ...... 94
Results ...... 95
BirA*-CTPS1 localizes to filaments upon treatment with DON ...... 95
BirA*CTPS promiscuously biotinylates proteins within cells ...... 96
BirA*-CTPS identifies known filament components—IMPDH2 and
CTPS1 and one candidate protein Histidine Ammonia Lyase ...... 97
Verification of HAL as a filament component ...... 102
Discussion ...... 104
Future directions ...... 107
Chapter 3: The biological role of filament assembly ...... 111
Introduction ...... 112
Materials and methods ...... 114
Cell culture and transfection ...... 114
Immunoblotting ...... 115 ix
Immunofluorescence ...... 115
Vectors ...... 116
Mutagenesis ...... 118
ULP1 Protein Purification and expression ...... 119
IMPDH Protein Purification and expression ...... 121
IMPDH activity assay ...... 123
Negatively Stained electron microscopy and particle
averaging/reconstruction...... 125
Electron cryo-microscopy and particle averaging/reconstruction ...... 126
[13C2, 15N]-glycine incorporation into AMP and GMP ...... 127
Results ...... 129
Mutants of IMPDH alter the ability of IMPDH2 to polymerize ...... 129
Polymerization does not alter enzymatic activity of IMPDH in vitro .... 133
Polymerization of IMPDH2 into filaments does alter guanine nucleotide
biosynthetic flux in vivo ...... 136
GTP bound human IMPDH2 takes on a collapsed octamer conformation
and ATP induced polymerization does not protect against allosteric GTP
inhibition...... 144
Discussion ...... 147
Future directions ...... 154 x
Chapter 4: Models describing possible roles for filament assembly ...... 156
Increased awareness of the significance of metabolic enzyme
polymerization ...... 157
Exploring the role of CTPS filament assembly ...... 158
Exploring the physiological role of IMPDH filament assembly ...... 159
Exploring the physiological role of IMPDH filament assembly: the
lymphocyte ...... 162
Exploring the physiological role of IMPDH filament assembly: tumor cells
...... 165
Exploring the physiological role of IMPDH filament assembly: open
versus closed octamer filaments ...... 168
Exploring the physiological role of filament co-assembly ...... 169
Conclusion ...... 172
LIST OF REFERENCES ...... 1
APPENDIX A ...... 168
APPENDIX B ...... 201
...... 217
xi
LIST OF TABLES
Table 1. Proteins unique to DON treated samples ...... 101
Table 2. IMP and NAD Km values for wild type, R356A, and Y12A ...... 135
xii
LIST OF ILLUSTRATIONS
Figure 1. The molecular structures of pyrimidine and purine bases...... 3
Figure 2 De novo purine nucleotide synthesis pathway...... 11
Figure 3. Pathways for purine catabolism in animals...... 14
Figure 4. De novo pyrimidine nucleotide synthesis pathway...... 19
Figure 5. Crystal structure of IMPDH...... 27
Figure 6. Octamers take on different conformations based on the type of nucleotide that binds...... 29
Figure 7. IMPDH2 filaments...... 39
Figure 8. One carbon metabolism in nucleotide synthesis...... 45
Figure 9. CTPS reaction...... 51
Figure 10. Crystal structure of CTPS...... 52
Figure 11. Molecular structures of CTPS inhibitors DON and Azaserine. .... 65
Figure 12. BirA* vector...... 90 xiii
Figure 13. BirA*-CTPS co-localizes with IMPDH2 in filaments...... 95
Figure 14. Biotinlyation of proteins in BirA*-CTPS expressing cells and non- expressing cells...... 96
Figure 15. Silver stain of lysates from untreated and DON treated cells...... 98
Figure 16. Coomassie dye stained gel of lysate from untreated and DON treated cells...... 99
Figure 17. HAL localizes to filaments upon DON treatment...... 102
Figure 18. Anti-HA antibody non-specifically cross reacts with filaments. . 103
Figure 19. pSMT3-Kan vector map ...... 117
Figure 20. Identification of filament disrupting and promoting point mutations of human IMPDH2...... 131
Figure 21. IMPDH2 polymerization does not alter its catalytic activity in vitro...... 134
Figure 22. . IMPDH filament formation does not alter GMP synthesis...... 139
Figure 23. Ligands influence IMPDH filament architecture by altering the conformation of IMPDH protomers ...... 142
Figure 24. Human IMPDH adopts the collapsed, inhibited conformation when bound to GTP...... 146
Figure 25. Non-polymerizing mutants Y12A and R356A prevent IMPDH polymerization in a dominant negative manner ...... 162 xiv
Figure 26. Role of IMPDH filament assembly in T-cell activation...... 164
Figure 27. Role of IMPDH filament assembly in cancer development...... 166
Figure 28. A hypothesis for the role of IMPDH, CTPS and HAL co-assembly in nucleotide synthesis...... 171
Figure 29. The schematic illustration of IMPDH catalyzing reaction...... 215
Figure 30. Graph showing the relative IMPDH activity in the IMPDH reaction with PARP-1 inhibitors (Olaparob and 5F02) compared to the control
(DMSO only)...... 216
Figure 31. Coomassie-stained gel of purified IMPDH2 wild type and mutant proteins...... 217
xv
ABSTRACT
Characterization of polymers of nucleotide biosynthetic enzymes
Sajitha Anthony
Jeffrey Peterson, Ph.D.
Mammalian cells contain various proteins that assemble into filamentous structures. Examples include cytoskeletal proteins such as actin and even metabolic enzymes like acetyl-CoA carboxylase. Within the last decade, CTP synthase (CTPS) and inosine monophosphate dehydrogenase
(IMPDH) have been found to assemble into filaments. Both are rate-limiting enzymes involved in nucleotide biosynthesis. Their assemblies are linked to nucleotide depletion. Filaments containing CTPS and/or IMPDH have been identified within the cells of different organisms such as bacteria, yeast, flies and mammals.
CTPS and IMPDH2 can co-assemble into filaments. To determine whether additional proteins can co-assemble with CTPS and IMPDH2, we implemented the Bio-ID system followed by mass spectrometry. Histidine xvi ammonia lyase was identified as a candidate and preliminary immunofluorescence studies found that it localizes to the filaments. Histidine ammonia lyase releases ammonia in its enzymatic reaction which may be used by CTPS which requires ammonia.
CTPS and IMPDH2 are also capable of assembling individually into filaments. CTPS filaments can regulate both cell shape and enzymatic activity within bacterial cells. On the other hand, the function of IMPDH2 polymerization is unclear. What is known is that IMPDH2 polymerization can be regulated by the binding of its allosteric regulators GTP and ATP which bind to the Bateman domain of the enzyme. ATP triggers the assembly of enzymatically active filaments composed of extended octamers. GTP, on the other hand, promotes enzymatically inactive octamers which are compressed in form and it is unknown whether these octamers can polymerize. To determine whether IMPDH2 polymerization changed
IMPDH2 enzymatic activity, point mutants of IMPDH2 that either prevent or reinforced octamer polymerization were engineered. No difference in activity, substrate affinity, or GTP inhibition sensitivity was found between polymerized and non-polymerized IMPDH2. Cellular isotope tracing experiments validated these results. Surprisingly, electron microscopy revealed that compact octamers can polymerize.
These findings demonstrate that IMPDH 2 polymerization has no effect on enzymatic activity. IMPDH is also involved in activities unrelated to xvii its enzymatic activity such as gene regulation. Polymerization may be a way to regulate its involvement in these activities
1
Chapter 1: Introduction
2
Nucleotide biosynthesizing enzymes form macromolecular structures within cells
Nucleotides fulfill many different roles within the cell, therefore the process of nucleotide biosynthesis is very important and must be strictly regulated (Garrett, 1995). Rate limiting nucleotide biosynthesizing enzymes such as CTP synthetase (CTPS) and inosine monophosphate dehydrogenase
(IMPDH) are indispensable for this regulation (Lane and Fan, 2015). Because of the critical role that IMPDH and CTPS play in nucleotide synthesis, their activities have become therapeutic targets in the fight against cancer, infections, and autoimmune disease (Yokota, 2002; McKeran and Watts,
1976; McCluskey et al., 2016 ). Within the past ten years, it has been discovered that each of these two enzymes can assemble into macromolecular structures within the cell (Carcamo et al., 2014; Liu et al., 2010; Ingerson-
Maher et al., 2010; Noree et al., 2010). These enzymes can assemble individually as well as together (Carcamo et al., 2014; Calise et al., 2016;
Keppeke et al., 2015). The structures are elongated and filamentous in appearance and can be found in both the nucleus and the cytoplasm
(Carcamo et al., 2014; Carcamo et al., 2011; Juda et al., 2014). Their assembly and disassembly has been linked to nucleotide levels within the cell
(Calise et al., 2016), but whether these structures regulate nucleotide synthesis or if filaments of IMPDH play any other biological role, is unclear.
It is also unknown if IMPDH and CTPS can assemble with additional 3 proteins. Therefore, the goal of my research was to address these two unknowns. In this first chapter, I will first discuss the important roles played by nucleotides within the cell, how nucleotides are produced and catabolized, as well as the important enzymes involved in these processes. Additionally, I will briefly discuss the diseases that could result if any of these nucleotide biosynthesizing or catabolizing processes malfunction. Then finally, a detailed of discussion about IMPDH and CTPS will occur. In chapters 2 and 3
I will discuss my research concerning these two enzymes and then finally, in chapter 4, I will discuss my overall models for IMPDH filament assembly as well as IMPDH and CTPS co-assembly.
The many roles of nucleotides in
cellular metabolism
The importance of nucleotides in
Figure 1. The molecular structures dividing and quiescent cells of pyrimidine and purine bases. A Nucleotides are the subunits nucleotide is a derivative of either one that make up the structure of nucleic acids such as DNA and RNA. DNA and RNA are two crucial molecules within the cell which contain genetic information. There are two major types of nucleotides: purines and pyrimidines (Figure 1). In quiescent cells, 4 nucleotides are needed to replace the RNA and protein that are constantly being turned over within the cell (Berg JM, 2002; GM., 2000; Hacker et al.,
2015; Lane and Fan, 2015). mRNA turnover, in particular, is an important part of gene expression within eukaryotic cells (Raghavan et al., 2002; Wang et al., 2002; Wilusz and Wilusz, 2004; Yang et al., 2003). Turnover is also important for regulating protein levels within a cell (Schoenberg and Maquat,
2012). In addition, mRNA production and decay are both needed to sustain cellular homeostasis (Schoenberg and Maquat, 2012). Transcripts that are abnormal or no longer needed are degraded to prevent the production of harmful proteins (Schoenberg and Maquat, 2012; Wilusz et al., 2001). Old proteins and misfolded proteins are constantly being replaced within the cell
(Schoenberg and Maquat, 2012; Wilusz et al., 2001). Nucleotides are also needed to maintain genomic integrity in quiescent cells (Lane and Fan, 2015).
For example, nucleotides are used by DNA repair pathways (Lane and Fan,
2015). In proliferating cells, the need for nucleotides is more obvious as these cells need to duplicate their DNA and synthesize RNA in order to divide
(Lane and Fan, 2015). Dividing cells must be able to make new nucleotides at the same rate that cell division occurs (Lane and Fan, 2015). Therefore, during late G1 phase of the cell cycle, there is an increase in the expression of genes that are involved in nucleotide synthesis (Hordern and Henderson,
1982; Lane and Fan, 2015). Nucleotide levels are also important in the quiescent, differentiated, neuronal cells of the nervous system. Although 5 quiescent and non-cycling, they have high metabolic activity and a high demand for ATP (Fasullo and Endres, 2015; Federico et al., 2012).
Nucleotides as intermediates in cellular communication
Nucleotides are also involved in cellular signaling as second messengers (Krawutschke et al., 2015; Seifert, 2015). Second messengers are produced within the cell as a response to extracellular signals (Krawutschke et al., 2015; Seifert, 2015). These second messengers then trigger a change within the cell (Krawutschke et al., 2015; Newton, 2016). Cyclic AMP is one of the most common second messengers within the cell (Krawutschke et al.,
2015). Cyclic GMP is another second messenger that commonly occurs and is used for intracellular signaling in the vascular and nervous system (Domek-
Lopacinska and Strosznajder, 2005; Krawutschke et al., 2015). Within the vascular system, Cyclic GMP’s regulation of the endothelial permeability and vascular tone lowers blood pressure (Krawutschke et al., 2015). Cyclic GMP also prevents the clumping of platelets (Krawutschke et al., 2015). The less commonly known second messengers cyclic CMP and cyclic UMP appear to be found in all mammals, but not in lower organisms such as E. coli (Hartwig et al., 2014; Seifert, 2015). This suggests that these molecules are relatively new and are possibly taking over activities that were not adequately fulfilled by the more ancient Cyclic AMP and Cyclic GMP within mammalian cells
(Hartwig et al., 2014; Seifert, 2015). 6
The role of nucleotides in the nervous system
Nucleotides serve as important messengers within the nervous system
(Averaimo and Nicol, 2014; Fasullo and Endres, 2015). For example, cAMP plays a crucial role in shaping the development and connectivity of neurons while ATP functions as a co-stimulator in motor, sensory-motor, hypothalamus, parasympathetic, and sympathetic nerves (Averaimo and
Nicol, 2014; Fasullo and Endres, 2015). ATP is also important for neuroprotection and plays a role in stem cell derived neuro-regeneration
(Averaimo and Nicol, 2014; Fasullo and Endres, 2015). Guanosine nucleotides are important for affecting glutamatergic neurotransmission via the stimulation of glial reuptake of L-glutamate (Deutsch et al., 2005; Fasullo and Endres, 2015).
Nucleotide sugars as activators of metabolites for lipid synthesis
Nucleotide sugars are used to activate metabolites which enables them to actively take part in different anabolic processes (Berg JM, 2002; Lane and
Fan, 2015). Examples of such nucleotide sugars include UDP-glucose and other UDP hexoses which are needed for carbohydrate synthesis, CDP- choline for lipid synthesis, GDP sugars such as GDP-mannose which is needed for glycosyltransferases, NAD(P) and FAD/FMN which are used for redox reactions, and ADP ribosylation which is used for many different regulatory functions (Berg JM, 2002). In order to create phospholipids, 7 diacylglyceride and alcohol must be combined, (Berg JM, 2002; Sarri et al.,
2011). In the de novo pathway, the diacylglycerol is activated. The pyrimidine nucleotide cytidine triphosphate (CTP) is needed in the first reaction which occurs between phosphatidate and CTP to form cytidine diphosphodiacylglyceride (CDP-diacylglycerol) (Berg JM, 2002). This activated product can then react with the hydroxyl group on the alcohol to form a phosphodiester linkage, thus creating phospholipids (Berg JM, 2002).
The type of phospholipid produced depends on the type of alcohol involved
(Berg JM, 2002). For example, if the alcohol is serine, the resulting phospholipid will be the phospholipid phosphatidyl serine (Berg JM, 2002).
Nucleotides as cofactors in reactions (S-Adenosyl-L-methionine)
S-Adenosyl-L-methionine, which is synthesized from ATP and methionine, is a cofactor involved in numerous enzyme-catalyzed reactions within the cell (Roje, 2006). It serves as a source of methyl groups for methylation reactions and also a precursor for the production of polyamines such as spermidine and spermine (Roje, 2006). In addition, it is a precursor for metal chelating compounds like nicotinamine and phytosiderophores
(Broderick et al., 2014; Roje, 2006).
8
Nucleotides as electron donors
The pyridine nucleotide NAD(H) is a mediator of electron transfer in many different catabolic and anabolic processes. NAD(H) is involved in the synthesis of ATP (Nakamura et al., 2012). NAD receives electrons and forms
NADH (Lodish, 2000; Nakamura et al., 2012). NADH then donates these electrons to the electron transport chain in the mitochondria where an electrochemical gradient is produced in order to convert ADP to ATP (Lodish,
2000; Nakamura et al., 2012).
Nucleotides as energy carriers (ATP, GTP)
Free energy within cells, is chemically stored, captured and transported (GM., 2000; Lehninger, 1993). The nucleotide triphosphates are important sources of chemical energy which are used to drive many different biochemical reactions within the cell (GM., 2000; Lehninger, 1993). Energy is released from these nucleotides by the hydrolysis of the terminal phosphate bond which releases around 30 kJ/mol of energy (Lehninger, 1993).
Of all the nucleotides, adenosine triphosphate (ATP) is the main carrier of this energy within cells, although UTP, GTP, and CTP are also used for particular reactions (Alberts, 2002). ATP is formed by reactions that are driven by energy released by oxidative breakdown of foodstuff (Alberts,
2002). Upon hydrolysis of the terminal phosphate, ATP sends energy through chemical pathways via the terminal phosphate group which is passed from 9 one molecule to the next (Lehninger, 1993). When this terminal phosphate is cleaved, ADP is left over. ADP is then recycled in the mitochondria to create more ATP molecules by the combination of ADP and an inorganic phosphate
(Pi).
ATP is often maintained at high concentrations (>1mM) in order to maintain cellular energy (Lane and Fan, 2015). There are other tissues in the body such as skeletal muscle and cardiac myocytes that maintain an ATP level above 5mM (Lane and Fan, 2015). This energy is then used to drive endergonic reactions. The other RNTPs are also kept at a certain concentration in order to activate and drive anabolic processes (Lane and
Fan, 2015). Usually, out of all the nucleotides, the concentration of ATP is the highest, followed by UTP, then GTP and finally CTP (Lane and Fan, 2015).
Within normal human cells, ATP is 2537+1,217 µM, GTP is 232+202 µM,
UTP 227+321 µM and CTP 83+133 µM (Traut, 1994). Within tumor cells,
ATP is 3134+2,135 µM, GTP 473+214 µM, UTP 686+542 µM, and CTP
402+252 µM (Traut, 1994).
The process of purine synthesis and catabolism in the cell
De novo nucleotide synthesis
The term de novo is Latin in origin and means “as if for the first time” or
“anew” (Merriam-Webster, 2017). As the name suggests, in this mode of 10 nucleotide synthesis, purine nucleotides are synthesized from basic components which include phosphoribosyl pyrophosphate (PRPP), glutamine, and glycine (Berg JM, 2002; Fasullo and Endres, 2015). Purines are constructed with the purine base already attached to the ribose ring, in other words, PRPP serves as the base on which purine nucleotides are built step by step (Figure 2) (Berg JM, 2002; Lane and Fan, 2015). Ten steps then follow which lead up to the creation of inosine monophosphate (IMP) (Figure 2)
(Berg JM, 2002). IMP can then go down the branch of either one of two pathways to produce the final product of GTP or ATP (Figure 2). 11
Figure 2 De novo purine nucleotide synthesis pathway. The purine base is built on the ribose ring PRPP by amidotransferase and transformylation reactions. IMP is the branch point for purine synthesis as it can be converted to either AMP or GMP via two separate arms of the pathway. The arm leading to AMP needs GTP as energy, while the arm leading to GMP needs ATP as energy (green arrows). In red are amino acids which contribute to purine nucleotide synthesis.
The salvage pathway
Free purine bases derived from the diet or the outside environment can
be attached to PRPP to create purine nucleoside monophosphates, rather
than being made from scratch (Berg JM, 2002; Lane and Fan, 2015). This is
referred to as the “salvage pathway” in contrast to the de novo pathway.
Whether cells use the salvage pathway or the de novo pathway depends on 12 factors such as growth conditions and the tissue type (Lane and Fan, 2015).
There are two enzymes that are involved in the recovery of purine bases: adenine phosphoribosyltransferase (APRT), which catalyzes the production of adenine monophosphate (AMP), and hypoxanthine-guanine phosphoribosyltransferase (HGPRT), which catalyzes the production of guanosine monophosphate (GMP) in addition to inosine monophosphate
(IMP) which is the precursor for both AMP and GMP (Berg JM, 2002; Torres and Puig, 2007).
APRT Adenine + PRPP AMP + PPi
HGPRT Guanine + PRPP GMP+ PPi
HGPRT Hypoxanthine + PRPP IMP + PPi
The salvage process saves a lot of energy in comparison with the de novo pathway (Berg JM, 2002). In order to create GMP and IMP, HGPRT salvages the free purine bases hypoxanthine (Torres and Puig, 2007).
Monophosphates can also be salvaged from the constant turnover of RNA within the cell (Lane and Fan, 2015). The breakdown of RNA and DNA within the cells releases nucleotide monophosphates, which can then be phosphorylated by nucleotide kinases (Lane and Fan, 2015).
13
Regulation of synthesis
Many of the enzymes that are involved in nucleotide biosynthesis are regulated allosterically, namely, by feedback inhibition of pathway products.
Feedback inhibition is a way to thwart the production of more nucleotides when the amount is more than what the cells can utilize (Lehninger, 1993).
In this way, the rate of nucleotide end product production is balanced to the needs of the cell (Lehninger, 1993). The synthesis of purines is regulated by
AMP and GMP inhibition of PRPP synthetase, which limits PRPP production
(Berg JM, 2002; Lane and Fan, 2015). Interestingly, HGPRT reduces the amount of de novo synthesis while concurrently salvaging bases because its products IMP and GMP both inhibit PRPP (Berg JM, 2002). PRPP amidotransferase is also inhibited allosterically by the binding of ATP, ADP, and AMP at one site on the enzyme and GTP, GDP, and GMP at another site
(Herve, 1989; Lane and Fan, 2015). On the other hand, PRPP amidotransferase activity is stimulated by PRPP (Johnson et al., 2002; Lane and Fan, 2015). Another locus of allosteric regulation occurs when IMP is converted to XMP or adenylosuccinate: AMP suppresses adenylosuccinate synthetase and GMP suppresses inosine monophosphate dehydrogenase
(IMPDH) (Berg JM, 2002; Lane and Fan, 2015). In addition to this, GTP is needed for AMP production and ATP is needed for GMP production (Figure 2, green arrows) (Berg JM, 2002). This co-dependent regulation balances adenine and guanine nucleotide synthesis within the cell because increasing 14 amounts of ATP results in increased GMP production and vice versa (Berg
JM, 2002). The production of deoxyribonucleotides for ribonucleotides is also allosterically regulated (Berg JM, 2002). The binding of dATP to ribonucleotide reductase inhibits its activity, while the binding of ATP reverses this inhibition (Berg JM, 2002).
Catabolism of purine nucleotides
The breakdown of nucleotides is necessary for the maintenance of nucleotide pools (Fasullo and Endres, 2015; Rampazzo et al., 2010). Purine
Figure 3. Pathways for purine catabolism in animals. The catabolism of the various purine nucleosides converges in the synthesis of uric acid.
15 bases cannot be completely broken down (Garrett, 1995). Purine nucleotides are first changed into nucleosides by intracellular nucleotidases (Figure 3)
(Champe, 1994; Garrett, 1995). Adenosine is then converted into inosine by adenosine deaminase (Figure 3) (Champe, 1994; Garrett, 1995). The purine bases and the sugars are then released from inosine, xanthosine, and guanosine. All the bases are then converted to xanthine which is then converted to the final product uric acid (Figure 3) (Champe, 1994; Fasullo and Endres, 2015; Garrett, 1995). When the enzymatic activity of xanthine oxidase is excessive, the result is abnormally high production of uric acid and reactive oxygen species (ROS) (Fasullo and Endres, 2015). Uric acid is not efficiently absorbed by the blood, which can result in crystal formation
(Fasullo and Endres, 2015; Steiger and Harper, 2014). These crystals can cause a type of inflammation known as gout (Fasullo and Endres, 2015;
Garrett, 1995; Steiger and Harper, 2014).
Purine nucleotide salvage deficiencies and neurological pathologies
Neurons are highly metabolically active even though they are non- cycling (Fasullo and Endres, 2015; Federico et al., 2012). As a result, they have a high demand for ATP (Federico et al., 2012). They rely more heavily on the salvage pathway, rather than the de novo pathway, for nucleotide production (Fasullo and Endres, 2015). In tissues of the nervous system, the reliance on the de novo pathway for nucleotide productions decreases as a 16 human approaches adulthood (Fasullo and Endres, 2015). Unsurprisingly,
HGPRT is localized to the cells of the nervous system and is found in high concentrations within the brain tissue (Ceballos-Picot et al., 2009; Fasullo and Endres, 2015). For this reason, mutations in genes which code for enzymes involved in the salvage pathway, particularly those that are found in high amounts within the nervous system and brain like HGPRT, result in neurological pathologies such as neurodegeneration, and behaviors like self- mutilation by biting (Fasullo and Endres, 2015).
The devastating disease called Lesch-Nyhan syndrome results from mutations in the gene that codes for HGPRT. Because the nervous system is severely impacted by the mutation in this gene, it is no surprise that a number of cognitive and behavioral abnormalities result from this disease
(Fasullo and Endres, 2015; Fu et al., 2014). In addition, there is excessive production of uric acid (Fu et al., 2014). This buildup of uric acid results in its crystallization in different areas of the body such as is in the urogenital system and in the joints (Fu et al., 2014). Urogenital system buildup results in renal failure often within the first ten to twenty years of the patients’ lives
(Fasullo and Endres, 2015; Fu et al., 2014).
17
Purine nucleotide salvage deficiencies and mitochondrial depletion syndromes
Mitochondrial depletion syndromes (MDS) are autosomal and rare in occurrence (El-Hattab and Scaglia, 2013; Mandel et al., 2001; Rahman and
Poulton, 2009). These disorders result in reduced energy production in tissues and organs that are affected (El-Hattab and Scaglia, 2013; Elpeleg,
2003; Rahman and Poulton, 2009). There are three major types of MDS: myopathic, encephalomyopathy, and hepatocerebral (Mandel et al., 2001).
The hepatocerebral form of the disease is a result of mutations in the deoxyguanosine kinase (dGUOK) gene (Mandel et al., 2001). There have been a total of 22 different point mutations, deletions, and insertions reported in this gene. DGUOK, which together with thymidine kinase-2 initiates the salvage pathway, is responsible for providing most of the deoxyribonucleotides required for mtDNA synthesis (Mandel et al., 2001).
This enzyme plays an important role in maintaining a balanced pool of deoxyribonucleotides within the mitochondria (Mandel et al., 2001).
Deficiencies of the purine de novo synthesis pathway which result in immunodeficiency
Mature T-lymphocytes, an important component of our immune system, need nucleotides for their survival (Quemeneur et al., 2003). When T- cells are activated in response to some immunological challenge, they need to 18 increase their pyrimidine and purine pools (Quemeneur et al., 2003). These nucleotides regulate the proliferation and the survival of these cells as well as the cell cycle progression (Quemeneur et al., 2003).
GTP in particular appears to be of extreme importance for T-cell function (Nguyen et al., 2015; Yokota, 2002). Reasons for this include the fact that G-protein intercepts the connecting of the T-cell receptor to phospholipase C7 during antigen induced receptor signal (Yokota, 2002). It is believed that guanine nucleotides are second messengers in the alternate
CD28 mediated pathway for T cell activation (Yokota, 2002). This is supported by the fact that the stimulation of CD28 results in increased intracellular cGMP levels (Yokota, 2002). The need for GTP for immune function is further demonstrated by the fact that if there is a deficiency of the enzyme purine nucleoside phosphorylase, an enzyme involved in purine nucleotide metabolism (Papinazath et al., 2011), guanine ribonucleotides levels are reduced and severe T-cell immune deficiency results (Yokota,
2002).
19
Pyrimidine synthesis and catabolism
De novo pyrimidine synthesis
Pyrimidine nucleotides are synthesized from basic components that
Figure 4. De novo pyrimidine nucleotide synthesis pathway. Unlike purines, the ring structure of the pyrimidine is assembled as a free base and not built on PRPP. PRPP is added to the first completely formed pyrimidine base, orotic acid, resulting in OMP. OMP is then converted into UMP, which becomes phosphorylated to UDP. UDP then eventually becomes dTMP or CTP.
include carbamoyl phosphate, aspartate, and PRPP (Figure 4) (Fasullo and
Endres, 2015; Levine et al., 1974). Unlike what occurs in purines, the pyrimidine base is created first and then attached to the activated ribose
PRPP (Figure 4) (Berg JM, 2002; Lane and Fan, 2015). 20
The pyrimidine salvage pathway
In the case of pyrimidine nucleotides, the base uracil can be salvaged by its direct conversion to UMP by the enzyme uracil phosphoribosyltransferase (Li et al., 2007; Moffatt and Ashihara, 2002). The nucleosides cytidine and uridine can be converted to CMP and UMP by the enzyme uridine kinase (Moffatt and Ashihara, 2002). UMP can then be phosphorylated by UMP/CMP kinase (CMPK1) to make UDP (Hsu et al.,
2005; Zauri et al., 2015). Nucleotide diphosphate kinase can then phosphorylate this and produce UTP (Lascu and Gonin, 2000; Vieira et al.,
2015). In the case of thymine, a ribose can be added to it by thymidine phophorylase to form the nucleoside thymidine (Elamin et al., 2016;
Spinazzola et al., 2002). Thymidine can also be taken up into cells via transport proteins (Zupanc and Horschke, 1996).
Thymidine kinase can phosphorylate thymidine to produce dTMP
(Zupanc and Horschke, 1996). Thymidylate kinase can phosphorylate dTMP to produce dTDP, which can be phosphorylated into dTTP (Zupanc and
Horschke, 1996). Cytidine can be converted by cytidine deaminase into uridine (Lane and Fan, 2015). Uridine-cytidine kinase can phosphorylate cytidine and produce CMP (Lane and Fan, 2015). CMPK1 can phosphorylate
CMP and produce CDP which can be phosphorylated by nucleoside diphosphate kinase to produce CTP (Zauri et al., 2015). Cytidine and 21 deoxycytidine can be salvaged via the enzyme cytidine deaminase
(Chabosseau et al., 2011; Lane and Fan, 2015; Zupanc and Horschke, 1996).
Regulation of pyrimidine synthesis
Pyrimidine synthesis can be regulated on the level of CAD gene expression (Evans and Guy, 2004). The expression of CAD is controlled at both the transcriptional and posttranscriptional level (Evans and Guy, 2004).
Expression increases when cells go from the resting phase into the proliferative phase (Evans and Guy, 2004).
The synthesis of pyrimidines is also subject to metabolic control via feedback inhibition. UTP inhibits the CPSII domain of the trifunctional enzyme CAD (which consists of carbamoyl-phosphate synthetase, aspartate transcarbamylase, and dihydroorotase) (Liao et al., 1986; Liu et al., 1994; Ng et al., 2015). On the other hand, PRPP allosterically activates CPSase activity (Evans and Guy, 2004). CTP inhibits the ATCase portion of CAD by binding to it and inducing a conformational change to an inactive form of the enzyme (Lehninger, 1993). ATP counteracts this CTP induced conformational shift (Lehninger, 1993). It has been found that CAD activity is also regulated by phosphorylation of PKA at two separate sites on the enzyme (Carrey et al.,
1985). Phosphorylation results in the activation of the carbamoyl phosphate synthetase activity by decrease the ATP Km (Carrey et al., 1985). UTP 22 inhibition is also greatly reduced as a result of this phosphorylation (Carrey et al., 1985).
The reduction of ribonucleotides to deoxyribonucleotides is regulated allosterically (Lehninger, 1993). There are two binding sites on the enzyme— one site regulates the overall activity of the enzyme, while the second site regulates the substrate specificity (Lehninger, 1993). The binding of dATP to the activity regulation site actually reduces the activity of ribonucleotide reductase because it signals that there is an excessive amount of deoxyribonucleotides (Lehninger, 1993). On the other hand, the binding of
ATP signals just the opposite, resulting in increased activity (Lehninger,
1993). The binding of ATP or dATP to the substrate specificity site increases the reduction of the pyrimidine nucleotides UDP and CDP (Lehninger, 1993).
TTP binding facilitates GDP reduction (Lehninger, 1993). As the amount of dGTP increases, this in turn will stimulatinke ATP reduction to dATP
(Lehninger, 1993).
Catabolism of pyrimidine nucleotides
The pyrimidine ring, unlike the purine, can be cleaved, opened, and degraded to structures that are easily absorbed by the bloodstream (Champe,
1994; Garrett, 1995). Pyrimidines are typically broken down by nucleotidases and nucleoside phosphorylases (Fasullo and Endres, 2015). These enzymes remove the base from the sugar (Champe, 1994; Fasullo and Endres, 2015). 23
Within mammalian cells, cytosine is broken down into uracil and then uracil is broken down by reduction to dihydrouracil (Canellakis, 1956; Fink et al.,
1956a; Fink et al., 1956b). The ring is then opened resulting in β- ureidopropionase acid and then decarboxylation occurs which produces NH3,
β-alanine, and CO2 (Canellakis, 1956; Champe, 1994; Fink et al., 1956a; Fink et al., 1956b; Garrett, 1995). β-alanine can be recycled and used as a precursor for acetyl-CoA synthesis (Champe, 1994; Garrett, 1995). This breakdown mainly occurs within the liver (Canellakis, 1956; Dagg et al.,
1964). Thymidine is broken down into aminoisobutyric acid, which can be recycled as a precursor for succinyl CoA synthesis (Champe, 1994; Garrett,
1995; Nielsen et al., 1974).
Pyrimidine nucleotide salvage deficiencies
There are a number of diseases that result from mutations in enzymes involved in the salvage of thymidine such as thymidine kinase II and thymidine phosphorylase (Ferraro et al., 2006). Thymidine kinase is necessary for mitochondrial dNTP synthesis (Fasullo and Endres, 2015).
Mutations in this gene cause muscular dysfunction which often results in death by respiratory failure as an infant (Fasullo and Endres, 2015; Kalko et al., 2014; Tyynismaa et al., 2012). Muscle cells require a lot of functioning mitochondria because of the large energy expenditure (Gouspillou and
Hepple, 2016). In addition to this, mitochondria are involved in the 24 regulation of a number of different processes in the muscle such as metabolism, energy sensitive signaling pathways, reactive oxygen species production and signaling, calcium homeostasis and apoptosis regulation
(Gouspillou and Hepple, 2016). Therefore, any mitochondrial dysfunction would largely manifest within the muscle tissue.
De novo pyrimidine nucleotide synthesis deficiencies
The rare genetic disorder called Miller syndrome can be traced to a mutation in dihydroorotate dehydrogenase (DHODH), an enzyme involved in the fourth step of pyrimidine synthesis, which occurs in the mitochondria (Ng et al., 2015). This disease is characterized by abnormalities of the ear and eye, micrognathia, skeletal abnormalities, developmental issues, and gastrointestinal irregularities (Ng et al., 2015). DHODH plays a role in NF-
кB activity, which is important during development (Kinoshita et al., 2011).
It is believed that the malfunctioning enzyme interferes with NF-кB during development and results in abnormal cell migration, reduced cellular proliferation, and an abnormally high amount of apoptosis (Kinoshita et al.,
2011). This then results in the physical abnormalities characteristic of this syndrome (Kinoshita et al., 2011).
Abnormalities in uridine monophosphate synthetase (UMPS), another gene involved in de novo pyrimidine nucleotide synthesis, result in hereditary orotic aciduria (Fasullo and Endres, 2015). This disease is characterized by 25 growth deficiency, intellectual impediments, megaloblastic anemia and the excretion of orotic acid in the urine (Ng et al., 2015). Symptoms can be relieved by administration of uridine (Becroft and Phillips, 1965; Haggard and Lockhart, 1967).
De novo pyrimidine nucleotide synthesis and immunodeficiency
Both B and T lymphocytes primarily depend on de novo nucleotide synthesis (Quemeneur et al., 2003) due to a lack of the salvage pathway enzyme HGPRT (Busuttil, 2014). It thus comes as no surprise that if key enzymes involved in de novo nucleotide synthesis are non-functioning or inhibited, impaired immune function results from lack of lymphocyte proliferation. It has been found that a lack of pyrimidines, in particular, induces apoptosis in activated lymphocytes (Quemeneur et al., 2003). A point mutation in one rate limiting enzyme involved in de novo nucleotide synthesis, CTP synthase 1 (CTPS1), causes immunological dysfunction
(Kucuk et al., 2016; Martin et al., 2014). The mutation affects a splice donor site resulting in the production of an abnormal transcript lacking exon 18
(Martin et al., 2014). This deleterious mutation results in CTPS1 not being expressed in the B and T cells of these patients (Martin et al., 2014). In the patients studied, frequent and severe infection, both bacterial and viral, were observed (Kucuk et al., 2016; Martin et al., 2014). Most of the viral infections could be traced to herpes viruses such as Epstein-Barr virus, and varicella 26 zoster virus (Kucuk et al., 2016; Martin et al., 2014). As a result of the
Epstein-Barr virus, some patients eventually developed non-Hodgkin lymphoma and some even died (Martin et al., 2014).
