invaluable aid in the sequencing and plasmids work. Alison Cameron-Vendrig for suggestions (Husain lab) for telling me about the subscript/superscript shortcut ...
Calcium-Independent Calmodulin-IQGAP1 Interaction may occur in G2/M Phase: Establishing Protocols for Detecting Cell Cycle-Dependent Calmodulin-IQGAP1 Interactions
by
Hussein Butt
© Copyright by Hussein Butt 2016
Calcium-Independent Calmodulin-IQGAP1 Interaction may occur in G2/M Phase: Establishing Protocols for Detecting Cell CycleDependent Calmodulin-IQGAP1 Interactions Hussein Butt
2016
Abstract From embryogenesis until death, the cell cycle contributes to both health and disease. The cell cycle regulated by intracellular signaling involving second messengers and numerous proteins pathways. Second messenger calcium (Ca2+), and calmodulin (CaM) are well established as regulators of the cell cycle, but the myriad of effects and diversity of partners has not been fully explored. Additionally, CaM-binding protein IQGAP1 appears well linked to various established pathways that regulate proliferation as well. However, how Ca2+, CaM and IQGAP1 work together to influence cell cycle progression is not understood. Here, we investigate the cellcycle–dependence and Ca2+-dependence of CaM-IQGAP1 interaction. Under low Ca2+ conditions, there appears to be an increase in CaM-IQGAP1 at 16 and 24 h post stimulation. Concomitantly, there is a decrease in CaM-IQGAP1 at these time points in the presence of 2 mM Ca2+. This suggests that CaM-IQGAP1 interaction may be an important mitosis and cytokinesis regulator.
ii
Acknowledgments I give thanks to the following people that made this work possible and aided in its development and completion: Mansoor Husain for the giving me the opportunity to be a part of the School of Graduate Studies and supervising this work. I have learned more than academics from you. The members of my graduate committee, Scott Heximer and Anthony Gramolini. The Department of Physiology for their patience and persistence while I completed this thesis. Talat Afroze (Husain lab) your work is the platform that mine thesis stands upon. Sonya K Hui (Husain lab), your findings initiated this project. Mohammed Hossein Noyan-Ashraf (Husain lab), you were the only person with which I could discuss immunoblotting issues. Masayoshi Ishida (Husain lab) for help guiding experimental design and provision of NIH3T3 and HEK293T cell lines. Dhanwantee Mundil (Husain lab) for direction with optimizing the tissue culture. Omar El-Mounayri (Husain lab) for contributing your time unduly to this work. Also your invaluable aid in the sequencing and plasmids work. Alison Cameron-Vendrig for suggestions (Husain lab) for telling me about the subscript/superscript shortcut. I would probably have carpel tunnel without it. Sarah K Steinbach for the information on bimolecular fluorescence complementation. iii
Abdul Momen, for the reassurance that this work had merit when it just perpetual failure. Filip Konecny (von Harsdorf lab) for the help with the live cell imaging HEKs. Anton Mihic (Ren-Ke Li lab) for the advice on thesis writing. I heard it, tried it, but never truly understood it until it was too late. Ren-Ke Li’s lab for use of plate reader and sonicator. Ben Neel’s lab, and Dalia Barsyte and Cheryl H Arrowsmith’s lab for use of Odyssey imager. Seema Nagaraj, Kevin Truong and the Truong lab, and Matthew J Smith, Mitsuhiko Ikura and the Ikura lab for providing plasmids. Thomas J Ribar and Anthony R Means (Duke University) for providing information on immunoblotting CaM protocol. Melissa Brierley (Roche) you have taught me more in a demonstration than most people could in years. Cyrus Chang (EMD Millipore Corporation) for help troubleshooting CaM immunoblotting. Joanna Szulimowski (Novus Biologicals Canada) for help troubleshooting CAMK2D immunoblotting. Tina Pitarresi (GE Healthcare) for information about Sepharose chromatography. Peter Grav, Daniel A Newman and (ELWS) for the lectures and reviewing my writing. The university is doing their students a disservice by not adding these classes to the curriculum and eliminating bad habits of writing.
iv
Table of Contents Abstract ........................................................................................................................................... ii Acknowledgments.......................................................................................................................... iii Table of Contents ............................................................................................................................ v List of Tables ................................................................................................................................. ix List of Figures ................................................................................................................................. x List of Abbreviations .................................................................................................................... xii Chapter 1 Introduction .................................................................................................................... 1 1.1.
Thesis Outline ..................................................................................................................... 2
Chapter 2 Background .................................................................................................................... 4 2.1.
Cell Division Cycle............................................................................................................. 4
2.2.
Calcium and Calmodulin .................................................................................................... 7
2.2.1.
Ca2+ Signaling ......................................................................................................... 7
2.2.2.
Intracellular Ca2+ Receptor and Adaptor Protein Calmodulin ................................ 8
2.2.3.
Ca2+ in Proliferation and Cell Cycle ....................................................................... 9
2.2.4.
CaM in Proliferation and Cell Cycle .................................................................... 11
2.3.
IQGAP1 ............................................................................................................................ 16
2.3.1.
IQGAP1, Cdc42 and Rac1.................................................................................... 16
2.3.2.
IQGAP1, the Cytoskeleton and Adhesion ............................................................ 17
2.3.3.
Novel Interactions with Ras-Superfamily Proteins............................................... 18
2.3.4.
IQGAP1 and Wnt Signal Transduction ................................................................ 19
2.3.5.
IQGAP1 and Other Signal Transduction Pathways.............................................. 20
2.3.6.
IQGAP1 in the Cell Cycle and Proliferation ........................................................ 21
2.3.6.1.
IQGAP1 in G1 Phase............................................................................... 21
2.3.6.2.
IQGAP1 in S Phase ................................................................................ 22 v
2.3.6.3. 2.3.7.
IQGAP1 in M Phase and Cytokinesis .................................................... 23
2.4.
IQGAP1 Regulation by CaM................................................................................ 25
Cardiovascular Disease ..................................................................................................... 27
2.4.1.
SMC Proliferation Contributes to Atherosclerosis and Restenosis ...................... 28
Chapter 3 Hypothesis .................................................................................................................... 30 3.1.
Rationale ........................................................................................................................... 30
3.2.
Objectives ......................................................................................................................... 31
3.3.
Summary ........................................................................................................................... 32
Chapter 4 Experimental Approach................................................................................................ 33 4.1.
Immunoblotting................................................................................................................. 33
4.1.1. 4.2.
Immunoblotting CaM............................................................................................ 33
Protein Binding Assays..................................................................................................... 34
4.2.1.
Co-immunoprecipitation ....................................................................................... 35
4.2.2.
CaM-Sepharose Pull-down ................................................................................... 36
4.2.3.
FRET..................................................................................................................... 37
4.3.
Cell Cycle Analysis........................................................................................................... 39
4.3.1. 4.3.1.1.
Synchronization by Serum Starvation/Stimulation............................................... 39
4.3.2.
MOVAS.................................................................................................. 40
4.3.2.1.
Fucci System ......................................................................................................... 41 Fucci Mouse and Cell Line..................................................................... 42
Chapter 5 Materials and Methods ................................................................................................. 43 5.1.
Cell Lines and Cell Culture............................................................................................... 43
5.2.
Serum Stimulation ............................................................................................................ 43
5.3.
Protein Extraction and Quantification .............................................................................. 43
5.4.
Immunoprecipitation......................................................................................................... 44
5.5.
Immunoblotting................................................................................................................. 44 vi
5.5.1.
Immunoblotting CaM............................................................................................ 45
5.6.
Coomassie Blue Staining .................................................................................................. 46
5.7.
Plasmids ............................................................................................................................ 46
5.7.1.
pTriEx-3 hCALM1-Cerulean ................................................................................ 46
5.7.2.
pEGFP-C1 Venus-mIqgap1 .................................................................................. 46
5.7.3.
Plasmid Verification ............................................................................................. 47
5.8.
Transfection ...................................................................................................................... 47
5.8.1.
Nucleofection........................................................................................................ 47
5.8.2.
X-fection ............................................................................................................... 48
Chapter 6 Results .......................................................................................................................... 49 6.1. Optimizing Seeding for Protein Yield and Cell Recovery from Serum Starvation/Stimulation.............................................................................................................. 49 6.2.
Optimizing Detection of IQGAP1 by Immunoblotting .................................................... 50
6.2.1. 6.3.
Anti-IQGAP1 Pulldown........................................................................................ 51
Optimizing Detection of CaM by Immunoblotting .......................................................... 52
6.3.1. CaM Requires Unique Conditions during Transfer Compared to SizeMatched, Control Protein COX4 ...................................................................................... 53 6.3.2.
Fixation with Glutaraldehyde Improves CaM Signal ........................................... 56
6.3.3.
Using Potassium Phosphate Transfer Buffers Improves CaM Signal .................. 57
6.3.4.
Optimized Detection of Endogenous CaM by Immunoblotting ........................... 58
6.4.
CaM and IQGAP1 have Stable Expression Following Serum Stimulation...................... 58
6.5.
IQGAP1 Appears to Prefer ApoCaM over Ca2+-CaM Late in the Cell Cycle ................. 59
6.6.
Expression of Fluorescent Protein-Tagged CaM and IQGAP1........................................ 59
6.6.1.
Nucleofection into MOVAS ................................................................................. 59
6.6.2.
Demonstrating Veracity of Expression Plasmids ................................................. 60
6.6.3.
Nucleofection into Fucci T1 Mouse Carotid SMC ............................................... 61
6.6.4.
X-fection into HEK293T ...................................................................................... 61 vii
6.7.
Expression of CaM and IQGAP1 in Other Cell Lines...................................................... 62
6.7.1.
Immunoblotting Potential Positive Controls for CaM and IQGAP1 Co-IP ......... 62
Chapter 7 Conclusions .................................................................................................................. 64 7.1.
Technical Findings............................................................................................................ 64
7.1.1.
Seeding and Protein Yield .................................................................................... 64
7.1.2.
IQGAP1 Detection................................................................................................ 64
7.1.3.
CaM Detection ...................................................................................................... 64
7.1.4.
Plasmids and Microscopy ..................................................................................... 65
7.2.
Biological Findings ........................................................................................................... 65
7.2.1.
Pattern of Expression for CaM and IQGAP1 ....................................................... 65
7.2.2.
Cell Cycle-Dependent Co-IP of ApoCaM-IQGAP1 Complexes.......................... 65
Chapter 8 Discussion .................................................................................................................... 67 8.1.
Technical Findings............................................................................................................ 67
8.1.1.
Seeding and Starvation Protocol ........................................................................... 67
8.1.2.
CaM Detection ...................................................................................................... 68
8.2.
Biological Findings ........................................................................................................... 71
8.2.1.
Limitations and Confirming the Results ............................................................... 73
8.2.2.
Alternative Approaches ........................................................................................ 74
8.3.
Future Experiments ........................................................................................................... 77
8.4.
Future Work ...................................................................................................................... 78
References................................................................................................................................... 108
viii
List of Tables Table 1: Primer sequences. ......................................................................................................... 100
ix
List of Figures Figure 1: Schematic of Ca2+-CaM actions in eukaryotic cell cycle.............................................. 80 Figure 2: Schematic of IQGAP1 protein structure and sites of protein-proteins interaction listed by domain...................................................................................................................................... 81 Figure 3: Seeding and cell densities during serum starvation/stimulation protocol. .................... 82 Figure 4: Optimizing anti-IQGAP1 antibody dilution.................................................................. 83 Figure 5: Immunoblotting endogenous IQGAP1 from MOVAS. ................................................ 84 Figure 6: Quantification and linear range of IQGAP1 signal from MOVAS............................... 85 Figure 7: Anti-IQGAP1 immunoprecipitation.............................................................................. 86 Figure 8: Standard immunoblotting is insensitive for detecting CaM. ......................................... 87 Figure 9: Cox4 can be easily detected by immunoblotting........................................................... 88 Figure 10: Comparing PVDF pore size for effects on protein adsorption.................................... 89 Figure 11: Transfer conditions affects detection of small proteins............................................... 90 Figure 12: Effect of different fixatives on CaM detection............................................................ 91 Figure 13: Glutaraldehyde fixation is incompatible with infrared signal detection. .................... 92 Figure 14: Potassium phosphate transfers greatly improve detection of rhCaM.......................... 93 Figure 15: Detection of endogenous CaM protein by IB.............................................................. 94 Figure 16: Quantification and linear range of CaM signal from MOVAS. .................................. 95 Figure 17: Expression of CaM and IQGAP1 following serum stimulation.................................. 96 Figure 18: Reciprocal co-IP of CaM and IQGAP1....................................................................... 97 Figure 19: Nucleofection of expression plasmids into MOVAS. ................................................. 99 x
Figure 20: Plasmid maps of fluorescent fusion constructs. ........................................................ 101 Figure 21: Excitation and emission spectra of the fluorophores................................................. 102 Figure 22: Nucleofection of expression plasmids into Fucci T1 mouse carotid SMC and imaging of Fucci reporters. ....................................................................................................................... 103 Figure 23: Transfection and expression of fusion constructed in HEK293T. ............................ 104 Figure 24: Detection of IQGAP1 and CaM in different cell lines. ............................................. 105 Figure 25: Detection of CAMK2D in different cell lines. .......................................................... 106 Figure 26: Detection of ERK in different cell lines. ................................................................... 107
xi
List of Abbreviations 1º, primary; 2º, secondary; APC, adenomatous polyposis coli; APC/C, anaphase promoting complex/cyclosome; apoCaM, Ca2+-free CaM; ATH, atherosclerosis; BiFC, bimolecular fluorescence complementation; Ca2+, calcium ions; CaM, calmodulin; CaM-CFP, CaM-Cerulean; CaMK, Ca2+/CaM-dependent protein kinase; cdk, cyclin-dependent kinase; CFP, cyan fluorescent protein; CHD, calponin-homology domain; CNBr, cyanogen bromide; Cox4; cytochrome C oxidase subunit 4 DVL, Dishevelled; EGTA, ethylene glycol tetraacetic acid; ERα, estrogen receptor α; xii
FRET, Förster resonance energy transfer; Fucci, fluorescent, ubiquitination-based cell cycle indicator; G0 phase, zeroth gap/resting phase; G1 phase, first gap phase; G2 phase, second gap phase; GAP, GTPase-activating proteins; GDP, guanosine diphosphate; GRD, RasGAP-related domain; GTP, guanosine triphosphate; HA, hyaluronic acid; IB, immunoblot; IP, immunoprecipitation; IQGAP, IQ motif-containing GTPase-activating proteins; kDa, kilodalton; KP, potassium phosphate; LC-MS/MS, liquid chromatography-tandem mass spectrometry; M phase, mitotic phase; MAP kinase, mitogen-activated protein kinase; O/N, overnight; pI, isoelectric point;
xiii
Rb, retinoblastoma protein; RGC, RasGAP C-terminal, rhCaM, recombinant human CaM; S phase, synthesis phase; SMC, vascular smooth muscle cell; TAg, large T antigen; YFP, yellow fluorescent protein; YFP-IQGAP1, Venus-IQGAP1;
xiv
1
Chapter 1 Introduction The clinical prognosis of cardiovascular disease has been significantly improved in recent years due to decades of research. Despite these great strides, cardiovascular disease remains an active area in both medical and research laboratories. Developments to improve diagnosis, pharmaceutical treatment, and surgical intervention have dramatically changed the disease from an acute incidence to a chronic disease. As a result, patients no long simply die with cardiovascular disease; instead they live with it for the rest of their lives. To live with cardiovascular disease means living with the risk of reoccurrence and lifelong management of one’s health. Recurrent cardiovascular events and medical management impose significant financial burden on medical systems and morbidity on patients. The ultimate goal is to have efficient and effective options to use in the clinic, and to increase the quality of living for patients.
For example, narrowing of arteries is treated with angioplasty. But the angioplasty treatment may result in re-narrowing of the vessel, worsening the original condition and requiring repeated angioplasty. The development of stents, especially drug-eluting stents, has largely reduced the incidence of restenosis [1-4]. However, to prevent thrombosis from occurring, patients are required to take anticoagulant indefinitely. Thus studies of restenosis remain a sizeable area of inquiry.
Restenosis develops from a growth of a fatty, fibrous tissue within a blood vessel. It is also marked with cellular infiltrate, particular immune cells and vascular smooth muscle cells (SMC). The eluted drugs from stents appear to minimize restenosis by inhibiting the proliferation of these SMC [5-7]. In order to identify and develop new elusion drugs to address restenosis, basic understanding of the proliferation is required. Cells proliferate through by passing through the cell cycle, a complex process involving numerous proteins and signaling pathways. Identifying novel regulators of the cell cycle and thus proliferation is important to finding alternatives to the current drugs and drug-eluting stents.
2
1.1. Thesis Outline The cell cycle is controlled by multiple signal pathways. But the one that is used repeatedly throughout the cell cycle is calcium signaling. This thesis attempts to identify protein-protein interactions that are controlled by calcium signaling and the cell cycle. From this work, a possible cell-cycle dependent interaction between calcium binding protein, calmodulin, and a known calmodulin target, IQGAP1, is identified.
In Chapter 2 below, relevant background information is reviewed. In § 2.1 Cell Division Cycle, a basic overview of the cell cycle is given, outlining the different phases and the key events that occur within each phase. Also, function and dysfunction of the cell cycle are accounted. This is followed by § 2.2 Calcium and Calmodulin, which describes the centrality of calcium biology to life in general, as well as specific roles to proliferation and the cell cycle. Here, the ubiquitous calcium-binding protein, calmodulin, is introduced and the roles that calmodulin is known to play in the cell cycle are illustrated. Through the diversity of calmodulin actions and targets for regulation, we have identified IQGAP1 as potentially favorable protein of interest for intervention and study. In § 2.3 IQGAP1 IQGAP1’s function as a scaffolding protein is discussed. As a scaffolding protein, IQGAP1 interacts with a myriad of proteins and connects various signaling pathways integrating them together. Here, these interactions are grouped by function and establishes what is known about IQGAP1, relative to the cell cycle. The powerful regulation that CaM exerts over IQGAP1 is also detailed. In § 2.4 Cardiovascular Disease, cardiovascular disease as it pertains to this work is summarized and evidence for proliferation of SMC contributing to restenosis is briefly explained.
Based on the guidance of Chapter 2, Chapter 3 Hypothesis proposes a framework through which a potential cell-cycle regulating calmodulin-IQGAP1 interaction would be tested and demonstrated. Chapter 4 Experimental Approach examines methods we may use to analyze the model and to prove our hypothesis. The techniques required must grant the ability to detect proteins, assay protein-protein interactions and discriminate between stages of the cell cycle. The specifics of the methods used in this thesis are elaborated in Chapter 5 Materials and Methods.
3
In Chapter 6 Results, the information that was gained by testing our hypothesis is reported. The path to detecting a cell-cycle dependent calmodulin-IQGAP1 interaction was wrought with problems at each step. How issues were encountered, addressed and overcome is recounted. Additionally data that establishes the possible existence and timing of an interaction between calmodulin and IQGAP1 within the cell cycle are shown.
In Chapter 7 Conclusions, the findings from Chapter 6 are reiterated and demarked. The findings are divided into § 7.1 Technical Findings that are related to technical optimizations that were discovered during this work, and § 7.2 Biological Findings that are related to the proposed hypothesis related to calmodulin and IQGAP1 regulating the cell cycle. Chapter 8 Discussion presents our findings in the context of what other authors have described. Similarly, this discussion is divided between the communication of optimization and procedures, and comparing biological findings to literature. The conclusion found from this work does not have the statistical significance to confidently prove the existence of a calmodulin-IQGAP1 interaction that depends on the cell cycle. The results only hint to a possibility. How one would untangle the truth and definitively test our hypothesis is proposed. Finally, the potential impact of this line of inquiry is reflected upon.
4
Chapter 2 Background
2.1. Cell Division Cycle Each time a mother cell replicates itself to create two daughter cells; the mother cell undergoes one round of the cell division cycle, better known as the cell cycle. The cell cycle is a complex, well-organized and concerted effort. This process is regulated by both external and internal factors [8]. Extracellular factors influence cells from outside. Such inputs include soluble factors like available nutrients, hormones, growth factors and growth inhibitors, as well as interaction with the extracellular matrix and signals between neighboring cells. These signals communicate through the cell membrane and are integrated by intracellular mechanisms. Surface receptors receive these signals and act downstream to activate second messengers and protein pathways. Together activators, regulators, inhibitors, effectors and enzymes produce refined changes to cellular machinery and architecture by transcription, translation, translocation and posttranslational modification. Pro-proliferative pathways promote growth and division, while cell cycle inhibitors counteract the effects of proliferative factors. The balance between the two is how the cell evaluates whether to remain at rest or to proceed through the cell cycle.
Once the decision to enter the cell cycle is made, the cell must (1) gather the required resources, (2) synthesize necessary proteins, (3) duplicate its genetic material, (4) segregate duplicated DNA, organelles and proteins, (5) and cleave the cytoplasm with the cell membrane creating two new cells. These steps each have activators that initiate the step, effectors that carry out the processes, regulators that ensure that the steps are carried out properly, and inhibitors that limit these processes that prevent excess activity.
During the cell cycle, the cell changes both in form and function. These changes are grouped into phases. The most obvious change is when the cell splitting from one mother cell into two daughter cells. The interval between divisions is the interphase. It is during the interphase when the cell duplicates its genome. But before DNA duplication occurs, the cell first grows by taking
5
up available nutrients. This is the first division of the interphase, the first gap phase or G1. In G1, gene and protein expression change dramatically. The cell removes cell cycle inhibitors, and synthesizes and activates regulatory proteins and enzymes necessary for DNA replication.
When the cell is ready to replicate its DNA, the cell enters the synthesis or S phase. In this phase replication machinery is assembled and the DNA opened. Each strand of double stranded DNA serves as the template for DNA polymerases to create the complementary strand. Since DNA is exposed, S phase is when DNA damage is detected and repaired [9-11]. Similarly, genome integrity is maintained by ensuring the DNA is replicated exactly once [12-14]. This creates two copies of DNA, one for each daughter cell. DNA synthesis is monitored by several proteins so that partially synthesized strands must be finished and completed DNA is not reinitiated. This regulation occurs so division proceeds normally and prevents an abnormal amount of DNA, or aneuploidy in the daughter cells. Once the DNA is successfully synthesized, the cell will proceed to cell division. Before division occurs, S phase is separated from mitosis by a second gap phase, G2. G2 finalizes the interphase with the cell ensuring the necessary internal environment for mitosis.
Following the interphase, the cell enters the mitotic or M phase. In M phase, the cell undergoes mitosis and cytokinesis. In mitosis, the DNA copies separate from each other. The DNA condenses further from the less compact chromatin into chromosomes. During this time, the nuclear envelope dissolves [15]. With the DNA exposed to the cytoplasm, microtubules are able to bind to chromosomes [16]. Once both sister chromatid are tethered, chromosomes are pulled to opposite poles along the microtubule spindles by motor proteins [17]. The equal pulling forces align the chromosomes along a central plane because genetic duplicates are held together at the centromere. The centromeres are cleaved and the sister chromosomes divided between the two poles of the mother cell [17]. Mitosis is completed by the formation of nuclei around the separated chromosomes. With the chromosomes divided, the mother cell divides in two by cytokinesis. The cell membrane pinches together due to action of the actomyosin contractile ring [18]. The contractile ring is a structure made primarily of actin and myosin protein complexes. The ring contracts to create the cleavage furrow at the center between the two ends of the cell. The
6
combination of the contractile ring and membrane trafficking create independent plasma membranes that separates the cytoplasm and other cellular components creating two cells [18, 19].
Completing M phase starts the cells in interphase again. Daughter cells can begin the cell cycle anew by repeating the process or enter a so-called resting or quiescent state, G0 [20]. In G0, the cell remains ‘outside’ of the cell cycle and is characterized by a lack of normal cell cycle marker proteins [21]. For some cell types, such as terminally differentiated neurons, they may never reenter the cell cycle again [8]. Other cells retain their proliferative ability, such as stem cells or hepatocytes. These cells wait until environmental conditions are once again suitable for the cell cycle. There is an additional state outside the cell cycle that cells can enter called senescence. Senescence is a consequence of irreparable damage to the cell and its genome. In this state, the cell will not replicate as a mechanism to preserve whole organism integrity and viability.
The cell cycle is the mechanism by which cells proliferate. Physiologically, proliferation is integral to development, growth and regeneration. Failure to initiate the cell cycle contributes to problems in wound healing, aging and disease [8]. Additionally, dysregulated passage through the cell cycle is also pathological. Proliferative intracellular signaling networks are key targets for mutations that create tumors and promoting proliferation appears to motivate five of the hallmark capabilities and characteristics of neoplastic cancer [22, 23]. Similar undesired tissue growth contributes to some types of cardiovascular pathologies [24].
7
2.2. Calcium and Calmodulin Calcium (Ca2+) is a divalent, soluble, alkali metal ion that is a necessary micronutrient for good health and well being [25]. About 99% of Ca2+ in the body is present in the teeth and bones as solid minerals [25]. This pool of Ca2+ is used for its structural and mechanical properties, but is biochemically inert. The remaining 1% of Ca2+ is aqueous and is present in blood and soft tissue. Aqueous Ca2+ is biologically available and its concentration is maintained within a narrow range. Because of this regulation, the solid phase also serves as Ca2+ storage that may be resorbed to maintain homeostasis at the cost of bone density. Aqueous Ca2+ serves several functions in the body. Aqueous Ca2+ is critical as it regulates blood pH and osmotic pressure [25]. Ca2+ also plays roles in membrane potential and depolarization of neurons and muscles. In neurons, Ca2+ stimulates the release of neurotransmitter and propagates neurological signals. In muscles, Ca2+ is required for contraction. Throughout the body, aqueous Ca2+ is used as a biochemical cofactor for enzymes and second messenger sensitive to multiple stimuli and producing varied downstream effects. Ca2+ regulates biological processes such as embryonic development, cell motility, cellular secretion, learning and memory, cell death, and proliferation [26, 27].
2.2.1.
Ca2+ Signaling
Ca2+ is able to elicit different cellular effects because it is highly controlled and tightly regulated. Under resting conditions, Ca2+ is stored outside cells in the extracellular space or inside organelles such as the mitochondria, endoplasmic reticulum and sarcoplasmic reticulum. Cytosolic Ca2+ is maintained at low concentrations, on the order of 10−8–10−7 M [20, 28-31]. Extracellular Ca2+ concentration is approximately 10−3 M [20, 31]. Maintenance of the distribution of Ca2+ requires energy as Ca2+ pumps move ions against the concentration gradient. When stimulated, Ca2+ channels briefly open and the ions diffuse down the concentration gradient. The influx can raise Ca2+ concentration to the magnitude of 10−5 M [20, 28-30]. The increase is temporary as Ca2+ is actively pumped back out of the cytosol. Transient Ca2+ influxes vary by magnitude, spatially, and temporally. The amplitude of the transients can be modulated by the amount of Ca2+ that enters the cell. Spatially, transients can be localized to small areas within the cell or occur across the whole cell [32]. Temporally, influxes can vary in both duration and
8
frequency. Cells are able to decode these complexities to elicit appropriate responses. As a result, Ca2+ signals are a ubiquitous second messenger for a variety of growth factors, hormones and pathways.
The multiplicity of cell types and the characteristics of Ca2+ influxes permit targeting of many different proteins [20]. Despite the large number of proteins and pathways that are sensitive to Ca2+ regulation, only a few proteins can bind to Ca2+ directly. Instead, the second messenger acts through a Ca2+-binding protein that acts as an adaptor. Once Ca2+-bound, the adaptor proteins bind to various targets, connecting the Ca2+ signal to the effector protein. These targets then cause the diverse phenotypic changes associated with Ca2+ signals. The variety of Ca2+-binding proteins add another layer of complexity to Ca2+ signaling. Ca2+ binding proteins include troponin C, the tissue-specific S100 protein family, and the ubiquitous calmodulin (CaM).
2.2.2.
Intracellular Ca2+ Receptor and Adaptor Protein Calmodulin
CaM is a monomeric, small (148 amino acids, 17 kDa), heat-stable, acidic (pI = 3.9–4.3) protein [33-36]. CaM is a cation-binding protein that acts as the intracellular Ca2+ receptor for cells. It is a dumbbell-shaped protein comprised of two pairs of EF-hands separated by a flexible αhelix [37]. An EF-hand is a helix-loop-helix motif that is able to bind one Ca2+ ion. With its four EF hands, CaM can bind up to four Ca2+ ions [36, 37]. Ca2+ binding in the EF-hands results in a conformational change in the central, flexible linker region that alters the ability for CaM to bind proteins [38]. The conformational change increases the α-helical content of the protein and exposes hydrophobic pockets [30, 36, 38-40]. These changes allow CaM to wrap around several target amino acid sequences. By binding Ca2+ and target sequences, allows CaM to sense the presence of Ca2+ inside the cell on behalf proteins that can not bind Ca2+ themselves and to act as an adaptor between Ca2+ signal and the target protein.
