Bundling of cytoskeletal actin by the formin FMNL1 contributes to celladhesion and migration
Item Type Dissertation
Authors Miller, Eric
Rights Attribution-NonCommercial-NoDerivatives 4.0 International
Download date 27/09/2021 05:11:17
Item License http://creativecommons.org/licenses/by-nc-nd/4.0/
Link to Item http://hdl.handle.net/20.500.12648/1760
Bundling of cytoskeletal actin by the formin FMNL1 contributes to cell adhesion and migration
Eric W. Miller
A Dissertation in the Department of Cell and Developmental Biology
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the College of Graduate Studies of State University of New York, Upstate Medical University
Approved ______Dr. Scott D. Blystone
Date______
i
Table of Contents
Title Page------i
Table of Contents------ii
List of Tables and Figures------vi
Abbreviations------viii
Acknowledgements------xiii
Thesis Abstract------xvi
Chapter 1: General Introduction------1
Introduction------2
Metastasis------3
History and Current Significance------3
Metastatic Cascade------4
Breast Cancer------4
Five Subtypes------4
Breast Cancer Metastasis------6
General Description of Metastatic Initiation------7
Cellular Locomotion------9
Introduction------9
History------10
Polarization------11
Protrusion and Adhesion Formation------12
Adhesion------14
Rear Retraction------16
Cancer Cell Invasion------16
EMT------16
ii
Migration During Invasion------17
Actin------19
G-Actin and F-Actin------19
Formins------21
Formin History------21
Formin Diversity and Subfamilies------22
General Formin Domain Structure------23
Formin Function------25
FMNL1------28
Structure and Functions------28
Localization of FMNL1------29
Binding Partners of FMNL------29
FMNL Proteins in Cancer------30
Formins as Actin Bundling Proteins------30
Detailed Formin History and Actin Bundle Organizations------30
Saccharomyces Cerevisiae – Bnr1------34
Schizosaccharomyces Pombe – Fus1 and Cdc12------35
Dictyostelium discoideum – ForC------37
Drosophila melanogaster – Cappuccino and DAAM------38
Mammals – Dia2, DAAM1, FMNL1-3, FHOD1, Delphilin------40
Arabidopsis thaliana – AFH1, AFH8, AFH14, AFH16------44
Oryza sativa – OsFH5 and OsFH16------47
Toxoplasma gondii – TgFRM1 and TgFRM2------48
The WH2 Domain------49
General Description------49
Non-formin WH2-domain Containing Proteins------50
iii
Mammalian Formins and WH2 Domains------51
Hypothesis/Signficance------52
References------53
Chapter 2: Expression patterns of human formins and FMNL1 alternative splice isoforms across multiple cell types------74
Introduction------75
Materials and Methods------78
Results------101
Discussion------135
References------138
Chapter 3: Two highly conserved amino acid residues of the FMNL1 FH2 domain are required for actin binding------147
Introduction------148
Materials and Methods------153
Results------154
Discussion------157
References------162
Chapter 4: The carboxy-terminus of the formin FMNL1ɣ bundles actin to potentiate adenocarcinoma migration------165
Abstract------166
Introduction------166
Materials and Methods------170
Results------179
Discussion------205
References------211
Chapter 5: General Discussion------217
iv
Introduction------218
FMNL1 and FMNL1 alternative splice isoform expression------219
2 conserved amino acid residues of FMNL1 are required for actin binding------224
The FMNL1ɣ C-terminus bundles actin to potentiate cancer migration------226
Concluding Remarks------237
References------239
v
List of Tables and Figures
Chapter 1: General Introduction
Figure 1 – The molecular subtypes of breast cancer------7
Figure 2 – The steps involved in cell migration------15
Figure 3 – Invasion, migration and intravasation into the circulatory system-----18
Figure 4 – Formin genes are located on a variety of chromosomes------23
Figure 5 – The FH2 domain moves processively with the barbed end of the
elongating actin filament while the FH1 domain recruits profilin-actin monomer
subunits for the growing barbed end------27
Chapter 2: Expression patterns of human formins and FMNL1 alternative splice isoforms across multiple cell types
Table 1 – The formin family of proteins------104
Figure 1 – Alignment of formin protein sequences------107
Figure 2 – Examples of qualitative PCR products for each formin ------108
Figure 3 – HeLa Cell RNA and Formin Expression------110
Figure 4 – Macrophage RNA and Formin Expression------113
Figure 5 – HCN Cell RNA and Formin Expression------116
Figure 6 – MDA-MB-231 Cell RNA and Formin Expression------119
Figure 7 – HK-2 Cell RNA and Formin Expression------122
Figure 8 – Meg-01 Cell RNA and Formin Expression------126
Figure 9 – Platelet RNA and Formin Expression------128
Figure 10 – Average FMNL1 expression varies across cell types------131
Figure 11 – The 3 FMNL1 alternative splice isoforms are expressed at the mRNA
level in MDA-MB-231 cells, macrophages, monocytes, and Meg-01 cells------134
vi
Chapter 3: Two highly conserved amino acid residues of the FMNL1 FH2 domain are required for actin binding
Table 1 – The highly conserved Ile and Lys residues of the FH2 domain------152
Figure 1 – The ability of FMNL1ɣ to bind actin is mediated by the FH2 domain
and not dependent on the C-terminal region------156
Figure 2 – FMNL1β and FMNL1ɣ bind actin, but FMNL1ɣFH2ø does not------157
Chapter 4: The carboxy-terminus of the formin FMNL1ɣ bundles actin to potentiate adenocarcinoma migration
Figure 1 – FMNL1ɣ and FMNL1ɣFH2ø inhibit actin assembly------190
Figure 2 – FMNL1ɣ is an efficient actin-bundling protein------193
Figure 3 – FMNL1 regulates breast cancer 2D cell motility------195
Figure 4 – Expression of GFP-tagged, full-length FMNL1 alternative splice
isoforms------197
Figure 5 – FMNL1 mediates cellular morphology, adhesion, and locomotion---200
Figure 6 – The formin FMNL1 is necessary for efficient invasion and displays a
unique WH2 domain structure------203
Chapter 5: General Discussion
Figure 1 – Proposed model of bundling mechanism incorporating WH2-mediated
actin nucleation------231
Figure 2 – Proposed model of bundling mechanism incorporating WH2-mediated
actin filament severing------232
Figure 3 – Proposed functions of FMNL1ɣ in the cell------235
vii
ABBREVIATIONS
ABM – Actin-binding mutant
ADF – Actin depolymerizing factor
ADP – Adenosine diphosphate
APC – Adenomatous polyposis coli
AR – Androgen receptor
Arp2/3 – Actin-related protein 2/3 complex
ATP – Adenosine triphosphate
BSA – Bovine serum albumin
Cdc42 – Cell division control protein 42 homolog cDNA – complementary deoxyribonucleic acid
Ck – Cytokeratins
Cobl – Cordon-bleu
CT – Carboxy-terminus
CXCL13 – C-X-C motif chemokine ligand 13
DAAM1 – Dishevelled associated activator of morphogenesis 1
DAAM2 – Dishevelled associated activator of morphogenesis 2
DABCO – 1,4-Diazabicyclo[2.2.2]octane
DAD – Diaphanous auto-regulatory domain
Dia1 – Diaphanous-related formin-1
Dia2 – Diaphanous-related formin-2
Dia3 – Diaphanous-related formin-3
DID – Diaphanous inhibitory domain
DMEM – Dulbecco’s Modified Eagle Medium
DMSO – Dimethylsulfoxide
viii
DNA – Deoxyribonucleic acid
DRF – Diaphanous-related formin
DTT - Dithiothreitol
E-cad – E-cadherin
ECM – Extracellular matrix
EDTA – Ethylenediaminetetraacetic acid
EGTA – Ethylene glycol-bis(β-aminoethyl ether)-N,N,N’,N’-tetraacetic acid
EGF – Epidermal growth factor
EGFR – Epidermal growth factor receptor
EMT – Epithelial-mesenchymal transition
Ena/VASP – Enabled/vasodilator-stimulated phosphoprotein
ER+ - Estrogen receptor positive
ErbB2 – Erb-B2 receptor tyrosine kinase 2
ERK – Extracellular signal-regulated kinase
F-actin – Filamentous actin
FAK – Focal Adhesion Kinase
FBS – Fetal bovine serum
FH1 – Formin homology 1 domain
FH2 – Formin homology 2 domain
FH2ø – Formin homology 2 domain-null
FH3 – Formin homology 3 domain
FHDC1 – FH2 domain-containing protein 1
FHOD1 – FH1/FH2 domain-containing protein 1
FHOD3 – FH1/FH2 domain containing protein 3
FITC – Fluorescein isothiocyanate
FMN1 – Formin 1
ix
FMN2 – Formin 2
FMNL1 – Formin-like protein 1
FMNL2 – Formin-like protein 2
FMNL3 – Formin-like protein 3
FRL – Formin-related protein identified in leukocytes
G-actin – Globular actin
GBD – GTPase binding domain
GFP – Green fluorescent protein
GPCR – G-protein coupled receptor
GST – Glutathione S-transferase
GTPase – Guanosine triphosphatase
HBSS – Hank’s Balanced Salt Solution
HER2+ - Human epidermal growth factor receptor2-enriched
HRP – Horseradish peroxidase
Hsp70 – Heat shock protein 70
IDC-NST – Invasive ductal carcinoma of no special type
IFNɣ - Interferon gamma
INF1 – Inverted formin 1
INF2 – Inverted formin 2
IPTG – Isopropyl β-D-1 thiogalactopyranoisde
KMEI – Imidazole, potassium chloride, magnesium chloride, EGTA
KO – Knock out
LysM – Lysozyme M
MAPK – Mitogen-activated protein kinase
MLCK – Myosin light-chain kinase
MMP – Matrix metalloproteinase
x
MRCK - Myotonic dystrophy kinase-related Cdc42-binding kinase
MT1-MMP – Membrane-type matrix metalloproteinase-1
MTOC – Microtubule organizing center
NPF – Nucleation promoting factor
PAKs - Serine/threonine p21-activating kinases
PARs – Partitioning proteins
PBS – Phosphate Buffered Saline
PCR – Polymerase chain reaction
PDGF – Platelet-derived growth factor
Pi – Inorganic phosphate
PI3K – Phosphoinositide 3-kinase
PIP – Phosphatidylinositol-phosphate
PKC – Protein kinase C
PR – Progesterone receptor
PTEN – Phosphatase and tensin homolog
PVDF – Polyvinylidene difluoride
Pyk2 – Protein tyrosine kinase 2 qRT-PCR – Quantitative real-time, reverse-transcriptase polymerase chain reaction
Rac – Ras-related C3 botulinum toxin substrate
Rho – Ras homolog gene family
RNA – Ribonucleic acid
ROCK – Rho-associated coiled-coil containing kinase
RT-PCR – Reverse-transcriptase polymerase chain reaction
WAVE – WASp-family verprolin-homologus protein
SDS – Sodium dodecyl sulfate
SDS-PAGE – Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
xi siRNA – Small interfering RNA
SMIFH2 – Small molecule inhibitor of Formin Homology 2 domain
Src – SRC proto-oncogene, non-receptor tyrosine kinase srGAP2 – SLIT-ROBO Rho GTPase-activating protein 2
TAMs – Tumor-associated macrophages
TBS – Tris Buffered Saline
TgACT1 – Toxoplasma gondii actin 1
TGFβ – Transforming growth factor beta
TgFRM1 – Toxoplasma gondii formin 1
TgFRM2 – Toxoplasma gondii formin 2
WASp – Wiskott-Aldrich Syndrome protein
WCA – WASp homology 2 domain – connecting domain – acidic domain
WH2 – WASp homology 2
WIP – WASP-interacting protein
WT – Wild-type
Zeb1 – Zinc finger E-box binding homeobox 1
xii
Acknowledgements
I started here at SUNY Upstate Medical University as a technician in the Blystone lab on December 8, 2008. It has now been close to ten years that I have been here and
I can honestly say I cannot imagine what life would be like if I had not first been offered a job here by my mentor, Dr. Scott Blystone. The guidance and insight he has provided me with is invaluable and I cannot even begin to express the appreciation I have for the opportunities he has provided me with here. I feel that the knowledge he has imparted on me not only made me a great scientist, but also aided me in becoming a better person as whole. I will absolutely miss working together after this many years, but I can honestly say that while I may be moving on to somewhere new, I am extremely grateful for the friendship we have developed, something I know will continue well into the future.
Another asset to me over the years has been my advisory committee, Dr.
Christopher Tuner and Dr. David Pruyne. Their knowledge and guidance is something that I will always be extremely thankful to have gained. I am extraordinarily appreciative of all the time they have put into seeing me succeed over the years. Additionally, I would especially like to thank the other members of my thesis defense committee: Dr.
John Copeland, Dr. Brian Haarer, and Dr. Paul Massa. I very much appreciate the time and effort they have provided me with in obtaining my degree.
SUNY Upstate Medical University, the College of Graduate Studies, and the Cell and Developmental Biology department have also been instrumental to my success. Dr.
Mark Schmitt, Terri Brown, and Jennifer Brennan were essential to my progress here.
Our chair, Dr. Joseph Sanger, provided me with both career guidance and funding to attend meetings which greatly enhanced my experience in science. The time Karen
Fontanella and Julie Arrigo provided me with in dealing with administrative tasks will always be appreciated as well.
xiii
All the friends I have made here at SUNY Upstate Medical University is something I will forever be appreciative of. I especially cherish the friendships I have formed with Robby Brooks and Neva Watson – they have provided me with so much support and the relationships I have with each of them, which I know will continue well into the future, have made me both a better scientist and better person.
My colleagues in the lab over the years have been wonderful, something I attribute to Scott always seeming to know who would best “fit” in the lab. Akos Mersich was a great help in the beginning of my career, taking the time to train and teach me while also being an extremely supportive friend. I truly value our friendship with each other and look forward to, hopefully, seeing each other more in the future. Lisi Krainer, who started almost one month after I began as a technician, provided me with friendship and support as well over the years and us growing as scientists together will always be something I remember.
Thank you to my family, especially my parents. None of this would be possible without both their support and patience. Their guidance has made me into someone who I am very proud to be and I cannot thank them enough for everything they have provided me with.
The support Ginny Grieb has provided me with during this time has been vital to my success here. She is not only an amazing scientist but an amazing person as well who I will always miss working with. The relationship we have formed over the years is priceless, and I know she will always been an essential part of my life.
And finally, I need to thank Matt Miller. Not only was he one of the best scientists
I have personally know, but one of the best people overall. He first started in the lab in the winter of 2009 and we quickly became best friends. I cannot begin to explain how lucky I feel knowing he was such an instrumental part of my life 9 years. I am so appreciative of the support, friendship, and knowledge he provided me with during this
xiv time. While he is no longer with us, the memories I have of our time together will always be one of my most priceless possessions.
xv
Thesis Abstract
Bundling of cytoskeletal actin by the formin FMNL1ɣ contributes to adenocarcinoma
adhesion and migration.
Author: Eric W. Miller
Sponsor: Scott D. Blystone
Metastasis is one of the leading causes of death in the world, affecting thousands every year. This is especially true of breast cancer, which can often result in the formation of secondary metastatic sites in the lung, liver, and bone marrow. There are many aspects to metastasis and an innumerable amount of molecular, biochemical, and cellular interactions contribute to its pathology. The ability of primary tumor cells to disseminate from the primary tumor, degrade the basement membrane, invade through the ECM, and eventually intravasate across the endothelial cell lining of the circulatory system or lymphatics requires a plethora of proteins, all working together in concert to achieve this. Nowhere in the cell is this more apparent than the actin cytoskeleton.
Locomotion of cells requires several alterations in the actin cytoskeleton component of the cellular machinery. Generally speaking, cells must be able to polarize, form protrusions, adhere to the substratum, translocate, and then retract their tail, repeating this process as they continue to navigate to their destination. While there are many underlying aspects to this activity, spatiotemporal rearrangements of the actin cytoskeleton are key to the successful cellular motility. The mechanics behind dynamic
xvi actin cytoskeletal modifications are varied and complex, demonstrating the requirement for a variety of actin-associated, regulatory proteins.
A crucial family of proteins involved in this process is the formin family of proteins. Formins are a relatively “new” group of actin modifiers which possess the unique ability to modify and generate linear actin filaments. While the members of this protein family all share some of the same actin modifying processes, many of these proteins also have functions exclusive to themselves. As a result, research into this field has blossomed and several novel features of different formins have been identified.
Furthermore, alternative splice isoforms of several formins are often expressed in a variety of cell types, with specific functions attributed to each.
The formin FMNL1 was originally identified in cells of a myeloid lineage and for many years was mostly thought to be involved in leukocyte adhesion and migration.
Indeed, our lab has characterized many of the functions of this protein in both human and murine macrophages. However, as a result of the work in this dissertation, we have generated sufficient evidence suggesting that FMNL1 not only plays a role in breast cancer migration, but also exhibits functions unique to a specific alternative splice isoform of this protein.
Our work on FMNL1 has pushed the field of study into this protein family in new directions. Herein, we have demonstrated that all three alternative splice isoforms of
FMNL1 are expressed in a variety of cell types and the FMNL1ɣ alternative splice isoform distinguishes itself from these isoforms via its ability to bundle linear actin filaments. Additionally, our data indicates that this is accomplished independently of the trademark FH2 domain, often thought to be the essential component of all formins.
More specifically, we have identified a unique amino acid sequence in the C-terminal region of this isoform that most likely regulates this function. As a result, we have not only identified a potential therapeutic target for the treatment of metastasis via inhibition
xvii of cellular locomotion, but also pushed the field of formin research into a novel direction by providing insight which may foster new hypotheses and challenge classical theories regarding the relationship between formins and actin
xviii
Chapter 1
General Introduction
1
Introduction
It is predicted that by the end of 2018, over 40,000 people in the United States will die as a result of metastases originating from primary tumors of the breast (Siegel et al., 2018). Therefore, it is imperative to gain a better understanding of the pathology of this disease in order to generate preventative strategies and develop potent therapeutics to treat this condition. While cellular locomotion is necessary for innumerable functions in life, it is this process that also drives metastasis. The foundation of cellular locomotion is the actin cytoskeleton, a spatiotemporally dynamic, complex structure that contributes to countless cellular processes ranging from intracellular transport to cellular motility to cytokinesis. While the actin cytoskeleton is only one-third of the cytoskeleton as a whole
(the other two components being the microtubule cytoskeleton and the intermediate filament cytoskeleton), it is arguably the most essential and functionally diverse of the three. The significance of the actin cytoskeleton is even more heightened when one takes into account how the aberrant regulation of this network can result in unwarranted cellular movement, much like what we observe with metastasis.
The regulation and modification of the actin cytoskeleton is governed by a multitude of actin-associated proteins. These proteins are required to be in precisely the right place at the right time throughout the cell in order to properly govern this filamentous network. Certain proteins can organize actin filaments in diverse structures, each of which are important for specific cellular functions. The formin family of proteins has recently proven to be one of the most powerful actin-associated protein families, many of them capable of modifying the actin cytoskeleton in several ways (Goode and
Eck, 2007, Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013).
Past work in our lab has focused on the formin FMNL1 and its unique role regulating actin dynamics in macrophage podosomes, adhesion structures required for these cells to adhere to the substratum and migrate through tissue. This dissertation will
2 focus primarily on how this same protein contributes to adhesion, migration, and invasion in a breast adenocarcinoma model, while also providing novel insight into the mechanism behind how FMNL1 can modify actin filaments in vitro.
Metastasis
History and Current Significance
In 1889, English surgeon Stephen Paget published his “seed and soil” hypothesis to describe metastasis, using the following analogy to describe the growth of secondary sites in the body: “When a plant goes to seed, its seeds are carried in all directions; but they can only live and grow if they fall on congenial soil” (Paget, 1889). For many years,
Stephen Paget’s theory was often challenged, especially following a theory posited by
James Ewing in 1928 when he suggested that metastasis occurs as a result of the physical surroundings, vasculature, and lymphatic arrangements between the primary tumor and the secondary metastatic sites (Ewing, 1928). 50 years would pass before it was demonstrated that while regional metastatic sites could be formed as a result of the physical and anatomical surroundings of the primary tumor, secondary metastatic sites in distant organs were specific to the type of cancer (Fidler and Kripke, 1977,
Sugarbaker, 1979, Hart and Fidler, 1980).
Now, in 2018, regardless of the diligent research and countless treatments available, cancer is still as prevalent as ever, second only to heart disease as the most common cause of death in the United States. It is predicted that in 2018 alone, 609,640
Americans will die of cancer and approximately 1.7 million new cases will be diagnosed
(Siegel et al., 2018). While 189 years have passed since metastasis was first identified and described by French gynecologist Jean Claude Recamier, it remains to this day the foremost challenge in treating cancer (Recamier, 1829, Steeg, 2006, Talmadge and
Fidler, 2010, Fidler and Kripke, 2015).
3
The Metastatic Cascade
The metastatic process itself, commonly referred to as the “metastatic cascade,” is often simplified as a linear sequence of steps: cancer cells invade the surrounding host tissue, enter into the circulatory system or lymphatics via intravasation, travel through the vasculature, inhabit the capillary bed of a distant, potential host organ site, extravasate from the vasculature, invade into this new tissue, and ultimately proliferate, forming micrometastases resulting in colonization (Steeg, 2006, Chaffer and Weinberg,
2011, Valastyan and Weinberg, 2011). However, in reality, metastasis is much more complex and each step of the metastatic cascade can be completed in a variety of manners. For example, intravasation could occur as a result of the loss of endothelial cell lining due to vascular mimicry or it may be due to co-migration of tumor cells with tumor-associated macrophages (TAMs) which can chemotactically guide tumor cells toward the vasculature (El Hallani et al., 2010, Harney et al., 2015, Yang et al., 2016,
Katt et al., 2018). However, when one observes the metastatic cascade as a whole, it becomes clear that tumor cells must be able to accomplish three underlying steps that serve as a foundation to this entire process: adhesion, migration, and invasion. Without successfully accomplishing these fundamental functions, tumor cells would be unable to travel to and colonize distant, secondary metastatic sites.
Breast Cancer
The Five Subtypes
While cancer can occur in a variety of organ and tissue types, breast cancer is the most common malignant disease amongst females in the United States. It is estimated that over 266,000 new cases will be diagnosed in 2018 and over 40,000 patients will perish as a result of this disease; nearly all these deaths will be due to metastasis (Rabbani and Mazar, 2007, Siegel et al., 2018). Breast cancer itself can be
4 classified into five main subtypes: luminal A, luminal B, human epidermal growth factor receptor 2-enriched (HER2+), basal-like, and normal breast-like (Figure 1) (Perou et al.,
2000, Sørlie et al., 2001, Sørlie et al., 2003, Sotiriou et al., 2003, Fadoukhair et al.,
2016). The luminal A and luminal B subtypes are both estrogen receptor positive (ER+) and predominately express the same genes as luminal breast epithelial cells, with luminal B subtypes also expressing mitosis- and cell proliferation-related genes at higher levels. While the HER2+ subtype expresses different ErbB2/HER2-related genes at a high level, these tumors can display other aspects of genetic diversity that makes them difficult to classify as one specific subtype (Russnes et al., 2017). For example, several primary tumors classified as HER2+ differ not only in their ER function, but also in regards to their copy number alterations and other genetic mutations (Staaf et al., 2010,
Ferrari et al., 2016, Russnes et al., 2017). The basal-like subtype is a heterogeneous group of tumors which include the subtype commonly referred to as “triple-negative,” where the tumor does not express ER, HER2, or progesterone receptor (PR).
Additionally, basal-like tumors often express different cytokeratins and other growth factors normally found in healthy breast myoepithelial cells. This subtype is associated with very poor prognosis and a limited survival rate of ~5-8 years following diagnosis
(Nielsen et al., 2004, Cheang et al., 2008, Badve et al., 2011, Russnes et al., 2017).
Only 5-10% of breast cancer tumors account for the normal breast-like subtype, the least characterized amongst the five subtypes. Gene expression of this subtype is similar to adipose tissue and can vary in regards to expression of ER, HER2, and PR.
Furthermore, while this subtype can be considered triple-negative if all three receptors are not expressed, it is still not considered basal-like due to lack of expression of different cytokeratins and other growth factors (Weigelt et al., 2010, Yersal and Barutca,
2014). While these five subtypes have been defined and characterized for several years, recent work has demonstrated that, in addition to the identification of rarer
5 subtypes, each of the five intrinsic subtypes have their own variations as well based on their molecular, genetic, and clinical profiles (Farmer et al., 2005, Prat et al., 2010, Guedj et al., 2012, Ringnér et al., 2013).
Breast Cancer Metastasis
Regardless of the subtype classification, breast cancer cells from the primary tumor can metastasize to several different tissues and organs, including, but not limited to, the brain, liver, lung, and bone, the most common of the metastatic sites (Coleman,
2001, Weilbaecher et al., 2011, Yates et al., 2017, Chen et al., 2018). However, breast cancer subtypes do display metastatic patterns based on their genetic profiles
(Kennecke et al., 2010, Cancer Genome Atlas Network, 2012, Wu et al., 2017). While bone metastasis is the most common metastatic site for all breast cancers, the luminal A and luminal B subtypes metastasize to bone at the highest rate compared to all other subtypes (Weilbaecher et al., 2011, Savci-Heijink et al., 2016). Brain metastasis most often occurs in patients with HER2+ and basal-like breast cancer subtypes, but other attributes, such as age and tumor differentatiation, can contribute to this as well (Smid et al., 2008, Kennecke et al., 2010, Witzel et al., 2016). The basal-like subtype is most often associated with lung metastasis while the luminal B subtype primarily found in liver metastases (Smid et al., 2008, Kimbung et al., 2016). Metastasis to the lymph nodes is often a predictor as to the risk of metastasis in other organs as tumor cells can utilize the lymphatics to travel to distant sites of the body. Nonetheless, the luminal A, luminal B, and HER2+ subtypes are most often associated with this secondary metastatic site
(Jatoi et al., 1999, He et al., 2015). Additionally, the histological classification of primary tumors allows for prediction of site-specific metastasis. Invasive ductal carcinomas, which originate from the ductal epithelium, most commonly metastasize to the lungs, lymph nodes, and brain while invasive lobular carcinomas, originating from epithelium
6 lining the lobules, preferentially metastasize to the gastrointestinal tract and ovaries
(Arpino et al., 2004, Makki, 2015, Chen et al., 2018).
Figure 1: Molecular subtypes of breast cancer. Multiple characteristics of the different breast cancer subtypes are displayed here. +/- indicates the expression of ER,
PR, HER2, or basal markers (EGFR, different cytokeratins). HER2 = human epidermal growth factor 2. EGFR = epidermal growth factor. Ck = cytokeratins. E-cad = E- cadherin. EMT = epithelial-mesenchymal transition. ER = estrogen receptor. PR = progesterone receptor. AR = androgen receptor. IDC-NST = invasive ductal carcinoma of no special type. Adapted from Geyer et al., 2009.
7
General Description of Metastatic Initiation
Broadly speaking, in order for the process of metastasis to begin, breast cancer cells must be able to migrate from the primary tumor, across the basement membrane, and into the stroma. During primary tumor development, the stroma becomes altered, resulting in changes in the composition of the extracellular matrix (ECM) and the number of cells within the ECM (Egeblad et al., 2010, Glentis et al., 2014, Clark and Vignjevic,
2015). These cells, which include fibroblasts, macrophages, endothelial cells, and pericytes, reorganize the connective tissue structure of the ECM via secretion of different proteins, growth factors, and cytokines (Wiseman and Werb, 2002, Cox and Erler, 2011,
Bonnans et al., 2014). Primary tumor cells react to these changes which results in further tumor growth and subsequent invasion (Levental et al., 2009, Burnier et al., 2011,
Lu et al., 2012, McCarthy et al., 2018). Detachment from the primary tumor requires tumor cells to modify both cell-cell and cell-extracellular matrix (ECM) adhesions, part of a process known as epithelial-mesenchymal transition (EMT). This process results in non-polarized epithelial cells shifting into a polarized, highly-invasive mesenchymal phenotype (Yilmaz and Christofori, 2009). EMT is primarily directed by specific transcription factors, especially Snail, Slug, and Zeb1, as well as a downregulation of E- cadherin, upregulation of N-cadherin, and alterations in the composition of the ECM (De
Craene and Berx, 2013, Lamouille et al., 2014, Nieto et al., 2016). Loss of E-cadherin expression drives the dissolution of cell-cell adhesions as this protein is essential for regulation of epithelial adherens junctions and other proteins necessary for maintaining proper cell-cell adhesions (Stockinger et al., 2001, Yilmaz et al., 2009). Integrins further drive EMT via interactions with the ECM which lead to alterations in expression of EMT- related transcription factors (Haraguchi et al., 2008, Yilmaz and Christofori, 2009).
Additionally, integrins recruit and direct various matrix metalloproteases (MMPs) to remodel the ECM, allowing cells to invade through the proteolytically-degraded ECM,
8 eventually reaching the intravasation phase of the metastatic cascade (Ellerbroek et al.,
2001, Cao et al., 2008).
Cellular Locomotion
Introduction
From this abbreviated description on the initiation of breast cancer metastasis, one can identify the three previously mentioned cellular functions required for this process to occur: adhesion, migration, and invasion. Without all three of these functional mechanisms, cells would not be able to disseminate from the primary tumor and travel to secondary metastatic sites. However, metastasis is not the only biological process that relies on adhesion, migration, and invasion. Development, immunosurveillance, and wound healing are just some of the processes that require these functions to work in concert with each other. For example, during the gastrulation phase of embryonic development, cells move as interconnected epithelial sheets over various distances
(Kurosaka and Kashina, 2008). Additionally, germ cells in the mouse migrate through the endoderm and travel to the somatic gonadal mesoderm, eventually forming the gonads (Bendel-Stenzel et al., 1998, Kunwar et al., 2006). Leukocytes in the circulatory system utilize a specialized form of adhesion and migration in order to respond to infection or injury (Nourshargh et al., 2010). However, unwarranted migration of immune cells can contribute to atherosclerosis (Jonasson et al., 1986, Hansson and Libby,
2006). Angiogenesis is a necessary physiological process that requires proper endothelial cell migration in order to generate new blood vessels, but can also contribute to malignant tumor formation and growth of metastases (Saaristo et al., 2000, Bryan and
D’Amore, 2007). Adhesion and migration of several cell types, ranging from platelets to neutrophils to fibroblasts, must be able to migrate to sites of burns or skin ulcers in order to contribute to the versatile process of wound repair (Singer and Clark, 1999).
9
History
Whether it be the development, homeostasis, or disease progression of a multicellular organism, cellular locomotion is a foundational function of life. In 1953,
Michael Abercrombie and John Heaysman at the University College of London first described how cellular movement of chicken fibroblasts was limited as a result of contact with neighboring cells, coining the term “contact inhibition” (Abercrombie and Heaysman,
1953). Abercrombie continued his work on the physical contact between cells during contact inhibition, observing that the lamellae of fibroblasts cease their ruffling behavior upon contact with each other resulting in increasing tension and retraction. This observation led him to hypothesize this was due to some form of adhesion event occurring between the cells (Abercrombie and Ambrose, 1958). Twelve years later,
Abercrombie and Heaysman, along with Susan Pegrum, published a series of papers describing the movement of fibroblasts, identifying membrane protrusions at the leading edge of the cell which they would term “lamellipodia;” this would later be identified as one of the major pieces of the cellular locomotion machinery (Abercrombie et al., 1970a,
Abercrombie et al., 1970b, Abercrombie et al., 1970c). Additionally, as a part of this series of papers, electron micrographs displayed not only the lamellipodia, but what was described as “electron-dense plaques containing longitudinal filaments.” These plaques would later be identified as focal adhesions, an essential component of the fundamental process of cellular adhesion (Abercrombie et al., 1971, Abercrombie et al., 1972).
Since these publications in the early 1970s, our understanding of cellular adhesion and locomotion has increased tremendously. Nonetheless, the more knowledge we have gained has only resulted in a deluge of more questions. In order to address the pathology of cancer cell metastasis, it is essential to understand the foundations of tumor cell invasion: adhesion and migration. As was previously described, adhesion and migration are essential to countless functions in regards to homeostasis, development,
10 and immunosurveillance as well. Therefore, adhesion and migration will further be addressed in a general manner, after which a detailed description of the cellular invasion process will follow.
Polarization
Cells frequently exhibit an inherent random motility prior to the establishment of polarity, the first step in directional migration (Figure 2). Polarization is often the result of cellular stimulation via different cues, such as growth factors (e.g. fibroblasts responding to platelet-derived growth factor (PDGF)), chemokines (e.g. B-cells responding to expression of CXCL13 by macrophages), or even the physical composition of the ECM
(e.g. fibroblast migration variation on rigidity gradients) (Seppä et al., 1982, Lo et al.,
2000, Ansel et al., 2002). However, cells can spontaneously polarize as well, such as in fish keratocytes where ROCK and myosin II activity regulate the perinuclear actin flow resulting in a reorganization of the actomyosin network and subsequent contractility at the cell rear (Yam et al., 2007). Regardless of the event that initiates cellular guidance, the establishment and regulation of a polarized cell is driven by several positive feedback loops at the cell front and inhibitory signals throughout the rest of the cell that restrict protrusion formation to the leading edge. Several molecules are involved in this process including the Rho-family GTPases, serine/threonine p21-activating kinases
(PAKs), the PAR (partitioning) family of proteins, the serine/threonine protein kinase C
(PKC) family, the phosphoinositide 3-kinases (PI3Ks), which produce PIP3 and
PI(3,4)P2, and PTEN which dephosphorylates them (Devreotes and Janetopoulos, 2003,
Etienne-Manneville and Hall, 2003, Li et al., 2003, Ridley et al., 2003, Srinivasan et al.,
2003). One of the main regulators of cell polarity, the Rho-family GTPase Cdc42, not only governs the direction of lamellipodia generation, but also interacts with the microtubule-organizing center (MTOC), allowing for translocation of the nucleus, various
11 organelles, and vesicles (Ridley et al., 2003, Maldonado and Dharmawardhane, 2018).
This localization is the result of Cdc42 interacting with a protein complex consisting of
PAR and PKC protein family members, as well MRCK (myotonic dystrophy kinase- related Cdc42-binding kinase) (Joberty et al., 2000, Ridley et al., 2003) . Also during this time, the phosphoinositides PIP3 and PI(3,4)P2, which are generated by PI3Ks, localize to the leading edge of the cell (Ridley et al., 2003, Xue and Hemmings, 2013).
Furthermore, Cdc42 can activate PAK1, which itself acts as a scaffold protein for Cdc42 activation. This Cdc42-PAK1 interaction, along with PIXα and the heterotrimeric GPCRs
(G-protein coupled receptors), effectively acts as a positive feedback loop allowing elevated Cdc42 activity at the leading edge and isolating PTEN activity to the sides and rear of the cell, whereupon interactions with myosin II allow for restriction of protrusion formation at the leading edge (Knaus and Bokoch, 1998, Chong et al., 2001, Ridley et al., 2003).
Protrusion and Adhesion Formation
Protrusion of the cell at the leading edge involves different structures which extend the cellular membrane, including lamellipodia, filopodia, podosomes, and invadopodia (Figure 2). This is mediated by the actin cytoskeleton, the primary force driving the formation of the structures that also governs the shape of the cell and the organization within it (Pollard and Borisy, 2003, Ridley et al., 2003). The unique polarity of the actin filament is the foundation for this as these polymers consist of the fast- growing barbed (+) end and the slow-growing pointed (-) end (Pollard and Borisy, 2003,
Pollard, 2017). As actin filaments polymerize at the barbed end, they generate a force which physically drives the cell membrane toward its migratory direction (Svitkina, 2018).
This process is cyclical and actin filaments at the leading edge are continuously disassembled from the pointed (-) end, resulting in a pool of actin monomer subunits
12 which can then be utilized for further filament assembly, a process referred to as treadmilling (Pollard and Borisy, 2003, Pollard, 2017). Actin filament organization within these structures varies as each structure has its own unique function (Pollard and
Borisy, 2003, Ridley et al., 2003, Ridley, 2011).
In the lamellipodia, the primary actin-driving force is the seven-subunit complex
Arp2/3 (actin-related protein 2/3) which works in concert with different NPFs (nucleation promoting factors) (Figure 2) (Ridley, 2011). This protein complex nucleates actin filaments and organizes them into dendritic networks at ~70° angles by capping the pointed (-) end and accelerating elongation of the barbed (+) end. Nucleation of actin filaments by Arp2/3 alone is kinetically unfavorable and, as a result, different NPFs with
WH2 (WASP homology 2) domains bind actin monomer subunits to aid in this process
(Goley and Welch, 2006). These NPFs are activated by different Rho-family GTPases, such as Cdc42 and Rac, while the WH2 domains are part of a larger WCA domain which interacts with Arp2/3 via an amphipathic connector region and acidic peptide. This interaction results in a structural change in the Arp2/3 complex, effectively allowing it to shuttle actin monomer subunits to the growing barbed end (Ridley et al., 2003, Goley and Welch, 2006, Campellone and Welch, 2010). Polymerization of actin filaments and their subsequent formation of filament bundles drives protrusion of the membrane. The region located behind the lamellopodia, the lamella, also aids in the migratory process by incorporating myosin II with the actin filaments, generating contractile actomyosin networks (Ridley, 2011, Svitkina, 2018).
Filopodia are thin, highly dynamic cellular protrusions composed of parallel F- actin bundles. These structures primarily serve as sensory structures of the motile cell, interacting with chemokines, growth factors, and the ECM (Ridley, 2011, Svitkina, 2018).
The parallel actin bundles primarily extend from the cellular cortex at the lamellipodia or through de novo actin nucleation at the plasma membrane. Several actin-associated
13 proteins have been implicated in the formation and regulation of these structures including Arp2/3, formins, Ena/VASP, and fascin (Pollard and Borisy, 2003, Ridley et al.,
2003, Campellone and Welch, 2010, Ridley, 2011, Svitkina, 2008).