Inosine monophosphate (IMPDH) and de novo nucleotide synthesis
The IMPDH reaction
IMPDH is responsible for taking IMP, which is the common precursor of both GTP and ATP, and catalyzing the first committed step to GTP synthesis: the conversion of IMP into xanthosine monophosphate (XMP)
(Figure 2) (Hedstrom, 2009). This process involves the catalysis of two different steps: first there is a dehydrogenase reaction which forms NADH and a covalent intermediate (E-XMP*) (Hedstrom, 2009). Second, there is a hydrolysis reaction in which E-XMP* is cleaved to release XMP (Hedstrom,
2009). This sends IMP down the arm of the pathway which creates the guanosine nucleotides. In this way, IMPDH is the gateway to guanosine nucleotide production. This de novo GMP producing pathway appears to exist in virtually every organism except for protozoan parasites like Giardia lamblia and Trichomonas vaginalis (Carlton et al., 2007; Morrison et al.,
2007). Both Trichomonas vaginalis and Giardia lamblia appear to lack the 27 ability to perform de novo nucleotide synthesis and instead rely on the salvage pathway (Jarroll et al., 1989; Smyth, 1995).
The structure of IMPDH
Generally, IMPDH monomers consist of two domains: the catalytic and the regulatory domain (Figure 5A)
A B C
Figure 5. Crystal structure of IMPDH. A) Monomer of IMPDH. In purple is the catalytic domain, in mustard is the CBS domain. B) Tetramer of IMPDH, each color depicts a single monomer. C) Octameric subunit of IMPDH, each color depicts one of the two tetramers. PDB ID 1NFB.
(Hedstrom, 2009; Makowska-Grzyska et al., 2012). The regulatory domain consists of a tandem CBS pair forming what is called a Bateman domain
(Kemp, 2004). The Bateman domain is not essential for IMPDH activity and there are a few organisms with IMPDH that does not contain them such as B. 28 burgdorferi, the parasite that causes Lyme disease in humans, and C. parvum, a parasite of the mammalian digestive tract (Hedstrom, 2009).
These monomers come together into tetramers (Figure 5B) (Hedstrom, 2009;
Sintchak et al., 1996). Each IMPDH tetramer has a central square-like region which is where the catalytic domains are located, while the CBS domains project from the tetramer corners, away from the catalytic domain (Figure
5B) (Hedstrom, 2009). The link between the catalytic domain and the subdomain is flexible and therefore these domains can take on various relative orientations which can vary by as much as 120 degrees (Colby et al.,
1999; Hedstrom, 2009). Labesse et al. (Labesse et al., 2013) discovered that in
Pseudomonas aeruginosa and human IMPDH1, two tetramers can interact to form octamers (Figure 5C). The octameric structure of Pseudomonas aeruginosa was recently solved by X-ray crystallography (Labesse et al.,
2013). Analytical ultracentrifugation also revealed these IMPDHpa to have a sedimentation coefficient that corresponds to an octamer (Labesse et al.,
2013). Size exclusion chromatography has also verified this octameric structure (Labesse et al., 2013). Furthermore, purified human IMPDH1 was revealed to form octamers (Labesse et al., 2013). Human IMPDH2 has been found to form octamers as well, but this octameric state has not been found for IMPDHs from all organisms (Alexandre et al., 2015). Alexandre et al.
(Alexandre et al., 2015) studied IMPDH from a number of different human pathogenic bacterial species. They found that there were some species with 29 octameric IMPDH, which they refer to as Class I IMPDHs (Alexandre et al.,
2015). But there were other species of bacteria that had tetrameric IMPDH
(Alexandre et al., 2015). These were referred to as Class II IMPDHs
(Alexandre et al., 2015).
The role of the Bateman domain and nucleotides in the regulation of
IMPDH activity
The Bateman domains are extremely conserved, being found in many
Figure 6. Octamers take on different conformations based on the type of nucleotide that binds. The binding of GTP to the Bateman domain results in a “closed” confirmation where the central catalytic domains are close to each other. This brings two finger domains together, resulting in decreased enzymatic activity. The binding of ATP on the other hand results in an “open” octamer state.
different types of proteins in a wide variety of organisms from archaebacteria 30 to eukaryotes (Ignoul and Eggermont, 2005; Thomas et al., 2012). They are known for their nucleotide binding properties such as those found in the master regulator of cellular energy—AMP kinase (AMPK) (Ignoul and
Eggermont, 2005). AMPK is activated by binding AMP, which signals low cellular ATP levels and thus low energy levels (Ignoul and Eggermont, 2005;
Xiao et al., 2013; Xiao et al., 2011). This results in the activation of catabolic pathways such as fatty acid oxidation and the halting of anabolic pathways such as fatty acid and cholesterol synthesis (Ignoul and Eggermont, 2005;
Mihaylova and Shaw, 2011). The importance of this domain is demonstrated by the fact that mutations in the Bateman domain have been associated with a number of diseases (Thomas et al., 2012) such as Wolff-Parkinson-White syndrome (which occurs due to mutations in AMPK’s Bateman domain), congenital myotonia (Bateman domain of CLCN1 chloride channel), and homocystinuria (Blair et al., 2001; Pusch, 2002; Shan et al., 2001). A missense mutation in IMPDH’s Bateman domain causes Leber congenital amaurosis and retinitis pigmentosa (Buey et al., 2015). These domains have also been associated with the maintenance of balanced nucleotide pools within the cell (Pimkin and Markham, 2008; Pimkin et al., 2009). It has been found that if this domain is deleted in bacterial IMPDH, the cell is no longer capable of maintaining balanced levels of ATP and GTP (Pimkin and
Markham, 2008; Pimkin et al., 2009). And although the Bateman domain is not necessary for IMPDH activity, it has been found that the binding of 31 nucleotides such as GTP and ATP to this domain of IMPDH in certain organisms influences activity (Buey et al., 2015; Labesse et al., 2013).
Labesse et al. (Labesse et al., 2013) found that ATP binding to two sites in the Pseudomonas aeruginosa CBS domain of IMPDH significantly increased
IMPDH activity, while ATP binding to human IMPDH1 had no effect on activity. Furthermore, they found that the binding of ATP induced a conformational change in the octamer, forcing it to take on an “open” conformation, where the enzymatic domains of the two tetramers are far apart (Labesse et al., 2013). Interestingly, guanine nucleotide binding has also been found to induce a conformational change but this results in inhibition of IMPDH activity in the eukaryotic protein (Buey et al., 2015).
Rather than the open octamer that ATP binding induces, guanine nucleotide binding induces a closed octamer conformation (Buey et al., 2015). This closed octamer brings the finger domains, which are four-stranded beta sheets projecting from the IMPDH monomers, into contact with each other
(Buey et al., 2015). This is believed to result in the protein losing its affinity for the substrate IMP thus leading to a loss of enzymatic activity (Buey et al.,
2015).
Activation of IMPDH by monovalent cations
It has been found that all IMPDHs are activated approximately hundredfold by potassium and other similarly sized monovalent cations such 32 as NH4+ or Rb+ (Xiang et al., 1996). Some IMPDHs are activated or inhibited by smaller ions like Na+ (Hedstrom, 2009; Heyde and Morrison, 1976; Xiang et al., 1996; Zhou et al., 1997). Human IMPDH2 is activated by K+, NH4+, Na+ and Li+ (Hedstrom, 2009; Heyde and Morrison, 1976; Xiang et al., 1996; Zhou et al., 1997). K+ does not appear to affect tetramer stability (Hedstrom, 2009;
Heyde and Morrison, 1976; Xiang et al., 1996; Zhou et al., 1997). K+ activates
IMPDH by acting as a “ball and socket” joint that expedites the conformational changes necessary to complete the catalytic cycle (Riera et al.,
2011).
Isoforms of IMPDH
In mammalian cells, there are two isoforms: IMPDH1 and IMPDH2.
These are encoded by two separate genes (Glesne et al., 1993; Gu et al., 1994;
Zimmermann et al., 1995). Eighty-four percent of the amino acids are identical between these two isoforms, but they are expressed differently in various tissues of the body (Gu et al., 1997). IMPDH2 is dominant in most tissues and has increased expression in proliferating cells (Gu et al., 1997;
Zimmermann et al., 1995). On the other hand, IMPDH1 generally has a low expression but is highly expressed in certain tissues like the brain, kidney, spleen, and pancreas (Gu et al., 1997). In the rat retina, it was found that
IMPDH2 is highly expressed during development, but after eye opening
IMPDH1 is highly expressed (Gunter et al., 2008; Thomas et al., 2012). The 33 expression of IMPDH1 is also dominant in the retina of adult humans
(Gunter et al., 2008). In addition to this, human photoreceptors have novel
IMPDH1 mRNAs which are a product of alternative splicing (Bowne et al.,
2006; Xu et al., 2008).
Moonlighting functions of IMPDH
Using Drosophila as the model organism, IMPDH cellular localization was found to be regulated by the cell cycle (Kozhevnikova et al., 2012). It was found to localize to the nucleus during the end of the S and throughout the
G2 phases (Kozhevnikova et al., 2012). Kozhevnikova et al. (Kozhevnikova et al., 2012) also found that oxidative stress results in the nuclear localization of
IMPDH. Via ChIP-qPCR it has been discovered, that IMPDH binds to all five histone genes, E2F, and various other genes, within the nucleus
(Kozhevnikova et al., 2012). These genes possess a variety of functions which include transcriptional regulation, signaling and development (Kozhevnikova et al., 2012). The knockdown of IMPDH results in a significant increase in the expression levels of histone mRNA and protein levels (Kozhevnikova et al., 2012). This depletion of IMPDH also results in the upregulation of the other IMPDH target genes (Kozhevnikova et al., 2012). These results suggest that IMPDH is acting as a transcriptional repressor (Kozhevnikova et al.,
2012). In addition, the Bateman domains of IMPDH have been found to bind to CT rich, single stranded regions within the genome (Kozhevnikova et al., 34
2012). Mutations in the Bateman domain that are associated with retinitis pigmentosa have been shown to abrogate the ability of IMPDH to bind to these CT rich single stranded DNA regions (Mortimer and Hedstrom, 2005;
Mortimer et al., 2008; Spellicy et al., 2007). These CT rich regions of the genome have a tendency to unwind into single strands and are associated with gene control (Kozhevnikova et al., 2012). IMPDH1 binds to the upper, single strand CT-rich regions within the loci of genes that it regulates such as the His3-His4 promoter, Mlc2 and E2F (Kozhevnikova et al., 2012).
IMPDH has also been found to associate with polyribosomes, which suggests that it could play some role in translation regulation (Cornuel et al.,
2002; McLean et al., 2004; Spellicy et al., 2007). Retinal IMPDH1 associates with rhodopsin mRNA-translating polyribosomes (Mortimer and Hedstrom,
2005; Mortimer et al., 2008; Spellicy et al., 2007). If there is any interference of rhodopsin expression, apoptosis of photoreceptor cells occurs (Mortimer and Hedstrom, 2005; Mortimer et al., 2008). In light of this, it is unsurprising that mutations in IMPDH1 and not IMPDH2 result in retinal diseases.
IMPDH associates with a number of different proteins including protein kinase B, glutamate dehydrogenase, and proteins involved in transcriptional regulation (Ingley and Hemmings, 2000). It has also been observed that the expression of the tumor suppressor p53 attenuates IMPDH activity (Ingley and Hemmings, 2000; Sherley, 1991).
35
Therapeutic targeting of IMPDH activity
Because pathogenic bacteria and viruses often rely on their rapid proliferation for successful infection, as a rate-limiting enzyme, IMPDH is an attractive therapeutic target. In addition, the reliance of lymphocytes on the de novo pathway makes the inhibition of IMPDH an effective immunosuppressant (McKeran and Watts, 1976; Quemeneur et al., 2003;
Sintchak et al., 1996). IMPDH1 and 2 gene amplifications and increased mRNA expression have been found in a variety of different cancers including, but not limited to, brain cancer, ovarian cancer, melanomas, cancer of the bladder, renal cell carcinomas, and uterine cancers (Konno et al., 1991; Majd,
2014; Nagai et al., 1991; Natsumeda et al., 1990). IMPDH inhibitors such as mycophenolic acid (MPA), ribavirin, and others are currently being used, are in clinical trials, or are beginning to be explored for use as clinical therapeutics for infections, immunosuppressants, and neoplastic disease
(Yokota, 2002).
MPA
MPA was first isolated and purified from spoiled corn (Bentley,
2000). It had demonstrated an anti-microbial ability as it was first found to inhibit the growth of Bacillus anthracis (Bentley, 2000). Because MPA is a strong inhibitor of mammalian IMPDH as well, it has not been used as an antibiotic (Hedstrom, 2009). MPA inhibits IMPDH by trapping the covalent 36 intermediate (E-XMP*) in the IMPDH reaction (Fleming et al., 1996). It wasn’t until many years after its discovery that MPA made its way into the clinic as an immunosuppressant drug, primarily used to prevent the rejection of transplanted organs in patients (Hedstrom, 2009). As lymphocytes of the immune system are dependent on de novo guanine nucleotide synthesis for their function (Quemeneur et al., 2003), the inhibition of IMPDH was a feasible way to achieve this goal. MPA has also been used for psoriasis treatment (Hedstrom, 2009). In addition, there is now interest in MPA as a potential anti-cancer treatment because of its antiangiogenic activity: it has been found to induce differentiation, replication arrest, and/or cell death in many different cancer cell lines such as breast, prostate, melanoma, leukemia, and neuroblastoma (Bacus et al., 1990; Chen and Pankiewicz,
2007; Floryk and Huberman, 2006; Floryk et al., 2004; Kiguchi et al., 1990;
Messina et al., 2004). In vitro studies have demonstrated that a lot of cancer cells produce GTP via the de novo nucleotide synthesis pathway by upregulating IMPDH (Natsumeda et al., 1984; Tong et al., 2009). GTP levels are also more elevated than those of ATP in many different cancer types
(Traut, 1994). Therefore, this antitumor effect of MPA is believed to be linked to its inhibition of IMPDH (Majd et al., 2014).
37
Ribavirin
Ribavirin, another IMPDH inhibitor, is a nucleoside analog (Witkowski et al., 1972). Ribavirin is one of the treatments given to Hepatitis C patients
(Gish, 2006). It is also used to treat respiratory syncytial virus (Wyde, 1998), and Lassa fever virus infections (McCormick et al., 1986). Once ribavirin is taken into the cell, it is activated by adenosine kinase and becomes ribavirin
5’-monophosphate (RVP) (Gish, 2006). In this form, it is an extremely effective inhibitor of IMPDH. However, RVP is then transformed even further. It becomes a ribavirin triphosphate (RTP) and, in this form, it is an inhibitor of RNA capping enzymes (Smith RA, 1980). As a triphosphate, it also has the ability to inhibit virus RNA polymerases (Cassidy and Patterson,
1989; Eriksson et al., 1977) and RTP can be used by viral RNA polymerase and become a part of the RNA (Crotty et al., 2000). Because ribavirin inhibits
IMPDH as well, there is a reduction of wild type competitor nucleotides and an increase in RTP incorporation (Crotty et al., 2000). Once incorporated, the base changes conformation from syn to anti and becomes locked in that form
(Crotty et al., 2000). In this form, it easily forms base pairs with incoming pyrimidines, which results in deleterious mutations, bringing the virus to
“error catastrophe” (Crotty et al., 2000). It is believed to be a combination of all of these properties that contribute to the anti-viral properties of ribavirin
(Graci and Cameron, 2006; Parker, 2005).
38
Other inhibitors of IMPDH
Another IMPDH inhibitor called mizoribine has been used as a drug for immunosuppression and rheumatoid arthritis (Ishikawa, 1999). It inhibits the proliferation of T and B cells (Yokota, 2002). Mizoribine is an imidazole nucleoside (Yokota, 2002). Once taken into cells, it is activated by phosphorylation by adenosine kinase and then inhibits IMPDH and GMP synthetase (Yokota, 2002).
Phenyl-oxazole urea inhibitors, similar to MPA, trap the covalent intermediate E-XMP*. Another such inhibitor, AVN944, has been found to have anti-proliferative effects against prostate cancer cell lines by triggering the arrest of the cell cycle in the S phase and also by inducing differentiation and apoptosis (Yokota, 2002).
Tiazofurin is a nucleoside analog that has strong antiviral and antitumor effects (Marquez et al., 1986). In vitro studies have shown it to be effective against myelogenous leukemia cells (Marquez et al., 1986).
Tiazofurin is phosphorylated by adenosine kinase, nicotinamide riboside kinase and/or 5’-nucleotidase, becoming a monophosphate (Fridland et al.,
1986; Saunders et al., 1990). The monophosphate is then transformed into the adenine dinucleotide TAD which can then bind and inhibit the E·IMP and E-XMP* forms of IMPDH (Hedstrom, 2009). The amount of TAD that builds up is what determines the effectiveness of tiazofurin (Hedstrom, 2009). 39
Benzamide riboside, like tiazofurin is converted into dinucleotide once inside of the cell (Hedstrom, 2009). In this form, it acts as a potent IMPDH inhibitor as is evident in the fact that it depletes guanine nucleotide pools within the cell (Hedstrom, 2009).
In a screen for natural, non-synthetic inhibitors of IMPDH, it was discovered that some unsaturated fatty acids inhibit both mammalian
IMPDH isoforms including linoleic acid and eicosadienoic acid (Mejia et al.,
2013; Mizushina et al., 2007).
IMPDH as filamentous structures
Physical characteristics of IMPDH
filaments
IMPDH is capable of assembling
itself into large, filamentous structures Figure 7. IMPDH2 filaments. Hek293 cells treated with 100nM within the cell (Figure 7). These filaments
methotrexate overnight then fixed are mostly cytoplasmic in location but can and stained with an anti-IMPDH2 antibody (green). One elongated also be found within the nucleus (Carcamo filament is indicated by the white et al., 2014; Carcamo et al., 2011; Juda et arrow, and a loop shaped filament is indicated by an orange al., 2014). Typically, there are only one or arrowhead. DAPI was used to visualize the nuclei (blue). two rods within a single cell (Carcamo et 40 al., 2011). The filaments can take on an elongated form (Figure 7, white arrow), as well as elongated rings (Figure 7, orange arrowhead) and circular or “hairpin” structures (Carcamo et al., 2011). Electron microscopy shows that the filaments are not encased within a membrane and therefore likely consist purely of protein (Juda et al., 2014). A single filament consists of individual fibers of protein with repeating subunits that are about 11 microns in length (Juda et al., 2014). Within mammalian cells, these filamentous structures are not associated with the Golgi complex, and do not contain proteins such as actin, tubulin, or vimentin (Carcamo et al., 2011; Ji et al.,
2006). These structures have also been ruled out as being primary cilia
(Carcamo et al., 2011; Willingham et al., 1987).
History of the discovery of IMPDH filaments
Filamentous IMPDH structures within cells were first observed in
1987 (Willingham et al., 1987). Using a monoclonal antibody, (Willingham et al., 1987) were able to identify filamentous, thread-like structures within several different cell types such as mouse and rat cell lines. The researchers referred to these structures as nematin (Willingham et al., 1987). Using double label immunofluorescence to anti-nematin and anti-tubulin or anti- vimentin, they found that nematin did not co-localize with cilia or vimentin
(Willingham et al., 1987). Immunoprecipitation was performed in an attempt to identify the protein with which the antibody was reacting (Willingham et 41 al., 1987). A 58 kD protein was immunoprecipitated (Willingham et al., 1987).
Today, it is suspected that the 58 kD protein immunoprecipitated by this group is IMPDH, because of the similarity in molecular weight.
Interestingly, (Ji et al., 2006) discovered that treatment with MPA triggers IMPDH to assemble into filaments. They found this to be the case in both normal and neoplastic cells (Ji et al., 2006). They hypothesized that filament assembly occurred as a result of MPA binding to IMPDH and causing a conformational change (Ji et al., 2006). However, the filament assembly was reversed upon treatment with GTP and partially reversed upon treatment with ATP, suggesting that assembly was associated with a reduction of purine nucleotide levels within the cell (Ji et al., 2006). The aggregation of IMPDH was also discovered to occur with purified protein (Ji et al., 2006). They found that IMPDH filaments did not co-localize with mitochondria, Golgi apparatus, endoplasmic reticulum, lipid bodies, or cytokeratins (Ji et al., 2006). In addition, just like the previously mentioned group, no other interacting proteins were found when co-immunoprecipitation and tandem affinity purification with tagged IMPDH were performed (Ji et al., 2006).
Two years later, (Gunter et al., 2008) explored the subcellular distribution of transfected IMPDH protein. Like (Ji et al., 2006) they found that treatment with IMPDH inhibitors resulted in the assembly of IMPDH2 into filaments, but unlike (Ji et al., 2006), they did not believe that this was 42 due solely to a conformational change induced by the binding of inhibitor
(Gunter et al., 2008). This was supported by their finding that when untreated CHO cells were transiently transfected with HA tagged IMPDH2, the recombinant protein had a diffuse pattern, while HA tagged IMPDH1 assembled into filaments (Gunter et al., 2008). In addition to this, when they inhibited the enzyme that acts just downstream of IMPDH, GMP synthetase, they found that IMPDH assembled into filaments (Gunter et al., 2008).
Furthermore, they quantified GTP levels after MPA treatment using HPLC and found that cellular GTP levels dropped by more than 80% (Gunter et al.,
2008). Based on these results, they concluded that filament assembly was tied to a lack of GTP nucleotides within the cell as opposed to a conformational change induced by drug binding (Gunter et al., 2008). And because uninhibited IMPDH still localized to the filament upon GMP synthetase inhibition, this also suggested that polymerized IMPDH could be enzymatically active.
Interestingly, (Carcamo et al., 2011) discovered filaments within HEp-
2 cells when they were stained with human autoantibodies from HCV patients who were receiving ribavirin treatment. They referred to these autoantibodies as “anti-Rods and Rings” (anti-RR) autoantibodies (Carcamo et al., 2011). As other groups before them, they found that these structures did not contain vimentin nor did they co-localize with cilia (Carcamo et al.,
2011). Furthermore, these structures did not contain tubulin or actin 43
(Carcamo et al., 2011). Using immunoprecipitation analysis, it was found that the autoantibody pulled down a 55 kD protein (Carcamo et al., 2011).
Farther analysis using immunoprecipitation-Western blot confirmed that this immunoprecipitated 55 kD protein was IMPDH (Carcamo et al., 2011). In addition to this, co-staining of cells with this autoantibody and anti- IMPDH showed that both co-localized to the same structure (Carcamo et al., 2011).
Thus, IMPDH2 was determined to be one of the proteins contained within these structures (Carcamo et al., 2011).
Regulation of IMPDH filament assembly
With few exceptions, IMPDH filaments do not assemble spontaneously within cultured mammalian cells and must be induced (Carcamo et al., 2011).
It has been observed that the assembly of IMPDH into these filaments occurs in response to a drop in purine nucleotide levels within the cell (Calise et al.,
2016). This drop in purine levels can occur by perturbation at different points along the purine nucleotide biosynthesis pathway such as by inhibition of
IMPDH. Case in point, these filaments have been found in peripheral blood mononuclear cells isolated from patients being treated with either MPA or ribavirin (Keppeke et al., 2016). Immunofluorescence and immunoprecipitation has been used to determine that the primary target of these autoantibodies is IMPDH (Carcamo et al., 2011; Keppeke et al., 2016).
Autoantibodies against IMPDH are seen in 6% of HCV patients after only one 44 month of treatment with ribavirin. But by the sixth month, greater than 47% of the patients have IMPDH autoantibodies (Keppeke et al., 2016). The amount of the autoantibody reaches its highest after twelve months of treatment with ribavirin (Keppeke et al., 2016). For unknown reasons, patients undergoing MPA treatment are much less likely to develop these autoantibodies than patients being treated with ribavirin (Keppeke et al.,
2016).
Methotrexate (MTX), a drug often used for cancer treatment, triggers the assembly of IMPDH into filaments within cultured mammalian cells
(Keppeke et al., 2016). Anti-RR autoantibodies have also been found in a patient undergoing MTX treatment for rheumatoid arthritis (Keppeke et al.,
2016). MTX perturbation occurs at a point earlier than IMPDH along the purine synthesis pathway. MTX is an anti-folate. It inhibits the enzyme dihydrofolate reductase (DHFR), which is necessary for the conversion of dihydrofolate into tetrahydrofolate (THF) as shown in Figure 8 (Goodsell,
1999; Rajagopalan et al., 2002). THF donates one carbon units for the production of purine nucleotides (Figure 8) (Yang and Vousden, 2016).
Aminopterin (AMT), another inhibitor of DHFR, has an even higher affinity for DHFR than MTX (Calise et al., 2016). 45
Figure 8. One carbon metabolism in nucleotide synthesis. Dietary folate is ingested and reduced to dihydrofolate. Dihydrofolate is then converted to tetrahydrofolate (THF) by the enzyme dihydrofolate reductase (DHFR). Serine hydroxymethyltransferase (SHMT) then converts THF to 5, 10-methylene-THF which is converted to 5, 10-methenyl-THF. 5, 10- methenyl-THF contributes one carbon unit to the purine ring. 5, 10-methenyl-THF is then converted to 10-formyl-THF which donates a second one carbon unit to the purine ring.
Therefore, AMT induces filament assembly more potently than MTX
(Calise et al., 2016). As is the case with the other modes of filament assembly,
MTX or AMT induced filaments can be inhibited from assembling or can be disassembled upon supplementation with guanosine which supports the idea that it is a drop in guanine nucleotide levels that triggers filament assembly
(Calise et al., 2016). In support of this idea, the addition of hypoxanthine to these MTX or AMT treated cells also prevented and disassembled IMPDH filaments (Calise et al., 2016). Hypoxanthine is converted in the salvage pathway to the precursor IMP which can then be used for GTP synthesis
(Calise et al., 2016; Champe, 1994; Garrett, 1995). Unsurprisingly, because
IMPDH is downstream of hypoxanthine in purine synthesis, if IMPDH is inhibited, hypoxanthine cannot prevent or reverse filament assembly (Calise et al., 2016). 46
It has also been found that IMPDH filaments assemble when cells are cultured in the standard culture media MEM which lacks serine and glycine
(Calise et al., 2016). So, this represents yet another perturbation along the purine biosynthesis pathway which can induce filament assembly. In the cell, an enzyme called serine hydroxymethyl transferase (SHMT) catalyzes a reversible reaction which takes serine and converts it into glycine
(Labuschagne et al., 2014). This reaction produces one carbon units which are then integrated into the THF cycle for use in nucleotide synthesis (Figure 8)
(Labuschagne et al., 2014). The addition of glycine alone to cells cultured in
MEM increases the amount of filaments assembled as compared to cells grown in just MEM (Calise et al., 2016). When Labuschagne et al.
(Labuschagne et al., 2014) traced the intracellular fate of exogenous radiolabeled glycine, they found that it did not enter the one carbon cycle for purine production; rather it was converted into serine. Mass spectrometry analysis of these glycine-fed cells also showed that the flux of glycine downstream to AMP, ADP, GMP, and GDP is reduced (Labuschagne et al.,
2014). This depletion of cellular purines could explain the increase in filament assembly because IMPDH filament assembly is linked to purine nucleotide depletion. On the other hand, it was found that the addition of serine alone to MEM cultured cells inhibited filament assembly (Calise et al.,
2016). The addition of serine would send the SHMT catalyzed reaction in the forward direction, towards the production of one carbon units, which are then 47 available for purine production (Figure 8) (Labuschagne et al., 2014).
Therefore, purine levels are maintained or increased within the cell and filaments need not assemble. On the other hand, the addition of glycine would send the reaction in the opposite direction, depleting one carbon units and thus reducing nucleotide production. This would explain the increased filament assembly upon glycine addition to MEM cultured cells.
The depletion of glutamine within a culture of mammalian cells also triggers a reversible assembly of IMPDH filaments (Calise et al., 2014).
Glutamine depletion demonstrates additional points of perturbation along the de novo purine biosynthesis pathway (Figure2) (Calise et al., 2014). That a lack of glutamine triggers filament assembly is unsurprising considering that there are three steps in purine synthesis that require the amine- nitrogen from glutamine to contribute to the purine ring. Therefore, without glutamine, there would be a lack of purine production, resulting in a purine deficit. Consistent with this idea, IMPDH filaments disassemble upon the addition of glutamine or guanosine to glutamine-depleted cells in culture
(Calise et al., 2014). It has been noted that the disassembly of filaments happens quite rapidly, within 15 min (Calise et al., 2014).
The triggering of filament assembly also occurs with purified IMPDH protein. Labesse et al. (Labesse et al., 2013) discovered that incubation of purified human IMPDH1 with ATP, triggers filament assembly.
48
The effect of IMPDH filament assembly on catalytic activity
The biological role that IMPDH filament assembly fulfills has remained elusive. Some have concluded that filament assembly was associated with some sort of change in the catalytic activity of the enzyme
(Labesse et al., 2013; Scott et al., 2004) while others have not seen an increase in enzymatic activity associated with filament assembly (Carr et al.,
1993; Labesse et al., 2013; Mortimer and Hedstrom, 2005; Thomas et al.,
2012). The question as to what the activity state of IMPDH is while in filament form will be addressed in chapter 3.
IMPDH can co-assemble with CTPS into filaments
CTPS, a rate limiting enzyme which catalyzes the conversion of UTP into CTP, is also capable of assembling with IMPDH into these filamentous structures (Figure 7) (Carcamo et al., 2011). Co-assembly of CTPS and
IMPDH was first discovered in mammalian cells (Carcamo et al., 2011).
Antibody staining of filaments in HEp-2 cells obtained by from INOVA
Diagnostics Inc. revealed that the filaments contained both IMPDH and
CTPS (Carcamo et al., 2014).
The co-assembly of IMPDH and CTPS usually only takes place under certain conditions (Keppeke et al., 2015). It has been found that if cells are treated with an inhibitor that perturbs both pyrimidine synthesis and purine synthesis such as 6-diazo-5-oxo-L-norleucine (DON), cells produce filaments 49 primarily containing IMPDH, filaments primarily containing CTPS and filaments with a mixture of both (Keppeke et al., 2015). The mixed filaments consist of filaments with various proportions of IMPDH and CTPS (Keppeke et al., 2015). The propensity to assemble mixed filaments upon DON treatment also appears to differ from cell type to cell type (Keppeke et al.,
2015). For example, HeLa cells are more likely to form mixed filaments than
COS-7 cells (Keppeke et al., 2015).
It is not clear whether IMPDH and CTPS are directly bound to each other when co-assembled in filaments. It is possible that an adaptor protein is present that links them (Aughey and Liu, 2015). But the fact that these proteins are so close to each other within these mixed filaments suggests that there may be some crosstalk between the pyrimidine and purine nucleotide pathways (Aughey and Liu, 2015). This crosstalk could involve the co- regulation of purine and pyrimidine nucleotide synthesis and/or regulation of the enzymatic activity of IMPDH and CTPS (Aughey and Liu, 2015). It is also possible that the assembly of IMPDH with CTPS within the filaments restricts the mobility of IMPDH and thus its ability to carry out certain functions that require it to relocate (Aughey and Liu, 2015). Such functions could include acting as a transcription factor within the nucleus (Aughey and
Liu, 2015). Because IMPDH acts as a transcriptional repressor for certain genes (Kozhevnikova et al., 2012), filament assembly may result in the increased expression of the genes that IMPDH regulates. 50
CTPS in pyrimidine synthesis
CTPS is a rate limiting enzyme that is involved in pyrimidine synthesis, namely, the production of CTP (Figure 4). The activity of CTPS was first described in 1955 (Endrizzi et al., 2004). This enzyme is a glutamine amidotransferase which catalyzes three reactions to convert UTP into CTP (Azzam and Liu, 2013; Endrizzi et al., 2004; Liu, 2010). The first reaction is an ATP dependent kinase reaction which occurs by the phosphorylation of UTP at the oxygen atom in position 4 (Figure 9, step 1).
This is then followed by a glutaminase reaction in which glutamine is converted into glutamate resulting in the release of ammonia (Figure 9, step
2). In vitro, ammonia can be used directly by CTPS (Endrizzi et al., 2004).
Finally, in a ligase reaction, the phosphate group on the oxygen is replaced by the ammonia (Figure 9, step 3) (Liu, 2010). CTPS is essential to all organisms from bacteria and yeast to mammals because it is responsible for producing the CTP nucleotide, and these cytosine nucleotides are needed for the production of the phospholipids of membranes (Goto et al., 2004; Han et al.,
2005).
51
Structural characteristic of CTPS
Human CTPS1 consists of 591 amino acids and has a molecular weight
UTP CTP
Figure 9. CTPS reaction. UTP is phosphorylated and then an amidotransferase reaction creates ammonia from glutamine. Ammonia is then sent tunneled through the inside of the enzyme to the synthetase domain where reacts with the intermediate and replaces the phosphate group becoming UTP.
of 67 kD. In its active form, CTPS is a tetramer (Figure 10B) made up of monomers containing two domains (Figure 10A & 10C) (Goto et al., 2004;
Habrian et al., 2016; Kursula et al., 2006). There is an N-terminal synthetase-amidoligase domain and a C-terminal glutaminase domain
(Habrian et al., 2016; Kursula et al., 2006). It is the glutaminase domain that detaches ammonia from glutamine and produces glutamate (Goto et al., 2004; 52
Kursula et al., 2006). The ammonia is then sent through a tunnel or channel from the glutaminase domain to the synthetase domain (Goto et al., 2004;
Habrian et al., 2016; Kursula et al., 2006). In the synthetase-amidoligase domain, the ligase reactions occur in which the ammonia replaces the phosphate that is currently attached to UTP and thus creates CTP (Kursula et al., 2006).
AA BB CC
Figure 10. Crystal structure of CTPS. Synthesis domain of CTPS (A) which tetramerize (B) to become the core of the enzyme, while the glutaminase domains (C) are on the outside of the tetramer, one projecting out from each of the four synthetase domains.
Isoforms and expression
Mammals have two isoforms of CTPS—CTPS1 and CTPS2 (Higgins et al., 2007). Both isoforms have a 74% amino acid identity (Kucuk et al., 2016).
Both isoforms are ubiquitously expressed although the role that each fulfills is not clear (Kucuk et al., 2016). Comparable expression levels of CTPS1 have been found in a variety of tissues including leukocytes, thymus, spleen, bone 53 marrow, intestine, colon, ovary, brain, heart, kidney, liver, lung, pancreas, skeletal muscle and naïve T-cells (Martin et al., 2014). In naïve T-cells,
CTPS2 expression is considerably higher than that of CTPS1 (Martin et al.,
2014). It has been discovered that mutations in CTPS1, which produce a non-functioning transcript, result in a severely impaired immune response even though CTPS2 is fully functional within the lymphocytes (Martin et al.,
2014). Normal T-cells which become activated have up to a twentyfold increase in the amount of CTPS1 expressed, as compared to naïve, inactivated T-cells (Martin et al., 2014). Although you also see an increase in
CTPS2 levels upon T-cell activation, the amount of increase is not anywhere near as much as what is seen with CTPS1 (Martin et al., 2014). This suggests that the functions of CTPS1 and CTPS2, at least in the case of immune function, are not interchangeable (Kucuk et al., 2016; Martin et al., 2014).
This supports the idea that within the T-cells, CTPS1 may be the primary isoform responsible for producing the bulk of CTP during the clonal expansion that occurs after T-cell activation.
The yeast cells Saccharomyces cerevisiae (S. cerevisiae) also have their own version of CTPS1 and CTPS2—URA7 and URA8 respectively (Higgins et al., 2007; Nadkarni et al., 1995). URA7 is more prevalently expressed than
URA8 and makes most of the CTP (Nadkarni et al., 1995; Park et al., 2003).