9
CaM levels are inversely dependent on Ca2+ [32]. Thus cells increase their sensitivity to calcium by increasing CaM levels when extracellular Ca2+ concentrations are low. Conversely, cells decrease their sensitivity to Ca2+ when Ca2+ is abundant by reducing CaM expression. This maintains a relatively constant level of downstream activation regardless of the availability of Ca2+. The model for Ca2+-dependent CaM function is that Ca2+-free CaM, (apoCaM) binds free Ca2+to create Ca2+-CaM complex, which can then affect downstream pathways [30, 41, 42]. Thus by mass action, increasing CaM concentration reduces the amount of intracellular Ca2+ needed to activate a given pathway [30, 41, 42].
CaM was originally discovered as an activator protein for Ca2+-dependent cyclic 3′,5′-nucleotide phosphodiesterases [33, 34]. It was historically called Ca2+-dependent regulatory or modulator protein and it was soon identified to regulate hundreds of proteins and enzymes [43]. Other Ca2+ binding proteins are generally limited to one or a few specific metabolic pathways or processes [44]. These proteins are less versatile because they are only found in vertebrates, and are tissue-restricted even in species where they are present [44]. In contrast, CaM is well-conserved throughout evolution with homologs found in all eukaryotes investigated, including simple, model organisms such as Saccharomyces cerevisiae (budding yeast), Schizosaccharomyces pombe (fission yeast), Dictyostelium discoideum (slime mold), Aspergillus nidulans, and Drosophila melanogaster. Deletion of CaM has proven it essential in S. pombe and S. cerevisiae [45]. Human and rodents have three CaM genes (CALM1, -2, -3) on separate chromosomes. The three loci produce identical CaM proteins that share 100% identity between mammalian species. In addition to being present in all eukaryotes, CaM has been found to be expressed in varying amount in all tissues. Deleting the CALM genes to investigate the importance of CaM protein in mammals would be difficult and probably only serve to reinforce the critical role CaM plays in eukaryotic biology [32].
2.2.3.
Ca2+ in Proliferation and Cell Cycle
Ca2+ and CaM regulate numerous and diverse processes and pathways. Many are cell-type specific. However, all cells are the product of cell division. Cells either divide to self-renew and
10
maintain tissue integrity, or are the terminally differentiated daughter cells of a dividing progenitor. The cell cycle is controlled by variety of growth factors and different cell types respond to different mitogens. Despite this variety in regulation, changes in Ca2+ and the requirement for CaM are commonly seen downstream of the growth factor simulation and multiple times during the cell cycle.
Both intracellular and extracellular stores of Ca2+ are necessary for cells be able to divide. Blocking Ca2+ pumps on the endoplasmic reticulum by pharmacological agents depletes intracellular Ca2+ stores. Depleting intracellular Ca2+ stores reversibly arrest cells in a G0-like state. Reducing the concentration of Ca2+ in the media from 1.0 mM to 0.1 mM reduces the rate of proliferation [46, 47]. Conversely, addition of CaCl2 causes DNA synthesis in contact-inhibited cells. Plasma Ca2+ is need for proliferation models in vivo, including induced thymic lymphocytes, bone marrow and partial hepatectomized liver [48-51]. Thymic lymphocytes initiate DNA synthesis following a sufficiently large injection of extracellular Ca2+ [51]. When the hematocrit of rats is reduced, the increase in proliferation of hematopoietic stem cells is accompanied by an increase of plasma Ca2+ [50]. If the rise in Ca2+ is blocked by thyroparathyroidectomy, there is no increase in the bone marrow cell division [50]. Conversely, the addition of the Ca2+, Ca2+-increasing parathyroid hormone or ionophore A-23187 to increase intracellular Ca2+ permits DNA synthesis in the bone marrow colony forming units [49, 51, 52]. Reducing plasma Ca2+ by parathyroidectomy or thyroparathyroidectomy causes hepatocytes to lose the proliferative response to partial hepatectomy [48]. CaCl2 is able to rescue this proliferative phenotype [53]. Tumors contain elevated amounts of Ca2+ than normal or host tissue [54].
Ca2+ is required to progress through the cell cycle at several points (Figure 1). Cells are sensitive to Ca2+ in both early G1 and the G1/S transition. These transitions are accompanied by transient intracellular Ca2+ increases. While Ca2+ appears necessary to exit G1 and initiate S phase, Ca2+ is not necessary during S phase. There is no evidence of Ca2+ transients or signaling, which suggests that S phase activities, including DNA synthesis are largely independent of Ca2+ metabolism. Additionally, depletion of Ca2+ results causes cells to arrest in either a quiescent-
11
like state, G1 or G2/M. This is consistent with the idea that cells make a major commitment to begin S phase and to continue through the cell cycle. Ca2+ transients are also observed within M phase, particularly the metaphase/anaphase transition and cytokinesis.
2.2.4.
CaM in Proliferation and Cell Cycle
Ca2+ actions in the proliferation and the cell cycle are primarily through CaM. Levels of CaM are elevated under conditions in which cells are undergoing growth and proliferation. Cultures have more CaM at subconfluent densities than at confluent densities [55]. Similar proliferationdependent elevation in CaM expression can be induced in vivo [56]. Chicken embryo fibroblasts and normal rat kidney cells transformed by Rous sarcoma virus, and NIH 3T3 transformed by SV40 also have increased CaM expression [55, 57-60]. Similarly, neoplastic cells have increased CaM content, and their level of elevation is correlated to their degree of malignancy [54, 56, 61]. This increased CaM expression contributes to the proliferative phenotype as inhibition of CaM by small molecule inhibitors, trifluoperazine and naphthalene sulfonamide, produces a dosedependent reduction in the ability of metastatic, breast cancer cells to form colonies, as well as other cancerous and non-cancerous cell lines [47, 62, 63]. As a consequence, transformed cells are able to proliferate in media with reduced Ca2+. By increasing their CaM content, the cells increase Ca2+ sensitivity and thus gaining autonomous proliferation. Increased Ca2+ sensitivity allows neoplastic cells to activate various enzymes and initiate pro-proliferative processes.
Expression of CaM is cell-cycle regulated. It is lowest in the first few hours following mitosis and increases again at the G1/S transition, reaching a plateau at S phase and are maintained until cytokinesis when it is divide between daughters [64, 65]. It appears that cell-cycle dependent CaM expression closely matches mRNA levels, with mRNA concentration increasing prior to S phase [66-68].
As CaM is the primary Ca2+ sensor, CaM is required at correspond point in the cell cycle as Ca2+ requirement (Figure 1). Increasing CaM expression shortens G1, accelerating the cell cycle,
12
while reducing CaM slows growth [69, 70]. When CaM is added, cell that are cell cycle arrested by low Ca2+ media, are able to overcome the inhibition and initiate S phase and beginning of DNA synthesis [47]. Conversely, entry into S phase is inhibited by both trifluoperazine and anti-CaM antibodies [47]. CaM continues to play diverse roles during G2, mitosis and cytokinesis.
CaM function depends on CaM binding partners and influence over downstream effects. CaM has multiple levels of influence on the cyclin/cyclin-dependent kinases (cdk) function. Cyclins are a family of proteins that have cell cycle-dependent expression. Cyclins regulate the cell cycle through their interactions with cdk proteins. In contrast, cdk protein expression is maintained throughout the cell cycle, but the protein is unstable. The interaction cyclin/cdk interaction stabilizes the cdk and activates its protein kinase function. Complexed together, cyclins/cdk permits cell cycle progression by targeting downstream proteins with phosphorylation. The earliest cyclin/cdk complex during the cell cycle is cyclin D/cdk4. Cyclin D and cdk4 expression and accumulation appears to be independent of CaM function [71]. However, neither protein possesses a nuclear localization signal. It appears that CaM acts as a chaperone to aid shuttling of the cyclin D/cdk4 complex into the nucleus [72]. Neither protein interacts with CaM. Instead, Hsp90 may act as a bridge between CaM and cyclin D/cdk4. Similarly, cdk inhibitor p21, is able to promote cyclin D/cdk4 activity by stabilizing cyclin D, assembling cyclin D/cdk4 complexes and regulating nuclear import. However, nuclear accumulation of p21 itself requires CaM [73].
By comparison, serum-induced expression and activation of cdk1, cdk2, cyclin A and cyclin E are all Ca2+-CaM dependent [71, 74]. Cyclin D/cdk4 and Cyclin E/cdk2 successively phosphorylate tumor suppressor, retinoblastoma protein (Rb). Hyper-phosphorylation of Rb prevents Rb from binds to and inhibits E2F1. E2F1 is a transcription factor that controls expression of genes necessary for S phase including as DNA polymerase. Cyclin E/cdk2 also contribute to assembly of pre-replication complexes. Cyclin A displaces cyclin E from cdk2. Cyclin A/cdk2 signals the commencement and progression of DNA synthesis. Cyclin A subsequently changes interactions from cdk2 to cdk1, which contributes to G2 progression. Thus inactivation of Rb and initiation of S phase is Ca2+-CaM–dependent.
13
Other possible targets for CaM in the cell cycle include calcineurin and Ca2+/CaM-dependent protein kinases (CaMK). Calcineurin is a Ca2+/CaM-activated protein phosphatase that activates NFAT by allowing nuclear entry. Calcineurin is important in T-cell activation and is targeted by cyclosporin A and tacrolimus. Expression of calcineurin peaks around S phase [75]. Inhibition of calcineurin by cyclosporin A causes G1 arrest, but the mechanism is unclear [32]. Cyclosporin A treatment causes an increase in p21 expression and a decrease in cyclins E and A [76, 77]. Calcineurin appears to promote cyclin D1 expression, but silent the cdk4 promoter [78-80]. This is odd as both are required during G1. There may be cell type-specific interactions or experimental considerations that need to be taken into account.
CaMK are a family of protein kinases from with multiple isoforms and splice variants. CaMKII plays multiple roles during cell cycle. The somatic isoform, CaMKII, is present in the nucleus during the interphase and is a part of the mitotic apparatus during M phase [81]. Expression of CaMK peaks at G1/S [82]. In mammals, CaMK inhibition causes G1 arrest, but depending on cell type used in the experiments, there are arrests categorically different [71, 83]. These arrests are similar to the early G1 and G1/S. CaMKII activity is also necessary during M phase. Inhibition of CaMKII in S phase-synchronized cells resulted in G2/M arrest [84]. CaMKII is involved in nuclear envelope breakdown [85]. CaMKII phosphorylates Cdc25, activating the phosphatase [86]. Cdc25 acts cdk1 to activate the cdk1/cyclin B complex and promote mitosis. While calcineurin and CaMK are attractive as Ca2+-CaM targets during the cell cycle, neither one of the proteins are essential for yeast [82, 87, 88].
In S phase, CaM accumulates in the nucleus and associates with the nuclear matrix. The accumulation of nuclear CaM reaches its maximum when DNA synthesis is at its peak. Nuclear CaM interacts with several proteins including myosin light chain kinase [89]. In addition, several components of DNA polyermase holoenzyme are CaM binding proteins and it seems that activation of DNA polymerase requires Ca2+-CaM [71, 90, 91]. However, once cells initiate DNA synthesis, CaM appears to be temporarily expendable. When cells in S phase are treated with
14
CaM antagonist, W-7, the cell are blocked in late G2 or early M phases [64]. The cells do not arrest in S phase. Therefore passing and leaving S phase and entering G2 are most likely independent of CaM. This coincides with Ca2+ requirement, as Ca2+ is necessary for exiting G1 and initiating S phase, as well as ending G2, but not requiring during S-G2 transition.
Entering mitosis from G2 requires CaM activation of PLK1. PLK1 is activated late in G2 and is a regulator of various mitotic events. The interaction occurs during G2/M, but not G1/S and is Ca2+-dependent [92]. The activity of PLK1 is inhibited by CaM-inhibitor, W-7, or transfecting of mutant CaM [92]. Conversely, overexpression of CaM increased PLK1 activity [92]. The interaction between CaM and PLK1 results in accelerated progression through M phase and increased Cdc25 hyper-phosphorylation.
CaM also regulates mitosis and cytokinesis. Repressing CaM expression in S. cerevisiae results in shorter mitotic spindles and chromosome loss. This results in defects in the segregation of chromosomes and cytokinesis. Similar defects in segregation are seen in S. pombe. CaM associates with microtubule spindles during normal mitosis and regulates tubulin [42, 59, 93]. Ca2+CaM inhibits microtubule assemble and promotes disassembly in vitro through a stoichiometric mechanism [42, 59]. Reducing CaM levels with RNAi or inhibiting CaM with calmidazolium results aberrant mitosis and cytokinesis [94, 95]. Defects include loss of mitotic spindle integrity, increase the number of polyploid cells, and induced bi-nucleate cells [94]. Actin organization, and microtubule assemble and disassemble are important processes during mitosis.
A CaM target during M phase may be Aurora B kinase. Aurora B regulates cytokinesis by stabilizing the cleavage furrow, delaying abscission to allow for proper chromosomal segregation. During cytokinesis, Aurora B is sent to be degraded by E3 ubiquitin ligase, SCFFBXL2 [94]. Aurora B interacts with CaM in a Ca2+-independent manner [94]. The CaM binding site on Aurora B overlaps with the FBXL2 recognition site on Aurora B. CaM binding to Aurora B blocks FBXL2 interaction and ubiquitination, thus protecting Aurora B protein from
15
degradation [94]. When the chromosomes fail to partition, CaM dissociates from the spindles and colocalizes with Aurora B [94]. Loss or inhibition of CaM results in decreased Aurora B levels, mitotic defects and polyploidy [94]. Thus CaM appears to sense failed chromosome segregation and delay cytokinesis by protecting Aurora B from ubiquitin-mediated degradation.
Additionally, FBXL2 weakly binds to CaM, and the interaction is enhanced by Ca2+ [94, 95]. FBXL2 can induce mitotic arrest in a Ca2+-dependent manner [95, 96]. FBXL2 also targets cyclin D3 for proteasomal degradation [95]. Similar to Aurora B, CaM binds to and protects cyclin D3 from SCFFBXL2. CaM interacts with cyclin D3 in the presence of EDTA but the interaction is disrupted with Ca2+ [95]. While Ca2+ dissociates CaM from cyclin D3, it also promotes the cyclin D3-FBXL2 interaction. Taken together, apoCaM maintains protein levels of both Aurora B and cyclin D3 by blocking FBXL2 binding, and preventing ubiquitinylation. Conversely, Ca2+-CaM may facilitate M-phase degradation of these proteins.
It is clear that Ca2+ and CaM play critical roles throughout the cell cycle. However, definitive and essential targets for Ca2+-CaM have been so far elusive. Also, studies on Ca2+ and CaM are far from having been exhaustive. Simple eukaryotic models have yielded neither essential Ca2+CaM–dependent targets, nor conserved mechanisms between the different species [32]. Considering how important the cell cycle is for all eukaryotic life, it would be expected that the cell cycle has a high degree of homology and was conserved through evolution. However, there has been considerable diversification of methods, mechanisms and regulation for cell cycle processes [20, 97]. Therefore, consensus between species may remain elusive.
16
2.3. IQGAP1 IQ motif-containing GTPase-activating proteins (IQGAP) are a family of proteins with three members in humans, IQGAP1, -2 and -3. IQGAP family proteins are present in S. pombe, S. cerevisiae, D. discoideum and C. elegans [98]. IQGAP1 was the first discovered and is the most studied [99]. This is due IQGAP1 being expressed constitutively and throughout the body [99, 100]. The less-studied IQGAP2 and IQGAP3 have expression restricted to specific tissues [101, 102]. IQGAP1 is a large (1 657 amino acids, 190 kDa), multi-domain protein [99] that interacts with dozens of proteins from diverse pathways (Figure 2). Based on its structure and function, IQGAP1 is described as a scaffold protein that improves signal transduction efficiency by bringing together proteins and facilitating their regulation. Since IQGAP1 interacts with various pathways, IQGAP1 allows cross-talk between these different pathways by integrating stimuli from different proteins to produce appropriate effects.
2.3.1.
IQGAP1, Cdc42 and Rac1
IQGAP1 was first discovered due to its homology to Ras GTPases activating proteins (GAP) [99]. However, neither Ras nor RhoA binds to IQGAP1 RasGAP-related domain (GRD) or RasGAP C-terminal (RGC) [103, 104]. Instead, these protein domains bind Cdc42 and Rac1 [103-106]. Ras, RhoA, Cdc42 and Rac1 all belong to the Ras superfamily of small GTPases. These signaling proteins regulate diverse biological functions including proliferation, trafficking, migration, adhesion, and cytoskeletal dynamics. Members of the Ras superfamily activate signal cascades when bound to guanosine triphosphate (GTP) and are inactive when bound to guanosine diphosphate (GDP). Typically, a GAP promotes the intrinsic but weak GTP-hydrolyzing function of Ras proteins, causing bound GTP to be converted to GDP, inactivating Ras signaling. Instead, IQGAP1 stabilizes Cdc42 and Rac1 in the active, GTP-bound form by both antagonizing real GAPs from binding, as well as directly inhibiting GTPase activity [103, 105-107].
17
2.3.2.
IQGAP1, the Cytoskeleton and Adhesion
IQGAP1 regulates Cdc42 and Rac1 by stabilizing their active forms and as well as acts as an effector for them by assembling the actin cytoskeleton [103-105, 108-110]. IQGAP1 binds directly to filamentous actin with its single calponin-homology domain (CHD) and can cause actin to polymerize [108, 109, 111, 112]. IQGAP1 also cross-links actin filaments by forming IQGAP1 dimers or oligomers with its unique IQGAP repeat domain and/or IQ domain [108, 111-114]. Additionally, IQGAP1 binds to E-cadherin and β-catenin with its RGC [115, 116]. Cadherins are a group of transmembrane proteins that maintain cell-cell adhesion by forming intercellular, homotypic dimers. The cytoplasmic tails of cadherin are linked to the actin cytoskeleton by α- and βcatenin. The connections are dynamically altered during remodeling, migration, and proliferation. IQGAP1 recruits β-catenin and cadherin to the leading edge, thus strengthening adhesion. IQGAP1 localizes to the cytoplasm, with strong staining co-localizing with cortical actin, adjacent to filopodia, ruffles and lamellipodia, but not stress fibers [98, 103, 104]. Therefore, IQGAP1 links actin cytoskeleton and adherens junctions [104].
However, in the absence of Cdc42 and Rac1 activation, IQGAP1 remains functional. Under these conditions, IQGAP1 weakens the actin cytoskeleton and cadherin-mediated adhesion by dissociating α-catenin from the cadherin-catenin complex. Therefore, IQGAP1 functions ambivalently depending on whether small GTPases are bound or not. This balance between IQGAP1-GTPase and apo-IQGAP1 regulates cell-cell adhesion and cell migration. Similarly, IQGAP1 has also been implicated in the adhesion and turnover of N-cadherin and VEcadherin [117, 118].
IQGAP1 also interacts with other components of the cytoskeleton. IQGAP1 has been demonstrated to interact with myosin components, but the consequences have not been explored [99, 100, 119, 120]. IQGAP1 C terminal interacts with tubulin plus end linker protein, CLIP170, and may function to link actin and microtubule function [121]. IQGAP1 has been demonstrated to be necessary for polarizing granules and orientating the microtubule organizing center in natural killer cells [122]. IQGAP1 also interacts with Dia1 with a region within the
18
GRC [123, 124]. This interaction was important for migration and phagocytosis as IQGAP1 appears necessary for Dia1 localization [123]. Additionally, IQGAP1 cooperates with Dia1 tethering microtubules [124]. Diaphanous proteins are formins, protein regulators of actin and microtubule that are often downstream of Rho GTPases. Silencing both of IQGAP1 and Rab6IP2 impaired microtubule capture by Dia proteins [124]. IQGAP1 also interacts with integrin α4, fibronectin and VCAM1 [125, 126]. Taken together, IQGAP1 can regulate various cytoskeleton-dependent processes including adhesion, migration, cell morphology, trafficking and secretion [127].
IQGAP1’s control over actin and adhesion is evident as these abilities are exploited during invasion and infection by bacteria such as Salmonella, Escherichia coli, Helicobacter pylori, Yersinia pestis as well as the Ebola virus egress [128-136].
2.3.3.
Novel Interactions with Ras-Superfamily Proteins
IQGAP1 binds other small GTPases, Ras subfamily member Rap1 and Arf6. IQGAP1 binds to Rap1 with its IQ-motifs instead of the GRD or RGC domain. IQGAP1 overexpression reduced adhesion-mediated and cAMP-dependent activation of Rap1, suggesting IQGAP1 negatively regulates Rap1 [137]. IQGAP1 binds Arf6 and seems that IQGAP1 is required for Arf6-dependent activation of Rac1 in malignant gliomas stimulated with growth factor [138].
In contrast to initial descriptions, IQGAP1 binds Ras-family member, K-Ras and improves its signaling activity [99, 139]. IQGAP1 binds K-Ras with its IQ-domain and binding is independent of K-Ras activation [104, 105, 139]. IQGAP1 does not bind H-Ras however. The discrepancy may be explained by early studies being unaware of IQGAP1’s anti-GAP function and divergence between H-, K- and N-Ras proteins. IQGAP1 simultaneously binds to Ras and Cdc42, suggesting a mechanism for linking Ras and Cdc42 in a signal cascade [103].
19
2.3.4.
IQGAP1 and Wnt Signal Transduction
IQGAP1 does more than transmitting signals between the cytoskeleton, adhesion complexes and microtubules. In addition to anchoring actin to cadherin, β-catenin transactivates gene expression during Wnt signaling. In the absence of Wnt signal, β-catenin is kept in adheres junctions. Labile β-catenin is phosphorylated by GSK-3β in a complex with axin and adenomatous polyposis coli (APC), which signals β-catenin for ubiquitin-dependent proteasomal degradation [140]. This removal maintains cytoplasmic β-catenin at a minimum. In the presence of Wnt simulation, GSK-3β is sequestered and β-catenin accumulates in the cytoplasm. Cytoplasmic β-catenin then translocates to the nucleus in a Rac1-dependent manner and recruits TCF/LEF family transcription factors and promotes transcription of target genes, such as cell cycle regulators, cyclin D1, and proto-oncogene, MYC [141, 142]. In addition to IQGAP1 regulating β-catenin by affecting the strength of cadherin, IQGAP1 can directly interact with β-catenin and APC, and has been demonstrated to support β-catenin trans-activation [143, 144]. IQGAP1 does this by stabilizing β-catenin in the cytoplasm, preventing proteasomal degradation and improving accumulation in the nucleus [143]. The effect of this was increased proliferation and migration. Additionally, IQGAP1 directly interacts with Dishevelled (DVL)-1,-2 and -3 in the region between the IQ motif and the GRD and mediates nuclear accumulation of DVL protein in Xenopus [145]. Dishevelled inhibits GSK-3β-axin-APC complex from inactivating β-catenin, thus disinhibiting Wnt signaling, as well as acts to stabilize β-catenin-containing trans-activation complex [146].
IQGAP1 also potentiates Wnt–β-catenin signaling through interactions with LGR4 and LRP5/6 [147]. LGR4 is the receptor for ligand, R-spondins. R-spondins through LGR4 disinhibits Wnt receptor expression and causes phosphorylation of Wnt co-receptor, LRP5/6. The LRP5/6 is targeted by MEK1/2 and phospho-LRP is a hallmark of Wnt activation. IQGAP1 binds to LGR4 with its GRD [147]. Upon stimulation IQGAP1 appears to be recruited into a super-complex comprised of DVL, MEK1/2, R-spondins-LGR4 [147]. In the absence of IQGAP1, LRP phosphorylation is reduced [147]. Activation with R-spondins also led increases association with mDia1 and N-WASP [147]. These interactions were necessary for migration and invasion of cancer cells. IQGAP1 appears to link both canonical and non-canonical Wnt signaling by bringing together Rho-GTPases Rac1 and Cdc42, actin, adhesion and various other cytoskeletal
20
components together, as well as directly influencing signaling transduction and permitting βcatenin trans-activation of target genes.
2.3.5.
IQGAP1 and Other Signal Transduction Pathways
IQGAP1 interacts with all three members in the mitogen-activated protein kinase (MAP kinase) cascade, B-Raf, Mek and Erk. Ras signal mediates cell cycle progression through Raf, MEK and MAPKs or ERKs. These signal pathways branch and receive crosstalk from different cellular pathways [148]. IQGAP1 forms complexes with K-Ras and B-Raf [139]. The IQ domain of IQGAP1 directly binds B-Raf and MEK and improves the kinase activity of both [149, 150]. IQGAP1 binds to Erk and increased phosphorylation of target protein, Elk-1[151]. Upon proper stimulation, IQGAP1 thus facilitates the activation of each kinase as well as enhances its phosphorylation function of the subsequent MAP kinase. Based on these findings, IQGAP1 acts as a scaffold protein for the K-Ras, B-Raf, Mek, Erk. To act as a scaffold, IQGAP1 recruits activator and target proteins, bringing them into close proximity and thus reinforcing signal transduction.
IQGAP1 interacts in complex with hyaluronic acid (HA) receptor, CD44 [152]. Stimulating cells with HA increases the IQGAP1-CD44 interaction. IQGAP1 promotes migration downstream of HA-CD44 through Cdc42 and actin. In addition, HA stimulation promotes IQGAP1-Erk2 binding, and phosphorylation and activation of Erk2 [152]. Downstream of HA-CD44, activated Erk2 phosphorylates Elk1 and estrogen receptor α (ERα). Phosphorylated Erk2, Elk1 and ERα target respective transcriptional elements which include cyclin D1. Knockdown of IQGAP1 by siRNA abrogates these downstream effects of HA-CD44 [152]. Since IQGAP1 mediates crosstalk between actin, MAP kinases, estrogen signals, this suggests that IQGAP1 is pivotal for downstream effects of HA-CD44.
Additionally, IQGAP1 interacts with other growth factor receptors. IQGAP1 binds VEGF receptor, VEGFR2 (KDR/Flk-1), and stimulating endothelial cells with VEGF promotes the
21
direct IQGAP1-VEGFR2 binding [153]. VEGF stimulation causes increased Rac1-IQGAP1 interaction as well as IQGAP1 phosphorylation and oxidation [153, 154]. Loss of IQGAP1 by siRNA diminishes effects downstream of VEGF/VEGFR2, including H2O2 production, proliferation, and cell migration [117, 153]. Similarly, IQGAP1 is known to interact with EGF receptor members ErbB1/HER1 and ErbB2/HER2 with its IQ domain [155, 156]. Activation of HER1 by EGF stimulation increases the extent of the interaction by activating PKC, which phosphorylates IQGAP1 [155]. Downstream of EGF and HER1, IQGAP1 acts to enhance signal strength by activating MAP kinases [155]. IQGAP1 also interacts with the ligand-independent EGF receptor, HER2 [156]. Knockdown of IQGAP1 causes reduced activation of AKT, as well as reduced expression and phosphorylation of HER2. Loss of IQGAP1 augmented the effects of anti-cancer biologic, trastuzumab. Trastuzumab-resistant cells have elevated IQGAP1 expression, which confers resistance to the drug [156]. IQGAP1 expression levels did not affect mRNA levels or stability suggesting IQGAP1 regulates HER2 expression by post-translational mechanisms [156].
Conversely IQGAP1 negatively regulates TGFβR. IQGAP1 binds directly to TGFβR2 via its RGC domain [157]. Both TGFβR1 and -2 expression levels correlate inversely to IQGAP1 expression, and shRNA against IQGAP1 results in increased expression of TGFβ-inducible target genes. TGFβ stimulation down-regulates TGFβR2 by increasing IQGAP1-TGFβR2 interaction, which results in internalization and lysosomal degradation [157].
2.3.6.
IQGAP1 in the Cell Cycle and Proliferation
2.3.6.1. IQGAP1 in G1 Phase Based on the interactions described above, it appears that IQGAP1 is strongly linked to proliferation and the cell cycle. IQGAP1 interacts with, targets or regulates many proteins and processes. Many of these are well known tumor suppressors, proto-oncoproteins or are otherwise active during cell cycle and proliferation. Cytoskeletal regulation, trafficking, adhesion, contact and migration are critical processes for a cell to decide and to carry out the cell cycle. Additionally, the aforementioned Ras-superfamily, Wnt pathway, growth factor receptors, MAP
22
kinases all play their respective roles in promoting growth, proliferation and cell cycle progression. Aberrations in Ras-related pathways are extremely common in neoplastic lesions. Ras requires Rho family GTPases, like Cdc42 and Rac1 to produce the transforming phenotypes. Wnt signals and β-catenin are critical in development and stem cell renewal, and this pathway is often dysregulated in colorectal and gastric cancers. In general growth factor receptors and the MAP kinases contribute to initiating events in proliferation. Downstream of these pathways are often oncogenes, cyclin D1 and MYC. These proteins are key regulators of passing through G1. Similarly, IQGAP1 appears to downregulate cyclin dependent kinase inhibitor, p27 [156]. These data suggest that IQGAP1 plays diverse, but overlapping roles in early G1 decisions to initiate cell divisions.