Podosomes and invadopodia are actin-rich structures the cell uses for both adhesion and degradation of the ECM. Podosomes are primarily found in monocytes, endothelial cells, and smooth muscle cells while invadopodia are found in cancer cells.
These structures have a similar composition with an actin-rich core surrounded by a plethora of different scaffolding proteins. While both structures have a similar width (0.5-
2.0μm) they vary in length with podosomes measuring between 0.5-2.0μm and invadopodia measuring longer than 2.0μm. Additionally, while both can secrete MMPs, they vary in their half-life, with podosomes having a half-life in the minute range and invadopodia having a half-life in the hour range (Murphy and Courtneidge, 2011).
Adhesion
Protrusions formed by the cell also contribute to its adhesive functions by providing a physical attachment to the substratum which, in turn, allows the cell to eventually generate the force required for proper locomotion (Figure 2) (Ridley et al.,
2003, Ridley, 2011, Svitkina, 2018). Adhesion complexes such as focal adhesion or podosomes are integrin-rich and contain an immense variety of adaptor proteins which can connect the actin cytoskeleton to the protrusions. Adhesions often form at the leading edge of the cell and their formation is mediated by Rho-family GTPases, such as
Rac1 and Cdc42. This also allows the lamellipodia and filopodia to efficiently stabilize
(Ridley et al., 2003, Wozniak et al., 2004, Ridley, 2011, Svitkina, 2018).
14
Figure 2: The steps involved in cell migration. The different steps of cell migration
(polarization, protrusion formation, adhesion formation, and rear retraction) are identified here along with illustrations demonstrating the cellular alterations in each step. Boxes list proteins involved in the different steps of the process. Adapted from Ridley et al.,
2003.
Integrins at adhesion sites can act as messengers, transmitting information concerning the substrate surface to the rest of the cell, such as ECM compostion and rigidity (Barczyk et al., 2010). The actual force transmission is the result of actomyosin contractility, where the interaction of myosin II with actin can generate piconewton forces. Mysoin II is regulated by both ROCK (Rho-associated coiled-coil protein kinase)
15 and MLCK (Myosin light-chain kinase), which, in turn, are regulated by Rho-family
GTPases and Ca2+ gradients, respectively (Pollard and Borisy, 2003, Ridley et al., 2003,
Ridley, 2011, Svitkina, 2018).
Rear Retraction
As the cell moves forward, the strong adhesions in the tail region will actually cause such high tension that the bond between the integrin and actin cytoskeleton breaks, and the integrin will subsequently undergo integrin endocytic recycling (Ridley et al., 2003, Maritzen et al., 2015, Nader et al., 2016). FAK (focal adhesion kinase), Src (a proto-oncogene tyrosine-protein kinase), and Ca2+ gradients contribute to the regulation of tail retraction and adhesion disassembly. As the tail is retracted, the change in the
Ca2+ gradient results in ERK activating calpain, a protease that can digest a variety of adhesion complex proteins, such as integrins and FAK (Ridley et al., 2003, Storr et al.,
2011).
Cancer Cell Invasion
EMT
In order to metastasize, epithelial cancer cells must be able to disseminate from the primary tumor (Stuelten et al., 2018). For this to be accomplished, E-cadherin must be downregulated, effectively allowing the cancer cell to disassemble cell-cell adhesions and detach from neighboring cells. It is during this time that an upregulation of mesenchymal markers occurs, including, but not limited to, N-cadherin, vimentin, fibronectin, and different integrins. TGFβ signaling also contributes to EMT, activating the ERK/MAPK pathways, PI3K, and Rho-family GTPases. Epithelial cell polarity will also be altered via LKB1-regulated polarization processes (Yamaguchi and Condeelis,
2007, Yilmaz and Christofori, 2009, Son and Moon, 2010).
16
Migration During Invasion
Following EMT, the cell can proceed to migrate via binding to the ECM followed by extension of the lamellipodia (Yamaguchi and Condeelis, 2007). Integrins along with different kinases and Rho-family GTPases allow for force transmission onto the ECM and enhancement of lamellipodia extension (Pollard and Borisy, 2003, Svitkina, 2018).
MMPs can also be secreted at this time, proteolytically degrading the ECM, effectively altering its structure so the cell can eventually move through this region (Figure 3)
(Yamaguchi and Condeelis, 2007, Leber and Efferth, 2009, Valastyan and Weingberg,
2011). Different actin-rich structures contribute to this movement, all of which are regulated by a variety of proteins. Two of these, the lamellipodia and invadopodia, are prominently featured in this process. Formation of both of these structures is driven by the actin cytoskeleton and polymerization machinery (Yamaguchi and Condeelis, 2007,
Ridley, 2011). Protrusion formation also requires free actin barbed ends which can allow for polymerization at specific sites in the cell (Pollard and Borisy, 2003, Pollard, 2017).
Rho family GTPases contribute to this migration in several ways: RhoA can mediate stress fiber formation, Rac1 contributes to lamellipodia formation, and Cdc42 is important for filopodia formation (Krugmann et al., 2001, Ehrlich et al., 2002, Ridley,
2011, Tojkander et al., 2012). Different adhesion molecules, such as laminins and integrins, are upregulated during this time along with actin-modifying proteins such as cofilin and Arp2/3 (Yamaguchi and Condeelis, 2007, Ridley, 2011).
17
Figure 3: Invasion, migration, and subsequent intravasation into the circulatory system. Cells from the primary tumor alter their morphology and detach from cell-cell contacts via EMT. Invadopodia degrade the basement membrane and ECM, allow the cell to migrate via the formation of different cellular protrusions. Cells eventually reach the vasculature, degrading the basement membrane surrounding the endoethelial cell lining and intravasate into the circulatory system. Adapted from Bravo-Cordero et al.,
2012.
Additionally, the formin family of proteins also contributes actin-modifying machinery which allows the cell to invade. Dia1 and FMNL1 have been shown to be important for regulating the cellular polarity of cancer cells via reorientation of the Golgi complex and the MTOC, both of which are activated downstream of different Rho family
18
GTPases such as RhoA and Cdc42 (Gomez et al., 2007, Narumiya et al., 2009, DeWard et al., 2010). Furthermore, Dia1 contributes to stress fiber formation in different cellular protrusions downstream of RhoA, regulates focal adhesion complex formation and turnover, and mediates leading edge dynamics via interaction with IQGAP (Tominaga et al., 2000, Brandt et al., 2007, Sarmiento et al., 2008, DeWard et al., 2010). Additionally,
Dia2 localizes with Src at invadopodia and regulates their formation in MDA-MB-231 cells (Lizárraga et al., 2009).
Actin
G-Actin and F-actin
Actin filaments (F-actin), along with microtubules and intermediate filaments, collectively form the cytoskeleton as a whole. Actin is widely conserved and is quite similar across all eukaryotic domains. Additionally, prokaryotes and archaea have genes for actin as well, demonstrating the significance of this protein throughout all of life
(Pollard and Borisy, 2003, Perrin and Ervasti, 2010). Eukaryotes have at least one gene that codes for actin, with mammals having six different genes for each actin isoform.
These include Ɣcyto-actin, βcyto-actin, αskeletal-actin, αcardiac-actin, αsmooth-actin, and Ɣsmooth- actin, which all share ~90% identity with each other (Perrin and Ervasti, 2010). The crystal structure of rabbit muscle monomeric actin (G-actin) was first described as a complex with DNase I in 1990 and since then our understanding of the protein has advanced exponentially (Kabsch et al., 1990).
In its monomeric state, actin is a 42kDa ATPase and belongs to the same structural superfamily as sugar kinases, hexokinases, Hsp70 proteins, and Arp (actin- related protein) proteins (Pollard and Borisy, 2003, Pollard, 2017). Actin is the most abundant protein in nearly all eukaryotic cells and concentrations in non-muscle cells can exceed 300μM (Pollard and Borisy, 2003). The formation of filaments first requires
19 the thermodynamically unfavorable step of nucleation, where monomers proceed to form dimers and trimers, upon which they effectively form an actin tetramer or “seed.”
Accelerated elongation can then occur with monomers effectively lengthening the filament at a rapid rate, with a critical concentration for monomer-binding to the barbed
(+) end being 0.1μM (Pollard and Borisy, 2003, Campellone and Welch, 2010, Pollard,
2017). These ATP-bound monomers are hydrolyzed in the filament itself, at a half-time of two seconds, as ADP-bound monomers depolymerize from the pointed (-) end
(Pollard and Borisy, 2003, Pollard, 2017).
The filament itself is a right-handed, helical, polarized polymer and elongation of the barbed end occurs at a rate of 11.6μM-1s-1 (Pollard and Borisy, 2003, Pollard, 2017).
All actin monomer subunits of the filament are oriented in the same direction. Nucleation of the actin filament is a very kinetically unfavorable process. Actin dimers and trimers are extremely unstable and dissociate in a matter of microseconds and milliseconds, respectively. Actin monomer sequestration proteins such as profilin and β-thymosin also contribute the inhibition of de novo actin polymerization (Pollard and Borisy, 2003,
Pollard, 2017, Svitkina, 2018). As a result, actin uses different nucleating proteins to allow for this process, including the Arp2/3 complex, formins, and WH2 domain- containing proteins (Campellone and Welch, 2010).
The importance of actin in such a large number of cellular functions is further revealed by the high concentration of actin monomer subunits within pools in the cytoplasm. As such a high concentration of monomeric subunits exceeds the critical concentration for actin filament polymerization, the cell utilizes different G-actin-binding proteins as regulators. These regulators, including, but not limited to profilin, thymosin
β4, and ADF/cofilin, mediate the sequestration of the available actin monomer subunits in the cell. In addition to this, they also have the ability to regulate the hydrolytic state of the monomeric subunits and their localization within the cell (Pollard and Borisy, 2003,
20
Pollard, 2017). New actin filaments can be generated from this pool of actin via actin nucleation and filament elongation. While a variety of proteins can contribute to this process, this dissertation will focus primarily on the formin family of proteins and their contributions toward modifying actin within the cell.
Formins
Formin History
Formins were initially identified following studies of the limb-deformity gene in mice and subsequent characterization of alternatively spliced mRNA transcripts of this gene (Kleinebrecht et al., 1982, Woychik et al., 1990). This allowed for the prediction of potential protein products which were termed “formins” due to their requirement in mice for proper limb and kidney formation (Kleinebrecht et al., 1982, Zeller et al., 1989, Maas et al., 1990, Messing et al., 1990). Research continued on these novel findings and evolutionarily conserved domains of protein products from different genes were identified in Drosophila melanogaster, Mus musculus, Gallus gallus domesticus, and
Saccharomyces cerevisiae, including the defining feature of formins, the FH2 (formin homology 2) domain. Formin family members were identified in other organisms and
Bni1 in S. cerevisiae was eventually identified as an actin-associated protein with the distinctive ability to nucleate and assemble unbranched actin filaments. Around this same time period, the interaction of formins with Rho-family GTPases was ascertained and since then the field has flourished (Castrillon and Wasserman, 1994, Chang et al.,
1997, Watanabe et al., 1997, Bione et al., 1998, Petersen et al., 1998, Swan et al., 1998,
Yayoshi-Yamamoto et al., 2000, Pruyne et al., 2002). Formins have since been recognized as some of the most powerful actin-modifying machines in the cell and the span of their functions ranges from cytokinesis to filopodia formation (Harris et al., 2010,
Bohnert et al., 2013a, Jaiswal et al., 2013). More recently, it has been established that
21 this diverse family of proteins interacts with microtubules as well, establishing themselves as remarkable modifiers of the cytoskeleton as a whole (Thurston et al.,
2012, Fernández-Barrera J and Alonso MA, 2018). However, the evolution of this field has only led to more questions concerning how these proteins operate and function, especially in regards to how they modify actin filaments.
Formin Diversity and Subfamilies
Since first being identified, formins have been extensively studied in several model systems ranging from Arabidopsis thaliana to Homo sapiens. While the number of formins in eukaryotes varies greatly depending on species, 15 genes encode for each formin in mammals, all of which can be separated into eight subfamilies: Dia
(Diaphanous-related formin), Daam (Disheveled-associated activator of morphogenesis),
FMNL (formin-like), INF (inverted formin), FHDC (FH2 domain-containing protein),
FHOD (FH1/FH2 domain-containing protein), FMN (formin), and Delphilin (Figure 4). In addition to the variation of their chromosomal position, several formins have multiple alternative splice isoforms as well (Higgs and Peterson, 2005, Rivero et al., 2005,
Schönichen and Geyer, 2010, Pruyne, 2016, Pruyne, 2017).
22
Gene No. of Formin Locus Isoforms Dia1 5q31.3 2 Dia2 Xq21.33 3 Dia3 13q21.2 7 Daam1 14q23.1 3 Daam2 6p21.2 1 FMNL1 17q21.31 3 FMNL2 2q23.3 4 FMNL3 12q13.12 3 INF2 14q32.33 2 FHDC1 4q31.3 1 FHOD1 16q22.1 1 FHOD3 18q12.2 2 FMN1 15q13.3 4 FMN2 1q43 1 Delphilin 7p22.1 1
Figure 4: Formin genes are located on a variety of chromosomes. The 15 human formin family members are listed here along with their gene locus and the number of their confirmed isoforms. Adapted from Schönichen and Geyer, 2010.
General Formin Domain Structure
Formins are high molecular weight, multi-domain proteins ranging in size from
~1000-1800 amino acid residues in length. The defining feature of formins is the ~350 amino acid residue FH2 domain, flanked N-terminally by the smaller FH1 (Figure 5).
The FH2 domain is essential to the dimeric formin structure that can bind to the barbed
(+) end of the actin filament. This results in a head-to-tail, “donut-shaped” homodimer that can remain bound to the growing barbed end, effectively acting as an anti-capping protein, allowing the acceleration of actin filament assembly. The FH1 domain is rich in poly-proline repeats that interact with profilin-bound actin monomers which can be recruited to accelerate actin filament elongation. C-terminal to the FH2 domain is a
23 region that can vary in sequence but does share a common region known as the DAD
(diaphanous auto-regulatory domain). This domain interacts with the GBD (GTPase- binding domain) at the furthest N-terminal end of the protein. The C-terminal region is also often home to other conserved domains, such as WH2 domains, although this is unique to specific formins. The region with the most variability, the FH3 domain, is located between the N-terminal GBD domain and the FH1 domain (Goode and Eck,
2007, Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013).
FH2 domains are unique in that they can not only nucleate and elongate actin filaments, but also bundle and sever these filaments as well (Harris et al., 2004, Harris et al., 2006, Goode and Eck, 2007, Vizcarra et al., 2014). This domain is also essential to the homodimerization of the formin. The N-terminal region of the FH2 domain contains what is termed a “lasso” subdomain, a name based on its molecular shape. The lasso subdomain of one formin monomer can then interact with the C-terminal “post” subdomain of the same formin monomer, resulting in the formin’s homodimeric state.
Between the lasso and post subdomains, a region of the molecule known as the “linker” provides the connection between the FH2 domains of the formins. It is here that two highly conserved residues, an isoleucine and lysine, are located which are essential for the formin to interact with actin (Xu et al., 2004, Goode and Eck, 2007, Schönichen and
Geyer, 2010, Breitsprecher and Goode, 2013).
The DAD domain is comprised of two unique structural features: an amphipathic helix and a central MDxLLxxL motif immediately followed by basic region that can differ widely in regards to amino acid residue composition and length. The helix region of the
DAD domain binds to the hydrophobic armadillo repeat region located within the FH3 domain while the basic region increases affinity of the FH3 domain with the GBD domain based on electrostatic interactions (Goode and Eck, 2007, Schönichen and Geyer, 2010,
Breitsprecher and Goode, 2013).
24
The mechanism of how the formin is released from its autoinhibited state by Rho- family GTPase binding remains elusive. Moreover, some studies have shown that interactions with Rho-family GTPases have little effect on formin activity (Han et al.,
2009, Miller et al., 2017). Additionally, mechanisms of activation other than Rho-family
GTPase binding may be necessary to fully activate the formin (Liu et al., 2008, Bartolini et al., 2016). It has also been suggested that some formins may be constiutively active.
For example, N-terminal myrisotylation of FMNL1 may result in the formin remaining in its open, active conformation while DIP-1 (diaphanous inhibitory protein 1) can actually bind to the FH2 domain directly, blocking formin activity (Han et al., 2009, Aspenström,
2010). As each formin behaves quite differently, in-depth analysis of this mechanism could very well demonstrate that the relationship between each formin and different Rho- family GTPases may be entirely unique.
Formin Function
Formins are often defined as actin nucleators, however, not all formins have the ability to nucleate and those that do can vary in how well they perform this action (Goode and Eck, 2007, Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013). Formin proteins have the ability to interact with the barbed end of actin filaments, which they can proceed to elongate and prevent capping by other proteins such as gelsolin (Figure 5)
(Zigmond et al., 2003). The exact mechanism of formin-mediated actin nucleation still remains to be resolved and is likely to be different for each formin. Currently, all studies that have demonstrated nucleation are in vitro and actual in vivo nucleation remains to be determined. Furthermore, when formin-mediated nucleation does occur, different
NPFs have been demonstrated to interact with formins in order to overcome this highly unfavorable reaction, such as DIP and APC (Aspenström, 2010, Okada et al., 2010).
Additionally, more evidence has suggested that the C-terminal region of the formin
25 contributes to stabilization of actin nucleation by binding actin monomers (Heimsath and
Higgs, 2012, Vig et al., 2017).
Filament assembly also varies between formins, and assembly rates greatly differ depending on the in vitro experimental conditions. Formin FH2 domains remain attached at the barbed end, processively adding actin monomer subunits and preventing capping proteins from binding to the barbed end. The formin moves with the barbed end as it is elongated in a “stair-stepping” fashion. The FH2 dimer can switch between an
“open” state where actin monomer subunits can be added and a “closed” state where addition of monomeric subunits is halted. The FH1 domain recruits profilin-bound actin monomer subunits during filament assembly, which can subsequently be added to the growing barbed end (Figure 5) (Goode and Eck, 2007, Paul and Pollard, 2009,
Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013). How actin monomer subunits are actually transferred from the FH1 domain to the FH2 domain remains to be determined.
While formins are frequently thought of as actin nucleators and elongators, they also modify filaments through bundling and severing activities. Indeed, FMNL1 is able to perform all of these actions (Harris et al., 2004, Harris et al., 2006). Dia2, FMNL2,
FMNL3, and FHOD1 have all been shown to have some form of bundling activity (Harris et al., 2006, Vaillant et al., 2008, Schönichen et al., 2013). Moreover, these additional activities of formins may rely on other domains within these proteins, such as WH2 domains, which have been shown to contribute to bundling in FMNL2 and FMNL3
(Vaillant et al., 2008).
26
Figure 5: The FH2 domain moves processively with the barbed end of the elongating actin filament while the FH1 domain recruits profilin-actin monomer subunits for addition to the growing barbed end. GBD = GTPase-binding domain.
DAD = Diaphanous autoregulatory domain. DID = Diaphanous inhibitory domain. FH =
Formin homology domain. N = N-terminus. C = C-terminus. Adapted from Chesarone et al., 2010.
27
FMNL1
Structure and Functions
Originally hypothesized to be predominately expressed in leukocytes, FMNL1 expression has now been demonstrated in a variety of cell and tissue types (Han et al.,
2009, Mersich et al., 2010, Colon-Franco et al., 2011, Gardberg et al., 2014). FMNL1 encodes for three different isoforms (FMNL1α, FMNL1β, FMNL1ɣ), all of which diverge as splice variants in the C-terminal region following at T1069 (Katoh and Katoh, 2003,
Han et al., 2009). FMNL1 is classically known as a regulator of the actin-rich adhesion complexes of macrophages known as podosomes (Mersich et al., 2010). This protein is essential to the stability of this complex, although this is not its only function (Mersich et al., 2010, Miller et al., 2017). Interestingly, FMNL1 seems to have a diverse array of functions not only in immunosurveillance, but cancer cell migration and development as well (Kitzing et al., 2010, Rosado et al., 2014).
FMNL1 is a primary example of the tremendous diversity of how a formin interacts with actin. This protein has been shown to nucleate, elongate, bundle, and sever actin filaments all while showing an indifference toward Rho-family GTPase interactions (Harris et al., 2004, Harris et al., 2006, Mersich et al., 2010, Miller et al.,
2017). While some studies have shown that FMNL1 interacts with RhoA, Rac1, and
Cdc42, others have demonstrated that these Rho-family GTPases have little to no impact on its activity (Yayoshi-Yamamoto et al., 2000, Seth et al., 2006, Gomez et al.,
2007, Mersich et al., 2010, Wang et al., 2015, Miller et al., 2017). Further adding to the complexities of this protein, the FH2 domain of FMNL1 has been shown to bind actin filaments via side-binding, where electrostatic interactions on the outside of the FH2 domain can interact with actin, allowing for binding or bundling activity (Harris et al.,
2004, Harris et al., 2006). FMNL1 also has a unique C-terminus, containing two putative, overlapping WH2 domains followed by five phenylalanine residues within very
28 close proximity to each other (Miller et al., 2017). In vivo studies have also demonstrated the importance of FMNL1 in regards to cellular functions, as widespread genetic deletion of this gene is embryonic lethal while conditional deletion in primary macrophages results in adhesion and migration defects (Miller et al., 2017).
Localization of FMNL1
Visualization of FMNL1 within cells has also demonstrated its variability.
Localization of both endogenous and exogenous expression has been observed not only to macrophage podosomes and pseudopods, but also to the nucleus, Golgi complex, lamella, lamellipodia, phagocytic cup, and myofibrils (Han et al., 2009, Mersich et al.,
2010, Colon-Franco et al., 2011, Naj et al., 2013, Rosado et al., 2014, Miller et al.,
2017). This could be attributed to its variety of interactions with actin as in vitro actin biochemistry experiments have demonstrated its ability to nucleate, elongate, bundle, and sever (Harris et al., 2004, Harriset al., 2006). This variance in localization could also be due to post-translational modifications. FMNL1 has been shown to be N-terminally myristoylated, resulting in its localization to the plasma membrane and subsequent contribution to membrane blebbing (Han et al., 2009).
Binding Partners of FMNL1
The promiscuous behavior of FMNL1 with its effectors and binding partners may also explain its diversity in function. As an inducer of microtubule acetylation, the functions of this protein may be broader than those of just the actin cytoskeleton. This is further supported by its ability to reorient the MTOC via Rac1 binding and evidence demonstrating its importance in spindle organization during mouse oocyte meiosis, where RhoA activates FMNL1 which can then influence GM130 expression (Gomez et al., 2007, Wang et al., 2015). FMNL1 also interacts srGAP2 (Slit-Robo GAP family
29 member 2) which results in a sterical hindrance of its severing function. This is the result of the SH3 domain of srGAP2 binding to the FH1 domain of Rac-activated FMNL1
(Mason et al., 2010). A bioinformatics analysis of the FMNL1 interactome and subsequent downstream experiments indicated that FMNL1ɣ interacts with the ~700kDa scaffold protein AHNAK1, mediating its localization to the plasma membrane (Han et al.,
2013). Additionally, on lipid droplets, FMNL1 has been shown to dictate assembly of non-muscle myosin IIa-decorated actin filaments (Pfisterer et al., 2017).
FMNL Proteins in Cancer
Several formins, especially those of the FMNL subfamily, have been implicated in cancer invasion and metastasis, which is not surprising considering their many cytoskeletal-based functions. In CRC cells, FMNL2 is required for migration, invasion, and metastasis, interacting with both Cdc42 and cortactin to drive invadopodia formation
(Ren et al., 2018). FMNL2 is also required for melanoma cell invasion, with one alternative splice isoform of this protein being upregulated (Péladeau et al., 2016).
Overexpression of FMNL3 can drive EMT in nasophayngeal carcinoma and deletion of this protein results in the lessening of the primary tumor to undergo EMT in xenografts
(Wu et al., 2017). Furthermore, depletion of FMNL1 in MDA-MB-231 cells results in an inhibition of invasion (Kitzing et al., 2010).
Formins as Actin Bundling Proteins
Detailed Formin History and Actin Bundle Organizations
In 1982, a study identifying phenotypes of mice homozygous for the mouse mutant limb-deformity gene laid the foundation for research into a new field of powerful, cytoskeleton-modifying proteins (Kleinebrecht et al., 1982). Further characterization of limb-deformity alternatively spliced mRNA transcripts predicted the potential protein
30 product structures (Woychik et al., 1990). This family of proteins was termed “formins” as a result of their contribution to the proper formation of both limbs and kidneys in mice
(Kleinbrecht et al., 1982, Zeller et al., 1989, Maas et al., 1990, Messing et al., 1990).
Subsequent research into the requirement of the diaphanous gene (diaphanous) for cytokinesis in Drosophila melanogaster revealed two evolutionarily conserved domains of protein products from both the limb-deformity gene (formin) of Mus muculus and
Gallus gallus domesticus and the Bni1 gene (Bni1) in Saccharomyces cerevisiae. The formin homology 1 (FH1) and formin homology 2 (FH2) domains were highly conserved in diaphanous, Bni1, and formin, with the FH2 domain eventually becoming the defining attribute of formin proteins (Castrillon and Wasserman, 1994). Analysis of sequence composition in proteins of other species allowed for identification of other formin family members, including, but not limited to, Cdc12 in Schizosaccharomyces pombe, Dia1 in
M. musculus and Homo sapiens, and Cyk-1 in Caenorhabditis elegans (Chang et al.,
1997, Watanabe et al., 1997, Bione et al., 1998, Swan et al., 1998). Eventually, Bni1 was identified as an actin-associated protein and hypothesized to participate in actin filament (F-actin) assembly (Evangelista et al., 1997). These data, in correlation with other studies identifying the N-terminal FH3 domain of formins and their interactions with
Rho-family GTPases, gave rise to the seminal paper first identifying the nucleation of unbranched actin filaments by a formin (Watanabe et al., 1997, Petersen et al., 1998,
Yayoshi-Yamamoto et al., 2000, Pruyne et al., 2002). This discovery of a novel protein family with the ability to nucleate unbranched actin filaments altered classical theories about how the actin cytoskeleton operates and introduced a family of proteins with an abundance of potential responsibilities in the cell. Sixteen years later, formins have been shown to not only nucleate linear actin filaments, but also to elongate, bundle, sever, and depolymerize them as well (Goode and Eck, 2007, Schonichen and Geyer,
2010, Breitpsrecher and Goode, 2013). These diverse actin-modifying activities of
31 formins designate these proteins as powerful instruments for the cell to use in order to modulate the cytoskeleton and, ultimately, fulfill a variety of essential cellular functions.
The foundational integrity of higher-ordered cellular structures, such as filopodia, lamellipodia, invadopodia, and microvilli, is dependent on not only the properties of the individual actin filaments themselves, but their aggregation into bundles as well (Mattila
PK and Lappalainen P, 2008, Brown and McKnight, 2010, Schoumacher et al., 2010,
Johnson et al., 2015). The manner in which these bundles are assembled, arranged, and oriented is contingent upon the function and abilities of the specific actin bundling proteins involved, as well as the subsequent involvement of cross-linking proteins, such as α-actinin and fascin (Bartles, 2000, Wagner et al., 2006, Courson and Rock, 2010).
While past studies have provided convincing evidence of Arp2/3 and Ena/VASP regulating actin filament bundle formation, recent work has demonstrated that specific formin proteins are also essential to this process (Vignjevic et al., 2003, Michelot et al.,
2005, Moseley and Goode, 2005, Harris et al., 2006, Schirenbeck et al., 2006, Quinlan et al., 2007, Vaillant et al., 2008, Li et al., 2010, Scott et al., 2011, Skillman et al., 2012,
Bohnert et al., 2013b, Jaiswal et al., 2013, Junemann et al., 2013, Schönichen et al.,
2013, Wang et al., 2013, Sun et al., 2017, Silkworth et al., 2018).
The formation of branched actin networks is primarily modulated by the seven- protein subunit complex Arp2/3 and families of nucleation promoting factors (NPFs) such as WASp (Wiskott-Aldrich syndrome protein) and Scar/WAVE (WASp-family verprolin- homologous protein) (Machesky and Insall, 1998, Mullins et al., 1998, Goley and Welch,
2006, Rotty et al., 2013). The force generated by these branched actin networks drive the mechanical and morphological properties of the cell, allowing for the completion of a variety of cellular functions (Goley and Welch, 2006, Rotty et al., 2013). However, different actin networks used for specific purposes can be generated within the cell as well. Crosslinked networks, which are also necessary for governing the morphological
32 and functional status of the cell, are different from the branched actin networks generated by the Arp2/3 complex in that they are produced from pre-formed actin filaments in conjunction with different proteins (Blanchoin et al., 2014). Cross-linking proteins vary in their ability and range to form networks. For example, while fimbrin can link actin filaments over a range of 10nm, the much larger protein filamin can link filaments up to 160nm (Klein et al., 2004, Stossel et al., 2001). Filaments that are cross- linked over a short distance, via proteins such as fascin, fimbrin, and α-actinin, are organized into tightly-packed bundles. These actin filament bundles can be organized in different fashions depending on the orientation of the filaments themselves (Matsudaira et al., 1983, Courson and Rock, 2010). Antiparallel actin bundle organizations are an important component of the actomyosin stress fibers necessary for the formation and regulation of adhesion complexes and are often formed by fimbrin and α-actinin (Laporte et al., 2012, Schwarz and Gardel, 2012). While parallel actin filament bundles can be formed via the Arp2/3 complex in the absence of capping protein, two other protein families are primarily considered to be important this mediating this specific type of filament organization: Ena/VASP and formins (Svitkina et al., 2003, Yang and Svitkina,
2011, Blanchoin et al., 2014). Ena/VASP proteins have the ability to generate unbranched actin filaments by enhancing barbed-end elongation and acting as anti- capping proteins for actin filaments. Additionally, this oligomeric protein has multiple actin binding sites, allowing for the bundling of filaments (Schirenbeck et al., 2006,
Trichet et al., 2008, Breitsprecher et al., 2011).
Classically, formins are thought of as actin nucleators which remain attached via their FH2 domain to the barbed end of the elongating actin filament, processively moving with this growing filament end as the FH1 domain binds profilin-bound monomeric actin which is added to the growing end of the filament (Goode and Eck, 2007, Paul and
Pollard, 2009, Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013). An
33 interesting component of formins is their ability to move from the barbed end of the filament to the sides, allowing for them to act as a crosslinker, organizing filaments into either parallel or antiparallel bundles (Harris et al., 2006). This versatile function makes formins unique in their bundling ability.
While this feature distinguishes formins from other crosslinking and bundling proteins, the mechanisms driving these actions remain to be fully elucidated. As many aspects of formin-mediated actin bundling have been revealed, much still remains to be discovered in regards to precisely how they bundle actin filaments. Interestingly, while several different formins can bundle, other actin-modifying abilities, such as nucleating, polymerizing, and severing, vary depending on protein. In conjunction with their diverse domain organization and how these domains interact with other proteins, defining formin functions has proven to be quite challenging. By providing a comprehensive review on the bundling function of formins, we hope to provide a systemic analysis that will allow us to identify patterns which may eventually contribute to the overall knowledge of this diverse family of proteins.
Saccharomyces cerevisiae – Bnr1
While two formins are expressed in S. cerevisiae, only one has been shown to bundle actin filaments (Kohno et al., 1996, Imamura et al., 1997, Moseley and Goode,
2005). Using low-speed co-sedimentation assays and electron microscopy, Bnr1, which localizes to the bud neck, was observed to display actin bundling activity at concentrations over 500nM (Moseley and Goode, 2005, Gao and Bretscher, 2009). The other formin expressed in S. cerevisiae, Bni1, localizes primarily to the bud tip as opposed to the bud neck where Bnr1 localizes (Vallen et al., 2000, Evangelista et al.,
2002, Sagot et al., 2002, Moseley et al., 2004, Pruyne et al., 2004, Moseley and Goode,
2005). This could explain why Bnr1 not only induces filament bundling but also acts as a
34 potent actin assembly protein. This bundling activity of Bnr1 was additionally confirmed with yeast actin as opposed to rabbit skeletal muscle actin, however, electron microscopy visually confirmed that the actin bundles are not well-organized (Wen and
Rubenstein, 2009). Interestingly, under low ionic strength conditions, Bnr1 could not bundle actin compared to high ionic strength conditions, indicating bundle formation might require higher ionic strengths which would allow access of the formin bundling motif to the surface of the actin filament. These data suggest that ionic strength- dependent conformation change in the formin or actin is required for the formin to interact with the filament side (Gao and Bretscher, 2009, Wen and Rubenstein, 2009).
Schizosaccharomyces Pombe – Fus1 and Cdc12
The fission yeast S. pombe expresses three different formins, Cdc12, For3, and
Fus1, which are important for cytokinesis, cell polarity, and mating, respectively (Chang et al., 1997, Petersen et al., 1998, Feierbach and Chang, 2001). Low-speed cosedimentation assays have demonstrated that Fus1-FH1-FH2 can bundle preassembled actin filaments with saturation occurring at ~0.5µM. This protein can also nucleate one filament per every two dimers and binds to barbed ends with a Kd of 0.2nM.
Compared to other fission yeast formins, it has a somewhat lower profilin-actin elongation rate of 5.0 subunits s-1 and dissociates from filaments more quickly than other yeast formins (Scott et al., 2011).
While Cdc12 is not able to bundle actin filaments on its own, a C-terminal fragment lacking the FH1 and FH2 domain but containing an oligomerization domain was able to bundle actin filaments in low-speed cosedimentation assays (Scott et al.,
2011, Bohnert et al., 2013b). Sid2 phosphorylates amino acid residues located C- terminally to the FH2 domain in Cdc12. Overexpression of mutants lacking the DAD domain and part of the C-terminus resulted in the formation of incomplete cytokinetic
35 rings and inhibited cytokinesis. A small, 87 amino acid residue sequence which includes part of the DAD domain and a region located C-terminally from it was identified as being sufficient for bundling while also overlapping with the region contributing to cytokinetic ring formation. Sequence analysis of this domain demonstrated that it was highly similar to RS domains important for oligomerization (Bohnert et al., 2013b). SDS-PAGE resolution of a fusion protein coding for this region of the molecule revealed that, in its native form, this region of the protein was multimeric. Additionally, in vitro binding assays of two separate fusion proteins made from this region indicated they self- associate (Boucher et al., 2001, Bohnert et al., 2013b).
Interestingly, cosedimentation assays of the region of Cdc12 located C-terminally to the FH2 domain and continuing on through to the C-terminus not only bound actin filaments, but bundled them as well. This bundling activity was confirmed visually via rhodamine phalloidin staining. While the oligomerization domain alone was able to bind actin filaments, it could not bundle them, indicating the entire C-terminal region of the molecule is necessary for this function. Microscopic analysis of different Cdc12 mutants demonstrated that oligomerization of the formin bundles actin early in the process of cytokinetic ring formation. Sid2 acts as a phosphoregulator of Cdc12 during cytokinesis, promoting multimer disassembly which is necessary for the maintenance of the cytokinetic ring. As a result, while Cdc12 in its native state is multimeric and functions as regulator of cytokinetic ring formation by mediating F-actin bundle formation, Sid2 can phosphorylate sites in the C-terminal region of this protein, resulting in disassembly of the multimer, halting actin bundling activity by this protein (Bohnert et al., 2013b).
36
Dictyostelium discoideum – ForC
D. discoideum expresses ten formins: ForA – ForJ. All D. discoideum formins have a similar domain structure as that of the metazoan formins with the exception of
ForC and ForI. While ForI lacks both an FH3 domain and a DAD domain, ForC lacks both an FH1 domain and the highly conserved isoleucine residue which has been shown to be essential for the actin binding in other formin family members (Shimada et al.,
2004, Xu et al., 2004, Otomo et al., 2004, Eichinger et al., 2005, Rivero et al., 2005,
Harris et al., 2006, Harris et al., 2010, Miller et al., 2017). ForC is also unique in that it has several stretches of highly repetitive amino acid residue sequences within the center of the FH2 domain (Junemann et al., 2013). Genome comparison demonstrated that the mammalian homologue with highest similarity in domain organization is FHOD1 (Dames et al., 2011).
Biochemical assays have demonstrated that the FH2 domain of ForC (ForC-FH2) does not stimulate spontaneous actin polymerization and actually inhibits assembly in a dose-dependent manner. Interestingly, ForC-FH2 does not inhibit depolymerization either, suggesting this protein may have a low affinity for barbed ends. In high-speed sedimentation assays, ForC-FH2 can bind F-actin with a KD ~ 0.9±0.2µM. Low-speed sedimentation assays have demonstrated that ForC-FH2 can bundle F-actin and that both the binding and bundling activities of this protein can be inhibited with increased
KCl concentrations. This suggests that FH2 domain function of ForC can vary based on ionic strengths and electrostatic interactions. Additionally, high-speed sedimentation assays where capping protein was included suggest that ForC-FH2 preferentially binds to actin filament sides (Junemann et al., 2013).