This is evident in the fact that a yeast cell containing only the URA8 gene produces only 36% of the CTP found in WT cells, while cells containing only 54 the URA7 gene maintained 78% (Nadkarni et al., 1995; Ozier-Kalogeropoulos et al., 1994). The fact that there is decrease in CTP within the cell upon deletion of either isotype, suggests some overlap in function, but the big difference in the percentage of decrease shows that they are not functionally identical. Nevertheless, these cells are capable of surviving if at least one functional CTPS is present (Nadkarni et al., 1995).
The regulation of CTPS activity
Allosteric regulation
Although the structural mechanisms concerning the regulation of
CTPS is not completely clear, it is known that all four nucleotides—UTP,
GTP, ATP, CTP—play a role in activating or inhibiting the enzyme (Habrian et al., 2016; Lunn et al., 2008). UTP and ATP prompt the active tetrameric form to assemble from the inactive dimers. Both UTP and ATP bind to an active site in the synthetase domain (Habrian et al., 2016; Lunn et al., 2008).
The inhibition of CTPS occurs by the feedback inhibition of its product CTP which binds on the synthetase domain (Habrian et al., 2016). CTP actually competes with UTP for binding (Habrian et al., 2016). GTP is necessary for the activation of glutamine hydrolysis, but at high concentrations, it can actually have an inhibitory effect (Habrian et al., 2016).
55
Regulation of CTP by post translational modifications
Both CTPS isoforms are also regulated by protein phosphorylation at several different sites (Kassel et al., 2010). Phosphopeptide analysis shows that protein kinase A, for example, phosphorylates human CTPS1 on many different serine residues (Han et al., 2005). Preliminary studies show that
CTPS2 is also phosphorylated by protein kinase A (Han et al., 2005). This phosphorylation was found to occur in S. cerevisiae cells expressing human
CTPS1 (Han et al., 2005) and it also occurred in vitro (Han et al., 2005). It was found that this phosphorylation correlated with an up to 2 fold increase in the cellular concentration of CTP (Han et al., 2005). The phosphorylation of the yeast version of CTPS1, URA7, is also regulated by protein kinase A, although unlike the human CTPS1, there is only one phosphorylation site for protein kinase A, Ser424 (Park et al., 1999).
In addition, in S. cerevisiae it was discovered that protein kinase C also phosphorylates purified CTPS on serine and threonine residues at multiple sites, such as Ser36, Ser330, Ser354 and Ser424 (Park et al., 2003). It was found that phosphorylation at one site on CTPS influenced the phosphorylation at another site in a hierarchical manner where multiple sites need to be phosphorylated before the kinase gains access to another site with a particular function (Park et al., 2003; Roach, 1991). Furthermore, phosphorylation at different sites had opposing effects on the enzymatic activity level (Park et al., 2003). It was found that phosphorylation at Ser36, 56
Ser354, and Ser454 resulted in elevated enzymatic activity (Park et al., 2003).
This increased activation was traced to an enhanced affinity of the enzyme for ATP and a reduced sensitivity to product inhibition by CTP (Park et al.,
2003). On the other hand, phosphorylation at Ser330 results in a reduction of activity (Park et al., 2003).
Interestingly, human CTPS1 becomes heavily phosphorylated in cells that are starved of serum overnight (Higgins et al., 2007). It was found that the kinase primarily responsible for this low serum phosphorylation is
Glycogen Synthase Kinase 3 (Higgins et al., 2007). The sites Ser574 and Ser575 were identified as being Glycogen Synthase Kinase 3 phosphorylation sites.
The phosphorylation of CTPS1 by Glycogen Synthase Kinase 3 under low serum conditions was associated with a decrease in CTPS1 enzymatic activity
(Higgins et al., 2007).
The targeting of CTPS activity as a therapeutic
Increased CTPS activity has been found within cancer cells (Hatse et al., 1999; Mandel et al., 1963; van den Berg et al., 1993). Under normal circumstances, CTPS is a tightly regulated enzyme because its product CTP is kept at low levels within cells (Barry et al., 2014; Hatse et al., 1999). CTPS activity has been found to be increased in hepatomas of slow, medium and rapid rates of growth (Kizaki et al., 1980) and a significant correlation was 57 found between the increased activity and the growth rate of the tumors and the transformational state of the cells (Kizaki et al., 1980). High concentrations of both uracil and CTP have been found in the peripheral blood cells of patients with acute lymphoblastic leukemia and non-Hodgkin lymphoma (van den Berg et al., 1993; Verschuur et al., 1998). This heightened amount of CTP was traced to increased CTPS expression (van den
Berg et al., 1993; Verschuur et al., 1998). CTPS activity has also been found to be elevated in colon cancer (Weber et al., 1980). Because of these strong links between high CTPS expression and cancer, inhibition of its activity as a treatment is of interest. In addition, the fact that CTPS plays a central role in the production of nucleic acids, phospholipids, and sialic acid, makes it a target of interest for treatment of viral and protozoal infections, as well as neoplastic diseases (McCluskey et al., 2016).
Cyclopentenylcytosine Triphosphate (CPEC) and c3Urd
CPEC is a carboxylic analog of cytidine and an extremely effective and specific competitive inhibitor of CTPS (Huang et al., 2004; Kang et al., 1989).
This analog diminishes CTP and dCTP pools within the cell (Huang et al.,
2004; Kang et al., 1989). CPEC has been demonstrated to inhibit leukemia cells originating from children with acute lymphocytic leukemia and acute myeloid leukemia (Huang et al., 2004; Kang et al., 1989). CPEC has also been effective against a variety of other cancer cells like such as tumor KB cells, 58
L1210 cells, HT-29 colon carcinoma cells, the promyelocytic leukemia cell like
HL60, SK-N-BE(2)c and SK-N-SH neuroblastoma cell lines, and glioblastoma cells (Huang et al., 2004; Kang et al., 1989). In addition to this, it has displayed in vivo antitumor activity in mice with L1210, P388 leukemias, B16 melanoma, and human colon carcinoma HT-29 tumors (Huang et al., 2004).
There is interest in using these drugs in combination with the chemotherapeutic agents ara-C and 5-aza-2-deoxycytidine because of the apparent synergistic effect that CPEC and c3Urd has with them (Hatse et al.,
1999).
Gemcitabine
Because CTPS binds a number of different nucleotides, nucleotide analogs have been explored as inhibitors for CTPS1 (McCluskey, 2015;
McCluskey et al., 2016). Gemcitabine is one such analog. This drug, which has the ability to induce cellular apoptosis, is currently being used as a drug treatment for a number of different solid tumors (Doi et al., 2017; McCluskey et al., 2016; Ulrickson et al., 2017).
Once gemcitabine is taken into the cell via a nucleoside transporter, it is converted into 5’-monophosphate (dF-dCMP) by deoxycytidine kinase and thymidine kinase 2, and then into 5’ diphosphate (dF-dCDF) (McCluskey et al., 2016; Mini et al., 2006). In this form, it inhibits ribonucleotide reductase 59 and depletes nucleotide pools (McCluskey et al., 2016). Further phosphorylation then creates the 5’ triphosphate (dF-dCTP) which is the most plentiful metabolite of them all, making up 85-90% of gemcitabine species within the cell (McCluskey et al., 2016). This triphosphate is then mostly taken into genomic DNA although a small amount is incorporated into
RNA and as a result, DNA polymerases are inhibited (McCluskey et al., 2016;
Mini et al., 2006). But the free form of this triphosphate that does not get incorporated into DNA or RNA can actually interact with a number of CTP binding enzymes such as CTP:phosphocholine cytidylyltransferase
(McCluskey et al., 2016). This then converts it to the cytotoxic gemcitabine- choline. The deamination of this then produces 2’,2’-difluoro-2’-deoxy-UMP
(dF-dUMP), which is much less toxic than gemcitabine, which inhibits thymidylate synthase and reduces dTMP pools (McCluskey et al., 2016).
What is interesting is that CTP pools have also been found to be depleted as a result of gemcitabine (McCluskey et al., 2016). The cause for this has been traced to the fact that dF-dUTP is a CTPS substrate and CTPS catalyzes the transformation of dF-dUTP into dF-dCTP (McCluskey et al.,
2016). This product then binds strongly to CTPS, with a 37 fold higher affinity than UTP binds, and potently inhibits CTPS activity (McCluskey et al., 2016). The fact that CTPS can regenerate dF-dCTP levels from dF-dUTP also results in a higher concentration of active drug within patients
(McCluskey et al., 2016). In addition, the potency of the drug is also increased 60 by the fact that dF-dCTP inhibits CTPS, thus depleting competing CTP pools within the cell (McCluskey et al., 2016).
Azaserine
“Glutamine addiction” describes a state of glutamine dependence that has been found in a number of cancers such as that of the lung, leukemias, lymphomas, multiple myeloma, triple negative breast cancer, and pancreatic cancer (Bolzoni et al., 2016; Rais et al., 2016; Wu et al., 1978). In order to proliferate, these cells need macromolecules such as nucleotides, amino acids and lipids (Barger and Plas, 2010; McKeehan, 1982; Newsholme et al., 1985).
Glutamine (and glucose) provides much of the carbon and nitrogen needed to create these macromolecules (Barger and Plas, 2010; McKeehan, 1982;
Newsholme et al., 1985). Therefore, if cancer cells have access to a glutamine supply, this results in unrestrained tumor growth (Eagle, 1955; Medina et al.,
1992; Rais et al., 2016; Souba, 1993). As a result of this dependence, the use of glutamine inhibitors has been explored as a cancer treatment (Rais et al.,
2016).
Azaserine was first isolated from Streptomyces fragilis and was characterized to have anti-tumor effects (Anderson et al., 1956; Greenlees,
1956; Tarnowski and Stock, 1957). Azaserine is a glutamine analog that inhibits many different glutamine utilizing reactions (Livingston et al., 1970).
One of the most studied pathways associated with azaserine’s anti-tumor 61 properties, is the de novo purine synthesis pathway (Livingston et al., 1970).
In this pathway, Azaserine inhibits the conversion of FGAR into FGAM—L- glutamine and ATP are required for this reaction (Livingston et al., 1970).
The mechanism by which Azaserine prevents FGAR from converting into
FGAM is believed to be that Azaserine binds to a sulfhydryl group on the enzyme that catalyzes this reaction, FGAR amidotransferase, thus preventing the attachment of L-glutamine to the enzyme (French, 1963). This renders the enzyme inactive. The binding of Azaserine to this enzyme is irreversible, as was concluded when it was discovered that L-glutamine was incapable of reversing this inhibition (Greenlees, 1956). In general, Azaserine irreversibly inhibits all enzymes dependent on glutamine by covalently binding to nucleophilic groups in the glutamine-binding site (Garrett, 1995).
In addition to this, Azaserine also affects pyrimidine nucleotide synthesis by preventing the conversion of UTP into CTP.
The de novo synthesis of adenine is also impeded by this inhibitor
(Barclay and Phillipps, 1966; Turker and Martin, 1985). This may be because
Azaserine reduces the production of NAD and NADP—the nitrogen of L- glutamine is a precursor for NAD’s amido nitrogen (Barclay and Phillipps,
1966; Fairbanks et al., 1995).
There is evidence supporting the idea that the inhibition of nucleotide production is not the only means by which this compound exerts its anti- tumor effects, particularly when it comes to certain tumor types. This was 62 discovered when growth inhibiting effects of Azaserine were still seen in
Ehrlich ascites tumors even though various purines and purine precursors such as FGAR were also present (Hedegaard and Roche, 1966). McFall et al.
(McFall et al., 1960) found that if you added guanosine to Ehrlich ascites tumors there was a great increase in the adenine and guanine cellular pools, but growth inhibition still remained. It has been suggested that the means by which these anti-tumor effects are implemented is by interference of
Azaserine with the cell’s ability to replicate its DNA (Simard and Bernhard,
1966). This possibility was supported by the observation that one group of investigators found that cells exposed to Azaserine experienced a large increase in size and amount of DNA. When these cells then attempted to divide, a lesion developed in the DNA which was the assumed cause of the subsequent cellular death (Hedegaard and Roche, 1966). The proliferation of cancerous cells needs energy in the form of ATP (Barger and Plas, 2010;
McKeehan, 1982; Newsholme et al., 1985).
DON
DON is a non-natural amino acid that like Azaserine is similar in structure to glutamine (Figure 11) (Levenberg et al., 1957; Livingston et al.,
1970; Rais et al., 2016). More precisely, it is a diazo analog of glutamine that was first purified from a strain of Streptomyces bacteria in the early 1950s
(Coffey et al., 1956; Kisner et al., 1980; Rais et al., 2016). Like Azaserine, 63
DON inhibits reactions requiring L-glutamine as a source of nitrogen for transamidation reactions (Kisner et al., 1980; Rais et al., 2016; Rosenbluth et al., 1976). Such reactions include those carried out by enzymes like glutaminase (Willis and Seegmiller, 1977), L-glutamine:D-fructose-6- phosphate transaminase (responsible for glucosamine synthesis) (Wu et al.,
2001), formyl glycinamide ribonucleotide (FGAR) amidotransferase
(Levenberg et al., 1957; Livingston et al., 1970; Moore and Lepage, 1957),
GMP synthase (Rodriguez-Suarez et al., 2007), NAD Synthetase (Barclay and
Phillipps, 1966) and CTPS (Kisner et al., 1980; Rais et al., 2016). All of these enzymes are inhibited by DON. DON is believed to act as a suicide inhibitor by entering the active glutaminase catalytic site of these enzymes and irreversibly and covalently binding there via alkylation (Alt et al., 2015;
Ortlund et al., 2000; Pinkus, 1977). From DON’s earliest discovery, it was pegged as potentially having anti-neoplastic properties (Nagai et al., 1991;
Xie et al., 2016) which have been confirmed by many later studies (Duvall,
1960; Eagle, 1955; Masterson and Nelson, 1965; Moore and Hurlbert, 1961;
Rais et al., 2016).
DON inhibits the glutamine utilizing enzyme, glutaminase (Willis and
Seegmiller, 1977). Glutaminase catalyzes the catabolism of glutamine into glutamate (Lee et al., 2016). Treatment of the human lymphoblast line WI-
L2 with DON has been found to reduce glutamine catabolism by over 95%
(Willis and Seegmiller, 1977). Glutaminase expression has been found to be 64 upregulated in a number of different cancer types and is positively correlated with the malignancy of the tumors and the rate of cellular growth (Huang et al., 2014; Lee et al., 2016; Lora et al., 2004; Medina et al., 1992; Qie et al.,
2014). In addition, it has been found that the knockdown of glutaminase in breast cancer cells greatly reduced their proliferation (Qie et al., 2014).
DON is an inhibitor of the first, rate limiting enzyme responsible for the production of glucosamine 6-phosphate from glutamine: L-glutamine: D- fructose-6-phosphate transaminase (Broschat et al., 2002; Wu et al., 2001).
Glucosamine 6-phosphate is a precursor for nitrogen containing sugars such as O linked N-acetylglucosamine (O-GlcNAc) (Bond and Hanover, 2015;
Roseman, 2001). O-GlcNAc modifications play a role in a number of metabolic disorders including cancer (Bond and Hanover, 2015).
NAD Synthetase, is another glutamine dependent enzyme (Tulpule,
1963; Wojcik et al., 2006) that can be inhibited by DON (Barclay and
Phillipps, 1966). NAD synthetase catalyzes the last step of NAD+ production
(De Ingeniis et al., 2012). 65
DON inhibits de novo purine synthesis by impeding the production of
FGAM from FGAR (Figure 2). This reaction occurs via the inhibition of the
enzyme which carries out this reaction: FGAR amidotransferase (Levenberg
et al., 1957; Livingston et al., 1970; Moore and Lepage, 1957) resulting in the
buildup of FGAR within tissues and a reduced amount of IMP being available
for the creation of AMP and GMP (Kisner et al., 1980). Another method of
purine metabolism inhibition by DON occurs further downstream in the
pathway via the inhibition of GMP Synthetase (Rodriguez-Suarez et al.,
Figure 11. Molecular structures of CTPS inhibitors DON and Azaserine. Both have structures that are similar to L-glutamine. 66
2007).
DON also inhibits the de novo synthesis of pyrimidine nucleotides
(Eidinoff, 1958; McKeehan, 1982; Moore and Hurlbert, 1961). One way this is done is by inhibition of the glutamine requiring reactions which converts
UTP into CTP (Eidinoff et al., 1958; Kisner et al., 1980; Levenberg et al.,
1957; Livingston et al., 1970). Interestingly, a much higher concentration of
DON, at least ten times as much, is required for pyrimidine synthesis inhibition as compared to what is required for purine synthesis inhibition
(Moore and Hurlbert, 1961). It has also been discovered that DON inhibits pyrimidine synthesis further up in the pathway via the inhibition of carbamoyl phosphate synthetase (Figure 4.), although the form of carbamoyl phosphate that uses ammonia appears to be found only in certain tissue types such as the mammalian liver (Brown and Cohen, 1960; Livingston et al., 1970).
CTPS as filamentous structures
The discovery of CTPS filaments
The ability of CTPS to assemble into filaments was independently discovered, more or less simultaneously, by three different groups in three different organisms (Liu, 2016). Nevertheless, the first publication concerning
CTPS filaments identified them within the Drosophila and this author first 67 coined the name cytoophidia for the filaments which means, “cell snake” (Liu,
2010, 2016). Liu et al. (Liu et al., 2010) first discovered these structures within the Drosophila female reproductive organs. Like most animals, the female Drosophila contains two ovaries (Jia et al., 2016). Each of these ovaries contains 16 structures referred to as ovarioles (Jia et al., 2016). Each ovariole consists of a string of 6 to 7 egg chambers developing in chronological order (Jia et al., 2016). Each of these egg chambers contains 15 large cells called nurse cells, and the egg chambers are enveloped by a thin epithelial layer of somatic follicle cells (Jia et al., 2016). It was within the egg chamber that large cytoophidia were found to be associated with the nurse cells while those short in length were discovered within the follicular cells (Liu, 2010).
The cytoophidia were also found in the oocyte (Liu, 2010). The large cytoophidia were called macrocytoophidia while the short ones were referred to as microcytoophidia (Liu, 2010). Subsequently, the author discovered these structures in a wide range of other tissues and cell types within the fly such as the cells of the brain, gut, testis, accessory glad, salivary gland, trachea, and lymph gland (Liu, 2010).
Two months after the Drosophila publication, the discovery of filaments within the Caulobacter crescentus (C. crescentus) was published by
(Ingerson-Maher et al., 2010). These CTPS filaments were found within the inner curvature of the cell when electron cryotomography was performed
(Ingerson-Mahar et al., 2010). CTPS was found to co-localize with another 68 curvature regulator protein within the filaments called crescentin. Both proteins worked together to regulate cellular shape (Ingerson-Mahar et al.,
2010).
That same year, (Noree et al., 2010) published that a screen of GFP labeled S. cerevisiae genes revealed that CTPS also assembled into filaments within these cells. The two CTPS isoforms—URA8 and URA7—were found to assemble together within these filaments (Noree et al., 2010). Fission yeast was also found to contain these CTPS filaments in both the cytoplasm and nucleus (Zhang et al., 2014).
The following year, (Carcamo et al., 2011) revealed that CTPS filaments could also assemble within mammalian cells such as normal human cells, human cancer cells, and spontaneously within mouse embryonic stem cells. These filaments can be found in both the nucleus as well as the cytoplasm of the cell (Carcamo et al., 2014; Gou et al., 2014; Zhang et al.,
2014).
Regulation of CTPS filament assembly
In all of the cell types studies thus far, the filaments have been found to be dynamic in that they are capable of assembly and disassembly under different conditions and/or are capable of translocating to different areas within the cell (Gou et al., 2014). The assembly of CTPS into filaments has 69 been studied in cells using live imaging of GFP labeled CTPS (Gou et al.,
2014). There were found to be multiple stages of filament assembly (Gou et al., 2014). First, there is a nucleation phase in which many foci form at the same time within the cytoplasm (Gou et al., 2014). In the second phase, these foci elongate and once they reach a certain length, they then go through a third phase of fusion (Gou et al., 2014). There are actually two different kinds of fusion: head-to-head fusion which increases the length of the filament, and there is side-by-side fusion, which increases the thickness of the filament
(Gou et al., 2014). When filaments fuse side by side, they slide towards one another, which suggest movement along the cytoskeleton (Gou et al., 2014). A fourth phase then occurs for those middle sized filaments, a phase referred to as bundling. In this phase, the filaments become very long and thick (Gou et al., 2014). At times, these long and thick filaments can then go on to circularize, where the ends come together (Gou et al., 2014).
In the case E. coli, (Barry et al., 2016) found that purified CTPS from this cell would assemble into filaments upon incubation with its product alone—the nucleotide CTP. None of CTPS’s substrates such as UTP, ATP, nor glutamine resulted in filament assembly (Barry et al., 2014). Filament assembly was determined using right-angle light scattering on a fluorimeter as well as by ultracentrifugation (Barry et al., 2014). The addition of increasing amounts of CTP caused an increased amount of CTPS to localize to the pellet after ultracentrifugation (Barry et al., 2014). This suggests that 70 an increasing amount of CTPS was polymerizing upon titration of CTP.
Furthermore, it was determined that direct binding of CTP to CTPS was necessary for this polymerization to take place because a non-CTP binding mutant of CTPS was unable to polymerize (Barry et al., 2014). Interestingly, incubation of the protein with DON resulted in the disassembly of CTPS, although if CTPS was incubated with both CTP and DON, filaments could still assemble (Barry et al., 2014). CTPS filaments resulting from DON and
CTP incubation would disassemble upon addition of CTPS substrates and could not reform (Barry et al., 2014). This demonstrates that the polymerization equilibrium of CTPS is regulated by the products and the substrates of CTPS. If CTPS is incubated with high concentrations of CTP, the subsequent addition of DON will not result in filament assembly.
It was discovered that the CTPS filaments found within the C. crescentus localize to different areas within the cells based on the cell cycle
(Ingerson-Mahar et al., 2010). Two types of cells result from the asymmetrical division of these bacteria: the stalked, progenitor “mother cell” and a swarmer progeny “daughter cell” (Curtis and Brun, 2010; Osley and
Newton, 1978; Skerker and Laub, 2004). Eventually, the swarmer cell elongates and differentiates into a stalked cell, but a stalked cell cannot regress into a swarmer cell (Curtis and Brun, 2010). A swarmer cell cannot go through cell division or replication but on the other hand, a fully matured stalked cell can (Curtis and Brun, 2010; Skerker and Laub, 2004). Thus, the 71 swarmer cell is in the G1 phase while the stalked cell is in the S phase
(Curtis and Brun, 2010; Osley and Newton, 1978; Skerker and Laub, 2004).
Then, in the late pre-divisional stage, the stalked cell will go into the G2 phase (Curtis and Brun, 2010; Skerker and Laub, 2004). Time lapse microscopy reveals that during the early phases of the cell cycle in the stalked cell, the CTPS filaments are short in appearance (Ingerson-Mahar et al., 2010). But, as the cells advance though the cell cycle, these filaments become longer and then re-localize to the inner curvature of the cell where they stay for the rest of the cell cycle (Ingerson-Mahar et al., 2010). The authors found that if CTPS was even slightly overexpressed within these cells, this would result in filaments that were longer and more widespread than usual, being found in about 63% of cells (Ingerson-Mahar et al., 2010).
When CTPS was highly overexpressed in the cells, the prevalence of filaments was even more increased, being found in about 76% of cells
(Ingerson-Mahar et al., 2010). Furthermore, the appearance of the filaments was even more striking—the filaments looked spread out and exuded from the inner curvature of the cell (Ingerson-Mahar et al., 2010). Therefore the expression level of CTPS strongly regulates the assembly of the filaments in the C. crescentus. When C. crescentus CTPS was expressed in E. coli, they found that it still assembled into filaments, suggesting that there are no species specific factors that regulate filament assembly in the C. crescentus
(Ingerson-Mahar et al., 2010). Interestingly, similar to what has been 72 observed in E. coli, the treatment of C. crescentus cells with DON results in the disassembly of the CTPS filaments, which may suggest that glutaminase activity of CTPS, is essential for C. crescentus CTPS assembly.
It was found that the CTPS in S. cerevisiae assembled foci upon incubation with the end product CTP (Noree et al., 2010). Like the E.coli
CTPS protein, they found that CTP had to bind to CTPS for filament assembly, as a strain of yeast cells expressing a non-binding mutant version of CTPS was incapable of filament assembly upon CTP incubation.
Interestingly though, this incubation resulted in an increase in foci formation
(Noree et al., 2010). ATP was also found to trigger foci assembly. Altogether, this suggests that nucleotides which promote tetramerization of the protein trigger foci assembly. It was also found that filaments assembled upon nutrient deprivation, namely, glucose deprivation (Noree et al., 2010).
Therefore, glucose availability plays a role in the regulation of CTPS filaments within these cells. In addition, the cellular energy state, in the form of ATP levels, played a part in CTPS filament regulation: the depletion of
ATP with sodium azide trigged filament assembly (Noree et al., 2010). The number of CTPS filaments within cells also increased upon treatment with a non-specific kinase inhibitor called staurosporine, suggesting that phosphorylation also plays a role in CTPS filament assembly within yeast cells (Noree et al., 2010). 73
Drosophila were found to have three isoforms of CTPS—isoforms A, B, and C (Azzam and Liu, 2013). It was discovered that only isoform C localizes to the filaments, although isoform A played an important role in filament formation. In addition, it was discovered that a small portion of the amino terminus of isoform C was indispensable for filament assembly although by itself it was not enough to assemble into filaments (Azzam and Liu, 2013).
The filaments contained within Drosophila tissue are spontaneously assembled, therefore they can be stained for and immediately observed under the microscope without any special conditions or drug treatment (Liu, 2010).
The microcytoophidia in the follicular cells have different lengths, based on the developmental stage (Liu, 2010). In follicle cells within the germanium, before stage 4, filaments are under 1 µm in length but they increase in size in the later stages (Liu, 2010). By stage 8, the follicular filaments are around 4
µm and then increase to 5 to 6 µm in stage 9 and 10 (Liu, 2010). The macrocytoophidia associated with the nurse cells also increase in size as development progresses (Liu, 2010). Macrocytoophidia in stage 5 nurse cells are anywhere from 10 to 20 µm in length while at the last stage of 10, they range from 40 to 50 µm in length (Liu, 2010). After the stage of 10, these macrocytoophidia begin to disassemble (Liu, 2010). The macrocytoophidium cannot be seen from stage 11-14 (Aughey et al., 2016; Liu, 2010).
Interestingly, (Strochlic et al., 2014) found that none of these differences in filament assembly within the egg chamber was due to any 74 differences in CTPS protein levels. CTPS protein levels also regulate the assembly of filaments within Drosophila. It was found that if CTPS levels are knocked down, the follicular filaments will disassemble (Chen et al., 2011).
On the other hand, the overexpression of CTPS promotes filament assembly
(Azzam and Liu, 2013). When CTPS is overexpressed, the filaments within the follicular cells become longer and thicker, while the filaments inside of the germline cells become exceedingly long and tangled (Azzam and Liu,
2013). Strochlic et al. (Strochlic et al., 2014) found that the kinase Ack localizes with CTPS in the nurse cell filaments but not in the follicular filaments. It was determined, using a loss of function Ack mutant, that Ack regulates CTPS filament assembly, although Ack was not found to directly phosphorylate CTPS (Strochlic et al., 2014). The protein that Ack phosphorylates in order to regulate CTPS activity is currently unknown
(Strochlic et al., 2014). In wild type flies, the filaments associated with the nurse cells appear no sooner than stage 2 and then begin to disappear at stage 11 and above (Strochlic et al., 2014). But, in the mutant flies, filaments are already apparent in the germarium which is before stage 2. This is a much earlier stage than usual (Strochlic et al., 2014). They also found that the filaments from these mutant flies remain assembled in late developmental stages where they would normally be disassembled (Strochlic et al., 2014). This shows that Ack plays a crucial role when it comes to 75 regulating the stage at which filaments assemble during development
(Strochlic et al., 2014).
Recently, (Pai et al., 2016) identified ubiquitination as a means of filament regulation within Drosophila. They found that when filaments in both the nurse cells as well as the follicular cells are treated with MG132, a proteosomal inhibitor, a reduction in the number of filaments can be observed
(Wang et al., 2015). The effect of this drug on the filaments was linked to a reduction ubiquitin because overexpression of ubiquitin abrogated this (Wang et al., 2015). This suggests that ubiquitin is a positive regulator of CTPS filament assembly within Drosophila. More specifically, it was discovered that the E3 ubiquitin ligase Cbl regulates filament formation by serving to stabilize the structure of the follicular cell CTPS filaments during endocycle stages (Wang et al., 2015). Using immunofluorescence staining one group found that myc, an oncogenic protein involved in cellular growth and proliferation, had increased protein expression during the stages of egg development where CTPS filaments were assembled (Aughey et al., 2016).
These stages are between 2 and 10b. Therefore, there is a correlation between myc expression and filament assembly within the filaments (Aughey et al.,
2016). In addition, it was discovered that the knockdown of myc expression results in a reduction of filament assembly within the egg chamber (Aughey et al., 2016). Conversely, they found that if they increased the expression of myc, this results in a 50% increase in the length of the filaments (Aughey et 76 al., 2016). Overexpression also results in filament assembly in late phases of egg chamber development where they normally would not be (Aughey et al.,
2016). This suggests that myc is a positive regulator of filament assembly within Drosophila cells (Aughey et al., 2016).
Unlike Drosophila, within mammalian cells, CTPS filaments are not typically assembled under normal culture conditions, but there are a few exceptions such as mouse 3T3 mouse fibroblast cells, Chinese hamster ovary cells (Stinton et al., 2013), rat NRK normal rat kidney epithelial cells, rat kangaroo epithelial cells, and mouse embryonic stem cells (Carcamo et al.,
2014). CTPS filaments can be triggered to assemble if subjected to certain conditions and treatments, for example, treatment with inhibitors such as the previously mentioned DON and azaserine (Carcamo et al., 2011; Keppeke et al., 2015). Both are capable of inducing CTPS filaments within cells (Chen et al., 2011). A phosphorylated form of Ack (Y284) has also been found to localize to the filaments contained within mammalian cells (Strochlic et al.,
2014). This suggests that Ack may regulate filament assembly within mammalian cells, just as it does in Drosophila. Although CTPS has been found to assemble with Ack as well as IMPDH2 within the filaments, it is currently unclear if there are additional proteins with which CTPS can assemble.
77
The biological role of CTPS filament assembly
Interestingly, (Barry et al., 2014) determined that an enzymatically inactive form of CTPS is contained within the filaments of the E. coli cell. The authors drew this conclusion when they saw a decrease in enzymatic activity concurrent to the CTP induced filament assembly as determined by absorbance level of CTP product (Barry et al., 2014). However, it appears that this inhibition must occur specifically by its product CTP in order for filaments to assemble because DON, which also inhibits CTPS activity, promotes filament disassembly. A possible mechanism for the inhibition of enzymatic activity upon filament assembly can be seen when one looks at the cryo-EM structure of the CTPS filament (Barry et al., 2014). The crystal structure shows that in the filamentous form, all the enzyme’s active sites are accessible to UTP, ATP, and glutamine, suggesting that active site blocking is not the mechanism by which the enzymatic activity is inhibited (Barry et al.,
2014). Rather, it is possible that there is obstruction of ammonia transfer between the glutaminase and synthetase active sites resulting from the filament structure disrupting the ability of the enzyme to undergo the conformational rearrangement necessary to undergo this transfer (Barry et al., 2014). In order to determine the role that filament assembly plays in cellular physiology, this group created a non-filament assembling mutant of
CTPS, CTPSE277R, which had a reduced ability to bind CTP (Barry et al.,
2014). They then replaced the wild type CTPS with this mutant within E.coli 78 cells (Barry et al., 2014). What they found was that the cells had reduced growth as compared to cells with wild type CTPS: wild type doubling was approximately 51 min while mutant doubling time was approximately 130 min (Barry et al., 2014). In addition to this, the mutant protein also had an effect on the pools of other nucleotides and their precursors and products
(Barry et al., 2014). For example, this mutant resulted in a decreased amount of the CTP precursor orotate, while CTP levels were increased as compared to wild type cells. Therefore, filaments are needed to regulate nucleotide metabolism and without them, growth defects result (Barry et al., 2014).
Interestingly, although CTP levels within mutant cells were elevated, isotope tracing showed that the mutant incorporated radio labeled cytidine into the
CTP pool at the same frequency as the wild type cell (Barry et al., 2014). This supports the idea that filament assembly is crucial for negative feedback inhibition within the cells which is needed in order to tightly regulate the
CTP level. Without the ability to assembly filaments, the uncontrolled buildup of CTP occurs despite having CTP production rates comparable to the filament assembling CTPS. The authors found that filamentous CTPS is more strongly inhibited by CTP as compared to the diffuse form (Barry et al.,
2014). But in order for CTPS to assemble into filaments, the authors found that it must first overcome a large nucleation barrier (Barry et al., 2014). It is by coupling CTPS enzymatic activity to polymerization with a large nucleation barrier that the cell is capable of maintaining sharp control over 79 activity (Barry et al., 2014). The cell is thus able to induce sudden and quick changes in CTPS enzymatic activity at will. In addition, the sequestering of the proteins into these filaments ensures that the protein is ready to be quickly reactivated by depolymerization (Barry et al., 2014).
When it comes to the C. crescentus, it appears as if the primary purpose of the CTPS within the filaments is to regulate the shape of the cell.
Interestingly, (Ingerson-Mahar et al., 2010) found that overexpression of
CTPS resulted in cells that were unusually straight or hooked with less end curvature. And when they reduced the levels of CTPS within the cells, they found that in addition to slow cell growth, the cells displayed over curvature by a sharp kink near the mid part of the cell (Ingerson-Mahar et al., 2010).
But the enzymatic activity of CTPS appeared to be separate from regulation of cellular shape (Ingerson-Mahar et al., 2010). Catalytically inactivating point mutations were created within the synthetase and the glutaminase domains (Ingerson-Mahar et al., 2010). The point mutation in the synthetase domain did not have an effect on filament assembly, but the point mutation in the glutaminase domain disrupted filament assembly (Ingerson-Mahar et al., 2010). The overexpression of the synthetase domain mutant, like overexpression of the wild type protein, resulted in unusually straight cell shape (Ingerson-Mahar et al., 2010). But the over expression of the glutaminase mutant, unlike what occurs with wild type protein, did not affect cell shape (Ingerson-Mahar et al., 2010). This demonstrates that CTPS 80 regulation of cell shape needs filamentous CTPS with a normally functioning glutaminase domain, but it does not have to be enzymatically active
(Ingerson-Mahar et al., 2010). In summation, evidence suggests that the biological role of filamentous CTPS is to regulate cell curvature within the C. crescentus.
In regards to Drosophila, whether or not the CTPS within the filaments is active or not remains a point of debate within the field. (Aughey et al., 2014) did metabolic profiling on flies that were either wild type, overexpressing CTPS, or DON treated. They found that there are four different patterns of expression that come about from overexpressed, wild type, and DON treated flies: low-low-high, high-low-low, low-high-high, and high-high-low respectively (Aughey et al., 2014). The group was especially interested in two of the four patterns—one of which included metabolites involved in amino acid metabolism, and the other one which included dipeptides (Aughey et al., 2014). They found that in these two groups, while
DON treatment greatly influences the levels of the metabolites in these two groups, there were no differences in that respect between the wild type and the overexpressing mutant (Aughey et al., 2014). Of particular interest was glutamine, which fell into the pattern of the group containing amino acid metabolites (Aughey et al., 2014). Although glutamine levels increased in the
DON treated flies, the wild type and the mutant flies have similar amounts
(Aughey et al., 2014). The similarity in level between the overexpression 81 mutant and the wild type supports the hypothesis that the CTPS assembled into filaments has no CTPS related enzymatic activity (Aughey et al., 2014).
In addition to this, the authors developed a mathematical model to describe the CTPS incorporation into filaments (Aughey et al., 2014). The authors’ model predicted that the concentration of CTP did not increase with CTPS in a linear fashion and that CTP levels actually quickly plateaued upon overexpression of the enzyme (Aughey et al., 2014). This demonstrates that the amount of CTP within the cell is quickly buffered upon filament assembly when CTPS is overexpressed (Aughey et al., 2014). This is due to the fact that when CTPS is overexpressed, most of it localizes to the filaments. On the other hand, when CTP levels drop, CTPS can be rapidly removed from the filaments. (Aughey et al., 2014). In summation, the authors concluded that the cell assembles CTPS within Drosophila into filaments in order to dampen the enzyme’s activity levels (Aughey et al., 2014). In this way, the filament is a means to tightly regulate diffuse, active CTP for the maintenance of homeostasis within the cell (Aughey et al., 2014).