In addition to mitogenic signaling, IQGAP1 regulates growth more directly. IQGAP1 complexes with mTORC and Akt to promote Akt activation, and promotes exocytosis and protein synthesis [158-160].IQGAP1 interacts with the rapamycin-sensitive MTORC1 complex through subunits mTOR and Raptor [159, 161]. These interactions were found to be cell cycle-dependent and were mutually exclusive with Cdc42 binding to IQGAP1. This switch from growthpromoting IQGAP1-MTORC1 to division-promoting IQGAP1-Cdc42 is controlled, at least in part, by IQGAP1 phosphorylation [110, 161, 162]. MTOR interfaces nutrient and growth factors signaling to regulate cell growth and division. IQGAP1 appears to keep the balance between cell growth and division. N-terminal increases size, C-terminal promotes cytokinesis [161]. This link between IQGAP1 and coupling growth/division is present in both S. cerevisiae and mammalian cells [159].
2.3.6.2. IQGAP1 in S Phase IQGAP1 appears to have a role in DNA replication and repair too. When the N-terminal half of IQGAP1 was expressed in fibroblasts localizes to the nucleus when expressed in fibroblasts [115]. Endogenous IQGAP1 accumulates in the nuclei in a minority of cells. IQGAP1 nuclear entry depends on GSK-3β activity and CRM1/XPO1 but is independent of β-catenin nuclear entry. The role of IQGAP1 in the nucleus is not well understood, but loss of IQGAP1 by knockdown
23
caused a delay in cells to progress through S phase. Arresting cells in S-phase by thymidine block or hydroxyurea, induces cells to accumulate nuclear IQGAP1 [163]. In response to DNA replication stress, IQGAP1 and actin translocate to the nucleus with similar timing dynamics, and the amount of nuclear IQGAP1 and actin is related to the degree of replication blockade [164]. Considering that nuclear actin has been demonstrated to be essential for RNA polymerase transcription and RNA export and chromatin remodeling, one might expect IQGAP1 regulates actin in the nucleus as it does in the cytoplasm. Therefore, IQGAP1 may regulate DNA and RNA in the nucleus. Additionally, IQGAP1 was shown to be in complex with PCNA and bind to RPAcoated single strand DNA [163, 165, 166]. IQGAP1 interacts with EEF1E1, a protein known to act in the ATM/ATR activation of p53 during DNA damage [167]. This hints at the possibility that IQGAP1 plays a direct role in DNA synthesis and repair.
2.3.6.3. IQGAP1 in M Phase and Cytokinesis The cytoskeleton is indispensible during M phase and cytokinesis. Microtubules are responsible for aligning and separating chromosomes during mitosis. The actin cytoskeleton and myosin are important in the cell cycle for organelle segregation and cytokinesis. IQGAP1 can regulate microtubules through CLIP-170, actin dynamics and potentially myosin. IQGAP1 regulates exocytosis and secretion, which adds new membrane to the dividing cell [160]. IQGAPs may continue have roles to play during M phase, as there is evidence for mitotic regulation. IQGAP1 changes localization during mitosis in human cells from diffuse cytoplasmic during the interphase to the midzone mitotic spindles and midbody during M phase [159]. During G1, IQGAP1 associates with Rac1 and -integrin, but dissociates from with Rac1–-integrin complex in G2/M phases [168]. This loss dissociation is coupled with a loss in the ability to crosslink actin. Additionally, IQGAP1 contains a ‘destruction box’ motif, RXXL, within its GRD [106]. This motif is recognized by the anaphase promoting ubiquitin ligase complex/cyclosome (APC/C). Although yeast Iqg1p does not have this motif, it is targeted by APC/C [169]
24
IQGAP family proteins are essential for budding in S. cerevisiae and cytokinesis D. discoideum. Cyk1p/Iqg1p, is essential for cell viability in S. cerevisiae and expression of the C-terminal of the IQGAP resulted in a multinucleated cells, consistent of a budding defect [170, 171]. Similarly, IQGAP proteins localize to the bud neck in D. discoideum and overexpression or underexpression produce giant and multinucleated cells in D. discoideum due to a defect in late cytokinesis [172]. Along these lines, Cdc42 was discovered in S. cerevisiae where it plays a role in budding. IQGAP protein in S. cerevisiae, Cyk1p/Iqg1p, is needed to assemble the actin for the actomyosin ring [170, 173]. This is driven by the CHD. The GRD appears to be required for contractility either by binding a Ras-superfamily protein and mediating signaling or through structural organization. [173] Therefore, in lower eukaryotes IQGAP family proteins have essential roles in cytokinesis.
Such findings strongly suggest that IQGAP1 plays some role in proliferation and the cell cycle, but the nature of this role is unclear. IQGAP1 and IQGAP2 knockout mice are both viable, but have increased risk of organ-specific cancer [174]. Also, reduced IQGAP1 in the tumor microenvironment has been found to promote tumor growth, metastasis, and worsens host survival [157]. These phenotypes suggest that IQGAP proteins are tumor suppressors. However, the increased hepatocellular carcinoma risk present in IQGAP2 knockout mice was largely due to a concomitant IQGAP1 increase [175]. Similarly, IQGAP1 is often elevated in several types of in cancerous tissue, and inhibiting IQGAP1 reduces the cancer phenotype [127, 139, 153, 158, 176]. Additionally, IQGAP1 may be potentially useful marker to diagnose H. pylori infection and measuring the malignancy of gastric cancers that the bacteria that causes them. Similarly, IQGAP1 may be diagnostic to malignancy in colon cancer. Uniquely, IQGAP3 appears to be dynamically expressed during development and regeneration [177, 178]. Together, these findings suggest that IQGAPs, and IQGAP1 are oncogenic. This ambivalence as tumor suppressor and oncogene underlines IQGAP1 as multifunctional protein that is context-dependent. IQGAP1 possess multiple mechanisms for stimulating proliferation and regulating proper growth and cell division.
25
2.3.7.
IQGAP1 Regulation by CaM
IQGAP1 is named as such after its IQ domain, which contains four, tandem IQ motifs. IQ motifs are 20-30 amino acid sequences that are able to bind to CaM-like proteins. The presence and avidity of IQ domain would suggest major part of IQGAP1 biology is its interaction with Ca2+binding, CaM-like proteins. Indeed, IQGAP1 binds to Ca2+ sensors S100A1, -B and -P [119, 179, 180]
. S100 proteins form homodimeric complexes with each monomer containing two EF-hands
that are able to bind zinc ions or Ca2+. IQGAP1 interacts with Ca2+-S100 proteins, but not zincbound forms. The consequences of such interactions have not been well explored as members of the S100 family have tissue-limited expression, unlike the ubiquitous Ca2+-binding protein, CaM.
All four IQ motifs in the IQ domain are able to bind CaM [99, 181]. An IQ motif is described as incomplete if the sequence binds CaM only in the presence of Ca2+. The motif is described as complete when a conserved arginine residue is present, which confers the ability to bind CaM in the absence of Ca2+. There are both incomplete and complete IQ motifs in IQGAP1, and consequently, IQGAP1 binds CaM in the presence and absence of Ca2+ [103, 181, 182]. IQ1 and IQ2 appear to only bind Ca2+-CaM, while IQ3 and IQ4 bind both Ca2+-CaM and apoCaM [181]. Furthermore, the sequence of containing IQ3 and IQ4 contains a functional 1-8-14 Ca2+-CaMbinding motif [183]. In addition, IQGAP1 has been reported to bind CaM with its CHD, but the physiological significance is not yet known [184]. IQGAP1 appears to be the major CaM-binding protein in normal breast cells and MCF-7 breast cancer cells under both Ca2+ and EGTA conditions [182].
Normally, available CaM is limiting, and IQGAP1 does not appear to be able to bind four CaM molecules simultaneously [182]. Consequently, CaM was found to be a substoichiometric subunit of IQGAP1 complexes at a ratio of one CaM per seventeen IQGAP1 molecules [98, 111]. CaM binding to IQGAP1 is incompatible with all the other binding partners to IQGAP1 that have been tested [100, 111, 116, 137, 138, 155, 179, 182, 184-187]. CaM exerts both competitive inhibition for overlapping binding sites or allosteric regulation of IQGAP1 molecules, depending on the protein partner. However, the change in affinity differs depending on the binding partner. For instance, Ca2+-
26
CaM prevents cdc42 binding, but apoCaM does not [182]. In contrast, apoCaM, not Ca2+-CaM prevents actin binding [98, 111]. Such findings suggest that CaM is a major, but transient regulator of IQGAP1 biology. Conversely, IQGAP1 appears to direct CaM subcellular localization and knockdown of IQGAP1 makes CaM-signaling more efficient [188]. Thus IQGAP1 is responsible for targeting and sequestering CaM and possibly scaffolding CaM signals. As both proteins appear to regulate proliferation and growth, we expect that CaM and IQGAP1 mutual regulation has consequences that are evident in cell cycle analysis.
IQGAP1 has multiple upstream and downstream binding partners both within and between molecular pathways. Within a pathway, IQGAP1 may function as a scaffold, bringing proteins from different levels of a signal cascade into close proximity, thus making the transduction more efficient. Between pathways, IQGAP1 joins different signal inputs to diverse, but common downstream effectors. However, the numbers of effectors that IQGAP1 can regulate suggests that it is also integrating these different upstream signals and regulates the effectors to create the balance required for a specific biological output. This function allows crosstalk between processes that must occur in conjunction, but do not otherwise overlap molecularly.
Several linkages have been made between IQGAP1 and mitogenic pathways. IQGAP1 is strongly linked to regulation of actin and microtubule cytoskeletons, both of which are important for the adhesion, cargo trafficking, DNA and organ segregation, and cytokinesis. Despite this, neither specific roles nor timing in the cell cycle has been demonstrated for IQGAP1. IQGAP1 has been demonstrated to enter the nucleus during S-phase arrest, but how it does so and what IQGAP1 may be doing in the nucleus is purely speculation. The expression pattern of IQGAP1 mRNA and IQGAP1 protein relative to the cell cycle is not known either. Sorting out the ambivalent nature of IQGAP1 and how IQGAP1 mediates crosstalk between the various pathways it facilitates is poorly understood. We recognize that IQGAP1 may be regulated by Ca2+-bound and Ca2+-free CaM, but the conditions that mediate CaM intervention, how CaM is able to mediate IQGAP1 despite substoichiometric interaction as well as the consequences of this interaction are not clear. It is also unknown the consequences of the interaction of IQGAP1 and apoCaM.
27
2.4. Cardiovascular Disease The cardiovascular system is organ system that comprises of the heart, arteries, veins, capillaries, lungs and blood. The cardiovascular system receives various homeostatic inputs from the nervous system, liver and kidneys. Together, the cardiovascular system is responsible for supplying the other organs and tissues of the body with nutrients and oxygen and removal of waste and carbon dioxide. The cardiovascular system also provides an avenue for the transmission of endocrine signals, the immune system as well as environmental toxins. Therefore, the cardiovascular system is critical to life as it integrates body into a single entity. Perturbation of any component of the cardiovascular system effects and strains the others organs. Thus disparate diseases ultimately lead to common endpoints and are grouped as cardiovascular diseases.
A major contributor to cardiovascular diseases is atherosclerosis (ATH) [189, 190]. ATH is dysfunction in the arteries characterized by hardening of the blood vessel and a narrowing of the vessel lumen. Major blood vessels consist of three layers. Adjacent to the lumen and in contact with the blood is the intima. The intima is a single layer of endothelial cells and the underlying connective tissue. This layer is adapted to withstand the shear of blood flow. Surrounding the intima is the media, which is made of elastic, fibrous extracellular matrices and vascular smooth muscle cells (SMC) that create them. The fibrous matrices give the vessel much of the physical rigidity and elasticity. This gives the vessel strength to withstand the forces exerted on it by blood pressure as well as elastic properties that dampen the pulsatile nature of blood flow. SMC in the vessel are able to contract or relax resulting in vasoconstriction or vasodilatation. Encasing the blood vessel in the outermost layer, the adventitia is made of mostly connective tissue.
Endothelial damage or dysfunction is recognized to initiate the development of ATH [8]. Many risks of ATH directly or indirectly damage the intima. Exposure to high levels of toxins from alcohol consumption, smoking, and toxic metabolites may damage endothelial cells. Additionally lipids and cholesterol may accumulate within and disrupting the integrity of the intima. Furthermore, irregular flow or high blood pressure also contributes to endothelial
28
dysfunction. Injured endothelial cells and exposed intimal matrices initiate inflammation and wound healing. In response immune cells and medial SMC infiltrate the intima. Typically, this process can be resolved and is self limiting. However, chronic exposure or an inability to heal the lesion creates plaques. A plaque is the accumulation of immune cells and SMC and the connective tissue they produce that form between the media and intima. Plaques form into the lumen and may occlude it, restricting blood flow. This is known as stenosis.
Clinically, when an artery narrows and blood flow is restricted, balloon angioplasty may be performed to reopen the vessel. Angioplasty may open the vessel, but also injures the endothelium by mechanical abrasion, as well as wounding the plaque and exposing it to the blood. In response, processes initiate wound healing once more. This may result in growth of the plaque and the vessel may narrow again. Thus the vessel is in stenosis again, or restenosis. As a preventative measure to restenosis, stents were developed. Stents are metal meshes inserted into the vessel to give structural support and keep the vessel open. While helpful, bare-metal stents do not eliminate restenosis. Consequently, drug-eluting stents have improved clinical outcomes from restenosis compared to bare-metal stents [1-4]. Drug-eluting stents appear to function by blocking proliferation of SMC. However, this also prevents proliferation of endothelial cells and thus endothelialization around the stent. As a stent is foreign object in the body, stents pose a risk for developing thrombi. Not only do drug-eluting stents not overcome issues for thrombosis formation, they may delay arterial healing and prolong thrombotic risk [3, 191]. Preventing restenosis still is a major area of research within the cardiovascular field [3].
2.4.1.
SMC Proliferation Contributes to Atherosclerosis and Restenosis
SMC play critical role in the development of ATH. SMC remain at rest, but are capable of restarting the cell cycle and proliferating as needed. They are sensitive to signals from PDGF, bFGF and IL-1. In response to these signals, they migrate into the intima, proliferate and deposit extracellular matrix. Formation of the neointima and its maturation are essential to recovery from wound healing. However, prolonged action by SMC also contributes to ATH. Particularly, the
29
dedifferentiation and proliferation of SMC appear most problematic [8, 190, 192, 193]. The proliferation of SMC in ATH neointima have been noted for their clonal nature and described as similar to a neoplasm [8].
The success of drug-eluting stents further illustrates involving SMC proliferation pathological. Sirolimus (rapamycin)-eluting stents have proven effective in preventing restenosis. Rapamycin is an anti-fungal, immunosuppressant and anti-proliferative compound. In humans, rapamycin acts on the mTOR, blocking protein synthesis, cell growth and entry into G1. It is theorized that rapamycin acts as an anti-proliferation agent, and prevents SMC from entering the cell cycle [5, 6]. Similarly, paclitaxel-coated stents also effectively prevent restenosis. Anti-cancer agent, paclitaxel stabilizes assembled microtubules preventing depolyermization. Paclitaxel is a mitotic inhibitor that blocks SMC proliferation and migration [5-7]. Similar strategies of manipulating the cell cycle through chemical means may improve the outcomes of ATH and cardiovascular disease.
30
Chapter 3 Hypothesis
3.1. Rationale Ca2+ and CaM are well known positive regulators for progression through the cell cycle. Both Ca2+ and CaM are necessary at several points during the cell cycle and reducing of either one result in decreased or stalled cell cycle progression. As such, there are multiple processes that Ca2+ and CaM participate in that promote proliferation.
IQGAP1 is known to interact with many pathways for controlling processes in the cell cycle, such as cytoskeletal dynamics, adhesion and trafficking. IQGAP1 interacts with both growth factor pathways and MAP kinase cascade to enhance their signals. IQGAP1 is also linked to DNA repair and Ras-family proteins. These interactions suggest that IQGAP1 has some involvement in the cell cycle either positively or negatively. Increased tumors incidence in IQGAP1 knockout animals suggests IQGAP1 is tumor suppressor. However, analysis of cancer cells suggest that IQGAP1 oncogenic. Together, evidence would seem IQGAP1 is ambivalent in the cell cycle, both positively or negatively depending on conditions and the phase of the cell cycle. Thus IQGAP1 poses as an interesting potential therapeutic target as it appears to be a central node for networks of protein pathways. Additionally, IQGAP1 seems to be strongly controlled by CaM interactions. Thus some if not most of these IQGAP1-mediated processes are subject to CaM regulation.
How the CaM-IQGAP1 interaction regulates the cell cycle is not known. CaM and IQGAP1 are both expressed throughout tissues in the body, as well as in diverse species. We expect an interaction to occur in almost all mammalian cell types, including human SMC. A cell cycledependent interaction is expected to occur as both proteins are present and are implicated in regulating the cell cycle. Additionally, preliminary work shows differential interaction of IQGAP1 with CaM depending on whether the cells are serum starved or serum stimulated. Since
31
Ca2+-CaM is a pro-proliferative species, and IQGAP1 is clinically oncogenic, the CaM-IQGAP1 interaction is expected to enhance cell cycle progression.
Our goal for this work is to gain insight into the cell cycle, in particular of SMC. If indeed SMC use CaM-IQGAP1 interaction in cell cycle progression, this interaction could be manipulated to alter the cells ability to proliferate. Pharmacological intervention that effects this interaction may be used as an alternative method for treating restenosis.
3.2. Objectives Based on our hypothesis, it is expected that both IQGAP1 and CaM are present and expressed in SMC. Due to the multitude of functions that IQGAP1 and CaM possess within the cell, the levels of protein expression are not expected to vary over the course of cell cycle. Despite the relatively constant expression of both proteins, it is expected that there will be a variation in the ability for IQGAP1 and CaM to interact. This interaction should correlate with the proliferative state of the SMC.
Since proteins with invariable levels of expression are expected to have variable interaction, there must be some sort of upstream regulation. Possible regulations may include posttranslational modifications of CaM and/or IQGAP1, adapter proteins that improve complex stability, loss of inhibitors, changes in protein localization, or sensitivity to Ca2+ concentrations.
Despite IQGAP1 being able to bind CaM under Ca2+ and Ca2+-free conditions, it is expected that Ca2+ will be relevant to the interaction. As Ca2+, CaM and IQGAP1 have each been demonstrated to increase proliferative potential by themselves; we expected that Ca2+ improves the CaMIQGAP1 interaction. This Ca2+-CaM–IQGAP1 species is expected to exist in greater abundance proliferating cells than in quiescent cells. If the interaction of IQGAP1 and CaM is inhibited or
32
enhanced, the ability for cells to progress through the cell cycle will be attenuated or augmented, respectively.
3.3. Summary Specifically, our hypothesis is that SMC express CaM and IQGAP1. When cells are proliferating, CaM and IQGAP1 interact in a cell cycle-dependent manner. This interaction is expected to require the presence of Ca2+. We do not expect presence or absence of CaMIQGAP1 complex to dependent on changes of protein expression. Instead, we predict posttranslational regulation to control the CaM-IQGAP1 interaction. The interaction of CaM and IQGAP1 predicted to promote passage through the cell cycle. The pro-proliferative nature of CaM-IQGAP1 is expected to be relevant in different tissues and cell types. In order to test our hypothesis fully, the following need to be demonstrated: 1. Detection SMC expression of CaM protein. 2. Detection SMC expression of IQGAP1 protein. 3. Investigate the protein levels of both CaM and IQGAP1 for cell-cycle dependency. 4. Determine whether an interaction between CaM and IQGAP1 occurs during the cell cycle. 5. Determine these CaM-IQGAP1 complexes require Ca2+. 6. Identify changes in regulation or modifications to CaM and/or IQGAP1. 7. Determine the effect CaM-IQGAP1 interaction on the proliferative ability of SMC. 8. Determine the mechanistic consequences of cell cycle CaM-IQGAP1 complexes. 9. Extend finding to other cell types. 10. Identify means for interfering with the CaM-IQGAP1 interaction that may be used clinically.
33
Chapter 4 Experimental Approach
4.1. Immunoblotting CaM and IQGAP1 are the protein products of their respective well-conserved genes. However, neither protein has known enzymatic or transcriptional activity. Also, the functional consequences of both proteins can affect multiple downstream signaling pathways. Many of these pathways are cell cycle-dependent or cell cycle regulators. Thus we can not select one to use for measuring experimental outputs without more information about the CaM-IQGAP1 interaction. Therefore, enzymatic conversion assays and transcription reporters are not useful for this study. Additionally, changes in transcriptional regulation are not expected to be informative for determining level of protein activity as both proteins are ubiquitously and constitutively expressed. Determining protein expression by measuring transcription is further complicated due to the fact that CaM is the product of several loci. Due to these limitations, CaM and IQGAP1 proteins must be measured directly in order to conduct experiments necessary to answer the hypothesis. Immunodetection allows for specific quantification of CaM and IQGAP1 by using antibodies raised against these protein fragments. In immunoblots (IB), proteins from cell extracts are separated by electrophoresis and protein analytically measured.
4.1.1.
Immunoblotting CaM
CaM has physical properties that make it harder to IB than other proteins. Generally, when performing one-dimensional electrophoresis, proteins are linearized by the combined effects of reducing disulfide bonds and detergents stabilizing exposed hydrophobic regions. Therefore the only physical property of consequence is molecular weight of individual proteins. Subsequently, the unfolded proteins are immobilized to a membrane to be probed with antibodies. CaM sequence contains no cysteine residues and thus CaM lacks cystine disulfide bonds [35, 194, 195]. Consequently, its three-dimensional structure is maintained entirely by the primary and secondary structure, and thus is unaffected by reducing agents. This is evident as engineered domains of CaM are able to take conformations approximate to the crystal structure of full length protein [196]. Additionally, CaM is able to bind Ca2+ in the presence of SDS [197, 198]. This binding
34
increases the α-helical content of CaM, making more compact and mobile during electrophoresis [30, 36]. Therefore the structure of this hydrophilic protein is tolerant of both reducing agents and detergent, and is not linearized during electrophoresis.
In addition to maintaining a compact and soluble conformation, its low molecular weight and acidic nature make CaM rapidly and efficiently transfer out of gels [199, 200]. However, these same properties make CaM adsorption to membranes poor. As a result, CaM does not adsorb to membranes well, nor does CaM remain adsorbed during the necessary, subsequent incubations [199]. As explained above, the lack of enzymatic or transcriptional activity limits the options to measure CaM experimentally, and these IB difficulties needed to be overcome.
Despite several scientific publications and manufacturer product information sheets seemingly neglecting these facts in their methods sections, possible solutions to overcoming these difficulties have been published [200-202]. However, as no protocol existed in our lab previously. Therefore one needed to be developed in order to have a method of reliable and repeatable detection.
4.2. Protein Binding Assays CaM and IQGAP1 lack enzymatic or transcriptional activity of their own. Instead, the biological effects of each protein are mediated by binding to various and diverse partners, and altering the function of these downstream effectors. The interaction and biochemical consequences of CaM binding IQGAP1 has been documented numerous times, but the specific biological conditions for this interaction have not been fully elucidated. Our hypothesis predicts variable interaction between these two proteins, which is dependent on the cell cycle. Additionally, as Ca2+ plays such a major role in the function of CaM and in the cell cycle, experiments must test the effects of Ca2+ on the interaction.
35
There are various methods that evaluate protein-protein interactions. The method chosen was reciprocal co-immunoprecipitation (IP) of CaM and IQGAP1. Additionally, purification of IQGAP1 using immobilized CaM was attempted. Expression plasmids of tagged CaM and IQGAP1 were received with the intent of developing fluorescent resonance energy transfer system, and potentially, combining the technique with the cell cycle reporter, Fucci (see § 4.3.2).
4.2.1.
Co-immunoprecipitation
In order to test our hypothesis that the interaction of CaM and IQGAP1 regulates the cell cycle, we need to be able to demonstrate that there is protein-protein interaction between CaM and IQGAP1. Protein-protein interactions are classically demonstrated by co-IP. By exploiting the highly specific reaction of antibodies with their epitopes target proteins, a protein of interest can be purified along with any proteins it may be complexed with during the reaction. Generally, IP is able to identify endogenous proteins as well as over-expressed or purified proteins. Being able to detect endogenous proteins out of the milieu of cellular proteins increases the confidence that the identified binding occurs in nature. The pull-down can be performed reciprocally, to strengthen the argument.
As binding reactions occur in vitro, co-IP experiments use protein lysates. This removes one form of protein regulation, localization. Interactions that would not occur inside intact cells because the proteins would not co-localize are made possible in the homogenized system. Also, proteins that highly co-localize are diluted by homogenization, potentially reducing detection. During the incubations, the extracted proteins are in native conformations. This allows the protein-protein interaction to occur. If the proteins are unfolded, the proteins may not interact. If the proteins are properly folded, the antibodies used to pull-down the complexes must react to an epitope that is present and exposed in the native protein. Additionally, antibody-antigen binding can not be inhibited by the complex, either by competing for the binding site or by a change in protein conformations. The buffer during the experiment must also be optimized to allow for antibody binding to its epitope as well as sufficiently re-creating the intracellular environment.
36
However, for our needs, the artificial conditions allow for manipulation of Ca2+ concentrations by using calcium salts and Ca2+ chelators.
This technique is particularly useful as it can be combined with IB. After the proteins complexes are purified, individual proteins can be identified by electrophoresis and mass spectrometry, or IB.
4.2.2.
CaM-Sepharose Pull-down
Historically, the privilege of having high-affinity antibodies against a given protein was rare. Thus the problem of using antibodies in a co-IP had to be circumvented. Affinity chromatography is a method for purifying proteins based on binding immobilized ligand. The ligand is immobilized to a solid phase and protein homogenate is added. Proteins that do not bind are removed by preliminary elutions. Bound proteins are retained, adsorbed to the solid substrate and may be collected by subsequent elutions. This technique can be performed in traditional columns or in a reaction set up similar to an IP. When mimicking an IP reaction, the ligand is on solid beads and is added to a greater volume of protein lysate with mixing. When performed this way, adequate antibodies are not necessary for complex pull-down. Instead, specificity is driven by the ability to bind to the ligand. This requires that the ligand is isolated to relative purity and that the ligand is fixed to the beads in a conformation that allows binding to that target. Additionally, one ligand may have several binding proteins. This technique can isolate all proteins that can bind to the ligand under the given conditions, not just that target of interest. As this technique depends on binding, the reaction solution could potentially interfere and must be optimized. Also, the target proteins must be in a conformation that allows recognition of the ligand. In many respects, affinity chromatography is similar to co-IP as they both depend on proteins binding to ligand. Affinity chromatography simplifies the process by chemically immobilizing the ligand to the beads while co-IP relies on additional antibody-ligand binding.
37
CaM has been readily isolated to purity [35, 194, 195]. Subsequently, CaM has been used for affinity chromatography by immobilizing it to cyanogen bromide (CNBr)-activated agarose beads to purify binding partners, to determine the cause and effect of this binding, and to identify additional CaM binding proteins [203-205]. Coupling proteins to agarose by CNBr is gentle to the protein, maximizing binding activity [206]. This is now available commercially as CaM Sepharose 4B (GE Healthcare Life Sciences). Due to the nature of CaM, buffer conditions should be manipulated to test the requirement of Ca2+ in binding. Adequate controls are required to disprove that the binding under either Ca2+ or Ca2+-free conditions are due to CaM binding to proteins from the homogenate, and not non-specific adsorption or retention by the structure of the agarose beads.
Currently, there is no CaM Sepharose 4B equivalent for IQGAP1 commercially available. CNBractivated Sepharose 4B is available. Therefore it would be possible to couple isolated IQGAP1 protein to create IQGAP1-conjugated beads. However, neither partial nor full-length IQGAP1 are commercially available. The coding sequence of IQGAP1 is large, thus making recombinant IQGAP1 a difficult prospect.
4.2.3.
FRET
Förster resonance energy transfer (FRET) is the mechanism that describes the phenomenon where a chromophore in an excited state is able to transfer or donate its energy to another, acceptor chromophore. The efficiency of this process rapidly decreases as the distance between the chromophores increases, which limits FRET to occurring only when the FRET pair is in close proximity. By using fluorescence imaging, computational analysis, one protein tagged with a FRET donor and another tagged with a FRET acceptor, we can quantifiably demonstrate that the two targets of interest are interacting. This method of demonstrating protein-protein interaction has the advantage of being able to detect the interaction in vivo, in transfected or chimeric cells. This eliminates the issue of proteins unfolding during isolation or antibody binding. Also, reactions occur in cytosol, which better represents living biology better than synthetic, lysis/reaction buffers.