TIRF microscopcy demonstrates that ForC-FH2 has the ability to form ~2-6 actin bundles within 30mins. These bundles are flexible and loose with filaments in both parallel and antiparallel orientations (Junemann et al., 2013). This loose bundle
37 phenotype observed may be explained by the net positive charge of ForC-FH2 and how it has a higher pI than the non-bundling protein Dia1 but a lower pI than Dia2, which can bundle (Harris et al., 2006, Junemann et al., 2013). Elongation rates were not affected and the number of filaments formed did not change either, indicating ForC-FH2 does not affect nucleation. Since the FH2 domains of other formins that bundle were previously predicted to need a positive net charge and ForC-FH2 has a higher pI than the non- bundler Dia1, but a lower pI than Dia2 which does bundle, this may explain why looser bundles were formed (Harris et al., 2006, Junemann et al., 2013). This is further supported by evidence demonstrating that while the FH2 domain of FHOD1 has a higher pI than ForC –FH2, it is required for bundling activity (Junemann et al., 2013,
Schönichen et al., 2013). This could indicate that ForC-FH2 is sequestering actin monomers, effectively impairing filament elongation, resulting in the formation of thick actin bundles. (Junemann et al., 2013)
Drosophila melanogaster – Cappuccino and DAAM
The D. melanogaster formin Cappucino (Capu), a member of the FMN formin subfamily, is essential during development for proper oocyte polarity and genetic deletion of this gene results in female sterility (Manseau and Schüpbach, 1989, Emmons et al., 1995). The mammalian homolog, FMN2, is important for maintenance of the F- actin mesh structure in mouse oocytes (Dahlgaard et al., 2007). Capu has a longer lasso region of 52 amino acid residues compared to other formins which normally have
10-20 amino acid residues between two conserved tryptophan residues (Yoo et al.,
2015). Deletion of this region results in higher bundling activity compared to WT CapuC when analyzed via low-speed cosedimentation assays (Yoo et al., 2015). While the C- terminal tail region of Capu is necessary for F-actin bundling, it is not sufficient. Low- speed sedimentation assays and TIRF microscopy have demonstrated that recombinant
38 protein expressing Capu-FH1-FH2-TAIL-COOH bundles actin filaments and titrating F- actin with increasing concentrations of Capu results in actin filaments in a transition state between bundled and unbundled (Vizcarra et al., 2014). This bundling activity of Capu, with KD = 6 ± 2nM, is very efficient, even compared to α-actinin (Wachsstock et al., 1993,
Quinlan et al., 2007). Interestingly, mutation of the conserved isoleucine residue does not affect the bundling activity of Capu, as bundles of mixed orientations are still formed.
The requirement of the C-terminal tail region of Capu for bundling activity is further supported by F-actin bundling analysis using a truncated mutant, Capu-FH1-FH2, which has little to no effect compared to Capu-FH1-FH2-TAIL-COOH. The C-terminal tail requirement for Capu-mediated bundling could indicate that proper binding and bundling of F-actin relies on not only FH2 domain regions, but specific sequences found in this part of the molecule which may be necessary for electrostatic interactions to take place
(Vizcarra et al., 2014). Bundling activity of Capu has also been observed to occur with the N-terminal region (GBD-FH3), demonstrating that different formin domains may have other functions as well (Rosales-Nieves et al., 2006).
DAAM is an essential component of the actin cytoskeleton for the development of several different tissue types in D. melanogaster, especially for the organization of actin cables during tracheal development. Deletion of DAAM results in the shortening of actin cables which become cross-linked as opposed to bundled as parallel filaments (Matusek et al., 2006). Additionally, fluorescence microscopy analysis has demonstrated that both
Drosophila DAAM-FH1-FH2 and DAAM-FH2 have the ability to induce F-actin bundling
(Barkó et al., 2010). Interestingly, high-speed sedimentation assays using the C- terminal region of DAAM have shown that this protein can bind to F-actin sides with a KD
= 38.9 ± 3.2μM and removal of the C-terminal region following the DAD domain significantly reduces its binding ability. Low-speed sedimentation assays have shown that DAAM bundles F-actin with a similar affinity to its binding activity and mutation of the
39 conserved isoleucine residue in the FH2 domain has no effect on this. Moreover, a truncated mutant encoding solely for the DAD and C-terminal region of DAAM displays bundling activity as well, albeit with a much lower efficiency (Vig et al., 2017).
Mammals – Dia2, DAAM1, FMNL1-3, FHOD1, Delphilin
High-speed sedimentation assays have shown that Dia2 binds to actin filament sides and low-speed sedimentation assays have demonstrated it can indeed bundle actin filaments, which has been confirmed via fluorescence and electron microscopy.
Additionally, this bundling activity is dependent on the proper ionic conditions as increasing salt concentrations can ultimately inhibit proper bundling activity by Dia2. The orientation of F-actin bundles formed by Dia2 are both parallel and antiparallel, as determined by myosin-S1 decoration experiments. Interestingly, bundling activity is not influenced by mutation of the conserved isoleucine residue that is essential for the FH2 domain to interact with actin. Dia2-mediated F-actin bundling may be occurring as a result of the formin dimer dissociating and recombining. An interesting aspect of Dia2- mediated bundling is that this activity is not exclusive to barbed end binding activity, as demonstrated by the mutation of the conserved isoleucine residue (Harris et al., 2006).
Dia2 also has a dual role in lamellipodia and filopodia bundle formation and may contribute to bundle formation in the contractile ring as well (Yang et al., 2007,
Watanabe et al., 2008). Indeed, deletion of Dia2 in B16F1 cells results in disorganization of F-actin networks at the lamellipodia and in the filopodia (Yang et al.,
2007).
DAAM1 is an essential component of some filopodia and loss of expression of
DAAM1 in cells results in defects in these cellular structures. Low-speed sedimentation assays and TIRF (total internal reflection fluorescence) microscopy have demonstrated that the C-terminal half of DAAM1 has the ability to bundle F-actin in a concentration-
40 dependent manner. F-actin bundles generated by DAAM1 are organized in a parallel arrangement and can be formed regardless if the actin filaments are growing or preformed. Transmission electron microscopy has demonstrated that DAAM1 produces densely-packed F-actin bundles. DAAM1 is especially interesting in that it is recruited by fascin, a well-characterized actin bundling protein, for the proper formation of filopodia, and actually regulates fascin expression during oocyte meiosis in mice (Jaiswal et al.,
2013, Lu et al., 2017).
FMNL1 has the ability to bind tightly to pre-formed actin filaments, with different isoforms having a binding affinity of KD < 0.1μM. Moreover, while this protein can accelerate actin filament polymerization, it actually retards barbed end elongation (even in the presenece of profilin) and slows actin filament re-annealing. FMNL1 is also dimeric and binds tightly to preformed actin filaments sides which allows for successful actin filament bundling (Harris et al., 2004, Harris et al., 2006). Low-speed sedimentation assays have demonstrated that FMNL1 has the ability to bundle both muscle and non-muscle actin in a similar fashion. Indeed, in the sarcomeres of mouse primary cardiomyocytes, deletion of FMNL1 results in disorganized and less densely pack actin filament bundles (Harris et al., 2006, Rosado et al., 2014). Fluorescence microscopy has confirmed that the FH2 domain of FMNL1 bundles actin filament and
Cy3-labeled FMNL1 localizes evenly throughout the actin bundles. Ionic and electrostatic interactions also play a role in FMNL1-mediated F-actin bundling as increasing concentrations of NaCl from 50mM to 150mM results in an inhibition of both bundling and side-binding activity. Myosin-S1 decoration experiments have also demonstrated that filaments in FMNL1-generated bundles are in a mixed orientation of both parallel and anti-parallel orientation. Low-speed sedimentation assays with FMNL1 where the conserved isoleucine residue has been mutated also demonstrate that this does not affect the bundling activity of this protein, suggesting that other regions of
41
FMNL1 are important for this function (Harris et al., 2006). Quantitative rheology analysis has shown that FMNL1 can cross-link and bundle filaments into highly viscoelastic solids (Esue et al., 2008). Indeed, when FMNL1 is absent, actin filament network formation is limited and networks that do form have low elasticity and low dynamic viscosity. However, in the presence of FMNL1, elasticity of F-actin networks is able to increase 25-fold and the rate of gelation is greatly reduced. These FMNL1- generated F-actin networks form viscoelastic solids and the tight cross-linking where actin filament movement is very limited. Additionally, while FMNL1 has also been shown to sever actin filaments, this does not significantly affect its ability to form bundles (Harris et al., 2004, Esue et al., 2008). Interestingly, mutation of the conserved FH2 isoleucine residue in FMNL1, which inhibits barbed end binding, only slightly enhances the stiffness of the FMNL1-generated actin filament networks. This suggests that other regions of
FMNL1 may be important for formin-mediated F-actin bundling (Esue et al., 2008).
A prominent feature of the FMNL family is that all three members not only have the ability to bundle F-actin, but they also contain a WH2 motif located at the C-terminus
(Harris et al., 2006, Vaillant et al., 2008, Heimsath and Higgs, 2012). This WH2 motif is required for actin bundling by both FMNL2 and FMNL3, similar to its requirement for bundling activity in espin (Loomis et al., 2006, Vaillant et al., 2008). Low-speed sedimentation assays have demonstrated that both FMNL2 and FMNL3 bundle actin filaments in vitro. Interestingly, removal of the DAD domain from FMNL2 reduces actin bundling and bundling activity is completely eliminated when both the DAD and WH2 domains are absent. FMNL3-mediated bundling slightly differs in that, while reduced actin bundling was observed with the elimination of the DAD domain, absence of the
DAD and WH2 domain still allowed for some bundling activity, albeit at reduced amounts
(Vaillant et al., 2008, Heimsath and Higgs, 2012). FMNL2-mediated barbed end binding, in conjunction with WH2 actin binding sites, could be essential for proper F-actin
42 bundling by this protein. This could also explain why a FMNL2 mutant lacking the DAD domain was not observed to localize to bundled F-actin in microspikes and the lamellipodial network in B16-F1 cells (Vaillant et al., 2008, Block et al., 2012).
Additionally, FMNL2 deletion in mouse primary cardiomyocytes results in a lack of Z- band formation and actin filaments display disorganized orientations (Rosado et al.,
2014). However, as this WH2 domain is required for F-actin bundling by FMNL2, the bundling activity of FMNL3 is not as dependent on this mechanism. This suggests that both proteins could be bundling F-actin using different mechanisms, where FMNL3 may primarily rely on the FH2 domain for this activity (Vaillant et al., 2008). Furthermore, in
FMNL3, the conserved isoleucine residue of the FH2 domain, which has been found in several formins to be essential for barbed end binding and proper actin filament assembly, can be mutated to inhibit barbed end binding, but this does not affect bundling, suggesting other regions of the protein may be essential for this process (Xu et al., 2004, Harris et al., 2006, Harris et al., 2010, Colon-Franco et al., 2011).
FHOD1 has previously been shown to induce stress fiber formation and regulate adhesion complex formation while high-speed sedimentation assays have indicated that multiple regions of the protein have the ability to bind F-actin (Gasteier et al., 2003,
Takeya and Sumimoto, 2003 Schönichen et al., 2006, Gardberg et al., 2013,
Schönichen et al., 2013). However, low-speed sedimentation assays indicated that the
C-terminal region was not able to bundle F-actin, unlike several other bundling formins.
TIRF microscopy did reveal that full-length FHOD1 and an N-terminal region of the protein were able to bundle actin, with enhanced bundling noticed with F-actin bundled by the N-terminal protein, suggesting that DID-DAD interactions may inhibit bundling by this region of the protein. This has been further confirmed by electron microscopy where
F-actin bundles induced by FHOD1 were detected even in the presence of gelsolin, suggesting side-binding activity may be necessary for bundle formation. Additionally,
43 strong localization to actin arcs and ventral stress fibers, where F-actin bundles are prominent, have been observed with FHOD1. This is interesting as data on other formins suggests that the FH2 domain and C-terminal region are primarily responsible for bundling activity. However, this is not the case with FHOD1 where some bundling activity does seem to rely on an N-terminal region between the FH3 and FH1 domains
(Schönichen et al., 2013).
Analysis of the interaction between Delphilin (FH1-FH2-COOH) and F-actin via
TIRF microscopy has shown that this formin acts as nucleator. However, filament size comparisons have demonstrated that Delphilin-nucleated filaments are shorter when compared to control due to the protein remaining bound to the barbed end as the number of filaments increases with increasing concentrations of protein when compared to actin alone. The barbed end binding of Delphilin is tight (Ki = 0.6nM) and seeded elongation assays have shown that it is indeed responsible for inhibition of F-actin elongation. Additionally, Delphilin is also able to bind weakly to filament sides, with only one dimer per every 20 actin monomers and a KD ~ 3.0μM. As a result, while bundling activity is present with this protein, it is quite weak, regardless of whether the FH1 domain is present (Silkworth et al., 2018).
Arabidopsis thaliana – AFH1, AFH8, AFH14, AFH16
A. thaliana has 21 formins, separated into two subfamilies known as Class I and
Class II (Cvrcková, 2000, Deeks et al., 2002, Cvrcková et al., 2004, Cheung and Wu,
2004, Favery et al., 2004, Van Damme et al., 2004, Ingouff et al., 2005, Blanchoin and
Staiger, 2010). Class I formins are distinct in that they have a putative signal peptide and predicted N-terminal transmembrane domain which may result in localization of the formin to the plasma membrane while Class II formins have a varying domain structure
(Cvrcková et al., 2004). AFH1 not only acts as a nucleator but also binds to filament
44 sides and bundles F-actin. Fluorescence microscopy has revealed that AFH1-FH1-FH2 organizes individual actin filaments into long actin bundles, which has been confirmed via low-speed sedimentation assays (Michelot et al., 2005). AFH1 is a powerful bundling protein, even more efficient than the formidable bundling protein VILLIN1, and has a strong affinity for binding to actin filament sides with a KD ~ 0.13μM (Huang et al., 2005,
Michelot et al., 2005). Interestingly, AFH1-FH2-COOH has a much weaker bundling activity than AFH1-FH1-FH2, suggesting that the FH1 domain is essential for proper bundle formation. Light scattering assays have demonstrated that this protein bundles
F-actin rapidly, data that support in vivo work demonstrating that overexpression of this protein induces actin cable formation in pollen tubes (Cheng and Wu, 2004, Michelot et al., 2005). As AFH1-FH1-FH2 has a strong preference for actin filament side-binding, this could result in actin bundle formation which would be important for the formation of actin cables at the plasma membrane or other areas of the cell (Michelot et al., 2005).
Confocal and electron microscopy have both been utilized to analyze the ability of AFH8 to bind F-actin sides and bundle actin filaments (Yi et al., 2005, Xue et al., 2011). AFH8-
FH1-FH2 and AFH8-FH2 both bundle filaments, with longer and thicker bundles generated by the FH1 domain containing protein. Low-speed sedimentation assays have confirmed this observation. Interestingly, the nucleotide state of the actin filaments actually affected the bundling activity of AFH8. Fluorescence microscopy and low-speed sedimentation assays have both demonstrated that both AFH8-FH1-FH2 and AFH8-FH2 are able to bundle ATP/ADP-Pi-bound F-actin less efficiently than ADP-bound F-actin, indicating that the nucleotide state of the actin may play an important role in how formin- mediated bundling occurs. AFH8-FH1-FH2 generated F-actin bundles also display a unique stellar structural phenotype not observed by AFH-FH2 generated bundles, which could be explained by the interaction between the FH1 domain and profilin. Indeed, with the addition of profilin, fluorescence microscopy analysis revealed that when compared
45 to AFH8-FH1-FH2 in the absence of profilin, AFH8-FH1-FH2 with profilin actually generated fewer actin bundles, suggesting that profilin may play a role in mediating bundling by this protein (Xue et al., 2011).
Class II formins AFH14 and AFH16 both have the ability to bind and bundle F- actin (Li et al., 2010, Wang et al., 2013). Actin filament bundling by AFH14-FH1-FH2 has been observed via both fluorescence and electron microscopy. While this protein can both initiate and assemble F-actin bundles, disassembly can be induced via the addition of salts, suggesting that bundle formation requires proper ionic and electrostatic conditions. Interactions with microtubules may affect F-actin bundle formation by AFH14 as well. TIRF microscopy has confirmed that AFH14-FH1-FH2-formed bundles actually disassemble upon the addition of taxol-stabilized microtubules, suggesting that the formin has a preference for microtubule binding. An intriguing aspect of AFH14 is its ability to bundle both F-actin and microtubules. High concentrations of formin have the ability to induce bundle formation of both actin filaments and microtubules, which, upon observation with TIRF microscopy, actually co-localize. This observation, which has also been identified in D. melanogaster with the formin Cappuccino, suggests that
AFH14 may have the uncommon ability to assemble bundles of actin filaments and microtubules, possibly as a results of overlapping microtubule- and F-actin-binding domains (Rosales-Nieves et al., 2006, Li et al., 2010).
High-speed sedimentation assays have demonstrated that AFH16-FH1-FH2 can bind F-actin with a KD = 2.66 ± 0.58μM. Bundling by AFH16-FH1-FH2 and AFH16-FH2 has been confirmed via both low-speed sedimentation assays and electron microscopy, where tightly packed bundles were observed. When compared to F-actin bundles formed by AFH8, fluorescence microscopy demonstrated that AFH16-formed bundles differently than AFH8-formed bundles, which displayed a stellar-like structural phenotype, possibly due to the higher nucleation ability of this protein (Xue et al., 2011,
46
Wang et al., 2013). Additionally, similar to AFH14, AFH16 has a preference for bundling microtubules over F-actin. AFH16-FH1-FH2-formed F-actin bundles were observed by fluorescence microscopy to actually disassemble in the presence of taxol-stabilized microtubules, suggesting induced detachment of the protein from pre-formed F-actin bundles. This preferential microtubule binding has also been confirmed when AFH16-
FH1-FH2 was exposed to both microtubules and F-actin simultaneously, as microtubule bundles would indeed form but F-actin bundles remained absent (Wang et al., 2013).
Oryza sativa – OsFH5 and OsFH15
Similar to other plants, O. sativa expresses several different formins separated into type I and type II (Blanchoin and Staiger, 2010, Zhang et al., 2011, Huang et al.,
2013, Sun et al., 2017). Actin bundling by OsFH5, a type II formin, has been confirmed via fluorescence microscopy and confirmed with low-speed sedimentation assays, with the ability comparable to AFH1 (Michelot et al., 2005, Zhang et al., 2011). Both OsFH5-
FH1-FH2 and OsFH5-FH2 also have F-actin side-binding abilities, with Os-FH1-FH2 binding to F-actin with a KD = 0.18 ± 0.04μM (Zhang et al., 2011).
OsFH15-FH1-FH2 has the ability to nucleate both actin and profilin-bound actin but also acts as an actin elongation inhibitor. High-speed sedimentation analysis demonstrates that OsFH15-FH1-FH2 binds actin with a KD of 0.82 ± 0.07μM. TIRF analysis has demonstrated that, unlike some other formin bundling proteins, OsFH15 can bind the barbed end and act as an actin filament capper. Low-speed sedimentation assays have shown that both OsFH15-FH1-FH2 and OsFH15-FH2 both have the ability to bundle actin filaments. This has been further confirmed via fluorescence microscopy, where both OsFH15-FH1-FH2 and OsFH15-FH2 were observed to initiate and assemble
F-actin bundles. Similar to AFH14, microtubules may influence the interaction between the formin and F-actin. TIRF analysis has revealed that F-actin bundles actually loosen
47 when incubated with taxol-stabilized microtubules, completely disassembling after four hours, demonstrating that the formin will actually detach from F-actin bundles and preferentially interact with microtubules (Sun et al., 2017).
Toxoplasma gondii – TgFRM1 and TgFRM2
T. gondii, an intracellular Apicomplexan parasite that causes toxoplasmosis, relies on a specialized form of myosin-based motility to rearrange F-actin and invade cellular hosts (Desmonts and Couvreur, 1974, Skillman et al., 2012). Interestingly, this protozoan alveolate only contains one actin isoform (TgACTI) and completely lacks the
Arp2/3 complex (Gordon and Sibley, 2005, Sahoo et al., 2006). Amongst other actin- associated proteins present in this organism, the formins TgFRM1, TgFRM2, and
TgFRM3 have been identified as well. While all three proteins have been shown to nucleate rabbit actin and promote motility, TgFRM3 has also been shown to bind T. gondii actin as well (Daher et al., 2010, Daher et al., 2012, Skillman et al., 2012). The
FH1-FH2 domains of TgFRM1 and TgFRM2 were both shown to enhance TgACTI filament polymerization and fluorescence microscopy confirmed that both proteins had the ability to form clusters of short actin filament bundles. Interestingly, visualization of
TgFRM1-FH1-FH2-modified TgACTI filaments via electron microscopy revealed short, interconnected networks of actin. Furthermore, the filaments themselves were thicker and less-striated compared to WT filaments, suggesting that the formin may be binding to filament sides. Additionally, the presence of large, globular protein at the filament connection nodes suggested that the formins may be forming higher-order oligomers in the presence of TgACTI. However, TgFRM2-FH1-FH2-modified TgACTI filaments do not display the same phenotype as longer, straighter bundled filaments were observed.
These side-binding observations are similar to those observed with FMNL1, Dia2, and
48
AFH1, suggesting that this activity could be responsible for the bundling effects observed with these two T. gondii formins (Michelot et al., 2005, Harris et al., 2006,
Esue et al., 2008, Baum et al., 2008, Skillman et al., 2012).
The WH2 domain
General Description
The WH2 domain has been identified in numerous actin-associated proteins, several of which have different or multiple functions (Chereau et al., 2005, Campellone and Welch, 2010, Carlier et al., 2013, Dominguez, 2016). This domain is generally defined as an actin monomer subunit-binding motif composed of ~17-35 amino acid residues that occur in single or tandem repeats. The domain consists of an N-terminal
α-helix, which binds to the hydrophobic-binding cleft at the barbed end of an actin monomer subunit, and an extended C-terminal region that can bind to different subdomains of the pointed end of the actin monomer subunit (Chereau et al., 2005).
This C-terminal region contains what is considered a defining feature of WH2 domains and imperative to its actin-binding ability, the LKKV/T motif (Symons et al., 1996,
Dominguez, 2016). Furthermore, the hydrolytic state of the actin monomer subunit is critical to its interaction with WH2 motifs as they preferentially bind ATP-bound actin monomer subunits. This binding interaction is tight, with some WH2 domain-containing proteins binding ATP-bound actin monomer subunits at the nanomolar level (Mattila et al., 2003, Chereau et al., 2006). WH2 domain-containing proteins were originally hypothesized to primarily nucleate actin, as the binding of 2-4 actin monomer subunits is essential for efficient actin nucleation (Pollard and Borisy, 2003, Campellone and Welch,
2010). However, recent studies have demonstrated they may have several other functions as well (Loomis et al., 2006, Vaillant et al., 2008). Due to the small size
(5kDa), short identifying motif, and its N-terminal helical structure, the specific definition
49 of the WH2 motif remains slightly ambiguous in the literature. Therefore, in this dissertation, proteins referred to as “WH2-like” and “WH2-related” will be generally classified as “WH2.”
Non-formin WH2-domain Containing Proteins
The Arp2/3 complex requires WH2-domain containing activators (NPFs) in order to undergo a conformational change that allows for actin nucleation (Rotty et al., 2013).
While these proteins differ in their molecular structure, they do share a common C- terminus which includes WH2 domains, along with proline-rich, central, and acidic domains (Ti et al., 2011, Wagner et al., 2013). The proline-rich domain allows for binding of profilin-bound actin monomer subunits while the central and acidic domains interact with the Arp2/3 complex itself. The WH2, central, and acidic domains are often referred together as WCA domains. The WH2 domains found in these proteins effectively act as actin monomer subunit recruiters for Arp2/3 complex-generated actin filaments (Campellone and Welch, 2010, Rotty et al., 2013, Dominguez, 2016). For example, WASP is an effective actin nucleator and one of the most well-characterized.
In its nascent state, WASP is found in an inactive, autoinhibited conformation. However, upon interaction with its activators WIP or NCK1/2 and stabilizing proteins such as
Cdc42, autoinhibition is alleviated and the WCA domain of WASP can then proceed to stimulate a conformational change in the Arp2/3 complex allowing for effective actin nucleation (Antón et al., 2007, Tomasevic et al., 2007, Campellone and Welch, 2010).
Cobl was initially identified as an actin nucleator essential for brain development in mice and later characterized as a regulator of microvillus length (Ahuja et al., 2007,
Wayt and Bretscher, 2014). This protein contains three WH2 domains, however, the second and third WH2 domains are separated by ~65 amino acid residues (Ahuja et al.,
2007, Dominguez, 2016). The interesting aspect of Cobl is its requirement of an N-
50 terminal lysine-rich domain located N-terminally to the first WH2 domain for actin nucleation (Husson et al., 2011). Moreover, the unique composition of the WH2 and surrounding domains of Cobl can also bind and sever actin filaments in addition to sequestering monomer subunits (Husson et al., 2011, Schwintzer et al., 2011).
Spire, originally identified in D. melanogaster, actually contains four tandem WH2 domains with linker regions that stabilize actin monomer subunit binding (Quinlan et al.,
2005, Rebowski et al., 2008). Spire interacts with the formin Capuccino in D. melanogaster to generate actin structures required for oogenesis. Spire binds to the formin via its KIND domain interaction with the formin C-terminus and homodimerizes, effectively allowing for two sets of WH2 repeats to interact with actin and enhancing actin nucleation activity (Dahlgaard et al., 2007, Dominguez, 2016).
Mammalian Formins and WH2 Domains
Recent work has demonstrated that different mammalian formins do contain
WH2 domains, primarily located in the C-terminus. This is true for all three FMNL subfamily members, INF2, Dia1, and DAAM1. Both FMNL2 and FMNL3 incorporate the
WH2 domain into their bundling function, with FMNL2 actually requiring this domain for proper F-actin bundling (Vaillant et al., 2018). INF2 also utilizes a C-terminal WH2 domain to sever filaments as well as contribute to both actin filament polymerization and depolymerization (Chhabra and Higgs, 2006). In Dia1, the WH2 domains contribute to actin nucleation activity while in DAAM1 they are essential for actin filament assembly
(Gould et al., 2011). We have recently identified two putative WH2 domains in FMNL1 which uniquely overlap with each other. These WH2 domains will be discussed in further detail in Chapter 4.
51
Hypothesis/Signficance
The actin cytoskeleton has many functions and proper regulation of this part of the cytoskeleton as a whole requires specific proteins to interact in a precise manner.
Errant reorganization of the actin cytoskeleton can result in cancer cells from a primary tumor disseminating, making their way through the metastatic cascade and eventually forming secondary metastatic sites. Therefore, it is important to understand how actin- modifying proteins operate and if upregulation affects cancer cell locomotion.
The formin family of proteins is still a relatively new field and much remains to be uncovered as to how formins operate. Classically, work has focused on nucleation and assembly of actin filaments. However, evidence suggests that formins can modify actin in several ways, including severing and bundling. Furthermore, domains other than the hallmark FH2 domain should be researched as it is becoming more and more clear that other parts of the formin molecule are necessary for their interactions.
Our lab has focused its work on the formin FMNL1 and have recently demonstrated that a specific splice isoform (FMNL1ɣ) is contributing to actin filament modification through a mechanism other than the FH2 domain in primary human macrophages. We hypothesize that FMNL1ɣ is actually bundling actin filaments through its WH2 domains which ultimately contributes to cellular adhesion and locomotion.
Using a breast cancer model and in vitro biochemical experiments, our goal is to elucidate the mechanism behind this bundling activity and determine how it contributes to cancer cell motility and adhesion.
52
References
Abercrombie M and Ambrose EJ (1958). Interference microscope studies of cell contacts in tissue culture. Exp Cell Res 15, 332-345.
Abercrombie M and Heaysman JE (1953). Observations on the social behaviour of cells in tissue culture. I. Speed movement of chick heart fibroblasts in relation to their mutual contacts. Exp Cell Res 5, 111-131.
Abercrombie M, Heaysman JE, Pegrum SM (1970a). The locomotion of fibroblasts in culture. I. Movements of the leading edge. Exp Cell Res 59, 393-398.
Abercrombie M, Heaysman JE, Pegrum SM (1970b). The locomotion of fibroblasts in culture. II. “Ruffling.” Exp Cell Res 60, 437-444.
Abercrombie M, Heaysman JE, Pegrum SM (1970c). The locomotion of fibroblasts in culture. III. Movements of particles on the dorsal surface of the leading lamella. Exp Cell Res 62, 389-398.
Abercrombie M, Heaysman JE, Pegrum SM (1971). The locomotion of fibroblasts in culture. IV. Electron microscopy of the leading lamella. Exp Cell Res 67, 359-367.
Abercrombie M, Heaysman JE, Pegrym SM (1972). The locomotion of fibroblasts in culture. V. Surface marking with concanavalin A. Exp Cell Res 73, 536-539.
Ahuja R, Pinyol R, Reichenbach N, Custer L, Klingensmith J Kessels MM, Qualmann B (2007). Cordon-bleu is an actin nucleation factor and controls neuronal morphology. Cell 131, 337-350.
Ansel KM, Harris RB, Cyster JG (2002). CXCL13 is required for B1 cell homing, natural antibody production, and body cavity immunity. Immunity 16, 67-76.
Antón IM, Jones GE, Wandosell F, Geha R, Ramesh N (2007). WASP-interacting protein (WIP): working in polymerization and much more. Trends Cell Biol 17, 555-562.
Arpino G, Bardou VJ, Clark GM, Elledge RM (2004). Infiltrating lobular carcinoma of the breast: tumor characteristics and clinical outcome. Breast Cancer Res 6, R149-156.
Aspenström P (2010). Formin-binding proteins: modulators of formin-dependent actin polymerization. Biochim Biophys Acta 1803, 174-182.
Badve S, Dabbs DJ, Schnitt SJ, Baehner FL, Decker T, Eusebi V, Fox SB, Ichihara S, Jacquemier J, Lakhani SR, Palacios J, Rakha EA, Richardson AL, Schmitt FC, Tan PH, Tse GM, Weigelt B, Ellis IO, Reis-Filho JS (2011). Basal-like and triple-negative breast cancers: a critical review with an emphasis on the implications for pathologists and oncologists. Mod Pathol 24, 157-167.
Barczyk M, Carracedo S, Gullberg D (2010). Integrins. Cell Tisue Res 339, 269-280.
53
Barkó S, Bugyi B, Carlier MF, Gombos R, Matusek T, Mihály J, Nyitrai M (2010). Characterization of the biochemical properties and biological function of the formin homology domains of Drosophila DAAM. J Biol Chem 285, 13154-13169.
Bartles JR (2000). Parallel actin bundles and their multiple actin-bundling proteins. Curr Opin Cell Biol 12, 72-78.
Bartolini F, Andres-Delgado L, Qu X, Nik S, Ramalingam N, Kremer L, Alonso MA, Gundersen GG (2016). An mDia1-INF2 formin activation cascade facitliated by IQGAP1 regulates stable microtubules in migrating cells. Mol Biol Cell 27, 1797-1808.
Baum J, Tonkin CJ, Paul AS, Rug M, Smith BJ, Gould SB, Richard D, Pollard TD, Cowman AF (2008). A malaria parasite formin regulates actin polymerization and localizes to the parasite-erythrocyte moving junction during invasion. Cell Host Microbe 3, 188-198.
Bendel-Stenzel M, Anderson R, Heasman J, Wylie C (1998). The origin and migration of primordial germ cells in the mouse.
Bione S, Sala C, Manzini C, Arrigo G, Zuffardi O, Banfi S, Borsani G, Jonveaux P, Philippe C, Zuccotti M, Ballabio A, Toniolo D (1998). A human homologue of the Drosophila melanogaster diaphanous gene is disrupted in a patient with premature ovarian failure: evidence for conserved function in oogenesis and implications for human sterility. Am J Hum Genet 62, 533-541.
Blanchoin L, Boujemaa-Paterski R, Sykes C, Plastino J (2014). Actin dynamics, architecture, and mechanics in cell motility. Physiol Rev 94, 235-263.
Blanchoin L and Staiger CJ (2010). Plant formins: diverse isoforms and unique molecular mechanism. Biochim Biophys Acta 1803, 201-206.
Block J, Breitsprecher D, Kühn S, Winterhoff M, Kage F, Geffers R, Duwe P, Rohn JL, Baum B, Brakebusch C, Geyer M, Stradal TE, Faix J, Rottner K (2012). FMNL2 drives actin-based protrusion and migration downstream of Cdc42. Curr Biol 22, 1005-1012.
Bohnert KA, Grzegorzewska AP, Willet AH, Vander Kooi CW, Kovar DR, Gould KL (2013b). SIN-dependent phosphoinhibition of formin multimerization controls fission yeast cytokinesis. Genes Dev 27, 2164-2177.
Bohnert KA, Willet AH, Kovar DR, Gould KL (2013a). Formin-based control of the actin cytoskeleton during cytokinesis. Biochem Soc Trans 41, 1750-1754.
Bonnans C, Chou J, Werb Z (2014). Remodelling the extracellular matrix in development and disease. Nat Rev Mol Cell Biol 15, 786-801.
Boucher L, Ouzounis CA, Enright AJ, Blencowe BJ (2001). A genome-wide survey of RS domain proteins. RNA 7, 1693-1701.
Brandt DT, Marion S, Griffiths G, Watanabe T, Kaibuchi K, Grosse R (2007). Dia1 and IQGAP1 interact in cell migration and phagocytic cup formation. J Cell Biol 178, 193- 200.
54
Bravo-Cordero JJ, Hodgson L, Condeelis J (2012). Directed cell invasion and migration during metastasis. Curr Opin Cell Biol 24, 277-283.
Breitsprecher D and Goode BL (2013). Formins at a glance. J Cell Sci 126, 1-7.
Breitsprecher D, Kiesewetter AK, Linkner J, Vinzenz M, Stradal TE, Small JV, Curth U, Dickinson RB, Faix J (2011). Molecular mechanism of Ena/VASP-mediated actin- filament elongation. EMBO J 30, 456-467.
Brown JW and McKnight CJ (2010). Molecular model of the microvillar cytoskeleton and organization of the brush border. PLoS One 5, e9406.
Bryan BA and D’Amore PA (2007). What tangled webs they weave: Rho-GTPase control of angiogenesis. Cell Mol Life Sci 64, 2053-2065.
Burnier JV, Wang N, Michel RP, Hassanain M, Li S, Lu Y, Metrakos P, Antecka E, Burnier MN, Ponton A, Gallinger S, Brodt P (2011). Type IV collagen-initiated signals provide survival and growth cues required for liver metastasis. Oncogene 30, 3766- 3783.
Campellone KG and Welch MD (2010). A nucleator arms race: cellular control of actin assembly. Nat Rev Mol Cell Biol 11, 237-251.
Cancer Genome Atlas Network (2012). Comprehensive molecular portraits of human breast tumours. Nature 490, 61-70.
Cao J, Chiarelli C, Richman O, Zarrabi K, Kozarekar P, Zucker S (2008). Membrane type 1 matrix metalloproteinase induces epithelial-to-mesenchymal transition in prostate cancer. J Biol Chem 283, 6232-6240.
Carlier MF, Pernier J, Avvaru BS (2013). Control of actin filament dynamics at barbed ends by WH2 domains: from capping to processive assembly.
Castrillon DH and Wasserman SA (1994). Diaphanous is required for cytokinesis in Drosophila and shares domains of similarity with the products of the limb deformity gene. Development 120, 3367-3377.
Chaffer CL and Weinberg RA (2011). A perspective on cancer cell metastasis. Sceince 331, 1559-1564.
Chang F, Drubin D, Nurse P (1997). cdc12p, a protein required for cytokinesis in fission yeast, is a component of the cell division ring and interacts with profilin. J Cell Biol 137, 169-182.
Cheang MC, Voduc D, Bajdik C, Leung S, McKinney S, Chia SK, Perou CM, Nielsen TO (2008). Basal-like breast cancer defined by five biomarkers has superior prognostic value than triple-negative phenotype. Clin Cancer Res 14, 1368-1376.
Chen W, Hoffmann AD, Liu H, Liu X (2018). Organotropism: new insights into molecular mechanisms of breast cancer metastasis. NPJ Precis Oncol 2, 4.
55
Chereau D, Kerff F, Graceffa P, Grabarek Z, Langsetmo K, Dominguez R (2005). Actin- bound structures of Wiskott-Aldrich syndrome protein (WASP)-homology domain 2 and the implications for filament assembly. Proc Natl Acad Sci USA 102, 16644-16649.
Chesarone MA, DuPage AG, Goode BL (2010). Unleashing formins to remodel the actin and microtubule cytoskeletons. Nat Rev Mol Cell Biol 11, 62-74.
Cheung AY and Wu HM (2004). Overexpression of an Arabidopsis formin stimulates supernumerary actin cable formation from pollen tube cell membrane. Plant Cell 16, 257-269.
Chong C, Tan L, Lim L, Manser E (2001). The mechanism of PAK activation. Autophosphorylation events in both regulatory and kinase domains control activity. J Biol Chem 276, 17347-17353.
Clark AG and Vignjevic DM (2015). Modes of cancer cell invasion and the role of the microenvironment. Curr Opin Cell Biol 36, 13-22.
Coleman RE (2001). Metastatic bone disease: clinical features, pathophysiology and treatments strategies. Cancer Treat Rev 27, 165-176.
Colon-Franco JM, Gomez TS, Billideau DD (2011). Dynamic remodeling of the actin cytoskeleton by FMNL1ɣ is required for structural maintenance of the Golgi complex. J Cell Sci 124, 3118-3126.
Courson DS and Rock RS (2010). Actin cross-link assembly and disassembly mechanics for α-actinin and fascin. J Biol Chem 285, 26350-26357.
Cox TR and Erler JT (2011). Remodeling and homeostasis of the extracellular matrix: implications for fibrotic diseases and cancer. Dis Model Mech 4, 165-178.
Cvrcková F (2000). Are plant formins integral membrane proteins? Genome Biol 1, RESEARCH001.
Cvrcková F, Novotný M, Pícková D, Zárský V (2004). Formin homology 2 domains occur in multiple contexts in angiosperms. BMC Genomics 5, 44.
Daher W, Klages N, Carlier MF, Soldati-Favre D (2012). Molecular characterization of Toxoplasma gondii formin 3, an actin nucleator dispensable for tachyzoite growth and motility. Eukaryot Cell 11, 343-352.
Daher W, Plattner F, Carlier MF, Soldati-Favre D (2010). Concerted action of two formins in gliding motility and host cell invasion by Toxoplasma gondii. PLoS Pathog 6, e1001132.
Dahlgaard K, Raposo AA, Niccoli T, St Johnston D (2007). Capu and Spire assembly a cytoplasmic actin mesh that maintains microtubule organization in the Drosophila oocyte. Dev Cell 13, 539-553.
56
Dames SA, Junemann A, Sass HJ, Schönichen A, Stopschinski BE, Grzeisek S, Faix J, Geyer M (2011). Structure, dynamics, lipid binding, and physiological relevance of the putative GTPase-binding domain of Dictyostelium formin C. J Biol Chem 286, 36907- 36920.