On the other hand, other groups such as (Strochlic et al., 2014), believe their findings support the idea that Drosophila filaments are enzymatically active. This group created mutant flies that had reduced CTPS expression.
The flies had reduced egg production and plasma membrane defects in the germline cells (Strochlic et al., 2014). This group then created a mutant form of human CTPS which lacked the ability to be inhibited by its end products 82
(E161K), rendering it constitutively active (Strochlic et al., 2014). When the mutant was expressed in Drosophila, they not only found that it assembled into filaments within the ovaries, it also rescued membrane defects and the sterility of the mutant flies (Strochlic et al., 2014). In addition, this group argued that filaments assemble within the egg chamber during a time when they are most needed metabolically (Strochlic et al., 2014). During the middle of oogenesis, a huge amount of nucleotides are needed for DNA and RNA synthesis as a result of the occurrence of multiple rounds of endoreplication from stage 2-10 (Dej and Spradling, 1999). This endoreplication results in a
1000 fold increase in DNA (Lee et al., 2009). In addition, during these very same stages, RNA production within the nurse cells is up, especially ribosomal RNA (Mermod et al., 1977). The fact that the CTPS filaments are there during a time that has a high demand for CTP, as well as the observation that a reduction of CTPS expression causes defects in oogenesis which can then be rescued with an enzymatically active form of CTPS that localizes to the filaments, supports the idea that these filaments are enzymatically active. In addition, it supports the hypothesis that they are assembled for the cause of promoting and supporting CTP biosynthesis and endoreplication.
The biological purpose that CTPS filaments serve within mammalian cells is not clear. CTPS tends to assemble with other proteins within the mammalian cells: with IMPDH, for example in HEp-2 cells (Carcamo et al., 83
2011) and with Ack in MCF7 cells (Strochlic et al., 2014). The role that co- assembly plays within the metabolism of the cells is unknown. Determining if there are additional filament components besides CTPS, IMPDH, and Ack, and then determining their function within this structure would help to resolve this question. The question of whether there are additional filament components will be discussed in the following chapter.
84
Chapter 2: Identification of CTPS1 associating proteins within the filaments
85
Introduction
The idea that the cytoophidium may not consist of only CTPS was first proposed by Liu (Liu, 2010) when he first observed the macrocytoophidium of the Drosophila. He saw that although the GFP-CTPS evenly labeled the whole length of many of the filaments, there were some filaments that had gaps in them (Liu, 2010). The lengths of these gaps were variable, but most of them were under 1 µm in length. The gaps were not in any consistent pattern; rather they were arbitrarily dispersed along the filaments (Liu,
2010). This suggests that a protein other than CTPS was occupying these
“gaps” which would explain why anti-CTPS antibody did not recognize it.
The finding that CTPS co-localized with another protein within
Drosophila macrocytoophidium came from Strochlic et al. (Strochlic et al.,
2014). Using immunofluorescence, they found that Ack, a non-receptor tyrosine kinase, localized with CTPS to macrocytoophidium within
Drosophila egg chambers (Strochlic et al., 2014). In addition, using immunofluorescence, they discovered that Ack1-phospho-Y284, co-localized with CTPS within a human breast cancer cell line, MCF7 (Strochlic et al.,
2014).
Evidence for the co-assembly of CTPS with yet another protein within mammalian cell filaments first came from Carcamo et al. (Carcamo et al.,
2011). Human autoantibodies were first used to identify filaments within
HEp-2 cells obtained from INOVA diagnostics. Further analysis using co- 86 staining with rabbit and anti-Drosophila CTPS antiserum demonstrated that
CTPS localized to these filaments (Carcamo et al., 2011). Then, in order to determine if IMPDH also localizes to the filaments, they stained the cells with rabbit anti-IMPDH and demonstrated that IMPDH co-localized with
CTPS within the filaments (Carcamo et al., 2011). In addition, IMPDH2 was immunoprecipitated from K562 cells using the autoantibody, It2006
(Carcamo et al., 2011).
Although IMPDH was identified as the target of the human autoantibody It2006 within the filaments, the immunoprecipitation of the
K562 cell extract using filament positive HCV patient sera demonstrate that six of the nine sera—603, 606, 608, 609, 610, and 612, do not immunoprecipitate IMPDH or CTPS (Carcamo et al., 2014). This suggests that additional proteins can localize with CTPS and IMPDH within mammalian cell filaments, but the identification of these proteins or whether or not this is actually the case remains to be determined.
One approach to determining filament components is the utilization of
Bio-ID (Roux et al., 2013; Roux et al., 2012). Bio-ID is a way to screen for protein-protein interactions which take place in living cells (Roux et al., 2013;
Roux et al., 2012). This identification system has been implemented by a number of groups as a means to discover and explore proteins that interact with a protein of interest (Firat-Karalar and Stearns, 2015; Kim et al., 2014;
Kim et al., 2016; Roux et al., 2012; Varnaite and MacNeill, 2016). 87
Bio-ID involves the use of biotin (vitamin H) which is a vitamin/coenzyme necessary for all life and is an important player in carboxylation reactions (Lehninger, 1993). Humans are incapable of synthesizing biotin, and must obtain it by dietary means, but plants, most bacteria and certain fungi can synthesize it themselves (Barat and Wu,
2007). There is a specific class of metabolic enzymes called biotin carboxylase and decarboxylases that are biotinylated by post translational modification
(Barat and Wu, 2007; Chapman-Smith and Cronan, 1999; Knowles, 1989;
Samols et al., 1988). Biotin carboxylases play an important role in lipogenesis, amino acid metabolism and the transduction of energy
(Chapman-Smith and Cronan, 1999). Typically, these enzymes take CO2 from bicarbonate and catalyze the movement of this carboxylate to organic acids resulting in the production of different cellular metabolites (Chapman-Smith and Cronan, 1999; Samols et al., 1988). This process is carried out by the use of biotin as a cofactor acting as a mobile carboxyl carrier (Chapman-Smith and Cronan, 1999). Biotinylation modification is rare within cells, there being only between 1 and 5 proteins species that get biotinylated within various organism (Cronan, 1990). This biotinylation is performed by a protein called biotin protein ligase (Barat and Wu, 2007). This enzyme catalyzes the covalent attachment of biotin to protein (Barat and Wu, 2007). In E. coli, this
35 kD protein regulates the biotinylation of acetyl-CoA carboxylase subunit
(Roux et al., 2012). The wild type version of the protein is highly selective and 88 extremely specific when it comes to biotinylation of its substrate (Roux et al.,
2012). Roux et al. (Roux et al., 2012) created a mutated version of this E.coli protein referred to as BirA*. The mutation R118G results in a version of the protein which has lost its high selectivity for biotinylation (Roux et al., 2012).
In vitro, BirA* has been found to promiscuously add biotin groups to the primary amines (mostly lysines) of proteins in a manner that is dependent on proximity (Roux et al., 2012). In this system, BirA* is fused to a protein of interest then expressed within the cells (Roux et al., 2013; Roux et al., 2012). The fusion protein biotinylates proteins near it within the cell
(Roux et al., 2013; Roux et al., 2012). Biotinylated protein can then be isolated using Streptavidin (Roux et al., 2012) or avidin (Barat and Wu,
2007). Streptavidin comes from the bacteria Steptomyces avidinii and has a high affinity for biotin (Hendrickson et al., 1989). Avidin is a glycoprotein that is isolated from the egg whites of chicken eggs (Marttila et al., 2000). It too has an extremely high affinity for biotin (Marttila et al., 2000). Proteins can be then identified with a method such as mass spectrometry (Hesketh et al., 2017; Roux et al., 2012). In order to determine if there are additional
CTPS interacting proteins within mammalian cell filaments, our lab implemented the Bio-ID system.
89
Materials and methods
Cell culture and transfection
HEK293 cells ( kindly given to us by Dr. Marc Kirschner) were cultured at
37°C in DMEM (Cellgro) supplemented with 10% FBS (Hyclone), 2mM L- glutamine (Gibco), 1% penicillin and 1% streptomycin. Cells were transfected using Lipofectamine 2000 in Opti-MEM Reduce Serum Medium
(ThermoFisher Scientific, # 51985034) as per manufacturer’s instruction.
Plasmid
BirA*-CTPS: The following primers containing BAMHI and XholI restriction site sequences 5’-CTTGCTCGAGATGAAGTACATTCTGGTTACTGG-3’ and
3’-TTATGGATCCTCAGTCSTGATTTATTGATGGAAACTTCA-5’ were used for PCR amplification of CTPS1 originally within in a pEGFP-C1 vector. The
PCR product of CTPS1 flanked by BAMHI and XholI restriction site sequences was then subcloned into pcDNA3.1 mycBioID (Addgene plasmid
#35700) C terminal to BirA* using BAMHI (New England Biolabs, #R3136s) and XhoI restriction site enzymes (New England Biolabs).
V5 and SBP tagged HAL: HAL gene originally from the entry vector hORFeome v8.1 (GE Healthcare, OHS6084-202635094) was cultured in Luria
Broth with 8% glycerol and then a mini prep was performed as per manufacturer’s instruction. This was then gateway cloned into two 90
destination vectors: pcDNA6.2/V5-DEST with C terminal V5 tag and
pcDNA5/FRT/TO with N-terminal HA and SBP tag.
Figure 12. BirA* vector. CTPS1 was subcloned into the vector C terminal to BirA* and myc by using XholI and BAMHI restriction site enzymes.
91
Transfection and cell culture
HEK293 cells were grown in 60 mm dishes with or without coverslips.
All the cells except for one dish without a coverslip were transfected with the above construct using lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions in Opti-MEM (ThermoFisher Scientific), a reduced serum media and left overnight. Transfected cells were then either left untreated or were incubated overnight with DON (Sigma-Aldrich) at a concentration of 200 µM. Cells were then incubated with biotin (Sigma-
Aldrich) at a concentration of 50 µM for 24 hours.
Immunofluorescence
Cells on coverslips were fixed with 4% formaldehyde for 20 min, permeabilized in 0.5% Triton X-100 for ten minutes and then blocked in Abdil
(PBS with 0.1% Triton X-100, 2% BSA. 0.1% Sodium azide) for 10 min.
Coverslips were then incubated with rabbit anti-IMPDH2 antibody (diluted
1:200 in Abdil, Abcam, ab129165) and anti-mouse myc (diluted 1:100 in
Abdil, Santa Cruz, sc-40) for an hour. Coverslips were then incubated with
Alexa Fluor® 488 goat anti rabbit (diluted 1:200 in Abdil, ThermoFisher) and was visualized using Texas Red Goat anti mouse (diluted 1:200 in Abdil,
ThermoFisher) in order to visualize IMPDH2 and myc tagged BirA*CTPS respectively. Histidine Ammonia Lyase (HAL) was identified using anti rabbit-HAL (1:100, Santa Cruz, sc-133646). Cellular DNA was detected using 92
DAPI. Images were taken using MetaVue software (Molecular Devices) controlling a CoolSnap ES camera (Photometrics) on a Nikon TE2000-U inverted microscope. The objective used was a 60x Plan Apo 1.40 NA Nikon.
Immunoblotting
Biotin incubated cells were lysed with a buffer consisting of 1% Triton X-100 and 50mM Tris pH 7.4. A small portion of the lysate was run on a 12.5% SDS
PAGE gel side by side to determine if cells expressing the construct had elevated protein biotinylation as compared to non-transfected cells. Protein was then transferred onto nitrocellulose blotting membranes (GE Healthcare,
#RPN203D) using a semi-dry transfer unit (Amersham) for 1 hour with the power supply set to 100 mA. The membranes were blocked overnight in 5%
BSA in tris-buffered saline with 0.4% Triton X-100. Biotinylated protein was detected by 2 hour room temperature incubation of membranes in peroxidase- conjugated streptavidin (diluted 1:2000 in 5% BSA, Cell Signaling, #3999).
Membranes were then washed 3 times in tris-buffered saline with 0.1%
Triton X-100. Santa Cruz Western Blotting Luminol Reagent (#sc-2048) was used for detection and images were developed on film.
93
Capture of biotinylated proteins
Cells were first washed 3 times with cold PBS. Cells were then lysed with lysis buffer (1% Triton X-100 and 50 mM Tris pH 7.4). Lysates were centrifuged for 5 min, 500 relative centrifugal force (rcf) at a temperature of
4°C. Supernatants were transferred to a new tube and SDS was added to a final concentration of 2%. Neutravidin Agarose Resin (ThermoFisher
Scientific, #29202) was prepared by washing 3 times in 10 ml of the lysis buffer, each time spun down 2000 rcf at 4°C for 5 minutes. Samples were then incubated with the resin at 4°C overnight. Beads were spun down at
5000 rcf for five minutes and depleted supernatant was pipetted off. Beads were then washed three times, 15 minutes each, with 2% SDS and spun down. 1x sample buffer was added to the beads and then samples were boiled for 5 minutes to release the biotinylated proteins.
Silver Stain
Two percent of the volume of the above samples was run on a 10% SDS
PAGE gel. The gel was then fixed in de-stain (50% MeOH, 10% Acetic acid,
0.05% formaldehyde, dH2O added up to 100 ml) for 15-30 minutes. The gel was washed with 50% EtOH and then incubated with shaking in 20 mg of
Na2S2O3 anhydrous in 100 ml of deionized water, followed by washing with deionized water three times, 30 seconds each. The gel was then incubated in silver stain (200 mg AgNO3, 100 µl formaldehyde, dH2O added up to 100 ml), 94
then washed in water for 15 seconds and developed in 3 ml of Na2S2O3 anhydrous solution, 6 g of Na2CO3, 100 µl of formaldehyde, with dH2O added up to 100 ml.
Identification of proteins using mass spectrometry
The remaining eluted protein samples were separated on a 10% SDS
PAGE gel and then gels were stained in Coomassie blue stain overnight in order to visualize any protein bands in the gel. Each of the two lanes (DON and untreated) were cut into 5 parts consisting of the following areas: 150 kD and above, 75-150 kD, 50-75 kD, 37-50 kD, and 25-37 kD. These pieces were then put into microcentrifuge tubes, snap frozen, and shipped on dry ice to the Taplin Biological Mass Spectrometry Facility (Harvard Medical School) and analysis was performed by Ross Tomaino.
95
Results
BirA*-CTPS1 localizes to filaments upon treatment with DON
In order for the
BirA*-CTPS construct to
be used as a tool for the
identification of Figure 13. BirA*-CTPS co-localizes with IMPDH2 in filaments. Immunofluorescence staining demonstrates additional filament that IMPDH (green) and BirA*CTPS (red) co-localize to components, we first had filaments upon treatment with DON (see white arrows). DAPI (blue) was used for nuclear staining. to determine if it
localized to the filaments. To this end, we transfected HEK-293 cells grown on glass coverslips with BirA*-CTPS1 and then on the following day, we treated them with 200 µM of DON overnight to induce the assembly of CTPS containing filaments. We then fixed and stained the cells. Immunofluorescence revealed that BirA*-CTPS1 localized with IMPDH to the filaments upon treatment with DON (Figure 13).
96
BirA*CTPS promiscuously biotinylates proteins within cells
In order to determine if the created fusion protein was useful for the identification of proteins in close proximity to CTPS1 within the filaments,
we had to determine if the BirA*-CTPS + - 100 BirA*-CTPS fusion protein Kd7 5 would promiscuously biotinylate 5 0 nearby cellular proteins. To do 3 7 this, we expressed the BirA*-
2 CTPS construct in a dish of 5 2 0 HEK-293 cells while leaving GAP DH another dish of cells Figure 14. BiotinlyationBiotin of proteins in BirA*- blot CTPS expressing cells and non-expressing untransfected. The following cells. Cells containing BirA*-CTPS have day, we gave both transfected markedly more biotinylated proteins than untransfected cells. This demonstrates that and non-transfected cells biotin cells containing the BirA*-CTPS construct more indiscriminately biotinylate proteins than and allowed the cells to incubate the untransfected cells. at 37ºC for a period of 24 hours
which allowed ample time for biotinylation of cellular proteins. We then lysed the cells and performed a
Western blot where we probed with streptavidin-HRP to determine and compare biotinylation levels of the BirA*-CTPS expressing cells and the wild type cells. This analysis revealed that BirA*-CTPS transfected cells had a high amount of biotinylation while the control cells had significantly less 97 biotinylation of proteins (Figure 14). These results indicate that the BirA*-
CTPS construct is less selective in its biotinylation of proximal cellular proteins than the endogenous, wild type biotin ligase of HEK293 cells.
Because wild type HEK293 cells, like other mammalian cells, contain only a few biotinylated proteins, there would still be some selective protein biotinylation occurring although the levels are nowhere near to what is seen in BirA-CTPS containing cells.
We were able to conclude that the fusion protein was a useful tool for identifying additional filament components because when BirA*-CTPS localizes to filaments within the cell, it is likely to biotinylate any other protein that may be within close proximity to it within the filaments.
BirA*-CTPS identifies known filament components—IMPDH2 and CTPS1 and one candidate protein Histidine Ammonia Lyase
It is currently unknown whether or not CTPS1 containing filaments in mammalian cells have additional proteins, other than IMPDH and Ack. In order to determine if there are additional proteins within the mammalian cell filaments, we transfected HEK293 cells with BirA*-CTPS and then the following day, we either treated them with DON overnight to induce filaments or left them untreated. We then incubated the lysates of these cells with Neutravidin resin in order to separate biotinylated proteins from non- biotinylated proteins. Neutravidin is the deglycosylated version of native 98
DON - + avidin (Marttila et al., 2000). This
250 kD removal of excess carbohydrate 150 100 75 results in a more neutral 50 isoelectric point and therefore less 37
25 nonspecific protein binding
Figure 15. Silver stain of lysates from (Marttila et al., 2000). We took a untreated and DON treated cells. A small amount of DON treated and untreated samples small amount of lysate from both were run in a 10% SDS PAGE gel and a silver the DON treated and the stain was performed to look for any DON specific protein biotinylation. Arrows are untreated control and separated pointing to a band that is more heavily this on a 10% SDS-PAGE gel. We biotinylated in the DON treated sample. then performed a silver stain analysis to determine if there were any differences in the protein biotinylation detectable by eye. Seeing this might indicate that certain proteins, which would normally not be in close proximity to CTPS, are localizing with the construct upon filament induction and getting biotinylated. We observed that both the DON treated and the untreated control had similar patterns of protein biotinylation. This does not conclusively mean that there are no other proteins assembling with CTPS 99
DON - +
200 kD 150 100 75
50
37
Figure 16. Coomassie dye stained gel of lysate from untreated
and DON treated cells. This gel was then cut into sections which
were then used for mass spec analysis as described in the text.
within the filaments, rather it may just be undetectable by eye in the gel
(Figure 15). In preparation for mass spec analysis, we then separated the remaining samples on a 10% SDS-PAGE gel and then identified protein bands by staining with Coomassie blue stain (Figure 16). We then cut the treated and untreated lanes into 5 sections each being sure to carefully section off any bands that were evident within one piece. Section 1 contained proteins >150 kD in weight +, section 2 contained protein bands 75-150 kD in weight, section #3 contained proteins 50-75 kD in weight, section #4 contained proteins 37-50 kD in weight, while section #5 contained proteins that were 25-35 kD in weight. We froze the gel sections and shipped them on dry ice for protein identification via mass spectrometry. Mass spectrometry 100 revealed a number of biotinylated proteins, including two of those known to be contained within the filaments—CTPS and IMPDH—although CTPS and
IMPDH showed up in both the DON treated and untreated samples.
Interestingly, another protein referred to as histidine ammonia lyase (HAL) not only was exclusively found
101
Table 1. Proteins unique to DON treated samples: Mass spectrometry revealed a number of proteins that were found only in the DON treated samples. The protein with the highest number of unique peptides was HAL (seen in red).
102
within the DON treated sample, but also had the highest number of unique peptides of all the proteins in the DON treated samples. HAL is an enzyme that releases ammonia during its reaction. Therefore, HAL may be localizing with CTPS within the filaments to provide ammonia for CTPS. CTPS requires ammonia in order to carry out its reaction.
Verification of HAL as a
filament component
In order to determine
if HAL localized to the
filament upon cellular
treatment with DON, we
treated HEK293 cells with
200 µM of DON overnight to
induce filament assembly Figure 17. HAL localizes to filaments upon DON treatment. HEK293 cells were and then performed treated with Don overnight and then fixed and stained with anti-HAL antibody immunofluorescence where we stained for endogenous
HAL. Immunofluorescence revealed that HAL localized to the filaments upon
DON treatment (Figure 17). 103
A
HA-HAL Anti-HA B C
Non-transfected HA-HAL Anti-HA No primary
Figure 18. Anti-HA antibody non-specifically cross reacts with filaments. (A) HEK293 cells were transfected with HA-SBP-HAL, treated with DON overnight, and then stained for HA in order to determine if the transgenic HAL localized to the filaments .Although filaments could be seen in transfected cells upon staining with HA. (B) Untransfected cells also displayed filaments upon staining with anti- HA this demonstrates that the HA antibody non-specifically labeled any filaments contained within the cells. (C) The control with no primary antibody displayed no filaments.
104
In order to verify these results, we transfected cells with HA-SBP-HAL and then the following day we treated them overnight with DON in order to induce filament assembly. We then performed immunofluorescence staining for HA to determine if the recombinant, tagged, HAL protein localized to the filaments.
Unfortunately, we discovered that the primary antibody, anti-HA, nonspecifically labeled filaments, therefore verification as to whether or not
DON induced CTPS filaments contain HAL remains to be determined (Figure
18).
Discussion
My study is the first to implement the Bio-ID system coupled with mass spectrometry to identify filament components. Using these two methods, I was able to identify the previously unidentified filament component candidate HAL. HAL was of great interest, not only because it had the highest number of unique peptides of all the proteins that were only biotinylated in the presence of DON, but also because of the role that it fulfills within the cell. HAL is the first enzyme involved in the degradation of histidine to glutamate (Schwede et al., 1999). HAL catalyzes the reaction that removes ammonia from L-histidine and results in the production of urocanic acid (Galpin, 1999). 105
The process of amino acid degradation occurs when they are released from protein digestion and turnover (Berg JM, 2002). This degradation mostly occurs within the liver (Berg JM, 2002). CTPS utilizes ammonia in the reaction that it catalyzes. But DON, being a glutamine analog, inhibits the ability of cells to use glutamine as an ammonia source. Therefore, it is interesting that HAL, which can create ammonia from a source other than glutamine, might be contained within the filaments. Perhaps it is providing ammonia for CTPS in the absence of glutamine.
Although Bio-ID is a very useful system because it detects protein- protein interactions within their natural cellular context and interactions that are weak and transient, there are some limitations to this system that could have affected our ability to detect filament components (Roux et al.,
2012). For example, cells are given an unnaturally high amount of biotin in order to enhance the biotinylation of the proximal proteins (Roux et al.,
2012). Biotin is covalently attached to primary amines, mainly lysines, and this results in a charge loss at these sites (Roux et al., 2012). This loss of charge could impede secondary modifications which may change the behavior of the fusion protein itself and the neighboring proteins (Roux et al., 2012).
This could affect what proteins get biotinylated and also the ability of the fusion protein itself to biotinylate proximal proteins (Roux et al., 2012). In addition, the fact that this system depends on the number and availability of primary amines in nearby proteins, and has an estimated maximum activity 106 radius of only 20-30 nm, means that a high amount of biotinylation on a protein does not necessarily correlate with a high abundance of an association. Likewise, the lack of biotinylation does not mean there is no interaction or proximity of a protein (Roux et al., 2012). This means that there may be additional proteins within the filament that BirA*-CTPS did not biotinylate. Also, this system cannot be used as validation for a protein- protein interaction. Rather, this system should be a screen to identify candidates, such as HAL, which should then be investigated and verified by other methods.
One may ask, why would the cell put a group of different proteins all together within a single structure? For example, two known filament associating proteins, CTPS and IMPDH, are from two completely different pathways. What is known is that proteins rarely act alone within cells and more often than not come together as “molecular machines” in order to carry out various biological processes (De Las Rivas and Fontanillo, 2010; Herce et al., 2013; Westermarck et al., 2013). Essentially all functions within the cells require proteins to interact with other proteins (De Las Rivas and Fontanillo,
2010; Herce et al., 2013; Westermarck et al., 2013). The correct assembly of the proteins is also of great importance and most often signaling cascade activation involves the assembly of multiprotein complexes (Westermarck et al., 2013). Therefore, it is possible that these proteins come together within the filaments in response to the triggering of some signaling cascade in which 107 they may take part. These proteins may come together in order to interact and regulate each other’s activity in a specific way that is needed for the current condition within the cell. In other words, perhaps those proteins that come together, under the condition that the cell brings them together, need to be regulated in a specific way in order for the cell to metabolically adapt to the changes that are occurring. If this is indeed the case, some follow-up questions could be, how exactly are these components being regulated once they are in contact with other proteins within the filament? And once assembled within the filaments, what task (or tasks) does this protein complex carry out? Additionally, how do the activities of these proteins differ when they are not assembled?
Within a single organism or cell type, the specific proteins that assemble together into these filaments may vary depending on the particular condition or need at the time. Moreover, this could differ from organism to organism as different organisms have different requirements and needs.
Therefore, determining the complete composition of the filaments may be a difficult if not impossible task to achieve.
Future directions
Another way to verify HAL as a filament component is to knock down protein levels of endogenous HAL from the cells, and then perform 108 immunofluorescence using the endogenous anti-HAL antibody to determine if it still localizes to the filaments. If the knockdown of HAL results in reduced amounts of HAL within the filaments as compared to cells without knockdown that would verify HAL is a filament component. This would be further verified if the introduction of recombinant HAL within these knockdown cells results in the reappearance of the protein within the filaments.
If HAL is verified as a component, whether or not HAL is a core component of the filaments should be determined. Protein levels can be knocked down within cells and then filaments induced with DON.
Immunofluorescence can then be performed to look at each of the currently known filament components to determine if filaments can still assemble. It may be that HAL is needed for IMPDH assembly into filaments, but not needed for CTPS assembly into filaments or vice versa. Alternatively, HAL may be necessary for any filament assembly to occur. If filaments can still assemble, that suggests that HAL is not a core component and is therefore not necessary for filament assembly. If filaments cannot assembly upon knockdown of HAL, that suggests that HAL is a core component and is necessary for filament assembly.
The function of HAL within the filaments should also be determined subsequent to its verification as a filament component. It is known that HAL releases ammonia as a result of the reaction it catalyzes. Ammonia is used by 109
CTPS in its reaction. Therefore, one could ask if HAL activity is necessary for the maintenance of the filament structure. In order to address this question, a mutant of HAL can be synthesized which renders it enzymatically inactive.
This can then be expressed within cells and the cells can be DON treated in order to induce filament assembly. Immunofluorescence can then be performed to determine if there are any changes to filament morphology. If there are changes to filament morphology, for example if there are little if any filaments assembled as compared to control cells with wild type HAL, this would suggest the HAL enzymatic activity is necessary for the maintenance of the filament structure.
A fusion protein consisting of BirA* and IMPDH can also be synthesized to determine if there are additional filament components. It is possible, that other filament components may be out of reach to BirA*-CTPS because they localize too far away from CTPS, but they may localize close enough to IMPDH for a BirA*-IMPDH fusion protein to biotinylate them.
Another way to identify filament components is to use IMP affinity purification. To do this, filaments can be induced within cells, and then cells can be gently lysed to release the filaments from the cell. To capture the filaments, IMP, a substrate of IMPDH, can be conjugated to sepharose beads.
The beads can then be packed into a column and the cell lysis applied to the column. The beads can be washed with the lysis buffer and then filaments can be eluted with a molar excess of IMP. The eluate can then be run on an 110
SDS PAGE gel and the bands that show up in a DON dependent manner can be excised. Mass spectrometry can be performed on the proteins in this gel piece and identified proteins can then be verified as FINS components using immunofluorescence.
Another method is subcellular fractionation which isolates the filaments from other cellular structures. This can be achieved by gradient centrifugation where components within the cell can be separated based on size and density.
Yet another approach is to do a genome screen for filament forming fluorescence tagged proteins within mammalian cells (Narayanaswamy et al.,
2009; Noree et al., 2010; Shen et al., 2016). It can then be determined if any of these proteins co-assemble with CTPS within the filaments.
In conclusion, there are many other methods that can be used in order to search for additional filament components, if they exist.
111
Chapter 3: The biological role of filament assembly
112
Introduction
It is currently known that IMPDH reversibly assembles into linear polymers as a purified protein in vitro and into larger filamentous bundles in cells (Juda et al., 2014). This enzyme catalyzes the rate-limiting step in de novo guanine nucleotide biosynthesis. The question as to why IMPDH assembles into filaments and the role that these filaments serve within the cell is a question that is currently foremost in the field. Whether filament assembly within the mammalian cell is tied to a change in the enzymatic activity of IMPDH, or whether the filaments are even enzymatically active are big questions that have yet to be answered clearly. A major challenge in evaluating the effect of assembly on IMPDH catalytic activity is the fact that
IMPDH is usually not polymerized in cultured cell lines grown in typical nutrient rich media. Filament assembly can be triggered, however, by treatment of cells with inhibitors of purine biosynthesis such as the IMPDH inhibitors MPA or ribavirin, inhibitors of enzymes acting both upstream and downstream of IMPDH such as methotrexate and DON, or by depleting the culture media of essential purine precursors such as glutamine or one-carbon units using methotrexate. This responsiveness of IMPDH to reduced flux through the guanine biosynthetic pathway suggests that assembly could be associated with a homeostatic mechanism to restore guanine nucleotide levels. Alternatively, the filaments may just serve as a place for the cells to 113 store inactive enzyme. The experimental conditions, however, have a big impact on IMPDH activity that is separate from filament assembly.
Remarkably, (Labesse et al., 2013) studied purified IMPDH from the
Pseudomonas aeruginosa and found that the addition of ATP, which triggers filament assembly in purified IMPDH protein, also resulted in an almost 3 fold increase in enzymatic activity. They found no such increase in the enzymatic activity of filamentous human IMPDH1 (Labesse et al., 2013).
Scott et al. (Scott et al., 2004) found that the addition of ATP resulted in a fourfold enhancement of the enzymatic activity in purified human IMPDH2.
However, a number of other groups were unable to reproduce this ATP induced increase in activity for neither IMPDH2 nor IMPDH1 (Carr et al.,
1993; Labesse et al., 2013; Mortimer and Hedstrom, 2005; Thomas et al.,
2012). Nevertheless, the conclusion that filamentous IMPDH is at least enzymatically active can still be drawn from these findings.
It was concluded by (Chang et al., 2015) that IMPDH filament assembly was associated with an upregulation of enzymatic activity. They determined that there was an association between filament assembly and active signaling pathways for cell growth within the cell (Chang et al., 2015).
They assessed the buildup of GTP within the cells after 3’-deazauridine treatment. 3’-deazauridine treatment inhibits CTPS and increased IMPDH filament assembly (Chang et al., 2015). They also found an increase in cellular GTP levels (Chang et al., 2015), but the problem with just measuring 114
GTP levels is that increased levels can result not only from increased de novo
IMPDH activity but also from increased RNA degradation, nucleotide salvage, or even reduced cellular GTP consumption.
Therefore, thus far there is no direct evidence to show that there is an upregulation in IMPDH activity and thus increased flux through the de novo purine synthesis pathways upon filament assembly. Nonetheless, when considering the previous research altogether, it seems to be agreed upon within the field that IMPDH in filamentous form (if assembly is not triggered by an IMPDH inhibitor) is enzymatically active. Therefore, the question of how IMPDH filament assembly affects catalytic activity remains to be answered which is what we sought to do in our study described below.
Materials and methods
Cell culture and transfection
HEK293 cells ( kindly given to us by Dr. Marc Kirschner) were cultured at
37°C in DMEM (Cellgro) supplemented with 10% FBS (Hyclone), 2mM L- glutamine (Gibco), 1% penicillin and 1% streptomycin. Cells were transfected using Lipofectamine 2000 in Opti-MEM Reduce Serum Medium
(ThermoFisher Scientific, # 51985034) as per manufacturer’s instruction.
115
Immunoblotting
SDS-PAGE gels were transferred by semi-dry transfer apparatus
(Amersham) to nitrocellulose membranes. Membranes were blocked using 5% milk in tris-buffered saline. Anti-myc (Santa Cruz), and anti-IMPDH (Abcam) and anti-GAPDH antibodies (Santa Cruz) were used and detected with peroxidase coupled secondary antibodies and enhanced using chemi- luminescence detection (Amersham).
Immunofluorescence
The R356A, Y12A, S275L and wild type constructs in the pcDNA3.1 vector were transfected into HEK293 cells grown on coverslips using Lipofectamine
2000 (Invitrogen, # 11668027) as per manufacturer’s suggestion. Cells were then treated with 10 µM of MPA the following afternoon in order to induce filaments. The following day, cells were fixed with 4% formaldehyde for 20 minutes at room temperature, permeabilized in 0.5% Triton X-100 for 10 minutes and then blocked in Abdil (phosphate buffered saline with 0.1%
Triton X-100, 2% BSA. 0.1% Sodium azide) for ten minutes. Cells were incubated with Anti-mouse myc (diluted 1:100 in Abdil, Santa Cruz, sc-40) for an hour and then for 30 minutes with secondary Texas red goat anti mouse antibody (diluted 1:200 in Abdil) to identify any transgenic IMPDH containing filaments. Nuclei were visualized using DAPI. Images were taken using MetaVue software (Molecular Devices) controlling a CoolSnap ES 116 camera (Photometrics) on a Nikon TE2000-U inverted microscope. The objective used was a 60x Plan Apo 1.40 NA Nikon.
Vectors
IMPDH: For expression in mammalian cells, IMPDH2 contained within a pcDNA3.1(+)/myc-His B vector was used. IMPDH2 being N terminal to the myc and 6 histidine tag. Expression is driven by the CMV promoter. For recombinant IMPDH2 expression in E. coli (for the purpose of protein purification), IMPDH contained within a pSMT3-Kan vector (derived from the pET-28a (+) vector) was used (Figure 19). This vector has a cleavable
Smt3/SUMO tag cloned between the NhI and BAMHI sites. This SUMO tag can be cleaved off using a protease called Ubiquitin-Like protein (ULP1).
IMPDH was cloned in C-terminal to this tag between the SacI and EagI restriction sites.
ULP1: The coding region for ULP1 (403-621) was amplified from the genomic
DNA of the S. cerevisiae genomic DNA (W303-1A) by PCR which generated the coding region with a 5’ NheI restriction site sequence and a 3’ SacI restriction site sequence. This was then cloned in the corresponding restriction sites within the pET28b-Kan vector (Novagen).
117
Figure 19. pSMT3-Kan vector map
118
Mutagenesis
Mutagenesis was conducted using QuickChange site-directed mutagenesis (Agilent) according to manufacturer’s instructions. Mutagenesis was conducted with the following forward primers and their corresponding reverse complements:
R356A 5’-CAGAGTATGCACGGGCCTTTGGTGTTCCG-3’
Y12A 5’-GATTAGTGGGGGCACGTCCGCAGTGCCAGACGACGGACTC-3’
S275L 5’- GGATGTAGTGGTTTTGGACTTATCCCAGGGAAATTCCATCTTC-
3’
Settings used for the thermocycler were as follows: Step 1: 95ºC for 2 minutes, Step 2: 95ºC for 30 seconds, Step 3: 55ºC for 1 minute, Step 4: 68ºC for 7.5 minutes, Step 5: 19 times to step 2, Step 6: 68ºC for 5 minutes, and
Step 7: hold at 4ºC. PCR products were then placed on ice for two minutes to cool down. Digestion of the parent plasmid was achieved by the incubation of the PCR product with 1 µl of dpn1 enzyme (New England Biolabs, #R0176L) at 37ºC for 2 hours. The success of the PCR was then checked by running 5 µl of the PCR product on a 1% agarose gel with an equivalent amount of the template as a comparison. Transformation of DH5α cells with pcDNA3.1
IMPDH2 vector was done as well as a transformation of BL21 (DE3)
Competent E. coli cells with pSMT3-Kan IMPDH2 plasmids. Then, both cell types were cultured separately in Luria Broth and then the DH5α cells 119 containing the pcDNA3.1 vector mutants were placed in agarose plates containing ampicillin at a concentration of 100 µg/ml, while the BL21 (DE3)
Competent E. coli cells containing the mutants in the pSMT3-Kan vector were placed on agarose plates containing kanamycin at a concentration of 30
µg/ml. These plates were incubated overnight at 37ºC. The following afternoon, three colonies from each plate were plucked and each was grown up in 2 ml of Luria Broth overnight at 37ºC in a rotating incubator. The following morning glycerol stocks were made with 360 µl of the cells from each of the clones which were then flash frozen and stored at -80ºC. Mini preps were done with the remaining cells from each clone and this was sent off for sequencing to ensure that the correct mutations had been made and that there were no PCR errors. One glycerol stock of the correctly sequenced clone from each mutant and cell type was saved, while the others were discarded.