38
Despite these points, it can be argued that FRET detection of protein-protein interactions does not re-create normal biology. First, the proteins are usually over-expressed by constitutively active promoters. This eliminates all regulation of protein expression, and thus regulation of protein-protein interaction that may occur prior to protein translation. This artificial expression pattern may lead to altered localization and aberrant biology. Second, the proteins are tagged with fluorescent proteins. These chimeric proteins may have altered localization due to altered placement of localization signals. The fluorescent proteins may sterically inhibit protein binding. Third, while FRET occurs in the intracellular environment of living cells, the cells can be influenced with chemicals, drugs and other compounds present in the culture media. This is relevant as this project deals with Ca2+ dependency. The cells may be treated with Ca2+ ionophores to increase intracellular Ca2+ concentrations, or Ca2+-free media or Ca2+ chelators to reduce the Ca2+ availability.
We were able to receive expression plasmids containing CaM and IQGAP1 tagged with modified cyan and yellow fluorescent proteins (CFP and YFP), respectively. CFP and YFP make a FRET pair, with CFP being the donor and YFP being the acceptor. However, neither had been published. Due to this, a cautious approach was taken and the plasmids were sequenced prior to use to determine plasmid structure and detect if mutations were present. Plasmids also needed to be transiently transfected into the cells of choice to get expression. Since FRET signal is based on optical detection, CFP and YFP expression needs to be relatively bright and the two signals need to give equivalent intensities. While Cerulean and Venus are bright variants of CFP and YFP, respectively [207, 208], the given intensities need to be matched by dosing the amount of each plasmid transfected. Microscopy conditions needed to be optimized as this was the first attempt at live cell imaging in the lab.
Additionally, as the goal was to combine the FRET detection of proteins binding with the cell cycle reporting system, Fucci, the two systems were combined to test the plausibility of combining the FRET pair with the red and green of Fucci (see § 4.3.2).
39
4.3. Cell Cycle Analysis Normally, a given population of cells will exist as a mixture of cells in different phases of the cell cycle, as each cell enters and progresses through the cycle at its own pace. But in order to study the cell cycle, we need to know where the cells are in the cycle. This can be overcome by arresting the cells, purifying by phase or single-cell analysis.
4.3.1.
Synchronization by Serum Starvation/Stimulation
For a cell to progress through the cell cycle, various steps must be completed. If a step is inhibited, then the cell stalls or arrests its proliferation, waiting for the problem to resolve. If the cell is beyond the inhibited step, it will continue until encountering said step in the subsequent round of the cell cycle. Therefore, when cells are arrested, it creates a phase-enriched population. There are a few ways to arrest the cell cycle. One of the easiest is to remove serum from the culture medium. Serum contains growth factors that maintain a viable, proliferating cell population. Most cells require external growth factors to provide the necessary signals to allow growth and cell division. As growth is an early process, occurring after M and within G1, removing growth factors arrests cells in G0/G1. Cells that are in S, G2 or M complete the cycle, as these phases do not require growth or growth factors to be completed. This population of G0/G1 arrested cells can be induced to re-enter the cell cycle by adding serum back to the cell. As the cells begin the cycle from the same phase, they enter subsequent phases as a group, and are thus synchronized compared to the asynchronous starting material.
While a simple technique, it needs to be optimized. Stimulating conditions should be a given as they are based on the growth conditions for normal cell culturing and passaging. However, the duration and severity of starvation needs to be taken into account to produce full arrest. Also, a highly synchronized population of cells needs to be produced upon release from arrest. The cells need to be in exponential growth, or logarithmic phase of growth. In this phase, the cells have the greatest potential to proliferate and thus should re-enter the cell cycle efficiently.
40
Synchronization is limited in usefulness as the synchronization often only lasts one phase. This makes it difficult to separate a cyclic event from a response to stimulation and removal of the stress of arrest that rises and falls with time. Additionally, treating with serum has a compounded effect of stimulating the serum response.
4.3.1.1. MOVAS In order to perform serum starvation, IP and IB, we needed a cell type to be a ready source of synchronized protein. In particular, for our interests, we needed a SMC type in order to mimic the biology of plaque formation and restenosis. For these reasons, we chose MOVAS, a large T antigen (TAg)-immortalized, murine, aortic smooth muscle cell-line created in this lab [209]. Simple culture conditions as well as rapid proliferation make it a suitable choice for use. Collecting cells for living animals to study the cell cycle would require induction of woundhealing or ex vivo manipulation, as most somatic cells including SMC are quiescent. This minimizes the superiority that in vivo work normally affords. Species differences are expected to be minimal as Ca2+ metabolism, cell cycle biology, and proteins of interest are well conserved between eukaryotes including mouse and human. CaM and IQGAP1 are also expressed ubiquitously. These facts imply that variations between SMC based on embryonic origin, which are often relevant [210], probably are minimal.
MOVAS are immortalized with TAg. As we are studying the cell cycle, this presents a possible limitation. TAg is a viral protein that sequesters retinoblastoma protein, and p53, hastening progression from G1 to S phase. If CaM and IQGAP1 biology is dependent on either process, then they will not behave normally. Also, TAg alters control of gene expression. Thus, patterns of CaM and IQGAP1 expression and interactions that are regulated by gene expression may also be altered. It has been demonstrated that immortalization results in cells that have the ability to proliferate at lower Ca2+ concentrations than normal cells. Transformation phenotypes positively correlate with increased CaM protein content. This may be overcome by additional experiments in other cell lines, alternate methods of immortalization or use of primary cells.
41
4.3.2.
Fucci System
The cell cycle is a complex process that involves various steps and changes to the cell that must occur before the cell can divide. These changes can be observed and detected by various methods. Such measurements and reporter systems can be used to determine the phase that individual cells are in and used for cytometry, cell sorting or individual cell analysis.
One of the newer reporter systems is fluorescent, ubiquitination-based cell cycle indicator (Fucci). Developed by Sakaue-Sawano et al., this system takes advantage of the cell cycledependent, periodic, reciprocal expression and ubiquitin ligase-mediated proteolysis of two gene products to visually indicate cell cycle phase [211]. The authors created a nuclear, red-fluorescent, G1 marker by creating a fusion protein with a protein fragment from Cdt1 [211]. Normally Cdt1 is expressed in the nuclei of cells until G1/S phase transition. At this time, E3 ubiquitin ligase complex, SCFSkp2, targets Cdt1 for proteasomal destruction. They ensured that the product of their construct expressed appropriately, both spatially and temporally [211]. They also ensured that the fusion proteins did not alter normal cell cycle progression.
Similarly, they created a second fusion protein to mark the remainder of the cycle, S/G2/M. This chimera is a GFP fused to a fragment of Geminin [211]. Geminin accumulates during S and remains until M phase when it is sent to the proteasome by a second ubiquitin ligase complex, the APC/C. The APC/C also targets Skp2 for proteolysis, inhibiting SCFSkp2 activity, promoting Cdt1 accumulation in G1. Conversely, a positive feedback involving SCFSkp2 action inhibits the APC/C, allowing geminin to accumulate. Together, SCFSkp2 and geminin reduce Cdt1 levels in S/G2/M.
The authors took advantage of this mutual inhibition of the ubiquitin ligases and the antiphase expression patterns of Cdt1 and geminin to elegantly mark the nuclei of cells in G1 red, and cells
42
in S/G2/M with green nuclei. During the G1/S transition, both colors are present, producing yellow if overlaid [211]. During the M/G1 transition, newly formed nuclei have faint or no coloration [211]. The Fucci system can be used to perform fluorescence-activated cell sorting to isolate populations of cells in different phases, imaged under a fluorescent microscope to observe the phases of individual cells or other, similar optical methods.
4.3.2.1. Fucci Mouse and Cell Line Sakaue-Sawano et al. took their Fucci system further to create a whole mouse expressing both constructs in somatic cell nuclei [211]. Our lab received this mouse and isolated carotid smooth muscle cells and immortalized them with TAg. This creates an in vitro model that can be transfected with CaM and IQGAP1 fusion proteins. The intention was to perform FRET using the fusion proteins and Fucci to detect CaM-IQGAP1 interaction as well as have a simple method for determining cell phase. This could alleviate the need for synchronization techniques. If synchronization was used, it would provide information about the cell phase, not merely the time since the addition of serum. We have not performed live-cell imaging on the Fucci samples. Thus optical optimization needed to be done. Additionally, the multicolor detection could be troubling with the red and green of Fucci, as well as the cyan and yellow FRET pair potentially overlapping in excitation and emissions wavelengths (Figure 21). With respect to this, SakaueSawano et al. were able to demonstrate combination of Fucci with intramolecular FRET sensor, Raichu-Ras [211]. The Raichu system uses CFP and YFP for its FRET pair, but is also includes a farnesyl moiety [212], limiting Raichu-Ras to the cytoplasm and cell membrane, away from the Fucci-stained nucleus. Normally, IQGAP1 is normally described as a cytoplasmic protein, concentrated at cell cortex and adhesion points. However, IQGAP1 is known to localize in the nucleus [115, 163, 164]. CaM is localized in both cytosolic and nuclear fractions.
43
Chapter 5 Materials and Methods
5.1. Cell Lines and Cell Culture MOVAS are aortic smooth muscle cells from adult C57BL/6 M. musculus that were transformed with SV40 TAg and selected by G418 resistance [209]. NIH 3T3 are fibroblasts from embryonic, male Swiss M. musculus that spontaneously immortalized during line establishment [213, 214]. HEK293T is an SV40 TAg expressing derivative of female, human embryonic kidney cells that were transformed by integration of human adenovirus type 5 DNA [215-218].
Cells were grown in complete media—Dulbecco’s modified Eagle’s medium supplemented with 10 % FBS (Wisent, Inc.) and 100 IU/ml penicillin and 100 µg/ml streptomycin (Wisent, Inc.)— at 37 °C, 5 % CO2.
5.2. Serum Stimulation MOVAS were seeded at 2 × 104 cells/cm2 and allowed to recover from passage O/N. The media was exchanged with starvation media—complete media lacking FBS—to cause growth arrest. Following two days serum starvation, cells were stimulated by adding fresh complete media. Cells were collected at the indicated time points as described in the following section.
5.3. Protein Extraction and Quantification Cells were washed twice in PBS, lysed in buffer A—50 mM Tris (pH 7.4), 150 mM NaCl, 1 % Triton X-100, 1 × protease inhibitor cocktail (Roche), 1 × phosphatase inhibitor cocktail 2 (Sigma) —and stored at −80 °C until use. Lysates were centrifuged at 15 000 × g for 5 min at 4 °C. Supernatants were quantified by Bradford assay (Sigma) against albumin from bovine serum (Sigma) as the standard. Absorption at 959 nm was measured by µQuant mircoplate spectrophotometer (Bio-Tek Instruments, Inc.).
44
Initial experiments used a different buffer, containing 50 mM Tris (pH 7.4), 250 mM NaCl, 5 mM EDTA, 0.1 % Igepal CA-630, 10 % glycerol, 0.1 mM DTT, 0.1 mM Na3VO4, 2 mM PMSF, 1 % phosphatase inhibitor cocktail 2 (Sigma), complete protease inhibitor (Roche). This buffer was commonly used in the lab prior for similar work. This buffer was deemed inappropriate for this project as the EDTA it contains may interfere with the Ca2+-depended binding that is being investigated.
5.4. Immunoprecipitation 300 µg of protein lysate were incubated with 2 µg of antibody in a total volume of 1 ml binding buffer—buffer A supplemented with 2 mM CaCl2, EGTA, or no additions as appropriate — rotating at 4 °C overnight (O/N). The antibodies used were rabbit anti-CaM (sc-5537, Santa Cruz Biotechnology, Inc), rabbit anti-IQGAP1 (sc-10792, Santa Cruz Biotechnology, Inc.) or normal rabbit IgG (sc-2027, Santa Cruz Biotechnology, Inc.). For each reaction, 50 µl of protein G agarose (Pierce) were washed thrice in matching binding buffer. The reactions were added to the resins and rotated for 2 h at 4 °C. Immune complexes were precipitated by centrifugation at 10 000 × g for 5 min at 4 °C. The immune complexes were washed thrice in matching binding buffer to remove non-specific bound proteins. After the final wash, bound proteins were isolated as above.
5.5. Immunoblotting Equal masses of protein were separated by SDS-PAGE on either 4–12 % Bis-Tris gels or 3–8 % Tris-acetate gels (Life Technologies) and transferred to PVDF (PerkinElmer, Inc.) according to the Life Technologies’ instructions. Membranes were blocked with 5 % skim milk in 0.1 % Tween 20, Tris-buffered saline for 1 h at RT.
45
Blots were incubated with antibodies at 4 °C O/N as follows: 1:10 000 rabbit anti-IQGAP1 (sc10792, Santa Cruz Biotechnology, Inc.), 1:400 rabbit anti-GAPDH (sc-25778, Santa Cruz Biotechnology, Inc.), 1:1 000 rabbit anti-p44/42 MAPK (137F5, 4695, Cell Signaling Technology), 1:500 mouse anti-CaMKII delta (1A8, H00000817-M02, Novus Biologicals), 1:5 000 rabbit anti-COX IV (4844, Cell Signaling Technology), 1:100 goat anti-OGG1/2 (sc-12075, Santa Cruz Biotechnology, Inc.). Antibodies were diluted in blocking buffer, except for anti-p44/42 MAPK which was diluted in 5 % BSA, as per manufacturer’s instructions.
The blots were washed thrice with 0.05 % Tween 20 in Tris-buffered saline for 5 min. Blots probed with rabbit 1º antibodies were incubated in 1:15 000 IRDye 680-conjugated donkey antirabbit IgG H + L (926-32223, LI-COR Biosciences) diluted in blocking buffer for 1 h. Blots were washed as above. Signal was detected using the Odyssey Imager and analyzed by Image Studio v2.0 (both from LI-COR Biosciences). For anti-CaMKIIδ antibody-probed blots, signal was detected with 1:10 000 HRP-conjugated goat anti-mouse IgG H + L (170-6516, Bio-Rad), Western Lightning-Plus ECL (PerkinElmer, Inc.) and X OMAT Blue X-ray film (Sigma). Densitometry was performed using GS-800 Calibrated Densitometer and Quantity One v 4.6.1 (both from Bio-Rad).
5.5.1.
Immunoblotting CaM
Gels were transferred to PVDF at 20 V O/N. Proteins were fixed on the membrane with glutaraldehyde for 45 min at RT. The membranes were blocked for 1 h at 37 °C. The blots were washed thrice with 0.05 % Tween 20 in TBS for 10 min then incubated with 1:2 000 mouse antiCaM (clone 6D4, 208696-200UL, Calbiochem) diluted in Tris-buffer saline for 1 h at 37 °C. Blots were washed again and incubated with HRP-conjugated goat anti-mouse IgG for 1 h at 37 °C. Blots were washed again and signal was detected with Western Lightning-Plus ECL and X OMAT Blue X-ray film.
46
As a positive control of CaM detection, full-length rhCaM (CALM2) from E. coli (Novus Biologicals) was used.
5.6. Coomassie Blue Staining Blots were air dried after proteins were electrotransferred to PVDF. Blots were stained with SimplyBlue SafeStain (Life Technologies) for < 2 min, and then washed thrice in water for 1 min.
5.7. Plasmids As an alternative method to the biochemical techniques of determining CaM-IQGAP1 interaction, expression plasmids containing CaM or IQGAP1 genes tagged with fluorescent protein were used.
5.7.1.
pTriEx-3 hCALM1-Cerulean
Full-length, human CALM1 cDNA was fused to 5′ end of the brighter eCFP variant, Cerulean. This construct was inserted into pTriEx-3 plasmid (Novagen, EMD Biosciences) using NcoI and XhoI restriction digests. This plasmid produces CaM-Cerulean (CaM-CFP) protein. This was a gift from Seema Nagaraj and Kevin Truong from Kevin Truong’s lab.
5.7.2.
pEGFP-C1 Venus-mIqgap1
The eGFP gene from the pEGFP-C1 plasmid (Clontech) was swapped out in favor for the brighter YFP variant, Venus. Full-length mouse Iqgap1 cDNA was inserted into the multicloning site of the modified plasmid using HindIII and SalI restriction digests. Iqgap1 was
47
inserted so that it is 3′ of Venus. This plasmid produces Venus-IQGAP1 (YFP-IQGAP1) protein. This was a gift from Matthew J Smith and Mitsuhiko Ikura from Mitsuhiko Ikura’s lab.
Human and mouse homologs of CaM and IQGAP1 have 100 % identical amino acid sequences, so mismatching the species of origin of the expression plasmids is not expected to hinder their interaction.
5.7.3.
Plasmid Verification
Both constructs were transformed into XL10-Gold Ultracompetent cells (Stratagene, Agilent) following the manufacture’s instructions, except SOC medium was used in place of NZY+ medium. Selection was performed using ampicillin- or kanamycin-LB agar plates and broth for pCALM1-Cerulean and pVenus-IQGAP1, respectively. Plasmids were purified from the bacteria using NucleoBond PC 500 plasmid DNA purification kit (Macherey-Nagel).
The plasmids were also sent to be sequenced by The Centre for Applied Genomics (Sick Kids, Toronto) using the primers listed in Table 1. The resulting sequences were aligned with the sequences for the genes and plasmids using Nucleotide BLAST (NCBI).
5.8. Transfection 5.8.1.
Nucleofection
MOVAS were transfected with plasmids using the Amaxa Basic Nucleofector Kit for primary SMC (Lonza) following manufacturer’s instructions, program A-033. Cells were fixed 24 and 48 h post-transfection with 4% paraformaldehyde. Fixed cells were stored in phosphate-buffered saline at −20 °C. Cells were permeabilized with 0.25 % Triton X-100 in PBS, counterstained with Hoechst and mounted in 50:50 glycerol:water. Cells were visualized under FluoView 1000 laser scanning confocal microscope (Olympus).
48
5.8.2.
X-fection
To confirm the fidelity of the constructs, they were transfected into HEK293T cells using the X-fect Transfection Reagent (Clontech) and fluorescent signal was visualized after 19 h of transfection by live cell fluorescent microscopy. Untransfected cells served as the negative control and pmaxGFP (Lonza) served as the positive control. Cells were visualized by live cell imaging in media.
49
Chapter 6 Results
6.1. Optimizing Seeding for Protein Yield and Cell Recovery from Serum Starvation/Stimulation Initial forays to test our hypothesis yielded poor results that could not be meaningfully interpreted (not shown). This suggested that the experimental conditions were not sufficiently optimized. The first issue identified was insufficient protein concentrations in lysates. This was critical to address as future experiments would include IP reactions. IP require large amounts of protein. In order to increase protein yield, the cell number present during lysing needed to be increased. This could be done by increasing the surface area and media during cell culture or the cell density. Increasing culture volume could be wasteful and unnecessarily cumbersome. The other solution would be to increase the cell density during experiments.
The rate of cell growth differs depending on the density of the cell population or confluency. So increasing the cell density may alter the proliferative state of the cell during the arresting protocol. For cell synchronization, cells should have exponential growth when arrested and stimulated to have efficient reentry into the cell cycle. In order to increase protein yield while maintaining experimentally necessary conditions for cell growth, the profiles for the cells under the serum starvation and stimulation needed to be determined.
Cells were grown under standard growth conditions, then trypsinized and collected. The cells were counted and seeded at varying initial cell densities as indicated (Figure 3). These cells were allowed to recover for 24 h and were serum starved for two days, followed by one day of stimulation in growth media. Each day of the protocol cells were trypsinized and counted by hemocytometer. At all seeding densities, the cells recover well from passage, as all cases demonstrate positive slope. Upon two days of serum starvation, proliferation slows. Two days are required as cell numbers still increase during the first day of starvation. For plates seeded at 1.0 × 104 cells/cm2, the decrease in proliferation is accompanied by cell death, evident by the
50
decrease in cell numbers. Upon addition of serum, cells were stimulated to proliferate again. Thus the protocol for arresting proliferation however, cell numbers needed to be optimized for protein yield and the ability to resume proliferation.
Initial attempts seeding at 1.0 × 104 cells/cm2, and using this seeding gave insufficient protein yields at both 0 and 24 h post stimulation. Seeding at 2.0 × 104 cells/cm2 or greater reliably yielded amounts of protein quantifiable by Bradford assay at these same time points (data not shown). Of the seeding densities that yielded protein, seeding with 2.0 × 104 cells/cm2 has the greatest slope between 3 and 4 days post-seeding and therefore the most rapid recovery when stimulated with serum. Seeding at 3.0 or 4.0 × 104 cells/cm2 gives smaller slopes, suggesting a slower recovery and less complete release from arrest. From these results, using 2.0 × 104 cells/cm2 was selected for future experiments, because arrest was complete and without cell death, reentry into the cell cycle was rapid and protein yield was adequate.
6.2. Optimizing Detection of IQGAP1 by Immunoblotting Once cell culture conditions were optimized, conditions for protein detection were tested. Initial IB of IQGAP1 resulted in saturated signal of the film, with multiple bands down the each lane and bands that bleed across the majority of the top part of the gel (see Figure 4, top panel). This suggests strong and non-specific binding by the 1º antibody to the proteins in the lysate. To correct this, optimization of the 1º dilutions used during incubations needed to be performed. To do this, lysates were ran in multiple instances of duplicate lanes and transferred. The instances were separated and incubated with varying dilutions of anti-IQGAP1 antibody ranging from 1:100 and 1:1 000, as per manufacturer’s recommendation (Figure 4, top panel). This did not yield optimal results, so dilutions were continued down to 1:50 000 (Figure 4, bottom panel). Greater dilutions yielded less staining non-specific and produced single, clear band. Staining appeared proportional to the amount of antibody present during 1º antibody incubations. Signal was detected with dilutions as great as 1:50 000. No signal was detected when the membrane was incubated in blocking buffer alone, i.e. no 1º antibody. These results suggest that IQGAP1 is
51
in high abundant in MOVAS and/or the antibody is very strong. The antibody needs to be diluted greatly to produce unique, specific band.
In order for quantifications to be meaningful, the linear range of detection needed to be determined. A range of masses of MOVAS lysate were loaded onto gels and separated by electrophoresis. Proteins were indirectly IB with anti-IQGAP1 antibody and anti-rabbit 2º antibody (Figure 5). The signal was quantified and plotted dependent on the protein loaded (Figure 6). The linear range was determined by linear regression and finding the largest range with an r2 ≥ 0.90. This range was between 1–10 µg of protein loading. By loading 3 µg of protein, this gives approximately 3-fold sensitivity in both directions.
6.2.1.
Anti-IQGAP1 Pulldown
IB detection for IQGAP1 yielded repeatable detection of a band at the approximate molecular weight of IQGAP1. However, the identity of this protein was not known. To confirm that the band was indeed IQGAP1 as well as testing the suitability of the anti-IQGAP1 to pulldown endogenous IQGAP1 protein, whole cell lysates incubated and precipitated with the antibody, or negative controls protein G resin only, or naïve rabbit IgG. Whole cell lysate serves as a positive control for IQGAP1 detection. The protein pellets were separated by electrophoresis. The gels were either stained with Coomassie Blue or immunoblotted as normal (Figure 7). Staining and IB both yielded a 190 kDa band present in whole cell lysate and anti-IQGAP1 precipitations. This band is not seen in either negative control. This is probably IQGAP1 and suggesting the antibody is specific to IQGAP1.
Both staining and IB also detect a 50 kDa band in the naïve IgG and anti-IQGAP1 conditions. The presence of immunoglobulin in both reactions as well as the molecular weight suggest that these bands are the IgG heavy chains, still able to react with 2º antibody after IB preparation. Staining with Coomassie Blue produced two bands that were present in all pulldown mixtures not detected by IB. The first protein band migrated to approximately 220 kDa and the second
52
was approximately 40 kDa. All Coomassie-stained bands were excised from the gel and sent for identification by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Mass spectrometry confirmed the identities of IQGAP1 and rabbit IgG heavy chain. No posttranslational modifications were found on the IQGAP1 fragments. Additionally mass spectrometry identified the 220 kDa bands as either myosin-9 or myosin-10 and the ~40 kDa bands as actin.
6.3. Optimizing Detection of CaM by Immunoblotting Initial IB for IQGAP1 produced excessive signal. This was easily remedied by typical optimizing typical parameters, the 1º antibody and loaded protein. In contrast, CaM yielded no signal and optimization remained elusive. Using standard protocols for detecting protein with the standing anti-CaM antibody used in house (sc-5537, Santa Cruz Biotechnology, Inc) did not yield signal (see Figure 8). Protein yields of the lysate optimized (see § 6.1) and the presence of protein in general was not an issue as the loading control GAPDH was detectable, and IQGAP1 detection was already optimized (see § 6.2).
CaM is expressed in all eukaryotic cells and is highly expressed in SMC. Despite this, CaM protein seemed absent from the membranes during IB. In order to optimize detection, a positive control for CaM protein was required. For this, purified recombinant human CaM (rhCaM) was used. To assess the sensitivity of detection, gels were loaded with a range of protein lysate as well as a range of rhCaM and probed with anti-CaM antibody (Figure 8). Similar to previous attempts, CaM could not be detected from lysate at any amount, including the maximum the gel well accepts, 50 µg. Signal from CaM protein could be detected from rhCaM when more than 0.1 µg. Taken together this means that CaM would need to be 0.2 % of total protein in a cell to be detectable. The highly abundant cytoskeletal protein, actin, is typically 1 % of total protein. As a second messenger, concentrations of CaM are much less. While variable between cell types, 0.1 % of total protein has been cited previously [219, 220]. Therefore the limit of detection is in the order of magnitude with the concentration of CaM in cells. Conversely, more protein than is
53
present in the cell is required for the development of signal. This makes endogenous protein undetectable with standard IB methods.
6.3.1.
CaM Requires Unique Conditions during Transfer Compared to Size-Matched, Control Protein COX4
CaM protein is not detectable by standard IB, while IQGAP1, and loading control, GAPDH are readily detectable. The molecular weights of IQGAP1 and GAPDH are 190 kDa and 37 kDa, respectively. Both proteins are much more massive than CaM with a molecular weight of 17 kDa. These size differences potentially disqualify them as useful controls for assessing the movements of CaM during IB.
Size can affect IB in two ways. First, size is inversely proportional to protein mobility through the gel during both electrophoretic separation and transfer to the membrane. Second, size is proportional to the strength of electrostatic forces that adsorb the protein to the membrane. Additionally, larger proteins tend to be more complex in folding, and are more easily unfolded to expose hydrophobic regions that further aid adsorption and retention on membranes.
Increased mobility rates of small proteins, such as CaM, mean low molecular weight proteins may be lost off the gel during electrophoresis. This can be rectified by reducing run time, at the expense of signal resolution. Migration may also be slowed by using a higher density polyacrylamide gel for electrophoresis. Runtime and density were not a concern as the CaMdetection protocol kept the faster migrating protein ladder markers and the loading dye on the gel (not shown).
During transfer low molecular weight proteins can be lost from the membrane as their high mobility allows them to move quickly out of the gel, onto and though the membrane. This is known as ‘blow-through’ and can be optimized by reducing the time and current of the transfer
54
or the addition of second membrane to catch the ‘blown-through’ proteins. The inclusion of a second membrane did not yield CaM protein by IB either (not shown). This suggests that there was a minimal loss of protein due to ‘blow-through’ or that CaM adsorbed to neither the first nor the second membranes. In order to solve the problem of CaM detection, the two possibilities of protein loss during transfer and the lack of CaM retention need to be parsed.
If ‘blow through’ of small proteins or a similar size-dependent loss of protein during transfers was occurring, other similarly sized proteins should also be subject to this loss and be difficult to detect by IB. To test this, mitochondrial protein cytochrome C oxidase subunit 4 (Cox4; 169 aa, 19.5 kDa) was used as size-controlled comparison. Cox4 was previously used in our laboratory in an unrelated line of inquiry (Figure 9). While Cox4 was readily detectable with the standard protocol, CaM was still not detectable.
Conditions for transfer were altered to determine whether they could be optimized for CaM detection. The first parameter of optimization was to reduce loss of protein by changing the membrane to a FluoroTrans W (Pall). FluoroTrans W is a preparation of PVDF that is supposed to be optimized for small protein detection, due to its smaller pore size of 0.2 µm compared to the more typical 0.45 µm. By decreasing the porosity it increase adsorption surface area as well as reduced protein permeability. Changing membranes did not appear to affect the intensity of protein staining by Coomassie Blue (Figure 10). Thus pore size does not appear to be a primary issue with respect to detection of CaM.