De Craene B and Berx G (2013). Regulatory networks defining EMT during cancer initiation and progression. Nat Rev Cancer 13, 97-110.
Deeks MJ, Hussey PJ, Davies B (2002). Formins: intermediates in signal-transduction cascades that affect cytoskeletal reorganization. Trends Plant Sci 7, 492-498.
Desmonts G and Couvreur J (1974). Congenital toxoplasmosis: a prospective study of 378 pregnancies. N Engl J Med 290, 1110-1116.
Devreotes P and Janetopoulos C (2003). Eukaryotic chemotaxis: distinctions between directional sensing and polarization.
DeWard AD, Eisenmann KM, Matheson SF, Alberts AS (2010). The role of formins in human disease. Biochim Biophys Acta 1803, 226-233.
Dominguez R (2016). The WH2 Domain and Actin Nulceation: Necessary but Insufficient. Trends Biochem Sci 41, 478-490.
Egeblad M, Rasch MG, Weaver VM (2010). Dynamic interplay between the collagen scaffold and tumor evolution. Curr Opin Cell Biol 22, 697-706.
Ehrlich JS, Hansen MD, Nelson WJ (2002). Spatio-temporal regulation of Rac1 localization and lamellipodia dynamics during epithelial cell-cell adhesion. Dev Cell 3, 259-270.
Eichinger L, Pachebat JA, Glöckner G, et al. (2005). The genome of the social amoeba Dictyostelium discoideum. Nature 435, 43-57.
Ellerbroek SM, Wu YI, Overall CM, Stack MS (2001). Functional interplay between type I collagen and cell surface matrix metalloproteinase activity. J Biol Chem 276, 24833- 24842.
El Hallani S, Boisselier B, Peglion F, Rousseau A, Colin C, Idbaih A, Marie Y, Mokhtari K, Thomas JL, Eichmann A, Delattre JY, Maniotis AJ, Sanson M (2010). A new alternative mechanism in glioblastoma vascularization: tubular vasculogenic mimicry. Brain 133, 973-982.
Emmons S, Phan H, Calley J, Chen W, James B, Manseau L (1995). Cappuccino, a Drosophila maternal effect gene required for polarity of the egg and embryo, is related to the vertebrate limb deformity locus. Genes Dev 9, 2482-2494.
Esue O, Harris ES, Higgs HN, Wirtz D (2008). The filamentous actin cross- linking/bundling activity of mammalian formins. J Mol Biol 384, 324-334.
Etienne-Manneville S and Hall A (2003). Cell polarity: Par6, aPKC and cytoskeletal crosstalk. Curr Opin Cell Biol 15, 67-72.
57
Evangelista M, Blundell K, Longtine MS, Chow CJ, Adames N, Pringle JR, Peter M, Boone C (1997). Bni1p, a yeast formin linking cdc42p and the actin cytoskeleton during polarized morphogenesis. Science 118-122.
Evangelista M, Pruyne D, Amberg DC, Boone C, Bretscher A (2002). Formins direct Arp2/3-inedependent actin filament assembly to polarize cell growth in yeast. Nat Cell Biol 4, 260-269.
Ewing J (1928). Neoplastic Diseases: A Treatise on Tumors. 3rd ed. Philadelphia (PA): WB Saunders Co. Ltd.
Fadoukhair Z, Zardavas D, Chad MA, Goulioti T, Aftimos P, Piccart M (2016). Evaluation of targeted therapies in advanced breast cancer: the need for large-scale molecular screening and transformative clinical trial designs. Oncogene 35, 1743-1749.
Farmer P, Bonnefoi H, Becette V, Tubiana-Hulin M, Fumoleau P, Larsimont D, Macgrogan G, Bergh J, Cameron D, Goldstein D, Duss S, Nicoulaz AL, Brisken C, Fiche M, Delorenzi M, Iggo R (2005). Identification of molecular apocrine breast tumours by microarray analysis. Oncogene 24, 4660-4671.
Favery B, Chelysheva LA, Lebris M, Jammes F, Marmagne A, De Almeida-Engler J, Lecomte P, Vaury C, Arkowitz RA, Abad P (2004). Arabidopsis formin AtFH6 is a plasma-associated protein upregulated in giant cells by parasitic nematodes. Plant Cell 16, 2529-2540.
Feierbach B and Chang F (2001). Roles of the fission yeast for3p in cell polarity, actin cable formation and symmetric cell division. Curr Biol 11, 1656-1665.
Fernández-Barrera J and Alonso MA (2018). Coordination of microtubule acetylation and the actin cytoskeleton by formins. Cell Mol Life Sci 75, 3181-3191.
Ferrari A, Vincent-Salomon A, Pivot X, et al. (2016). A whole-genome sequence and transcriptome on HER2-positive breast cancers. Nat Commun 7, 12222.
Fidler IJ and Kripke ML (1977). Metastasis from preexisting variant cells within a malignant tumor. Sceince 197, 893-895.
Fidler IJ and Kripke ML (2015). The challenge of targeting metastasis. Cancer Metastasis Rev 34, 635-641.
Gao L and Bretscher A (2009). Polarized growth in budding yeast in the absence of localized formin. Mol Biol Cell 20, 2540-2548.
Gardberg M, Heuser VD, Iljin K, Kampf C, Uheln M, Carpen O (2014). Characterization of leukocyte formin FMNL1 expression in human tissues. J Histochem Cytochem 62, 460-470.
Gardberg M, Kaipio K, Lehtinen L, Mikkonen P, Heuser VD, Talvinen K, Iljin K, Kampf C, Uhlen M, Grénman R, Koivisto M, Carpén O (2013). FHOD1, a formin upregulated in epithelial-mesenchymal transition, participates in cancer cell migration and invasion. PLoS One 8, e74923.
58
Gasteier JE, Madrid R, Krautkrämer E, Schröder S, Muranyi W, Benichou S, Fackler OT (2003). Activation of the Rac-binding partner FHOD1 induces actin stress fibers via a ROCK-dependent mechanism. J Biol Chem 278, 38902-38912.
Geyer FC, Lopez-Garcia MA, Lambros MB, Reis-Filho JS (2009). Genetic characterization of breast cancer and implications for clinical management. J Cell Mol Med 13, 4090-4103.
Glentis A, Gurchenkov V, Matic Vignjevic D (2014). Assembly, heterogeneity, and breaching of the basement membranes. Cell Adh Migr 8, 236-245.
Goley ED and Welch MD (2006). The Arp2/3 complex: an actin nucleator comes of age. Nat Rev Mol Cell Biol 7, 713-726.
Gomez TS, Kumar K, Medeiros RB, Shimizu Y, Leibson PJ, Billadeau DD (2007). Formins regulate the actin-related protein 2/3 complex-independent polarization of the centrosome to the immunological synapse. Immunity 26, 177-190.
Goode BL and Eck MJ (2007). Mechanism and function of formins in the control of actin assembly. Annu Rev Biochem 76, 593-627.
Gordon JL and Sibley LD (2005). Comparative genome analysis reveals a conserved family of actin-like proteins in apicomplexan parasites. BMC Genomics 6, 179.
Guedj M, Marisa L, de Reynies A, Orsetti B, Schiappa R, Bibeau F, MacGrogan G, Lerebours F, Finetti P, Longy M, Bertheau P, Bertrand F, Bonnet F, Martin AL, Feugeas JP, Bièche I, Lehmann-Che J, Lidreau R, Birnbaum D, Bertucci F, de Thé H, Theillet C (2012). A refined molecular taxonomy of breast cancer. Oncogene 31, 1196-1206.
Han Y, Eppinger E, Schuster IG, Weigand LU, Liang X, Kremmer E, Peschel C, Krackhardt AM (2009). Formin-like 1 (FMNL1) is regulated by N-terminal myristoylation and induces polarized membrane blebbing. J Biol Chem 284, 33409-33417.
Han Y, Yu G, Sarioglu H, Caballero-Martinez A, Schlott F, Ueffing M, Haase H, Peschel C, Krackhardt AM (2013). Proteomic investigation of the interactome of FMNL1 in hematopoietic cells unveils a role in calcium-dependnent membrane plasticity.
Hansson GK and Libby P (2006). The immune response in atherosclerosis: a double- edged sword. Nat Rev Immunol 6, 508-519.
Haraguchi M, Okubo T, Miyashita Y, Miyamoto Y, Hayashi M, Crotti TN, McHugh KP, Ozawa M (2008). Snail regulates cell-matrix adhesion by regulation of the expression of integrins and basement membrane proteins. J Biol Chem 283, 23514-23523.
Harney AS, Arwert EN, Entenberg D, Wang Y, Guo P, Qian BZ, Oktay MH, Pollard JW, Jones JG, Condeelis JS (2015). Real-time imaging reveals local, transient vascular permeability, and tumor cell intravasation stimulated by the TIE2hi macrophage-derived VEGFA. Cancer Discov 5, 932-943.
Harris ES, Gauvin TJ, Heimsath EG, Higgs HN (2010). Assembly of filopodia by the formin FRL2 (FMNL3). Cytoskeleton (Hoboken) 67, 755-772.
59
Harris ES, Li F, Higgs HN (2004). The mouse formin, FRLα, slows actin filament barbed end elongation, competes with capping protein, accelerates polyermization from monomers, and severs filaments. J Biol Chem 279, 20076-20087.
Harris ES, Rouiller I, Hanein D, Higgs HN (2006). Mechanistic differences in actin bundling activity of two mammalian formins, FRL1 and mDia2. J Biol Chem 281, 14383- 14392.
Hart IR and Fidber IJ (1980). Role of organ selectivity in the determination of metastatic patterns of B16 melanoma. Cancer Res 40, 2281-2287.
He ZY, Wu SG, Yang Q, Sun JY, Li FY, Lin Q, Lin HX (2015). Breast Cancer Subtype is Associated With Axillary Lymph Node Metastasis: A Retrospective Cohort Study. Medicine (Baltimore) 94, e2213.
Heimsath EG Jr and Higgs HN (2012). The C terminus of formin FMNL3 accelerates actin polymerization and contains a WH2 domain-like sequence that binds both monomers and filament barbed ends. J Biol Chem 287, 3087-3098.
Higgs HN and Peterson KJ (2005). Phylogenetic analysis of the formin homology 2 domain. Mol Biol Cell 16, 1-13.
Huang J, Kim CM, Xuan YH, Liu J, Kim TH, Kim BK, Han CD (2013). Formin homology 1 (OsFH1) regulates root-hair elongation in rice (Oryza sativa). Planta 237, 1227-1239.
Huang S, Robinson RC, Gao LY, Matsumoto T, Brunet A, Blanchoin L, Staiger CJ (2005). Arabidopsis VILLIN1 generates actin filament cables that are resistant to depolymerization. Plant Cell 17, 486-501.
Husson C, Renault L, Didry D, Pantaloni D, Carlier MF (2011). Cordon-Bleu uses WH2 domains as multifunctional dynamizers of actin filament assembly. Mol Cell 43, 464-477.
Imamura H, Tanaka K, Hihara T, Umikawa M, Kamei T, Takahashi K, Sasaki T, Takai Y (1997). Bni1p and Bnr1p: downstream targets of the Rho family small G-proteins which interact with profiling and regulate actin cytoskeleton in Saccharomyces cerevisiae. EMBO J 16, 2745-2755.
Ingouff M, Fitz Gerald JN, Guérin C, Robert H, Sørensen MB, Van Damme D, Geelen D, Blanchoin L, Berger F (2005). Plant formin AtFH5 is an evolutionarily conserved actin nucleator involved in cytokinesis. Nat Cell Biol 7, 374-380.
Jaiswal R, Breitsprecher D, Collins A, Corrêa IR Jr, Xu MQ, Goode BL (2013). The formin Daam1 and fascin directly collaborate to promote filopodia formation. Curr Biol 23, 1373-1379.
Jatoi I, Hilsenbeck SG, Clark GM, Osborne CK (1999). Significance of axillary lymph node metastasis in primary breast cancer. J Clin Oncol 17, 2334-2340.
Joberty G, Petersen C, Gao L, Macara IG (2000). The cell-polarity protein Par6 links Par3 and the atypical protein kinase C to Cdc42. Nat Cell Biol 2, 531-539.
60
Johnson HE, King SJ, Asokan SB, Rotty JD, Bear JE, Haugh JM (2015). F-actin bundles direct the initiation and orientation of lamellipodia through adhesion-based signaling. J Cell Biol 208, 443-455.
Jonasson L, Holm J, Skalli O, Bondjers G, Hansson GK (1986). Regional accumulation of T cells, macrophages, and smooth muscle cells in the human atherosclerotic plaque. Arteriosclerosis 6, 131-138.
Junemann A, Winterhoff M, Nordholz B, Rottner K, Eichinger L Gräf R, Faix J (2013). ForC lacks canonical formin activity but bundles actin filaments and is required for multicellular development of Dictyostelium cells. Eur J Cell Biol 92, 201-212.
Kabsch W, Mannherz HG, Suck D, Pai EF, Holmes KC (1990). Atomic structure of the actin:DNase I complex. Nature 347, 37-44.
Katoh M and Katoh M (2003). Identification and characterization of human FMNL1, FMNL2 and FMNL3 genes in silico. Int J Oncol 22, 1161-1168.
Katt ME, Wong AD, Searson PC (2018). Dissemination from a Solid Tumor: Examining the Multiple Parallel Pathways. Trends Cancer 4, 20-37.
Kennecke H, Yerushalmi R, Woods R, Cheang MC, Voduc D, Speers CH, Nielsen TO, Gelmon K (2010). Metastatic behavior of breast cancer subtypes. J Clin Oncol 28, 3271- 3277.
Kimbung S, Johansson I, Danielsson A, Veerla S, Egyhazi Brage S, Frostvik Stolt M, Skoog L, Carlsson L, Einbeigi Z, Lidbrink E, Linderholm B, Loman N, Malmström PO, Söderberg M, Walz TM, Fernö M, Hatschek T, Hedenfalk I, TEX study group (2016). Transcriptional Profiling of Breast Cancer Metastasis Identifies Liver Metastasis- Selective Genes Associated with Adverse Outcome in Luminal A Primary Breast Cancer. Clin Cancer Res 22, 146-157.
Kitzing TM, Wang Y, Pertz O, Copeland JW, Grosse R (2010). Formin-like 2 drives amoeboid invasive cell motility downstream of RhoC. Oncogene 29, 2241-2248.
Klein MG, Shi W, Ramagopal U, Tseng Y, Wirtz D, Kovar DR, Stiager CJ, Almo SC (2004). Structure of the actin crosslinking core of fimbrin. Structure 12, 999-1013.
Kleinebrecht J, Selow J, Winkler W (1982). The mouse mutant limb-deformity (ld). Anat Anz 152, 313-324.
Knaus UG and Bokoch GM (1998). The p21Rac/Cdc42-activated kinases (PAKs). Int J Biochem Cell Biol 30, 857-862.
Kohno H, Tanaka K, Mino A, Umikawa M, Imamura H, Fujiwara T, Fujita Y, Hotta K, Qadota H, Watanabe T, Ohya Y, Takai Y (1996). Bni1p implicated in cytoskeletal control is a putative target of Rho1p small GTP binding protein in Saccharomyces cerevisiae. EMBO J 15, 6060-6068.
61
Krugmann S, Jordens I, Gevaert K, Driessens M, Vandekerckhove J, Hall A (2001). Cdc42 induces filopodia by promoting the formation of an IRSp53:Mena complex. Curr Biol 11, 1645-1655.
Kunwar PS, Siekhaus DE, Lehmann R (2006). In vivo migration: a germ cell perspective. Annu Rev Cell Dev Biol 22, 237-265.
Kurosaka S and Kashina A (2008). Cell biology of embryonic migration. Birth Defects Res C Embryo Today 84, 102-122.
Lamouille S, Xu J, Derynck R (2014). Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol 15, 178-196.
Laporte D, Ojkic N, Vavylonis D, Wu Q (2012). α-Actinin and fimbrin cooperate with myosin II to organize actomyosin bundles during contractile-ring assembly. Mol Biol Cell 23, 3094-3110.
Leber MF and Efferth T (2009). Molecular principles of cancer invasion and metastasis. Int J Oncol 34, 881-895.
Levental KR, Yu H, Kass L, Lakins JN, Egeblad M, Erlier JT, Fong SF, Csiszar K, Giaccia A, Weninger W, Yamauchi M, Gasser DL, Weaver VM (2009). Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell 139, 891- 906.
Li Y, Shen Y, Cai C, Zhong C, Zhu L, Yuan M, Ren H (2010). The type II Arabidopsis Formin14 interacts with microtubules and microfilaments to regulate cell division. Plant Cell 22, 2710-2726.
Li Z, Hannigan M, Mo Z, Liu B, Lu W, Wu Y, Smrcka AV, Wu G, Li L, Liu M, Huang CK, Wu D (2003). Directional sensing requires Gβɣ-mediated PAK1 and PIXα-dependent activation of Cdc42. Cell 114, 215-227.
Liu W, Sato A, Khadka D, Bharti R, Diaz H, Runnels LW, Habas R (2008). Mechanism of activation of the formin protein DAAM1. Proc Natl Acad Sci USA 105, 210-215.
Lizárraga F, Poincloux R, Romao M, Montagnac G, Le Dez G, Bonne I, Rigaill G, Raposo G, Chavrier P (2009). Diaphanous-related formins are required for invadopodia formation and invasion of breast tumor cells. Cancer Res 69, 2792-2800.
Lo CM, Wang HB, Dembo M, Wang YL (2000). Cell movement is guided by the rigidity of the substrate. Biophys J 79, 144-152.
Loomis PA, Kelly AE, Zheng L, Changyaleket B, Sekerková G, Mugnaini E, Ferreira A, Mullins RD, Bartles JR (2006). Targeted wild-type and jerker espins reveal a novel, WH2-domain-dependent way to make actin bundles in cells. J Cell Sci 119, 1655-1665.
Lu P, Weaver VM, Werb Z (2012). The extracellular matrix: a dynamic niche in cancer progression. J Cell Biol 196, 395-406.
62
Lu Y, Zhang Y, Pan MH, Kim NH, Sun SC, Cui XS (2017). Daam1 regulates fascin for actin assembly in mouse oocyte meiosis. Cell Cycle 16, 1350-1356.
Maas RL, Zeller R, Woychik RP, Vogt TF, Leder P (1990). Disruption of formin-encoding transcripts in two mutant limb deformity alleles. Nature 346, 853-855.
Machesky LM and Insall RH (1998). Scar1 and the related Wiskott-Aldrich syndrome protein, WASP, regulated the actin cytoskeleton through the Arp2/3 complex. Curr Biol 8, 1347-1356.
Makki J (2015). Diversity of Breast Carcinoma: Histological Subtypes and Clinical Relevance. Clin Med Insights Pathol 8, 23-31.
Maldonado MDM and Dharmawardhane S (2018). Targeting Rac and Cdc42 GTPases in cancer. Cancer Res 78, 3101-3111.
Manseau LJ and Schüpbach T (1989). Cappuccino and spire: two unique maternal- effect loci required for both the anteroposterior and dorsoventral patterns of the Drosophila embryo. Genes Dev 3, 1437-1452.
Martizen T, Schachtner H, Legler DF (2015). On the move: endocytic trafficking in cell migration. Cell Mol Life Sci 72, 2119-2134.
Mason FMN, Heimsath EG, Higgs HN, Soderling SH (2011). Bi-modal regulation of a formin by srGAP2. J Biol Chem 286, 6577-6586.
Matsudaira P, Mandelkow E, Renner W, Hersterberg LK, Weber K (1983). Role of fimbrin and villin in determining the interfilament distances of actin bundles. Nature 301, 209-214.
Mattila PK and Lappalainen P (2008). Filopodia: molecular architecture and cellular functions. Nat Rev Mol Cell Biol 9, 446-454.
Mattila PK, Salminen M, Yamashiro T, Lappalainen P (2003). Mouse MIM, a tissue- specific regulator of cytoskeletal dynamics, interacts with ATP-actin monomers through its C-terminal WH2 domain. J Biol Chem 278, 8452-8459.
Matusek T, Djiane A, Jankovics F, Brunner D, Mlodzik M, Mihály J (2006). The Drosophila formin DAAM regulates the tracheal cuticle pattern through organizing the actin cytoskeleton. Development 133, 957-966.
McCarthy JB, El-Ashry D, Turley EA (2018). Hyaluronan, Cancer-Associated Fibroblasts and the Tumor Microenvironment in Malignant Progression. Front Cell Dev Biol 6, 48.
Mersich AT, Miller MR, Chkourko H, Blystone SD (2010). The formin FRL1 (FMNL1) is an essential component of macrophage podosomes. 67, 573-585.
Messing A, Behringer RR, Slapjak JR, Lemke G, Palmiter RD, Brinster RL (1990). Insertional mutation at the ld locus (again!) in a line of transgenic mice. Mouse Genome 87, 107.
63
Michelot A, Guérin C, Huang S, Ingouff M, Richard S, Rodiuc N, Staiger CJ, Blanchoin L (2005). The formin homology 1 domain modulates the actin nucleation and bundling activity of Arabidopsis FORMIN1. Plant Cell 17, 2296-2313.
Miller MR, Miller EW, Blystone SD (2017). Non-canonical activity of the podosomal formin FMNL1ɣ supports immune cell migration. J Cell Sci 130, 1730-1739.
Moseley JB and Goode BL (2005). Differential activities and regulation of Saccharomyces cerevisiae formin proteins Bni1 and Bnr1. J Biol Chem 280, 28023- 28033.
Moseley JB, Sagot I, Manning AL, Xu Y, Eck MJ, Pellman D, Goode BL (2004). A conserved mechanism for Bni1- and mDia1-induced actin assembly and dual regulation of Bni by Bud6 and profilin. Mol Biol Cell 15, 896-907.
Mullins RD, Heuser JA, Pollard TD (1998). The interaction of Arp2/3 complex with actin: nucleation, high affinity pointed end capping, and formation of branching networks of filaments. Proc Natl Acad Sci USA 95, 6181-6186.
Murphy DA and Courtnedige SA (2011). The ‘ins’ and ‘outs’ of podosomes and invadopodia: charateristics, formation, and function. Nat Rev Mol Cell Biol 12, 413-426.
Nader GP, Ezratty EJ, Gundersen GG (2016). FAK, talin and PIPKIɣ regulate endocytosed integrin activation to polarize focal adhesion assembly. Nat Cell Biol 18, 491-503.
Narumiya S, Tanji M, Ishizaki T (2009). Rho signaling, ROCK and mDia1, in transformation, metastasis and invasion. Cancer Metastasis Rev 28, 65-76.
Nielsen TO, Hsu FD, Jensen K, Cheang M, Karaca G, Hu Z, Hernandez-Boussard T, Livasy C, Cowan D, Dressler L, Akslen LA, Ragaz J, Gown AM, Gilks CB, van de Rijn M, Perou CM (2004). Immunohistochemical and clinical characterization of the basal-like subtype of invasive breast carcinoma. Clin Cancer Res 10, 5367-5374.
Nieto MA, Huang RY, Jackson RA, Thiery JP (2016). EMT: 2016. Cell 166, 21-45.
Nourshargh S, Hordijk PL, Sixt M (2010). Breaching multiple barriers: leukocyte motility through venular walls and the interstitium. Nat Rev Mol Cell Biol 11, 366-378.
Okada K, Bartolin F, Deaconescu AM, Moseley JB, Dogic Z, Grigorieff N, Gundersen GG, Goode BL (2010). Adenomatous polyposis coli protein nucleates actin assembly and synergizes with the formin mDia1. J Cell Biol 189, 1087-1096.
Otomo T, Tomchick DR, Otomo C, Pancal SC, Machius M, Rosen MK (2005). Structural basis of actin filament nucleation and processive capping by a formin homology 2 domain. Nature 433, 488-494.
Paget S (1889). The distribution of secondary growths in cancer of the breast. Lancet 1, 99-101.
64
Paul AS and Pollard TD (2009). Review of the mechanism of processive actin filament elongation by formins. Cell Motil Cytoskeleton 66, 606-617.
Péladeau C, Heibein A, Maltez MT, Copeland SJ, Copeland JW (2016). A specific FMNL2 isoform is up-regulated in invasive cells. BMC Cell Biol 17, 32.
Perou CM, Sørlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, Pollack JR, Ross DT, Johnsen H, Akslen LA, Fluge O, Pergamenschikov A, Williams C, Zhu SX, Lønning PE, Børresen-Dale AL, Brown PO, Botstein D (2000). Molecular portraits of human breast tumors. Nature 406, 747-752.
Perrin BJ and Ervasti JM (2010). The actin gene family: function follows isoform. Cytoskeleton 67, 630-634.
Petersen J, Nielsen O, Egel R, Hagan IM (1998). FH3, a domain found in formins, targets the fission yeast formin Fus1 to the projection tip during conjugation. J Cell Biol 141, 1217-1228.
Pfisterer SG, Gateva G, Horvath P, Pirhonen J, Salo VT, Karhinen L, Varjosalo M, Ryhanen SJ, Lappalainen P, Ikonen E (2017). Role for formin-like 1-dependent acto- myosin assembly in lipid droplet dynamics and lipid storage. Nat Commun 8, 14858.
Pollard TD (2017). What we know and do not know about actin. Handb Exp Pharmacol 235, 331-347.
Pollard TD and Borisy GG (2003). Cellular motility driven by assembly and disassembly of actin filaments. Cell 112, 453-465.
Prat A, Parker JS, Karginova O, Fan C, Herschkowitz JI, He X, Perou CM (2010). Phenotypic and molecular characterization of the claudin-low intrinsic subtype of breast cancer. Breast Cancer Res 12, R68.
Pruyne D (2016). Revisiting the Phylogeny of the Animal Formins: Two New Subtypes, Relationships with Multiple Wing Hairs Proteins, and a Lost Human Formin. PLoS One 11, e0164067.
Pruyne D (2017). Probing the origins of metazoan formin diversity: Evidence for evolutionary relationships between metazoan and non-metazoan formin subtypes. PLoS One 12, e0186081.
Pruyne D, Evangelista M, Yang C, Bi E, Zigmond S, Bretscher A, Boone C (2002). Role of formins in actin assembly: nucleation and barbed-end association. Science 297, 612- 615.
Pruyne D, Gao L, Bi E, Bretscher A (2004). Stable and dynamic axes of polarity use distinct formin isoforms in budding yeast. 15, 4971-4989.
Quinlan ME, Heuser JE, Kerkhoff E, Mullins RD (2005). Drosophila Spire is an actin nucleation factor. Nature 433, 382-388.
65
Quinlan ME, Hilgert S, Bedrossian A, Mullins RD, Kerkhoff E (2007). Regulatory interactions between two actin nucleators, Spire and Cappuccino. J Cell Biol 179, 117- 128.
Rabbani SA and Mazar AP (2007). Evaluating distant metastases in breast cancer: from biology to outcomes. Cancer Metastasis Rev 26, 663-674.
Rebowski G, Boczkowska M, Hayes DB, Guo L, Irving TC, Dominguez R (2008). X-ray scattering study of actin polymerization nuclei assembled by tandem W domains. Proc Natl Acad Sci USA 105, 10785-10790.
Recamier JC (1829). Recherches sur la Traitment du Cancer sur la Compression Methodique Simple ou Combinee et sur l’Histoire Generale de la Meme Maladie. 2nd ed.
Ren XL, Qiao YD, Li JY, Li XM, Zhang D, Zhang XJ, Zhu XH, Zhou WJ, Shi J, Wang W, Liao WT, Ding YQ, Liang L (2018). Cortactin recruits FMNL2 to promote actin polymerization and endosome motility in invadopodia formation. Cancer Lett 419, 245- 256.
Ridley AJ (2011). Life at the leading edge. Cell 145, 1012-1022.
Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR (2003). Cell migration: integrating signals from front to back. Science 302, 1704-1709.
Ringnér M, Staaf J, Jönsson G (2013). Nonfamilial breast cancer subtypes. Methods Mol Biol 973, 279-295.
Rivero F, Muramoto T, Meyer AK, Urushihara H, Uyeda TQ, Kitayama C (2005). A comparative sequence analysis reveals a common GBD/FH3-FH1-FH2-DAD architecture in formins from Dictyostelium, fungi, and metazoan. BMC Genomics 6, 28.
Rosado M, Barber CF, Berciu C, Feldman S, Birren SJ, Nicastro D, Goode BL (2014). Critical roles for multiple formins during cardiac myofibril development and repair. Mol Biol Cell 25, 811-827.
Rosales-Nieves AE, Johndrow JE, Keller LC, Magie CR, Pinto-Santini DM, Parkhurst SM (2006). Coordination of microtubule and microfilament dynamics by Drosophila Rho1, Spire and Cappuccino. Nat Cell Biol 8, 367-376.
Rotty JD, Wu C, Bear JE (2013). New insights into the regulation and cellular functions of the Arp2/3 complex. Nat Rev Mol Cell Biol 14, 7-12.
Russnes HG, Lingjærde OC, Børresen-Dale AL, Caldas C (2017). Breast Cancer Molecular Stratification: From Intrinsic Subtypes to Integrative Clusters. Am J Pathol 187, 2152-2162.
Saaristo A, Karpanen T, Alitalo K (2000). Mechanisms of angiogenesis and their use in the inhibition of tumor growth and metastasis. Oncogene 19, 6122-6129.
66
Sagot I, Rodal AA, Moseley J, Goode BL, Pellman D (2002). An actin nucleation mechanism mediated by Bni1 and profilin. Nat Cell Biol 4, 626-631.
Sahoo N, Beatty WL, Heuser JE, Sept D, Sibley LD (2006). Unusual kinetic and structural properties control rapid assembly and turnover of actin in parasite Toxoplasma gondii. Mol Biol Cell 17, 895-906.
Sarmiento C, Wang W, Dovas A, Yamaguchi H, Sidani M, El-Sibai M, Desmarais V, Holman HA, Kitchen S, Backer JM, Alberts A, Condeelis J (2008). WASP family members and formin proteins coordinate regulation of cell protrusions in carcinoma cells. J Cell Biol 180, 124-1260.
Savci-Heijink CD, Halfwerk H, Koster, J, van de Vijver MJ (2016). A novel gene expression signature for bone metastasis in breast carcinomas. Breast Cancer Res Treat 156, 249-259.
Schirenbeck A, Arasada R, Bretschneider T, Stradal TE, Schleicher M, Faix J (2006). The bundling activity of vasodilator-stimulated phosphoprotein is required for filopodium formation. Proc Natl Acad Sci USA, 103, 7694-7699.
Schönichen A, Alexander M, Gasteier JE, Cuesta FE, Fackler OT, Geyer M (2006). Biochemical characterization of the diaphanous autoregulatory interaction in the formin homology protein FHOD1. J Biol Chem 281, 5084-5093.
Schönichen A and Geyer M (2010). Fifteen formins for an actin filament: a molecular view on the regulation of human formins. Biochim Biophys Acta 1803, 152-163.
Schönichen A, Mannherz HG, Behrmann E, Mazur AJ, Kühn S, Silván U, Schoenenberger CA, Fackler OT, Raunser S, Dehmelt L, Geyer M (2013). FHOD1 is a combined actin filament capping and bundling factor that selectively associates with actin arcs and stress fibers. J Cell Sci 126, 1891-1901.
Schoumacher M, Goldman RD, Louvard D, ViGnjevic DM (2010). Actin, microtubules, and vimentin intermediate filaments cooperate for elongation of invadopodia. J Cell Biol 189, 541-556.
Schwarz US and Gardel ML (2012). United we stand: integrating the actin cytoskeleton and cell-matrix adhesions in cellular mechanotransduction. J Cell Sci 125, 3051-3060.
Schwintzer L, Koch N, Ahuja R, Grimm J, Kessels MM, Qualmann B (2011). The functions of the actin nucleator Cobl in cellular morphogenesis critically depend on syndapin I. EMBO J 30, 3147-3159.
Scott BJ, Neidt EM, Kovar DR (2011). The functionally distinct fission yeast formins have specific actin-assembly properties. Mol Biol Cell 22, 3826-3839.
Seppä H, Grotendorst G, Seppä S, Schiffmann E, Martin GR (1982). Platelet-derived growth factor in chemotactic for fibroblasts. J Cell Biol 92, 584-588.
67
Seth A, Otomo C, Rosen MK (2006). Autoinhibition regulates cellular localization and actin assembly activity of the diaphanous-related formins FRLα and mDia1. J Cell Biol 174, 701-713.
Shimada A, Nyitrai M, Vetter IR, Kühlmann D, Bugyi B, Narumiya S, Geeves MA, Wittinghofer A (2004). The core FH2 domain of diaphanous-related formins in an elongated actin binding protein that inhibits polymerization. Mol Cell 13, 511-522.
Siegel R, Miller KD, Jemal A (2018). Cancer statistics, 2018. CA Cancer J Clin 68, 7-30.
Silkworth WT, Kunes KL, Nickel GC, Phillips ML, Quinlan ME, Vizcarra CL (2018). The neuron-specific formin Delphilin nucleates nonmuscle actin but does not enhance elongation. 29, 610-621.
Singer AJ and Clark RA (1999). Cutaneous wound healing. N Engl J Med 341, 738-746.
Skillman KM, Daher W, Ma CI, Soldati-Favre D, Sibley LD (2012). Toxoplasma gondii profilin acts primarily to sequester G-actin while formins efficiently nucleate actin filament formation in vivo. 51, 2486-2495.
Son H and Moon A (2010). Epithelial-mesenchymal transition and cell invasion. Toxicol Res 26, 245-252.
Sørlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Hastie T, Eisen MB, van de Rijn M, Jeffrey SS, Thorsen T, Quist H, Matese JC, Brown PO, Botstein D, Lønning PE, Børresen-Dale AL (2001). Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications.
Sørlie T, Tibshirani R, Parker J, Hastie T, Marron JS, Nobel A, Deng S, Johnsen H, Pesich R, Geisler S, Demeter J, Perou CM, Lønning PE, Brown PO, Børresen-Dale AL, Botstein D (2003). Repeated observation of breast tumor subytpes in independent gene expression data sets. Proc Natl Acad Sci USA 100, 8418-8423.
Sotiriou C, Neo SY, McShane LM, Korn EL, Long PM, Jazaeri A, Martiat P, Fox SB, Harris AL, Liu ET (2003). Breast cancer classification and prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci USA 100, 10393- 10398.
Smid M, Wang Y, Zhang Y, Sieuwerts AM, Yu J, Klijn JG, Foekens JA, Martens JW (2008). Subtypes of breast cancer show preferential site of relapse. Cancer Res 68, 3108-3114.
Srinivasan S, Wang F, Glavas S, Ott A, Hofmann F, Aktories K, Kalman D, Bourne HR (2003). Rac and Cdc42 play distinct roles in regulating PI(3,4,5)P3 and polarity during neutrophil chemotaxis. J Cell Biol 160, 375-385.
Staaf J, Ringnér M, Vallon-Christersson J, Jönsson G, Bendahl PO, Holm K, Arason A, Gunnarsson H, Hegardt C, Agnarsson BA, Luts L, Grabau D, Fernö M, Malmström PO, Johannson OT, Loman N, Barkardottir RB, Borg A (2010). Identifcation of subtypes in human epidermal growth factor receptor 2-positive breast cancer reveals a gene signature prognostic of outcome. J Clin Oncol 28, 1813-1820.
68
Steeg PS (2006). Tumor metastasis: mechanistic insights and clinical challenges. Nat Med 12, 895-904.
Stockinger A, Eger A, Wolf J, Beug H, Fosiner R (2001). E-cadherin regulates cell growth by modulating proliferation-dependent beta-catenin transcriptional activity. J Cell Biol 154, 1185-1196.
Storr SJ, Carragher NO, Frame MC, Parr T, Martin SG (2011). The calpain system and cancer. Nat Rev Cancer 11, 364-374.
Stossel TP, Condeelis J, Cooley L, Hartwig JH, Noegel A, Schleicher M, Shapiro SS (2001). Filamins as integrators of cell mechanics and signalling. Nat Rev Mol Cell Biol 2, 138-145.
Stuelten CH, Parent CA, Montell DJ (2018). Cell motility in cancer invasion and metastasis: insights from simple model organisms. Nat Rev Cancer 296-312.
Sugarbaker EV (1979). Cancer metastasis: a product of tumor-host interactions. Curr Probl Cancer 3, 1-59.
Sun T, Li S, Ren H (2017). OsFH15, a class I formin, interacts with microfilaments and microtubules to regulate grain size via affecting cell expansion in rice. Sci Rep 7, 6538.
Svitkina T (2018). The actin cytoskeleton and actin-based motility. Cold Spring Harb Perspect Biol 10, a018267.
Svitkina TM, Bulanova EA, Chaga OY, Vignjevic DM, Kojima S, Vasiliev JM, Borisy GG (2003). Mechanism of filopodia intiation by reorganization of a dendritic network. J Cell Biol 3, 409-421.
Swan KA, Severson AF, Carter JC, Martin PR, Schanbel H, Schnabel R, Bowerman B (1998). cyk-1: a C. elegans FH gene required for a late step in embryonic cytokinesis. J Cell Sci 111, 2017-2027.
Symons M, Derry JM, Karlak B, Jiang S, Lemahieu V, Mccormick F, Francke U, Abo A (1996). Wiskott-Aldrich syndrome protein, a novel effector for the GTPase CDC42Hs, is implicated in actin polymerization. Cell 84, 723-734.
Swan KA, Severson AF, Carter JC, Martin PR, Schanbel H, Schnabel R, Bowerman B (1998). cyk-1: a C. elegans FH gene required for a late step in embryonic cytokinesis. J Cell Sci 111, 2017-2027.
Takeya R, Taniguchi K, Narumiya S, Sumimoto H (2008). The mammalian formin FHOD1 is activated through phosphorylation by ROCK and mediates thrombin-induced stress fiber formation in endothelial cells. EMBO J 27, 618-628.
Talmadge JE and Fidler IJ (2010). AACR Centennial Series: The Biology of Cancer Metastasis: Historical Perspective. Cancer Res 70, 5649-5669.