ULP1 Protein Purification and expression
Glycerol stocks: ULP1 plasmid was transformed into BL21 (DE3) Competent
E. coli cells. Cells were cultured in 1 ml of Luria Broth for an hour, and then streaked on a kanamycin containing agar plate. The plate was incubated overnight at 37ºC. The following day, a colony was plucked from the plate and placed into 1 ml of Luria Broth with 30 µg/ml kanamycin. This was incubated 120 overnight, while shaking, at 37ºC. The following day, 320 µl was placed into
780 µl of sterile 50% glycerol, was flash frozen, then stored at -80ºC.
Overexpression: 10 ml of Luria Broth was inoculated with cells from either a glycerol stock of from a plate. This was grown overnight at 37ºC. The following day, the 10 ml cultures were added to 1 L of pre-autoclaved Luria
Broth containing 50 µg/ml of kanamycin. This was grown at 37ºC until the
OD600 reached 2. Cells were induced with 1 mM of IPTG at 20ºC overnight while shaking. The following morning, the cells were pelleted and frozen at -
80 ºC until purification or purification was begun immediately.
Purification: Pellets were resuspended in 35 ml of lysis buffer (pH 7.6- 25 mM
Tris, 500 mM NaCl, 10 mM imidazole, 10% glycerol, 1 mM BME). Cells were then lysed using a homogenizer. Lysates were cleared using Lynx 6000 at
14,000 rpm, for 45 minutes at 5 ºC. Cleared lysate was then put into a nickel
(Ni)-column (either gravity or pre-packed). For the gravity column, the column was then washed 3 times in lysis buffer. The cleared lysate was incubated with 5 ml column for 1 hour. Column was washed 5 times. A pre- packed nickel bead column was washed thoroughly with lysis buffer and then protein was eluted with 3-5 column volumes of elution buffer (pH 7.6- 25 mM
Tris, 300 mM NaCl, 400 mM imidazole, 10% glycerol, 1 mM BME). The concentration was measured using a Bradford protein assay. The protein was then concentrated to 1-3 ml using an Amicon spin filter. The protein was loaded onto a Superdex 200 column. A gel of the fractions was then run and 121 the fractions containing the ULP1 were selected. The pooled fractions were then concentrated to the desired volume or concentration. The protein was then flash frozen and stored at -80 ºC.
IMPDH Protein Purification and expression
Overexpression: Sterile pipette tips were gently scraped on the top of the
R356A, Y12A, wild type, and S275L BL21 (DE3) glycerol stock cells and then the tips were each placed into a 25 ml of Luria Broth containing 50 µg/ml of
Kanamycin and cultured overnight at 37ºC. The following day, each of these
25 ml cultures was pipetted into flasks containing 1 L of pre-autoclaved Luria
Broth with Kanamycin at a concentration of 50 ng/ml. These cells were grown up at 37ºC until they reached an OD600 of 0.8. Flasks were then cooled on ice for approximately 5 minutes, and then induced with 1 mM IPTG
(GoldBio.com, #12481C50) for 4 hours at 30ºC. Cells were then pelleted for 5 minutes at a speed of 8000 x g (rotor: Beckman Coulter, JA-10; centrifuge:
Beckman Coulter Avanti J-E). Pellets were then stored at -80ºC.
Purification: Thawed pellets were re-suspended in 20 ml of lysis buffer per liter (pH8-50mM KPO4, 300mM KCl, 20mM imidazole, 0.8M urea).
Resuspended pellets were poured into a beaker with a stirring rod, benzonase
(Millipore, #2774110) was added to help break up the clumps, and the beakers were slowly stirred on ice for ten minutes. Cells were then homogenized using the Emulsiflex-C3 (Avestin) with 3 cycles of 10-15 122 thousand psi. Lysates were then cleared at a speed of 18,000 rpm for 20 min
(rotor: Beckman Coulter JA 25.5; centrifuge: Beckman Coulter Avanti J-E).
During the spinning, 1 ml per liter of nickel-NTA agarose beads (Qiagen,
30210) were washed in 10 column volumes of dH2O once and then twice in 10 column volumes of lysis buffer in a 50 ml centrifuge tube. Twenty µl samples of each of the cleared lysates were then taken out and set aside. One ml of beads per liter was then added to the cleared lysates. This was then gently rotated at 4ºC for 2 hours. Beads were then harvested by centrifugation of lysates at 4000 rpm for 10 minutes (rotor: JS 5.3). Twenty µl of each depleted sample were set aside. The remaining depleted sample was discarded. Beads were then washed with ten column volumes of lysis buffer a total of three times. Elution buffer (pH 8- 50mM KPO4 , 300 mM KCl, 500 mM imidazole) was then added to the beads at a ratio of 1:1 bead volume to elution buffer volume. Beads were then rotated at 4°C for 15 minutes. The eluate was collected then this was repeated four more times to collect four more eluates subsequently for 5 minutes each. Protein was then separated on a 10% SDS
PAGE gel then Coomassie stained to look at protein amounts in cleared lysate, depleted lysate, and eluates. Concentrations of each eluate were also determined using a Bradford protein assay. Eluates containing the most protein were then pooled. The concentration of each pooled protein was then determined using a Bradford. A 20 µl sample of each protein type was then taken out and set aside. One mg of ULP1 was then added for every 100 mg of 123
IMPDH protein. The proteins were rotated slowly overnight at 4ºC to allow cleavage of the SUMO tag from the IMPDHs to occur. The following day, the proteins alongside their pre-cleaved samples were separated on a 10% SDA-
PAGE gel to verify that the SUMO tag had been cleaved from the protein.
The concentrations of the proteins were once again determined using a
Bradford protein assay. Based on this concentration, the volume that the proteins were to be concentrated to (3K Amicon Ultra centrifugal filter unit, #
Z740186) was determined. The protein was then filtered by syringe driven filtration (Millex-GP, Z359904) then loaded onto a Superdex 200 or a
Superose 6 gel filtration column pre-equilibrated in gel filtration buffer: (50 mM Tris, 100 mM KCl, 1 mM DTT, pH 7.4). A gel of the fractions was then run. Fractions that were most pure and had the most protein were pooled together. The concentrations of the pooled samples were then determined using a Bradford protein assay. The proteins were aliquoted, flash frozen in liquid nitrogen, and stored at -80ºC.
IMPDH activity assay
NADH standard curve: Fivefold dilutions of NADH were created and four different amounts of NADH from these dilutions were used for the standard curve: 0.016, 0.0032, 0.00064, and 0.000128 µmoles. The corresponding fluorescence level at each of these amounts of NADH was determined using a fluorimeter (Cary Eclipse). This was then plotted as fluorescence (AU) versus 124
NADH (µmol). A line was fitted to the scatter plot and an equation was derived from this.
Determination of IMPDH specific activity: Reaction mixtures containing
IMPDH2 were placed into a cuvette and the amount of NADH production over time was determined using a fluorimeter (Cary Eclipse) which determines the fluorescence over a period of 3 minutes (approximately 9 measurements per minute). Each of these fluorescence readings at each time point was then plugged into the y value of standard equation to determine the x value, which is the corresponding µmoles of NADH at the given time point. The µmoles of NADH versus the time was plotted and a line was fit to the scatter plot. An equation was then derived from this line. The slope of this line, which is the µmoles of NADH produced per minute, was used to determine the specific activity which was written as “µmoles of
NADH/min/mg”.
IMPDH Activity assays: To determine the specific activities of the different
IMPDH constructs, IMPDH activity buffer (50 mM tris, 100 mM KCl, 1 mM
DTT, pH 7.4) with 5 mM NAD and 3 mM IMP was mixed together in a microcentifuge tube. The reaction was then initiated by the addition of 13.8
µg of protein bringing the reaction mixture to a final volume of 100 µl. In the experiments where ATP or ATP and substrate were added, 13.8 µg of protein was pre-incubated with 1 mM ATP at room temperature for ten minutes and 125 the reaction was initiated with 5 mM NAD and 3 mM IMP or 13.8 µg of
IMPDH was incubated with 1 mM ATP and 5 mM NAD or 1 mM ATP and 3 mM IMP and then the reaction was initiated with various concentrations of the second substrate. In the case where protein was incubated with 1 mM
ATP, 3 mM IMP and GTP, after the 10 minute incubation of 13.8 µg of
IMPDH with 1 mM ATP and 3 mM IMP, IMPDH was subsequently incubated with various concentrations of GTP for 10 min before the initiation of the reaction. All activity assay reactions had a final volume of 100 µl.
NADH production was measured in real time in a 100 µl cuvette (1 cm path length) in a Cary Eclipse fluorimeter (excitation: 340 nm, emission: 440 nm) equipped with a temperature controlled multi-cuvette holder maintained at
37°C. The specific activity for each protein was then determined under each condition using the previously described procedure. The specific activities were normalized to the specific activity value without ATP incubation, or the specific activity value without substrate incubation.
Negatively Stained electron microscopy and particle averaging/reconstruction
Aliquots of purified protein, including wild WT IMPDH and the mutants Y12A, R356A, and S275L, were diluted to 2.5 µM in assembly buffer
(50 mM Tris, 100 mM KCl, 1 mM DTT, pH 7.4) and incubated for 15 minutes at room temperature in the presence of various concentrations of ATP, GTP, 126
IMP, and NAD+, as outlined in Figures 20D and 23C. Following incubation, samples were applied to glow-discharged continuous carbon film EM grids and negatively stained using 1% uranyl acetate. Grids were imaged by transmission electron microscopy using an FEI Tecnai G2 Spirit operating at
120 kV and a Gatan Ultrascan 4000 CCD via the Leginon software package
(Suloway, et al., 2005). Image processing, including CTF estimation, particle picking, and two-dimensional reference-free classification, was performed using the Appion and RELION software packages (Lander et al., 2009;
Scheres, 2012). The image data collected from all ATP/GTP/IMP/NAD+ combinations were combined and processed en masse (77,525 helical repeats in total). Following 2D classification, the resulting class averages were manually grouped according to the observed rise into one of three categories:
“compressed” (rise ~94 Angstrom (Å)), “extended” (rise ~114 Å), or
“undetermined” (filament ends, distorted and bent octamers, and poorly aligned particles). The proportion of helical repeats from each experimental dataset belonging to these three categories was then back-calculated.
Electron cryo-microscopy and particle averaging/reconstruction
Purified Y12A protein was diluted in assembly buffer to 4 µM, and incubated at room temperature in the presence of 0.1 mM ATP and 0.1 mM
GTP for 15 minutes. Following incubation, the sample was applied to a glow- discharged C-Flat holey carbon EM grid (Protochips), and blotted/plunged 127 into liquid ethane using a Vitrobot plunging apparatus (FEI). Imaging was performed using an FEI Tecnai G2 F20 operating at 300 kV and a Gatan K-2 direct electron detector via the Leginon software package (Suloway, et al.,
2005). Image processing, including CTF estimation, particle picking, two- dimensional reference-free classification, three-dimensional classification and
3D refinement, was performed using the Appion and RELION software packages (Lander et al., 2009; Scheres, 2012). The map resolution, as determined by the relion_postprocess script using the Fourier shell correlation (FSC) = 0.143 criterion was 8.7 Å. The final electron density map was further sharpened using a b-factor of -987.
[13C2, 15N]-glycine incorporation into AMP and GMP
HEK-293 cells were maintained in DMEM (Cellgro) containing 10%
FBS (HyClone) and 2 mM L-glutamine (Gibco). 24 hours prior to the glycine addition, two 10-cm dishes of 40% confluent HEK-293 cells were transfected with either the empty vector pcDNA 3.1/myc-His B or S275L IMPDH2 or
Y12A IMPDH2 constructs using Lipofectamine-2000 (ThermoFisher
Scientific) as per manufacturer’s instructions. The next day, cells were washed once in Dulbecco’s PBS (Gibco) and the media of the cells was replaced with DMEM containing 10% dialyzed FBS, 2 mM L-glutamine, and
1.2 mM [13C2, 15N]-glycine (Cambridge Isotope Laboratories). After 6 hours, cells were washed once with Dulbecco’s PBS, released from plates with 0.25% 128 trypsin-EDTA, harvested, washed twice more with Dulbecco’s PBS, and counted. Cells from paired tissue culture dishes were pooled, resulting in 1-
10x106 cells per sample.
Samples were then processed as described (Laourdakis et al., 2014).
Briefly, 70 l of a solution containing 0.5 M perchloric acid and 200 M of the internal standard [13C9, 15N3]-CTP was added to each cell pellet. The cells were vortexed for 10 seconds and incubated on ice for 20 minutes. Lysates were then neutralized with 7 l of 5 M potassium hydroxide, vortexed for 10 seconds, and incubated on ice for an additional 20 minutes. Cell debris was harvested by centrifugation at 11,000 x g for 10 minutes at 4°C.
Supernatants were further clarified using Amicon Ultra 0.5 ml centrifugal filters (3 kD MWCO) as per manufacturer’s instructions.
Collected extracts were analyzed by ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) using a Waters
Acquity UPLC HSS T3 column (1.8 mm, 2.1 mm x 50 mm). Quantitative analysis of nucleotides was performed using the multiple reaction monitoring
(MRM) mode, and the peaks of nucleosides and nucleotides were analyzed using MassLynx V4.1. In addition, calibration standards were freshly prepared for every sample set analysis, and linear calibration curves were established. All reported data represent the mean and standard error of three biological replicates. P values were calculated using two-tailed unpaired
Student’s t-tests in Graphpad Prism. 129
For each replicate, samples for immunofluorescence and Western blotting were prepared in parallel to samples for nucleotide extraction. Cells for immunofluorescence were stained as described previously for total
IMPDH and myc tagged protein except that DRAQ5 (Biolegend) was used as the nuclear stain. Images were obtained using a Leica DM 5000 microscope coupled to a Leica TCS SP5 confocal laser scanner. For Western blot analysis, cells were lysed in RIPA buffer supplemented with Complete protease inhibitor cocktail (Roche). Proteins were separated by SDS-PAGE and transferred to nitrocellulose. Blots were stained with anti-IMPDH2 antibody
EPR8365(B) (Abcam).
Results
Mutants of IMPDH alter the ability of IMPDH2 to polymerize
In order to determine the effect of IMPDH2 polymerization on catalytic activity, we sought to identify residues within the putative octamer polymerization interface that when mutated to alanine would disrupt
IMPDH2 polymerization. We used the IMPDH2 crystal structure (PDB code:
1NF7) which contains stacked octamers that resemble IMPDH filaments
(Labesse et al., 2013). Working with the assumption that these crystal contacts are the same as polymerization interfaces in IMPDH2 filaments, we engineered a series of mutants that we predicted would block polymerization 130
(Figure 20A). The myc-tagged mutants were then individually transfected into HEK293 cells. This was done to ensure that there were similar levels of each mutant being expressed within the cells, therefore, in subsequent assays, any observed differences between the behaviors of the mutants would not be attributed to a difference in protein expression. To our satisfaction, we found that the expression levels of the wild type and mutant constructs were comparable as were the endogenous IMPDH2 expressions within transfected cells (Figure 20B). In wild type, and mutant expressing cells, filament assembly was then induced by treatment with the IMPDH inhibitor MPA, overnight. This was then visualized using immunofluorescence staining with anti-myc antibodies. Of the amino acid residues selected for mutagenesis
(threonine 10, aspartate 15, tyrosine 12 (Y12A), and arginine 356 (R356A)) 131
Figure 20. Identification of filament disrupting and promoting point mutations of human IMPDH2. (A) The human IMPDH2 octamer (PDB file 1NF7). One octamer (color: greens-catlytic domains,pinks-regulatory Bateman domains) is shown with a subset of crystal packing neighbors (gray). Residues tyrosine 12 (red) and arginine 356 (orange) located at crystal contact sites near the presumed filament assembly interface are indicated. (B) Immunoblot of HEK293 cell lysates with antibodies directed against myc-IMPDH2, total IMPDH, or loading control glyceraldehyde 3-phosphate dehydrogenase (GAPDH). (C) Immunofluorescence images of HEK293 cells transiently transfected with the indicated IMPDH2-myc construct stained with anti-myc antibodies. Cells were either treated with 10 µM mycophenolic acid to induce IMPDH2 assembly or not as indicated. Images are representative of 3 experiments. Scale bar, 20 µm. (D) Negative stain electron microscopy of purified, recombinant IMPDH2 proteins incubated in the presence of 5 mM NAD, with and without 1 mM ATP. Images representative of 2 separate experiments. Scale bar 50 nm. 132 only Y12A and R356A strongly disrupted filament induction by MPA while wild type assembled as expected (Figure 20C).
We then wanted to confirm that these mutations were acting directly on IMPDH2 self-assembly (as opposed to disrupting protein folding or interactions with assembly partners), so we used electron microscopy to directly visualize recombinant wild type and mutant IMPDH2. The wild type protein appears as a mixture of octamers and tetramers with occasional short filaments in the absence of ligands, and formed long polymers in the presence of ATP (Figure 20D). In the absence of ligands, Y12A and S356A appear as a mixture of octamers and tetramers without any short polymers, and no polymers were observed in the presence of ATP, indicating that the mutations directly interfere with polymerization (Figure 20D).
Another mutation, S275L, was identified by another group during a study of human IMPDH1 sequence polymorphisms (Wu et al., 2010). This mutation results in constitutive IMPDH1 filament assembly when expressed in cells (Tse-Yu-Wu, 2011). We created this mutation in IMPDH2 for the purposes of our study and confirmed the propensity of S275L to assemble in transfected cells, even without MPA present (Figure 20C). Unlike the wild type protein, S275L robustly polymerizes in both the presence and absence of
ATP, which demonstrated that the propensity of S275L to polymerize is an intrinsic result of the mutation (Figure 20D). S275L is localized adjacent to the NAD+ binding site and the basis for its constitutive filamentation is 133 unclear. Nevertheless, this construct served as a useful experimental counterpoint to the non-assembling mutants.
Polymerization does not alter enzymatic activity of IMPDH in vitro
We then wanted to determine if polymerization altered the enzymatic activity of IMPDH2. To that end, we assessed the catalytic activity of purified, recombinant wild type and mutant IMPDH2 by monitoring NADH production in real time by fluorescence (Figure 21A & 21B). The increase of
NADH fluorescence over time resulting from NAD, IMP and IMPDH2 dependent reduction of NAD to NADH, can be detected with high specificity, using a fluorimeter. This can be used to quantify enzymatic activity (Figure
21B). Initial reaction rates were measured under conditions where saturating amounts of the substrates NAD and IMP were used in the absence of ATP.
We observed no evidence for increased activity of wild type IMPDH2 under polymerization conditions (Figure 21C). Similarly, non-polymerizing
IMPDH2 mutants were unaffected by ATP (Figure 21C). These results confirmed our conclusion that these mutations likely do not disrupt protein folding. 134
Figure 21. IMPDH2 polymerization does not alter its catalytic activity in vitro. (A) Schematic of the IMPDH reaction illustrating production of NADH, which can be detected by fluorescence. (B) NADH fluorescence was monitored in reactions lacking NAD+, IMP, IMPDH (no enzyme), or in a reaction containing all reaction components (Complete). (C) IMPDH activity was measured for wild type and mutant forms under non-assembling conditions. All mutants showed no significant difference in activity compared to wild type (two-sided Student’s t test; P > 0.05). Bars denote mean and standard deviation. (D) IMPDH activity assays as in (C) were conducted in the absence or presence of ATP to induce assembly. ATP did not significantly alter activity in any case (two-sided Student’s t test). (E & F) Initial reaction rates are plotted as a function of substrate concentration for IMP or NAD+ comparing wild type IMPDH2 with non- assembling mutants under polymerization conditions (1 mM ATP) or for wild type in the presence or absence of ATP (G&H). Michaelis constants determined from these titrations are given in Supplemental Table 1. 2-5 replicates per substrate concentration.
135
Purified wild type IMPDH2 has been found to polymerize upon incubation with its allosteric effector ATP (Labesse et al., 2013). But the non- polymerizing mutants do not polymerize in the presence of ATP (Figure 20D).
To determine if filament assembly had an effect on IMPDH2 enzymatic activity, we measured the activity of each variant in the presence or the absence of ATP (Figure 21D). We found no significant difference in enzymatic activity between polymerized IMPDH2 and non-polymerized IMPDH2
(Figure 21D). Taken together, the comparable activity of IMPDH2 under polymerizing and non-polymerizing conditions indicates that assembly of purified IMPDH2 does not affect specific activity.
Measuring the catalytic activity of IMPDH2 under saturating substrate concentrations as we previously did might mask subtle differences in IMPDH2 substrate affinity in the polymer compared with free enzyme.
Therefore, we determined Mechaelis constants (Km) of mutant and wild type
IMPDH2 by measuring the initial reaction velocity at a range of substrate concentrations under polymerizing and non-polymerizing conditions. The
Table 2. IMP and NAD Km values for wild type, R356A, and Y12A 136 mutant and wild type IMPDH2 affinity for IMP (Figure 21E) and NAD
(Figure 21F) were indistinguishable. This demonstrated that wild type
IMPDH2 polymers and unassembled mutants have comparable substrate affinities. Consistent with this, we found that triggering wild type IMPDH2 polymerization with ATP also did not affect IMP affinity (Figures 21G, also see Table 2). A 10-fold increase in the Km(app.) for NAD+ is likely due to an effect of ATP that is separate from polymerization since no difference in
NAD+ substrate affinity was observed comparing wild type and non- polymerizing mutants (Figure 21F). We conclude that substrate affinity is not substantially altered by IMPDH2 polymerization.
Polymerization of IMPDH2 into filaments does alter guanine nucleotide biosynthetic flux in vivo
Next, we considered whether IMPDH2 polymerization might alter its catalytic activity in the cellular context in a way not detectable for the purified protein in vitro, for example by regulating IMPDH2 interaction with other proteins. To answer this question, we established a way to monitor IMPDH2 activity in live cells by monitoring the incorporation of isotopically labeled glycine provided in the culture media, into GMP, a downstream metabolic product of IMPDH, or AMP as a control. Glycine is an important purine nucleotide precursor. Because IMPDH2 is rate limiting for guanine nucleotide synthesis, we hypothesized that if there were any changes 137 in its activity it would affect the rate of labeled glycine incorporation into the
GMP pool.
For our experimentation, we compared the activity the filament promoting mutant, S275L, with the non-polymerizing mutant Y12A.
Importantly, these forms exhibit identical catalytic parameters in vitro
(Figures 21, C-H). HEK293 cells were transfected with each construct or empty vector and were immunostained with antibodies against IMPDH or anti-myc antibodies specific to the transfected proteins. Total IMPDH expression was increased by ~ 2-fold in IMPDH2-transfected cells compared to empty vector-transfected cells, evident from the increased immunofluorescence signal (Figure 22A) and by immunoblot (Figure 22B).
S275L and Y12A expressed at similar levels. As expected, S275L-transfected cells showed robust IMPDH2 filaments whereas Y12A-expressing cells showed diffuse, cytoplasmic staining (Figure 22A).
To determine if cells expressing S275L mutant had increased nucleotide synthesis as compared to cells expressing Y12A, cultures of transfected cells were incubated for 6 hours in culture medium containing
13 15 [ C2, N]-glycine at a 3 fold molar excess over unlabeled glycine. This resulted in a three unit mass increase in glycine-derived metabolites which allowed for the quantitation of pulse-labeled AMP/GMP. The nucleotides were extracted and quantified via liquid chromatography-mass spectrometry. 138
Unlike the percentage of isotopically labeled AMP, the isotopically labeled
GMP fraction was modestly 139
Figure 22. . IMPDH filament formation does not alter GMP synthesis. (A) Representative images of HEK293 cells transfected with either empty vector or vectors expressing either S275L or Y12A IMPDH2. Transfection of S275L IMPDH2 results in constitutive filament formation while transfection of Y12A IMPDH2 abrogates filament formation of both transfected and endogenous IMPDH. (B) Representative immunoblot of total IMPDH expression in transfected HEK293 cells. The higher molecular weight band is the transfected protein whereas the lower band is endogenous IMPDH. (C) Incorporation of
[13C2, 15N]-glycine into AMP and GMP. Center bars represent mean and error bars represent standard error for three biological replicates conducted on different days. Indicated p values were calculated with a two-tailed, unpaired Student’s t test. No statistical differences (P < 0.05) were found between samples for incorporation of [13C2, 15N]-glycine into AMP pools.
140 increased in both IMPDH2-transfected cells as compared to empty vector control which was consistent with an increase in total IMPDH activity
(Figure 22C). Interestingly, the percent of GMP incorporating labeled glycine, and therefore synthesized by way of IMPDH, was indistinguishable in S275L and Y12A expressing cells (Figure 22C). This finding demonstrates that polymerized and non-polymerized IMPD2 have comparable catalytic activity in vivo.
GTP effects on IMPDH octamer conformation and catalytic activity
A recent study found that eukaryotic IMPDHs are subject to feedback inhibition by low millimolar GDP or GTP (Buey et al., 2015). The crystal structure of Ashbya gossypii IMPDH bound to GDP (PDB: 4Z87) revealed a collapsed conformation of the IMPDH octamer mediated by the binding of three GDP molecules to sites within the Bateman domain (Buey et al., 2015).
Two sites overlap with the canonical Bateman domain ATP-binding sites, implying that competition between adenine and guanine nucleotides for these allosteric regulatory sites and/or binding of a guanine nucleotide to the third site can trigger a dramatic conformational change in the IMPDH octamer from an expanded, catalytically active form to a collapsed, inactive form.
Residues that mediate GTP binding are conserved in eukaryotic but not prokaryotic IMPDHs, though the ability of human IMPDH to adopt the collapsed state has not been reported. To examine this directly we carried out 141 a comprehensive investigation of IMPDH2 conformational states using negative stain and cryo-electron microscopy. We sought to determine which conformations the human octamer can adopt in the free and filament contexts and what role ligands play in determining these conformations.
We explored a range of combinations and concentrations of ATP, GTP, and substrates and assessed the structural consequences using negative stain images and two-dimensional, reference-free classification of IMDPH2 filaments. We identified four major classes of filament segments: expanded, collapsed, bent, and ‘poorly aligning’ (Figure 23, A-D). The spacing between octamer centers in the expanded conformation is ~110 Å compared to ~95 Å in the collapsed conformation. The bent conformation is well-resolved with one side matching the spacing of the collapsed conformation and the other side matching the expanded conformation, suggesting that in some cases octamers may occupy conformational states intermediate in the transition between expanded and collapsed, with IMPDH2 monomers present in mixed states within a single octamer. We defined as 'poorly aligning' segments that were assigned to classes without detailed features. This finding demonstrates the surprising plasticity of the filaments to accommodate both expanded and collapsed octamer conformations. We quantified the frequency of these classes in the presence or absence of ATP, GTP and substrates (Figure 23B).
This analysis shows that binding of either substrate shifts the conformational equilibrium toward the expanded octamer, and that GTP promotes the 142
Figure 23. Ligands influence IMPDH filament architecture by altering the conformation of IMPDH protomers (A & B) (A) Representative class averages of IMPDH filaments in four different conformational states, with the height of one IMPDH octamer indicated. (B) Quantification of the fraction of IMPDH helical segments in each conformational state as a function of ligand concentrations. (C) Negative stain EM reconstruction of IMPDH filament structures in the expanded substrate-bound (0.1 mM ATP, 2 mM NAD+) and collapsed GTP- bound (0.1 mM ATP, 0.1 mM GTP) conformations. The refined helical rotation are rise for the expanded filament are 30° and 111 Å, and for the collapsed filament 35.5° and 94 Å, consistent with the rise measured directly from two-dimensional class averages in (A). (D) The heights of crystal structures of P. aeruginosa IMPDH-ATP and A. gossypii IMPDH-GDP closely match the refined helical rise of the human IMPDH filaments in the expanded and collapsed states, respectively.
143 collapsed conformation, suggesting that the balance of GTP and substrates tunes the conformational state of the IMPDH filament.
We performed three-dimensional reconstructions of IMPDH2 filaments under conditions where the equilibrium was shifted most strongly toward the expanded conformation (0.1 mM ATP, 3 mM IMP, 5 mM NAD+) or the collapsed conformation (0.1 mM ATP, 0.1 mM GTP) (Figure 23C), yielding structures at ~20 Å resolution. The structures reveal helical polymers formed from stacked octamers that differ significantly from each other in their refined helical rise (111 Å for expanded and 95 Å for the collapsed state), consistent with the 2-dimensional class averages. In the collapsed conformation there is a 30.5° rotation between octamers and the collapsed filament exhibits an additional 5° of rotation between octamers. It appears that the filament assembly contacts between octamers are relatively fixed, and the differences in helical symmetry arise primarily from differences in the connection between the regulatory Bateman domain and the catalytic domain.
Strikingly, neither filament conformation matches the ~130 Å spacing of the human IMPDH2 octamer in its crystal lattice (Figure 20A). Rather, the expanded conformation most closely resembles the ATP-bound active octamer conformation first reported for Pseudomonas IMPDH (Labesse et al., 2013), while the collapsed conformation is most similar to the Ashbya GDP-bound inhibited form (Figure 23D) (Buey et al., 2015). The identification of 144 conditions that shift the conformation of IMPDH between collapsed and expanded conformations suggests the potential for cooperativity in the transition of polymerized octamers between states. This model would imply that polymers may function to coordinate octamer conformational changes en masse.
GTP bound human IMPDH2 takes on a collapsed octamer conformation and ATP induced polymerization does not protect against allosteric GTP inhibition
We now wished to confirm that GTP-bound human IMPDH2 adopts a collapsed conformation similar to Ashbya IMPDH with high-resolution cryo- electron microscopy. We determined the 8.7 Å structure of GTP-bound Y12A
IMPDH2 (Figure 24A). At this resolution alpha-helices are well-defined, allowing us to model the overall conformation of the protein by rigid-body fitting of the human IMPDH catalytic and Bateman domains into the EM density (Figures 24B & 24C). Comparison of this atomic model to the Ashyba
GDP-bound IMPDH2 crystal structure (PDB ID 4Z87) unambiguously demonstrates that under these conditions Y12A assumes the collapsed conformation (Figure 24D) and not the extended form (Figure 24E).
Since both polymerized and octomeric IMPDH2 can adopt the expanded and collapsed conformations, we hypothesized that polymerization 145 might stabilize the expanded conformation, reducing negative feedback inhibition and allowing higher steady state GTP accumulation in certain biological contexts. To test this, we measured the effect of GTP on the activity of wild type and mutant IMPDH2 under polymerization conditions. Contrary to our hypothesis, we observed comparable low millimolar GTP-mediated inhibition of all three forms (Figure 24F). Although we cannot rule out subtle differences in the kinetics or cooperativity of inhibition in the two contexts.
146
Figure 24. Human IMPDH adopts the collapsed, inhibited conformation when bound to GTP.(A) Cryo-EM reconstruction of human Y12A IMPDH2 non-polymerizing mutant in the presence of 0.1 mM ATP and 0.1 mM GTP, at 8.7 Å resolution. Both catalytic (greens) and Bateman (pinks) domains are well resolved. (B) The atomic model of IMPDH2-ATP- GTP, generated by fitting the catalytic domain and Bateman domains into the cryo-EM structure as two separate rigid bodies. (C) Close-up view of a single IMPDH monomer fit in the cryo-EM structure. (D) A monomer of the A. gossypii IMPDH-GDP crystal structure (gray) (PDB ID 4Z87) aligned to the human IMPDH2-ATP-GTP model (color) on the catalytic domain (green). A rotation of ~3° relates the human Bateman domain (pink) to the A. gossypii structure. (E) The same alignment as (D), but to the P. aueriginosa IMPDH-ATP monomer, in which the Bateman domains are related by a 14° rotation. (F) Dose-dependent inhibition of wild type and mutant IMPDH2 activity by GTP. IMPDH2 was incubated with 1 mM ATP and 3 mM NAD+ for 10 minutes at room temperature to promote polymerization and then the indicated concentration of GTP was added for an additional 10 minutes prior to reaction initialization by addition of 5 mM IMP. Mean values are plotted and were fit to a three-parameter dose-response curve. Each data point represents 3-6 replicate reactions. 95% confidence intervals are plotted as dotted lines of the corresponding color. (G) Model: human IMPDH exists in a conformational equilibrium between an expanded active conformation and a collapsed inactive conformation that can be shifted by binding to substrates or GTP. Unlike other metabolic filaments, the conformational equilibrium of IMPDH is unaffected by polymerization because the filament form can accommodate both active and inactive conformations.
147
Discussion
IMPDH is a rate limiting enzyme that is crucial for the process of de novo purine nucleotide synthesis. IMPDH catalyzes the reaction which converts IMP into XMP which is the first committed step in guanine nucleotide production. Because of this, it is considered to be the gatekeeper for the production of guanine nucleotides therefore making it essential to almost all organisms studied thus far.
IMPDH also has important clinical relevance. Highly proliferative cells, such as cancer cells, have a great demand for guanine nucleotides and this demand cannot be met using the salvage pathway (Natsumeda et al.,
1984). In fact studies have shown that IMPDH2 in particular is highly upregulated in cancers such as leukemias (Konno et al., 1991; Nagai et al.,
1991) This makes IMPDH an enzyme of consequence in cancer.
Within the past ten years the finding that this enzyme is capable of assembling into filamentous structures has been of great interest. Within mammalian cells, it was found that IMPDH can be triggered to polymerize within mammalian cells by use of IMPDH in inhibitors (Ji et al., 2006). In addition, these IMPDH filaments were found within the peripheral blood mononuclear cells of Hepatitis C patients undergoing ribavirin treatment
(Carcamo et al., 2011). In vitro, IMPDH filament assembly was also found to be triggered by ATP (Labesse et al., 2013). 148
Studies have found that the incubation of purified P. aeruginosa
IMPDH with ATP is associated with increased enzymatic activity (Labesse et al., 2013) which suggests that polymerization of this bacterium species upregulates IMPDH activity. On the other hand, ATP had no effect on the activity of Ashbya gossypii IMPDH (Buey et al., 2015), which suggests that the effect of ATP on IMPDH activity might vary from species to species.
Nevertheless, how polymerization impacts human IMPDH2 enzymatic activity was not clear. Furthermore, previous studies examining IMPDH2 enzymatic activity have failed to directly tie IMPDH2 polymerization with enzymatic activity. Our study sought to remedy this.
Unlike previous studies, our study aimed to directly tie the polymerization of human IMPDH2 to its enzymatic activity level. In order to do this, we utilized electron microscopy to capture a visual representation of human IMPDH2—polymerized as well as non-polymerized—under the exact experimental conditions which we measured enzymatic activity. The inability of the non-polymerizing mutants to assemble in vitro showed that these mutations specifically affected inter-octamer contacts and were not influenced by cellular factors just as we hypothesized. We then compared the activity of polymerized and non-polymerized human IMPDH2 and found that they were comparable. These results are in contradiction to the findings of a previous group which reported a greater than fourfold increase in human
IMPDH2 activity upon incubation with ATP (Scott et al., 2004) although they 149 did not connect this with IMPDH2 polymerization. On the other hand, our findings are consistent with the more recent findings from other labs which have been unable to corroborate the finding that ATP increased human
IMPDH2 activity (Buey et al., 2015; Thomas et al., 2012). In addition, as of now, no groups have found any increase in activity upon human IMPDH1 polymerization (Buey et al., 2015; Labesse et al., 2013; Mortimer and
Hedstrom, 2005). One group has proposed a possible explanation for the irreproducible results from the single lab. One possibility is that the increased activity could be the result of contaminating nucleotidases
(Mortimer and Hedstrom, 2005). Such contamination can results from incomplete purification. The stimulation of IMPDH activity by various nucleotides has been traced to nucleotidase contamination (Holmes et al.,
1974).