Further tests on protein transfer time and conditions were conducted. The reasoning was the more time the transfer is left running, the more time proteins have to make their way through the PVDF membrane. Thus reducing transfer time reduced the time that proteins have to ‘blow through’. Transfer time could not be easily increased as the transfer buffer may break down with run time and the small buffer volume afforded by the XCell II blot module. Additionally, washing the gel before transfer was tested as an optimizing step. Washing removes running
55
buffer and prepares the gel for transfer. There are benefits and drawbacks of washing the gel before transfer, because the presence of SDS from the running buffer may be advantageous or disadvantageous to transfer. SDS can help the movement of proteins out of the gel in the same manner that the detergent aids electrophoresis. SDS binds to proteins, exposing hydrophobic pockets and unfolding the protein, therefore linearizing the polypeptide. This binding is able to mask the charged residues or neutrality of the protein and create fairly uniform length:charge ratio. However, SDS solubilizes proteins in the buffer, inhibiting adhesion to the membrane. Washing the gel is also thought to aid transfer as the methanol in the transfer buffer dehydrates the gel and fixes proteins in position. However, washing the gel may potentially proteins wash out, thus reducing sensitivity. Additionally, washing the gel increases handling time therefore increasing the potential for gel damage and contamination.
Attempting to optimize the transfer conditions yielded similarities and differences between CaM and Cox4. For both proteins, longer transfer times improved detection (Figure 11; top row and bottom right). Cox4 from lysate is detectable from at 15 min transfer, but signals are stronger from samples transferred for 30 and 60 min. Similarly, rhCaM produces poor or no signal after 15 min transfers compared to 30 min transfers. Signal from 30 min transfers are weaker than 60 min transfers. Therefore reducing the transfer time does not appear to improve detection by IB. While loading gels into the transfer directly from electrophoresis allowed detection of Cox4, washing away running buffer prevents detection of Cox4 (Figure 11; bottom left). In contrast, rhCaM is detectable under both conditions, but detection is improved by washing (Figure 11; bottom right). Endogenous CaM yields no signal under any condition (Figure 11; right column). From these results, a few conclusions can be made. Since reducing transfer time worsens protein detection, ‘blow-through’ does not explain the lack of signal. Also, washing the gel prior to transfer is an important parameter that affects subsequent detection of protein. Normally considered as a minor step, washing the gel directly impacts transfers and may make or break an IB detection. For some proteins, washing worsens IB detection. In the particular case of Cox4, washing abolished signal. In contrast, CaM benefits from washing. Due to these contradictory effects, the effectiveness of gel washing needs to be assessed based on individual protein, or protein families basis. Additionally, variation in protocol, such as buffer compositions, may also
56
alter the benefit of gel washing. Specifically, these observations predicate that optimization of CaM IB protocols will require conditions that impede the detection of other proteins.
6.3.2.
Fixation with Glutaraldehyde Improves CaM Signal
The need to develop a protocol for reliable, repeatable detection of endogenous CaM spurred a search for a successful method. Literature searches and perusal of methods yielded little insight. In contrast, several papers that include details of IB excluded clonality or source of their antiCaM antibodies. Fortuitously, unrelated search uncovered a passive mention of fixatives during the IB detection of CaM protein [73]. This was surprising and initiated searches for the use of fixatives during IB of CaM. The search yielded several methodological papers trying to overcome specific problem of detecting CaM and related metalloproteins. Two solutions were proposed that seemed to be an improvement over the current system. The first solution was fixing the proteins to the membrane [199, 201, 202, 221]. The reasoning is that CaM may bind to the membrane initially, but is insufficiently hydrophobic to remain bound. Indeed, loss of CaM proteins from solid substrate has been demonstrated [199, 221]. The second was a change of transfer buffer [200, 201, 222, 223]. Altering the transfer buffer is thought to give better initial membrane adsorption. Additionally, a manufacturer that includes similar solutions in anti-CaM antibody methods was also found [224-226]. Similar to the dearth of information in literature, no other antiCaM antibodies manufactures include any deviations from their typical protocols in their product information sheets. This even includes other manufacturers that carry the same clones.
Fixing proteins to membranes is an old-fashioned remedy for poor signal based on applying immunohistochemistry and immunocytochemistry techniques to IB. In immunohistochemistry and immunocytochemistry, different fixatives are available and their selection may depend on the target in question. To test which fixing agent would aid in the detection of CaM replicate lanes of rhCaM and protein lysate were transferred, fixed with different fixing agents, then blocked and probed as usual (Figure 12). As expected from previous attempts, the modern standard using no fixing gave no signal. Interestingly, neither 100 % methanol nor 0.4 % PFA produced better detection of CaM. Only 0.2 % glutaraldehyde fix gave signal for rhCaM as well
57
as, for the first time, signal for endogenous CaM. This highlights the idiosyncratic nature of CaM. PFA is commonly used as the default fixing agent for fluorescent labeling tissue or cells. Here, PFA failed to improve detection. In contrast, glutaraldehyde, which is largely constrained to electron microscopy due to auto-fluorescence [227], is required for CaM IB. This autofluorescence caused by glutaraldehyde fixation disqualifies the use of LI-COR Odyssey infrared imager as a valid method for detecting probes (Figure 13). Therefore, traditional, horseradish peroxidase, chemiluminescence and x-ray film were used.
6.3.3.
Using Potassium Phosphate Transfer Buffers Improves CaM Signal
The second solution for detection of CaM protein by IB was to move away from standard transfer. Proposed solutions included adding NaCl or CaCl2 to Towbin transfer buffer, or performing the transfer buffer with a potassium phosphate buffer (KP). Adding salt to the transfer buffer increases the higher ionic strength of the transfer buffer. The goal is that this will encourage hydrophobic interactions with the membrane and greater adsorption [222]. Adding CaCl2 or using KP buffer is thought to improve transfer by precipitating detergent from the solution as the calcium or potassium dodecyl sulfate salt has a lower solubility than the sodium salt [200, 223].Phosphate ions benefit transfer also [200]. KP buffer necessitates overnight transfers as it has higher conductance than normal transfer buffer. However, lower voltage may encourage CaM binding to the membrane [200].
Adding 2 mM CaCl2 [200, 201] to Towbin transfer buffer did not improves signal when combined with glutaraldehyde fixation (not shown). In contrast, KP buffer combined with glutaraldehyde fixation greatly improve CaM-specific signal (Figure 14).
58
6.3.4.
Optimized Detection of Endogenous CaM by Immunoblotting
In order to have sufficient sensitivity to detect fluctuations in protein expression, the linear range of CaM detection was determined performing IB on a range of masses of MOVAS lysates. Proteins were indirectly IB with anti-CaM antibody and anti-mouse 2º antibody (Figure 15). Signal was quantified and plotted dependent on protein loading (Figure 16). The linear range was determined by linear regression and finding the largest range with an r2 ≥ 0.90. This range was from the detection limit of 10 µg to the maximum loading of 50 µg. By loading 23 µg of protein, a 2-fold sensitivity in detection could be achieved. Similarly, detection from rhCaM was plotted versus the input protein. Values were used as inputs into the equation of the linear regression equation to give approximations of the CaM content of MOVAS. Assuming that endogenous CaM behaves similarly to purified rhCaM, the CaM yield from MOVAS is 0.002 1 ± 0.000 4 µg of CaM per 1 µg of whole cell lysate or CaM content is approximately 0.21 % of total protein.
6.4. CaM and IQGAP1 have Stable Expression Following Serum Stimulation With protocols for the detection of IQGAP1 and CaM, we were able to assess the expression of both proteins following serum starvation/stimulation, and thus the pattern of expression relative to the cell cycle. MOVAS were starved of serum for 2 days, and then stimulated with serum. Protein lysates were collected at different time points, IB was performed for CaM, IQGAP1 and GAPDH, then signal was quantified (Figure 17). There were no statistically significant differences between group means as determined by one-way ANOVA [IQGAP1 F(6, 24) = 0.698, p = 0.654, CaM F(6, 19) = 1.323, p = 0.295, GAPDH F(6, 19) = 1.305, p = 0.302]. Protein expression for CaM, IQGAP1 and GAPDH show no pattern of expression relative to time of serum stimulation.
59
6.5. IQGAP1 Appears to Prefer ApoCaM over Ca2+-CaM Late in the Cell Cycle To test the hypothesis of a cell-cycle dependent interaction between CaM and IQGAP1, reciprocal co-IP reactions were performed on serum starved/stimulated MOVAS lysate. Reactions were incubated in the presence of 2 mM EGTA to create a Ca2+-free binding conditions, or 2 mM CaCl2 to identify Ca2+-dependent interactions (Figure 18). Resulting statistical analysis yielded no significant differences between group means as determined by oneway ANOVA [FIP CaM EGTA (3, 8) = 2.143, p = 0.173 (Figure 18B); Fco-IP IQGAP1 EGTA(3, 8) = 0.624, p = 0.619 (Figure 18C); Fco-IP IQGAP1 Ca2+ (3, 7) = 0.301, p = 0.824 (Figure 18C); FIP IQGAP1 EGTA (3, 8) = 0.231, p = 0.872 (Figure 18E); FIP IQGAP1 Ca2+ (3, 7) = 0.703, p = 0.58 (Figure 18E); F co-IP CaM Ca2+
(3, 8) = 0.465, p = 0.714(Figure 18F)].
Despite the development of CaM IB protocols and ready detection of CaM from EGTA reactions, samples from Ca2+-containing reactions yield no CaM signals (Figure 18 A and D). While the analysis for cell cycle expression did not identify specific cell cycle expression for CaM, under Ca2+-free conditions, there is a trend for greater precipitation of CaM at 16- and 24 h compared to 0- and 8 h of serum stimulation (Figure 18 B). With this increase, there is a minor trend for increased Ca2+-independent co-IP of IQGAP1 at as serum stimulation increases under (Figure 18C). Concomitantly, there appears to be a decrease in Ca2+-dependent interaction between CaM and IQGAP1. Likewise, despite roughly equal IP of IQGAP1 (Figure 18E), in the presence of EGTA, there tends to be increased CaM at 16- and 24 h (Figure 18F). This may suggest that during later phases of the cell cycle, there is an increase in apoCaM-IQGAP1 interactions.
6.6. Expression of Fluorescent Protein-Tagged CaM and IQGAP1 6.6.1.
Nucleofection into MOVAS
In effort to avoid the difficulties that were experienced with IB CaM and preparing for reciprocal co-IP, attempts were made to visualize the interaction using fluorescent microscopy. Being over
60
4.9 kbp in size, the IQGAP1 cDNA was expected to be difficult to isolate from cells within our laboratory. Fortuitously, YFP-tagged IQGAP1 plasmids had already been prepared by a neighboring laboratory [228]. We were able to receive both a plasmid to express CFP-tagged CaM, pCALM1-Cerulean, and a plasmid to express YFP-tagged IQGAP1, pVenus-Iqgap1 (see § 5.7 and Figure 20 for details). Together, CFP and YFP are a FRET pair, allowing for both two-color detection of the proteins by fluorescent microscopy as well as FRET. Potentially, this could also be combined with the Fucci system for detecting progression through the cell cycle. However, the live cell imaging required for FRET is not a technique that has been used this laboratory. Plasmids were transformed into competent E. coli. Successful transformation was screened on using antibiotic plates. Two clones of each plasmid were selected, expanded and isolated.
To test the expression plasmids, each plasmid was singly transfected into MOVAS using Amaxa Basic Nucleofector Kit for primary SMC (Lonza) according to manufacturer instructions. The kit included pmaxGFP as a positive control for successful transfection of fluorescent proteins. The kit also suggests using four additional programs, D-033, P-013, P-024, U-25, for electroporation of SMC. Neither pCALM1-Cerulean nor pVenus-Iqgap1 yielded signal using appropriate fluorescent channels by microscopy (not shown). Additionally, the transfection of positive control pmaxGFP did not produce green fluorescent labeling of cells (Figure 19 top). Cells were present by DIC imaging (not shown) as well as Hoechst staining (Figure 19 bottom).
6.6.2.
Demonstrating Veracity of Expression Plasmids
Communicating back with donor laboratories revealed that neither plasmid had been published prior. Combined with the lack of signal from the transfections we need to verify the identity and fidelity of the constructs. The plasmids were sequenced using the primers listed in Table 1 and this information was combined with detailed of the construct backbone to create plasmids maps (Figure 20).
61
6.6.3.
Nucleofection into Fucci T1 Mouse Carotid SMC
Assured of the identities of the plasmids, an attempt to combine the FRET pair with Fucci cell cycle reporter was made. This multicolor detection could be have difficulties as the red and green of Fucci, as well as the cyan and yellow FRET pair have overlapping excitation and emissions wavelengths (Figure 21). However, multicolor fluorescent detection is not uncommon and was used by the creators of Fucci [211]. The fluorescent constructs were nucleofected into Fucci T1 mouse carotid SMC and imaged using live cell microscopy (Figure 22).
The presence of cells was confirmed by DIC. For the red channel, there appeared to be red nuclear staining, suggesting that the cells had G1 DNA content. However, a minority of the cells had this staining. The majority of cells had an absence of red nuclear staining, with some cells demonstrating cytoplasmic staining. This is atypical for Fucci mKO2-Cdt1(30–120) [211]. For the green channel, there was no nuclear staining. Signal from mAG1-hGem(1–60) would be expected to be abundant because low levels of red nuclei suggests cells are passing through the cell cycle [211]. There appeared to be cytoplasmic and extracellular green signal. The extracellular green signal appeared to be greatest above and below the plane of the cells. Considering this pattern of staining and the cell type, this is probably auto-fluorescence of extracellular components secreted by the smooth muscle cells, particularly elastin byproducts. Additionally there was evidence of yellow signal from cells not transfected with pVenus-Iqgap1. There were also instances of signal appearing to match signal from other channels. This would suggest that there was overlapping excitation between the fluorophores.
6.6.4.
X-fection into HEK293T
Failing to demonstrate successful expression of reporter genes from the plasmid, a simpler system was chosen. The plasmids were transfected into the notoriously easy to transfect HEK293T cell line using the X-fect transfection kit (Clontech) and signal visualized after 19 h by live cell fluorescent microscopy. This system successfully transfected the plasmids into the cells demonstrated by presence of cyan fluorescence in cells incubated with pCALM1-Cerulean and yellow fluorescence in cells incubated with pVenus-Iqgap1 (Figure 23). Transfection with
62
pmaxGFP also was successful as the cells fluoresced green. This was not due to any inherent fluorescence of HEK293T, the media or the culture ware as cells that did not undergo transfection did not fluoresce in any of the spectrum channels.
6.7. Expression of CaM and IQGAP1 in Other Cell Lines The literature of CaM, IQGAP1 their cell cycle and proliferative effects are mainly targeting endpoints in cancer, not cardiovascular disease. Additionally, focusing our work on SMC has led to difficulties transfecting SMC. Switching to another, simpler cell type, such as HEK293T, may lead to greater progress that can be brought back to SMC. Additionally, switching cell types may make the work more appealing to others and thus have a greater impact. With this in mind, other cell types were tested for the presence and the ability to detect CaM and IQGAP1 in these cell types. Lysate from MOVAS, NIH3T3 and HEK293T were probed for both proteins (Figure 24).
6.7.1.
Immunoblotting Potential Positive Controls for CaM and IQGAP1 Co-IP
Critical to this line of inquiry is the demonstration of CaM-IQGAP1 complex formation. To definitively demonstrate specific protein-protein interactions between CaM and IQGAP1, we will need positive controls for co-IP. CaM has a variety of Ca2+-CaM binding proteins to choose from, including CaMKII. The CaMKII isoform, CAMK2D, is highly expressed in cardiovascular system [229, 230]. To identify the protein by IB, MOVAS, NIH3T3 and HEK293T were probed with anti-CAMK2D antibody (Figure 25). The expression levels of CAMK2D appear to be greatest in NIH3T3, and least in HEK293T.
Similarly, a positive control for anti-IQGAP1 co-IP is required. Naïvely, actin would be a default choice. However, myosin and actin were identified in non-specific pulldown reactions (Figure 7). Therefore actin is inappropriate for use as a positive control for IQGAP1 binding. The GTPases are well described binding partners of IQGAP1, but the interaction depends on the
63
activation of the enzymes. ERK2 was described as bound to IQGAP1 in untreated human breast epithelial cells, and was demonstrated to occur in HEK-293H [151]. ERK is already used in the laboratory for unrelated work. Anti-ERK antibody was used to probe MOVAS, NIH3T3 and HEK293T (Figure 26). The cell lines used appear to have comparable levels of expression of ERK proteins.
64
Chapter 7 Conclusions
7.1. Technical Findings Technical difficulties were experienced at every step on the path from cell preparation to detection of protein signals for IQGAP1 and CaM. Even attempts to curtail these issues by changing experimental techniques were met with their own set of problems. This demonstrates the novelty of this work to our laboratory. Any experiment intended to test a hypothesis must be optimized for the system being tested, or the system must be changed to match the ability of the experimenters.
7.1.1.
Seeding and Protein Yield
For MOVAS, following the protocol described in § 5.2, 2.0 × 104 cells/cm2 is required for adequate recovery from serum starvation and enough cells on a reasonable sized culture ware to yield sufficient protein for IP and IB.
7.1.2.
IQGAP1 Detection
Optimized conditions for detection of IQGAP1 protein from MOVAS and using antibody sc-10792 require 1:10 000 dilutions and 3 µg of lysate. This dilution factor is greater than manufacturer recommendations. This suggests that the antibody may be strong affinity for IQGAP1 or that IQGAP1 is abundant in MOVAS, or both. While the antibody is strong, there is high level of cross-reactivity and non-specific interactions that need to be minimized before.
7.1.3.
CaM Detection
CaM protein is physically unique compared to most other proteins. These unique physical properties render CaM protein difficult to detect by IB. However, knowledge about CaM
65
predates protein electrophoresis and IB. As a consequence, others researchers have attempted to IB CaM, encountered these problems, and developed optimized conditions to address the limitations. These optimizations are old-fashioned, but are still necessary despite advancements in IB technique, improved membranes, more sensitive detection and high-affinity antibodies. As long as laboratories try to study CaM and use IB, these optimization steps can not be ignored Since CaM will continue to have the same physical properties. Despite the effort to develop this methodology for detection of CaM protein by IB, detection appears to fail when the samples contain Ca2+.
7.1.4.
Plasmids and Microscopy
The plasmids when expressed in cells yield fluorescent protein-tagged construct proteins. Combined with the sequencing suggests a failure for the Amaxa Nucleofection kits in introducing them into MOVAS or the Fucci T1 mouse carotid SMC. Additionally, the extracellular fluorescence seen with the Fucci T1 mouse carotid SMC suggests that SMC require fixation and a mounting agent to reduce background fluorescence.
7.2. Biological Findings 7.2.1.
Pattern of Expression for CaM and IQGAP1
In MOVAS, it the expression levels of CaM and IQGAP1 protein do not exhibit any pattern for related to serum starvation/stimulation. This implies that neither protein has cell-cycle dependent expression patterns.
7.2.2.
Cell Cycle-Dependent Co-IP of ApoCaM-IQGAP1 Complexes
While statistical analysis did not identify significant differences between any of the time points in either EGTA or Ca2+ conditions, there appears to be a tendency for increased reciprocal precipitation of CaM and IQGAP1 under Ca2+-free conditions. This increased apoCaM-IQGAP1
66
complex occurs concurrently with a decrease in co-IP of IQGAP1 from Ca2+-CaM. There is also an increased precipitation of CaM as cells proliferate in the absence of Ca2+.
67
Chapter 8 Discussion
8.1. Technical Findings 8.1.1.
Seeding and Starvation Protocol
Confluency, subconfluent is a term that is useful as it communicates, but it does not offer useful meaning between cell different types or cell lines. Also, assessment of confluency is either an estimate of visual coverage, and is therefore a shorthand/shortcut, or is derived from cell counts and densities, which can be communicated specifically. Cell density values carry less ambiguity about the procedure. However density is still limited to individual cell types. Descriptions the goal of cell density is designed to achieve, along with the density value so that experiments. Various goals may be achieving cell monolayers, rapid proliferation, rapid growth or sensitivity to treatments.
Similarly treatments such as cell synchronization need to be optimized on a cell type basis. For instance, MOVAS require 0 % serum and 2 days starvation for arrest. Conversely, a primary cell culture would not be able to tolerate such severe media conditions. A complete deprivation of serum would cause cell death and proliferation would be arrested with reduction to 0.1 or 0.25 % serum present in the media. Similarly, cells may arrest within shorter period of time or require longer period to achieve synchronous cell populations.
Additionally, there are different methods for causing cell arrest and cell-cycle synchronous populations of cells. Serum starvation arrests cells in G0 and G1. Other methods such as cycloheximide, thymidine, hydroxyurea, mitotic shake, nocodazole, colchicine cause arrest by different mechanisms are in different phases. Additional methods for cell cycle arrest could be used to demonstrate a specific cell cycle response and not a delayed effect in response to the reintroduction of serum.
68
This work did not analyze the extent to which the arrested cells were synchronous. Instead it is based on past use of this technique in the laboratory [209]. To do this for MOVAS or other cell types, flow cytometry and FACS could be used to assess the number of cells that are in each phase of the cell cycle at different time points.
8.1.2.
CaM Detection
Detection of CaM and similar metal-binding proteins require distinct protocols for detection by immunoblotting. Ca2+ buffers during transfer are thought to remove detergent from CaM proteins [200, 223]. However, CaM increases hydrophobicity when bound to Ca2+ [39]. This would make CaM proteins more adherent to hydrophobic membranes. Glutaraldehyde is superior to other fixatives probably because of the high α-helical content of CaM and glutaraldehyde fixes in part by deforming α-helical structures. Autofluroescence of glutaraldehyde fixatives is a known phenomenon [227]. Altering transfers, fixing, including host serum proteins are known methods for improving protein detection. These were more common in the past as high-affinity antibodies make them largely unnecessary or a hindrance.
CaM has been known for decades and studied extensively. Recent articles and manufacturer information sheets do not make mention of divergent techniques. There are three possibilities. The first possibility is that these scientists have vastly improved their detection compared to our laboratory and the past researchers that have published these more primitive techniques. This is unlikely as CaM proteins have not changed physical characteristics and these characteristic are the major source of trouble. Additionally, the work here demonstrates that many modern developments have not addressed issues that are needed to detect CaM. The second option is that publishers are aware of these techniques and are using them, but are not publishing or referencing their methods. They could be doing this to hinder competitors. Evidence for this is that many papers include detailed information about other antibodies used throughout the papers, but exclude anti-CaM antibodies. However, this is still ethically problematic as would be in bad faith and go against the principle of communicating meaningful results to further scientific endeavor. The third option is that authors are not aware that CaM requires these techniques.
69
Published papers include results of what the authors attribute to CaM but they are not using optimized techniques. Based on this work, this means that their results are dubious. This doubt may be the reason why method sections omit to include information for anti-CaM antibodies. The authors are aware their results are doubtful, but believe the antibody sensitivity to be a fault. This also means that every publication that perpetuates the idea that CaM is just another protein, and misleads future researchers. If this is true, then it calls into question the merits of peerreview. Peers researchers that work on CaM should be sufficiently knowledgeable to be aware of the difficulties of CaM IB and catch these exclusions. However, the interest in CaM has faltered in recent years as well as a diversification of fields of study. Therefore reviewers of CaM studies are peers by proxy of another field, such as cardiovascular disease, and are thus ignorant of CaM biology.
Roger Y Tsien, winner of the Nobel Prize in chemistry for developing GFP into a vastly used tool as it is today, initially worked with Ca2+. He had already worked with Fura-2 and Indo-1, fluorescent dyes capable of imaging Ca2+ [231]. Similarly, upon developing GFP, he used it for demonstrating CaM activity [232]. Additionally, Sepharose-CaM remains a useful tool. Primarily it is used for purification and chromatography of proteins tagged with a Ca2+-dependent CaMbinding protein. This allows for simple association and dissociation in the presence of Ca2+ ions or Ca2+ chelators. However, CaM-Sepharose was initially developed and remains useful for studying CaM. Both fluorescent proteins and CaM-Sepharose have a new life as useful tools in general biochemistry, but they started as workarounds to inability to reliably detect CaM and CaM activity.
The methods used to detect CaM activity (e.g. modified IB, purification, chromatography, GFP) are not unique to CaM. These and other more elaborate techniques were common place when high-affinity antibodies were not available for any protein. However, with the advent of the Genome Project and manufacturers carrying stocks of high-affinity antibodies, these techniques have been large forgotten and have not passed on to the next generation of biologist. While these methods are labor-intensive and are largely antiquated, they remain useful, when road blocks are encountered.
70
Proteins are more difficult to analyze than nucleic acids. Nucleic acids differ only in length, sequence and melting point. Analysis and manipulation is simple and fairly uniform. PCR, Southern, northern blots, sequencing and genetic engineering are relatively simple and require only a handful of tools are extremely powerful
In comparison, differ in form and function. Many proteins are enzymes and activity may be measured. However, many proteins are not enzymes. This includes both proteins, CaM and IQGAP1, limiting proteins analysis to IB. Every step in IB is idiosyncratic. Every protein is unique. Every 1º antibody is unique. Often multiple antibodies may be needed for one protein to identify different fragments or posttranslational modifications. The multitude if gel compositions, densities, running buffers, transfer buffers, blocking, membranes, signal production and signal recording, stripping. Every step is critical and needs to be optimized for the samples, and questions at hand. While this is widely known and manufacturers do a decent job of having worked out excellent general guidelines, the older methods such as fixation, use of various serums, deviations from traditional or widely published reagents, are being forgotten. The issue with CaM have been discussed and largely solved in the past, but failure to communicate this in subsequent reports is creating serious problems and needs to be addressed.
Washing is important as SDS reduces binding of calmodulin to membrane [223]. These properties are also found in other metalloproteins such as S100 proteins (11 kDa), hemoglobin (15 kDa), metallothioneins (6 kDa), and synuclein (14 kDa) [199, 201, 202, 221]. However, the prototypical Ca2+ binding protein, troponin C (18 kDa), does not have these blotting problems [223].
These old-fashioned improvements as well as the simple addition of washing the gel before transfer may not improve detection of other proteins. Evidence is that Cox4 signal worsened upon washing. Similarly, smaller sized proteins such as unrelated histones (10 kDa) are categorically different from CaM. CaM is a soluble protein that chiefly interacts with positively
71
charged Ca2+ and possibility other cations. In contrast, histones interact with negatively charged DNA, so may not benefit from these improvements.
8.2. Biological Findings This work asked whether the cell cycle is regulated by interactions between CaM and IQGAP1. Also under investigation was how the interaction is influenced by the presence of Ca2+. Ca2+ and CaM, and IQGAP1 regulate proliferation and Ca2+ and CaM regulate several cell cycle processes [46, 47, 55, 57-60]. Also, CaM binds IQGAP1 and this binding regulates IQGAP1 interaction with other proteins [100, 111, 116, 137, 138, 155, 179, 182, 184-187]. However, the timing and role of CaM binding IQGAP1 is not well known. The results demonstrated here show CaM and IQGAP1 may interact more in later stages of cell cycle, the G2 and M phases. Late phase activity of IQGAP1 has precedence, as yeast IQGAP family proteins arrange and separate chromatids in mitosis and loss of IQGAP1 results in failure to separate chromosomes and genetic instability [170, 171]. These homologs also form part of the contractile ring and subsequently participate in cytokinesis S. pombe, and budding in S. cerevisiae. However, similar links between late cell cycle and IQGAP1 have not been demonstrated in higher species. These steps are also regulated by CaM [42, 59, 93]. Taken together, the results and literature suggest that CaM and IQGAP1 act together to separate the DNA and dividing the daughter cells. Overlapping phenotypes of CaM RNAi and IQGAP1 knockout have been observed previously [172]. However, the IQGAP1 knockout mouse is viable and only has a slightly higher risk of developing tissuespecific cancer [174]. These two facts suggest that IQGAP1 is expendable because IQGAP1 is a redundant protein or plays a minor, facilitative role in the cell cycle. Homology between IQGAP1 and IQGAPs -2 and -3 are 62 % and %, respectively [101]. Conclusion that IQGAP1 is expendable is not new [105]. Despite the similarity in protein structure and sharing several binding partners, there is evidence of IQGAP functions diverging [102, 178]. However, the potency for IQGAP1 as a pharmaceutical target still remains as IQGAP1 is a major signaling node.