Thurston SF, Kulacz WA, Shaikh S, Lee JM, Copeland JW (2012). The ability to induce microtubule acetylation is a general feature of formin proteins. PLoS One 7, e48041.
69
Ti SC, Jurgenson CT, Nolen BJ, Pollard TD (2011). Structural and biochemical characterization of two binding sites for nucleation-promoting factor WASp-VCA on Arp2/3 complex. Proc Natl Acad Sci USA 108, E463-471.
Tojkander S, Gateva G, Lappalainen P (2012). Actin stress fibers – assembly, dynamics and biological roles. J Cell Sci 125, 1855-1864.
Tomasevic N, Jia Z, Russell A, Fujii T, Hartman JJ, Clancy S, Wang M, Beraud C, Wood KW, Sakowicz R (2007). Differential regulation of WASP and N-WASP by Cdc42, Rac1, Nck, and PI(4,5)P2. Biochemistry 46, 3494-3502.
Tominaga T, Sahai E, Chardin P, McCormick F, Courtneidge SA, Alberts AS (2000). Diahphanous-related formins bridge Rho GTPase and Src tyrosine kinase signaling. Mol Cell 5, 13-25.
Trichet L, Sykes C, Plastino J (2008). Relaxing the actin cytoskeleton for adhesion and movement with Ena/VASP. J Cell Biol 181, 19-25.
Trost M, English L, Lemieux S, Courcelles M, Desjardins M, Thibault P (2009). The phagosomal proteome in INFɣ-activated macrophages. Immunity 30, 143-154.
Vaillant DC, Copeland SJ, Davis C, Thurston SF, Abdennur N, Copeland JW (2008). Interaction of the N- and C-terminal autoregulatory domains of FRL2 does not inhibit FRL2 activity.
Valastyan S and Weinberg RA (2011). Tumor metastasis: molecular insights and evolving paradigms. Cell 147, 275-292.
Vallen EA, Caviston J, Bi E (2000). Roles of Hof1p, Bni1p, Bnr1p, and myo1p in cytokinesis in Saccharomyces cerevisiae. Mol Biol Cell 11, 593-611.
Van Damme D, Bouget FY, Van Poucke K, Inzé D, Geelen D (2004). Molecular dissection of plant cytokinesis and phragmoplast structure: a survey of GFP-tagged proteins. Plant J 40, 386-398.
Vig AT, Földi I, Szikora S, Migh E, Gombos R, Tóth MÁ, Huber T, Pintér Rm Talián GC, Mihály J, Bugyi B (2017). The activities of the C-terminal region of the formin protein disheveled-associated activator of morphogenesis (DAAM) in actin dynamics.
Vignjevic D, Yarar D, Welch MD, Peloquin J, Svitkina T, Borisy GG (2003). Formation of filopodia-like bundles in vitro from a dendritic actin network. J Cell Biol 160, 951-962.
Vizcarra CL, Bor B, Quinlan ME (2014). The role of formin tails in actin nucleation, processive elongation, and filament bundling. J Biol Chem 289, 30602-30613.
Wachsstock DH, Schwartz WH, Pollard TD (1993). Affinity of alpha-actinin for actin determines the structure and properties of actin filament gels. Biophys J 65, 205-214.
Wagner AR, Luan Q, Liu SL, Nolen BJ (2013). Dip defines a class of Arp2/3 complex activators that function without preformed actin filaments. Curr Biol 23, 1990-1998.
70
Wagner B, Tharmann R, Haase I, Fischer M, Bausch AR (2006). Cytoskeletal polymer networks: the molecular structure of cross-linkers determines macroscopic properties. Proc Natl Acad Sci USA, 103, 13974-13978.
Wang F, Zhang L, Duan X, Zhang GL, Wang ZB, Wang Q, Xiong B, Sun SC (2015). RhoA-mediated FMNL1 regulates GM130 for actin assembly and phosphorylates MAPK for spindle formation in mouse oocyte meiosis. Cell Cycle 14, 2835-2843.
Wang J, Zhang Y, Wu J, Meng L, Ren H (2013). AtFH16, an Arabidopsis type II formin, binds and bundles both microfilaments and microtubules, and preferentially binds to microtubules. J Integr Plant Biol 55, 1002-1015.
Watanabe N, Madaule P, Reid T, Ishizaki T, Watanabe G, Kakizuka A, Saito Y, Nakao K, Jockusch BM, Narumiya S (1997). p140mDia, a mammalian homolog of Drosophila diaphanous, is a target protein for Rho small GTPase and is a ligand for profilin. EMBO J, 3044-3056.
Watanabe S, Ando Y, Yasuda S, Hosoya H, Watanabe N, Ishizaki T, Narumiya S (2008). mDia2 induces the actin scaffold for the contractile ring and stabilizes its position during cytokinesis in NIH 3T3 cells. Mol Biol Cell 19, 2328-2338.
Wayt J and Bretscher A (2014). Cordon Bleu serves as a platform at the basal region of microvilli, where it regulates microvillar length through its WH2 domains. Mol Biol Cell 25, 2817-2827.
Weigelt B, Mackay A, A’hern R, Natrajan R, Tan DS, Dowsett M, Ashworth A, Reis-Filho JS (2010). Breast cancer molecular profiling with single sample predictors: a retrospective analysis. Lancet Oncol 11, 339-349.
Weilbaecher KN, Guise TA, McCauley LK (2011). Cancer to the bone: a fatal attraction. Nat Rev Cancer 11, 411-425.
Wen KK and Rubenstein PA (2009). Differential regulation of actin polymerization and structure by yeast formin isoforms. J Biol Chem 284, 16776-16783.
Wiseman BS and Werb Z (2002). Stromal effects on mammary gland development and breast cancer. Science 296, 1046-1049.
Witzel I, Oliveira-Ferrer L, Pantel K, Müller V, Wikman H (2016). Breast cancer brain metastases: biology and new clinical perspectives. Breast Cancer Res 18, 8.
Woychik RP, Maas RL, Zeller R, Vogt TF, Leder P (1990). ‘Formins’: proteins deduced from the alternative transcripts of the limb deformity gene. Nature 346, 850-853.
Wozniak MA, Modzelewska K, Kwong L, Keely PJ (2004). Focal adhesion regulation of cell behavior. BIochim Biophys Acta 1692, 103-119.
Wu Q, Li J, Zhu S, Wu J, Chen C, Liu Q, Wei W, Zhang Y, Sun S (2017). Breast cancer subtypes predict the preferential site of distant metastases: a SEER based study. Oncotarget 8, 27990-27996.
71
Wu Y, Shen Z, Wang K, Ha Y, Jia Y, Ding R, Wu D, Gan S, Li R, Luo B, Jiang H, Jie W (2017). High FMNL3 expresion promotes nasopharyngeal carcinoma cell metastasis: role in TGF-β1-induced epithelia-to-mesenchymal transition. Sci Rep 7, 42507.
Xu Y, Moseley JB, Sagot I, Poy F, Pellman D, Goode BJ, Eck MJ (2004). Crystal structures of a Formin-Homology 2 domain reveal a tethered dimer architecture. Cell 116, 711-723.
Xue G and Hemmings BA (2013). PKB/Akt-dependent regulation of cell motility. J Natl Cancer Inst. 105, 393-404.
Xue XH, Guo CQ, Du F, Lu QL, Zhang CM, Ren HY (2011). AtFH8 is involved in root development under effect of low-dose latrunculin B in dividing cells. Mol Plant 4, 264- 278.
Yam PT, Wilson CA, Ji L, Hebert B, Barnhart EL, Dye NA, Wiseman PW, Danuser G, Theriot JA (2007). Actin-myosin network reorganization breaks symmetry at the cell rear to spontaneously initiate polarized cell motility. J Cell Biol 178, 1207-1221.
Yamaguchi H and Condeelis J (2007). Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochim Biophys Acta 1773, 645-652.
Yang C, Czech L, Gerboth S, Kojima S, Scita G, Svitkina T (2007). Novel roles of formin mDia2 in lamellipodia and filopodia formation in motile cells. PLoS Biol 5, e317.
Yang JP, Liao YD, Mai DM, Xie P, Qiag YY, Zheng LS, Wang MY, Mei Y, Meng DF, Xu L, Cao L, Yang Q, Yang XX, Wang WB, Peng LX, Huang BJ, Qian CN (2016). Tumor vasculogenic mimicry predicts poor prognosis in cancer patients: a meta-analysis. Angiogenesis 19, 191-200.
Yates LR, Knappskog S, Wedge D, et al. (2017). Genomic Evolution of Breast Cancer Metastasis and Relapse. Cancer Cell 32, 169-184.
Yayoshi-Yamamoto S, Taniuchi I, Watanabe T (2000). FRL, a novel formin-related protein, binds to Rac and regulates cell motility and survival of macrophages. Mol Cell Biol 20, 6872-6881.
Yersal O and Bartuca S (2014). Biological subtypes of breast cancer: Prognostic and therapeutic implications. World J Clin Oncol 5, 412-424.
Yi K, Guo C, Chen D, Zhao B, Yang B, Ren H (2005). Cloning and functional characterization of a formin-like protein (AtFH8) from Arabidopsis. Plant Physiol 138, 1071-1082.
Yilmaz M and Christofori G (2009). EMT, the cytoskeleton, and cancer cell invasion. Cancer Metastasis Rev 28, 15-33.
Yoo H, Roth-Johnson EA, Bor B, Quinlan ME (2015). Drosophila Cappuccino alleles provide insight into formin mechanism and role in oogenesis. Mol Biol Cell 26, 1875- 1886.
72
Zeller R, Jackson-Grusby L, Leder P (1989). The limb deformity gene is required for apical ectodermal ridge differentiation and anteroposterior limb pattern formation. Genes Dev 3, 1481-1492.
Zhang Z, Zhang Y, Tan H, Wang Y, Li G, Liang W, Yuan Z, Hu J, Ren H, Zhang D (2011). RICE MORPHOLOGY DETERMINANT encodes the type II formin FH5 and regulates rice morphogenesis. Plant Cell 23, 681-700.
Zigmond SH, Evangelista M, Boone C, Yang C, Dar AC, Sicheri F, Forkey J, Pring M (2003). Formin leaky cap allows elongation in the presence of tight capping proteins. Curr Biol 13, 1820-1823.
73
Chapter 2
Expression patterns of human formins and FMNL1 alternative splice isoforms across multiple cell types.
Parts of this work were previously published in:
Krainer EC, Ouderkirk JL, Miller EW, Miller MR, Mersich AT, Blystone SD (2013). The multiplicity of human formins: Expression patterns in cells and tissues. Cytoskeleton 70,
424-438.
This chapter focuses on the work I performed for this project in the context of this dissertation.
74
Introduction
The formin family of actin-associated proteins is highly conserved across eukaryotes, ranging from plants to yeast to mammals (Higgs and Peterson, 2005,
Pruyne, 2016, Pruyne, 2017). Their numerous and diverse actin-modifying abilities are essential for a wide range of cellular functions. This is especially prevalent in mammals, where formins have been shown to not only be important in development, homeostasis, and immunosurveillance, but in cancer and disease as well (DeWard et al., 2010,
Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013). Upregulation of formin expression has been observed in several types of cancer, including breast adenocarcinoma, nasopharyngeal cancer, and colorectal carcinoma (Zhu et al., 2008,
DeWard et al., 2010, Lin and Windhorst, 2016, Wu et al., 2017).
The formin family is extremely diverse, especially in humans where 15 different formins are expressed (Schönichen and Geyer, 2010). Past studies have identified formin expression levels in different cell and tissue types using different analytical techniques (Gardberg et al., 2010a, Gbadegsin et al., 2012, Gardberg et al., 2014, Dutta and Maiti, 2015). This has often led to different views on the variance of formin expression levels in specific cell and tissue types, resulting in some confusion in the literature. This is especially true of the FMNL subfamily, whose expression has classically been thought to be limited to cells of a hematopoietic origin (Favaro et al.,
2003, Katoh and Katoh, 2003, Block et al., 2012, Han et al., 2013, Gardberg et al.,
2014). As a result of past and current studies, we now know this to be inaccurate, as all three members of this subfamily have been shown to be expressed in cell and tissue types as varied as leukocytes, cardiac muscle, and carcinomas (Gardberg et al., 2014,
Rosado et al., 2014, Chen et al., 2018).
The expression levels of nearly all 15 human formins have been previously correlated with some type of cancer cell migration and invasion (DeWard et al., 2010,
75
Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013). This makes sense as formins have been shown to be crucial regulators of two-thirds of the cytoskeleton as a whole: the actin cytoskeleton and microtubule cytoskeleton (Chesarone et al., 2009,
Copeland et al., 2016, Henty-Ridilla et al., 2016, Fernández-Barrera and Alonso, 2018).
It should be noted that FHOD3 has been implicated in intermediate filament interactions, however a current search of the literature shows no other prevalent evidence suggesting formins regulate the intermediate filament cytoskeleton (Kanaya et al., 2005). Highly motile cells, such as invasive cancer cells or leukocytes, require actin- and microtubule- based machinery to regulate and drive locomotion, allowing cancer cells to metastasize to secondary sites in the body and leukocytes to travel to sites of inflammation and infection (Yamaguchi and Condeelis, 2007, Pollard and Cooper, 2009). A major component of this machinery is the podosme in macrophages and invadopodia in cancer cells. These are actin-rich adhesion structures which not only connect the ECM with the cytoskeleton and support migration, but also release MMPs which degrade the ECM, allowing for the cell to invade through the surrounding tissue. While both have similar functions, they do vary in size and protein composition, with invadopodia usually being longer and having a greater half-life (Murphy and Coutneidge, 2012).
Formins are unique in that not only can they generate linear actin filaments, but several have the ability to modify F-actin in other manners (Goode and Eck, 2007,
Schönichen and Geyer, 2010, Breitsprecher and Goode, 2013). FMNL1 is especially versatile, demonstrating the ability to nucleate, assemble, bundle, and sever F-actin
(Harris et al., 2004, Harris et al., 2006). As a result of this versatility, FMNL1 can be especially valuable to highly locomotive cell types. This would also explain the requirement of FMNL1 expression for macrophage adhesion as well, as this ability is essential to the main function of the cell (Mersich et al., 2010, Miller and Blystone, 2015,
Miller et al., 2017).
76
FMNL1 is also unique in that three alternative splice isoforms of this protein have been characterized (Yayoshi-Yamamoto, 2000, Katoh and Katoh, 2003a, Han et al.,
2013). Interestingly, all three diverge in the C-terminal region and we have shown that the ɣ alternative splice isoform is an essential regulator of macrophage migration and podosome formation (Miller et al., 2017). Additionally, this alternative splice isoform was shown to regulate Golgi structure as well (Colón-Franco et al., 2011). Isoforms have been identified in other formins as well, many of which also have specific functions, similar to FMNL1ɣ (Iskratsch et al., 2010, Stastna et al., 2012, Péladeau et al., 2016).
Expression levels, whether at the mRNA or protein level, provide essential data and clues as to the function of different genes. A comprehensive analysis of expression levels can allow for a more complete understanding of the genes in question. Since formin expression levels have been determined using multiple different techniques in past studies, we have performed a complete analysis of mRNA expression levels of this family of proteins in 22 different cell and tissue types using consistent methodologies
(Katoh and Katoh, 2003a-c, Kitzing et al., 2010, Gardberg et al. 2014). Herein, I have highlighted the work I contributed to this publication, along with a new analysis of formin expression patterns, as well as other data demonstrating FMNL1 alternative splice isoform expression in some of the different cell types.
This qRT-PCR (quantitative real-time reverse transcriptase polymerase chain reaction) analysis provides particularly unique data on FMNL1. While we observed high expression in different hematopoietic, kidney, and neuronal cells, it was also found to be high in breast cancer cells. We have previously shown that a specific alternative splice isoform of FMNL1 is imperative for leukocyte adhesion and migration (Mersich et al.,
2010, Miller and Blystone, 2015, Miller et al., 2017). Additionally, FMNL1 expression has been shown to correlate with enhanced invasion in breast cancer cells as well (Han et al., 2009, Kitzing et al., 2010, Gardberg et al., 2014). This led me to hypothesize that
77
FMNL1 may be necessary for proper and efficient cancer cell adhesion and migration.
Furthermore, as adhesion and migration are key components of cancer cell invasion, this could implicate FMNL1 as an important factor in the metastatic cascade.
Materials and Methods
Protein Alignment
The 15 members of the human formin family of actin binding proteins (Table 1,
Fig. 1) were aligned using constraint‐based multiple alignment tool (COBALT). For individual comparisons, basic local alignment search tool for proteins was used with gapping of 10%.
Cells and RNA
The following cells were purchased from ATCC (Manassas, VA): HeLa
(immortalized human cervical cancer cell line), Meg‐01 (human megakaryocytes from chronic myelogenous leukemia), HCN (human cortical neurons). The following cells were gifts: HK‐2 (immortalized proximal tubule epithelial cell line from human kidney) from Thomas R. Welch, M.D.; MDA‐MB‐231 (human breast cancer cell line from epithelial adenocarcinoma) from Christopher E. Turner, Ph.D. Monocytes that were differentiated into macrophages and platelets were isolated from healthy human blood as previously described (Blystone et al., 1997, Jones et al., 1998, Mersich et al., 2010).
RNA Purification
RNA was extracted and purified from 5x106 to 20x106 cells using RNeasy®Mini
Kit from QIAGEN (Valencia, CA), in keeping with clinical laboratory standards for qRT-
PCR as laid out in “Quantification of mRNA using real‐time RT‐PCR” and “The MIQE guidelines: minimum information for publication of quantitative real‐time PCR
78 experiments” (Nolan et al., 2006, Bustin et al., 2009). Cells grown in suspension were pelleted at 1000rpm for 5mins. using a Hermle Z400K centrifuge (Hermle Labortechnik,
Wehingen, Germany) at room temperature (RT). For adherently grown cells, growth media was removed and cells were incubated at 37°C in 0.25% Trypsin‐EDTA (1X),
Phenol Red (GIBCO® Life Technologies, Grand Island, NY) for 5mins. and pelleted at
1000rpm for 5 mins. at RT. Cell pellets were disrupted by adding 600μL QIAGEN buffer
RLT, containing a high concentration of guanidine isothiocycanate to facilitate RNA binding to a silica membrane in subsequent steps, and the resulting lysate was homogenized by passing it through a blunt, 20‐gauge needle five times. Subsequently,
600μL 70% ethanol from Pharmaco‐AAPER (Pharmaco, Brookfield, CT; AAPER,
Shelbyville, KY) was added to the lysate and mixed by pipetting. The sample was then transferred to the RNeasy spin column, which was placed in a collection tube and centrifuged for 15secs. at 14,000rpm using an Eppendorf 5415C centrifuge (Hamburg,
Germany). To remove nonspecifically bound carbohydrates, proteins, fatty acids, and
RNA molecules smaller than 200 bases from the silica membrane, 700μL QIAGEN buffer RW1, containing guanidine salt and ethanol, was added to the column and centrifuged at 14,000rpm for 15secs. Next, 500μL buffer RPE was added to the column and centrifuged at 14,000rpm for 15secs., followed by another 500μL of buffer RPE and centrifugation at 14,000rpm for 2mins. After placing the column in a new collection tube, it was centrifuged for 1min. at 14,000rpm to ensure the elimination of ethanol. RNase‐ free water (30μL) was then added to the column, which was placed in a new collection tube and centrifuged at 14,000rpm for 1min., and repeated with the eluate from this step.
RNA Quantification
RNA was quantified using Agilent RNA 6000 Nano Kit in the Agilent 2100
Bioanalyzer (Santa Clara, CA). RNA Nano 6000 reagents were warmed to RT for
79
30mins. in the dark. Prior to running the experimental chip, the electrodes of the bioanalyzer were decontaminated by adding 350μL RNAseZap and 350μL RNAse‐free deionized (DI) water into the RNAseZap and RNAse‐free DI water electrode cleaner chip, respectively. The RNAseZap electrode cleaner chip was placed into the bioanalyzer, and incubated with the lid closed for 1min., and the chip was removed and replaced with the RNAse‐free DI water electrode cleaner chip, which was incubated for
10secs. with the lid closed. After removing the chip, the bioanalyzer was left open to dry for 10secs. To prepare the Gel Matrix, 550μL of gel was filtered through the filter column and centrifuged at 4000rpm for 10mins., and once filtered, 65μL of gel were put into one tube. Next, the dye concentrate was vortexed for 10secs. and 1μL was added to the
65μL of filtered gel matrix. After vortexing thoroughly, the mixture was centrifuged at
13,000xg for 10mins. A new chip was placed on the chip priming station and 9μL of the gel‐dye mixture were added to the G well, the plunger was pulled up to the 1mL mark, the station closed, and the plunger depressed and incubated for 30secs. A properly sealed syringe gasket was indicated by the plunger's retraction to 0.3mL, and after
5secs., the plunger was pulled back to 1mL and the priming station was opened. Gel‐ dye mixture (9μL) was added to each G well and 5μL of nanomarker buffer was added to each of the 12 sample wells and the ladder well. RNA 6000 ladder (1μL of 150μg/ml) was added to the ladder well and 1μL of sample into each sample well, and the chip was vortexed at 240rpm for 1min. before it was inserted into the bioanalyzer. The assay for total eukaryote RNA was selected in the 2100 Expert software for each analysis, which calculated RNA quality and concentration, expressed as an RNA integrity number (RIN) using an algorithm based on a Bayesian analysis of nine distinct features of the electropherogram of 1208 samples of RNA of varying degrees of degradation using the
Agilent 2100 Bioanalyzer system. The RNA extracted from some cell types was found to have a low RIN, which was due to the low concentration of RNA in the sample, and not
80 due to degraded RNA as can be seen from gels showing integrity of major 28S and 18S
RNA. However, because the algorithm used to compute the RIN incorporates nine features of the electropherogram, the low concentration yielded a low RIN (Schroeder et al., 2006).
Reverse Transcription PCR
For RT‐PCR, 1μg of RNA, determined as described above, was used to prepare
DNA using the SuperScript®III First Strand Synthesis System for RT‐PCR from
Invitrogen (St. Louis, MO) (Ståhlberg et al., 2004). For each reaction, 1μg of RNA,
2.5μL dNTP mix, and 2.5μL oligo(dT)20 were added to a final volume of 25μL diethylpyrocarbonate‐treated water and incubated first at 70°C for 5mins., and then on ice for 5mins. Following these incubations, 22.5μL 10x RT buffer, 20μL MgCl2, 10μL
DTT, and 5μL RNase OUT were added, and the mixture was incubated at 42°C for
2mins. After adding 2.5μL of SuperScript®III, the mixture was incubated at 42°C for
50mins., 70°C for 15mins. to terminate the reaction, and on ice for 5mins. Next, 2.5μL of
RNase H were added to degrade any remnant RNA and incubated at 37°C for 20mins.
Qualitative PCR
For each qualitative PCR reaction, 1μL each of specific forward and reverse primers at 20pM were mixed with 2μL of cDNA. PCR was run using MJ Research PTC‐
100 Programmable Thermal Controller from Bio‐Rad (Hercules, CA). After 1min. at
94°C, 94μL of the following was added: 10μL TAQ DNA polishing buffer B, 6μL MgCl2,
2μL dNTP, 0.8μL of 5,000 units/mL Taq DNA polymerase, all from Fisher Scientific
(Hampton, NH), 5μL dimethyl sulfoxide from Sigma‐Aldrich (St. Louis, MO), and 71.2μL water, followed by a drop of mineral oil from Fisher Scientific. After a total of 2mins. at
81
94°C, followed 35 cycles of 30secs. at 55°C, 1min. at 72°C, and 1min. at 92°C, and one
2min. cycle at 94°C.
Quantitative Real‐Time PCR
Quantitative real‐time PCR was performed using Eco Real‐Time PCR System from Illumina, Inc. (San Diego, CA) and iQ™SYBR®Green Supermix from Bio‐Rad. For each qRT-PCR cell analysis, 500μL of iQ™SYBR®Green Supermix was combined with
50μL of cDNA and 360μL of distilled water and mixed carefully. For each formin analyzed, 1μL of forward specific primer, 1μL of reverse specific primer, each at 1.25pM, the design of which is detailed in this Materials and Methods section, and 18μL of the cDNA mixture were combined. Forward and reverse 18S primer (1μL) was used at a concentration of 10pM. The qRT-PCR reaction was programmed as follows: 2.5mins. at
95°C, 40 cycles of 15secs. at 95°C, 20secs. at 55°C, 30secs. at 72°C, concluded with
15secs. each at 55°C, 60°C, and 95°C. The data were analyzed and graphed using
Microsoft Excel (Redmond, WA).
Primer Design
Custom primer pairs were designed and synthesized for each formin to span the
FH1 and FH2 domain linkage of each formin. The products of each primer pair range from 179bp to 355bp, averaging 307bp. Each set of formin primers, except for FMN1, spans at least one intron, with an average intron size of 9745bp, and ranging from 83bp to 107733bp. The product sizes corresponded to the combined size of the exons, showing that no genomic DNA was present in the PCR reactions. Intron‐exon boundaries in the primer product regions are indicated by an underlined amino acid.
Oligo(dT)20 was used for RT‐PCR, resulting in retrotranscription starting from the C‐ terminal end only. The primers used for qRT-PCR and qualitative PCR are equidistant
82 from the C‐terminus, with the exception of the inverted formin INF1 (FHDC1) and FMN1.
A component of the oligo design stemmed from our search for comparable regions within target templates across all formins, following RNA and DNA alignments, as performed using COBALT.
83
Primers used for qPCR
DAAM1
NP_055807 1078 aa linear PRI 28-JAN-2012
NM_014992 4256 bp mRNA linear PRI 28-JAN-2012
5′- GCC CGA GAA CAA ACT GGA AGG -3′
Bp 1979 - 1999
AA 620 - 627
5′-GGG CAG ATC TTC CTG TTC GTC C -3′
Bp 2276 - 2297
AA 718 - 724 product size 319 bp, spans exons 15-19, spans introns sized 1659, 7358, 4902, 1243
84
DAAM2
NP_001188356 1068 aa linear PRI 09-SEP-2011
NM_015345 6252 bp mRNA linear PRI 28-NOV-2011
5′- GGA GCG TGT CCC TGG CAC CGT ATG G -3′
Bp 2048-2072
AA 616-623
5′- CCA GCA TGT CCT TAG CAA GGT CCT CC -3′
Bp 2318-2343
AA 706-713 product size 296 bp, spans exons 16-18, bridges introns sized 1086, 2577
85
Delphillin
NP_001138590 1211 aa linear PRI 15-AUG-2011
NM_001145118 4632 bp mRNA linear PRI 15-AUG-2011
5′- GGA CAT GAG GTC AGA GGC TAT TGG(sequence = AACTGCTCCCACGA) -3′
Bp 981-990
AA 328-331
5′- CCA GCA CAG GAA CAT CGA CAC CC -3′
Bp 1260-1282
AA 421-427 product size 301 bp
86
Dia 1
NP_005210 1272 aa linear PRI 18-DEC-2011
NM_005219 5804 bp mRNA linear PRI 18-DEC-2011
5′ – GCTTGTGGCTGAGGACCTCTCCC - 3′
Bp 2499 - 2521
AA 786 - 793
5′ – GCTGTTCTGACTGAGTCTATGATC – 3′
Bp 2791 - 2814
AA 884 - 891 product size 316 bp, spans exons 16-20, bridges introns sized 1451, 489, 36994, 4637
87
Dia 2
NP_006720 1101 aa linear PRI 04-MAR-2012
NM_001042517 4812 bp mRNA linear PRI 21-NOV-2011
5′ – GATCAGACCTCATGAAATGACTG – 3′
Bp 2178 - 2200
AA 645 - 652
5′ – CGGTTGGCAGAGTCTATGATTCAG – 3′
Bp 2461 - 2484
AA 742 - 749 product size 307 bp, spans exons 17-20, bridges introns sized 45073, 8525, 4318
88
Dia 3
NP_001035982 1193 aa linear PRI 21-NOV-2011
NM_007309 3782 bp mRNA linear PRI 04-MAR-2012
5′ – GGTCAAAGATTGAACCCACAG – 3′
Bp 2324 – 2344
AA 652 – 657
5′ – GCTGAGTGAGGCTTTAATTCAGAACC – 3′
Bp 2622 – 2647
AA 749 – 756 product size 324 bp, spans exons 16-20, bridges introns sized 6944, 107733, 2124,
24432
89
FHOD1
NP_037373 1164 aa linear PRI 05-FEB-2012
NM_013241 3865 bp mRNA linear PRI 05-FEB-2012
5′ – CGTGACGTGAAGCTGGCTGGGGG – 3′
Bp 2006 - 2028
AA 632 - 638
5′ – CCATGATGCCCACGGAGGAAGAGC – 3′
Bp 2325 - 2348
AA 740 - 746 product size 343 bp, spans exons 13-14, bridges intron sized 1462
90
FHOD3
NP_079411 1439 aa linear PRI 28-JAN-2012
NM_025135 4942 bp mRNA linear PRI 28-JA N-2012
5′- CCG ACG CTG CAG AGA ATT CCT GTG G -3′
Bp 2881 - 2905
AA 929 - 936
5′- CGATCCACCGATGAGGAGAAGC -3′
Bp 3159 - 3185
AA 1022 - 1090 product size 305 bp, spans exons 16-17, bridges intron sized 11930
91
FMN1
NP_001096654 1196 aa linear PRI 30-JAN-2012
NM_001103184 12355 bp mRNA linear PRI 30-JAN-2012
5′ – GCTGAAGAAGGGGGCTACCGC – 3′
Bp 846 – 866
AA 992 - 999
5′ – GGAGAGTGGGAGTGGCCTTCG – 3′
Bp 1004 – 1024
AA 1083 - 1090 product size 179 bp
92
FMN2
NP_064450 1722 aa linear PRI 30-OCT-2011
NM_020066 6440 bp mRNA linear PRI 30-OCT-2011
5′- GCC TCT TTA CTG GAC CAG G -3′
Bp 4107 - 4215
AA 1302 - 1309
5′- CGA GTT CGT CTG ACT GTG C -3′
Bp 4444 - 4462
AA 1393 - 1399 product size 356 bp, spans exons 5-8, bridges introns sized 37845, 15638, 1895
93
FRL1 (FMNL1)
NP_005883 1100 aa linear PRI 19-NOV-2011
NM_005892 3973 bp mRNA linear PRI 19-NOV-2011
5′ – GCACTGAAACCCAGCCAGATCACC – 3′
Bp 2145 - 2168
AA 649 - 656
5′ – GGAAGTCCAGGCCCAGAGCCTGC – 3′
Bp 2433 - 2445
AA 742 - 748 product size 301 bp, spans exons 16-18, bridges introns sized 652, 470
94
FRL3 (FMNL2)
Q8IVF7 FRL3_HUMAN 1028 aa linear PRI 22-FEB-2012
NM_175736 11192 bp mRNA linear PRI 28-JAN-2012
5′- GCA CTG AAA CCC AAC CAG ATC AGT GGC -3′
Bp 1966 – 1992
AA 579 – 587
5′- GCG CAT CAG GCA CTC CAC GAA GTC C -3′
Bp 2259 – 2292
AA 676 – 686 product size 327 bp, spans exons 16-18, bridges introns sized 289, 750
95
FRL2 (FMNL3)
NP_443137 1092 aa linear PRI 28-JAN-2012
NM_052905 5575 bp mRNA linear PRI 28-JAN-2012
5′- GCT CTG AAG CCC AAT CAG ATC AAT GGC -3′
Bp 2263 - 2290
AA 633 - 641
5′- GGT AGG AAC CGC ATC AAG CAT TCC -3′
Bp 2566 - 2589
AA 734 - 740 product size 327 bp, spans exons 16-18, bridges introns sized 962, 1571
96
INF1 (FHDC1)
Q9C0D6 FHDC1_HUMAN 1143 aa linear PRI 22-FEB-2012
NM_033393 6480 bp mRNA linear PRI 18-DEC-2011
5′- GGA CCT TGG CAG CCA GGC AGG -3′
Bp 425 – 441
AA 118 – 122
5′- CGC AAG GTC TCT GAT CCA TAA TGC -3′
Bp 681 – 704
AA 203 – 209 product size 280 bp, spans exons 1-3, bridges introns sized 9943, 656
97
WHIF1 (INF2)
NP_071934 1249 aa linear PRI 26-FEB-2012
NM_022489 4725 bp mRNA linear PRI 26-FEB-2012
5′- GCT GCC ATC CAA CGT GGC ACG TGA GC -3′
Bp 1856 - 1881
AA 572 - 577
5′- CGT GCT TCT CGG GAA GGA GCT TAA GG -3′
Bp 2168 - 2193
AA 676 - 683 product size 338 bp, spans exons 8-11, bridges introns sized 433, 83, 548
98
FMNL1 Alternative Splice Isoform Primer Design
MDA-MB-231
FMNL1Δ FMNL1ɣ/α FMNL1β
Forward Primer: 5'- Forward Primer: 5'- Forward Primer: 5'-
GCACTGAAACCCAGCCAG CCGCCAGATCCTGGA CCGGGCACTGACGGGC
ATCACC-3' GATTGTCCTGG-3' CGGTGCCT-3'
Reverse Primer: 5'- Reverse Primer: 5'- Reverse Primer: 5'-
CCGGCGGCCGCTATGTGA GCATCTCTTCTCCCAG CGAGAGGTCGGAGGTG
TGATGTCTTCAATGG-3' GCTGGCC-3' ACCTGCAGTGGG-3'
Predicted Size: 1143bp Predicted Size: Predicted Size: 1312bp
1029/855bp
Macrophage
FMNL1Δ FMNL1ɣ/α FMNL1β
Forward Primer: 5'- Forward Primer: 5'- Forward Primer: 5'-
GCACTGAAACCCAGCC GCACTGAAACCCAGCCA GGAGAACGAATCCATGG
AGATCACC-3' GATCACC-3' CCAAGATTGC-3'
Reverse Primer: 5'- Reverse Primer: 5'- Reverse Primer: 5'-
CCGGCGGCCGCTATGT CGCAGATCGCGGCCGCT GCGGCCGCTCAGCCTGT
GATGATGTCTTCAATG AGAGGGGCATCTCTTCT CCACGGCCCCACACCTT
G-3' CC-3' TT-3'
Predicted Size: 1143bp Predicted Size: Predicted Size: 1312bp
1200/1026bp
99
Monocyte
FMNL1Δ FMNL1ɣ/α FMNL1β
Forward Primer: 5'- Forward Primer: 5'- Forward Primer: 5'-
GCACTGAAACCCAGCC GCACTGAAACCCAGCCA GGCCCAGGAGTGAAGG
AGATCACC-3' GATCACC-3' CCAAGAAGCC-3'
Reverse Primer: 5'- Reverse Primer: 5'- Reverse Primer: 5'-
CCGGCGGCCGCTATGT CGCAGATCGCGGCCGCT GCGGCCGCTCAGCCTGT
GATGATGTCTTCAATG AGAGGGGCATCTCTTCT CCACGGCCCCACACCTT
G-3' CC-3' TT-3'
Predicted Size: 1143bp Predicted Size: Predicted Size: 1132bp
1200/1026bp
Meg-01
FMNL1Δ FMNL1ɣ/α FMNL1β
Forward Primer: 5'- Forward Primer: 5'- Forward Primer: 5'-
GCACTGAAACCCAGCCA GGCCTGGATGTGCTG CCGCCAGATCCTGGAGAT
GATCACC-3' CTCGAGTACC-3' TGTCCTGG-3'
Reverse Primer: 5'- Reverse Primer: 5'- Reverse Primer: 5'-
CCGGCGGCCGCTATGTG GGGATGTCCGCTTGC CCTGCAGAAGCGGCCGC
ATGATGTCTTCAATGG-3' CGGTGCGGGCCG-3' TACAGCGAGAGGTCGG-3'
Predicted Size: 1143bp Predicted Size: Predicted Size: 887bp
1302/1276bp
100
Data Analysis qRT-PCR analyses were run in triplicate and the results were averaged. Data within one standard deviation were used for the expression analysis. The percent of 18S expression of each formin was calculated as the inverted natural logarithm of the ratio of average formin expression to average 18S expression and multiplied by 100 [1 ‐ LN
(average formin expression / average 18S expression) x 100]. ANOVA and T‐tests were performed on raw data, p<0.05 for specific comparisons. ANOVA was performed on all data points listed in Table 2 with p<0.05.
Results
Using qRT-PCR in conjunction with site-specific primers, seven different human cell types were analyzed: HeLa, macrophages, HCN, MDA-MB-231, HK-2, Meg-01, and platelets. All RNA samples were isolated and purified from whole cell lysates as formin localization, and their expression in different organelles, can greatly vary from cell type to cell type. By following “The MIQE guidelines: minimum information for publication and quantitative real time PCR experiments,” we determined the best primer choice would be
Oligo(dT)20. This standard primer was optimal for our experiments as it hybridizes to the poly(A) tail of mRNA which coincides with our primers targeting more of the C-terminal region of the different formins. Additionally, we chose 18s RNA subunit as the reference for formin expression levels. This subunit is widely used as a marker and generally considered as a proper standard. Therefore, all mRNA expression levels are presented as a percentage of total 18s RNA expression. Primer design was conducted following an alignment of all 15 human formin proteins and subsequent sequence analysis.
Interestingly, we determined that the region between the FH1 and FH2 domains would be the best option for primer targeting design since this region’s sequence composition is quite unique between formins while at the same time retaining a similar length.
101
While the entirety of this study focused on 22 different cell and tissue types, my personal work focused on the seven listed previously. All protein and gene accession numbers for each of the 15 human formins, as well as references for each along with supporting information, can be observed in Table 1. The site-specific oligonucleotides designed and used for these experiments were the result of formin protein sequence alignement analysis (Figure 1). Using qualitative PCR, all of our primers for the 15 human formins exhibited products at the correct predicted sizes (Figure 2). However, there was one exception with Delphilin, as no product was observed using its uniquely designed primer pairs. This protein has been shown to be expressed in the Purkinje cells of the cerebellar cortex, but even the use of human cerebellar RNA-derived cDNA did not allow for detection on an agarose gel.