Substrate affinity is an important aspect of enzymatic activity.
Previous to our study, there have been no studies concerning the impact that human IMPDH2 polymerization has on the affinity of the substrates NAD and IMP. In this study we compared the NAD and the IMP affinity for polymerized and non-polymerized human IMPDH2 and to our surprise we found that there was no increase in the substrate affinity for polymerized human IMPDH2. We hypothesized that there would be increased affinity of substrates upon IMPDH2 polymerization because we believed that filament assembly increased nucleotide synthesis. Therefore, when we saw no increase 150 in polymerized IMPDH2 activity, we believed that increased substrate affinity was the method that polymerized human IMPDH2 would use instead to increase nucleotide synthesis. We found that ATP actually decreased the affinity that NAD had for human IMPDH2, an affect that was reported by
(Labesse et al., 2013) for P. aeruginosa IMPDH. Labesse et al. (Labesse et al.,
2013) compared the affinity of NAD and IMP for P. aeruginosa IMPDH with and without ATP and found that ATP increased the affinity that IMP had for
IMPDH almost 50 fold, while it had the opposite effect on NAD. We believe this reduced affinity of NAD+ for IMPDH2 in the presence of ATP is likely due to competitive binding of ATP to the NAD+ site rather than an effect of polymerization since no difference in NAD+ substrate affinity was observed comparing wild type and non-polymerizing mutants.
In addition to human IMPDH2 being allosterically regulated by ATP, it is also allosterically regulated by GTP (Buey et al., 2015). GTP binds to three sites on the Bateman domain of IMPDH and induces the aforementioned collapsed or closed octamer confirmation where the finger domains of the monomers from both tetramers are brought into contact with each other. This conformational change results in inhibition of IMPDH enzymatic activity (Buey et al., 2015). The authors discussed in great detail this “closed” octamer conformation of IMPDH, which contrasts with the open octamer conformation promoted by ATP. We discovered that filaments could also be composed of collapsed octamers in addition to the previously 151 discovered open octamer filaments. This discovery displays the flexibility of the inter-octamer contacts within the filaments, which allows them to accommodate both types of conformations, which differ by ~20 Å in helical rise. The ability of IMPDH2 filaments to accommodate multiple functional states of the enzyme in different conformations is unexpected. Recent structures of the metabolic filament CTP synthase have shown that, while different ligands influence the conformation of the enzyme, a single conformation is stabilized in the filament, creating a direct link between assembly and enzyme activity (Barry et al. 2014). Thus, the ability of IMPDH filaments to accommodate different functionally important conformational states while remaining assembled in the filament is unusual. The fact that we found filament segments enriched in one conformation or the other, however, suggests that there may be cooperativity in the conversion of polymerized octamers between the two conformations. In addition, we discovered a small population of filaments which contained bent octamers.
This suggests that in certain cases octamers might occupy conformational states that are intermediate in the transition between extended and collapsed, with IMPDH2 monomers existing in mixed states within a single octamer. We hypothesized that ATP competed for binding with GTP at the two canonical nucleotide binding sites in the Bateman domain. Furthermore, we hypothesized the polymerization triggered by ATP would stabilize the open confirmation of the octamers, thus making it difficult for GTP to bind 152 and inhibit the enzyme. Our results did not support our hypothesis. By way of electron microscopy, we discovered a previously unknown phenomenon which was that closed octamers polymerized in wild type IMPDH2. The closed octamers from the filament disrupting mutants were incapable of polymerizing, but both polymerizing and non-polymerizing IMPDH2 activities were inhibited by GTP. This demonstrated that closed octamers, whether in filamentous form or in separate units, result in IMPDH2 having reduced activity levels. This would explain why there was no difference between polymerized and non-polymerized IMPDH2 with respect to GTP inhibition.
So why does IMPDH2 polymerize? The answer to this question may lie in the moonlighting functions that IMPDH possesses which have nothing to do with its enzymatic activity. IMPDH is capable of moving back and forth from the cytoplasm to the nucleus and this movement is regulated by the cell cycle as well as cellular conditions (Kozhevnikova et al., 2012). While in the cytoplasm, IMPDH has been observed to be associated with polyribosomes
(Mortimer et al., 2008), which suggests that it may have some sort of involvement in translation regulation. Within the Drosophila, IMPDH localizes to the nucleus in response to increased oxidative stress and during the G1 phase of cell cycle (Kozhevnikova et al., 2012). Depletion of deoxynucleotide pools also results in the nuclear buildup of IMPDH
(Kozhevnikova et al., 2012). Within the nucleus, IMPDH is known to act as a 153 transcriptional repressor for a variety of different genes by binding directly to regulatory sequences within the genome (Kozhevnikova et al., 2012). Genes that are regulated by IMPDH within the Drosophila include all four of the histone genes and E2F which is the major regulator of the G1/S transition, and other genes involved mainly in transcriptional regulation, signaling, and development (Kozhevnikova et al., 2012). The effects of IMPDH inhibition can be seen both on a transcriptional and protein level (Kozhevnikova et al.,
2012). In light of this, it is easy to imagine that filament assembly and disassembly may be a method that the cell uses to regulate IMPDH2 participation in these activities. Assembling IMPDH into filaments may serve as an “on” switch by limiting the ability of IMPDH to act as a transcriptional repressor. This could result in the prolonged expression of a number of genes all at once. Alternatively, the disassembly of IMPDH may be an “off” switch, resulting in the efficient repression of a number of genes at once. Thus, filament assembly and disassembly could be a way for the cell to induce relatively quick changes in metabolism by influencing the expression of a group of genes at one time in response to sudden changes. Filament assembly and disassembly has been found to take place within as little as 1-15 minute’s time (Calise et al., 2014; Juda et al., 2014).
154
Future directions
The question as to whether or not polymerization regulates IMPDH involvement in gene expression can be tested using S275L mutants as well as the non-polymerizing mutants such as Y12A. These mutants can be expressed in cells and then the mRNA as well as the protein levels of the
IMPDH2 regulated genes can be determined using qPCR and a Western blot, respectively. We would expect that if filament assembly does indeed regulate
IMPDH2’s involvement in gene regulation, then we would see a difference between the gene expression levels when these two cell types are compared.
Because IMPDH2 is constitutively compartmentalized into filaments within the S275L expressing cells, we would expect that IMPDH would be unable to transfer to the nucleus to suppress gene expression. This would result in a higher number of transcripts and more protein expression for IMPDH regulated genes such as the histone genes and E2F. On the other hand, because IMPDH2 cannot assemble into filaments in Y12A expressing cells and thus is always free to relocate to the nucleus and suppress the activity of its target genes, I expect that the gene expression of the histone genes, E2F, etc. would have lower transcript levels as well as lower protein expression as compared to S275L expressing cells.
We can also determine if IMPDH2 ability to bind to DNA is affected by filament assembly by performing chromatin immunoprecipitation in S275L cells and in Y12A cells. If polymerization of IMPDH2 prevents its ability to 155 bind to DNA, there would be reduced DNA binding in the S275L cells as compared to the Y12A cells.
156
Chapter 4: Models describing possible roles for filament assembly 157
Increased awareness of the significance of metabolic enzyme polymerization
The discovery of assemblies within cells is increasing quickly and it has been noted that metabolic enzymes are more likely than other proteins to form these assemblies (O'Connell et al., 2012; Wilson and Gitai, 2013; Yeates et al., 2010). This suggests that the assembly of metabolic structures serves an important role in cellular metabolism (Petrovska et al., 2014). The assembly of metabolic enzymes can be found in a variety of organisms such as bacteria, yeast, flies, and mammals (Ingerson-Mahar et al., 2010; Liu, 2010;
Noree et al., 2010). In the yeast cell, glutamine synthetase as well as yeast
CTPS1—Ura7—have each been found to assemble into their own separate structures (Noree et al., 2010). These assemblies occur in response to a depletion of carbon sources (i.e. glucose) and/or the energy source ATP (Noree et al., 2010).
The assembly of metabolic enzymes has also been discovered within mammalian cells. Some examples include acetyl-CoA carboxylase upon exposure to citrate (Beaty and Lane, 1983) and of course the nucleotide producing enzymes IMPDH and CTPS which were the focus of the work in this thesis.
158
Exploring the role of CTPS filament assembly
Most studies in the literature concerning CTPS assembly within mammalian cells involve the use of inhibitors, such as DON, which directly inhibit CTPS activity. However, it has also been demonstrated that the inhibition of IMPDH results in the assembly of CTPS filaments in a small percentage of cells (Chang et al., 2015). CTPS has been found within the filaments of HEp-2 cells obtained from a specific manufacturer called INOVA
Diagnostics Inc. as well, but how these filaments came to be assembled within these cells is unknown (Carcamo et al., 2011). More recently, using purified protein, a group at the University of Washington found that polymerized human CTPS actually has increased catalytic activity as compared to its non-polymerized form (unpublished). This is the complete opposite of what was found for E.coli CTPS (Barry et al., 2014). In addition to this, the University of Washington group also found that human CTPS actually polymerized in the presence of its substrates UTP, ATP, and GTP, while it disassembled in the presence of GTP (unpublished). Again, just the opposite was found for E.coli CTPS—CTP induced CTPS polymerization in this bacteria while substrates induced the disassembly of CTPS (Barry et al.,
2014). The University of Washington group determined that these differences between the behaviors of the human and E. coli CTPS were due to drastic dissimilarities in the structure of the filaments that these two organism form
(unpublished). This difference in structure was traced to an insert located 159 within the GAT domain of human CTPS1 which mediates human filament assembly (unpublished). This GAT-GAT assembly interaction utilizes an interface that is different from what is used by E.coli enzyme (unpublished).
Exploring the physiological role of IMPDH filament assembly
The bulk of studies carried out on the polymerization of nucleotide biosynthetic enzymes in mammalian cells have focused on IMPDH filaments.
Under typical, nutrient rich culture conditions, with few exceptions (i.e. mouse embryonic stem cells), IMPDH is not constitutively polymerized within mammalian cells. But, IMPDH will polymerize in response to depleted cellular GTP pools. The depletion of GTP can result from perturbation at a number of different points along the purine nucleotide biosynthesis pathway
—glutamine depletion (glutamine is required upstream of IMPDH in the de novo pathway), THF depletion (also required upstream of IMPDH), inhibition of GMP synthetase (an enzyme downstream of IMPDH) and of course, inhibition of IMPDH itself. Nevertheless, the purpose that IMPDH polymerization fulfills within the mammalian cell has been unclear. We attempted to address this unknown using purified recombinant IMPDH2. As a result of our analysis, we finally determined that the ATP triggered assembly of human IMPDH2 did nothing to alter the protein’s catalytic properties. This included enzymatic activity (Figure 21D), substrate affinity
(Figure 21E-H), and the ability to resist allosteric inhibition (Figure 24F). 160
The in vitro finding that polymerization did not alter enzymatic activity was confirmed in vivo using isotopically labeled glycine. Therefore, despite us having answered some very important questions concerning the catalytic activity of IMPDH filaments, we are still left with an incomplete understanding regarding the purpose of IMPDH polymerization.
Alternatively, filament assembly may serve to regulate IMPDH in a way that has nothing to do with its enzymatic activity. Furthermore, filament assembly may serve to affect the expression of other genes. Kozhevnikova et al. (Kozhevnikova et al., 2012) have demonstrated that within Drosophila cells, IMPDH can bind to single stranded DNA and regulates the expression of a number of different genes by acting as a transcriptional repressor. These
IMPDH regulated genes include the five histones, E2F (a transcription factor important for cell cycle advancement), as well as other genes involved in development, signaling, and transcriptional regulation such as JIL1 and Mlc2
(Kozhevnikova et al., 2012). It has also been demonstrated that human
IMPDH is capable of binding to DNA and RNA in vitro and in vivo and that human IMPDH is much more selective when it comes to the sequences that it binds to as compared to microbial IMPDH (McLean et al., 2004). This supports the idea that human IMPDH may bind to specific regulatory sequences of genes within the cell. Furthermore, analogous to what was found in the Drosophila (Kozhevnikova et al., 2012), McLean et al. (McLean et al., 2004) found that human IMPDH is located in both the cytoplasm and 161 the nucleus of cells. Therefore, it is possible that just like the Drosophila
IMPDH, human IMPDH regulates genes expression and a study to determine the genes that human IMPDH regulates would be important. To that end, first a ChIP assay would have to be performed to determine where within the genome human IMPDH binds. This would clarify to which regulatory sequences of which genes IMPDH binds. As the next step, it would have to be determined in what way IMPDH regulates these genes, by acting as a transcriptional activator, or by acting as a transcriptional repressor as found in the Drosophila, or both. Therefore, polymerization may be a way for the cell to prevent IMPDH2 from inhibiting or activating the expression of certain genes that the cell may need to be turned on—or off—under the given conditions. The non-polymerizing IMPDH2 mutants—R356A and Y12A—are not only incapable themselves of assembling into filaments (Figure 25, anti- myc staining), they also inhibit the ability of endogenous, wild type IMPDH from assembling into filaments (Figure 25, anti IMPDH staining). This means that R356A and Y12A are dominant negative making them ideal for addressing the purpose of human IMPDH polymerization in cells.
162
Figure 25. Non-polymerizing mutants Y12A and R356A prevent IMPDH polymerization in a dominant negative manner Cells with high expression of the mutant protein (indicated with the white arrow) show that not only are the mutants incapable of self-polymerization (anti- myc staining), they also prevent wild type, endogenous from assembling polymerizing (anti- IMPDH staining).
Exploring the physiological role of IMPDH filament assembly: the lymphocyte
Interestingly, our lab has demonstrated that upon activation of lymphocytes isolated from the spleen of mice, IMPDH assembles into 163 filaments (unpublished). An immediate assumption would be that in this cellular context, polymerization does increase IMPDH activity in order to accommodate the huge increase in guanine nucleotide demand that comes with clonal expansion. Another possibility is that polymerization does not increase enzymatic activity but rather serves a function separate from the enzymatic activity of IMPDH. The redistribution of IMPDH could trigger a change in the expression of other genes on a transcriptional (McLean et al.,
2004; Kozhevnikova et al., 2012 ) or a translational level (Mortimer et al.,
2008) which may facilitate the clonal expansion of the lymphocytes. As a transcriptional regulator, the redistribution of IMPDH into filaments may allow the expression and/or inhibition of certain genes necessary for the lymphocyte to successfully undergo clonal expansion and mount an immune response. In order to determine this, first a ChIP assay can be done to determine to which one of the regulation sequences of which genes the mouse
IMPDH binds. It can then be determined if IMPDH acts as a repressor or an activator for the expression of these genes. Once this has been determined, non-polymerizing mutants (Y12A) can be expressed within a set of T-cells, while expressing an equivalent amount of wild type IMPDH into another group of T-cells. The T-cell receptors on both groups of cells can then be stimulated and gene expression in cells expressing the non-polymerizing mutant Y12A and wild type IMPDH can be compared using RNAseq, paying special attention to differences found in IMPDH regulated genes. A model for 164 filament assembly in T-cells could be that naïve, resting T-cells, have diffuse
IMPDH which works to inhibit and/or activate genes who’s expression, or inexpression, are important for maintaining a resting state (Figure 26A). For example, IMPDH may suppress genes that promote proliferation and cell growth such as E2F, while it promotes the expression of genes that negatively regulate cell growth such as p21. Upon activation of the T-cell by the presentation of the antigen, the naïve T-cells are signaled to assemble
IMPDH into filaments, thus sequestering the protein and making it unavailable to influence gene expression (Figure 26B). These genes then either become activated or become inactivated which promotes the rapid proliferation of T-cells.
Figure 26. Role of IMPDH filament assembly in T-cell activation.(A) In its diffuse form, IMPDH may be free to activate or repress genes within the T-cell which maintains its naïve state. (B) Stimulation of the T-cell receptor, may signal the assembly of IMPDH into filaments, thus preventing it from regulating the expression of genes it regulates in its non-polymerized form. As a result, T-cells are free to proliferate and therefore undergo clonal expansion.
165
Exploring the physiological role of IMPDH filament assembly: tumor cells
Interestingly our lab discovered IMPDH filaments in many, but not all, of the tumor cells occupying the cortex area of tumors within mice. These tumors were extracted from SCID mice that had been injected with MDA-
MB-231 triple negative breast cancer cells (unpublished). Highly proliferative cells have been found to be heavily dependent on the de novo pathway to provide the guanine nucleotide needed (Majd et al., 2014), so unsurprisingly, it has been found that various cancers upregulate IMPDH gene expression
(Tong et al., 2009). Thus, in this particular cellular context, it would be useful to first compare the enzymatic activity of IMPDH in tumor cells containing filaments with adjacent tumor cells containing no filaments. This can be done by first expressing the Y12A mutant in one set of MDA-MB-231 cells and expressing the S275L mutant in another set of MDA-MB-231 cells. The incorporation of isotopically labeled glycine into a product immediately downstream of IMPDH, such as XMP or GMP can then be quantified and compared between these two cell types. It is possible that just as we discovered for HEK cells, no difference in IMPDH activity will be found when these two MDA-MB-231 cells types are compared. On the other hand, just as might be possible in the lymphocytes, the redistribution or relocation of
IMPDH into filaments may serve to regulate the ability of IMPDH to 166 influence the expression of certain genes. A change in the metabolism or advancement/development of the tumor may prompt the
Figure 27. Role of IMPDH filament assembly in cancer development.
assembly of IMPDH into filaments and thus alter the expression of certain genes which are required to accommodate these changes (Figure 27). Such changes could be a sudden acceleration of growth rate or a tumor cell switching to an invasive phenotype (Figure 27). 167
Therefore, one can imagine, if IMPDH polymerization is triggered, it might be unable to suppress genes involved in cell cycle advancement, be unable to suppress the expression of developmental genes such as Notch and
Wnt (Andersson et al., 2011; Cadigan and Nusse, 1997) which often become activated in tumor cells (Cadigan and Nusse, 1997; D'Angelo et al., 2015; Hu et al., 2012; Wang et al., 2014), or be unable to suppress the expression of other genes that promote other tumorgenic properties such as invasion
(Figure 27). At the same time, it may be unable to promote genes that negatively regulate cell cycle progression and other tumorgenic properties
(Figure 27).
Another important experiment would be to create stable MDA-MB-231
triple negative breast cancer cell lines expressing S275L or Y12A. Gene expression levels between filament assembling tumor cells (expressing
S275L) and tumor cells without IMPDH polymerization (expressing Y12A) can then be compared using RNAseq. Furthermore, it can be determined if there are any differences in the tumorigenic properties of filament containing tumor cells (cells expressing S275L mutant), and non-filament containing tumor cells (cells expressing Y12A mutant). Properties that can be compared include metastatic capability, invasion abilities, proliferation, migration capabilities, and non-adherent cell growth. These can be analyzed in vitro, within cells in culture, but can also be examined in vivo by injecting these 168
S275L and Y12A expressing cells within SCID mice and then examining and comparing tumor development (i.e. tumor size/weight and metastases) within the S275L MDA-MB-231 cell injected mice, and the Y12A MDA-MB-231 cell injected mice. Additionally, we can determine if any discovered differences in properties can be traced to particular genes that IMPDH regulates. This can be achieved by knocking down the expression level, or overexpression of these
IMPDH regulated genes, one at a time in wild type MDA-MB-231 cells, and then testing tumorigenic properties to determine which of these properties are affected.
Exploring the physiological role of IMPDH filament assembly: open versus closed octamer filaments
Our work has demonstrated that not only does ATP induce conformational changes in human IMPDH, the nucleotide GTP does as well.
We found that the binding of GTP to human IMPDH2 induced a closed or compressed octamer conformation. Furthermore, these closed octamers of the wild type protein assembled themselves into filaments. In addition to this, a bent or mixed octamer filament was also discovered (Figure 23). The biological significance of these closed octamer filaments is unknown considering that both closed individual and polymerized octamers experience the same inhibition of activity (Figure 24F). It is possible that the assembly of the closed octamers into filaments serves no purpose. The octamers of 169 purified wild type IMPDH, whether open or closed, may just have a natural proclivity to assemble themselves into filaments in vitro due to the symmetry of the enzyme’s octameric (and tetrameric) structure. The fact that wild type
IMPDH can still spontaneously polymerize into octamers and even short filaments in the absence of ATP does support this idea. But, upon exposure to
ATP, basically all of the IMPDH protein is octameric and contained within long filaments. Therefore, it may be that because the addition of ATP promotes more of the protein to assemble into octamers by stabilizing the octameric structure, more robust filament assembly naturally follows. On the other hand, without ATP, less of the protein is found in octamers, and therefore filaments are smaller and more infrequent. It would be interesting to determine which filament types—open, closed, or mixed—bundles into the huge filaments that are seen using immunofluorescence within cells (Figure
23). It may be possible that the composition of the bundles varies depending on how the filament is induced. For example, it is possible that the composition of the IMPDH filaments assembled within activated T-cells is different from that of the filaments induced in PBMC by ribavirin therapy.
Exploring the physiological role of filament co-assembly
The assembly of IMPDH with CTPS in cells may serve an entirely different purpose than when assembled alone. It is possible that when assembled with CTPS, there are other proteins, perhaps CTPS itself, which 170 alter the enzymatic activity of IMPDH by increasing its activity. We have identified a candidate filament component—histidine ammonia lysase
(HAL)—which is known to be involved in histidine catabolism. This enzyme takes histidine and converts it into urocanic acid. Most importantly, during this reaction, ammonia is released. Interestingly, CTPS utilizes ammonia during its reaction and the protonated form of ammonia, ammonium, actually increases IMPDH activity 100-fold. It is possible that under nutrient rich conditions within the cell, IMPDH, CTPS and HAL are diffuse and have little to no contact with each other within the cells (Figure 28, step 1). A drop in the levels of precursor necessary for nucleotide synthesis, or an increased demand for these precursors, such as increased glutamine demand in cancer cells, may trigger the assembly of CTPS and IMPDH into filaments. HAL may then be attracted to this filamentous form of the proteins or be recruited to assemble with CTPS and IMPDH in order to become a source of ammonia for CTPS and activate IMPDH activity (Figure 28, steps 2 & 3). In this way,
HAL would stimulate both CTPS and IMPDH to increase the nucleotide production under this starved state (Figure 28, step 4). Also, ammonia is a by-product of catabolism in general, and can become toxic to cells (Slivac et al., 2010). The fact that this by-product can be used as a source for CTPS to produce pyrimidines and can enhance the enzymatic activity of IMPDH may be a way for the cell to counteract this toxicity and promote survival by increasing the nucleotide pools needed for cellular growth. 171
1 2 3
4
Figure 28. A hypothesis for the role of IMPDH, CTPS and HAL co-assembly in nucleotide synthesis.(1) In nutrient enriched conditions, CTPS, IMPDH and HAL are diffuse and have no contact with each other. (2) Upon depletion of nucleotide precursors, all three proteins co-assemble within filaments. (3) HAL may then provide the ammonia needed to activate IMPDH and for CTPS to carry out its reaction. (4) This then results in enhanced nucleotide production
172
Conclusion
There is still much to learn in regards to the polymerization of IMPDH and CTPS into filaments. These structures merit study because they likely have clinical implications. That they have clinical importance is supported by the fact that they assemble within the PBMC of patients undergoing drug therapy, they assemble within activated lymphocytes, and they can be found within tumor cells. By discovering the significance of this assembly within cells, we might be able to better treat neoplastic diseases as well as diseases of the immune system. At the very least, this knowledge would expand our general understanding concerning how polymerization of metabolic enzymes regulates their function and activity within the cell.
LIST OF REFERENCES
173
LIST OF REFERENCES
Alberts, B., Johnson, A., Lewis, J.,et al (2002). Molecular Biology of the Cell, 4th edition edn (New York: Garland Science).
Alexandre, T., Raynal, B., and Munier-Lehmann, H. (2015). Two classes of bacterial IMPDHs according to their quaternary structures and catalytic properties. PLoS One 10, e0116578.
Alt, J., Potter, M.C., Rojas, C., and Slusher, B.S. (2015). Bioanalysis of 6- diazo-5-oxo-l-norleucine in plasma and brain by ultra-performance liquid chromatography mass spectrometry. Anal Biochem 474, 28-34.
Anderson, L.E., Ehrlich, J., Sun, S.H., and Burkholder, P.R. (1956). Strains of Streptomyces, the sources of azaserine, elaiomycin, griseoviridin, and viridogrisein. Antibiot Chemother (Northfield) 6, 100-115.
Andersson, E.R., Sandberg, R., and Lendahl, U. (2011). Notch signaling: simplicity in design, versatility in function. Development 138, 3593-3612.
Aughey, G.N., Grice, S.J., and Liu, J.L. (2016). The Interplay between Myc and CTP Synthase in Drosophila. PLoS Genet 12, e1005867.
Aughey, G.N., Grice, S.J., Shen, Q.J., Xu, Y., Chang, C.C., Azzam, G., Wang, P.Y., Freeman-Mills, L., Pai, L.M., Sung, L.Y., et al. (2014). Nucleotide synthesis is regulated by cytoophidium formation during neurodevelopment and adaptive metabolism. Biol Open 3, 1045-1056. 174
Aughey, G.N., and Liu, J.L. (2015). Metabolic regulation via enzyme filamentation. Crit Rev Biochem Mol Biol 51, 282-293.
Averaimo, S., and Nicol, X. (2014). Intermingled cAMP, cGMP and calcium spatiotemporal dynamics in developing neuronal circuits. Front Cell Neurosci 8, 376.
Azzam, G., and Liu, J.L. (2013). Only one isoform of Drosophila melanogaster CTP synthase forms the cytoophidium. PLoS Genet 9, e1003256.
Bacus, S.S., Kiguchi, K., Chin, D., King, C.R., and Huberman, E. (1990). Differentiation of cultured human breast cancer cells (AU-565 and MCF-7) associated with loss of cell surface HER-2/neu antigen. Mol Carcinog 3, 350- 362.
Barat, B., and Wu, A.M. (2007). Metabolic biotinylation of recombinant antibody by biotin ligase retained in the endoplasmic reticulum. Biomol Eng 24, 283-291.
Barclay, R.K., and Phillipps, M.A. (1966). Effects of 6-Diazo-5-oxo-L- norleucine and Other Tumor Inhibitors on the Biosynthesis of Nicotinamide Adenine Dinucleotide in Mice. Cancer research 26, 282-286.
Barger, J.F., and Plas, D.R. (2010). Balancing biosynthesis and bioenergetics: metabolic programs in oncogenesis. Endocrine-Related Cancer 17, R287- R304.
Barry, R.M., Bitbol, A.F., Lorestani, A., Charles, E.J., Habrian, C.H., Hansen, J.M., Li, H.J., Baldwin, E.P., Wingreen, N.S., Kollman, J.M., et al. (2014). Large-scale filament formation inhibits the activity of CTP synthetase. eLife 3, e03638.
Beaty, N.B., and Lane, M.D. (1983). Kinetics of activation of acetyl-CoA carboxylase by citrate. Relationship to the rate of polymerization of the enzyme. The Journal of biological chemistry 258, 13043-13050.
175
Becroft, D.M., and Phillips, L.I. (1965). HEREDITARY OROTIC ACIDURIA AND MEGALOBLASTIC ANAEMIA: A SECOND CASE, WITH RESPONSE TO URIDINE. Br Med J 1, 547-552.
Bentley, R. (2000). Mycophenolic Acid: a one hundred year odyssey from antibiotic to immunosuppressant. Chem Rev 100, 3801-3826.
Berg JM, T.J., Stryer L. ( 2002). Biochemistry, 5th edition edn (New York: W H Freeman).
Blair, E., Redwood, C., Ashrafian, H., Oliveira, M., Broxholme, J., Kerr, B., Salmon, A., Östman-Smith, I., and Watkins, H. (2001). Mutations in the γ2 subunit of AMP-activated protein kinase cause familial hypertrophic cardiomyopathy: evidence for the central role of energy compromise in disease pathogenesis. Human Molecular Genetics 10, 1215-1220.
Bolzoni, M., Chiu, M., Accardi, F., Vescovini, R., Airoldi, I., Storti, P., Todoerti, K., Agnelli, L., Missale, G., Andreoli, R., et al. (2016). Dependence on glutamine uptake and glutamine addiction characterize myeloma cells: a new attractive target. Blood 128, 667-679.
Bond, M.R., and Hanover, J.A. (2015). A little sugar goes a long way: the cell biology of O-GlcNAc. J Cell Biol 208, 869-880.
Bowne, S.J., Liu, Q., Sullivan, L.S., Zhu, J., Spellicy, C.J., Rickman, C.B., Pierce, E.A., and Daiger, S.P. (2006). Why do mutations in the ubiquitously expressed housekeeping gene IMPDH1 cause retina-specific photoreceptor degeneration? Invest Ophthalmol Vis Sci 47, 3754-3765.
Broderick, J.B., Duffus, B.R., Duschene, K.S., and Shepard, E.M. (2014). Radical S-adenosylmethionine enzymes. Chem Rev 114, 4229-4317.
Broschat, K.O., Gorka, C., Page, J.D., Martin-Berger, C.L., Davies, M.S., Huang, H.-c., Gulve, E.A., Salsgiver, W.J., and Kasten, T.P. (2002). Kinetic Characterization of Human Glutamine-fructose-6-phosphate Amidotransferase I: POTENT FEEDBACK INHIBITION BY GLUCOSAMINE 6-PHOSPHATE. Journal of Biological Chemistry 277, 14764-14770.
176
Brown, G.J., Jr., and Cohen, P.P. (1960). Comparative biochemistry of urea synthesis. 3. Activities of urea-cycle enzymes in various higher and lower vertebrates. The Biochemical journal 75, 82-91.
Buey, R.M., Ledesma-Amaro, R., Velazquez-Campoy, A., Balsera, M., Chagoyen, M., de Pereda, J.M., and Revuelta, J.L. (2015). Guanine nucleotide binding to the Bateman domain mediates the allosteric inhibition of eukaryotic IMP dehydrogenases. Nat Commun 6, 8923.
Busuttil, R.W.a.K., G.B (2014). Transplantation of the Liver, 3 edn (Saunders).
Cadigan, K.M., and Nusse, R. (1997). Wnt signaling: a common theme in animal development. Genes & development 11, 3286-3305.
Calise, S.J., Carcamo, W.C., Krueger, C., Yin, J.D., Purich, D.L., and Chan, E.K. (2014). Glutamine deprivation initiates reversible assembly of mammalian rods and rings. Cell Mol Life Sci 71, 2963-2973.
Calise, S.J., Purich, D.L., Nguyen, T., Saleem, D.A., Krueger, C., Yin, J.D., and Chan, E.K. (2016). 'Rod and ring' formation from IMP dehydrogenase is regulated through the one-carbon metabolic pathway. J Cell Sci 129, 3042- 3052.
Canellakis, E.S. (1956). Pyrimidine metabolism. I. Enzymatic pathways of uracil and thymine degradation. The Journal of biological chemistry 221, 315- 322.
Carcamo, W.C., Calise, S.J., von Muhlen, C.A., Satoh, M., and Chan, E.K. (2014). Molecular cell biology and immunobiology of mammalian rod/ring structures. Int Rev Cell Mol Biol 308, 35-74.
Carcamo, W.C., Satoh, M., Kasahara, H., Terada, N., Hamazaki, T., Chan, J.Y., Yao, B., Tamayo, S., Covini, G., and von Mühlen, C.A. (2011). Induction of cytoplasmic rods and rings structures by inhibition of the CTP and GTP synthetic pathway in mammalian cells. PloS one 6, e29690.
Carlton, J.M., Hirt, R.P., Silva, J.C., Delcher, A.L., Schatz, M., Zhao, Q., Wortman, J.R., Bidwell, S.L., Alsmark, U.C., Besteiro, S., et al. (2007). Draft 177 genome sequence of the sexually transmitted pathogen Trichomonas vaginalis. Science 315, 207-212.
Carr, S.F., Papp, E., Wu, J.C., and Natsumeda, Y. (1993). Characterization of human type I and type II IMP dehydrogenases. The Journal of biological chemistry 268, 27286-27290.
Carrey, E.A., Campbell, D.G., and Hardie, D.G. (1985). Phosphorylation and activation of hamster carbamyl phosphate synthetase II by cAMP-dependent protein kinase. A novel mechanism for regulation of pyrimidine nucleotide biosynthesis. Embo j 4, 3735-3742.
Cassidy, L.F., and Patterson, J.L. (1989). Mechanism of La Crosse virus inhibition by ribavirin. Antimicrob Agents Chemother 33, 2009-2011.
Ceballos-Picot, I., Mockel, L., Potier, M.C., Dauphinot, L., Shirley, T.L., Torero-Ibad, R., Fuchs, J., and Jinnah, H.A. (2009). Hypoxanthine-guanine phosphoribosyl transferase regulates early developmental programming of dopamine neurons: implications for Lesch-Nyhan disease pathogenesis. Hum Mol Genet 18, 2317-2327.
Chabosseau, P., Buhagiar-Labarchède, G., Onclercq-Delic, R., Lambert, S., Debatisse, M., Brison, O., and Amor-Guéret, M. (2011). Pyrimidine pool imbalance induced by BLM helicase deficiency contributes to genetic instability in Bloom syndrome. Nature Communications 2, 368.
Champe, P.C.a.H., R.A. (1994). Biochemistry (Philadelphia: Lippincott-Raven Publishers).
Chang, C.C., Lin, W.C., Pai, L.M., Lee, H.S., Wu, S.C., Ding, S.T., Liu, J.L., and Sung, L.Y. (2015). Cytoophidium assembly reflects upregulation of IMPDH activity. J Cell Sci 128, 3550-3555.
Chapman-Smith, A., and Cronan, J.E., Jr. (1999). Molecular biology of biotin attachment to proteins. J Nutr 129, 477s-484s.
Chen, K., Zhang, J., Tastan, O.Y., Deussen, Z.A., Siswick, M.Y., and Liu, J.L. (2011). Glutamine analogs promote cytoophidium assembly in human and Drosophila cells. J Genet Genomics 38, 391-402. 178
Chen, L., and Pankiewicz, K.W. (2007). Recent development of IMP dehydrogenase inhibitors for the treatment of cancer. Curr Opin Drug Discov Devel 10, 403-412.
Coffey, G.L., Ehrlich, J., Fisher, M.W., Hillegas, A.B., Kohberger, D.L., Machamer, H.E., Rightsel, W.A., and Roegner, F.R. (1956). 6-Diazo-5-oxo-L- norleucine, a new tumor-inhibitory substance. I. Biologic studies. Antibiot Chemother (Northfield) 6, 487-497.
Colby, T.D., Vanderveen, K., Strickler, M.D., Markham, G.D., and Goldstein, B.M. (1999). Crystal structure of human type II inosine monophosphate dehydrogenase: implications for ligand binding and drug design. Proc Natl Acad Sci U S A 96, 3531-3536.
Cornuel, J.F., Moraillon, A., and Gueron, M. (2002). Participation of yeast inosine 5'-monophosphate dehydrogenase in an in vitro complex with a fragment of the C-rich telomeric strand. Biochimie 84, 279-289.
Cronan, J.E. (1990). Biotination of proteins in vivo. A post-translational modification to label, purify, and study proteins. Journal of Biological Chemistry 265, 10327-10333.
Crotty, S., Maag, D., Arnold, J.J., Zhong, W., Lau, J.Y., Hong, Z., Andino, R., and Cameron, C.E. (2000a). The broad-spectrum antiviral ribonucleoside ribavirin is an RNA virus mutagen. Nat Med 6, 1375-1379.
Curtis, P.D., and Brun, Y.V. (2010). Getting in the Loop: Regulation of Development in Caulobacter crescentus. Microbiology and Molecular Biology Reviews 74, 13-41.