These G2/M interactions between CaM and IQGAP1 occurred more frequently under Ca2+-free conditions than in Ca2+-containing buffer. Ca2+ is required for most CaM interactions. Unusually,
72
CaM can bind IQGAP1 as both Ca2+-CaM and apoCaM. Considering that Ca2+-CaM allows passage through all phases of the cell cycle; it is surprising that IQAGP1 prefers apoCaM at late cell cycle. One possible explanation is that Ca2+-CaM interacts with other binding partners, and these proteins out compete IQGAP1 for available Ca2+-CaM. Instead, IQGAP1 binds to apoCaM between Ca2+ influxes. Ca2+ signals are fleeting compared to the longer Ca2+-free intervals. Therefore IQGAP1 preferring apoCaM in G2/M implies there may be more frequent and/or more stable interactions. In general, either Ca2+-CaM or apoCaM binding IQGAP1 prevents IQGAP1 from binding other binding partners. Many IQGAP1 binding partners transmit signals, and the transmitted signals are strengthened when the signal transducers bind IQGAP1. The signal proteins, such as β-catenin, or MTORC1, convey messages for cell cycle entry and growth pathways; all early events in the cell cycle. Therefore, apoCaM may act specifically in the late cell cycle to prevent inappropriate scaffolding and signaling. MAP kinases are active during mitosis [233, 234] Also, IQGAP1 binding partners APC, ARF6, and CLIP170 have been individually identified as playing roles in kinetochores and cytokinesis [16, 19, 235]. Therefore, is is possible that CaM is directing IQGAP1 from one set of partners to another. Additionally, IQGAP1 contains a destruction motif that might target the protein for ubiquitinylation by the APC/C and proteasomal degradation [106]. Therefore, apoCaM binding to and inhibiting IQGAP1 may be the first steps of a larger process that results in IQGAP1 degradation. However, IQGAP1 expression levels show no reduction in our experiments. The steady expression levels may be a result of MOVAS being an immortalized cell line, which may have altered protein turnover.
Similar Ca2+-independent, M-phase CaM interactions an have been observed with Aurora B and cyclin D3 [94, 95]. When chromosomes fail to segregate, CaM colocalizes with Aurora B and antagonizes SCFFBLX2 from ubiquitinylating Aurora B, thus protecting the kinase from degradation [94]. IQGAP1 sequence contains has a destruction motif that suggest IQGAP1 is a substrate for APC/C and M-phase degradation, but no M-phase degradation has been described yet [106]. Perhaps CaM plays a similar role with M-phase IQGAP1 and prevents APC/C-regulated degradation by inducing conformational change. In contrast to most F-box proteins, FBLX2 does not target ‘degrons’, but targets CaM-binding IQ motifs directly [95]. Instead of APC/C, SCFFBXL2 may target IQGAP1 for ubiquitinylation during M-phase. Additionally, FBXL2 interacts with CaM and this enhanced ubiquitinylation of cyclin D3 [95].
73
8.2.1.
Limitations and Confirming the Results
Conclusions from the findings presented here are hampered several technical limitations. A primary difficulty is that Ca2+-containing samples yield poor detection of immunoprecipitated CaM. CaM bound to Ca2+ changes shape as a part of its biological function [30, 36, 38-40]. These changes may alter or hide epitopes, which would prevent antibody binding [236]. Because these changes are relevant to antibody affinity, anti-Ca+2-activated CaM have been raised and are commercially available. Loss of binding should prevent both detection and pull down. However, anti-CaM antibodies were able to co-IP IQGAP1 and IQGAP1 does not precipitate when mixed with Protein G and rabbit gamma immunoglobins. Therefore, the antibodies probably bind and pull down Ca2+-CaM successfully. More likely is that Ca2+-CaM complexes are more difficult to immobilize than apoCaM and were absent from the membrane. Since CaM maintains its structure through cystine reduction and SDS denaturing, Ca2+-CaM can remain bound to Ca2+ during SDS-PAGE. Additionally, Ca2+-CaM migrates faster than apoCaM due to the more compact conformation [30, 35, 36, 194, 195, 197, 198]. CaM usually migrates at 17–20 kDa, while Ca2+CaM migrates at a rate of 15 kDa. Similarly, the increased mobility may alter the rate escape from the gel or retention of Ca2+-CaM proteins during transfer. The optimized, CaMimmunodetection protocol was developed with apoCaM and appears that the protocol is insufficient for Ca2+-CaM detection. To verify this shortcoming, rhCaM or cell lysates can be immunoblotted with and without the samples containing added Ca2+. If samples containing Ca2+ proves to hinder detection, a solution must be found. A protocol can be tailored specifically for Ca2+-CaM, or before electrophoresis, Ca2+ may be removed from samples. Precipitated protein pellets can be solubilized in buffer containing EGTA to chelate Ca2+ away from CaM, or be subjected to dialysis to remove Ca2+ from the samples.
To confirm that apoCaM and IQGAP1 interact at in a cell cycle-dependent manner, additional experiments need to be performed. The results currently lack statistical significance. Statistical strength would improve with more repetitions. Also, the specificity the pull down reactions needs to be demonstrated by including controls for CaM and IQGAP1 binding. The results show that IQGAP1 is only isolated by specific antibodies, not protein G or naïve antibodies. However,
74
a similar negative control has not been demonstrated with CaM. Positive binding controls would demonstrate the validity of co-IP. Both proteins have several interacting proteins to from which to choose. Actin binds IQGAP1, but actin precipitated non-specifically in the pulldown reactions. So the pull downs need to be optimized so that actin pulldown is specific or another protein must be chosen. Erk1/2 also binds to IQGAP1 and does so in resting cells, but successfully IQGAP1-Erk1/2 co-IP seems to require a fixing step [151]. A positive control for CaM co-IP would also act as a control for Ca2+ requirement, as most binding partners only bind under Ca2+ conditions. CaM binds to and activates CaM-dependent kinases (CaM) in a Ca2+dependent manner. CaM kinase family of proteins consists of several isoforms and splice variants. CaMKIIδ is an isoform expressed in the cardiovascular system [229, 230]. However, CaM kinases can also become independent of CaM by autophosphorylation.
8.2.2.
Alternative Approaches
Immunoprecipitation and immunoblotting CaM have proven to be difficult. Limitations of working with CaM may be overcome with engineered tools. The need for specific antibodies could be alleviated with tagged proteins. The tags could serve as the target for the pulldown reaction as well as for detection. Additionally, tagged CaM would probably make detection less problematic as transfer and retention would depend less on the unique characteristics of CaM and more on the protein tag. Caution need to be taken so that the tag does not interfere with protein folding, structure and biological activity. Fluorescent protein-tagged expression plasmids could be used for this purpose. Using tagged proteins, either added to the reactions or expressed from plasmids, introduces exogenous proteins. Exogenous proteins become problematic if the binding is dependent on the dose of proteins or post-translational modifications.
Another method that can be used is immobilizing proteins for affinity chromatography. Purified protein can be affixed to commercially-available, CNBr-activated Sepharose, or similar activated matrices [206]. Attaching purified CaM [35, 194, 195] to CNBr-Sepharose creates CaM-Sepharose [203], which is commercially available. Using CaM-Sepharose, IQGAP1 could be pulled down to determine the ability for IQGAP1 to bind CaM, in the presence or absence of Ca2+. To similarly
75
pull down CaM, CNBr-Sepharose can use be used to create IQGAP1-Sepharose. Making IQGAP1-Sepharose requires isolated IQGAP1. IQGAP1 may be isolated by anti-IQGAP1 pull down, by synthesizing recombinant proteins, or by purifying endogenous IQGAP1 from cell lysates by using CaM-Sepharose and EGTA [182]. With both CaM-Sepharose and IQGAP1Sepharose, reciprocal pull downs can be performed. The negative control for Sepharose-bound pull down would be blanked CNBr-Sepharose. To blank CNBr-Sepharose, the resin needs to be fixed to a non-specific protein, such as BSA or casein, or other amino-containing molecules like Tris or glycine. While using immobilized proteins removes the need for antibodies during pull down, antibodies are still required for detection. Additionally, the bound protein is exogenous to the cell lysates introduces the same problems described earlier.
A third method to work around the difficulties of studying CaM is fluorescent microscopy. Fluorescent immunohistochemistry or imaging fluorescent fusion proteins avoids transferring proteins by being performed in situ. Both methods are done on fixed samples, mirroring what needs to be done to prevent loss of CaM proteins. Unlike the protein precipitation methods, fluorescent imaging does not demonstrate binding directly. At best, fluorescent detection demonstrates that CaM and IQGAP1 co-localize. Co-localization is required for protein-protein interactions, but proteins that co-localize are not necessarily bound together. However, visualizing interactions in an intact cell provides additional information about where the interaction occurs. Knowing where the CaM and IQGAP1 are when they interact could potentially give hints about their function and could guide future experiments. Both proteins can be found at the cortex, in the cytoplasm or in the nucleus. By using fusion proteins and live cell imaging, living cells can be viewed, avoiding most problems working with CaM. In fact, one of the early uses of fluorescent fusion proteins was observing Ca2+ [232]. Once conditions are optimized, protein-protein interaction could be view in vivo using FRET. Manipulating the system to determine Ca2+ dependency of CaM-IQGAP1 interaction is more difficult than the in vitro binding assays, but can be done using Ca2+ ionophores to increase the intracellular Ca2+ concentration and cell permeable chelators to reduce Ca2+ availability. However, GFP-fusion protein studies must always be interpreted cautiously and with the caveat that tagged proteins have typically not been shown to have full biological function [235]
76
To overcome the overlap of Fucci and FRET excitation and emission energies, FRET may be substituted with another live cell, protein interaction assay, bimolecular fluorescence complementation (BiFC). In this technique, a single fluorophore is expressed as two halves, each fused to a protein of interest [237, 238]. When proteins are near enough to form a complex, the Nand C-terminal halves of the fluorophore may interact and reassemble a functional fluorescent protein, which is detected by emitted light. This simplifies the microscopy by eliminating one, going from four wavelengths to three. This also reduces background noise as there is no emission until the fluorophore is reconstituted [239]. However, as the biochemical complementation is largely stable, BiFC can only measure increases in protein interactions, not decreases or dynamic changes [239-241]. Therefore, unlike FRET, complementation methods such as BiFC can not detect any further dynamics after the initial interaction in a single assay [241]. Further studies blocking the cell cycle at different phases would allow for multiple, intra-cycle interactions. The duration of interactions can be further investigated by additional affinity chromatography or precipitation reactions.
As of yet, our results do not differentiate between a cell cycle-dependent interaction and an increase in interaction initiated by serum stimulation. To make this distinction, additional time points are needed, extrapolated to visualize additional cycles of division, and interpolated to increase the ability to see variations at a finer time resolution. However, the longer a time points is from synchronization, the more the cells becomes asynchronous. An alternative is to arrest cells using another mechanism of stalling cycle progression. Arresting agent, cycloheximide, also block cells in G1 by inhibiting protein synthesis. Other agents and synchronizing methods enrich the cell population in other parts of the cell cycle. Hydroxyurea, thymidines and cytosine arabinoside all stall DNA replication for S phase arrest. Microtubule inhibitors like nocodazole or mitotic shake off arrest cells in M phase. Using redundant synchronizing methods that vary by mechanism of arrest, and having corroborating results would strengthen evidence for cell cycledependent CaM-IQGAP1 interaction and debase the possibility that differences are due to adding serum or drug treatment. Another alternative is to determine the CaM-IQGAP1 interaction from cells purify from an asynchronous population based on cell cycle phase. Cell cycle phase can be
77
indicated by DNA content, protein markers, or Fucci signal. Taking cells from a mixed population would eliminate variables introduced by treatment. However, asynchronous cells have a skewed distribution of cells favoring G1. To get samples representing S, G2 and M phase populations, sufficient cells must be collected.
Another limitation is that the study conducted with an immortalized cell line. Immortalized cells have lost or deregulated cell cycle regulation. It is known that Ca2+ signaling, CaM and IQGAP1 are altered from normal biology in immortalized or malignant states. We may have observed a consequence of immortalization as CaM levels did not appear to change with respect to the cell cycle as would be suspected. Repeating our experiments in a primary cell type would overcome this limitation.
8.3. Future Experiments What we have demonstrated so far is only a correlation between CaM-IQGAP1 and cell cycle. We need to determine whether this interaction has consequence by disrupting it. It would be easy to overexpress either CaM or IQGAP1. Knocking down or out CaM would prove difficult given that there are three genes. The difference CALM loci also do not align by BLAST. It should be possible to knock down IQGAP1 by RNA interference as there is only one IQGAP1 gene and there is also a knockout mouse [174]. Both Ca2+-CaM and apoCaM can be disrupted using ionophores or chelators to increase or decrease cytosolic Ca2+ concentrations. There are also small molecules and proteins that are CaM inhibitors. These approaches would disrupt all CaM function and potentially results in greater confounding artifacts than meaningful results. Also, their mechanisms of the CaM inhibitors are not completely understood, and may not inhibit CaM, but its effectors instead. While there are no described inhibitors of IQGAP1 per se, there have been publications of mutations and truncations of IQGAP1 having significant variations in function. Our understanding of IQGAP1 as modular protein with separate domains and motifs allows for disruption of these individual domains and motifs, while leaving the rest of the protein function arguably untouched. Such experiments would be able to determine the consequence of mutated IQGAP1 with CaM and narrow down the possible implications of the CaM-IQGAP1
78
interaction. Of particular interest would be modification of the IQ domain or individual motifs as they are though to be the major sites of CaM binding. Through deletion or conversion of complete motifs to incomplete and vise versa, we can shed light onto the biology of CaM and IQGAP1 in the cell cycle.
8.4. Future Work Presently, the IQGAP family is poorly defined. They are identified as homologs to S. pombe sar1 and the presence of any number of IQ motifs. Many members lack other domains that are seen in IQGAP1. We do not know how well knowledge gained from lower eukaryotes can be extrapolated to mammalian systems [98]. Some of these IQGAP homologs appear to be bona fide of Ras GAPs and not inhibitors of GTPase hydrolysis. Despite this, there seems to be conservation in function as actin and cytokinetic regulators. Also, cytokinesis has several proteins that are grouped into different groups. IQGAP1 has demonstrated functions that fall under more than one of these categories [18]
IQGAP1 should be a protein of interest for many areas of research due to its interaction with diverse signaling pathways, many of which are disease pathways of their own, such as Wnt, MAP kinases, DNA repair etc [143-147, 149, 150, 163, 165, 166]. Additionally, we are aware of significant overlap, branching and crosstalk of signaling pathways that makes the biological system able to respond to the myriad of environmental conditions and insults that the body is subjected to on a daily basis. Many of these diverse downstream targets are known IQGAP1-interacting proteins. Theses potential interactions further illustrate that IQGAP1 is poorly understood and needs much due attention. By searching these avenues for understanding, we may be able to develop more complete, unified theories about biology from the splintered and fractured understanding we have today to develop superior tools for diagnosis and greater interventions for treatment.
Many binding partners of IQGAP1 are implicated in the development of cancer, and it seems reasonable to postulate that several of these may contribute to the mechanism by which IQGAP1
79
enhances tumorigenesis. Studies like these may uncover the pleiotropism of IQGAP1 function. Examination of all these is a large undertaking, which is beyond the scope of the present work.
Passage through the cell cycle is needed for development, tissue homeostasis and wound healing. Cell cycle failure or dysregulated passage contributes to degenerative diseases, cancer Alzheimer’s disease and cardiovascular diseases [22, 242].
The cell cycle is a conserved biological process, but there is diversity in how cells accomplish these broad goals [13, 15, 97, 243, 244]. Broadly is mitosis and meiosis, which share much of their molecular machinery. Additionally, there is open, intermediate and closed mitosis depending on whether the nuclear envelope dissolves. The formation of coenocytes is an additional divergence, because nuclear replication occurs without cytokinesis. Cells also possess more than one method of dividing cells in two. Cells also divide by one of two methods: substrate-independent cytokinesis and a substrate-dependent, traction-mediated cytofission [245]. Thus a simply unified theory will probably remain elusive, but much will have to be done to resolve the every study brings understanding of the regulators that control the cell cycle. These regulators are the key for developing improved diagonistics and more effective treatments.
80
Figure 1: Schematic of Ca2+-CaM actions in eukaryotic cell cycle. The cell cycle is a complicated process requiring concerted actions of numerous proteins. An ion-protein complex, Ca2+-CaM, is required multiple times in the cell cycle to progress during proliferation. Ca2+-CaM is needed when (clockwise) entering into the first gap (G1) phase from a resting state (G0), progressing into synthesis (S) phase, the end of the second gap (G2) phase, arranging and separating chromosomes in mitosis (M) phase, and dividing the mother cell in two daughters. See text for more information.
81
Figure 2: Schematic of IQGAP1 protein structure and sites of protein-proteins interaction listed by domain. IQGAP1 is a large (1 657 aa), multi-domain protein that uses its multiple domains to interact with dozens of proteins. Understanding how the interactions take place at a molecular level may aid understanding IQGAP1 functions at cellular and organism levels. The domains of IQGAP1 are able to interact with different protein bind partners, as listed. CHD, calponin homology domain; WW, polyproline domain; IQ, tandem IQ motif domain; R, arginine-containing complete IQ motif; X, incomplete IQ motif; GRD, RasGAP-related domain; RGC, RasGAP C-terminal; * denote self-association. See text for more information.
82
recovery
starvation
stimulation
cell density (cells/cm2)
1.5E+05
1.0 x 10^4
1.0E+05
2.0 x 10^4 2.5 x 10^4 3.0 x 10^4 4.0 x 10^4
5.0E+04
0.0E+00 0
1
2
3
4
time since seeding (days)
Figure 3: Seeding and cell densities during serum starvation/stimulation protocol. MOVAS were seeded at initial density in the range of 1.0 × 104–4.0 × 104 cells/cm2 as indicated. The cells were then starved of serum for 2 days, followed by 10% serum stimulation for 24 h. Cells were counted by hemocytometer at daily intervals. Points represent mean cell counts and standard deviations at each time point. n = 3–5.
83
Figure 4: Optimizing anti-IQGAP1 antibody dilution. Equal amounts of MOVAS cell lysate were loaded in duplicate onto gels, separated by electrophoresis and transferred to PVDF. The membranes were incubated with anti-IQGAP1in varying dilutions. This is a representative blot from two replicates.
84
Figure 5: Immunoblotting endogenous IQGAP1 from MOVAS. MOVAS cell lysate was loaded at indicated amounts, immunoblotted with anti-IQGAP1 antibody, and signal was quantified to determine the linear range of detection. Probing without 1º antibody (no 1º antibody) was used as a negative control for 2º antibody cross reactions and background noise. This is a representative blot from three replicates.
85
Figure 6: Quantification and linear range of IQGAP1 signal from MOVAS. To determine the linear range of detection of IQGAP1, a range of MOVAS lysate was probed with antiIQGAP1 (Figure 5), and the signal quantified and plotted versus the loaded protein from MOVAS (filled squares ■, error bars represent standard deviation). The curve highlights the saturation of detected signal. The thick line indicates linear regression for the range of loaded protein from 0–10 µg, r2 = 0.954, n = 3.
86
Figure 7: Anti-IQGAP1 immunoprecipitation. To demonstrate the specificity of the antiIQGAP1 antibody and test its use in immunoprecipitation reactions, whole cell lysate was immunoprecipitated with protein G resin only, naïve rabbit IgG or anti-IQGAP1 antibody and separated by electrophoresis. The gels were either stained with Coomassie blue (right) or immunoblotted and probed with anti-IQGAP1 (left). Prominent Coomassie bands were excised (red boxes) from the gel and sent to be identified by LC-MS/MS. Proteins names accompany the identified bands. Immunoblotting signal from IQGAP1 is only present in whole cell lysate or anti-IQGAP1 reactions. The IgG heavy chains from the antibodies in the pulldowns produce bands at ~50 kDa. Note the presence of actin. This is a representative blot from two replicates.
87
Figure 8: Standard immunoblotting is insensitive for detecting CaM. Ranges of cell lysate and rhCaM protein were immunoblotted, probing for CaM. CaM is undetectable in cell lysate at all amounts of protein loaded onto the gel. Signal was detected from purified rhCaM when in samples containing at least 0.1 µg CaM. This is a representative blot from two replicates.
88
Figure 9: Cox4 can be easily detected by immunoblotting. 20 µg of heart extracts from control and treated mice were immunoblotted for Cox4 (19.5 kDa). Bands of Cox4 are readily detectable by immunoblotting from these samples.
89
Figure 10: Comparing PVDF pore size for effects on protein adsorption. In order to improve detection of CaM protein, tests were performed to determine the effect of reduced PVDF pore size. Gels were transferred to the either typical 0.45 µm PVDF (PVDF) or 0.2 µm PVDF (FluoroTrans W) and stained with Coomassie Blue. This is a representative blot from two replicates.
90
Figure 11: Transfer conditions affects detection of small proteins. Replicate lanes of 30 µg whole cell lysate (protein) or 0.1 µg rhCaM (calmodulin) were separated by SDS-PAGE. The gels were either transferred to PVDF directly after electrophoresis (no washing) or washed in transfer buffer to remove running buffer prior to transfer (gel washed). Length of transfer was varied 15, 30 or 60 minutes. Blots were the probed with anti-COX 4 and anti-CaM antibodies. Red boxes indicate the area where bands of Cox4 or CaM should appear. This is a representative blot from two replicates.
91
Figure 12: Effect of different fixatives on CaM detection. Literature has suggested that immunoblotting of CaM requires the proteins are fixed following transfer. However, the fixative used differs for different proteins. Replicate lanes of 30 µg whole cell lysate (protein) or 0.1 µg rhCaM immunoblotted. Following transfer, but prior to blocking, membranes were fixed in the described fixing agent. This is a representative blot from three replicates.
92
Figure 13: Glutaraldehyde fixation is incompatible with infrared signal detection. Transferred membranes with 0.1 µg rhCaM and 30 µg whole cell lysate were fixed with glutaraldehyde and blocked. The blotted membranes were not exposed to 1º or 2º antibodies. The membrane was imaged on LI-COR Odyssey Imager at both 700 nm and 800 nm channels. Glutaraldehyde-fixed proteins produce signal detectable on both channels. This interferes with detection of CaM from cells. This is a representative blot from two replicates.
93
Figure 14: Potassium phosphate transfers greatly improve detection of rhCaM. Literature suggests that transfers in potassium phosphate (KP) buffer improve detection of CaM A range of rhCaM masses were ran on SDS-PAGE and then subjected to two different immunoblotting protocols. In the left panels, gels were immunoblotted using our standard protocol (see § 5.5). Right panels were washed and transferred at 20 V overnight in 25 mM potassium phosphate buffer and fixed in place with 0.2 % glutaraldehyde. Top panels were stained with Coomassie Blue to detect the presence of protein. Bottom panels were probed with anti-CaM antibodies. Signal is absent in standard immunoblotting, even at 1 000 ng, while as low as 100 ng and 30 ng of rhCaM loading can be detected by Coomassie staining and immunoblotting, respectively. This is a representative blot from three replicates.
94
Figure 15: Detection of endogenous CaM protein by IB. After developing a protocol to successfully detect rhCaM, tests to determine the sensitivity for detection and the conditions for further experiments. A range of rhCaM and cell lysates masses were immunoblotted using the developed protocol. Endogenous CaM is readily detectable with 10 µg of cell lysate. This is a representative blot from three replicates.
95
rhCaM (µg) 0.05
0.1
0.15
0.2
signal
20
10
0 0
50
100
MOVAS lysate (µg)
Figure 16: Quantification and linear range of CaM signal from MOVAS. To determine the linear range of detection of CaM, a range of MOVAS lysate was probed with anti-CaM (Figure 15), and the signal quantified and plotted versus the loaded protein from MOVAS (open circles ○, error bars represent standard deviation; lower x-axis). Similarly, detection of rhCaM is plotted versus input (filled squares ■, error bars represent standard deviation; upper x-axis). The line indicates linear regression for the range of loaded protein from 10–50 µg, r2 = 0.956, n = 3.
96
Figure 17: Expression of CaM and IQGAP1 following serum stimulation. To determine whether CaM and IQGAP1 have cell-cycle dependent expression, protein lysate was collected following serum starved/stimulated at different time points. A) Immunoblotting results from IQGAP1, CaM and GAPDH. This is a representative blot from six replicates. Quantified expression levels of B) IQGAP1, C) CaM, and D) GAPDH protein relative to 0 h serum stimulation and standard deviations. There were no statistically significant differences between group means as determined by one-way ANOVA.
97
Figure 18: Reciprocal co-IP of CaM and IQGAP1. Lysate was collected from MOVAS serum starved/stimulated for 0, 8, 16 or 24 h. The protein was immunoprecipitated with A) anti-CaM or D) anti-IQGAP1 antibodies and protein G resin under both 2 mM EGTA (EGTA; black bars in B, C and E, F) or CaCl2 (Ca2+; white bars in B, C and E, F). These are representative blots from three replicates. Bar graphs B and C are means normalized to 0 h and standard deviations of quantified data of the precipitated CaM and co-IP IQGAP1, respectively. Bar graphs E and F are means normalized to 0 h and standard deviations quantified data of the precipitated IQGAP1 and
98
co-IP CaM, respectively. There were no statistically significant differences between group means as determined by one-way ANOVA.
99
Figure 19: Nucleofection of expression plasmids into MOVAS. MOVAS cells were electroporated with pmaxGFP expression plasmids using Amaxa Nucleofection Kit using the suggested programs. Each program was attempted twice. No GFP signal was detected in any program on 1 or 2 days after treatment despite cells presence, evident by Hoechst staining.
100
pCALM1-Cerulean sequencing Primer (Direction)
Sequence
pVenus-Iqgap1 sequencing Primer (Direction)
Sequence
N-CALM1 (+)
GATCAGCTGACCGAAGAACAG
N-IQGAP1 (+)
GCTCCGTCCTGGATAATGAG
N-CALM1 (−)
TGCAATTCAGCTTCTGTTGG
N-IQGAP1 (−)
GACTCCGTTTCTAAGGCCCT
C-CALM1 (+)
AGGCATTCCGAGTCTTTGAC
C-IQGAP1 (+)
GGTCTCCAAAAAGCCTAGGG
C-CALM1 (−)
TCCGTCTCCATCAATATCTGC
C-IQGAP1 (−)
AATTGGTTTGCCTGAAGGTC
T7 primer (+)
TAATACGACTCACTATAGGG
T7 terminator (−)
ACCCCTCAAGACCCGTTTAG
Table 1: Primer sequences. The primers used to sequence the two plasmids.
101
Figure 20: Plasmid maps of fluorescent fusion constructs. We received expression plasmids containing fluorophore-tagged proteins to use for transfection. The plasmids express a fusion construct of a fluorescent protein and one of our proteins of interest, calmodulin or IQGAP1. Maps were made by combining known plasmid backbone and cDNA sequences, and results from sequences analysis at joining sites. A) pTriEx-3 hCALM1-Cerulean map. Human CALM1 cDNA was fused 5′ to modified, enhanced cyan fluorescent protein, Cerulean gene. Fusion construct was inserted into pTriEx-3 backbone using NcoI and XhoI. Plasmids were prepared by transforming E. coli and expression selected for ampicillin LB agar. Plasmids were verified by restriction digests with NcoI and EcoRI. B) pEGFP-C1 Venus-mIqgap1 map. The EGFP gene from the plasmid was swapped out for modified, enhanced yellow fluorescent protein, Venus. Mouse Iqgap1 cDNA was inserted 3′ of the fluorescent protein at the multi-cloning site using HindIII and SalI. Plasmids were prepared by transforming E. coli and expression selected for kanamycin LB agar. Plasmids were verified by restriction digests with NcoI, XhoI and KpnI.
102
Figure 21: Excitation and emission spectra of the fluorophores. There is overlap of the excitation and emission spectra of the fluorophores that would be used in the proposed experiment that would combine Fucci and FRET signals. Fucci cell cycle indicator includes signal from Azami Green and Kusabira Orange. The FRET pair includes Cerulean and Venus protein tags. Excitation spectra are outlined in dashed lines (---). Emission spectra are outlined in solid lines and filled in (—). A. Images from SpectraViewer (Invitrogen). B. Image from Spectrum Viewer (BD Biosciences).
103
Figure 22: Nucleofection of expression plasmids into Fucci T1 mouse carotid SMC and imaging of Fucci reporters. Fucci T1 mouse carotid SMC were transfected using Amaxa Nucleofection kits for SMC. Cells were imaged using live cell microscopy. The cells were transfected with the listed plasmids. Untransfected are did not undergo a transfection protocol. CALM1-CFP is the pCALM1-Cerulean expression plasmid. YFP-Iqgap1 is the pVenus-Iqgap1 expression plasmid. A selection of the conditions is shown. Transfections were performed in triplicate.
104
Figure 23: Transfection and expression of fusion constructed in HEK293T. HEK293T cells were transfected with the indicated expression plasmid using the X-fect Transfection Reagent kit. Cells were visualized by live-cell imaging the indicated filters. pmaxGFP is a GFP expression plasmid used a positive control. CALM1-CFP is the pCALM1-Cerulean expression plasmid. YFP-Iqgap1 is the pVenus-Iqgap1 expression plasmid. Normal HEK293T that were not incubated with plasmids and transfection reagents (No Treatment) were used as a negative control. Signals from transfected cells indicate plasmids are successfully introduced into the cells and constructs are being expressed. Transfections were done in duplicate.