102
Table 1
103
Table 1 (Continued)
104
Table 1: The formin family of proteins. The seven formin families with all their members are listed, as well as the alternate nomenclature for each formin. The 15 mammalian formins analyzed in this study are grouped into families based on FH2 phylogeny with gene names and accession numbers provided. The amino acid and genomic accession sequences used for alignment and oligonucleotide design are listed.
105
Figure 1
106
Figure 1: Alignment of formin protein sequences. Formin protein sequences from Table 1 were analyzed using COBALT. Amino acid residues highlighted in light grey are translations of forward primers and amino acid residues highlighted in black are translations of reverse primers used for RT-PCR and qRT-PCR as described in
Methods. The vertical line indicates the start of the FH2 domain. The C-terminal end of the FH1 domain is not depicted in this alignment, and is located upstream of the amino acids indicating the forward primers used for qRT-PCR. The underlined amino acids indicate the location of intron-exon borders between forward and reverse primers. For information regarding intron and extron length and position, see Materials and Methods.
107
Figure 2
Figure 2: Examples of qualitative PCR products for each formin. PCR performed on DNA from RT-PCR to demonstrate that the primer pairs for each formin, as listed in
Materials and Methods, produce unique and specific products of predicted sizes.
Product sizes are above each band in number of base pairs. RT-PCR products of formin primers were analyzed using electrophoresis on a 2% agarose gel and visualized using ethidium bromide under ultraviolet light. The gel was photographed using Polaroid
GelCam (Minnetonka, MN) and Fujifilm FP-3000B (Tokyo, Japan) instant black and white film. The resulting image was scanned using an HP ScanJet5300C (Palo Alto,
CA) and processed using the Microsoft Scanning Wizard and Microsoft Powerpoint
(Redmond, WA). A single example from U2OS cells of multiple trials in several cell types is shown as a representative example.
108
HeLa Cells
HeLa cells are adenocarcinoma cells originating from epithelial tissue of the cervix and have been classically used for research since the 1950s (Masters, 2002). qRT-PCR analysis demonstrated an average total formin expression level of 35.8%, with
Dia1, FHOD1, and FHDC1 all exhibiting the highest expression levels between 50-52%
(Figure 3B). Interestingly, Dia1 and Dia2 showed expression levels between 40-50%, but the other member of this subfamily, Dia3, only demonstrated expression levels of
27%. A similar pattern was observed for FHOD1 and FHOD3, where their expression levels, 51% and 13%, respectively, showed a wide discrepancy. This is interesting as
Dia3 shares a ~40-50% sequence identity with both Dia1 and Dia2, while FHOD3 is
~40% identical to FHOD1 and ~20% to Dia1 (Schönichen and Geyer, 2010). However, this makes some sense as both Dia1 and Dia2 have been shown to be essential for cancer cell migration and invasion while Dia3 has strongly been attributed to regulating microtubules and kinetochores (Kitzing et al., 2007, Guo et al., 2011, Pettee et al.,
2014). Additionally, FHOD1 is upregulated when cancer cells undergo EMT and subsequently migrate and invade, but FHOD3 is primarily attributed to sarcomere organization in cardiomyocytes (Kan-o et al., 2012, Gardberg et al., 2013).
FMNL1 average expression in HeLa cells was the second highest, tied with platelets, when evaluating all seven cell types (Figure 10). FMNL1, more specifically the
FMNL1ɣ alternative splice isoform, has been shown to regulate Golgi structural architecture in HeLa cells (Colón-Franco et al., 2011). Furthermore, it has been shown to form a complex with srGAP2 in HeLa cells, where activated FMNL1 and srGAP2 co- localize to the cell membrane where it mediates membrane ruffling (Mason et al., 2010).
Expression of FMNL1 could be related to these regulatory functions, which are both essential to cell migration, as well as other cellular processes.
109
Figure 3
A
B
110
Figure 3: HeLa Cell RNA and Formin Expression. (A) HeLa cell RNA was obtained as described in Materials and Methods. The “Electropherogram Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and 28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the “Electropherogram
Summar” is a visual representation of RNA degradation. The RNA Integrity Number
(RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of
Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta
Clara, CA) (Schroeder et al., 2006). (B) qRT-PCR was performed as described in
Materials and Methods. The level of expression of all formins in HeLa cells was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit.
Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
111
Macrophages
Past studies from our lab have shown that upon differentiation from monocytes,
FMNL1 mRNA expression in macrophages is increased six-fold. This was confirmed at the protein level as well, where FMNL1 expression increased 108% when comparing monocytes to macrophages (Mersich et al., 2010). However, when comparing the expression levels of all formins, FMNL1 is not the highest, even in its own subfamily, where FMNL3 is expressed ~10% higher (Figure 4B). The two most highly expressed formins are from two different subfamilies: Dia1 and FHOD1. While the FMNL formins have been primarily attributed to leukocytes, this observation demonstrates that other formins may have substantial roles in different macrophage cellular functions (Favaro et al., 2003, Katoh and Katoh, 2003a, Block et al., 2012, Han et al., 2013, Gardberg et al.,
2014). These results make sense as well since Dia1 has been shown to interact with
IQGAP1 for proper phagocytic cup formation in macrophages and FHOD1 has exhibited macrophage podosome regulatory functions (Brandt et al., 2007, Panzer et al., 2016).
Dia2, DAAM2, and FMN2 expression was under 10% for each formin. These low levels are to be expected for DAAM2 and FMN2 as they are more important for developmental activities as opposed to immunosurveillance (Azoury et al., 2008, Lee and Deneen,
2012). Overall, macrophages have the lowest level of average formin expression, however, their limited, albeit essential, role in immunosurveillance could explain these observations.
Interestingly, the average FMNL1 expression for macrophages is one of the lowest of the seven described here at 26% (Figure 10). While FMNL1 is important for regulating essential macrophage functions, the mRNA expression level may not directly correlate with protein expression and function. Furthermore, FMNL1 function may not require high amounts of FMNL1 protein, which could also explain the low mRNA levels as well.
112
Figure 4
A
B
113
Figure 4: Macrophage RNA and Formin Expression. (A) Macrophage RNA was obtained as described in Materials and Methods. The “Electropherogram Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and
28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the “Electropherogram
Summar” is a visual representation of RNA degradation. The RNA Integrity Number
(RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of
Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta
Clara, CA) (Schroeder et al., 2006). (B) qRT-PCR was performed as described in
Materials and Methods. The level of expression of all formins in macrophages was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit.
Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
114
HCN Cells
HCN cells are a human cortical neuronal cell line and they display formin expression levels between 17-50%, with FHOD1 being the most highly expressed
(Figure 5B). Quite surprisingly, Delphilin had the lowest expression level, which is interesting as it has been associated with GluRδ2 signaling in Purkinje cells, although this has been shown to be isoform-specific (Miyagi et al., 2002, Matsuda et al., 2006).
The high FHOD1 expression was followed by members of the INF subfamily of formins,
INF1 (FHDC1) and INF2 (WHIF1). The 50% expression level of FHOD1 in the brain is interesting as a recent study demonstrated no FHOD1 expression in neurons or glial cells using immunofluorescence (Gardberg et al., 2013). In the Dia subfamily, Dia1 was expressed the highest at 38%. Past studies have indicated that Dia1 regulates synaptic vesicle endocytosis in hippocampal neurons (Soykan et al., 2017). Dia1 expression has also been correlated with axon elongation in cerebellar granule neurons (Arakawa et al.,
2003). The FMN subfamily was expressed at an average of 29% with FMN2 being expressed 4% higher than FMN1. Nonetheless, FMN1 is essential for dendritogenesis and synaptogenesis in mouse hippocampal neurons. Overexpression of FMN1 in these cells increases primary dendrite and glutamatergic synaptic input number, while downregulation results in morphological and synaptic alterations (Simon-Areces et al.,
2011).
The FMNL subfamily was expressed at an average of 31.3% with FMNL2 expression being the highest at 35%. FMNL1 was the lowest expressed of this subfamily at 26% (Figure 5B). However, a past study on FMNL1 protein expression in neurons found no expression via immunohistochemistry (Gardberg et al., 2014). It should also be noted that, while this study did focus on mammalian formins, the singular
FMNL subfamily D. melanogaster homologue has been identified as a regulator of axon growth in the mushroom bodies of flies (Dollar et al., 2016).
115
Figure 5
A
B
116
Figure 5: HCN Cell RNA and Formin Expression. (A) HCN RNA was obtained as described in Materials and Methods. The “Electropherogram Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and 28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the “Electropherogram Summar” is a visual representation of RNA degradation. The RNA Integrity Number (RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta Clara, CA) (Schroeder et al., 2006).
(B) qRT-PCR was performed as described in Materials and Methods. The level of expression of all formins in HCN cells was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit. Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
117
MDA-MB-231
These breast adenocarcinoma cells exhibited an overall average formin expression level of 34%. The Dia and FMNL subfamilies both had an average 39% expression level. This is to be expected in such a highly motile and invasive cell line, as both these subfamilies of proteins have been implicated in a tremendous amount of cellular locomotion functions and cancer invasion (Han et al., 2009, Lizárraga et al.,
2009, Kitzing et al., 2010, Vega et al., 2011, Kim et al., 2016). Furthermore, the average expression level of both the FHOD and INF subfamilies is strikingly similar, at 43.5% and
37.5%, respectively (Figure 6B). DAAM1 is also expressed at a comparable level of
32%. As all these proteins have been implicated in some aspect of cancer cell migration and invasion, a question arises as to why DAAM2 and FMN2 are expressed at the lowest levels. This is a curious observation as several formins are involved not only in cancer-related cellular activities, but homeostatic and developmental functions as well.
This is most likely explained by their limited functions in development, although DAAM2 was recently identified as regulator of gliomagenesis (Lee and Deneen, 2012,
Sahasrabudhe et al., 2016, Zhu et al., 2017).
While FMNL2 expression has been associated with regulation of migration and invasion in breast cancer cells, and it is the highest expressed of the three FMNL subfamily members, high FMNL1 expression has also been observed in breast cancer
(Kitzing et al., 2010, Gardberg et al., 2014). Immunohistochemical analysis of one of the rarest but most aggressive forms of breast cancer (the basal subtype, of which MDA-
MB-231 cells belong to) has shown high FMNL1 protein expression in epithelial and inflammatory cells in human breast cancer tumors (Gardberg et al., 2014). Also in this same study, FMNL1 mRNA expression was not very high when compared to other
FMNL family members, demonstrating that while mRNA expression may be low, this may not correspond to protein function (Gardberg et al., 2014).
118
Figure 6
A
B
119
Figure 6: MDA-MB-231 Cell RNA and Formin Expression. (A) MDA-MB-231 RNA was obtained as described in Materials and Methods. The “Electropherogram
Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and 28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the
“Electropherogram Summar” is a visual representation of RNA degradation. The RNA
Integrity Number (RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta
Clara, CA) (Schroeder et al., 2006). (B) qRT-PCR was performed as described in
Materials and Methods. The level of expression of all formins in MDA-MB-231 cells was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit.
Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
120
HK-2 Cells
HK-2 cells are a proximal tubule epithelial cell line which was isolated and immortalized from a healthy adult human kidney (Ryan et al., 1994). This cell line had the highest overall expression from the seven different types studied here at 51%.
Unsurprisingly, the previously mentioned study identifying zero FHOD1 expression in the brain (while we found mRNA expression to be highest there) found similar results for the podocytes, mesangial cells, and tubular epithelium of the kidney (Gardberg et al., 2013).
There are several possible explanations for these data, including the reasoning that high mRNA expression does not always correlate to high protein expression. Additionally, the immunofluorescence analysis could be supported with subsequent Western blots to confirm the observation that this protein is indeed not expressed (Gardberg et al., 2013).
DAAM2 is also expressed above the 60% level, which could be explained by its requirement for proper Wnt signaling during kidney development (Figure 7B)
(Goggolidou et al., 2014). The high overall expression could be attributed to several factors, including the pattern of predominately high mRNA expression throughout several types of epithelial tissue and their role in different renal pathologies (Chhabra and Higgs, 2006, Dettenhofer et al., 2008, Sun et al., 2011, Krainer et al., 2013,
Subramanian et al., 2016).
The FMNL subfamily had an average overall expression of 46%, which, when compared to other subfamily members, is lower. However, compared to other cell and tissue types, this is quite high. This is interesting as a past qRT-PCR analysis on
FMNL1 expression demonstrated low levels of expression of all three FMNL1 alternative splice isoforms in the kidney compared to other cell and tissue types (Han et al., 2009).
However, FMNL1 protein expression of all three FMNL1 alternative splice isoforms was observed in this cell type following transfection, with FMNL1ɣ primarily localizing to the cell membrane and membrane blebs (Han et al., 2009).
121
Figure 7
A
B
122
Figure 7: HK-2 Cell RNA and Formin Expression. (A) HK-2 RNA was obtained as described in Materials and Methods. The “Electropherogram Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and 28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the “Electropherogram Summar” is a visual representation of RNA degradation. The RNA Integrity Number (RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta Clara, CA) (Schroeder et al., 2006).
(B) qRT-PCR was performed as described in Materials and Methods. The level of expression of all formins in HK-2 cells was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit. Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
123
Meg-01 Cells and Platelets
Meg-01 cells, a megakaryocyte cell line isolated from human bone marrow, and platelets, which we isolated from peripheral human blood, both exhibited some similar trends in their formin expression levels, albeit with some minor discrepancies (Figure 8B and Figure 9B). Furthermore, as platelets are products of megakaryocytes, both cell types will be discussed in this single section.
Dia1 was the most highly expressed formin in Meg-01 cells (Figure 8B). While this formin was also expressed at high levels in platelets, FHOD1 was expressed at a slightly elevated level (51% compared to 54%). This is a curious result as FHOD3 was expressed at extremely low levels in both Meg-01 cells and platelets (6% and 17%, respectively) and the FHOD subfamily share a 43.4% sequence identity (Schönichen and Geyer, 2010). However, high FHOD1 expression in both megakaryocytes and platelets has previously been observed at the protein level and has been suggested to play a role in regulating actin stress fibers in these cells (Thomas et al., 2011). The
FMNL subfamily, which has often been attributed primarily to cells of hematopoietic origin, exhibited similar average levels of expression when comparing Meg-01 cells to platelets, at 34% and 36%, respectively. DAAM1 and Dia1 have both been previously identified as regulators of actin assembly in platelets, which could explain their high expression levels in these cells, although Dia1 knockout mice do not display altered platelet function (Eisenmann et al., 2007, Higashi et al., 2008). The overall average formin expression level of these cells were slightly higher than macrophages and, similar to those cells, could be attributed to their limited function in the immune response.
The average FMNL subfamily expression in Meg-01 cells and platelets was
34.3% and 36%, respectively. FMNL2 expression was the highest in both Meg-01 cells and platelets, at 47% and 39%, respectively (Figure 8B and Figure 9B). This correlates with a past study demonstrating very high FMNL2 protein expression in megakaryocytes
124
(Gardberg et al., 2010b). FMNL1 itself was not the most highly expressed of the 3
FMNL subfamily members, with Meg-01 cells expressing FMNL1 at 29% and platelets at
38% (Figure 8B and Figure 9B). However, past studies have shown FMNL1 to be highly expressed at the mRNA level in bone marrow myeloid cells (Gardberg et al., 2014).
Additionally, other mRNA studies have demonstrated high levels of FMNL1 in platelets
(Zuidscherwoude et al., 2018).
125
Figure 8
A
B
126
Figure 8: Meg-01 Cell RNA and Formin Expression. (A) Meg-01 RNA was obtained as described in Materials and Methods. The “Electropherogram Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and 28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the “Electropherogram
Summar” is a visual representation of RNA degradation. The RNA Integrity Number
(RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of
Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta
Clara, CA) (Schroeder et al., 2006). (B) qRT-PCR was performed as described in
Materials and Methods. The level of expression of all formins in Meg-01 cells was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit.
Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
127
Figure 9
A
B
128
Figure 9: Platelet RNA and Formin Expression. (A) Platelet RNA was obtained as described in Materials and Methods. The “Electropherogram Summary” visualizes the analysis of total RNA, where peaks between and underneath the 18s and 28s ribosomal subunits indicate degraded RNA, and higher peaks correspond to a higher concentration of RNA. The rectangular box to the right of the “Electropherogram Summar” is a visual representation of RNA degradation. The RNA Integrity Number (RIN) is calculated using an algorithm based on Bayesian analysis of nine distinct feature of the electropherogram of 1208 samples of RNA of varying degrees of Degradation using the Agilent 2100 bioanalyzer system (Agilent Technologies, Stanta Clara, CA) (Schroeder et al., 2006).
(B) qRT-PCR was performed as described in Materials and Methods. The level of expression of all formins in platelets was averaged and is displayed as a percentage of expression of the 18s ribosomal subunit. Data are reported as the average percentage of 18s ribosomal subunit, p < 0.05.
129
FMNL1 Expression
The highest level of FMNL1 mRNA expression was observed in kidney cells at a level of 45%, followed by platelets and cervical cancer cells both at 38%, breast cancer cells at 32%, megakaryocytes at 29%, and macrophages and neurons both at 26%.
FMNL1 protein expression has been observed at higher levels in the kidney before, primarily in the arteriolar smooth muscle cells located there (Schuster et al., 2007,
Gardberg et al., 2014). The two cancer cell lines, HeLa and MDA-MB-231, have similar
FMNL1 expression levels, and both have been shown to have high protein expression levels as well (Arjonen et al., 2011, Colón-Franco et al., 2011, Gardberg et al., 2014).
The lower macrophage expression level is interesting as we know this FMNL1 is an essential component for the adhesive and migratory abilities of these cells. However, this is the same level as HCN cells, and FMNL1 has been shown to regulate srGAP2 in cortical neuron migration, demonstrating that this low mRNA expression level may not directly correlate with protein function (Guerrier et al., 2009, Mason et al., 2011).
130
Figure 10
Figure 10: Average FMNL1 expression varies across cell types. The average
FMNL1 expression for each cell type is displayed. qRT-PCR was performed as described in Materials and Methods. Data are reported as the average percentage of the 18s ribosomal subunit, p < 0.05
FMNL1 Alternative Splice Isoform Expression
FMNL1 expresses three different alternative splice isoforms, all of which diverge in the C-terminal region at the T1069 residue: FMNL1α, FMNL1β, and FMNL1ɣ (Kato and Kato, 2003, Han et al., 2009). We have previously shown that these alternative splice isoforms are unique and do not all function similarly, both in vitro and in vivo
(Mersich et al., 2010, Miller et al., 2017). Past studies by other groups have confirmed this as well, demonstrating that a specific alternative splice isoform of FMNL regulates
131
Golgi complex architecture and that N-terminal myristoylation of another alternative splice isoform is required for membrane blebbing (Han et al., 2009, Colon-Franco et al.,
2011). This led us to question if the mRNA expression levels of the different FMNL1 alternative splice isoforms vary between cell types, as this protein has been shown to have different functions depending on the cell type.
Primer design for this analysis proved challenging as we often were unable to confirm FMNL1 alternative splice isoform expression with the same set of primers for each cell type. By altering both forward and reverse primer design, we were eventually able to generate primers which demonstrated expression of all three FMNL1 alternative splice isoforms in MDA-MB-231 cells, 1° human monocytes, 1° human macrophages, and Meg-01 cells using RT-PCR. As a result of using combinations of different forward and reverse primers, products sizes varied depending on both isoform and primer combination (Figure 11). All primer sequences and predicted product sizes are listed in
Materials and Methods. FMNL1 alternative splice isoform mRNA expression was exhibited in all four cell types at the correct predicted sizes (Figure 11). Additionally, for each set of RT-PCR reactions, a control set was also performed which is labeled
“FMNL1Δ.” The FMNL1Δ reaction correlates to a PCR reaction using the same forward and reverse primers for each cell type, resulting in the same size product. The amplified product exhibited here is an 1143bp product which begins N-terminally of the FH1 domain and continues until the splice site T1069. This internal region of FMNL1 expression was confirmed in each of these four cell types.
FMNL1 Alternative Splice Isoform qRT-PCR Analysis
After confirming the expression of all three FMNL1 alternative splice isoforms in these four cell types, we attempted to quantitate their mRNA expression before analyzing any other cell types as a result of the difficulties in primer design. We
132 speculated that as a result of the difficulties encountered during RT-PCR analysis, similar issues may arise for the more sensitive qRT-PCR experiments. Indeed, regardless of using the same qRT-PCR protocol as we did for analyzing formin expression, we were unable to quantify mRNA expression levels in these cell types.
Following subsequent rounds of experiments and a time-consuming troubleshooting process, we decided to forego continuing quantitation of FMNL1 alternative splice isoform mRNA in these cell types.
133
Figure 11
Figure 11: The three FMNL1 alternative splice isoforms are expressed at the mRNA level in MDA-MB-231 cells, macrophages, monocytes, and Meg-01 cells.
Qualitative RT-PCR images are displayed for four different cell types. Product sizes vary due to the alterations in primer design. FMNL1Δ is not a control RT-PCR reaction encoding for an internal region of FMNL1. Bands are separated via agarose gel electrophoresis and visualized using UV light in conjunction with ethidium bromide staining.
134
Discussion
Herein, we have described the mRNA expression patterns of the 15 human formin family members in seven different cell types. While this work is a component of a larger analysis published in 2013, it nonetheless provides useful information on formin expression and, more specifically, expression of FMNL1, which has proven beneficial for the formin field of research (Krainer et al., 2013). As a result, this discussion will highlight only the work presented here and how it applies to FMNL1 expression in breast adenocarcinoma.
MDA-MB-231 formin expression averages at 34.13% while the FMNL subfamily as a whole exhibits an average expression level of 38.7%. FMNL1 itself is expressed on average at a level of 32.0% in MDA-MB-231 cells. Interestingly, this is one of the lower formins expressed in this cell type, the FMNL subfamily average of 38.7% is quite similar to the Dia subfamily (39.3%) and FHOD subfamily (38.5%). Specifically in this cell type, both of these other subfamilies have been shown to play a role in cell migration and invasion. For example, Dia1 regulates intracellular trafficking of MT1-MMP, Dia2 interacts with RhoA downstream of CXCL12 to promote amoeboid phenotype switching, and Dia3 is required for invadopodia formation (Lizárraga et al., 2009, Kim et al., 2016,
Wyse et al., 2017). Additionally, FHOD1 regulates cell morphology and migration while
FHOD3 acts downstream of RCP-α5β1 integrin trafficking to regulate invasion (Jurmeister et al., 2012, Paul et al., 2015a).
Both of these subfamilies, which are expressed at an average level similar to that of the FMNL subfamily, are required for proper cell migration in this specific cell type.
Moreover, FMNL2 and FMNL3 are both expressed at the protein level in MDA-MB-231 cells (Gardberg et al., 2014). FMNL2 actually plays a pivotal role in β1 integrin endocytosis in these cells, a key component of cell migration (Paul et al., 2015b, Wang et al., 2015). Cancer invasion in other cell types is also regulated by these proteins and
135 correlates with their expression levels, such as FMNL2 in colorectal and breast cancer,
FMNL3 in prostate and nasopharyngeal cancer, and both FMNL2 and FMNL3 in melanoma (Kitzing et al., 2010, Li et al., 2010, Vega et al., 2011, Gardberg et al., 2016,
Wu et al., 2017).
The cancer regulatory properties of these proteins are further compounded when one takes into account FMNL1 expression. Enhanced FMNL1 mRNA and protein expression levels have previously been observed in MDA-MB-231 cells. An in silico analysis of the FMNL1 transcriptome showed increased levels in MDA-MB-231 cells which was further confirmed via Western blot and immunofluorescence analysis (Han et al., 2013). Immunohistochemical analyses of eight different breast tumors confirmed
FMNL1 expression in malignant epithelial cells, as well as lymphocytes and macrophages found within the tumor (Gardberg et al., 2014). Interestingly, the FMNL1ɣ alternative splice isoform was specifically found to localize to the membrane and induce blebbing in MDA-MB-231 (Han et al., 2009). Furthermore, FMNL1 is upregulated in nasopharyngeal carcinomas, where it contributes to invasion via epigenetic upregulation of MTA1, and in leukemia, where it has been shown to regulate both proliferation and migration in two different leukemia cell lines via Rac1 interactions (Favaro et al., 2013,
Chen et al., 2018).
While MDA-MB-231 cells and macrophages are of two separate origins and have different functions, they do share a commonality in that they are both highly motile cell types. Macrophages are essential for immunosurveillance and are found within nearly every tissue type in the body. Their locomotion is essential to immune system functionality and their modes of migration and invasion share many features with that of
MDA-MB-231 cells. Both of these cell types use specialized structures to invade through the ECM, effectively altering its composition through the use of MMPs. For example, MMP-1 (collagenase-1), MMP-3 (stromelysin), and MMP-9 (gelatinase B) are
136 used by macrophages for host defense and normal tissue remodeling and repair, while
MDA-MB-231 cells use these proteases to degrade the ECM, allowing for efficient invasion through tissue and eventual intravasation into the circulatory system or lymphatics (Cury et al., 1988, Welgus et al., 1990, Campbell et al., 1991, Lindenmeyer et al., 1997, Phromoni et al., 2009, Steenport et al., 2009, Liu et al., 2012). Additionally, both cell types express MT1-MMP (membrane-type-1 matrix metalloprotease) and MMP-
2 (gelatinase A), which are regulated by the formin Dia1 in MDA-MB-231 cells (Hayashi et al., 2000, Gonzalo and Arroyo, 2010, Kim et al., 2016, Kim and Rhee, 2016).
The results of this mRNA expression analysis correlate well with our previous work demonstrating that FMNL1 is required for proper macrophage adhesion and migration (Mersich et al., 2010, Miller and Blystone, 2015). In conjunction with the previously mentioned studies, this insinuates that FMNL1 could very well be essential to
MDA-MB-231 cell migration and subsequent invasion. Additionally, the expression of all three FMNL1 alternative splice isoforms is observed in MDA-MB-231 cells, monocytes, macrophages, and Meg-01 cells. While we were unable to obtain quantitative results, these qualitative results demonstrate that expression of all three alternative splice isoforms is observed in cell types of different embryonic origin. Furthermore, MDA-MB-
231 cells and macrophages exhibit similar patterns in regards to formin protein expression and function in migration and invasion. The data presented here demonstrate that formin expression and function could be linked regardless of cell type and that FMNL1 could very well be imperative to MDA-MB-231 cell migration, regardless of classical theories that this protein is predominately important for leukocytes and other cells of hematopoietic origin.
137
References
Arakawa Y, Bito H, Furuyashiki T, Tsuji T, Takemoto-Kimura S, Kimura K, Nozaki K, Hashimoto N, Narumiya S (2003). Control of axon elongation via an SDF-1α/Rho/mDia pathway in cultured cerebellar granule neurons. J Cell Biol 161, 381-391.
Arjonen A, Kaukonen R, Ivaska J (2011). Filopodia and adhesion in cancer cell motility. Cell Adh Migr 5, 421-430.
Azoury J, Lee KW, Georget V, Rassinier P, Leader B, Verlhac MH (2008). Spindle positiong in mouse oocytes relies on a dynamic meshwork of actin filaments. Curr Biol 18, 1514-1519.
Block J, Breitsprecher D, Kühn S, Winterhoff M, Kage F, Geffers R, Duwe P, Rohn JL, Baum B, Brakebusch C, Geyer M, Stradal TE, Faix J, Rottner K (2012). FMNL2 drives actin-based protrusion and migration downstream of Cdc42. Curr Biol 22, 1005-1012.
Blystone SD, Williams MP, Slater SE, Brown EJ (1997). Requirement of integrin β3 tyrosine 747 for β3 tyrosine phosphorylation and regulation of αvβ3 avidity. J Biol Chem 272, 28757-28762.
Brandt DT, Marion S, Griffiths G, Watanabe T, Kaibuchi K, Grosse R (2007). Dia1 and IQGAP1 interact in cell migration and phagocytic cup formation. J Cell Biol 178, 193- 200.
Breitsprecher D and Goode BL (2013). Formins at a glance. J Cell Sci 126, 1-7.
Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wittwer CT (2009). The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55, 611-622.
Campbell EJ, Cury JD, Shaprio SD, Goldberg GI, Welgus HG (1991). Neutral proteinases of human mononuclear phagocytes. Cellular differentiation markedly alters cell phenotype for serine proteinases, metalloproteinases, and tissue inhibitor of metalloproteinases. J Immunol 146, 1286-1293.
Castrillon and Wasserman (1994). Diaphanous is required for cytokinesis in Drosophila and shares domains of similarity with the products of the limb deformity gene. Development 120, 3367-3377.
Chae SC, Inazu Y, Amagi A, Maeda Y (1998). Underexpression of a novel gene, DIA2, impairs the trainsition of Dictyostelium cells from growth to differentiation. Biochem Biophys Res Commun 252, 278-283.
Chen WH, Cai MY, Zhang JX, Wang FW, Tang LQ, Liao YJ, Jin XH, Wang CY, Guo L, Jiang YG, Ren CP, Mai HQ, Zeng MS, Kung HF, Qian CN, Xie D (2018). FMNL1 mediates nasopharyngeal carcinoma cell aggressiveness by epigenetically upregulating MTA1. Oncogene, doi: 10.1038/s41388-018-0351-8.
138
Chesarone MA, DuPage AG, Goode BL (2010). Unleashing formins to remodel the actin and microtubule cytoskeletons. Nat Rev Mol Cell Biol 11, 62-74.
Chhabra ES and Higgs HN (2006). INF2 is a WASP homology 2 motif-containing formin that severs actin filaments and accelerates both polymerization and depolymerization. J Biol Chem 281, 26754-26767.
Colón-Franco JM, Gomez TS, Billadeau DD (2011). Dynamic remodeling of the actin cytoskeleton by FMNL1ɣ is required for structural maintenance of the Golgi complex. J Cell Sci 124, 3118-3126.
Copeland SJ, Thurston SF, Copeland JW (2016). Actin- and microtubule-dependent regulation of Golgi morphology by FHDC1. Mol Biol Cell 27, 260-276.
Cury JD, Campbell EJ, Lazarus CJ, Albin RJ, Welgus HG (1988). Selective up- regulation of human alveolar macrophage collagenase production by lipopolysaccharide and comparison to collagenase production by fibroblasts. J Immunol 141, 4306-4312.
Dettenhofer M, Zhou F, Leder P (2008). Formin 1-isoform IV deficient cells exhibit defects in cell spreading and focal adhesion formation. PLoS One 3, e2497.
DeWard AD, Eisenmann KM, Matheson SF, Alberts AS (2010). The role of formins in human disease. Biochim Biophys Acta 1803, 226-233.
Dollar G, Gombos R, Barnett AA , Sanchez Hernandez D, Maung SM, Mihály J, Jenny A (2016). Unique and overlapping functions of formins FRL and DAAM during ommatidial rotation and neuronal development in Drosophila. Genetics 202, 1135-1151.
Dutta P and Maiti S (2015). Expression of multiple formins in adult tissues and during developmental stages of mouse brain. Gene Expr Patterns 19, 52-59.
Eisenmann KM, West RA, Hildebrand D, Kitchen SM, Peng J, Sigler R, Zhang J, Siminovitch KA, Alberts AS (2007). T cell responses in mammalian diaphanous-related formin mDia1 knock-out mice. J Biol Chem 282, 25152-25158.
Favaro PM, de Souza Medina S, Traina F, Bassères DS, Costa FF, Saad ST (2003). Human leukocyte formin: a novel protein expressed in lymphoid malignancies and associated with Akt. Biochem Biophys Res Commun 311, 365-371.
Favaro P, Traina F, Machado-Neto JA, Lazarini M, Lopes MR, Pereira JK, Costa FF, Infante E, Ridley AJ, Saad ST (2013). FMNL1 promotes proliferation and migration of leukemia cells. J Leukoc Biol 94, 503-512.
Fernández-Barrera J and Alonso MA (2018). Coordination of microtubule acetylation and the actin cytoskeleton by formins. Cell Mol Life Sci 75, 3181-3191.
Gardberg M, Heuser VD, Iljin K, Kampf C, Uhlen M, Carpén O (2014). Characterization of Leukocyte Formin FMNL1 Expression in Human Tissues. J Histochem Cytochem 62, 460-470.
139
Gardberg M, Heuser VD, Koskivuo I, Koivisto M, Carpén O (2016). FMNL2/FMNL3 formins are linked with oncogenic pathways and predict melanoma outcome. J Pathol Clin Res 2, 41-52.
Gardberg M, Kaipio K, Lehtinen L, Mikkonen P, Heuser VD, Talvinen K, Iljin K, Kampf C, Uhlen M, Grénman R, Kovisto M, Carpén O (2013). FHOD1, a formin upregulated in epithelial-mesenchymal transition, participates in cancer cell migration and invasion. PLoS One 8, e74923.
Gardberg M, Talvinen K, Kaipio K, Iljin K, Kampf C, Uhlen M, Carpén O (2010a). Characterization of Diaphnous-related formin FMNL1 in human tissues. BMC Cell Biol 11, 55.
Gardberg M, Talvinen K, Kaipio K, Iljin K, Kampf C, Uhlen M, Carpén O (2010b). Characterization of Diaphahous-related formin FMNL2 in human tissues. BMC Cell Biol 11, 55.
Gbadegesin RA, Lavin PJ, Hall G, Bartkowiak B, Homstad A, Jiang R, Wu G, Byrd A, Lynn K, Wolfish N, Ottai C, Stevens P, Howell D, Conlon P, Winn MP (2012). Inverted formin 2 mutations with variable expression in patients with sporadic and hereditary focal and segmental glomerulosclerosis. Kidney Int 81, 94-99.
Goggolidou P, Hadjirin NF, Bak A, Papakrivopoulou E, Hilton H, Norris DP, Dean CH (2014). Atmin mediates kidney morphogenesis by modulating Wnt signaling. Hum Mol Genet 23, 5303-5316.
Gonzalo R and Arroyo AG (2010). MT1-MMP: A novel component of the macrophage cell fusion machinery. Commun Integr Biol 3, 256-259.
Goode BL and Eck MJ (2007). Mechanism and function of formins in the control of actin assembly. Annu Rev Biochem 76, 593-627.
Guerrier S, Coutinho-Budd J, Sassa T, Gresset A, Jordan NV, Chen K, Jin WL, Frost A, Polleux F (2009). The F-BAR domain of srGAP2 induces membrane protrusions required for neuronal migration and morphogenesis. Cell 138, 990-1004.
Guo Y, Cheng L, Ahmad S, Mao Y (2011). Formin mDia3: A novel target for Aurora B kinase. Bioarchitecture 1, 88-90.
Habas R, Kato Y, He x (2001). Wnt/Frizzled activation of Rho regulates vertebrate gastrulation and requires a novel formin homology protein DAAM1. 107, 843-854.
Han Y, Eppinger E, Schuster IG, Weigand LU, Liang X, Kremmer E, Peschel C, Krackhardt AM (2009). Formin-like 1 (FMNL1) is regulated by N-terminal myristoylation and induces polarized membrane blebbing. J Biol Chem 284, 33409-33417.
Han Y, Yu G, Sarioglu H, Caballero-Martinez A, Schlott F, Ueffing M, Haase H, Peschel C, Krackhardt AM (2013). Proteomic investigation of the interactome of FMNL1 in hematopoietic cells unveils a role in calcium-dependent plasticity. J Proteomics 78, 72- 82.
140
Harris ES, Li F, Higgs HN (2004). The mouse formin, FRL1α, slows actin filament barbed end elongation, competes with capping protein, accelerates polymerization from monomers, and severs filaments. J Biol Chem 279, 20076-20087.
Harris ES, Gauvin TJ, Heimsath EG, Higgs HN (2010). Assembly of filopodia by the formin FRL2 (FMNL3). Cytoskeleton (Hoboken) 67, 755-772.
Harris ES, Rouiller I, Hanein D, Higgs HN (2006). Mechanistic differences in actin bundling activity of two mammalian formins, FRL1 and mDia2. J Biol Chem 281, 14383- 14392.
Hayashi K, Horikoshi S, Osada S, Shofuda K, Shirato I, Tomino Y (2000). Macrophage- derived MT1-MMP and increased MMP-2 activity are associated with glomerular damage in crescentic glomerulonephritis. J Pathol 191, 299-305.
Henty-Ridilla JL, Rankova A, Eskin JA, Kenny K, Goode BL (2016). Accelerated actin filament polymerization from microtubule plus ends. Science 352, 1004-1009.
Higashi T, Ikeda T, Shirakawa R, Kondo H, Kawato M, Horiguchi M, Okuda T, Okawa K, Fukai S, Nureki O, Kita T, Horiuchi H (2008). Biochemical characterization of the Rho GTPase-regulated actin assembly by diaphanous-related formins, mDia1 and DAAM1, in platelets. J Biol Chem 283, 8746-8755.
Higgs HN and Peterson KJ (2005). Phylogenetic analysis of the formin homology 2 domain. Mol Biol Cell 16, 1-13.
Iskratsch T, Lange S, Dwyer J, Kho AL, dos Remedios C, Ehler E (2010). Formin follows function: a muscle-specific isoform of FHOD3 is regulated by CK2 phosphorylation and promotes myofibril maintenance. J Cell Biol 191, 1159-1172.
Jones SL, Wang J, Turck CW, Brown EJ (1998). A role for the actin-bundling protein L- plastin in the regulation of leukocyte integrin function. Proc Natl Acad Sci USA 95, 9331- 9336.
Jurmeister S, Baumann M, Balwierz A, Keklikoglou I, Ward A, Uhlmann S, Zhang JD, Wiemann S, Sahin Ö (2012). MicroRNA-200c represses migration and invasion of breast cancer cells by targeting actin-regulatory proteins FHOD1 and PPM1F. Mol Cell Biol 32, 633-651.
Kanaya H, Takeya R, Takeuchi K, Watanabe N, Jing N, Sumimoto H (2005). FHOS2, a novel formin-related actin-organizing protein, probably associates with the nestin intermediate filament. Genes Cells 10, 655-678.