D'Angelo, R.C., Ouzounova, M., Davis, A., Choi, D., Tchuenkam, S.M., Kim, G., Luther, T., Quraishi, A.A., Senbabaoglu, Y., Conley, S.J., et al. (2015). Notch reporter activity in breast cancer cell lines identifies a subset of cells with stem cell activity. Molecular cancer therapeutics 14, 779-787.
Dagg, C.P., Coleman, D.L., and Fraser, G.M. (1964). A GENE AFFECTING THE RATE OF PYRIMIDINE DEGRADATION IN MICE. Genetics 49, 979- 989. 179
De Ingeniis, J., Kazanov, M.D., Shatalin, K., Gelfand, M.S., Osterman, A.L., and Sorci, L. (2012). Glutamine versus ammonia utilization in the NAD synthetase family. PLoS One 7, e39115.
De Las Rivas, J., and Fontanillo, C. (2010). Protein-protein interactions essentials: key concepts to building and analyzing interactome networks. PLoS computational biology 6, e1000807.
Dej, K.J., and Spradling, A.C. (1999). The endocycle controls nurse cell polytene chromosome structure during Drosophila oogenesis. Development 126, 293-303.
Deutsch, S.I., Long, K.D., Rosse, R.B., Mastropaolo, J., and Eller, J. (2005). Hypothesized deficiency of guanine-based purines may contribute to abnormalities of neurodevelopment, neuromodulation, and neurotransmission in Lesch-Nyhan syndrome. Clin Neuropharmacol 28, 28- 37.
Doi, T., Hamaguchi, T., Shitara, K., Iwasa, S., Shimada, Y., Harada, M., Naito, K., Hayashi, N., Masada, A., and Ohtsu, A. (2017). NC-6004 Phase I study in combination with gemcitabine for advanced solid tumors and population PK/PD analysis. Cancer Chemother Pharmacol.
Domek-Lopacinska, K., and Strosznajder, J.B. (2005). Cyclic GMP metabolism and its role in brain physiology. J Physiol Pharmacol 56 Suppl 2, 15-34.
Duvall, L.R. (1960). New agent data summaries: 6-diazo-5-oxo-L-norleucine. Cancer Chemother Rep 7, 86-98.
Eagle, H. (1955). Nutrition needs of mammalian cells in tissue culture. Science 122, 501-514.
Eidinoff, M.L., Knoll, J.E., Marano, B., and Cheong, L. (1958). Pyrimidine Studies. Cancer research 18, 105-109.
180
Eidinoff, M.L., Knoll, J.E., Marano, B., Cheong, L. (1958). Pyrimidine Studies I. Effect of DON (6-Diazo-5-oxo-L-norleucine) on Incorporation of Precursors into Nuleic Acid Pyrimidines. Cancer research 18.
El-Hattab, A.W., and Scaglia, F. (2013). Mitochondrial DNA depletion syndromes: review and updates of genetic basis, manifestations, and therapeutic options. Neurotherapeutics 10, 186-198.
Elamin, Y.Y., Rafee, S., Osman, N., KJ, O.B., and Gately, K. (2016). Thymidine Phosphorylase in Cancer; Enemy or Friend? Cancer Microenviron 9, 33-43.
Elpeleg, O. (2003). Inherited Mitochondrial DNA Depletion. Pediatr Res 54, 153-159.
Endrizzi, J.A., Kim, H., Anderson, P.M., and Baldwin, E.P. (2004). Crystal structure of Escherichia coli cytidine triphosphate synthetase, a nucleotide- regulated glutamine amidotransferase/ATP-dependent amidoligase fusion protein and homologue of anticancer and antiparasitic drug targets. Biochemistry 43, 6447-6463.
Eriksson, B., Helgstrand, E., Johansson, N.G., Larsson, A., Misiorny, A., Noren, J.O., Philipson, L., Stenberg, K., Stening, G., Stridh, S., et al. (1977). Inhibition of influenza virus ribonucleic acid polymerase by ribavirin triphosphate. Antimicrob Agents Chemother 11, 946-951.
Evans, D.R., and Guy, H.I. (2004). Mammalian pyrimidine biosynthesis: fresh insights into an ancient pathway. The Journal of biological chemistry 279, 33035-33038.
Fairbanks, L.D., Bofill, M., Ruckemann, K., and Simmonds, H.A. (1995). Importance of ribonucleotide availability to proliferating T-lymphocytes from healthy humans. Disproportionate expansion of pyrimidine pools and contrasting effects of de novo synthesis inhibitors. The Journal of biological chemistry 270, 29682-29689.
Fasullo, M., and Endres, L. (2015). Nucleotide salvage deficiencies, DNA damage and neurodegeneration. Int J Mol Sci 16, 9431-9449.
181
Federico, A., Cardaioli, E., Da Pozzo, P., Formichi, P., Gallus, G.N., and Radi, E. (2012). Mitochondria, oxidative stress and neurodegeneration. J Neurol Sci 322, 254-262.
Ferraro, P., Nicolosi, L., Bernardi, P., Reichard, P., and Bianchi, V. (2006). Mitochondrial deoxynucleotide pool sizes in mouse liver and evidence for a transport mechanism for thymidine monophosphate. Proc Natl Acad Sci U S A 103, 18586-18591.
Fink, K., Cline, R.E., Henderson, R.B., and Fink, R.M. (1956a). Metabolism of thymine (methyl-C14 or -2-C14) by rat liver in vitro. The Journal of biological chemistry 221, 425-433.
Fink, R.M., McGaughey, C., Cline, R.E., and Fink, K. (1956b). Metabolism of intermediate pyrimidine reduction products in vitro. The Journal of biological chemistry 218, 1-7.
Firat-Karalar, E.N., and Stearns, T. (2015). Probing mammalian centrosome structure using BioID proximity-dependent biotinylation. Methods Cell Biol 129, 153-170.
Fleming, M.A., Chambers, S.P., Connelly, P.R., Nimmesgern, E., Fox, T., Bruzzese, F.J., Hoe, S.T., Fulghum, J.R., Livingston, D.J., Stuver, C.M., et al. (1996). Inhibition of IMPDH by mycophenolic acid: dissection of forward and reverse pathways using capillary electrophoresis. Biochemistry 35, 6990- 6997.
Floryk, D., and Huberman, E. (2006a). Mycophenolic acid-induced replication arrest, differentiation markers and cell death of androgen-independent prostate cancer cells DU145. Cancer letters 231, 20-29.
Floryk, D., Tollaksen, S.L., Giometti, C.S., and Huberman, E. (2004). Differentiation of human prostate cancer PC-3 cells induced by inhibitors of inosine 5'-monophosphate dehydrogenase. Cancer research 64, 9049-9056.
French, T.C., et al. (1963). Azaserine-reactive Sulfhydryl Group of 2- Formamido-IVribosylacetamide S-Phosphate : L-Glutamine Amido-ligase (Adenosine Diphosphate) The Journal of biological chemistry 288. 182
Fridland, A., Connelly, M.C., and Robbins, T.J. (1986). Tiazofurin metabolism in human lymphoblastoid cells: evidence for phosphorylation by adenosine kinase and 5'-nucleotidase. Cancer research 46, 532-537.
Fu, R., Ceballos-Picot, I., Torres, R.J., Larovere, L.E., Yamada, Y., Nguyen, K.V., Hegde, M., Visser, J.E., Schretlen, D.J., Nyhan, W.L., et al. (2014). Genotype-phenotype correlations in neurogenetics: Lesch-Nyhan disease as a model disorder. Brain 137, 1282-1303.
Galpin, J.D., Ellis, B.E., and Tanner, M.E. (1999). The Inactivation of Histidine Ammonia-Lyase by L-Cysteine Oxygen: Modification of the Electrophilic Center. Journal of American Chemical Society, 10840.
Garrett, R.H., Grisham, C. M. Biochemistry, 6th edn (Boston, MA: Cengage Learning).
Garrett, R.H.a.G., C.M (1995). Biochemistry (Saunders College Publishing).
Gish, R.G. (2006). Treating HCV with ribavirin analogues and ribavirin-like molecules. J Antimicrob Chemother 57, 8-13.
Glesne, D., Collart, F., Varkony, T., Drabkin, H., and Huberman, E. (1993). Chromosomal localization and structure of the human type II IMP dehydrogenase gene (IMPDH2). Genomics 16, 274-277.
GM., C. (2000). The Cell: A Molecular Approach. 2nd edition (Sunderland (MA): Sinauer Associates).
Goodsell, D.S. (1999). The molecular perspective: methotrexate. Oncologist 4, 340-341.
Goto, M., Omi, R., Nakagawa, N., Miyahara, I., and Hirotsu, K. (2004). Crystal structures of CTP synthetase reveal ATP, UTP, and glutamine binding sites. Structure 12, 1413-1423.
Gou, K.M., Chang, C.C., Shen, Q.J., Sung, L.Y., and Liu, J.L. (2014). CTP synthase forms cytoophidia in the cytoplasm and nucleus. Exp Cell Res 323, 242-253. 183
Gouspillou, G., and Hepple, R.T. (2016). Editorial: Mitochondria in Skeletal Muscle Health, Aging and Diseases. Frontiers in Physiology 7.
Graci, J.D., and Cameron, C.E. (2006). Mechanisms of action of ribavirin against distinct viruses. Rev Med Virol 16, 37-48.
Greenlees, J.a.L., G.A. (1956). Purine Biosynthesis and Inhibitors in Ascites Cell Tumors. Cancer research 16, 5.
Gu, J.J., Kaiser-Rogers, K., Rao, K., and Mitchell, B.S. (1994). Assignment of the human type I IMP dehydrogenase gene (IMPDH1) to chromosome 7q31.3- q32). Genomics 24, 179-181.
Gu, J.J., Spychala, J., and Mitchell, B.S. (1997). Regulation of the Human Inosine Monophosphate Dehydrogenase Type I Gene: UTILIZATION OF ALTERNATIVE PROMOTERS. Journal of Biological Chemistry 272, 4458- 4466.
Gunter, J.H., Thomas, E.C., Lengefeld, N., Kruger, S.J., Worton, L., Gardiner, E.M., Jones, A., Barnett, N.L., and Whitehead, J.P. (2008). Characterisation of inosine monophosphate dehydrogenase expression during retinal development: differences between variants and isoforms. Int J Biochem Cell Biol 40, 1716-1728.
Habrian, C., Chandrasekhara, A., Shahrvini, B., Hua, B., Lee, J., Jesinghaus, R., Barry, R., Gitai, Z., Kollman, J., and Baldwin, E.P. (2016). Inhibition of Escherichia coli CTP Synthetase by NADH and Other Nicotinamides and Their Mutual Interactions with CTP and GTP. Biochemistry 55, 5554-5565.
Hacker, S.M., Buntz, A., Zumbusch, A., and Marx, A. (2015). Direct Monitoring of Nucleotide Turnover in Human Cell Extracts and Cells by Fluorogenic ATP Analogs. ACS Chem Biol 10, 2544-2552.
Haggard, M.E., and Lockhart, L.H. (1967). Megaloblastic anemia and orotic aciduria. A hereditary disorder of pyrimidine metabolism responsive to uridine. Am J Dis Child 113, 733-740.
184
Han, G.S., Sreenivas, A., Choi, M.G., Chang, Y.F., Martin, S.S., Baldwin, E.P., and Carman, G.M. (2005). Expression of Human CTP synthetase in Saccharomyces cerevisiae reveals phosphorylation by protein kinase A. The Journal of biological chemistry 280, 38328-38336.
Hartwig, C., Bahre, H., Wolter, S., Beckert, U., Kaever, V., and Seifert, R. (2014). cAMP, cGMP, cCMP and cUMP concentrations across the tree of life: High cCMP and cUMP levels in astrocytes. Neurosci Lett 579, 183-187.
Hatse, S., De Clercq, E., and Balzarini, J. (1999). Role of antimetabolites of purine and pyrimidine nucleotide metabolism in tumor cell differentiation. Biochem Pharmacol 58, 539-555.
Hedegaard, J., and Roche, J. (1966). [On the antagonism between the inhibitors of the biosynthesis of purines: azaserine and cycloserine, and the natural formiminated derivatives during the in vivo growth of cells of the Ehrlich ascites tumor]. C R Seances Soc Biol Fil 160, 539-544.
Hedstrom, L. (2009). IMP dehydrogenase: structure, mechanism, and inhibition. Chem Rev 109, 2903-2928.
Hendrickson, W.A., Pahler, A., Smith, J.L., Satow, Y., Merritt, E.A., and Phizackerley, R.P. (1989). Crystal structure of core streptavidin determined from multiwavelength anomalous diffraction of synchrotron radiation. Proc Natl Acad Sci U S A 86, 2190-2194.
Herce, H.D., Deng, W., Helma, J., Leonhardt, H., and Cardoso, M.C. (2013). Visualization and targeted disruption of protein interactions in living cells. Nature communications 4, 2660.
Herve, G. (1989). Allosteric Enzymes, 1 edn (Boca Raton, FL: CRC Press).
Hesketh, G.G., Youn, J.Y., Samavarchi-Tehrani, P., Raught, B., and Gingras, A.C. (2017). Parallel Exploration of Interaction Space by BioID and Affinity Purification Coupled to Mass Spectrometry. Methods Mol Biol 1550, 115-136.
Heyde, E., and Morrison, J.F. (1976). Studies on inosine monophosphate dehydrogenase. An associating-dissociating system. Biochim Biophys Acta 429, 635-644. 185
Higgins, M.J., Graves, P.R., and Graves, L.M. (2007). Regulation of human cytidine triphosphate synthetase 1 by glycogen synthase kinase 3. The Journal of biological chemistry 282, 29493-29503.
Holmes, E.W., Pehlke, D.M., and Kelley, W.N. (1974). Human IMP dehydrogenase. Kinetics and regulatory properties. Biochim Biophys Acta 364, 209-217.
Hordern, J., and Henderson, J.F. (1982). Comparison of purine and pyrimidine metabolism in G1 and S phases of HeLa and Chinese hamster ovary cells. Can J Biochem 60, 422-433.
Hsu, C.H., Liou, J.Y., Dutschman, G.E., and Cheng, Y.C. (2005). Phosphorylation of Cytidine, Deoxycytidine, and Their Analog Monophosphates by Human UMP/CMP Kinase Is Differentially Regulated by ATP and Magnesium. Mol Pharmacol 67, 806-814.
Hu, Y.Y., Zheng, M.H., Zhang, R., Liang, Y.M., and Han, H. (2012). Notch signaling pathway and cancer metastasis. Advances in experimental medicine and biology 727, 186-198.
Huang, F., Zhang, Q., Ma, H., Lv, Q., and Zhang, T. (2014). Expression of glutaminase is upregulated in colorectal cancer and of clinical significance. Int J Clin Exp Pathol 7, 1093-1100.
Huang, M., Wang, Y., Collins, M., and Graves, L.M. (2004). CPEC induces erythroid differentiation of human myeloid leukemia K562 cells through CTP depletion and p38 MAP kinase. Leukemia 18, 1857-1863.
Ignoul, S., and Eggermont, J. (2005). CBS domains: structure, function, and pathology in human proteins. Am J Physiol Cell Physiol 289, C1369-1378.
Ingerson-Mahar, M., Briegel, A., Werner, J.N., Jensen, G.J., and Gitai, Z. (2010). The metabolic enzyme CTP synthase forms cytoskeletal filaments. Nat Cell Biol 12, 739-746.
186
Ingley, E., and Hemmings, B.A. (2000). PKB/Akt interacts with inosine-5' monophosphate dehydrogenase through its pleckstrin homology domain. FEBS Lett 478, 253-259.
Ishikawa, H. (1999). Mizoribine and mycophenolate mofetil. Curr Med Chem 6, 575-597.
Jarroll, E.L., Manning, P., Berrada, A., Hare, D., and Lindmark, D.G. (1989). Biochemistry and metabolism of Giardia. J Protozool 36, 190-197.
Ji, Y., Gu, J., Makhov, A.M., Griffith, J.D., and Mitchell, B.S. (2006). Regulation of the interaction of inosine monophosphate dehydrogenase with mycophenolic Acid by GTP. The Journal of biological chemistry 281, 206-212.
Jia, D., Xu, Q., Xie, Q., Mio, W., and Deng, W.M. (2016). Automatic stage identification of Drosophila egg chamber based on DAPI images. Sci Rep 6, 18850.
Juda, P., Smigova, J., Kovacik, L., Bartova, E., and Raska, I. (2014). Ultrastructure of cytoplasmic and nuclear inosine-5'-monophosphate dehydrogenase 2 "rods and rings" inclusions. J Histochem Cytochem 62, 739- 750.
Kalko, S.G., Paco, S., Jou, C., Rodriguez, M.A., Meznaric, M., Rogac, M., Jekovec-Vrhovsek, M., Sciacco, M., Moggio, M., Fagiolari, G., et al. (2014). Transcriptomic profiling of TK2 deficient human skeletal muscle suggests a role for the p53 signalling pathway and identifies growth and differentiation factor-15 as a potential novel biomarker for mitochondrial myopathies. BMC Genomics 15, 91.
Kang, G.J., Cooney, D.A., Moyer, J.D., Kelley, J.A., Kim, H.Y., Marquez, V.E., and Johns, D.G. (1989). Cyclopentenylcytosine triphosphate. Formation and inhibition of CTP synthetase. The Journal of biological chemistry 264, 713- 718.
Kassel, K.M., Au da, R., Higgins, M.J., Hines, M., and Graves, L.M. (2010). Regulation of human cytidine triphosphate synthetase 2 by phosphorylation. The Journal of biological chemistry 285, 33727-33736.
187
Kemp, B.E. (2004). Bateman domains and adenosine derivatives form a binding contract. The Journal of clinical investigation 113, 182-184.
Keppeke, G.D., Calise, S.J., Chan, E.K., and Andrade, L.E. (2015). Assembly of IMPDH2-based, CTPS-based, and mixed rod/ring structures is dependent on cell type and conditions of induction. J Genet Genomics 42, 287-299.
Keppeke, G.D., Prado, M.S., Nunes, E., Perazzio, S.F., Rodrigues, S.H., Ferraz, M.L., Chan, E.K., and Andrade, L.E. (2016). Differential capacity of therapeutic drugs to induce Rods/Rings structures in vitro and in vivo and generation of anti-Rods/Rings autoantibodies. Clin Immunol 173, 149-156.
Kiguchi, K., Collart, F.R., Henning-Chubb, C., and Huberman, E. (1990). Induction of cell differentiation in melanoma cells by inhibitors of IMP dehydrogenase: altered patterns of IMP dehydrogenase expression and activity. Cell Growth Differ 1, 259-270.
Kim, D.I., Birendra, K.C., Zhu, W., Motamedchaboki, K., Doye, V., and Roux, K.J. (2014). Probing nuclear pore complex architecture with proximity- dependent biotinylation. Proc Natl Acad Sci U S A 111, E2453-2461.
Kim, D.I., Jensen, S.C., and Roux, K.J. (2016). Identifying Protein-Protein Associations at the Nuclear Envelope with BioID. Methods Mol Biol 1411, 133-146.
Kinoshita, F., Kondoh, T., Komori, K., Matsui, T., Harada, N., Yanai, A., Fukuda, M., Morifuji, K., and Matsumoto, T. (2011). Miller syndrome with novel dihydroorotate dehydrogenase gene mutations. Pediatr Int 53, 587-591.
Kisner, D.L., Catane, R., and Muggia, F.M. (1980). The rediscovery of DON (6-diazo-5-oxo-L-norleucine). Recent Results Cancer Res 74, 258-263.
Kizaki, H., Williams, J.C., Morris, H.P., and Weber, G. (1980). Increased cytidine 5'-triphosphate synthetase activity in rat and human tumors. Cancer research 40, 3921-3927.
Knowles, J.R. (1989). The mechanism of biotin-dependent enzymes. Annu Rev Biochem 58, 195-221.
188
Konno, Y., Natsumeda, Y., Nagai, M., Yamaji, Y., Ohno, S., Suzuki, K., and Weber, G. (1991). Expression of human IMP dehydrogenase types I and II in Escherichia coli and distribution in human normal lymphocytes and leukemic cell lines. Journal of Biological Chemistry 266, 506-509.
Kozhevnikova, E.N., van der Knaap, J.A., Pindyurin, A.V., Ozgur, Z., van Ijcken, W.F., Moshkin, Y.M., and Verrijzer, C.P. (2012). Metabolic enzyme IMPDH is also a transcription factor regulated by cellular state. Mol Cell 47, 133-139.
Krawutschke, C., Koesling, D., and Russwurm, M. (2015). Cyclic GMP in Vascular Relaxation: Export Is of Similar Importance as Degradation. Arterioscler Thromb Vasc Biol 35, 2011-2019.
Kucuk, Z.Y., Zhang, K., Filipovich, L., and Bleesing, J.J. (2016). CTP Synthase 1 Deficiency in Successfully Transplanted Siblings with Combined Immune Deficiency and Chronic Active EBV Infection. J Clin Immunol 36, 750-753.
Kursula, P., Flodin, S., Ehn, M., Hammarstrom, M., Schuler, H., Nordlund, P., and Stenmark, P. (2006). Structure of the synthetase domain of human CTP synthetase, a target for anticancer therapy. Acta Crystallogr Sect F Struct Biol Cryst Commun 62, 613-617.
Labesse, G., Alexandre, T., Vaupre, L., Salard-Arnaud, I., Him, J.L., Raynal, B., Bron, P., and Munier-Lehmann, H. (2013). MgATP regulates allostery and fiber formation in IMPDHs. Structure 21, 975-985.
Labuschagne, C.F., van den Broek, N.J., Mackay, G.M., Vousden, K.H., and Maddocks, O.D. (2014). Serine, but not glycine, supports one-carbon metabolism and proliferation of cancer cells. Cell Rep 7, 1248-1258.
Lander, G.C., Stagg, S.M., Voss, N.R., Cheng, A., Fellmann, D., Pulokas, J., Yoshioka, C., Irving, C., Mulder, A., Lau, P.W., Lyumkis, D., Potter, C.S., and Carragher, B. (2009). Appion: an integrated, database-driven pipeline to facilitate EM image processing. J Struct Biol 166, 95-102.
Lane, A.N., and Fan, T.W. (2015). Regulation of mammalian nucleotide metabolism and biosynthesis. Nucleic Acids Res 43, 2466-2485. 189
Laourdakis, C.D., Merino, E.F., Neilson, A.P., and Cassera, M.B. (2014). Comprehensive quantitative analysis of purines and pyrimidines in the human malaria parasite using ion-pairing ultra-performance liquid chromatography-mass spectrometry. Journal of chromatography B, Analytical technologies in the biomedical and life sciences 967, 127-133.
Lascu, I., and Gonin, P. (2000). The Catalytic Mechanism of Nucleoside Diphosphate Kinases. Journal of Bioenergetics and Biomembranes 32, 237- 246.
Lee, H.O., Davidson, J.M., and Duronio, R.J. (2009). Endoreplication: polyploidy with purpose. Genes & development 23, 2461-2477.
Lee, S.Y., Jeon, H.M., Ju, M.K., Jeong, E.K., Kim, C.H., Park, H.G., Han, S.I., and Kang, H.S. (2016). Dlx-2 and glutaminase upregulate epithelial- mesenchymal transition and glycolytic switch. Oncotarget 7, 7925-7939.
Lehninger, A., Nelson, D., Cox, M. (1993). Principles of Biochemistry, Second edn (New York: Worth Publishers).
Levenberg, B., Melnick, I., and Buchanan, J.M. (1957). Biosynthesis of the purines. XV. The effect of aza-L-serine and 6-diazo-5-oxo-L-norleucine on inosinic acid biosynthesis de novo. The Journal of biological chemistry 225, 163-176.
Levine, R.L., Hoogenraad, N.J., and Kretchmer, N. (1974). A review: biological and clinical aspects of pyrimidine metabolism. Pediatr Res 8, 724- 734.
Li, J., Huang, S., Chen, J., Yang, Z., Fei, X., Zheng, M., Ji, C., Xie, Y., and Mao, Y. (2007). Identification and characterization of human uracil phosphoribosyltransferase (UPRTase). J Hum Genet 52, 415-422.
Liao, W.S., Heller, R., Green, P., and Stark, G.R. (1986). Regulation of carbamoyl phosphate synthetase-aspartate transcarbamoylase- dihydroorotase gene expression in growing and arrested cells. The Journal of biological chemistry 261, 15577-15581. 190
Liu, J.L. (2010). Intracellular compartmentation of CTP synthase in Drosophila. J Genet Genomics 37, 281-296.
Liu, J.L. (2016). The Cytoophidium and Its Kind: Filamentation and Compartmentation of Metabolic Enzymes. Annual review of cell and developmental biology 32, 349-372.
Liu, X., Guy, H.I., and Evans, D.R. (1994). Identification of the regulatory domain of the mammalian multifunctional protein CAD by the construction of an Escherichia coli hamster hybrid carbamyl-phosphate synthetase. The Journal of biological chemistry 269, 27747-27755.
Livingston, R.B., Venditti, J.M., Cooney, D.A., and Carter, S.K. (1970). Glutamine antagonists in chemotherapy. Adv Pharmacol Chemother 8, 57- 120.
Lodish, H., Berk, A., Sipursky, S.L. (2000). Molecular Cell Biology, 4th edition edn (New York: W. H. Freeman).
Lora, J., Alonso, F.J., Segura, J.A., Lobo, C., Marquez, J., and Mates, J.M. (2004). Antisense glutaminase inhibition decreases glutathione antioxidant capacity and increases apoptosis in Ehrlich ascitic tumour cells. Eur J Biochem 271, 4298-4306.
Lunn, F.A., MacDonnell, J.E., and Bearne, S.L. (2008). Structural requirements for the activation of Escherichia coli CTP synthase by the allosteric effector GTP are stringent, but requirements for inhibition are lax. The Journal of biological chemistry 283, 2010-2020.
Majd, N., et. al (2014). A Review of the Potential Utitlity of Mycophenolate Mofetil as a Cancer Therapeutic. Journal of Cancer Research 2014.
Majd, N., Sumita, K., Yoshino, H., Chen, D., Terakawa, J., Daikoku, T., Kofuji, S., Curry, R., Wise-Draper, T.M., Warnick, R.E., et al. (2014). A Review of the Potential Utility of Mycophenolate Mofetil as a Cancer Therapeutic. Journal of Cancer Research 2014, 12.
191
Makowska-Grzyska, M., Kim, Y., Wu, R., Wilton, R., Gollapalli, D.R., Wang, X.K., Zhang, R., Jedrzejczak, R., Mack, J.C., Maltseva, N., et al. (2012). Bacillus anthracis inosine 5'-monophosphate dehydrogenase in action: the first bacterial series of structures of phosphate ion-, substrate-, and product- bound complexes. Biochemistry 51, 6148-6163.
Mandel, H., Szargel, R., Labay, V., Elpeleg, O., Saada, A., Shalata, A., Anbinder, Y., Berkowitz, D., Hartman, C., Barak, M., et al. (2001). The deoxyguanosine kinase gene is mutated in individuals with depleted hepatocerebral mitochondrial DNA. Nat Genet 29, 337-341.
Mandel, P., Wintzerith, M., Klein-Pete, N., and Mandel, L. (1963). Comparative Investigation of the Free Nucleotides of an Ascitic Hepatoma and of Normal or Regenerating Liver. Nature 198, 1000-1001.
Marquez, V.E., Tseng, C.K., Gebeyehu, G., Cooney, D.A., Ahluwalia, G.S., Kelley, J.A., Dalal, M., Fuller, R.W., Wilson, Y.A., and Johns, D.G. (1986). Thiazole-4-carboxamide adenine dinucleotide (TAD). Analogues stable to phosphodiesterase hydrolysis. J Med Chem 29, 1726-1731.
Martin, E., Palmic, N., Sanquer, S., Lenoir, C., Hauck, F., Mongellaz, C., Fabrega, S., Nitschke, P., Esposti, M.D., Schwartzentruber, J., et al. (2014). CTP synthase 1 deficiency in humans reveals its central role in lymphocyte proliferation. Nature 510, 288-292.
Marttila, A.T., Laitinen, O.H., Airenne, K.J., Kulik, T., Bayer, E.A., Wilchek, M., and Kulomaa, M.S. (2000). Recombinant NeutraLite avidin: a non- glycosylated, acidic mutant of chicken avidin that exhibits high affinity for biotin and low non-specific binding properties. FEBS Lett 467, 31-36.
Masterson, J.G., and Nelson, J.H., Jr. (1965). The role of chemotherapy in the treatment of gynecologic malignancy. Am J Obstet Gynecol 93, 1102-1111.
McCluskey, G.a.B., S. (2015). Gemcitabine is a Competitive Inhibitor of Cytidine-5'-triphosphate Synthase (CTPS) that Induces Enzyme Filament Polymerization. The FASEB Journal 29.
McCluskey, G.D., Mohamady, S., Taylor, S.D., and Bearne, S.L. (2016). Exploring the Potent Inhibition of CTP Synthase by Gemcitabine-5'- Triphosphate. Chembiochem 17, 2240-2249. 192
McCormick, J.B., King, I.J., Webb, P.A., Scribner, C.L., Craven, R.B., Johnson, K.M., Elliott, L.H., and Belmont-Williams, R. (1986). Lassa fever. Effective therapy with ribavirin. N Engl J Med 314, 20-26.
McKeehan, W.L. (1982). Glycolysis, glutaminolysis and cell proliferation. Cell Biol Int Rep 6, 635-650.
McKeran, R.O., and Watts, R.W. (1976). Use of phytohaemagglutinin stimulated lymphocytes to study effects of hypoxanthine-guanine phosphoribosyltransferase (HGPRT) deficiency on polynucleotide and protein synthesis in the Lesch-Nyhan syndrome. J Med Genet 13, 91-95.
McLean, J.E., Hamaguchi, N., Belenky, P., Mortimer, S.E., Stanton, M., and Hedstrom, L. (2004). Inosine 5'-monophosphate dehydrogenase binds nucleic acids in vitro and in vivo. The Biochemical journal 379, 243-251.
Medina, M.A., Sanchez-Jimenez, F., Marquez, J., Rodriguez Quesada, A., and Nunez de Castro, I. (1992). Relevance of glutamine metabolism to tumor cell growth. Mol Cell Biochem 113, 1-15.
Mejia, E.J., Magranet, L.B., De Voogd, N.J., Tendyke, K., Qiu, D., Shen, Y.Y., Zhou, Z., and Crews, P. (2013). Structures and cytotoxic evaluation of new and known acyclic Ene-Ynes from an American Samoa Petrosia sp. Sponge. J Nat Prod 76, 425-432.
Mermod, J.J., Jacobs-Lorena, M., and Crippa, M. (1977). Changes in rate of RNA synthesis and ribosomal gene number during oogenesis of Drosophila melanogaster. Dev Biol 57, 393-402.
Merriam-Webster (2017). (Springfield, MA: Merriam-Webster Incorporated).
Messina, E., Gazzaniga, P., Micheli, V., Barile, L., Lupi, F., Agliano, A.M., and Giacomello, A. (2004). Low levels of mycophenolic acid induce differentiation of human neuroblastoma cell lines. Int J Cancer 112, 352-354.
Mihaylova, M.M., and Shaw, R.J. (2011). The AMPK signalling pathway coordinates cell growth, autophagy and metabolism. Nat Cell Biol 13, 1016- 1023. 193
Mini, E., Nobili, S., Caciagli, B., Landini, I., and Mazzei, T. (2006). Cellular pharmacology of gemcitabine. Ann Oncol 17 Suppl 5, v7-12.
Mizushina, Y., Dairaku, I., Yanaka, N., Takeuchi, T., Ishimaru, C., Sugawara, F., Yoshida, H., and Kato, N. (2007). Inhibitory action of polyunsaturated fatty acids on IMP dehydrogenase. Biochimie 89, 581-590.
Moffatt, B.A., and Ashihara, H. (2002). Purine and pyrimidine nucleotide synthesis and metabolism. Arabidopsis Book 1, e0018.
Moore, E.C., and Hurlbert, R.B. (1961). Biosynthesis of RNA cytosine and RNA purines: differential inhibition by diazo-oxonorleucine. Cancer research 21, 257-261.
Moore, E.C., and Lepage, G.A. (1957). In vivo sensitivity of normal and neoplastic mouse tissues to azaserine. Cancer research 17, 804-808.
Morrison, H.G., McArthur, A.G., Gillin, F.D., Aley, S.B., Adam, R.D., Olsen, G.J., Best, A.A., Cande, W.Z., Chen, F., Cipriano, M.J., et al. (2007). Genomic minimalism in the early diverging intestinal parasite Giardia lamblia. Science 317, 1921-1926.
Mortimer, S.E., and Hedstrom, L. (2005). Autosomal dominant retinitis pigmentosa mutations in inosine 5'-monophosphate dehydrogenase type I disrupt nucleic acid binding. The Biochemical journal 390, 41-47.
Mortimer, S.E., Xu, D., McGrew, D., Hamaguchi, N., Lim, H.C., Bowne, S.J., Daiger, S.P., and Hedstrom, L. (2008). IMP Dehydrogenase Type 1 Associates with Polyribosomes Translating Rhodopsin mRNA. Journal of Biological Chemistry 283, 36354-36360.
Nadkarni, A.K., McDonough, V.M., Yang, W.L., Stukey, J.E., Ozier- Kalogeropoulos, O., and Carman, G.M. (1995). Differential biochemical regulation of the URA7- and URA8-encoded CTP synthetases from Saccharomyces cerevisiae. The Journal of biological chemistry 270, 24982- 24988.
194
Nagai, M., Natsumeda, Y., Konno, Y., Hoffman, R., Irino, S., and Weber, G. (1991). Selective Up-Regulation of Type II Inosine 5′-Monophosphate Dehydrogenase Messenger RNA Expression in Human Leukemias. Cancer research 51, 3886-3890.
Nakamura, M., Bhatnagar, A., and Sadoshima, J. (2012). Overview of pyridine nucleotides review series. Circ Res 111, 604-610.
Narayanaswamy, R., Levy, M., Tsechansky, M., Stovall, G.M., O'Connell, J.D., Mirrielees, J., Ellington, A.D., and Marcotte, E.M. (2009). Widespread reorganization of metabolic enzymes into reversible assemblies upon nutrient starvation. Proceedings of the National Academy of Sciences 106, 10147- 10152.
Natsumeda, Y., Ohno, S., Kawasaki, H., Konno, Y., Weber, G., and Suzuki, K. (1990). Two distinct cDNAs for human IMP dehydrogenase. Journal of Biological Chemistry 265, 5292-5295.
Natsumeda, Y., Prajda, N., Donohue, J.P., Glover, J.L., and Weber, G. (1984). Enzymic capacities of purine de novo and salvage pathways for nucleotide synthesis in normal and neoplastic tissues. Cancer research 44, 2475-2479.
Newsholme, E.A., Crabtree, B., and Ardawi, M.S.M. (1985). The role of high rates of glycolysis and glutamine utilization in rapidly dividing cells. Bioscience Reports 5, 393-400.
Newton, A.C., Bootman, M.D., Scott, J.D. (2016). Second Messengers. Cold Spring Harbor Perspectives in Biology August 2016.
Ng, B.G., Wolfe, L.A., Ichikawa, M., Markello, T., He, M., Tifft, C.J., Gahl, W.A., and Freeze, H.H. (2015). Biallelic mutations in CAD, impair de novo pyrimidine biosynthesis and decrease glycosylation precursors. Hum Mol Genet 24, 3050-3057.
Nguyen, L.X.T., Lee, Y., Urbani, L., Utz, P.J., Hamburger, A.W., Sunwoo, J.B., and Mitchell, B.S. (2015). Regulation of ribosomal RNA synthesis in T cells: requirement for GTP and Ebp1. Blood 125, 2519-2529.
195
Nielsen, H.R., Sjolin, K.E., Nyholm, K., Baliga, B.S., Wong, R., and Borek, E. (1974). Beta-aminoisobutyric acid, a new probe for the metabolism of DNA and RNA in normal and tumorous tissue. Cancer research 34, 1381-1384.
Noree, C., Sato, B.K., Broyer, R.M., and Wilhelm, J.E. (2010). Identification of novel filament-forming proteins in Saccharomyces cerevisiae and Drosophila melanogaster. J Cell Biol 190, 541-551.