105
Figure 24: Detection of IQGAP1 and CaM in different cell lines. With the possibility of experiments moving away from MOVAS to simpler cell systems, the ability to detect CaM and IQGAP1 in various cell lines needed to be determined. IQGAP1 and CaM are easily detected in both HEK293T and NIH3T3 cell lines.
106
Figure 25: Detection of CAMK2D in different cell lines. There was no standing positive control for anti-CaM pulldown in the laboratory. CAMKII are activated by Ca2+-CaM, and CAMK2D is present in SMC [229, 230]. Preliminary blots were performed to test the presence of CAMK2D in various cell lines. NIH3T3 has the highest expression of CAMK2D out of the three cell lines tested. This is a representative blot from two replicates.
107
Figure 26: Detection of ERK in different cell lines. Actin precipitated in the anti-IQGAP1 pulldown tests (Figure 7) invalidating it as a positive control for co-IP. Since ERK2 was described as bound to IQGAP1 in unstimulated conditions [151] it is a putative positive control. Preliminary blots were performed to test the presence of ERK in various cell lines. Anti-ERK antibody yields two bands, one for each ERK isoform. This is a representative blot from two replicates.
108
References
1.
Bicknell KA, Surry EL, Brooks G. "Targeting the cell cycle machinery for the treatment of cardiovascular disease". J Pharm Pharmacol, 2003; 55(5): 571-91.
2.
Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O'Shaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, Jaeger JL, Kuntz RE, SIRIUS Investigators. "Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery". N Engl J Med, 2003; 349(14): 1315-23.
3.
Lemos PA, Serruys PW, Sousa JE. "Drug-eluting stents: cost versus clinical benefit". Circulation, 2003; 107(24): 3003-7.
4.
Stone GW, Ellis SG, Cox DA, Hermiller J, O'Shaughnessy C, Mann JT, Turco M, Caputo R, Bergin P, Greenberg J, Popma JJ, Russell ME, TAXUS-IV Investigators. "One-year clinical results with the slow-release, polymer-based, paclitaxel-eluting TAXUS stent: the TAXUS-IV trial". Circulation, 2004; 109(16): 1942-7.
5.
Sollott SJ, Cheng L, Pauly RR, Jenkins GM, Monticone RE, Kuzuya M, Froehlich JP, Crow MT, Lakatta EG, Rowinsky EK, Kinsella JL. "Taxol inhibits neointimal smooth muscle cell accumulation after angioplasty in the rat". J Clin Invest, 1995; 95(4): 1869– 76.
6.
Heldman AW, Cheng L, Jenkins GM, Heller PF, Kim D-W, Ware M, Jr, Nater C, Hruban RH, Rezai B, Abella BS, Bunge KE, Kinsella JL, Sollott SJ, Lakatta EG, Brinker JA, Hunter WL, Froehlich JP. "Paclitaxel stent coating inhibits neointimal hyperplasia at 4 weeks in a porcine model of coronary restenosis". Circulation, 2001; 103(18): 2289-95.
7.
Axel DI, Kunert W, Göggelmann C, Oberhoff M, Herdeg C, Küttner A, Wild DH, Brehm BR, Riessen R, Köveker G, Karsch KR. "Paclitaxel inhibits arterial smooth muscle cell proliferation and migration in vitro and in vivo using local drug delivery". Circulation, 1997; 96(2): 636-45.
8.
Kumar V, Cotran RS, Robbins SL, Basic Pathology. 7 ed. 2003, Philadelphia: Saunders.
9.
Karanam K, Kafri R, Loewer A, Lahav G. "Quantitative live cell imaging reveals a gradual shift between DNA repair mechanisms and a maximal use of HR in mid-S phase". Mol Cell, 2012; 47(2): 320-9.
10.
Branzei D, Foiani M. "The DNA damage response during DNA replication". Curr Opin Cell Biol, 2005; 17(6): 568-75.
11.
Lisby M, Rothstein R, Mortensen UH. "Rad52 forms DNA repair and recombination centersduring S phase". Proc Natl Acad Sci U S A, 2001; 98(15): 8276-82.
12.
Vaziri C, Saxena S, Jeon Y, Lee C, Murata K, Machida YJ, Wagle N, Hwang DS, Dutta A. "A p53-dependent checkpoint pathway prevents rereplication". Mol Cell, 2003; 11(4): 997-1008.
109
13.
Machida YJ, Hamlin JL, Dutta A. "Right place, right time, and only once: replication initiation in metazoans". Cell, 2005; 123(1): 13-24.
14.
Remus D, Diffley JF. "Eukaryotic DNA replication control: lock and load, then fire". Curr Opin Cell Biol, 2009; 21(6): 771-7.
15.
De Souza CPC, Osmani SA. "Mitosis, not just open or closed". Eukaryot Cell, 2007; 6(9): 1521-7.
16.
Chan GK, Liu S-T, Yen TJ. "Kinetochore structure and function". Trends Cell Biol, 2005; 15(11): 589-98.
17.
Blow JJ, Tanaka TU. "The chromosome cycle: coordinating replication and segreghation. Second in the cycles review series". EMBO Rep, 2005; 6(11): 1028-34.
18.
Glotzer M. "The molecular requirements for cytokinesis". Science, 2005; 307(5716): 1735-9.
19.
Albertson R, Riggs B, Sullivan W. "Membrane traffic: a driving force in cytokinesis". Trends Cell Biol, 2005; 15(2): 92-101.
20.
Whitfield JF, Boynton AL, MacManus JP, Sikorska M, Tsang BK. "The regulation of cell proliferation by calcium and cyclic AMP". Mol Cell Biochem, 1979; 27(3): 155-79.
21.
Blow JJ, Hodgson B. "Replication licensing – defining the proliferative state?". Trends Cell Biol, 2002; 12(2): 72-8.
22.
Hanahan D, Weinberg RA. "The hallmarks of cancer". Cell, 2000; 100(1): 57-70.
23.
Hanahan D, Weinberg RA. "Hallmarks of cancer: the next generation". Cell, 2011; 144(5): 646-74.
24.
Kirshenbaum LA, Abdellatif M, Chakraborty S, Schneider MD. "Human E2F-1 reactivates cell cycle progression in ventricular myocytes and represses cardiac gene transcription". Dev Biol, 1996; 179(2): 402-11.
25.
Thompson J, Manore M, Sheeshka J, Nutrition: A Functional Approach. Canadian ed. 2007, Toronto: Pearson Education Canada.
26.
Levin RM, Weiss B. "Binding of trifluoperazine to the calcium-dependent activator of cyclic nucleotide phosphodiesterase". Mol Pharmacol, 1977; 13(4): 690-7.
27.
Berridge MJ, Lipp P, Bootman MD. "The versatility and universality of calcium signalling". Nat Rev Mol Cell Biol, 2000; 1(1): 11-21.
28.
"Calmodulin—Ubiquitious Cell Rgulator". Lancet, 1982; 320(8291): 192-3.
29.
Rasmussen H, Goodman DB, Tenenhouse A. "The role of cyclic AMP and calcium in cell activation". CRC Crit Rev Biochem, 1972; 1(1): 95-148.
30.
Wolff DJ, Poirier PG, Brostrom CO, Brostrom MA. "Divalent cation binding properties of bovine brain Ca2+-dependent regulator protein". J Biol Chem, 1977; 252(12): 4180-17.
31.
Hickie RA, Wei J-W, Blyth LM, Wong DYW, Klaassen DJ. "Cations and calmodulin in normal and neoplastic cell growth regulation". Can J Biochem Cell Biol, 1983; 61(8): 934-41.
110
32.
Kahl CR, Means AR. "Regulation of cell cycle progression by calcium/calmodulindependent pathways". Endocr Rev, 2003; 24(6): 719-36.
33.
Cheung WY. "Cyclic 3',5'-nucleotide phosphodiesterase: Demonstration of an activator". Biochem Biophys Res Commun, 1970; 38(3): 533-8.
34.
Kakiuchi S, Yamazaki R. "Calcium dependent phosphodiesterase activity and its activating factor (PAF) from brain: studies on cyclic 3',5'-nucleotide phosphodiestease (III)". Biochem Biophys Res Commun, 1970; 41(5): 1104-10.
35.
Lin YM, Liu YP, Cheung WY. "Cyclic 3’:5’-nucleotide phosphodiesterase: purification, characterization, and active form of the protein activator from bovine brain". J Biol Chem, 1974; 249(15): 4943-54.
36.
Dedman JR, Potter JD, Jackson RL, Johnson JD, Means AR. "Physicochemical properties of rat testis Ca2+-dependent regulator protein of cyclic nucleotide phosphodiesterase: Relationship of Ca2+-binding, conformational changes, and phosphodiesterase activity". J Biol Chem, 1977; 252(23): 8415-22.
37.
Crouch TH, Klee CB. "Positive cooperative binding of calcium to bovine brain calmodulin". Biochemistry, 1980; 19(16): 3692-8.
38.
Chou JJ, Li S, Klee CB, Bax A. "Solution structure of Ca2+–calmodulin reveals flexible hand-like properties of its domains". Nat Struct Biol, 2001; 8(11): 990-7.
39.
LaPorte DC, Wierman BM, Storm DR. "Calcium-induced exposure of a hydrophobic surface on calmodulin". Biochemistry, 1980; 19(16): 3814-9.
40.
Liu YP, Cheung WY. "Cyclic 3':5'-nucleotide phosphodiesterase: Ca2+ confers more helical conformation to the protein activator". J Biol Chem, 1976; 251(14): 4193-8.
41.
Brostrom CO, Wolff DJ. "Calcium-dependent cyclic nucleotide phosphodiesterase from glial tumor cells". Arch Biochem Biophys, 1974; 165(2): 715-27.
42.
Nishida E, Kumagai H, Ohtsuki I, Sakai H. "The interactions between calcium-dependent regulator protein of cyclic nucleotide phosphodiesterase and microtubule proteins: I. Effect of calcium-dependent regulator protein on the calcium sensitivity of microtubule assembly". J Biochem, 1979; 85(5): 1257-66.
43.
Orosz F, Horváth I, Ovádi J. "New anti-mitotic drugs with distinct anti-calmodulin activity". Mini Rev Med Chem, 2006; 6(10): 1145-57.
44.
Babu YS, Sack JS, Greenhough TJ, Bugg CE, Means AR, Cook WJ. "Three-dimensional structure of calmodulin". Nature, 1985; 315(6014): 37-40.
45.
Davis TN, Urdea MS, Masiarz FR, Thorner J. "Isolation of the yeast calmodulin gene: Calmodulin is an essential protein". Cell, 1986; 47(3): 423-31.
46.
Boynton AL, Whitfield JF. "Calcium requirements for the proliferation of cells infected with a temperature-sensitive mutant of Rous sarcoma virus". Cancer Res, 1978; 38(5): 1237-40.
47.
Jones A, Boynton AL, MacManus JP, Whitfield JF. "Ca-calmodulin mediates the DNAsynthetic response of calcium-deprived liver cells to the tumor promoter TPA". Exp Cell Res, 1982; 138(1): 87-93.
111
48.
Swierenga SHH, Whitfield JF, Boynton AL, MacManus JP, Rixon RH, Sikorska M, Tsang BK, Walker PR. "Regulation of proliferation of normal and neoplastic rat liver cells by calcium and cyclic AMP". Ann N Y Acad Sci, 1980; 349: 294-311.
49.
Rixon RH. "Mitotic activity in the bone marrow of rats and its relation to the level of plasma calcium". Curr Mod Biol, 1968; 2(2): 68-74.
50.
Perris AD, MacManus JP, Whitfield JF, Weiss LA. "Parathyroid glands and mitotic stimulation in rat bone marrow after hemorrhage". Am J Physiol, 1971; 220(3): 773-8.
51.
Perris AD, Whitfield JF, Rixon RH. "Stimulation of mitosis in bone marrow and thymus of normal and irradiated rats by divalent cations and parathyroid extract". Radiat Res, 1967; 32(3): 550-63.
52.
Gallien-Lartigue O. "Calcium and ionophore A-23187 as initiators of DNA replication in the pluripotent haemopoietic stem cell". Cell Tissue Kinet, 1976; 9(6): 533-40.
53.
Perris AD, Whitfield JF, Tölg PK. "Role of calcium in control of growth and cell division". Nature, 1968; 219(5153): 527-9.
54.
Wei J-W, Morris HP, Hickie RA. "Positive correlation between calmodulin content and hepatoma growth rates". Cancer Res, 1982; 42(7): 2571-4.
55.
Watterson DM, Van Eldik LJ, Smith RE, Vanaman TC. "Calcium-dependent regulatory protein of cyclic nucleotide metabolism in normal and transformed chicken embryo fibroblasts". Proc Natl Acad Sci U S A, 1976; 73(8): 2711-5.
56.
MacManus JP, Braceland BM, Rixon RH, Whitfield JF, Morris HP. "An increase in calmodulin during growth of normal and cancerous liver in vivo". FEBS Lett, 1981; 133(1): 99-102.
57.
LaPorte DC, Gidwitz S, Weber MJ, Storm DR. "Relationship between changes in the calcium dependent regulatory protein and adenylate cyclase during viral transformation". Biochem Biophys Res Commun, 1979; 86(4): 1169-77.
58.
Chafouleas JG, Pardue RL, Brinkley BR, Dedman JR, Means AR. "Regulation of intracellular levels of calmodulin and tubulin in normal and transformed cells". Proc Natl Acad Sci U S A, 1981; 78(2): 996-1000.
59.
Marcum JM, Dedman JR, Brinkley BR, Means AR. "Control of microtubule assemblydisassembly by calcium-dependent regulator protein". Proc Natl Acad Sci U S A, 1978; 75(8): 3771-5.
60.
Cheung WY. "Calmodulin plays a pivotal role in cellular regulation". Science, 1980; 207(4426): 19-27.
61.
Wei J-W, Hickie RA. "Increased content of calmodulin in Morris hepatoma 5123 t.c.(h)". Biochem Biophys Res Commun, 1981; 100(4): 1562-8.
62.
Wei J-W, Hickie RA, Klaassen DJ. "Inhibition of human breast cancer colony formation by anticalmodulin agents: trifluoperazine, W-7, and W-13". Cancer Chemother Pharmacol, 1983; 11(2): 86-90.
63.
Horwitz SB, Chia GH, Harracksingh C, Orlow S, Pifko-Hirst S, Schneck J, Sorbara L, Speaker M, Wilk EW, Rosen OM. "Trifluoperazine inhibits phagocytosis in a macrophagelike cultured cell line". J Cell Biol, 1981; 91(3 Pt 1): 798-802.
112
64.
Sasaki Y, Hidaka H. "Calmodulin and cell proliferation". Biochem Biophys Res Commun, 1982; 104(2): 451-6.
65.
Chafouleas JG, Bolton WE, Hidaka H, Boyd AE, III, Means AR. "Calmodulin and the cell cycle: Involvement in regulation of the cell-cycle progression". Cell, 1982; 28(1): 4150.
66.
Colomer J, Agell N, Engel P, Alberola-Ila J, Bachs O. "Calmodulin expression during proliferation activation of human T lymphocytes". Cell Calcium, 1993; 14(8): 609-18.
67.
Agell N, Pujol MJ, López-Girona A, Bosch M, Rosa JL, Bachs O. "Calmodulin expression during rat liver regeneration.". Hepatology, 1994; 20(4 Pt 1): 1002-8.
68.
Chafouleas JG, Lagacé L, Bolton WE, Boyd AE, III, Means AR. "Changes in calmodulin and its mRNA accompany reentry of quiescent (G0) cells into the cell cycle". Cell, 1984; 36(1): 73-81.
69.
Rasmussen CD, Means AR. "Calmodulin is involved in regulation of cell proliferation". EMBO J, 1987; 6(13): 3961-8.
70.
Rasmussen CD, Means AR. "Calmodulin is required for cell-cycle progression during G1 and mitosis". EMBO J, 1989; 8(1): 73-82.
71.
Takuwa N, Zhou W, Takuwa Y. "Calcium, calmodulin and cell cycle progression". Cell Signal, 1995; 7(2): 93-104.
72.
Taulés M, Rius E, Talaya D, López-Girona A, Bachs O, Agell N. "Calmodulin is essential for cyclin-dependent kinase 4 (Cdk4) activity and nuclear accumulation of cyclin D1-Cdk4 during G1". J Biol Chem, 1998; 273(50): 33279-86.
73.
Taulés M, Rodrı́guez-Vilarrupla A, Rius E, Estanyol JM, Casanovas O, Sacks DB, PérezPayá E, Bachs O, Agell N. "Calmodulin binds to p21Cip1 and is involved in the regulation of its nuclear localization". J Biol Chem, 1999; 274(35): 24445-8.
74.
Takuwa N, Zhou W, Kumada M, Takuwa Y. "Ca2+-dependent stimulation of retinoblastoma gene product phosphorylation and p34cdc2 kinase activation in serumstimulated human fibroblasts". J Biol Chem, 1993; 268(1): 138-45.
75.
Plochocka-Zulinska D, Rasmussen G, Rasmussen CD. "Regulation of calcineurin gene expression in Schizosaccharomyces pombe: dependence on the ste11 transcription factor". J Biol Chem, 1995; 270(42): 24794-9.
76.
Khanna A, Hosenpud J. "Cyclosporine induces the expression of the cyclin inhibitor p21". Transplantation, 1999; 67(9): 1262-8.
77.
Tomono M, Toyoshima K, Ito M, Amano H, Kiss Z. "Inhibitors of calcineurin block expression of cyclins A and E induced by fibroblast growth factor in Swiss 3T3 fibroblasts". Arch Biochem Biophys, 1998; 353(2): 374-8.
78.
Kahl CR, Means AR. "Calcineurin regulates cyclin D1 accumulation in growthstimulated fibroblasts". Mol Biol Cell, 2004; 15(4): 1833-42.
79.
Baksh S, Widlund HR, Frazer-Abel AA, Du J, Fosmire S, Fisher DE, DeCaprio JA, Modiano JF, Burakoff SJ. "NFATc2-mediated repression of cyclin-dependent kinase 4 expression". Mol Cell, 2002; 10(5): 1071-81.
113
80.
Schneider G, Oswald F, Wahl C, Greten FR, Adler G, Schmid RM. "Cyclosporine inhibits growth through the activating transcription factor/cAMP-responsive elementbinding protein binding site in the cyclin D1 promoter". J Biol Chem, 2002; 277(46): 43599-607.
81.
Ohta Y, Ohba T, Miyamoto E. "Ca2+/calmodulin-dependent protein kinase II: localization in the interphase nucleus and the mitotic apparatus of mammalian cells". Proc Natl Acad Sci U S A, 1990; 87(14): 5341-5.
82.
Alemany V, Sanchez-Piris M, Bachs O, Aligue R. "Cmk2, a novel serine/threonine kinase in fission yeast". FEBS Lett, 2002; 524(1-3): 79-86.
83.
Minami H, Inoue S, Hidaka H. "The effect of KN-62, Ca2+/calmodulin dependent protein kinase II inhibitor on cell cycle". Biochem Biophys Res Commun, 1994; 199(1): 241-8.
84.
Planas-Silva MD, Means AR. "Expression of a constitutive form of calcium/calmodulin dependent protein kinase II leads to arrest of the cell cycle in G2.". EMBO J, 1992; 11(2): 507-17.
85.
Baitinger C, Alderton J, Poenie M, Schulman H, Steinhardt R. "Multifunctional Ca2+/calmodulin-dependent protein kinase is necessary for nuclear envelope breakdown.". J Cell Biol, 1990; 111(5 Pt 1): 1763-73.
86.
Patel R, Holt M, Philipova R, Moss S, Schulman H, Hidaka H, Whitaker M. "Calcium/calmodulin-dependent phosphorylation and activation of human Cdc25-C at the G2/M phase transition in HeLa cells". J Biol Chem, 1999; 274(12): 7958-68.
87.
Sugiura R, Sio SO, Shuntoh H, Kuno T. "Calcineurin phosphatase in signal transduction: lessons from fission yeast". Genes Cells, 2002; 7(7): 619-27.
88.
Yoshida T, Toda T, Yanagida M. "A calcineurin-like gene ppb1+ in fission yeast: mutant defects in cytokinesis, cell polarity, mating and spindle pole body positioning". J Cell Sci, 1994; 107(Pt 7): 1725-35.
89.
Serratosa J, Pujol MJ, Bachs O, Carafoli E. "Rearrangement of nuclear calmodulin during proliferative liver cell activation.". Biochem Biophys Res Commun, 1988; 150(3): 1162-9.
90.
Hammond RA, Foster KA, Berchthold MW, Gassmann M, Holmes AM, Hübscher U, Brown NC. "Calcium-dependent calmodulin-binding proteins associated with mammalian DNA polymerase alpha". Biochim Biophys Acta, 1988; 951(2-3): 315-21.
91.
Rixon RH, Isaacs RJ, Whitfield JF. "Control of DNA polymerase-alpha activity in regenerating rat liver by calcium and 1 alpha,25(OH)2D3". J Cell Physiol, 1989; 139(2): 354-60.
92.
Dai G, Qian Y, Chen J, Meng F-L, Pan F-Y, Shen W-G, Zhang S-Z, Xue B, Li C-J. "Calmodulin activation of polo-like kinase 1 is required during mitotic entry". Biochem Cell Biol, 2013; 91(5): 287-94.
93.
Means AR, Dedman JR. "Calmodulin-an intracellular calcium receptor". Nature, 1980; 285(5760): 73-7.
94.
Mallampalli RK, Glasser JR, Coon TA, Chen BB. "Calmodulin protects Aurora B on the midbody to regulate the fidelity of cytokinesis". Cell Cycle, 2013; 12(4): 663–73.
114
95.
Chen BB, Glasser JR, Coon TA, Mallampalli RK. "F-box protein FBXL2 exerts human lung tumor suppressor-like activity by ubiquitin-mediated degradation of cyclin D3 resulting in cell cycle arrest". Oncogene, 2012; 31(20): 2566-79.
96.
Chen BB, Glasser JR, Coon TA, Mallampalli RK. "FBXL2 is a ubiquitin E3 ligase subunit that triggers mitotic arrest". Cell Cycle, 2011; 10(20): 3487-94.
97.
Kearsey SE, Cotterrill S. "Enigmatic variations: divergent modes of regulating eukaryotic DNA replication". Mol Cell, 2003; 12(5): 1067-75.
98.
Machesky LM. "Cytokinesis: IQGAPs find a function". Curr Biol, 1998; 8(6): R202-R5.
99.
Weissbach L, Settleman J, Kalady MF, Snijders AJ, Murthy AE, Yan Y-X, Bernards A. "Identifacation of a human RasGAP-related protein containing calmodulin-binding motifs". J Biol Chem, 1994; 269(32): 20517-21.
100.
Weissbach L, Bernards A, Herion DW. "Binding of myosin essential light chain to the cytoskeleton-associated protein IQGAP1". Biochem Biophys Res Commun, 1998; 251(1): 269-76.
101.
Brill S, Li S, Lyman CW, Church DM, Wasmuth JJ, Weissbach L, Bernards A, Snijders AJ. "The Ras GTPase-activating-protein-related human protein IQGAP2 harbors a potential actin binding domain and interacts with calmodulin and Rho family GTPases". Mol Cell Biol, 1996; 16(9): 4869-78.
102.
Wang S, Watanabe T, Noritake J, Fukata Y, Yoshimura T, Itoh N, Harada T, Nakagawa M, Matsuura Y, Arimura N, Kaibuchi K. "IQGAP3, a novel effector of Rac1 and Cdc42, regulates neurite outgrowth". J Cell Sci, 2007; 120(Pt 4): 567-77.
103.
Hart MJ, Callow MG, Souza B, Polakis P. "IQGAP1, a calmodulin-binding protein with a rasGAP-related domain, is a potential effector for cdc42Hs". EMBO J, 1996; 15(12): 2997-3005.
104.
Kuroda S, Fukata M, Kobayashi K, Nakafuku M, Nomura N, Iwamatsu A, Kaibuchi K. "Identification of IQGAP as a putative target for the small GTPases, Cdc42 and Rac1". J Biol Chem, 1996; 271(38): 23363-7.
105.
McCallum SJ, Wu WJ, Cerione RA. "Identification of a putative effector for Cdc42Hs with high sequence similarity to the RasGAP-related protein IQGAP1 and a Cdc42Hs binding partner with similarity to IQGAP2". J Biol Chem, 1996; 271(36): 21732-7.
106.
Kurella VB, Richard JM, Parke CL, LeCour LF, Jr, Bellamy HD, Worthylake DK. "Crystal structure of the GTPase-activating protein-related domain from IQGAP1". J Biol Chem, 2009; 284(22): 14857-65.
107.
Swart-Mataraza JM, Li Z, Sacks DB. "IQGAP1 is a component of Cdc42 signaling to the cytoskeleton". J Biol Chem, 2002; 277(27): 24753-63.
108.
Fukata M, Kuroda S, Fujii K, Nakamura T, Shoji I, Matsuura Y, Okawa K, Iwamatsu A, Kikuchi A, Kaibuchi K. "Regulation of cross-linking of actin filament by IQGAP1, a target for Cdc42". J Biol Chem, 1997; 272(47): 29579-83.
109.
Erickson JW, Cerione RA, Hart MJ. "Identification of an actin cytoskeletal complex that includes IQGAP and the Cdc42 GTPase". J Biol Chem, 1997; 272(39): 24443-7.
115
110.
Grohmanova K, Schlaepfer D, Hess D, Gutierrez P, Beck M, Kroschewski R. "Phosphorylation of IQGAP1 modulates its binding to Cdc42, revealing a new type of Rho-GTPase regulator". J Biol Chem, 2004; 279(47): 48495-504.
111.
Bashour A-M, Fullerton AT, Hart MJ, Bloom GS. "IQGAP1, a Rac- and Cdc42-binding protein, directly binds and cross-links microfilaments". J Cell Biol, 1997; 137(7): 155566.
112.
Mateer SC, Morris LE, Cromer DA, Benseñor LB, Bloom GS. "Actin filament binding by a monomeric IQGAP1 fragment with a single calponin homology domain". Cell Motil Cytoskeleton, 2004; 58(4): 231-41.
113.
Ren J-G, Li Z, Crimmins DL, Sacks DB. "Self-association of IQGAP1: Characterization and functional sequelae". J Biol Chem, 2005; 280(41): 34548-57.
114.
Fukata M, Nakagawa M, Itoh N, Kawajiri A, Yamaga M, Kuroda S, Kaibuchi K. "Involvement of IQGAP1, an effector of Rac1 and Cdc42 GTPases, in cell-cell dissociation during cell scattering". Mol Cell Biol, 2001; 21(6): 2165-83.
115.
Kuroda S, Fukata M, Nakagawa M, Fujii K, Nakamura T, Ookubo T, Izawa I, Nagase T, Nomura N, Hideki Tani, Shoji I, Matsuura Y, Yonehara S, Kaibuchi K. "Role of IQGAP1, a target of the small GTPases Cdc42 and Rac1, in regulation of E-cadherinmediated cell-cell adhesion". Science, 1998; 281(5378): 832-5.
116.
Li Z, Kim SH, Higgins JMG, Brenner MB, Sacks DB. "IQGAP1 and calmodulin modulate E-cadherin function". J Biol Chem, 1999; 274(53): 37885-92.
117.
Yamaoka-Tojo M, Tojo T, Kim HW, Hilenski L, Patrushev NA, Zhang L, Fukai T, Ushio-Fukai M. "IQGAP1 mediates VE-cadherin–based cell–cell contacts and VEGF signaling at adherence junctions linked to angiogenesis". Arterioscler Thromb Vasc Biol, 2006; 26(9): 1991-7.
118.
Sharma M, Henderson BR. "IQ-domain GTPase-activating protein 1 regulates -Catenin at membrane ruffles and its role in macropinocytosis of N-cadherin and adenomatous polyposis coli". J Biol Chem, 2007; 282(11): 8545–56.
119.
Pathmanathan S, Elliott SF, McSwiggen S, Greer B, Harriott P, Irvine GB, Timson DJ. "IQ motif selectivity in human IQGAP1: binding of myosin essential light chain and S100B ". Mol Cell Biochem, 2008; 318(1-2): 43-51.
120.
Havugimana PC, Hart GT, Nepusz T, Yang H, Turinsky AL, Li Z, Wang PI, Boutz DR, Fong V, Phanse S, Babu M, Craig SA, Hu P, Wan C, Vlasblom J, Dar V-u-N, Bezginov A, Clark GW, Wu GC, Wodak SJ, Tillier ERM, Paccanaro A, Marcotte EM, Emili A. "A census of human soluble protein complexes". Cell, 2012; 150(5): 1068-81.
121.
Fukata M, Watanabe T, Noritake J, Nakagawa M, Yamaga M, Kuroda S, Matsuura Y, Iwamatsu A, Perez F, Kaibuchi K. "Rac1 and Cdc42 capture microtubules through IQGAP1 and CLIP-170". Cell, 2002; 109(7): 873-85.