Kan-o M, Takeya R, Taniguchi K, Tominaga R, Sumimoto H (2012). Expression and subcellular localization of mammalian formin FHOD3 in the embryonic and adult heart. PLoS One 7, e34765.
Katoh M and Katoh M (2003a). Identification of human FMNL1, FMNL2, and FMNL3 genes in silico. Int J Oncol 22, 1161-1168.
141
Katoh M and Katoh M (2003b). Identification and characterization of human DAAM2 gene in silico. Int J Oncol 22, 915-920.
Katoh M and Katoh M (2003c). Identification and characterization of human GRID2IP gene and rat GRID2IP gene in silico. Int J Mol Med 12, 1015-1019.
Katoh M and Katoh M (2004a). Identification and characterization of human FHDC1, mouse FHDC1, and zebrafish FHDC1 genes in silico. Int J Mol Med 13, 929-934.
Katoh M and Katoh M (2004b). Identification and characterization of the human FMN1 gene in silico. Int J Mol Med 14, 121-126.
Katoh M and Katoh M (2004c). Characterization of FMN2 gene at human chromosome 1q43. Int J Mol Med 14, 469-474.
Kim D, Jung J, You E, Ko P, Rhee S (2016). mDia1 regulates breast cancer invasion by controlling membrane type 1-matrix metalloproteinase localization. Oncotarget 7, 17829- 17843.
Kim D and Rhee S (2016). Matrix metalloproteinase-2 regulates MDA-MB-231 breast cancer cell invasion induced by active mammalian diaphanous-related formin 1. Mol Med Rep 14, 277-282.
Kitzing TM, Sahadevan AS, Brandt DT, Knieling H, Hannemann S, Fackler OT, Grosshans J, Grosse R (2007). Positive feedback between Dia1, LARG, and RhoA regulates cell morphology and invasion. Genes Dev 21, 1478-1483.
Kitzing TM, Wang Y, Pertz O, Copeland JW, Grosse R (2010). Formin-like 2 drives amoeboid invasive cell motility downstream of RhoC. Oncogene 29, 2441-2448.
Krainer EC, Ouderkirk JK, Miller EW, Miller MR, Mersich AT, Blystone SD (2013). The multiplicity of human formins: Expression patterns in cells and tissues. Cytoskeleton (Hoboken) 70, 424-438.
Lee HK and Deneen B (2012). DAAM2 is required for dorsal patterning via modulation of canonical Wnt signaling in the developing spinal cord. Dev Cell 22, 183-196.
Li Y, Zhu X, Zeng Y, Wang J, Zhang X, Ding YQ, Liang L (2010). FMNL2 enhances invasion of colorectal carcinoma by inducting epithelial-mesenchymal transition. Mol Cancer Res 8, 1579-1590.
Lin YN and Windhorst S (2016). Diaphanous-related formin 1 as a target for tumor therapy. Biochem Soc Trans 44, 1289-1293.
Lindenmeyer F, Legrand Y, Menashi S (1997). Upregulation of MMP-9 expression in MDA-MB-231 tumor cells by platelet granular membrane. FEBS Lett 418, 19-22.
Liu H, Kato Y, Erzinger SA, Kiriakova GM, Qian Y, Palmieri D, Steeg PS, Price JE (2012). The role of MMP-1 in breast cancer growth and metastasis to the brain in a xenograft model. BMC Cancer 12, 583.
142
Lizárraga F, Poincloux R, Romao M, Montagnac G, Le Dez G, Bonne I, Rigaill G, Raposo G, Chavrier P (2009). Diaphanous-related formins are required for invadopodia formation and invasion of breast tumor cells. Cancer Res 69, 2792-2800.
Mason FM, Heimsath EG, Higgs HN, Soderling SH (2011). Bi-modal regulation of a formin by srGAP2. J Biol Chem 286, 6577-6586.
Masters JR (2002). HeLa cells 50 years on: the good, the bad and the ugly. Nat Rev Cancer 2, 315-319.
Matsuda K, Matsuda S, Gladding CM, Yuzaki M (2006). Characterization of the δ2 glutamate receptor-binding protein Delphilin: Splicing variants with differential palmitoylation and an additional PDZ domain. J Biol Chem 281, 25577-25587.
Mersich AT, Miller MR, Chkourko H, Blystone SD (2010). The formin FRL1 (FMNL1) is an essential component of macrophage podosomes. Cytoskeleton (Hoboken) 67, 573- 585.
Miller MR and Blystone SD (2015). Human macrophages utilize the podosome formin FMNL1 for adhesion and migration. Cellbio 4, 1-11.
Miller MR, Miller EW, Blystone SD (2017). Non-canonical activity of the podosomal formin FMNL1ɣ supports immune cell migration. J Cell Sci 130, 1730-1739.
Miyagi Y, Yamashita T, Fukaya M, Sonoda T, Okuno T, Yamada K, Watanabe M, Nagashima Y, Aoki I, Okuda K, Mishina M, Kawamoto S (2002). Delphilin: a novel PDZ and formin homology domain-containing protein that synaptically colocalizes and interacts with glutamate receptor δ2 subunit. J Neurosci 22, 803-814.
Murphy DA and Courtneidge SA (2011). The ‘ins’ and ‘outs’ of podosomes and invadopodia: characteristics, formation and function. Nat Rev Mol Cell Biol 12, 413-426.
Nolan T, Hands RE, Bustin SA (2006). Quantification of mRNA using real-time RT-PCR. Nat Protoc 1, 1559-1582.
Panzer L, Trübe L, Klose M, Joosten B, Slotman J, Cambi A, Linder S (2016). The formins FHOD1 and INF2 regulate inter- and intra-structural contractility of podosomes. J Cell Sci 129, 298-313.
Paul NR, Allen JL, Chapman A, Morlan-Mairal M, Zindy E, Jacquemet G, Fernandez del Ama L, Ferizovic N, Green DM, Howe JD, Ehler E, Hurlstone A, Caswell PT (2015a). α5β1 integrin recycling promotes Arp2/3-indpendent cancer cell invasion via the formin FHOD3. J Cell Biol 210, 1013-1031.
Paul NR, Jacquemet G, Caswell PT (2015b). Endocytic Trafficking of Integrins in Cell Migration. Curr Biol 25, R1092-1105.
Péladeau C, Heibein A, Maltez MT, Copeland SJ, Copeland JW (2016). A specific FMNL2 isoform is up-regulated in invasive cells. BMC Cell Biol 17, 32.
143
Peng J, Wallar BJ, Flanders A, Swiatek PJ, Alberts AS (2003). Disruption of the Diaphanous-related formin DRF1 gene encomding mDia1 reveals a role for Drf3 as an effector for Cdc42. Curr Biol 13, 534-545.
Pettee KM, Dvorak KM, Nestor-Kalinoski AL, Eisenmann KM (2014). An mDia2/ROCK signaling axis regulates invasive egress from epithelial ovarian cancer spheroids. PLoS One 9, e90371.
Phromnoi K, Yodkeeree S, Anuchapreeda S, Limtrakul P (2009). Inhibition of MMP-3 activity and invasion of the MDA-MB-231 human invasive breast carcinoma cell line by bioflavonoids. Acta Pharmacol Sin 30, 1169-1176.
Pollard TD and Cooper JA (2009). Actin, a central player in cell shape and movement. Science 326, 1208-1212.
Pruyne D (2016). Revisiting the phylogeny of the animal formins: Two new subtypes, relationships with multiple wing hairs proteins, and a lost human formin. PLoS One 11, e0164067.
Pruyne D (2017). Probing the origins of metazoan formin diversity: Evidence for evolutionary relationships between metazoan and non-metazoan formin subtypes. PLoS One 12, e0186081.
Rosado M, Barber CF, Berciu C, Feldman S, Birren SJ, Nicastro D, Goode BL (2014). Critical roles for multiple formins during cardiac myofibril development and repair. Mol Biol Cell 25, 811-827.
Ryan MJ, Johnson G, Kirk J, Fuerstenberg SM, Zager RA, Torok-Storb B (1994). HK-2: an immortalized proximal tubule epithelial cell line from normal adult human kidney. Kidney Int 45, 48-57.
Sahasrabudhe A, Ghate K, Mutalik S, Jacob A, Ghose A (2016). Formin 2 regulates the stabilization of filopodial tip adhesions in growth cones and affects neuronal outgrowth and pathfinding in vivo. Development 143, 449-460.
Schönichen A and Geyer M (2010). Fifteen formins for an actin filament: a molecular view on the regulation of human formins. Biochim Biophys Acta 1803, 152-163.
Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber M, Gassmann M, Lightfoot S, Menzel W, Granzow M, Ragg T (2006). The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Mol Biol 7, 3.
Schuster IG, Busch DH, Eppinger E, Kremmer E, Milosevic S, Hennard C, Kuttler C, Elwart JW, Frankenberger B, Nössner E, Salat C, Bogner C, Borkhardt A, Kolb HJ, Krackhardt AM (2007). Allorestricted T cells with specificity for the FMNL1-derived peptide PP2 have potent antitumore activity against hematologic and other malignancies. Blood 110, 2931-2939.
Simon-Areces J, Dopazo A, Dettenhofer M, Rodriguez-Tebar A, Garcia-Segura LM, Arevalo MA (2011). Formin1 mediates the induction of dendritogenesis and synaptogenesis by neurogenin3 in mouse hippocampal neurons. PLoS One 6, e21825.
144
Soykan T, Kaempf N, Sakaba T, Vollweiter D, Goerdeler F, Puchkov D, Kononenko NL, Haucke V (2017). Synaptic vesicle endocytosis occurs on multiple timescales and is mediated by formin-dependent actin assembly. Neuron 93, 854-866.
Ståhlberg A, Håkannson J, Xian X, Semb H, Kubista M (2004). Properties of the reverse transcription reaction in mRNA quantification. Clin Chem 50, 509-515.
Stastna J, Pan X, Wang H, Kollmannsperger A, Kutscheidt S, Lohmann V, Grosse R, Fackler OT (2012). Differing and isoform-specific roles for the formin DIAPH3 in plasma membrane blebbing and filopodia formation. Cell Res 22, 728-745.
Steenport M, Khan KM, Du B, Barnhard SE, Dannenberg AJ, Falcone DJ (2009). Matric metalloproteinase (MMP)-1 and MMP-3 induce macrophage MMP-9: evidence for the role of TNF-α and cyclooxygenase-2. J Immunol 183, 8119-8127.
Subramanian B, Sun H, Yan P, Charooratana VT, Higgs HN, Wang F, Lai KV, Valenzuela DM, Brown EJ, Schlöndorff JS, Pollak MR (2016). Mice with mutant INF2 show impaired podocyte and slit diaphragm integrity in response to protamine-induced kidney injury. Kidney Int 90, 363-372.
Sun H, Schlondorff JS, Brown EJ, Higgs HN, Pollak MR (2011). Rho activation of mDia formins is modulated by an interaction with inverted formin 2 (INF2). Proc Natl Acad Sci USA 108, 2933-2938.
Thomas SG, Calaminus SD, Machesky LM, Alberts AS, Watson SP (2011). G-protein coupled and ITAM receptor regulation of the formin FHOD1 through Rho kinase in platelets. J Thromb Haemost 9, 1648-1651.
Vega FM, Fruhwirth G, Ng T, Ridley AJ (2011). RhoA and RhoC have distinct roles in migration and invasion by acting through different targets. J Cell Biol 193, 655-665.
Wang Y, Arjonen A, Pouwels Jm Ta H, Pausch P, Bange G, Engel U, Pan X, Fackler OT, Ivaska J, Grosse R (2015). Formin-like 2 promototes β1-integrin trafficking and invasive motility downstream of PKCα. Dev Cell 34, 475-483.
Welgus HG, Campbell EJ, Cury JD, Eisen AZ, Senior RM, Wilhelm SM, Goldberg GI (1990). Netural metalloproteinases produced by human mononuclear phagocytes. Enzyme profile, regulation, and expression during cellular development. J Clin Invest 86, 1496-1502.
Westendorf JJ, Mernaugh R, Hiebert SW (1999). Identification and characterization of a protein containing homology (FH1/FH2) domains. Gene 232, 173-182.
Wu Y, Shen Z, Wang K, Ha Y, Lei H, Jia Y, Ding R, Wu D, Gan S, Li R, Luo B, Jiang H, Jie W (2017). High FMNL3 expression promotes nasopharyngeal carcinoma cell metastasis: role in TGF-β1-induced epithelia-to-mesenchymal transition. Sci Rep 7, 42507.
Wyse MM, Goicoechea S, Garcia-Mata R, Nestor-Kalinoski AL Eisenmann KM (2017). mDia2 and CXCL12/CXCR4 chemokine signaling intersect to drive tumor cell amoeboid morphological transitions. Biochem Biophys Res Commun 484, 255-261.
145
Yamaguchi H and Condeelis J (2007). Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochim Biophys Acta 1773, 642-652.
Yayoshi-Yamamoto S, Taniuchi I, Watanabe T (2000). FRL, a novel formin-related protein, binds to Rac and regulates cell motility and survival of macrophages. Mol Cell Biol 20, 6872-6881.
Zhu W, Krishna S, Garcia C, Lin CJ, Mitchell BD, Scott KL, Mohila CA, Creighton CJ, Yoo SH, Lee HK, Deneen B (2017). DAAM2 driven degradation of VHL promotes gliomagenesis. Elife 6, e31926.
Zhu XL, Liang L, Ding YQ (2008). Overexpression of FMNL2 is closely related to metastasis of colorectal cancer. Int J Colroectal Dis 23, 1041-1047.
Zuidscherwoude M, Green HLH, Thomas SG (2018). Formin proteins in megakaryocytes and platelets: regulation of actin and microtubule dynamics. Platelets 18, 1-8.
146
Chapter 3
Two highly conserved amino acid residues of the FMNL1 FH2 domain are required for actin binding.
Parts of this work were previously published in:
Miller MR, Miller EW, Blystone SD (2017). Non-canonical activity of the podosomal formin FMNL1ɣ supports immune cell migration. J Cell Sci 130, 1730-1739.
This chapter focuses on the work I performed for this project in the context of this dissertation.
147
Introduction
FMNL1 is unique among the formin family members as it has the ability to bundle, sever, assemble, and cap actin filaments. Classically, formin activity has been ascribed predominately to the highly conserved FH2 domain. More specifically, two hydrophobic amino acid residues have been identified as key regulators of formin-actin interaction: an aliphatic isoleucine residue and a positively charged, polar lysine residue
(Table 1). I1431 and K1601 were first identified in the S. cerevisiae formin Bni1 and have since been shown as critical regulators of formin function in several members of this protein family (Xu et al., 2004).
We previously have demonstrated that FMNL1 is required for proper human macrophage adhesion and migration in vitro, verifying its regulation of the podosomes of these cells. This led us to question if these functional contributions of FMNL1 were also true in vivo. In order to test this hypothesis, we developed a murine FMNL1 knockout model. We generated FMNL1 floxed mice with loxP sites flanking exons 4, 5, and 6 of
FMNL1, which, upon interaction with Cre recombinase, results in an excision of all three exons and a subsequent frameshift mutation, ultimately resulting in an invalid transcript of FMNL1 (Miller et al., 2017).
We bred floxed FMNL1 mice with mice expressing Cre recombinase under different promoters, including the Ella promoter, which results in global gene deletion, and the Lyz2 promoter, which results in gene deletion in cells of a myeloid lineage.
Remarkably, we determined that global depletion of FMNL1 resulted in embryonic lethality at a 95% confidence interval (CI). This suggests that FMNL1 plays necessary roles in development. Additionally, developmental functions are not widespread amongst formins as only deletion of Dia2, FHOD3, and DAAM1 have been shown to be embryonically lethal (Miller et al., 2017).
148
As a result of this embryonic lethality, and as we were questioning the function of
FMNL1 specifically in macrophages, we altered the breeding strategy by utilizing a different mouse model in conjunction with our FMNL1 floxed mice. Using LysMCre mice, which express Cre recombinase under the Lyz2 promoter, we conditionally deleted
FMNL1 gene expression in mouse macrophages. The resulting mice were healthy, with no significant alterations in the peripheral blood or other tissues. However, we did observe a reduction of Kupffer cell residency in the liver, suggesting that FMNL1 is required for normal residential macrophage tissue distribution. Using an in vivo inflammation model and 2D migration assay, we were also able to determine that
FMNL1 is required for proper macrophage migration. Furthermore, we confirmed that depletion of FMNL1 expression resulted in a significant reduction of the number of macrophages forming podosomes and a significant increase in the surface area of these cells, indicating FMNL1 is required for both podosome formation and regulation of cellular morphology (Miller et al., 2017).
In order to confirm deletion of FMNL1 in FMNL1-depleted macrophages, we performed reconstitution experiments with human FMNL1. In humans, three alternative splice isoforms of FMNL1 are actually expressed: FMNL1α, FMNL1β, and FMNL1ɣ.
These unique isoforms diverge at the C-terminal DAD domain of FMNL1, more precisely at T1069. FMNL1α and FMNL1ɣ actually share a common end sequence, however,
FMNL1ɣ retains a 58 amino acid residue sequence before this shared region. FMNL1β has an entirely unique sequence which is slightly longer than FMNL1α (Katoh and
Katoh, 2003, Han et al., 2009). Reconstitution of FMNL1-depleted macrophages with full-length FMNL1β and FMNL1ɣ demonstrated that the migration defect observed in
FMNL1-null macrophages could be rescued specifically with FMNL1ɣ but not FMNL1β.
Additionally, using fluorescence microscopy, we confirmed that only FMNL1ɣ localized to
149 macrophage podosomes while FMNL1β displayed a diffuse localization throughout the cytoplasm (Miller et al., 2017).
Two highly conserved residues of the FH2 domain have been shown to be essential for formin-actin interactions in several formin family members (Xu et al., 2004).
These residues correspond with I720 and K871 in FMNL1 (Table 1). We questioned whether these residues were required for the function of FMNL1ɣ in macrophage podosomes. We mutated both residues, I720A and K871A, in a full-length, GFP-
FMNL1ɣ vector and reconstituted FMNL1-null macrophages using this plasmid. We observed that even with these two residues mutated, GFP-FMNL1ɣ still localized to podosomes and was able to rescue the previously observed migration defect (Miller et al., 2017).
Similar mutations of these conserved FH2 domain residues have been made in other formins and different effects have been observed at both the cellular and biochemical level (Table 1). The FHOD3 mutations of I1127A or K1273D results in loss of stress fiber formation in HeLa cells and sarcomere disorganization in myofibrils of cardiomyocytes (Taniguchi et al., 2009). The isoleucine mutation in DAAM1 (I698A) results in a loss of stress fiber formation ability in NIH 3T3 cells, and mutating both amino acid residues (I698A and K847D) inhibits actin filament assembly in vitro (Lu et al., 2007, Liu et al., 2008). Similar results were observed with Dia1, where the corresponding isoleucine and lysine mutations were found to inhibit actin filament assembly in vitro and morphological defects were observe in U2OS cells expressing this mutant (Otomo et al., 2005a, Hotulainen and Lappalainen, 2006, Daou et al., 2014).
Expressing Dia3 with this isoleucine mutation results in inhibition of bleb formation and reduction of serum response factor transcriptional activity (Stastna et al., 2012). In
FHDC1, mutating the conserved isoleucine residue, I180A, results in a loss of Golgi dispersion abilities, demonstrating that the FH2 domain is essential for this activity
150
(Copeland et al., 2016). Furthermore, expression of this same isoleucine mutant in
FHDC1 inhibits regulation of actin dynamics in cilia of NIH 3T3 cells (Copeland et al.,
2018). Even in Delphilin, actin assembly is abolished with I718A and slightly reduced with K868A in vitro (Silkworth et al., 2018). However, not all formins are affected by these mutations, as expression of INF2 with mutations of the isoleucine and the lysine residues does not affect F-actin distribution in Jurkat cells (Andrés-Delgado et al., 2012).
Analysis of the effects of these mutations on the FMNL subfamily have also been performed. For example, while FMNL2 (I704A) and FMNL3 (I649A) still co-localize with
Cdc42, perturbation of actin accumulations are observed with expression of WT FMNL2 and FMNL3. Additionally, Co-IP experiments demonstrate abolishment of F-actin binding (Kage et al., 2017). In FMNL3, the conserved isoleucine residue has been mutated to an alanine (I649A) and subsequent downstream experiments demonstrated that that this did indeed affect actin filament elongation and F-actin bundling activity.
Barbed end binding activity was eliminated with this mutation and actin filament elongation was inhibited as well, effectively eradicating filopodia assembly in Jurkat,
300.19, and NIH 3T3 cells (Harris et al., 2010). This same mutation in FMNL3 also affects actin assembly and abolishes barbed end binding in vitro (Heimsath and Higgs,
2012, Thompson et al., 2013).
In FMNL1, effects of these mutations have also been observed biochemically and at the cellular level. A previous study demonstrated that in mouse FMNL1β
(originally referred to as mouse FRL1α), this mutant protein inhibits actin filament elongation similar to the WT construct in vitro. This mutation also results in rapid dissociation from the barbed end of actin filaments but does not affect bundling activity
(Harris et al., 2006). Expression of FMNL1ɣ with this mutation is also necessary for
Golgi structural maintenance regulation (Colón-Franco et al., 2011).
151
The localization of FMNL1ɣ-ABM to macrophage podosomes and its ability to rescue the previously observed migratory defect in FMNL1-null macrophages demonstrated that the FH2-mediated barbed end binding was dispensable for FMNL1ɣ function in these cells. In order to confirm that barbed end binding was indeed abolished, we performed different actin pull-down assays using recombinant fusion proteins coding for the different FMNL1 alternative splice isoforms. These constructs coded for the FH1, FH2, and entire C-terminal region. These pull-down assays demonstrated that mutation of the conserved isoleucine and lysine residues results in abolition of FH2-mediated actin filament interactions.
Table 1
Table 1: The highly conserved Ile and Lys residues of the FH2 domain. Originally identified in Bni1, the highly conserved Ile and Lys residues of the FH2 domain are aligned here for each of the 15 H. sapiens formins along with Bni1 from S. cerevisiae and FMNL1 from M. musculus. The conserved isoleucine and lysine residues are highlighted in red and numbers next to each sequence indicate the starting and ending amino acid residue number for the regions of the FH2 domains identified here.
152
Materials and Methods
Human macrophages were derived from peripheral blood monocytes as described previously (Miller and Blystone, 2015). Whole-cell RNA was extracted and purified from primary human macrophages with the RNeasy Mini Kit (Qiagen, Hilden,
Germany). Primary human macrophage cDNA was generated from RNA using the
SuperScript III First-Strand Synthesis System for reverse-transcriptase (RT)-PCR (Life
Technologies). Full-length FMNL1 was cloned into a previously generated pEGFP vector from primary human macrophage cDNA (Chandoke et al., 2004). FMNL1α
(accession: NM_005892), FMNL1β (accession: BC001710), and FMNL1ɣ
(accession: FJ534522) were cloned from primary human macrophage cDNA via PCR.
The truncation mutant GFP-FMNL1Δ, truncated at the splice site T1069, was cloned via
PCR using the forward primer 5’-GCACTGAAACCCAGCCAGATCACC-3’ and the reverse primer 5’-CCGGCGGCCGCTATGTGATGATGTCTTCAATGG-3’. A stop codon and a NotI site were introduced at the C-terminal splice site T1069 via PCR mutagenesis. A NotI site was introduced into each isoform following the stop codon at the C-terminus via PCR mutagenesis and was used in conjunction with a BsiWI site for transfer to pEGFP-FMNL1. The N-terminal XbaI site and C-terminal BamHI site were used to transfer FMNL1β and FMNL1ɣ into pLVX-AcGFP-C1 (Clontech). All FMNL1 constructs were verified by DNA sequencing.
Two substitution mutations, I720A and K871A, were generated to eliminate actin binding by PCR ‘sewing’ into the FMNL1ɣ isoform. For the I720A mutation, primers used for the first reaction were forward primer 5′-
CCGAATGCCACTCTTGAACTGGGTGGC-3′ and reverse primer 5′-
CGCAGGGTGGCGGCCAAGTTCTTGGCCCGG-3′. The second reaction primers were forward primer 5′-CGGGCCAAGAACTTGGCCGCCACCCTGCG-3′ and reverse primer
153
5′-CGAGGCTGATCAGCGGGTTTAAACG-3′. The combining reaction primers were forward primer 5′-CCGAATGCCACTCTTGAACTGGGTGGC-3′ and reverse primer 5′-
CGAGGCTGATCAGCGGGTTTAAACG-3′. For the K871A mutation, the primers used for the first reaction were forward primer 5′-CCGAATGCCACTCTTGAACTGGGTGGC-3′ and reverse primer 5′-GCTTGCGATCAGTCGAGGCCATCTCCAACAGC-3′. The second reaction primers were 5′-GCTGTTGGAGATGGCCTCGACTGATCG-3′ and reverse primer 5′-CGAGGCTGATCAGCGGGTTTAAACG-3′. The combining reaction primers were forward primer 5′-CCGAATGCCACTCTTGAACTGGGTGGC-3′ and reverse primer 5′-CGAGGCTGATCAGCGGGTTTAAACG-3′. Nucleotides in bold indicate introduced changes to the sequence. The PCR product containing both substitution mutations was inserted into the pLVX-AcGFP-FMNL1ɣ lentiviral vector between SfiI and NotI sites. The GFP–FMNL1ɣ actin-binding mutant (ABM/FH2ø) was transformed into DH10β competent cells (Life Technologies), and mutations were confirmed by DNA sequencing. Constructs encoding GST fusion proteins of FMNL1ɣ and the FMNL1ɣ-FH2ᴓ mutant were prepared by performing PCR. For solubility, these constructs encode FMNL1 beginning with the FH1 domain and continuing through to the
C-terminus. Fusion proteins were grown in Escherichia coli using standard techniques and harvested with glutathione–agarose. Agarose-bound FMNL1ɣ-CT and FMNL1ɣ-
FH2ᴓ-CT were incubated with lysate from MDA-MB-231 cells, washed six times in PBS and precipitates were subjected to SDS-PAGE and western blotting to demonstrate loss of actin binding in the FH2ᴓ mutant.
Results
The highly conserved Ile and Lys residues of the FH2 domain of formins are required for
FMNL1ɣ-mediated actin binding.
154
Past studies have previously identified two highly conserved amino acid residues of the FH2 domain of formins are necessary for formin-mediated barbed end binding (Xu et al., 2004, Otomo et al., 2005b). We performed actin pull-down assays to directly assess the requirement of these highly conserved Ile and Lys residues in FMNL1-actin interactions. Using GST-tagged fusion proteins coding for the FH1 domain of FMNL1 and ending at the C-terminus bound to glutathione-sepharose, we directly tested the ability of FMNL1Δ, FMNL1ɣ, and FMNL1ɣFH2ø to bind actin. Both supernatants and pellets from these assays were separated via SDS-PAGE electrophoresis, probed for with α-actin 1° antibody, and visualized using an HRP-conjugated 2° antibody and enhanced chemiluminescence (ECL). FMNL1Δ, the mutant truncated at the splice site
T1069, and FMNL1ɣ were both able to directly bind actin in these pull-down assays
(Figure 1). FMNL1ɣFH2ø was unable to bind actin, indicated by the absence of detectable actin observed in the pellet lane. These data suggest that the highly conserved Ile and Lys residues of the FMNL1 FH2 domain are required for actin-binding.
This analysis also demonstrates that truncation of the C-terminal region of FMNL1 does not affect the binding of this formin to actin, further implicating the importance of the FH2 domain in this function.
155
Figure 1
Figure 1: The ability of FMNL1ɣ to bind actin is mediated by the FH2 domain and not dependent on the C-terminal region. Representative Western blot demonstrating that the ability of FMNL1ɣ to bind actin is regulated by the FH2 domain. Actin pull- downs were performed as described in Materials and Methods, separated via SDS-
PAGE, and probed for with α-Actin 1° antibody. Numbers on the left indicate MW in kDa. S = supernatant, P = pellet.
The amino acid residue composition of the FMNL1 alternative splice isoform C-terminal region does not affect the actin-binding ability of this protein.
After confirming that the highly conserved Ile and Lys residues of FMNL1 are required for actin-binding, we next questioned whether the C-terminus of the FMNL1 alternative splice isoforms may affect the ability of this protein to bind actin. Using the same actin pull-down assays that were previously described, we observed the actin- binding abilities of FMNL1β, FMNL1ɣ, and FMNL1ɣFH2ø using GST-tagged fusion proteins coding for the FH1 and FH2 domains of these different alternative splice isoforms and ending at their C-termini. Both supernatants and pellets were separated via SDS-PAGE electrophoresis, probed for with α-actin 1° antibody, and visualized with
156
HRP-conjugated 2° antibody and ECL. Both FMNL1β and FMNL1ɣ showed actin in the pellet, demonstrating their ability to properly actin (Figure 2). However, FMNL1ɣFH2ø was unable to bind actin, indicated by the lack of actin in the pellet lane (Figure 2). This further confirms that the FH2 domain is required for FMNL1-mediated actin-binding.
Additionally, the confirmation that both FMNL1β and FMNL1ɣ pull-down actin suggests that the variance in C-terminal amino acid residue composition does not affect the actin- binding ability of the protein.
Figure 2
Figure 2: Both FMNL1β and FMNL1ɣ bind actin, but FMNL1ɣFH2ø does not.
Representative Western blot demonstrating that the ability of FMNL1β and FMNL1ɣ to bind actin is regulated by the FH2 domain. GST-FMNL1-CT pull-downs were performed as described in Materials and Methods, separated via SDS-PAGE, and probed for with
α-Actin. Numbers on the left indicate MW in kDa. S = supernatant, P = pellet.
Discussion
Herein, we have demonstrated that the highly conserved isoleucine (I720) and lysine (K871) residues of the FH2 domain of FMNL1ɣ are required for barbed end
157 binding of the formin to F-actin. Furthermore, as a whole, this paper demonstrates that the FH2-mediated barbed end binding of FMNL1ɣ is not required for regulation of macrophage podosome function, effectively demonstrating this is reliant on another region of the protein.
This study represents one of the first reports demonstrating that the highly conserved isoleucine and lysine residues of the FH2 domain of a specific alternative splice isoform of FMNL1 are required for actin-binding. Furthermore, this is also one of the first reports demonstrating this with actin from whole cell lysates as opposed to isolated and purified muscle actin. This confirmation that the FH2 domain is required for binding of FMNL1 to actin isolated from cells confirms previous work identifying the FH2 domain as the essential component of formin-actin interactions (Shimada et al., 2004,
Xu et al., 2004, Otomo et al., 2005b, Harris et al., 2006).
Regardless of the C-terminal region amino acid composition of the FMNL1 alternative splice isoforms, both FMNL1β and FMNL1ɣ were able to successfully bind actin. FMNL1Δ, the mutant truncated at the splice site is also able to bind actin, confirming the FH2 domain is a key regulator of this function. However, past studies have implicated the C-terminal region of some formins, especially of the FMNL subfamily, as actin-interaction regions of the protein (Chhabra and Higgs, 2006, Vaillant et al., 2008, Gould et al., 2011, Heimsath and Higgs, 2012). As FMNL1Δ had no deleterious effect on actin-binding, this demonstrates that the C-terminal region of
FMNL1 may not be required for this function. However, we do find this somewhat peculiar since the FMNL1ɣ C-terminal region actually contains two predicted WH2 domains.
WH2 motifs have been identified in several actin-associated proteins and seem to have several functions (Campellone and Welch, 2010, Carlier et al., 2013,
Dominguez, 2016). Compared to other WH2 motifs, those predicted in FMNL1ɣ are
158 unique in that their amino acid composition is quite different from those identified in
“classic” WH2 domain-containing proteins, such as WASp and JMY (Dominguez et al.,
2016). This unique amino acid residue sequence composition could be contributing to
FMNL1 function not through potent actin-binding as one would observe with the FH2 domain, but perhaps other actin-modifying activities. For instance, FMNL3 and INF2 have both exhibited actin filament severing activity that could rely on their WH2 domains
(Chhabra and Higgs, 2006, Heimsath and Higgs, 2012). Additionally, the C-terminal region may still be able to interact with actin, but at a much lower level than the FH2 domain. Furthermore, it could be important for binding actin monomers which are eventually recruited to the FH2 domain-bound barbed end. This makes sense since even though barbed end binding could be abolished via FH2 domain mutations, the
WH2 domain would still bind monomeric actin. As a result, while the formin could still bind G-actin, it would no longer be able to bind the barbed ends of F-actin, resulting in the lack of actin we observe in the Western blot analysis of the GST-FMNL1-CT pull- down assays.
This distinct amino acid residue sequence of two overlapping WH2 domains is exclusive to FMNL1ɣ and could explain the actin-modifying functions we have observed.
Much like Spire and FMNL3, WH2 motifs in FMNL1ɣ could be capping actin filaments and preventing assembly (Bosch et al., 2007, Carlier et al., 2013, Heimsath and Higgs,
2012). While the question of how this unique WH2 domain contributes to the bundling process remains to be determined, the work here advances understanding of this process and the mechanism behind WH2 domain function in formins.
Previous studies in mouse FMNL1β demonstrated that mutation of the conserved isoleucine residue results in a reduction of barbed end binding but does not affect the bundling or elongation activities of this protein. This mutation also resulted in mouse
FMNL1β losing its ability to compete with capping protein by dissociating at a higher rate
159 than the WT-FMNL1β. This high dissociation rate, in addition to other data demonstrating that bundling activity of FMNL1β is affected by salt concentration, suggest that this protein is binding to the sides of actin filaments with amino acid residues on the outside of the FH2 domain via electrostatic interactions (Harris et al., 2006). This would make sense as the highly conserved isoleucine and lysine residues are located on the inside of the FH2 domain (Xu et al., 2004, Otomo et al., 2005b, Harris et al., 2006).
The effects of this isoleucine mutation have also been observed at the cellular level using a HeLa cell model system. Depletion of FMNL1 expression in these cells results in Golgi fragmentation which can only be rescued by the FMNL1ɣ alternative splice isoform, not FMNL1α or FMNL1β. Rescue attempts via expression of FMNL1ɣ with either an I720A or K871D mutation does not rescue the fragmented Golgi phenotype observed, indicating that barbed end interactions of FMNL1ɣ with F-actin is required for regulation of Golgi complex structure (Colón-Franco et al., 2011). In contrast to our own data, this shows that FMNL1ɣ requires FH2 domain-mediated barbed end binding to rescue cellular structures affected by depletion of endogenous
FMNL1. Indeed, this suggests that in our own study on macrophage podosomes, while similar to this study in that a specific alternative splice isoform of FMNL1 is required for regulation of a specific cellular structure, different regions of the molecule are necessary for this. Whether this difference is due to cell type or cellular structure remains to be determined.
Taken as a whole with the entirety of this published work, these data demonstrate that the primary function of FMNL1 is not dependent on FH2 domain- mediated barbed end binding. Indeed, other functions of FMNL1 have been suggested such as severing and bundling. Interestingly, the severing activity of FMNL3 has been shown to not be affected when barbed end binding is inhibited via mutation of the conserve Ile residue. However, mutating specific amino acid residues of the WH2
160 domain in the C-terminus does inhibit the ability of FMNL3 to sever actin filaments
(Heimsath and Higgs, 2012). Furthermore, mouse FMNL1β has previously been shown to act as a severing protein, a function which has been shown to require the WH2 domain in other proteins (Harris et al., 2004, Chhabra and Higgs, 2006, Jiao et al.,
2014). Previous studies have also demonstrated that while mutation of the conserved
Ile of the FH2 domain does indeed inhibit barbed end binding, this does not affect the ability of mouse FMNL1β to bundle actin filaments, actually increasing bundling activity 2 fold compared to WT FMNL1β (Harris et al., 2006). In FMNL3, deletion of the WH2 domains found in the C-terminus reduces actin bundling activity and in FMNL2, these mutations result in complete abrogation of any bundling activity (Vaillant et al., 2008).
Moreover, WH2 and WH2-like domains have actually been shown to be essential for some proteins to form actin filament bundles (Loomis et al., 2006, Millard et al., 2007).
Future studies examining these different actin-modifying functions and what role the
FMNL1 C-terminal region plays in them will allow us to not only distinctly identify a unique function of this protein, but also demonstrate that formin-actin interactions are not entirely based on the FH2 domain.
161
References
Andrés-Delgado L, Antón OM, Bartolini F, Ruiz-Sáenz A, Correas I, Gundersen GG, Alonso MA (2012). INF2 promotes the formation of detyrosinated microtubules necessary for centrosome reorientation in T cells. J Cell Biol 198, 1025-1037.
Campellone KG and Welch MD (2010). A nucleator arms race: cellular control of actin assembly. Nat Rev Mol Cell Biol 11, 237-251.
Carlier MF, Pernier J, Avvaru BS (2013). Control of actin filament dynamics at barbed ends by WH2 domains: from capping to permissive and processive assembly. Cytoskeleton 70, 540-549.
Chhabra ES and Higgs HN (2006). INF2 is a WASP homology 2 motif-containing formin that severs actin filaments and accelerates both polymerization and depolymerization. J Biol Chem 281, 26754-26767.
Colón-Franco JM, Gomez TS, Billadeau DD (2011). Dynamic remodeling of the actin 0cytoskeleton by FMNL1ɣ is required for structural maintenance of the Golgi complex. J Cell Sci 124, 3118-3126.
Copeland SJ, Thurston SF, Copeland JW (2016). Actin- and microtubule-dependent regulation of Golgi morphology by FHDC1. Mol Biol Cell 27, 260-276.
Copeland SJ, McRae A, Guarguaglini G, Tinkle-Mulcahy L, Copeland JW (2018). Actin- dependent regulation of cilia length by the inverted formin FHDC1. Mol Biol Cell 29, 1611-1627.
Daou P, Hasan S, Breitsprecher D, Baudelet E, Camoin L, Audebert S, Goode BL, Badache A (2014). Essential and nonredundant roles for Diaphanous formins in cortical microtubule capture and directed cell migration. Mol Biol Cell 25, 658-668.