O'Connell, J.D., Zhao, A., Ellington, A.D., and Marcotte, E.M. (2012). Dynamic reorganization of metabolic enzymes into intracellular bodies. Annual review of cell and developmental biology 28, 89-111.
Ortlund, E., Lacount, M.W., Lewinski, K., and Lebioda, L. (2000). Reactions of Pseudomonas 7A glutaminase-asparaginase with diazo analogues of glutamine and asparagine result in unexpected covalent inhibitions and suggests an unusual catalytic triad Thr-Tyr-Glu. Biochemistry 39, 1199-1204.
Osley, M.A., and Newton, A. (1978). Regulation of cell cycle events in asymmetrically dividing cells: functions required for DNA initiation and chain elongation in Caulobacter crescentus. J Bacteriol 135, 10-17.
Ozier-Kalogeropoulos, O., Adeline, M.T., Yang, W.L., Carman, G.M., and Lacroute, F. (1994). Use of synthetic lethal mutants to clone and characterize a novel CTP synthetase gene in Saccharomyces cerevisiae. Mol Gen Genet 242, 431-439.
Pai, L.M., Wang, P.Y., Lin, W.C., Chakraborty, A., Yeh, C.T., and Lin, Y.H. (2016). Ubiquitination and filamentous structure of cytidine triphosphate synthase. Fly (Austin) 10, 108-114.
Papinazath, T., Min, W., Sujiththa, S., Cohen, A., Ackerley, C., Roifman, C.M., and Grunebaum, E. (2011). Effects of purine nucleoside phosphorylase deficiency on thymocyte development. Journal of Allergy and Clinical Immunology 128, 854-863.e851.
Park, T.S., O'Brien, D.J., and Carman, G.M. (2003). Phosphorylation of CTP synthetase on Ser36, Ser330, Ser354, and Ser454 regulates the levels of CTP and phosphatidylcholine synthesis in Saccharomyces cerevisiae. The Journal of biological chemistry 278, 20785-20794. 196
Park, T.S., Ostrander, D.B., Pappas, A., and Carman, G.M. (1999). Identification of Ser424 as the protein kinase A phosphorylation site in CTP synthetase from Saccharomyces cerevisiae. Biochemistry 38, 8839-8848.
Parker, W.B. (2005). Metabolism and antiviral activity of ribavirin. Virus Res 107, 165-171.
Petrovska, I., Nuske, E., Munder, M.C., Kulasegaran, G., Malinovska, L., Kroschwald, S., Richter, D., Fahmy, K., Gibson, K., Verbavatz, J.M., et al. (2014). Filament formation by metabolic enzymes is a specific adaptation to an advanced state of cellular starvation. eLife.
Pimkin, M., and Markham, G.D. (2008). The CBS subdomain of inosine 5'- monophosphate dehydrogenase regulates purine nucleotide turnover. Mol Microbiol 68, 342-359.
Pimkin, M., Pimkina, J., and Markham, G.D. (2009). A regulatory role of the Bateman domain of IMP dehydrogenase in adenylate nucleotide biosynthesis. The Journal of biological chemistry 284, 7960-7969.
Pinkus, L.M. (1977). Glutamine binding sites. Methods Enzymol 46, 414-427.
Pusch, M. (2002). Myotonia caused by mutations in the muscle chloride channel gene CLCN1. Hum Mutat 19, 423-434.
Qie, S., Chu, C., Li, W., Wang, C., and Sang, N. (2014). ErbB2 activation upregulates glutaminase 1 expression which promotes breast cancer cell proliferation. J Cell Biochem 115, 498-509.
Quemeneur, L., Gerland, L.M., Flacher, M., Ffrench, M., Revillard, J.P., and Genestier, L. (2003). Differential control of cell cycle, proliferation, and survival of primary T lymphocytes by purine and pyrimidine nucleotides. J Immunol 170, 4986-4995.
Raghavan, A., Ogilvie, R.L., Reilly, C., Abelson, M.L., Raghavan, S., Vasdewani, J., Krathwohl, M., and Bohjanen, P.R. (2002). Genome-wide analysis of mRNA decay in resting and activated primary human T lymphocytes. Nucleic Acids Res 30, 5529-5538. 197
Rahman, S., and Poulton, J. (2009). Diagnosis of mitochondrial DNA depletion syndromes. Arch Dis Child 94, 3-5.
Rais, R., Jancarik, A., Tenora, L., Nedelcovych, M., Alt, J., Englert, J., Rojas, C., Le, A., Elgogary, A., Tan, J., et al. (2016). Discovery of 6-Diazo-5-oxo-l- norleucine (DON) Prodrugs with Enhanced CSF Delivery in Monkeys: A Potential Treatment for Glioblastoma. J Med Chem 59, 8621-8633.
Rajagopalan, P.T., Zhang, Z., McCourt, L., Dwyer, M., Benkovic, S.J., and Hammes, G.G. (2002). Interaction of dihydrofolate reductase with methotrexate: ensemble and single-molecule kinetics. Proc Natl Acad Sci U S A 99, 13481-13486.
Rampazzo, C., Miazzi, C., Franzolin, E., Pontarin, G., Ferraro, P., Frangini, M., Reichard, P., and Bianchi, V. (2010). Regulation by degradation, a cellular defense against deoxyribonucleotide pool imbalances. Mutat Res 703, 2-10.
Riera, T.V., Zheng, L., Josephine, H.R., Min, D., Yang, W., and Hedstrom, L. (2011). Allosteric activation via kinetic control: potassium accelerates a conformational change in IMP dehydrogenase. Biochemistry 50, 8508-8518.
Roach, P.J. (1991). Multisite and hierarchal protein phosphorylation. The Journal of biological chemistry 266, 14139-14142.
Rodriguez-Suarez, R., Xu, D., Veillette, K., Davison, J., Sillaots, S., Kauffman, S., Hu, W., Bowman, J., Martel, N., Trosok, S., et al. (2007). Mechanism-of-action determination of GMP synthase inhibitors and target validation in Candida albicans and Aspergillus fumigatus. Chem Biol 14, 1163-1175.
Roje, S. (2006). S-Adenosyl-L-methionine: beyond the universal methyl group donor. Phytochemistry 67, 1686-1698.
Roseman, S. (2001). Reflections on Glycobiology. Journal of Biological Chemistry 276, 41527-41542.
198
Rosenbluth, R.J., Cooney, D.A., Jayaram, H.N., Milman, H.A., and Homan, E.R. (1976). DON, CONV and DONV-II. Inhibition of L-'asparagine synthetase in vivo. Biochem Pharmacol 25, 1851-1858.
Roux, K.J., Kim, D.I., and Burke, B. (2013). BioID: a screen for protein- protein interactions. Curr Protoc Protein Sci 74, Unit 19.23.
Roux, K.J., Kim, D.I., Raida, M., and Burke, B. (2012). A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. J Cell Biol 196, 801-810.
Samols, D., Thornton, C.G., Murtif, V.L., Kumar, G.K., Haase, F.C., and Wood, H.G. (1988). Evolutionary conservation among biotin enzymes. The Journal of biological chemistry 263, 6461-6464.
Sarri, E., Sicart, A., Lazaro-Dieguez, F., and Egea, G. (2011). Phospholipid synthesis participates in the regulation of diacylglycerol required for membrane trafficking at the Golgi complex. The Journal of biological chemistry 286, 28632-28643.
Saunders, P.P., Spindler, C.D., Tan, M.T., Alvarez, E., and Robins, R.K. (1990). Tiazofurin is phosphorylated by three enzymes from Chinese hamster ovary cells. Cancer research 50, 5269-5274.
Scheres, S.H.W. (2012). RELION: Implementation of a Bayesian approach to cryo-EM structure determination. Journal of Structural Biology 180, 519-530
Schoenberg, D.R., and Maquat, L.E. (2012). Regulation of cytoplasmic mRNA decay. Nat Rev Genet 13, 246-259.
Schwede, T.F., Retey, J., and Schulz, G.E. (1999). Crystal structure of histidine ammonia-lyase revealing a novel polypeptide modification as the catalytic electrophile. Biochemistry 38, 5355-5361.
Scott, J.W., Hawley, S.A., Green, K.A., Anis, M., Stewart, G., Scullion, G.A., Norman, D.G., and Hardie, D.G. (2004). CBS domains form energy-sensing modules whose binding of adenosine ligands is disrupted by disease mutations. J Clin Invest 113, 274-284. 199
Seifert, R. (2015). cCMP and cUMP: emerging second messengers. Trends Biochem Sci 40, 8-15.
Shan, X., Dunbrack Jr, R.L., Christopher, S.A., and Kruger, W.D. (2001). Mutations in the regulatory domain of cystathionine β–synthase can functionally suppress patient-derived mutations in cis. Human Molecular Genetics 10, 635-643.
Shen, Q.J., Kassim, H., Huang, Y., Li, H., Zhang, J., Li, G., Wang, P.Y., Yan, J., Ye, F., and Liu, J.L. (2016). Filamentation of Metabolic Enzymes in Saccharomyces cerevisiae. J Genet Genomics 43, 393-404.
Sherley, J.L. (1991). Guanine nucleotide biosynthesis is regulated by the cellular p53 concentration. The Journal of biological chemistry 266, 24815- 24828.
Simard, R., and Bernhard, W. (1966). [The phenomenon of nucleolar segregation: specificity of action of certain antimetabolites]. Int J Cancer 1, 463-479.
Sintchak, M.D., Fleming, M.A., Futer, O., Raybuck, S.A., Chambers, S.P., Caron, P.R., Murcko, M.A., and Wilson, K.P. (1996). Structure and mechanism of inosine monophosphate dehydrogenase in complex with the immunosuppressant mycophenolic acid. Cell 85, 921-930.
Skerker, J.M., and Laub, M.T. (2004). Cell-cycle progression and the generation of asymmetry in Caulobacter crescentus. Nat Rev Micro 2, 325- 337.
Slivac, I., Blajić, V., Radošević, K. et al. (2010). Influence of different ammonia, lactate and glutamine concentrations on CCO cell growth. Cytotechnology 62: 585.
Smith RA, K.W. (1980). In: Developments in Antiviral Therapy. ( London: Academic Press).
Smyth, J.D. (1995). Introduction to Animal Parasitology, 3 edn (Cambridge University Press). 200
Souba, W.W. (1993). Glutamine and cancer. Ann Surg 218, 715-728.
Spellicy, C.J., Daiger, S.P., Sullivan, L.S., Zhu, J., Liu, Q., Pierce, E.A., and Bowne, S.J. (2007). Characterization of retinal inosine monophosphate dehydrogenase 1 in several mammalian species. Mol Vis 13, 1866-1872.
Spinazzola, A., Marti, R., Nishino, I., Andreu, A.L., Naini, A., Tadesse, S., Pela, I., Zammarchi, E., Donati, M.A., Oliver, J.A., et al. (2002). Altered Thymidine Metabolism Due to Defects of Thymidine Phosphorylase. Journal of Biological Chemistry 277, 4128-4133.
Steiger, S., and Harper, J.L. (2014). Mechanisms of spontaneous resolution of acute gouty inflammation. Curr Rheumatol Rep 16, 392.
Stinton, L.M., Myers, R.P., Coffin, C.S., and Fritzler, M.J. (2013). Clinical associations and potential novel antigenic targets of autoantibodies directed against rods and rings in chronic hepatitis C infection. BMC Gastroenterol 13, 50.
Strochlic, T.I., Stavrides, K.P., Thomas, S.V., Nicolas, E., O'Reilly, A.M., and Peterson, J.R. (2014). Ack kinase regulates CTP synthase filaments during Drosophila oogenesis. EMBO Rep 15, 1184-1191.
Suloway, C., Pulokas, J., Fellmann, D., Cheng, A., Guerra, F., Quispe, J., Stagg, S., Potter, C.S.,and Carragher, B. (2005). Automated molecular microscopy: the new Leginon system. J Struct Biol 151, 41-60.
Tarnowski, G.S., and Stock, C.C. (1957). Effects of Combinations of Azaserine and of 6-Diazo-5-oxo-l-norleucine with Purine Analogs and Other Antimetabolites on the Growth of Two Mouse Mammary Carcinomas. Cancer research 17, 1033-1039.
Thomas, E.C., Gunter, J.H., Webster, J.A., Schieber, N.L., Oorschot, V., Parton, R.G., and Whitehead, J.P. (2012). Different characteristics and nucleotide binding properties of inosine monophosphate dehydrogenase (IMPDH) isoforms. PLoS One 7, e51096.
201
Tong, X., Zhao, F., and Thompson, C.B. (2009). The molecular determinants of de novo nucleotide biosynthesis in cancer cells. Current opinion in genetics & development 19, 32-37.
Torres, R.J., and Puig, J.G. (2007). Hypoxanthine-guanine phosophoribosyltransferase (HPRT) deficiency: Lesch-Nyhan syndrome. Orphanet J Rare Dis 2, 48.
Traut, T.W. (1994). Physiological concentrations of purines and pyrimidines. Mol Cell Biochem 140, 1-22.
TSE-YU-WU (2011). Pharmacogentics and Pharmacogenomics of Mycophenolic Acid. In Molecular Pharmacology and Experimental Therapeutics (Mayo Graduate School, College of Medicine), pp. 231.
Tulpule, P.G. (1963). A comparative study of the synthesis of nicotinamide nucleotides by erythrocytes of some vertebrates. The Biochemical journal 87, 301-305.
Turker, M.S., and Martin, G.M. (1985). Induction of adenine salvage in mouse cell lines deficient in adenine phosphoribosyltransferase. Mol Cell Biol 5, 2662-2668.
Tyynismaa, H., Sun, R., Ahola-Erkkila, S., Almusa, H., Poyhonen, R., Korpela, M., Honkaniemi, J., Isohanni, P., Paetau, A., Wang, L., et al. (2012). Thymidine kinase 2 mutations in autosomal recessive progressive external ophthalmoplegia with multiple mitochondrial DNA deletions. Hum Mol Genet 21, 66-75.
Ulrickson, M.L., Puri, A., Lindstrom, S., Cassaday, R.D., De Padova, N., and Becker, P.S. (2017). Gemcitabine and docetaxel as a novel treatment regimen for blastic plasmacytoid dendritic cell neoplasm. Am J Hematol. van den Berg, A.A., van Lenthe, H., Busch, S., de Korte, D., Roos, D., van Kuilenburg, A.B., and van Gennip, A.H. (1993). Evidence for transformation- related increase in CTP synthetase activity in situ in human lymphoblastic leukemia. Eur J Biochem 216, 161-167.
202
Varnaite, R., and MacNeill, S.A. (2016). Meet the neighbors: Mapping local protein interactomes by proximity-dependent labeling with BioID. Proteomics 16, 2503-2518.
Verschuur, A.C., van Gennip, A.H., Muller, E.J., Voute, P.A., and van Kuilenburg, A.B. (1998). Increased activity of cytidine Triphosphate synthetase in pediatric acute lymphoblastic leukemia. Advances in experimental medicine and biology 431, 667-671.
Vieira, P.S., de Giuseppe, P.O., Murakami, M.T., and de Oliveira, A.H.C. (2015). Crystal structure and biophysical characterization of the nucleoside diphosphate kinase from Leishmania braziliensis. BMC Structural Biology 15, 2.
Wang, P.Y., Lin, W.C., Tsai, Y.C., Cheng, M.L., Lin, Y.H., Tseng, S.H., Chakraborty, A., and Pai, L.M. (2015). Regulation of CTP Synthase Filament Formation During DNA Endoreplication in Drosophila. Genetics 201, 1511- 1523.
Wang, Y., Liu, C.L., Storey, J.D., Tibshirani, R.J., Herschlag, D., and Brown, P.O. (2002). Precision and functional specificity in mRNA decay. Proc Natl Acad Sci U S A 99, 5860-5865.
Wang, Z., Da Silva, T.G., Jin, K., Han, X., Ranganathan, P., Zhu, X., Sanchez- Mejias, A., Bai, F., Li, B., Fei, D.L., et al. (2014). Notch signaling drives stemness and tumorigenicity of esophageal adenocarcinoma. Cancer research 74, 6364-6374.
Weber, G., Lui, M.S., Takeda, E., and Denton, J.E. (1980). Enzymology of human colon tumors. Life Sci 27, 793-799.
Westermarck, J., Ivaska, J., and Corthals, G.L. (2013). Identification of protein interactions involved in cellular signaling. Molecular & cellular proteomics : MCP 12, 1752-1763.
Willingham, M.C., Richert, N.D., and Rutherford, A.V. (1987). A novel fibrillar structure in cultured cells detected by a monoclonal antibody. Exp Cell Res 171, 284-295.
203
Willis, R.C., and Seegmiller, J.E. (1977). The inhibition by 6-diazo-5-oxo-l- norleucine of glutamine catabolism of the cultured human lymphoblast. J Cell Physiol 93, 375-382.
Wilson, M.Z., and Gitai, Z. (2013). Beyond the cytoskeleton: mesoscale assemblies and their function in spatial organization. Current opinion in microbiology 16, 177-183.
Wilusz, C.J., and Wilusz, J. (2004). Bringing the role of mRNA decay in the control of gene expression into focus. Trends Genet 20, 491-497.
Wilusz, C.J., Wormington, M., and Peltz, S.W. (2001). The cap-to-tail guide to mRNA turnover. Nat Rev Mol Cell Biol 2, 237-246.
Witkowski, J.T., Robins, R.K., Sidwell, R.W., and Simon, L.N. (1972). Design, synthesis, and broad spectrum antiviral activity of 1- -D-ribofuranosyl-1,2,4- triazole-3-carboxamide and related nucleosides. J Med Chem 15, 1150-1154.
Wojcik, M., Seidle, H.F., Bieganowski, P., and Brenner, C. (2006). Glutamine- dependent NAD+ synthetase. How a two-domain, three-substrate enzyme avoids waste. The Journal of biological chemistry 281, 33395-33402.
Wu, G., Haynes, T.E., Li, H., Yan, W., and Meininger, C.J. (2001). Glutamine metabolism to glucosamine is necessary for glutamine inhibition of endothelial nitric oxide synthesis. The Biochemical journal 353, 245-252.
Wu, M.C., Arimura, G.K., and Yunis, A.A. (1978). Mechanism of sensitivity of cultured pancreatic carcinoma to asparaginase. Int J Cancer 22, 728-733.
Wu, T.Y., Peng, Y., Pelleymounter, L.L., Moon, I., Eckloff, B.W., Wieben, E.D., Yee, V.C., and Weinshilboum, R.M. (2010). Pharmacogenetics of the mycophenolic acid targets inosine monophosphate dehydrogenases IMPDH1 and IMPDH2: gene sequence variation and functional genomics. British journal of pharmacology 161, 1584-1598.
Wyde, P.R. (1998). Respiratory syncytial virus (RSV) disease and prospects for its control. Antiviral Res 39, 63-79.
204
Xiang, B., Taylor, J.C., and Markham, G.D. (1996). Monovalent cation activation and kinetic mechanism of inosine 5'-monophosphate dehydrogenase. The Journal of biological chemistry 271, 1435-1440.
Xiao, B., Sanders, M.J., Carmena, D., Bright, N.J., Haire, L.F., Underwood, E., Patel, B.R., Heath, R.B., Walker, P.A., Hallen, S., et al. (2013). Structural basis of AMPK regulation by small molecule activators. Nat Commun 4, 3017.
Xiao, B., Sanders, M.J., Underwood, E., Heath, R., Mayer, F.V., Carmena, D., Jing, C., Walker, P.A., Eccleston, J.F., Haire, L.F., et al. (2011). Structure of mammalian AMPK and its regulation by ADP. Nature 472, 230-233.
Xie, C., Jin, J., Bao, X., Zhan, W.H., Han, T.Y., Gan, M., Zhang, C., and Wang, J. (2016). Inhibition of mitochondrial glutaminase activity reverses acquired erlotinib resistance in non-small cell lung cancer. Oncotarget 7, 610- 621.
Xu, D., Cobb, G., Spellicy, C.J., Bowne, S.J., Daiger, S.P., and Hedstrom, L. (2008). Retinal isoforms of inosine 5'-monophosphate dehydrogenase type 1 are poor nucleic acid binding proteins. Arch Biochem Biophys 472, 100-104.
Yang, E., van Nimwegen, E., Zavolan, M., Rajewsky, N., Schroeder, M., Magnasco, M., and Darnell, J.E., Jr. (2003). Decay rates of human mRNAs: correlation with functional characteristics and sequence attributes. Genome Res 13, 1863-1872.
Yang, M., and Vousden, K.H. (2016). Serine and one-carbon metabolism in cancer. Nat Rev Cancer 16, 650-662.
Yeates, T.O., Crowley, C.S., and Tanaka, S. (2010). Bacterial microcompartment organelles: protein shell structure and evolution. Annual review of biophysics 39, 185-205.
Yokota, S. (2002). Mizoribine: mode of action and effects in clinical use. Pediatr Int 44, 196-198.
205
Zauri, M., Berridge, G., Thezenas, M.L., Pugh, K.M., Goldin, R., Kessler, B.M., and Kriaucionis, S. (2015). CDA directs metabolism of epigenetic nucleosides revealing a therapeutic window in cancer. Nature 524, 114-118.
Zhang, J., Hulme, L., and Liu, J.L. (2014). Asymmetric inheritance of cytoophidia in Schizosaccharomyces pombe. Biol Open 3, 1092-1097.
Zhou, X., Cahoon, M., Rosa, P., and Hedstrom, L. (1997). Expression, purification, and characterization of inosine 5'-monophosphate dehydrogenase from Borrelia burgdorferi. The Journal of biological chemistry 272, 21977-21981.
Zimmermann, A.G., Spychala, J., and Mitchell, B.S. (1995). Characterization of the Human Inosine-5′-monophosphate Dehydrogenase Type II Gene. Journal of Biological Chemistry 270, 6808-6814.
Zupanc, G.K., and Horschke, I. (1996). Salvage pathway of pyrimidine synthesis: divergence of substrate specificity in two related species of teleostean fish. Comparative biochemistry and physiology Part B, Biochemistry & molecular biology 114, 269-274.
APPENDIX A
206
APPENDIX B
Use of Inosine monophosphate dehydrogenase activity assay to determine the specificity of PARP-1 inhibitors
Sajitha Anthony, Jeffrey R. Peterson, Yingbiao Ji Cancer Biology Program, Fox Chase Cancer Center, Philadelphia, PA 19111
To whom correspondence should be addressed: Dr. Yingbiao Ji, Fox Chase Cancer Center, 333 Cottman Avenue, Philadelphia, PA 19111, Telephone: (215) 728-7408; Fax: (215) 728-2412 E-mail: [email protected] 207
Running head: IMPDH assay for the specificity of PARP-1 inhibitors.
Key words: Inosine monophosphate dehydrogenase (IMPDH), enzyme, enzymatic activity, NAD, NADH, autofluorescence, reduction, oxidation, rate limiting.
Abstract
Inosine monophosphate dehydrogenase (IMPDH) is a rate-limiting enzyme involved in purine nucleotide biosynthesis. It is responsible for catalyzing the oxidation of inosine monophosphate (IMP) into xanthosine monophosphate
(XMP). Concurrently, the cofactor NAD+ is reduced to NADH. Poly (ADP- ribose) polymerase 1 (PARP-1) also utilizes NAD+ as a substrate to synthesize poly(ADP-ribose). It has been demonstrated that inhibition of PARP-1 activity can be an effective cancer therapeutic. . However, most PARP-1 inhibitors, including Olaparib, were developed as NAD+ analogs. Therefore, these inhibitors likely interfere with other NAD+-dependent pathways such as the one involved in de novo purine metabolism. In this chapter, we 208 describe a method to quantitatively measure IMPDH activity by taking advantage of the autofluorescence of the product NADH. We use this method to analyze the effects of Olaparib and non-NAD+-like PARP-1 inhibitor (5F02) on IMPDH activity. We found that Olaparib, unlike 5F02, significantly inhibits IMPDH activity in a dose-dependent manner. Our results suggest that IMPDH inhibition is an off-target effect of Olaparib treatment. The consequences of this effect should be addressed by future clinical studies.
1 Introduction
Poly(ADP-ribose) polymerase 1 (PARP-1) uses NAD+ as the substrate to assemble long and branching polymer named as poly(ADP-ribose) (pADPr)
(1). PARP-1 regulates a range of biological processes, including DNA repair and transcriptional control (2, 3). Deregulation of PARP-1 activity has been observed in multiple cancer types (4). PARP-1 inhibitors have been proven effective in eliminating BRAC1/2-associated ovarian and prostate cancer cells
(5, 6). However, current PARP-1 inhibitors have been developed as NAD+ competitors (7). Because NAD+ is utilized by the cell in a variety of metabolic processes (8), it is very likely that NAD+-dependent PARP-1 inhibitors interfere with other NAD+-dependent pathways, which leads to undesirable side-effects (9). New PARP-1 inhibitors have been recently developed based on H4-dependent route of PARP-1 activation (10, 11). These inhibitors are 209 expected to show superior specificity and efficacy relative to the conventional ones (11). We have used an IMPDH activity assay to compare the specificity of these new PARP-1 inhibitors to that of the commonly used clinical drug,
NAD+-dependent Olaparib.
IMPDH catalyzes the oxidation of IMP into XMP [Figure 1]. The cofactor
NAD+ is simultaneously reduced into NADH [Figure 1]. There are a number of conventional approaches to quantifying NADH. Colorimetric or bioluminescent assays are often used to determine total intracellular levels of
NAD+ and NADH, as well as their ratio (15). Because this method relies on assessing the total NAD+ levels in the cell lysate and a number of other enzymes reduce NAD+ in their reactions, it is not specific to IMPDH activity.
In this chapter, we describe a method which determines the specific activity of IMPDH directly. This involves the use of a fluorimeter to measure the buildup of fluorescence as NADH is being produced by purified IMPDH protein. We used this IMPDH activity assay to test whether PARP-1 inhibitors, the NAD+ analog Olaparib, and non-NAD+-like 5F02, reduce the production of NADH in the IMPDH reaction. This would be suggestive of
NAD+ competition (11). We found that Olaparib decreases the IMPDH activity significantly in a dosage-dependent manner in vitro, supporting our hypothesis that NAD+-dependent PARP-1 inhibitors interfere with NAD+- 210 dependent purine metabolism [Figure 2] (11). In contrast, treatment with
5F02 did not result in any significant reduction of NADH production in the
IMPDH assay [Figure 2] (11). Altogether, these results support the idea that
NAD+-independent inhibitors are more specific to PARP-1 than the currently used Olaparib (11).
2 Materials
2.1 Lysis buffer 30 mL of 50 mM Tris pH 8, 300 mM NaCl, 10 mM imidazole, 10mM β- ME, 800 mM urea 10% glycerol.
2.2 Wash buffer 2 mL of 50 mM Tris pH 8, 300 mM NaCl, 15 mM imidazole, 10 mM β-ME, 800 mM urea, 10% glycerol.
2.3 Elution buffer 2 ml of 50 mM Tris pH 8, 300 mM NaCl, 250 mM imidazole, 10 mM β- ME, 800 mM urea, 10% glycerol.
2.4 Dialysis buffer 2 L of 50 mM Tris pH8, 1mM DTT, 100 mM NaCl, 10% glycerol, 800 mM Urea.
2.5 Purified His tagged WT IMPDH
2.6 IMPDH reaction mixture master mix 211
1. Prepare a 2x solution of IMPDH activity buffer consisting of 100 mM
Tris pH 7.4, 200 mM KCl, and 20 mM MgCl2.
2. Prepare a 10 mM stock solution of NAD in H2O. 3. Prepare a 0.3 M stock solution of IMP in 250 mM Tris pH 7.4 4. Now create a “master mix” consisting of 1x IMPDH activity buffer, 3mM IMP and 1 mM NAD. For five replicates of four samples (20 in total), mix 1 ml of 2x IMPDH activity buffer (50 µl x 20), 20 µl of 0.3 IMP stock solution(1µl x 20) and finally 200 µl of a 50 mM NAD stock solution (10 µl x 20).
2.7 Cary Eclipse fluorescence spectrophotometer (Agilent Technologies)
3 Methods
3.1 Purification of His tagged WT IMPDH 1. Allow 100 µL of BL21 (DE3) cells to thaw on ice. Mix 1 µl of purified DNA of IMPDH in pET-28a(+) vector with cells and allow to sit on ice to 30 minutes. 2. Heat shock at 42°C for 40 seconds. Place back on ice for 2 minutes. 3. Add 1 ml of Luria Broth to tube of cells/DNA and grow at 37°C in shaking incubator for 1 hour. 4. Take 10 0µl of cells and spread of agar plate with Kanamycin concentration of 50 µg/ml. Put in 37°C incubator overnight. 5. Pluck single colony from agar plate using pipette tip and place in 20 ml of Luria Broth with Kanamycin concentration of 50 µg/ml. 6. Grow cells overnight at 37°C. 7. Add 25 ml culture to 1L of autoclaved Luria Broth. Add kanamycin to final concentration of 50 µg/ml and place in shaking incubator at 37°C
8. Allow cells to grow to OD600. 9. Induce with 1 mM IPTG for 3.5 hours at 30°C. 212
10. Pellet cells at 4000 rpm, for 20 minutes, at 4°C. Pour of supernatant, and freeze pellets at -80°C overnight. 11. Re-suspend pellet in 30 ml of Lysis buffer 12. Sonicate on ice with output control= 4, 50% duty cycle for 1 set of 10 pulses and one constant for 8 seconds. 13. Pellet cell debris by centrifugation for 15 minutes at 18000 rpm at 4°C
14. Wash 1 ml of beads in dH2O and then lysis buffer. Add to lysate. 15. Rotate lysates for 3 hours at 4°C. 16. Harvest beads at 1000 x g for 10 minutes at 4°C. 17. Wash beads 3 times with 15 ml of lysis buffer and 2 times with 1 ml of wash buffer. Elute protein from beads by rotating beads in elution buffer at 4°C for 15 minutes. 18. Repeat this three times, collecting the protein after each elution. You should have a total of 3 eluates at the end. 19. Analyze protein elutions on a 10% SDS PAGE gel to decide which eluates to pool together. 20. Dialyze protein in snake skin tubing overnight and then twice for one hour in approximately 600 ml of dialysis buffer each time.
3.2 IMPDH activity assay 1. Prepare the following samples each in a separate 1.5 ml microcentrifuge tube:
Without treatment: 61 µl of master mix + 34.4 µl of dH2O With 0.5 mM Olaparib: 61 µl of master mix +0.5 µl of Olaparib (from
100mM stock) + 33.9 µl dH2O. With 0.5 mM 5F02: 61 µl of master mix + 0.5 µL of 5F02 (from 100mM
stock) + 33.9 µl dH2O.
Vehicle only: 61µl of master mix +0.5 µL of DMSO + 33.9 µl dH2O . 213
2. Set the fluorimeter to measure NADH concentrations at 37°C with an emission wavelength of 440 nm and an excitation wavelength of 340 nm. Set measurement time for ten minutes. 3. Add 4.6 µl (1µg) of His-tagged IMPDH protein (0.22 µg/µl) to each of the above samples to start the reaction. Flick the tube slightly to mix and then add to a 100 µl quartz cuvette. Place the cuvette in the fluorimeter to begin measurements. 4. Create a standard curve for NADH florescence by plotting fluorescence vs various µM concentrations of NADH. Use the equation from this line to determine the specific activity of IMPDH (µM NADH/min/mg) for each of the above reactions. Significant differences are then determined using the Student’s t-test. 5. Repeat the steps 1-4 with increasing concentrations of Olaparib and 5F02 such as1mM, 2 mM, 4 mM, 6 mM and 8 mM. An example of the result is shown in Figure 2 (11).
4. Notes
1. It is best to obtain the enzymatic readings of protein immediately after purification. 2. The protein should be stored in aliquots of the volume you typically use in one day. 3. Aliquots should be flash frozen in a methanol and dry ice bath then stored at -80 degrees. 4. When defrosting protein, it is best to do it quickly using the warmth of one’s hands. Once close to being completely melted, flick tube, and then place on ice. 5. Keep protein on ice during entire use.
References 214
1. D’Amours D, Desnoyers S, Silva D, Poirier GG (1999). Poly(ADP- ribosyl)ation reactions in the regulation of nuclear functions. Biochem J 1999:342:249-268 2. Ji Y, Tulin AV (2010). The roles of PARP-1 in gene control and cell differentiation. Current Opinion in Genetics & Development 20:512-518 3. Kraus WL, Hottiger MO (2013) PARP-1 and gene regulation: progress and puzzles. Molecular Aspects of Medicine 34:1109-1123. 4. Rouleau M, Patel A, Hendzel MJ, Kaufmann SH, Poirier GG (2010). PARP inhibition: PARP-1 and beyond. Nature Reviews Cancer 10:293-301. 5. Fong PC, Boss DS, Yap TA, Tutt A, Wu P, Mergui-Roelvink M, Mortimer P, Swaisland H, Lau A, O'Connor MJ, Ashworth A (2009). Inhibition of poly (ADP-ribose) polymerase in tumors from BRCA mutation carriers. New England Journal of Medicine 361:123-134. 6. Mateo J, Carreira S, Sandhu S, Miranda S, Mossop H, Perez-Lopez R, Nava Rodrigues D, Robinson D, Omlin A, Tunariu N, Boysen G (2015). DNA-repair defects and olaparib in metastatic prostate cancer. New England Journal of Medicine 373:1697-1708. 7. Wahlberg E (2012). Family-wide chemical profiling and structural analysis of PARP and tankyrase inhibitors. Nat. Biotechnology 30: 283-288 8. Cantó C, Menzies KJ, Auwerx J (2015). NAD+ metabolism and the control of energy homeostasis: a balancing act between mitochondria and the nucleus. Cell metabolism 22: 31-53. 9.Tulin A (2011). Re-evaluating PARP-1 inhibitor in cancer. Nat. Biotechnology 29:1078-1079 10. Kotova E, Pinnola AD, Tulin AV (2011) Small-molecule Collection and High-throughput Colorimetic Assay to Identify PARP-1 Inhibitors. Methods Mol. Biol. 780:491-516. 11. Thomas C, Ji Y, Lodhi N, Kotova E, Pinnola AD, Golovine K, Makhov P, Pechenkina K, Kolenko V, Tulin AV (2016) Non-NAD-Like poly(ADP- 215
Ribose) Polymerase-1 Inhibitors effectively Eliminate Cancer in vivo. EBioMedicine. 2016 Oct 4. pii: S2352-3964(16)30456-X. doi: 10.1016/j.ebiom.2016.10.001. [Epub ahead of print] 12. Te HS, Randall G, Jensen DM (2007). Mechanism of action of ribavirin in the treatment of chronic hepatitis C. Gastroenterol Hepatol 3:218-225. 13. Ji Y, Gu J, Makhov AM, Griffith JD, Mitchell BS (2005). Regulation of the interaction of inosine monophosphate dehydrogenase with mycophenolic acid by GTP. The Journal of Biological Chemistry 281: 206-212. 14. Majd N, Sumita K, Yoshino H, Chen D, Terakawa J, Daikoku T, Kofuj S, Curry R, Wise-Draper TM, Warnich RE, Guarnaschelli J, Sasaki AT (2014). A review of the potential utility of mycophenolic mofetil as a cancer therapeutic. Journal of Cancer Research 2014:1-12 15. Shetty PK, Galeffi F, Turner DA (2014). Nicotinamide pre-treatment ameliorates NAD (H) hyoeroxidation and improves neuronal function after severe hypoxia. Neurobiology of Disease 62:469-78.
Figure 29. The schematic illustration of IMPDH catalyzing reaction.IMPDH uses NAD as the substrate to convert IMP to XMP with the production of NADH
216
Figure 30. Graph showing the relative IMPDH activity in the IMPDH reaction with PARP-1 inhibitors (Olaparob and 5F02) compared to the control (DMSO only). Under 1mM NAD concentration condition in the reaction, Olaparib inhibits IMPDH activity significantly with the 2-3 fold decrease in a dosage-dependent manner. In contrast, Non-NAD-like inhibitor 5F02 only showed very little inhibition effect on IMPDH activity even in the highest dosage (8mM). *P<0.05;
**P<0.01.
APPENDIX B
217
APPENDIX B
Figure 31. Coomassie-stained gel of purified IMPDH2 wild type and mutant proteins.