122.
Kanwar N, Wilkins JA. "IQGAP1 involvement in MTOC and granule polarization in NK-cell cytotoxicity". Eur J Immunol, 2011; 41(9): 2763-73.
123.
Brandt DT, Marion S, Griffiths G, Watanabe T, Kaibuchi K, Grosse R. "Dia1 and IQGAP1 interact in cell migration and phagocytic cup formation". J Cell Biol, 2007; 178(2): 193-200.
116
124.
Daou P, Hasan S, Breitsprecher D, Baudelet E, Camoin L, Audebert S, Goode BL, Badache A. "Essential and nonredundant roles for Diaphanous formins in cortical microtubule capture and directed cell migration". Mol Biol Cell, 2014; 25(5): 658-68.
125.
Humphries JD, Byron A, Bass MD, Craig SE, Pinney JW, Knight D, Humphries MJ. "Proteomic analysis of integrin-associated complexes identifies RCC2 as a dual regulator of Rac1 and Arf6.". Sci Signal, 2009; 2(87): ra51.
126.
Byron A, Humphries JD, Craig SE, Knight D, Humphries MJ. "Proteomic analysis of α4β1 integrin adhesion complexes reveals α-subunit-dependent protein recruitment". Proteomics, 2012; 12(13): 2107-14.
127.
Jadeski L, Mataraza JM, Jeong H-W, Li Z, Sacks DB. "IQGAP1 stimulates proliferation and enhances tumorigenesis of human breast epithelial cells". J Biol Chem, 2008; 283(2): 1008-17.
128.
Krishnan S, Fernandez GE, Sacks DB, Prasadarao NV. "IQGAP1 mediates the disruption of adherens junctions to promote Escherichia coli K1 invasion of brain endothelial cells". Cell Microbiol, 2012; 14(9): 1415-33.
129.
Kim H, White CD, Li Z, Sacks DB. "Salmonella enterica serotype Typhimurium usurps the scaffold protein IQGAP1 to manipulate Rac1 and MAPK signalling". Biochem J, 2011; 440(3): 309-18.
130.
Brown MD, Bry L, Li Z, Sacks DB. "Actin pedestal formation by enteropathogenic Escherichia coli is regulated by IQGAP1, calcium, and calmodulin". J Biol Chem, 2008; 283(50): 35212-22.
131.
Brown MD, Bry L, Li Z, Sacks DB. "IQGAP1 regulates Salmonella invasion through interactions with actin, Rac1, and Cdc42". J Biol Chem, 2007; 282(41): 30265-72.
132.
Conlin VS, Curtis SB, Zhao Y, Moore EDW, Smith VC, Meloche RM, Finlay BB, Buchan AMJ. "Helicobacter pylori infection targets adherens junction regulatory proteins and results in increased rates of migration in human gastric epithelial cells". Infect Immun, 2004; 72(9): 5181-92.
133.
Buss C, Müller D, Rüter C, Heusipp G, Schmidt MA. "Identification and characterization of Ibe, a novel type III effector protein of A/E pathogens targeting human IQGAP1". Cell Microbiol, 2009; 11(4): 661-77.
134.
Osman MA, Bloom GS, Tagoe EA. "Helicobacter pylori-induced alteration of epithelial cell signaling and polarity: a possible mechanism of gastric carcinoma etiology and disparity". Cytoskeleton (Hoboken), 2013; 70(7): 349-59.
135.
Lu J, Qu Y, Liu Y, Jambusaria R, Han Z, Ruthel G, Freedman BD, Harty RN. "Host IQGAP1 and Ebola virus VP40 interactions facilitate virus-like particle egress". J Virol, 2013; 87(13): 7777-80.
136.
Chung LK, Philip NH, Schmidt VA, Koller A, Strowig T, Flavell RA, Brodsky IE, Bliska JB. "IQGAP1 is important for activation of caspase-1 in macrophages and is targeted by Yersinia pestis type III effector YopM". mBio, 2014; 5(4): e01402-e14.
137.
Jeong H-W, Li Z, Brown MD, Sacks DB. "IQGAP1 binds Rap1 and modulates its activity". J Biol Chem, 2007; 282(28): 20752-62.
117
138.
Hu B, Shi B, Jarzynka MJ, Yiin J-J, D'Souza-Schorey C, Cheng S-Y. "ADP-ribosylation factor 6 regulates glioma cell invasion through the IQ-domain GTPase-activating protein 1-Rac1-mediated pathway". Cancer Res, 2009; 69(3): 794-801.
139.
Matsunaga H, Kubota K, Inoue T, Isono F, Ando O. "IQGAP1 selectively interacts with K-Ras but not with H-Ras and modulates K-Ras function". Biochem Biophys Res Commun, 2014; 444(3): 360-4.
140.
Jamieson C, Sharma M, Henderson BR. "Wnt signaling from membrane to nucleus: catenin caught in a loop". Int J Biochem Cell Biol, 2012; 44(6): 847–50.
141.
Wu X, Tu X, Joeng KS, Hilton MJ, Williams DA, Long F. "Rac1 activation controls nuclear localization of β-catenin during canonical Wnt signaling". Cell, 2008; 133(2): 340-53.
142.
Zhu G, Wang Y, Huang B, Liang J, Ding Y, Xu A, Wu W. "A Rac1/PAK1 cascade controls -catenin activation in colon cancer cells". Oncogene, 2012; 31(8): 1001-12.
143.
Wang Y, Wang A, Wang F, Wang M, Zhu M, Ma Y, Wu R. "IQGAP1 activates Tcf signal independent of Rac1 and Cdc42 in injury and repair of bronchial epithelial cells". Exp Mol Pathol, 2008; 85(2): 122-8.
144.
Watanabe T, Wang S, Noritake J, Sato K, Fukata M, Takefuji M, Nakagawa M, Izumi N, Akiyama T, Kaibuchi K. "Interaction with IQGAP1 links APC to Rac1, Cdc42, and actin filaments during cell polarization and migration". Dev Cell, 2004; 7(6): 871-83.
145.
Goto T, Sato A, Shimizu M, Adachi S, Satoh K, Iemura S, Natsume T, Shibuya H. "IQGAP1 functions as a modulator of Dishevelled nuclear localization in Wnt signaling". PLoS ONE, 2013; 8(4): e60865.
146.
Gan X-q, Wang J-y, Xi Y, Wu Z-l, Li Y-p, Li L. "Nuclear Dvl, c-Jun, β-catenin, and TCF form a complex leading to stabiLization of β-catenin–TCF interaction". J Cell Biol, 2008; 180(6): 1087-100.
147.
Carmon KS, Gong X, Yi J, Thomas A, Liu Q. "RSPO–LGR4 functions via IQGAP1 to potentiate Wnt signaling". Proc Natl Acad Sci U S A, 2014; 111(13): E1221-E9.
148.
Ewen ME. "Relationship between Ras pathways and cell cycle control". Prog Cell Cycle Res, 2000; 4: 1-17.
149.
Ren J-G, Li Z, Sacks DB. "IQGAP1 modulates activation of B-Raf". Proc Natl Acad Sci U S A, 2007; 104(25): 10465-9.
150.
Roy M, Li Z, Sacks DB. "IQGAP1 is a scaffold for mitogen-activated protein kinase signaling". Mol Cell Biol, 2005; 25(18): 7940-52.
151.
Roy M, Li Z, Sacks DB. "IQGAP1 Binds ERK2 and Modulates Its Activity". J Biol Chem, 2004; 279(17): 17329-37.
152.
Bourguignon LYW, Gilad E, Rothman K, Peyrollier K. "Hyaluronan-CD44 interaction with IQGAP1 promotes Cdc42 and ERK signaling, leading to actin binding, Elk1/estrogen receptor transcriptional activation, and ovarian cancer progression". J Biol Chem, 2005; 280(12): 11961-72.
153.
Yamaoka-Tojo M, Ushio-Fukai M, Hilenski L, Dikalov SI, Chen YE, Tojo T, Fukai T, Fujimoto M, Patrushev NA, Wang N, Kontos CD, Bloom GS, Alexander RW. "IQGAP1,
118
a novel vascular endothelial growth factor receptor binding protein, is involved in reactive oxygen species–dependent endothelial migration and proliferation". Circ Res, 2004 95(3): 276-83. 154.
Kaplan N, Urao N, Furuta E, Kim S-J, Razvi M, Nakamura Y, McKinney RD, Poole LB, Fukai T, Ushio-Fukai M. "Localized cysteine sulfenic acid formation by vascular endothelial growth factor: role in endothelial cell migration and angiogenesis". Free Radic Res, 2011; 45(10): 1124-35.
155.
McNulty DE, Li Z, White CD, Sacks DB, Annan RS. "MAPK scaffold IQGAP1 binds the EGF receptor and modulates its activation". J Biol Chem, 2011; 286(17): 15010–21.
156.
White CD, Li Z, Dillon DA, Sacks DB. "IQGAP1 protein binds human epidermal growth factor receptor 2 (HER2) and modulates trastuzumab resistance". J Biol Chem, 2011; 286(34): 29734-47.
157.
Liu C, Billadeau DD, Abdelhakim H, Leof E, Kaibuchi K, Bernabeu C, Bloom GS, Yang L, Boardman L, Shah VH, Kang N. "IQGAP1 suppresses TβRII-mediated myofibroblastic activation and metastatic growth in liver". J Clin Invest, 2013; 123(3): 1138-56.
158.
Chen F, Zhu H-H, Zhou L-F, Wu S-S, Wang J, Chen Z. "IQGAP1 is overexpressed in hepatocellular carcinoma and promotes cell proliferation by Akt activation". Exp Mol Med, 2010; 42(7): 477-83.
159.
Tekletsadik YK, Sonn R, Osman MA. "A conserved role of IQGAP1 in regulating TOR complex 1". J Cell Sci, 2012; 125(Pt 8): 2041-52.
160.
Rittmeyer EN, Daniel S, Hsu S-C, Osman MA. "A dual role for IQGAP1 in regulating exocytosis". J Cell Sci, 2008; 121(Pt 3): 391-403.
161.
Wang J-B, Sonn R, Tekletsadik YK, Samorodnitsky D, Osman MA. "IQGAP1 regulates cell proliferation through a novel CDC42-mTOR pathway". J Cell Sci, 2009; 122(Pt 12): 2024-33.
162.
Li Z, McNulty DE, Marler KJM, Lim L, Hall C, Annan RS, Sacks DB. "IQGAP1 promotes neurite outgrowth in a phosphorylation-dependent manner". J Biol Chem, 2005; 280(14): 13871-8.
163.
Johnson M, Sharma M, Brocardo MG, Henderson BR. "IQGAP1 translocates to the nucleus in early S-phase and contributes to cell cycle progression after DNA replication arrest". Int J Biochem Cell Biol, 2011; 43(1): 65-73.
164.
Johnson MA, Sharma M, Mok MTS, Henderson BR. "Stimulation of in vivo nuclear transport dynamics of actin and its co-factors IQGAP1 and Rac1 in response to DNA replication stress". Biochim Biophys Acta, 2013; 1833(10): 2334-47.
165.
Maréchal A, Li J-M, Ji XY, Wu C-S, Yazinski SA, Nguyen HD, Liu S, Jiménez AE, Jin J, Zou L. "PRP19 transforms into a sensor of RPA-ssDNA after DNA damage and drives ATR activation via a ubiquitin-mediated circuitry". Mol Cell, 2014; 53(2): 235-546.
166.
Stark C, Breitkreutz B-J, Reguly T, Boucher L, Breitkreutz A, Tyers M. "BioGRID: a general repository for interaction datasets". Nucleic Acids Res, 2006; 34(Database issue): D535-D9.
119
167.
Kristensen AR, Gsponer J, Foster LJ. "A high-throughput approach for measuring temporal changes in the interactome.". Nat Methods, 2012; 9(9): 907-9.
168.
Nakajima E, Suzuki K, Takahashi K. "Mitotic dissociation of IQGAP1 from Rac-bound 1-integrin is mediated by protein phosphatase 2A". Biochem Biophys Res Commun, 2005; 326(1): 249-53.
169.
Ko N, Nishihama R, Tully GH, Ostapenko D, Solomon MJ, Morgan DO, Pringle JR. "Identification of yeast IQGAP (Iqg1p) as an anaphase-promoting-complex substrate and its role in actomyosin-ring-independent cytokinesis". Mol Biol Cell, 2007; 18(12): 513953.
170.
Epp JA, Chant J. "An IQGAP-related protein controls actin-ring formation and cytokinesis in yeast". Curr Biol, 1997; 7(12): 921-9.
171.
Lee S, Escalante R, Firtel RA. "A Ras GAP is essential for cytokinesis and spatial patterning in Dictyostelium". Development, 1997; 124(5): 983-96.
172.
Adachi H, Takahashi Y, Hasebe T, Shirouzu M, Yokoyama S, Sutoh K. "Dictyostelium IQGAP-related protein specifically involved in the completion of cytokinesis". J Cell Biol, 1997; 137(4): 891-8.
173.
Shannon KB, Li R. "The multiple roles of Cyk1p in the assembly and function of the actomyosin ring in budding yeast". Mol Biol Cell, 1999; 10(2): 283-96.
174.
Li S, Wang Q, Chakladar A, Bronson RT, Bernards A. "Gastric hyperplasia in mice lacking the putative Cdc42 effector IQGAP1". Mol Cell Biol, 2000; 20(2): 697-701.
175.
Schmidt VA, Chiariello CS, Capilla E, Miller F, Bahou WF. "Development of hepatocellular carcinoma in Iqgap2-deficient mice is IQGAP1 dependent". Mol Cell Biol, 2008; 28(5): 1489–502.
176.
Ostrower ST, Fischer TC, Smith RV, Belbin TJ, Bent JP, Parikh SR. "Gene expression in the oropharynx of children exposed to secondhand smoke". Laryngoscope, 2010; 120(12): 2467-72.
177.
Nojima H, Adachi M, Matsui T, Okawa K, Tsukita S, Tsukita S. "IQGAP3 regulates cell proliferation through the Ras/ERK signalling cascade". Nat Cell Biol, 2008; 10(8): 971-8.
178.
Kunimoto K, Nojima H, Yamazaki Y, Yoshikawa T, Okanoue T, Tsukita S. "Involvement of IQGAP3, a regulator of Ras/ERK-related cascade, in hepatocyte proliferation in mouse liver regeneration and development". J Cell Physiol, 2009; 220(3): 621-31.
179.
Mbele GO, Deloulme JC, Gentil BJ, Delphin C, Ferro M, Garin J, Takahashi M, Baudier J. "The zinc- and calcium-binding S100B interacts and co-localizes with IQGAP1 during dynamic rearrangement of cell membranes". J Biol Chem, 2002; 277(51): 49998-50007.
180.
Heil A, Nazmi AR, Koltzscher M, Poeter M, Austermann J, Assard N, Baudier J, Kaibuchi K, Gerke V. "S100P is a novel interaction partner and regulator of IQGAP1". J Biol Chem, 2011; 286(9): 7227-38.
181.
Li Z, Sacks DB. "Elucidation of the interaction of calmodulin with the IQ motifs of IQGAP1". J Biol Chem, 2003; 278(6): 4347–52.
120
182.
Joyal JL, Annan RS, Ho Y-D, Huddleston ME, Carr SA, Hart MJ, Sacks DB. "Calmodulin modulates the interaction between IQGAP1 and Cdc42: Identification of IQGAP1 by nanoelectrospray tandem mass spectrometry". J Biol Chem, 1997; 272(24): 15419-25.
183.
Jang D-J, Ban B, Lee J-A. "Characterization of novel calmodulin binding domains within IQ Motifs of IQGAP1". Mol Cells, 2011; 32(6): 511-8.
184.
Ho Y-D, Joyal JL, Li Z, Sacks DB. "IQGAP1 integrates Ca2+/calmodulin and Cdc42 signaling". J Biol Chem, 1999; 274(1): 464-740.
185.
Briggs MW, Li Z, Sacks DB. "IQGAP1-mediated stimulation of transcriptional coactivation by -catenin is modulated by calmodulin". J Biol Chem, 2002; 277(9): 745365.
186.
Mateer SC, McDaniel AE, Nicolas V, Habermacher GM, Lin M-JS, Cromer DA, King ME, Bloom GS. "The mechanism for regulation of the F-actin binding activity of IQGAP1 by calcium/calmodulin". J Biol Chem, 2002; 277(14): 12324-33.
187.
Ren J-G, Li Z, Sacks DB. "IQGAP1 integrates Ca2+/calmodulin and B-Raf signaling". J Biol Chem, 2008; 283(34): 22972-82.
188.
Psatha MI, Razi M, Koffer A, Moss SE, Sacks DB, Bolsover SR. "Targeting of calcium: calmodulin signals to the cytoskeleton by IQGAP1". Cell Calcium, 2007; 41(6): 593-605.
189.
Vinereanu D. "Risk factors for atherosclerotic disease: present and future". Herz, 2006; 31 Suppl 3: 5-24.
190.
Rivard A, Andrés V. "Vascular smooth muscle cell proliferation in the pathogenesis of atherosclerotic cardiovascular diseases". Histol Histopathol, 2000; 15(2): 557-71.
191.
Finn AV, Nakazawa G, Joner M, Kolodgie FD, Mont EK, Gold HK, Virmani R. "Vascular responses to drug eluting stents: importance of delayed healing". Arterioscler Thromb Vasc Biol, 2007; 27(7): 1500-10.
192.
Zettler ME, Pierce GN. "Cell cycle proteins and atherosclerosis". Herz, 2000; 25(2): 1007.
193.
Andrés V, Castro C. "Antiproliferative strategies for the treatment of vascular proliferative disease". Curr Vasc Pharmacol, 2003; 1(1): 85-98.
194.
Teo TS, Wang TH, Wang JH. "Purification and properties of the protein activator of bovine heart cyclic adenosine 3',5'-monophosphate phosphodiesterase". J Biol Chem, 1973; 248(2): 588-95.
195.
Watterson DM, Harrelson WG, Jr, Keller PM, Sharief F, Vanaman TC. "Structural similarities between the Ca2+-dependent regulatory proteins of 3':5'-cyclic nucleotide phosphodiesterase and actomyosin ATPase". J Biol Chem, 1976; 251(15): 4501-13.
196.
Finn BE, Evenäs J, Drakenberg T, Waltho JP, Thulin E, Forsén S. "Calcium-induced structural changes and domain autonomy in calmodulin". Nat Struct Biol, 1995; 2(9): 777-83.
197.
Burgess WH, Jemiolo DK, Kretsinger RH. "Interaction of calcium and calmodulin in the presence of sodium dodecyl sulfate". Biochim Biophys Acta, 1980 623(2): 257-70.
121
198.
Heydorn WE, Creed GJ, Jacobowitz DM. "Observations and implications on the migration of calmodulin in a two-dimensional gel system". Electrophoresis, 1987; 8(5): 251-2.
199.
Van Eldik LJ, Wolchok SR. "Conditions for reproducible detection of calmodulin and S100 in immunoblots". Biochem Biophys Res Commun, 1984; 124(3): 752-9.
200.
McKeon TA, Lyman ML. "Calcium ion improves electrophoretic transfer of calmodulin and other small proteins". Anal Biochem, 1991; 193(1): 125-30.
201.
Mizzen CA, Cartel NJ, Yu WH, Fraser PE, McLachlan DR. "Sensitive detection of metallothioneins-1, -2 and -3 in tissue homogenates by immunoblotting: a method for enhanced membrane transfer and retention". J Biochem Biophys Methods, 1996; 32(2): 77-83.
202.
Suzuki Y, Takeda Y, Ikuta T. "Immunoblotting conditions for human hemoglobin chains". Anal Biochem, 2008; 378(2): 218-20.
203.
Watterson DM, Vanaman TC. "Affinity chromatography purification of a cyclic nucleotide phosphodiesterase using immobilized modulator protein, a troponin C-like protein from brain". Biochem Biophys Res Commun, 1976; 73(1): 40-6.
204.
Klee CB, Krinks MH. "Purification of cyclic 3',5'-nucleotide phosphodiesterase inhibitory protein by affinity chromatography on activator protein coupled to Sepharose". Biochemistry, 1978; 17(1): 120-6.
205.
Grand RJA, Perry SV. "Calmodulin-binding proteins from brain and other tissues". Biochem J, 1979; 183(2): 285-95.
206.
Porath J, Axén R, Ernback S. "Chemical coupling of proteins to agarose". Nature, 1967; 215(5109): 1491-2.
207.
Rizzo MA, Springer GH, Granada B, Piston DW. "An improved cyan fluorescent protein variant useful for FRET". Nat Biotechnol, 2004; 22(4): 445-9.
208.
Nagai T, Ibata K, Park ES, Kubota M, Mikoshiba K, Miyawaki A. "A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications". Nat Biotechnol, 2002; 20(1): 87-90.
209.
Afroze T, Husain M. "c-Myb-binding sites mediate G1/S-associated repression of the plasma membrane Ca2+-ATPase-1 promoter". J Biol Chem, 2000; 275(12): 9062-9.
210.
Majesky MW. "Developmental basis of vascular smooth muscle diversity". Arterioscler Thromb Vasc Biol, 2007; 27(6): 1248-58.
211.
Sakaue-Sawano A, Kurokawa H, Morimura T, Hanyu A, Hama H, Osawa H, Kashiwagi S, Fukami K, Miyata T, Miyoshi H, Imamura T, Ogawa M, Masai H, Miyawaki A. "Visualizing spatiotemporal dynamics of multicellular cell-cycle progression". Cell, 2008; 132(3): 487-98.
212.
Mochizuki N, Yamashita S, Kurokawa K, Ohba Y, Nagai T, Miyawaki A, Matsuda M. "Spatio-temporal images of growth-factor-induced activation of Ras and Rap1". Nature, 2001; 411(6841): 1065-8.
213.
Todaro GJ, Green H. "Quantitative studies of the growth of mouse embryo cells in culture and their development into established lines". J Cell Biol, 1963; 17(2): 299-313.
122
214.
Stavropoulos N, Rowntree RK, Lee JT. "Identification of developmentally specific enhancers for Tsix in the regulation of X chromosome inactivation". Mol Cell Biol, 2005; 25(7): 2757-69.
215.
Graham FL, Smiley J, Russell WC, Nairn R. "Characteristics of a human cell line transformed by DNA from human adenovirus type 5". J Gen Virol, 1977; 36(1): 59-74.
216.
DuBridge RB, Tang P, Hsia HC, Leong P-M, Miller JH, Calos MP. "Analysis of mutation in human cells by using an Epstein-Barr Virus shuttle system". Mol Cell Biol, 1987; 7(1): 379-87.
217.
Bernstein E, Muratore-Schroeder TL, Diaz RL, Chow JC, Changolkar LN, Shabanowitz J, Heard E, Pehrson JR, Hunt DF, Allis CD. "A phosphorylated subpopulation of the histone variant macroH2A1 is excluded from the inactive X chromosome and enriched during mitosis". Proc Natl Acad Sci U S A, 2008; 105(5): 1533-8.
218.
Bierne H, Tham TN, Batsche E, Dumay A, Leguillou M, Kernéis-Golsteyn S, Regnault B, Seeler JS, Muchardt C, Feunteun J, Cossart P. "Human BAHD1 promotes heterochromatic gene silencing". Proc Natl Acad Sci U S A, 2009; 106(33): 13826-31.
219.
Kakiuchi S, Yasuda S, Yamazaki R, Teshima Y, Kanda K, Kakiuchi R, Sobue K. "Quantitative determinations of calmodulin in the supernatant and particulate fractions of mammalian tissues". J Biochem, 1982; 92(4): 1041-8.
220.
Wu X, Bers DM. "Free and bound intracellular calmodulin measurements in cardiac myocytes". Cell Calcium, 2007; 41(4): 353-64.
221.
Lee BR, Kamitani T. "Improved immunodetection of endogenous -synuclein". PLoS ONE, 2011; 6(8): e23939.
222.
Rochette-Egly C, Daviaud D. "Calmodulin binding to nitrocellulose and Zetapore membranes during electrophoretic transfer from polyacrylamide gels". Electrophoresis, 1985; 6(5): 235-8.
223.
Hincke MT. "Conditions for improved adsorption of calmodulin to nitrocellulose: Detection by 45Ca binding". Electrophoresis, 1988; 9(7): 303-6.
224.
Millipore. "Anti-Calmodulin (Ab-2) Mouse mAb (6D4)". [Internet]. Millipore, [updated: 1999 March 11; cited 2012 February 13]. Available from: .
225.
Millipore. "Anti-Calmodulin (Ab-1) Mouse mAb (2D1)". [Internet]. Millipore, [updated: 1999 March 11; cited 2012 February 13]. Available from: .
226.
Hulen D, Baron A, Salisbury JL, Clarke M. "Production and specificity of monoclonal antibodies against calmodulin from Dictyostelium discoideum". Cell Motil Cytoskeleton, 1991; 18(2): 113-22.
227.
Collins JS, Goldsmith TH. "Spectral properties of fluorescence induced by glutaraldehyde fixation". J Histochem Cytochem, 1981; 29(3): 411-4.
228.
Smith MJ, Hardy WR, Li G-Y, Goudreault M, Hersch S, Metalnikov P, Starostine A, Pawson T, Ikura M. "The PTB domain of ShcA couples receptor activation to the cytoskeletal regulator IQGAP1". EMBO J, 2010; 29(5): 884-96.
123
229.
Li W, Li H, Sanders PN, Mohler PJ, Backs J, Olson EN, Anderson ME, Grumbach IM. "The multifunctional Ca2+/calmodulin-dependent kinase II (CaMKII) controls neointima formation after carotid ligation and vascular smooth muscle cell proliferation through cell cycle regulation by p21". J Biol Chem, 2011; 286(10): 7990-9.
230.
Mercure MZ, Ginnan R, Singer HA. "CaM kinase II2-dependent regulation of vascular smooth muscle cell polarization and migration". Am J Physiol Cell Physiol, 2008; 294(6): C1465-C75.
231.
Grynkiewicz G, Poenier M, Tsien RY. "A new generation of Ca2+ indicators with greatly improved fluorescence properties". J Biol Chem, 1985; 260(6): 3440-50.
232.
Miyawaki A, Llopis J, Heim R, McCaffery JM, Adams JA, Ikura M, Tsien RY. "Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin". Nature, 1997; 388(6645): 882-7.
233.
Chiri S, De Nadai C, Ciapa B. "Evidence for MAP kinase activation during mitotic division". J Cell Sci, 1998; 111(Pt 17): 2519-27.
234.
Philipova R, Larman MG, Leckie CP, Harrison PK, Groigno L, Whitaker M. "Inhibiting MAP kinase activity prevents calcium transients and mitosis entry in early sea urchin embryos". J Biol Chem, 2005; 280(26): 24957–67.
235.
Maiato H, DeLuca J, Salmon ED, Earnshaw WC. "The dynamic kinetochore-microtubule interface". J Cell Sci, 2004; 117(Pt 23): 5461-77.
236.
Fisher DD, Cyr RJ. "Calcium levels affect the ability to immunolocalize calmodulin to cortical microtubules". Plant Physiol, 1993; 103(2): 543-51.
237.
Kerppola TK. "Design and implementation of bimolecular fluorescence complementation (BiFC) assays for the visualization of protein interactions in living cells". Nat Protoc, 2006; 1(3): 1278-86.
238.
Kerppola TK. "Bimolecular fluorescence complementation (BiFC) analysis as a probe of protein interactions in living cells.". Annu Rev Biophys, 2008; 37: 465-87.
239.
Kerppola TK. "Visualization of molecular interactions by fluorescence complementation". Nat Rev Mol Cell Biol, 2006; 7(6): 449-56.
240.
Bracha-Drori K, Shichrur K, Katz A, Oliva M, Angelovici R, Yalovsky S, Ohad N. "Detection of protein–protein interactions in plants using bimolecular fluorescence complementation". Plant J, 2004; 40(3): 419-27.
241.
Kerppola TK. "Complementary methods for studies of protein interactions in living cells". Nat Methods, 2006; 3(12): 969-71.
242.
Bonda DJ, Lee H-p, Kudo W, Zhu X, Smith MA, Lee H-g. "Pathological implications of cell cycle re-entry in Alzheimer disease". Expert Rev Mol Med, 2010; 12: e19.
243.
Azimzadeh J, Wong ML, Downhour DM, Sánchez Alvarado A, Marshall WF. "Centrosome loss in the evolution of planarians". Science, 2012; 335(6067): 461-3.
244.
Lloyd C, Chan J. "Not so divided: the common basis of plant and animal cell division". Nat Rev Mol Cell Biol, 2006; 7(2): 147-52.
124
245.
Fukui Y, De Lozanne A, Spudich JA. "Structure and function of the cytoskeleton of a Dictyostelium myosin-defective mutant". J Cell Biol, 1990; 110(2): 367-78.