Dominguez R (2016). The WH2 Domain and Actin Nucleation: Necessary but Insufficient. Trends Biochem Sci 41, 478-490.
Gould CJ, Maiti S, Michelot A, Graziano BR, Blanchoin L, Goode BL (2011). The formin DAD domain plays dual roles in autoinhibition and actin nucleation. Curr Biol 21, 384- 390.
Han Y, Epiinger E, Schuster IG, Weigand LU, Liang X, Kremmer E, Peschel C, Krackhardt AM (2009). Formin-like 1 (FMNL1) is regulated by N-terminal myristoylation and induces polarized membrane blebbing. J Biol Chem 284, 33409-33417.
Jiao Y, Walker M, Trinick J, Pernier J, Montaville P, Carlier MF (2014). Mutagenetic and electron microscopy analysis of actin filament severing by Cordon-Bleu, a WH2 domain protein. Cytoskeleton 71, 170-183.
Harris ES, Li F, Higgs HN (2004). The mouse formin, FRLα, slows actin filament barbed end elongation, competes with capping protein, accelerates polymerization from monomers, and severs filaments. J Biol Chem 279, 20076-20087.
162
Harris ES, Rouiller I, Hanein D, Higgs HN (2006). Mechanistic differences in actin bundling activity of two mammalian formins, FRL1 and mDia2. J Biol Chem 281, 14383- 14392.
Harris ES, Gauvin TJ, Heimsath EG, Higgs HN (2010). Assembly of filopodia by the formin FRL 2 (FMNL3). Cytoskeleton 67, 755-772.
Heimsath EG Jr and Higgs HN (2012). The C-terminus of formin FMNL3 accelerates actin polymerization and contains a WH2 domain-like sequence that binds both monomers and filament barbed ends. J BIol Chem 287, 3087-3098.
Hotulainen P and Lappalainen P (2006). Stress fibers are generated by two distinct assembly mechanisms in motile cells. J Cell Biol 173, 383-394.
Kage F, Steffen A, Ellinger A, Ranftler C, Gehre C, Brakebusch C, Pavelka M, Stradal T, Rottner K (2017). FMNL2 and -3 regulate Golgi architecture and anterograde transport downstream of Cdc42. Sci Rep 7, 9791.
Katoh M and Katoh M (2003). Identification and characterization of human FMNL1, FMNL2 and FMNL3 genes in silico. Int J Oncol 22, 1161-1168.
Liu W, Sato A, Khadka D, Bharti R, Diaz H, Runnels LW, Habas R (2008). Mechanism of activation of the formin protein DAAM1. Proc Natl Acad Sci USA 105, 210-215.
Loomis PA, Kelly AE, Zheng L, Changyaleket B, Sekerková G, Mugnaini E, Ferreira A, Mullins RD, Barltes JR (2006). Targeted wild-type and jerker espins reveal a novel, WH2-domain-dependent way to make actin bundles in cells. J Cell Sci 119, 1655-1665.
Lu J, Meng W, Poy F, Maiti S, Goode BL, Eck MJ (2007). Structure of the FH2 domain of DAAM1: Implications for formin regulation of actin assembly. J Mol Biol 369, 1258-1269.
Millard TH, Dawson J, Machesky LM (2007). Characterization of IRTKS, a novel IRSp53/MIM family actin regulator with distinct filament bundling properties. J Cell Sci 120, 1663-1672.
Miller MR, Miller EW, Blystone SD (2017). Non-canonical activity of the podosomal formin FMNL1ɣ supports immune cell migration. J Cell Sci 130, 1730-1739.
Otomo T, Otomo C, Tomchick DR, Machius M, Rosen MK (2005a). Structural basis of Rho GTPase-mediated activation of the formin mDia1. Mol Cell 18, 273-281.
Otomo T, Tomchick DR, Otomo C, Panchal SC, Machius M, Rosen MK (2005b). Structural basis of actin filament nucleation and processive capping by a formin homology 2 domain. Nature 433, 488-494.
Shimada A, Nyitrai M, Vetter IR, Kühlmann D, Bugyi B, Narumiya S, Geeves MA, Wittinghofer A (2004). The core FH2 domain of diaphanous-related formins is an elongated actin binding protein that inhibits polymerization. Mol Cell 13, 511-522.
163
Silkworth WT, KunesKL, Nickel GC, Phillips ML, Quinlan ME, Vizcarra CL (2018). The neuron-specific formin Delphilin nucleates nonmuscle actin but does not enhance elongation. Mol Biol Cell 29, 610-621.
Stastna J, Pan X, Wang H, Kollmannsperger A, Kutscheidt S, Lohmann V, Grosee R, Fackler OT (2012). Differing and isoform-specific roles for the formin DIAPH3 in plasma membrane blebbing and filopodia formation. Cell Res 22, 728-745.
Taniguchi K, Takeya R, Suetsugu S, Kan-O M, Narusawa M, Shiose A, Tominaga R, Sumimoto H (2009). Mammalian formin FHOD3 regulates actin assembly and sarcomere organization in striated muscles. J Biol Chem 284, 29873-29881.
Thompson ME, Heimsath EG, Gauvin TJ, Higgs HN, Kull FJ (2013). FMNL3 FH2-actin structure gives insight into formin-mediated actin nucleation and elongation. Nat Struct Mol Biol 20, 111-118.
Vaillant DC, Copeland SJ, Davis C, Thurston SF, Abdennur N, Copeland JW (2008). Interaction of the N- and C-terminal autoregulatory domains of FRL2 does not inhibit FRL2 activity. J Biol Chem 283, 33750-33762.
Xu Y, Moseley JB, Sagot I, Poy F, Pellman D, Goode BL, Eck MJ (2004). Crystal structures of a formin homology-2 domain reveal a tethered dimer architecture. Cell 116, 711-723.
164
Chapter 4
The carboxy-terminus of the formin FMNL1ɣ bundles actin to potentiate adenocarcinoma migration
This work is currently in revision to the Journal of Cellular Biochemistry
Submitted: May 30, 2018
Miller EW and Blystone SD (2018). The carboxy-terminus of FMNL1ɣ bundles actin to potentiate adenocarcinoma migration. J Cell Biochem, in revision.
165
Abstract
The formin family of proteins contributes to spatiotemporal control of actin cytoskeletal rearrangements during motile cell activities. The FMNL subfamily exhibits multiple mechanisms of linear actin filament formation and organization. Here we report novel actin-modifying functions of FMNL1 in breast adenocarcinoma migration models.
FMNL1 is required for efficient cell migration and its three isoforms exhibit distinct localization. Suppression of FMNL1 protein expression results in a significant impairment of cell adhesion, migration, and invasion. Overexpression of FMNL1ɣ, but not FMNL1β or FMNL1α, enhances cell adhesion independent of the FH2 domain and
FMNL1ɣ rescues migration in cells depleted of all three endogenous isoforms. While
FMNL1ɣ inhibits actin assembly in vitro, it facilitates bundling of filamentous actin independent of the FH2 domain. The unique interactions of FMNL1ɣ with filamentous actin provide a new understanding of formin domain functions and its effect on motility of diverse cell types suggests a broader role than previously realized.
Introduction Developmental, homeostatic and pathologic movement of cells requires a dynamic actin cytoskeleton. The organization of actin fibers in these processes takes numerable forms and is guided directly and indirectly by many proteins. Formins have been implicated in cell motility and exhibit direct regulation of actin dynamic events such as nucleation, elongation, capping, bundling and severing (Pruyne et al., 2002, Zigmond et al., 2003, Harris et al., 2004, Moseley et al., 2004, Chhabra and Higgs, 2006, Harris et al., 2006, Kovar, 2006, Schonichen and Geyer, 2010, Bravo-Cordero et al., 2012, Fife et al., 2014, Vizcarra et al., 2014). Recently we described a role for the formin FMNL1 in macrophage migration using a murine deletion model (Miller et al., 2017). In macrophages, FMNL1 associates with actin in the podosome, an adhesion structure with
166 invasive capabilities (Mersich et al., 2010, Miller et al., 2017). The loss of FMNL1 severely handicapped macrophage migration and decreased podosome stability, but could be rescued selectively by the gamma isoform of FMNL1 (Miller et al., 2017). In this study, we sought to understand the mechanism of action of FMNL1ɣ in regulating cell migration and determine the extent of these effects in other cell types expressing this formin.
The 15 human formins are divided into seven subfamilies: diaphanous-related formins (Dia), disheveled-associated activators of morphogenesis (DAAM), formin-like proteins (FMNL), formin homology domain-containing proteins (FHOD), inverted formins
(INF), Delphilin, and original formins (FMN). The Dia, DAAM, FMNL, and FHOD subfamilies all bear a similar domain architecture, illustrated in Figure 1A (Schonichen and Geyer, 2010, Krainer et al., 2013). These homodimeric proteins dimerize in a head- to-tail fashion at the highly-conserved, ~400 amino acid residue formin homology 2
(FH2) domain that is canonically responsible for promoting barbed end (+ end) elongation of actin filaments (F-actin) (Pruyne et al., 2002, Moseley et al., 2004, Xu et al., 2004, Harris and Higgs, 2006, Goode and Eck, 2007, Paul and Pollard, 2009). The diaphanous auto-regulatory domain (DAD) is located C-terminally to the FH2 domain and interacts with armadillo repeat regions found in the formin homology 3 (FH3) domain, located N-terminally of the FH1 domain. This interaction, which results in autoinhibition of the formin, can be alleviated by binding of Rho-family GTPases to the
GTPase-binding domain (GBD), located N-terminally to the FH3 domain. In vitro, Rho- family GTPase binding enhances FH2-dependent actin elongation, likely through unfolding of the protein into an open conformation (Watanabe et al., 1999, Otomo et al.,
2005, Nezami et al., 2006, Lammers et al., 2008).
Members of the FMNL subfamily (FMNL1, FMNL2, FMNL3) have been shown to regulate actin dynamics by nucleation, elongation, capping, severing, and/or bundling
167
(Harris et al., 2004, Harris and Higgs, 2006, Harris et al., 2006, Goode and Eck, 2007,
Esue et al., 2008, Vaillant et al., 2008, Block et al., 2012, Kage et al., 2017,B). FMNL1 is of particular interest as this formin, in vitro, exhibits each of these actin filament modifying actions (Harris et al., 2004, Harris et al., 2006, Goode and Eck, 2007, Esue et al., 2008). Three alternative splice isoforms of FMNL1 have been identified, two of which diverge following the C-terminal DAD region: FMNL1α, FMNL1β, and FMNL1ɣ.
While FMNL1α and FMNL1ɣ share a common end sequence of 30 amino acids,
FMNL1β has a unique sequence (Figure 1A). A 58 amino acid residue retention following the splice site T1069 in FMNL1ɣ differentiates it from FMNL1α (Yayoshi-
Yamamoto et al., 2000, Katoh and Katoh, 2003, Han et al., 2009, Han et al., 2013).
This subfamily is reported to interact with several Rho-family GTPases, including
RhoA, RhoC, Rac1, and Cdc42 (Yayoshi-Yamamoto et al., 2000, Seth et al., 2006,
Gomez et al., 2007, Kitzing et al., 2010, Mersich et al., 2010, Vega et al., 2011, Block et al., 2012, Favaro et al., 2013, Wang et al., 2015, Kage et al., 2017,A, Miller et al., 2017).
We have reported that expression of constitutively active or kinase dead forms of
GTPases had no effect on FMNL1 localization to podosomes (Mersich et al., 2010). No changes in GTPase levels were detected upon genetic deletion of FMNL1 in murine macrophages (Miller et al., 2017). It has also been suggested that FMNL1ɣ may not be subject to GTPase conformational regulation. FMNL1α and FMNL1β include the nonpolar amino acid residues isoleucine and leucine, respectively, at position 1071, while FMNL1ɣ specifies the positively-charged, polar amino acid residue lysine at this position. Based on the hydrophobic interactions involved with the binding of the DID and
DAD in Dia1, this lysine residue in FMNL1 could be significant in releasing the formin from its autoinhibited state (Nezami et al., 2006, Lammers et al., 2008, Han et al., 2009).
Reports have demonstrated that FMNL3 is not subject to DID-DAD autoregulation
(Vaillant et al., 2008). Given the specificity of the rescue of macrophage migration and
168 localization of FMNL1 to podosomes only by the gamma isoform, these observations suggest that the responsible activity lies within the alternatively spliced domain and may not be subject to regulation by DID-DAD interactions.
The beta isoform of murine FMNL1 does not facilitate actin assembly through elongation, exhibiting an inhibitory effect unperturbed by the presence of profilin that is attributed to barbed end accumulation (Harris et al., 2004). FMNL1β has been shown to bundle actin filaments, using side-binding activity that is in competition with barbed-end capping. Alleviation of processivity through FH2 mutation suggested that barbed-end binding and bundling were mutually exclusive yet both dependent upon FH2 domains
(Harris et al., 2006). Murine FMNL1 exhibits two isoforms, corresponding to human
FMNL1 alpha and beta. A third isoform, gamma, is present in humans. Interestingly, only the gamma isoform of FMNL1 rescues macrophage migration in FMNL1-null mice, but the effects of FMNL1ɣ on elongation and bundling of actin are previously undescribed (Miller et al., 2017).
FMNL1ɣ sequence analysis reveals potential tandem, overlapping Wiskott-
Aldrich syndrome homology 2 (WH2) domains within its unique 58 amino acid residues
(Figure 1A). In other proteins, these highly conserved WH2 motifs bind actin and facilitate its nucleation and assembly, however, overlapping WH2 motifs have not been reported or previously described (Quinlan et al., 2005, Chhabra and Higgs, 2006,
Chereau et al., 2005, Haglund et al., 2010, Gould et al., 2011, Gaucher et al., 2012,
Heimsath and Higgs, 2012).
Here we examined the interactions of FMNL1ɣ with actin, in comparison with the alpha and beta isoforms also found in human cells, to understand its unique contribution to macrophage migration and podosome stability. As FMNL1ɣ is expressed in other cells, we examined its requirement for migration in the well-characterized metastatic breast cancer line, MDA-MB-231. We find that similar to reports of murine FMNL1β,
169 human FMNL1ɣ suppresses actin assembly, but in a manner seemingly independent of barbed-end binding. We report that all three isoforms of FMNL1 exhibit actin bundling capability. Notably, bundling is substantially more efficient by the gamma isoform of
FMNL1. Suppression of FMNL1 expression inhibits cancer cell migration and invasion, but is rescued by the gamma isoform. Finally, all three isoforms of FMNL1 are expressed in MDA-MB-231 cells and each displays unique localization within cells.
FMNL1ɣ appears near the leading edge of migrating cells, slightly inside the lamellipodia. We propose that FMNL1ɣ bundling of actin contributes to actin structures involved in the migration of diverse cell types. These activities are likely due to unique sequences in its spliced region which bear homology to WH2 motifs.
Materials and Methods
Cells, Antibodies, and Reagents
MDA-MB-231 cells were purchased from ATCC (Manassas, VA). Rabbit polyclonal α-FMNL1 primary antibody (Novus Biologicals, Littleton, CO) was used at a concentration of 0.05µg/mL for both Western blot analysis and immunofluorescence.
This antibody is targeted against an internal region of FMNL1, and, as a result, is not isoform-specific. Goat polyclonal α-Transaldolase primary antibody (Santa Cruz
Biotechnology, Dallas, TX) was used to detect controls for Western blot analysis at a concentration of 0.2µg/mL. Goat polyclonal α-Dia1 and goat polyclonal α-FHOD1 antibodies (both from Santa Cruz Biotechnology) were used for Western blotting at a concentration of 2.0µg/mL. Rabbit polyclonal α-Dia2 primary antibody (Novus
Biologicals) was used for Western blotting at a concentration of 1.0µg/mL. Lysis buffer was composed of 50mM Tris, 5mM MgCl2, 150mM NaCl, and 1% Triton X-100 (all from
Thermo Fisher Scientific, Waltham, MA) supplemented with 1mM phenymethylsulfonyl fluoride (PMSF), 10µg/mL aprotinin, and 10µg/mL leupeptin (all from Sigma-Aldrich, St.
170
Louis, MO). All oligonucleotides were synthesized and purchased from Thermo Fisher
Scientific and all restriction endonucleases were purchased from New England BioLabs
(Ipswich, MA).
Cloning and Plasmid Generation
All FMNL1 alternative splice isoforms and mutants were generated from primary human macrophage cDNA as was previously described (Miller et al., 2017). The truncation mutant GFP-FMNL1Δ was cloned via PCR using the forward primer 5’-
GCACTGAAACCCAGCCAGATCACC-3’ and the reverse primer 5’-
CCGGCGGCCGCTATGTGATGATGTCTTCAATGG-3’. A stop codon and a NotI site were introduced at the C-terminal splice site T1069 via PCR mutagenesis. This NotI site was used in conjunction with a BsiWI site for transfer into a previously generated pEGFP vector (Mersich et al., 2010). FMNL1Δ was then transferred to pLVX-AcGFP1-C1
(Clontech, Mountain View, CA) via an N-terminal XbaI site and a C-terminal BamHI site.
GFP-FMNL1α was generated in the same manner but using the reverse primer 5’-
CGCAGATCGCGGCCGCTAGAGGGGCATCTCTTCTCC-3’. GFP-FMNL1β and GFP-
FMNL1ɣ were cloned as previously described. GFP-FMNL1ɣFH2ᴓ was cloned as previously described but under the name GFP-FMNL1ɣABM (Miller et al., 2017). GFP-
FMNL1ɣR was generated via PCR mutagenesis and cloned into the original pEGFP vector. The FMNL1 siRNA s2227 (Thermo Fisher Scientific) targeting region 5’-
AGGTCATTGCTGAGAAGTA-3’ beginning at K884 was mutated as follows: 5’-
AGGTGATAGCAGAGAAGTA-3’. Purified CYK1-CT was a generous gift from Dr. David
Pruyne. The GST fusion protein constructs pGEX-FMNL1ɣ-CT and pGEX-
FMNL1ɣFH2ᴓ-CT (originally named pGEX-FMNL1ɣABM) were generated as previously described (Miller et al., 2017). The GST fusion protein constructs encoding for FMNL1Δ-
CT, FMNL1α-CT, and FMNL1β-CT were cloned into pGEX-4T-1 (GE Healthcare Life
171
Sciences, Pittsburgh, PA) by introducing a BglII site via PCR mutagenesis into the multiple cloning site (MCS) and, in conjunction with a NotI site, transferring the FMNL1-
CT regions into this modified vector. All FMNL1 GST fusion constructs begin at the FH1 domain (A536) and end at their respective carboxy-termini.
Generation of Stable Cell Lines
MDA-MB-231 cells were infected with lentivirus coding for all full-length, GFP- tagged FMNL1 alternative splice isoforms and mutants as was previously described for bone marrow-derived macrophages (BMDMs) (Miller and Blystone, 2015,B). The only exception to this was cells expressing GFP-FMNL1ɣR. This construct was transiently transfected via lipofection with jetPEI (Polyplus-transfection, Illkirch, France) according to manufacturer’s protocol. The lentiviral expression vector pLVX-AcGFP1-C1 contains a puromycin resistance gene which allows for positive antibiotic selection of cells expressing the GFP-tagged, full-length FMNL1 alternative splice isoforms and mutants.
72hrs. post-infection, puromycin (Thermo Fisher Scientific) was added at a concentration of 0.10µg/mL. Over the course of 4-6 weeks, puromycin concentration was gradually increased with each cell passage to 2.0µg/mL. The pEGFP vector that FMNL1ɣR was cloned into contains a zeocin resistance gene which allows for positive antibiotic selection of cells expressing this mutant. 48hrs. post-transfection, zeocin (Thermo
Fisher Scientific) was added at a concentration on 1.0µg/mL. Over an eight week time- course, zeocin concentration was gradually increased with each cell passage, ending at
400µg/mL. Cells were then maintained in complete media with their respective positive selection antibiotics. In order to determine GFP expression levels, 100,000 MDA-MB-
231 cells expressing full-length, GFP-tagged FMNL1 alternative splice isoforms and mutants were lysed with 25µL of 2M KOH (Sigma-Aldrich) and transferred to a 96-well plate (Corning, Inc.). Immediately following lysis, lysate fluorescence was measured at
172
Ex. 488nm and Em. 509nm on a SpectraMax Gemini EM microplate reader (Molecular
Devices, Sunnyvale, CA). Lysates from cells expressing GFP alone were set at 100% and the fluorescence measurements from all other cells expressing the different GFP constructs were calculated as a percentage of GFP alone. Experiments were performed in quadruplicate and repeated three times to quantify expression levels.
Western Blot Analysis
MDA-MB-231 cells were lysed with lysis buffer overnight (o/n) at 4°C. Lysate protein concentration was determined using a Pierce BCA Protein Assay Kit (Thermo
Fisher Scientific) according to manufacturer’s protocol. Equal amounts of lysate were loaded onto 9% SDS-PAGE polyacrylamide gels and resolved via electrophoresis.
Separated protein was transferred to a PVDF membrane (Millipore, Burlington, MA), stained with Coomassie (Sigma-Aldrich), and blocked with 3% bovine serum albumin
(BSA) (Sigma-Aldrich). Transfer was probed for FMNL1 and transaldolase as a loading control. Protein was visualized using HRP-conjugated secondary antibodies (Jackson
ImmunoResearch Laboratories, West Grove, PA) and enhanced chemiluminescence
(ECL) (GE Healthcare Life Sciences). Densitometry analysis was performed in ImageJ.
Invasion and Transmigration Assays
MDA-MB-231 cells were transduced via lipofection with INTERFERin (Polyplus- transfection) in conjunction with 400nM control siRNA, 400nM FMNL1 siRNA, 400nM
Dia1 siRNA, 400nM Dia2 siRNA (all from Thermo Fisher Scientific) or 400nM FHOD1 siRNA (Santa Cruz Biotechnology) for 48hrs. Pharmacological FH2-domain inhibition of all formins in cells was accomplished using 30µM SMIFH2 (Sigma-Aldrich) for 1hr.
DMSO (Sigma-Aldrich) was used as a vehicle control. Adherent cells were dissociated with trypsin and all cells were counted via staining with 0.04% trypan blue (Sigma-
173
Aldrich) and a hemocytometer (Reichert Technologies, Depew, NY). 20,000 cells from each treatment were transferred to rehydrated, 8.0µm porous Matrigel invasion chambers (Corning, Inc., Corning, NY) in serum-free media in 24-well plates (Corning,
Inc.). Cell invasion was induced using 10% fetal bovine serum (FBS) (Lampire
Biological Laboratories, Ottsville, PA) as a chemoattractant and permitted for 24hrs.
Non-invading cells were removed from the membrane and cells that had successfully invaded were fixed with ice-cold 3.7% formaldehyde (Thermo Fisher Scientific) at 4°C.
Membranes were allowed to dry for 30mins. and stained using a differential-staining kit
(IMEB, San Marcos, CA). Membranes were removed with a scalpel and inverted on a drop of immersion oil (Cargille Laboratories, Cedar Grove, NJ) on a 25 x 75 x 1mm frosted, glass slide (Globe Scientific, Paramus, NJ). Another drop of immersion oil was added to the top of the membrane, after which it was covered with a 12mm No. 1 coverslip (Thermo Fisher Scientific). Cells were counted via light microscopy with a
Leica Galen III microscope affixed with a 40x objective (Leica Microsystems, Wetzlar,
Germany). Five fields from each coverslip were counted and all experiments were performed in triplicate. Transmigration assays were performed as described for invasion assays except uncoated 8.0µm porous Transwell inserts (Corning, Inc.) were used as opposed to Matrigel invasion chambers and only 5,000 cells from each treatment were utilized. Western blot analysis was performed in conjunction with invasion assays to confirm successful protein expression silencing. Two-sided Student’s t-test was used for statistical analysis.
RT-PCR and Fusion Protein Purification
MDA-MB-231 whole-cell RNA was extracted and purified with the RNeasy Mini
Kit (Qiagen, Hilden, Germany). MDA-MB-231 cDNA was generated from this purified
RNA using the SuperScript III First-Strand Synthesis System (Thermo Fisher Scientific).
174
All final PCR products were amplified with the same forward primer targeted at an internal region of FMNL1 before the splice site T1069: 5’-
GCACTGAAACCCAGCCAGATCACC-3’. The reverse primer for FMNL1Δ was 5’-
CCGGCGGCCGCTATGTGATGATGTCTTCAATGG-3’. As FMNL1α (Accession:
NM_005892) and FMNL1ɣ (Accession: FJ534522) share the same common end sequence, the same reverse primer sequence was used for amplification of both products: 5’-CGCAGATCGCGGCCGCTAGAGGGGCATCTCTTCTCC-3’. The reverse primer used for amplification of FMNL1β (Accession: BC001710) was 5’-
CCTGCAGAAGCGGCCGCTACAGCGAGAGGTCGG -3’. All PCR reactions were performed in a PTC-100 Programmable Thermal Controller (MJ Research Inc.,
Waltham, MA). Final amplification products were run out on a 1% agarose gel (Thermo
Fisher Scientific) and visualized with ethidium bromide (Sigma-Aldrich) and a 312nm transilluminator (Thermo Fisher Scientific). Images were captured with a Polaroid
GelCam (Polaroid Corporation, Minnetonka, MN) on FP-3000B FujiFilm black and white prints (FujiFilm, Minato, Tokyo, Japan).
All pGEX-FMNL1-CT alternative splice isoform vectors were transformed into
Rosetta competent cells (Millipore) using standard techniques. Transformed Escherichia coli were grown to OD600 ~ 0.6 – 0.8 at 37°C and then incubated for 1hr. at 16°C. Fusion protein expression was induced with 0.2mM Isopropyl β-D-1-thiogalactopyranoside
(IPTG) (Thermo Fisher Scientific, Waltham, MA) o/n at 16°C. E. coli were lysed, sonicated, and centrifuged. Fusion protein was isolated from lysate with glutathione
Sepharose 4B (GE Healthcare Life Sciences). Fusion protein-bound Sepharose was washed and protein was cleaved at the thrombin cut site with a Thrombin Cleavage
Capture Kit (Millipore). Purified protein was kept on ice and concentration was determined with a Pierce BCA Protein Assay Kit. Protein was used immediately and not stored for future use.
175
Fluorescence Microscopy, Cell Morphology, and Motility Analysis
MDA-MB-231 cells were permitted to adhere overnight to 12mm No. 1 coverslips coated with 100µg/mL poly-L-lysine (Sigma-Aldrich) and 100µg/mL type I collagen
(Corning Inc.) at 37°C and 5% CO2. Coverslips were washed 3x with PBS (Thermo
Fisher Scientific) followed by simultaneous fixation and permeabilization with 3.7% formaldehyde and 1% Triton X-100, respectively, for 5mins. at room temperature (RT).
Coverslips were quenched with 0.4M glycine (Thermo Fisher Scientific) for 20mins. and blocked with 3% BSA for 1hr at RT. Coverslips were then washed 3x with 0.05%
Tween-20 (Thermo Fisher Scientific) and incubated with α-FMNL1 primary antibody for
2hrs. at 37°C. followed by incubation for 1hr. at 37°C with 2µg/mL 488-conjugated goat
α-rabbit IgG DyLight (Thermo Fisher Scientific) and 14.0nM rhodamine phalloidin
(Cytoskeleton, Inc., Denver, CO) to visualize endogenous FMNL1 expression and F- actin, respectively. Coverslips were then washed 2x with 0.05% Tween-20, 1x with dH2O, and affixed to 25 x 75 x 1mm frosted, glass slides with 8µL anti-fade mounting media consisting of 0.5M Tris, pH 8.5, 10.5% Poly(vinyl alcohol) (Sigma-Aldrich), 26.4% glycerol (Thermo Fisher Scientific), and 2.5% 1,4-Diazabicyclo[2.2.2]octane (DABCO)
(Sigma-Aldrich). Microscopic analysis was performed using a Nikon Eclipse E800 fluorescence microscope (Nikon, Melville, NY) with a 60x objective and images were captured with a Hamamatsu ORCA-ER digital camera (Bridgewater, NJ). NIS-Elements
(Nikon) was used for image editing and analysis. Microscopy on GFP-expressing, stable cells was performed as described above without any antibody staining. Cell perimeter and surface area were measured with NIS-Elements. 20 cells were analyzed per experiment. Two-sided Student’s t-test was used for statistical analysis. For cell motility analysis, FMNL1 expression was silenced in MDA-MB-231 cells, as was previously described, in MatTek No. 1.5 glass bottom culture dishes (MatTek Corporation, Ashland,
MA). 48hrs. post-siRNA transduction, media was replaced and dishes with adherent
176 cells remaining were transferred to a bio-chamber set at 37°C and 5% CO2, a component of the Leica AF6000 Deconvolution System (Leica, Wetzlar, Germany).
Using deconvolution microscopy, eight individual cells per experiment were imaged every 10mins. for 16hrs. These cells were tracked and analyzed for both velocity and average accumulated distance traveled using the tracking plugin and chemotaxis tool in
ImageJ. Rose plot mapping was performed in Microsoft Excel. Two-sided Student’s t- test was used for statistical analysis.
Adhesion Assays
MDA-MB-231 cells and cells expressing GFP-FMNL1ɣR were depleted of endogenous FMNL1 expression via siRNA transduction as was previously described.
50,000 cells were permitted to adhere to the bottom of a 96-well plate for 1hr. at 37°C.
Media was aspirated and adherent cells were washed 3x with PBS. Cells were fixed with 3.7% formaldehyde for 10mins. at RT and stained with 0.05% crystal violet (Sigma-
Aldrich) for 30mins. at RT. Cells were washed 3x with PBS and incubated with 100µL of
100% CH3OH (Thermo Fisher Scientific) for 1min. while mixing. Absorbance was measured at 550nm with an EMax Precision Microplate Reader (Molecular Devices,
Sunnyvale, CA). Untreated cell adhesion was set at 100% and all other conditions are reported as a percentage of this maximum. Stable cell line adhesion analysis was performed in a similar manner without siRNA transduction. WT cells were set at 100% and all other cell lines were measured as a percentage of this maximum. Experiments for all adhesion assays were performed in quadruplicate three separate times. Student’s t-test was used for statistical analysis.
177
Actin Biochemistry and Microscopy
Actin was purified from rabbit skeletal muscle acetone powder (Pel-Freez
Biologicals, Rogers, AR) as was previously described (Spudich and Watt, 1971).
Pyrene-labeled actin (Cytoskeleton, Inc.) was combined with unlabeled actin at a 1:5 ratio, resulting in 20% pyrene-labeled actin. Actin was dialyzed against G-buffer composed of 2.0mM Tris-HCl, pH 8.0 (Thermo Fisher Scientific), 0.1mM CaCl2 (Sigma-
Aldrich), 0.2mM ATP (Sigma-Aldrich), and 0.1mM DTT (Millipore) at 4°C for 24hrs. Actin was transferred to 7x20mm centrifuge tubes (Beckman Coulter, Brea, CA) and centrifuged in a TLA-100 rotor (Beckman Coulter) at 100,000RPM to isolate G-actin from
F-actin. G-actin was removed and concentration was determined with a Pierce BCA
Protein Assay Kit. 3.0µM G-actin was primed with 10x EGTA/MgCl2 composed of 10mM
EGTA (Sigma-Aldrich) and 1mM MgCl2 for 2mins. on ice. G-actin was combined with either 500nM purified FMNL1-CT fusion protein or 100nM CYK1-CT and polymerization was induced by the addition of 10x KMEI composed of 100mM imidazole, pH 7.0
(Sigma-Aldrich), 500mM KCl (Thermo Fisher Scientific), 10mM MgCl2, and 10mM EGTA.
Reactions were transferred to a 96-well plate and fluorescence was measured at Ex.
365nm and Em. 407nm on a SpectraMax Gemini EM Microplate Reader. Fluorescence measurements were taken every 10secs. for 400secs. Raw data was compiled in
Microsoft Excel and plotted as fluorescence (RFUs) vs. time (secs.). A least-squares regression analysis was performed and actin assembly rates in nmol/s were determined from the equation S* = (S x Mt) / (fmax – fmin) where S* is slope, S is raw slope, Mt is the concentration of available actin monomers, fmax is fluorescence of fully-polymerized actin, and fmin is fluorescence of unpolymerized actin (Pollard, 1986, Pollard and Cooper,
1986, Li and Higgs, 2003). For low-speed co-sedimentation assays, 10µM actin was purified as previously described, dialyzed against G-buffer for 24hrs., and centrifuged at
14,000 x g for 45mins., all at 4°C. F-actin generation was induced by the addition of 10x
178
KMEI and incubated at 4°C o/n. 10µM unlabeled phalloidin (Sigma-Aldrich) was added to the reaction for filament stabilization and incubated on ice for 2hrs. Stabilized F-actin was combined with 500nM purified FMNL1-CT fusion protein and incubated on ice for
1hr. The actin-protein reaction was centrifuged at 10,000 x g for 20mins. at 4°C.
Reactions were separated via SDS-PAGE and transferred to a PVDF membrane.
Coomassie staining was used to visualize protein. Actin distribution was determined via densitometry analysis using ImageJ. For visual analysis of actin filament bundling,
3.0µm F-actin was purified as previously described. 1.0µm rhodamine phalloidin was added to the reaction. Following a 2hr. incubation on ice, stabilized, rhodamine-stained,
F-actin was combined with 500nM purified FMNL1-CT fusion protein on ice for 1hr. 4µL of the reaction was transferred to 25 x 75 x 1mm frosted, glass slides coated with
100µg/mL poly-L-lysine. Samples were allowed to adsorb to the slide and then covered with 12mm No.1 coverslips coated with 100µg/mL poly-L-lysine. Microscopy was immediately performed using a Nikon Eclipse E800 fluorescence microscope affixed with a 100x objective. Images were captured with a Hamamatsu ORCA-ER digital camera.
NIS-Elements was used for image editing and analysis.
Results
FMNL1ɣ inhibits actin assembly independent of FH2 domain function.
Three human isoforms of FMNL1 differ only in their carboxy-terminus, however the gamma isoform singularly rescues knockout macrophage’s ability to migrate using normal podosomes (Figure 1A) (Miller et al., 2017). In order to identify actin regulatory functions of FMNL1ɣ that could explain its isoform-specific role in macrophage migration, we used pyrene-fluorescence actin assembly assays. By combining pyrene-labeled, monomeric actin with purified FMNL1ɣ fusion proteins (Figure 1A, B), we measured the effects of FMNL1ɣ on actin filament assembly rates by recording fluorescence over time.
179
We utilized the Caenorhabditis elegans formin CYK-1, known for its potent actin assembly activity, as a positive control (Neidt et al., 2008). This fusion protein encodes for the FH1-FH2-COOH region of CYK-1 and is named CYK1-CT. We found that the presence of similarly constructed FMNL1ɣ-CT significantly reduced actin assembly rates when compared to actin alone, similar to reports of murine FMNL1β (Figure 1C, D, G)
(Harris et al., 2004, Harris et al., 2006). Under these same conditions, CYK1-CT exerted an increase in the assembly of actin (Figure 1C, D, G).
Studies analyzing the FH2 domain of the Saccharomyces cerevisiae formin
Bni1p and its interaction with actin filament barbed ends have shown that two amino acid residues of this domain are essential for this action: I1431 and K1601 (Xu et al., 2004).
In FMNL1, these two conserved amino acid residues, I720 and K871, have been shown to be essential for FMNL1 to remain associated with the actin filament barbed end, as well as maintain the structural integrity of the Golgi complex (Harris et al., 2006, Colon-
Franco et al., 2011). Interestingly, we have previously demonstrated that the inhibition of barbed end binding of actin filaments through these FH2 domain mutations does not impede the function or localization of FMNL1ɣ with macrophage podosomal actin, nor its ability to rescue migration in FMNL1-null macrophages (Miller et al., 2017). To test whether FH2 domain engagement of barbed ends contributed to the inhibition of actin assembly by FMNL1ɣ, we prepared a fusion protein incorporating the disabling FH2 mutations (Figure 1A, B). FMNL1ɣ-CT and FMNL1ɣFH2ᴓ-CT both exhibited similar inhibition of actin assembly rates, 0.48 nmol/s and 0.34 nmol/s, respectively (Figure 1E,
F, H). These rates were significantly reduced compared to the control actin assembly rate of 1.84 nmol/s (Figure 1G). The similarity of behavior between FMNL1ɣ-CT and
FMNL1ɣFH2ᴓ-CT suggests that the actin filament barbed end binding activity typically ascribed to the FH2 domain does not contribute to the suppression of actin assembly by
FMNL1ɣ seen in this assay. Together with the specific rescue of knockout macrophage
180 migration by FMNL1ɣ, this implicates an actin interaction specific to the gamma isoform in these processes.
The C-terminus of FMNL1ɣ efficiently bundles actin filaments in a dose-dependent manner.
Based on our observation that FMNL1ɣ-CT and FMNL1ɣFH2ᴓ-CT both reduced actin assembly rates by 74% and 81% respectively, we questioned what actin filament regulatory function FMNL1ɣ was performing. The reduction of assembly rates led us to investigate whether FMNL1ɣ interacts with existing actin filaments using a low-speed co- sedimentation assay for actin bundling. We generated additional constructs which encode for GST-tagged fusion proteins of FMNL1Δ, truncated at the beginning of the alternative splicing, FMNL1α, and FMNL1β, all beginning at the FH1 domain and continuing through to the C-terminus. Using these fusion proteins, purified as in Figure
1B, in conjunction with FMNL1ɣ-CT and FMNL1ɣFH2ᴓ-CT, allowed us to assess the role of the C-terminal region of FMNL1 in actin filament interactions.
We observed a significant increase in the amount of actin in the pellet fraction when actin filaments were allowed to interact with all FMNL1-CT isoforms (Figure 2A).
FMNL1Δ-CT displayed a similar pellet fraction to that of the control sample, verifying that the C-terminus of FMNL1 is required for FMNL1-mediated actin filament interaction
(Figure 2A). When compared to the control fractions or the FMNL1Δ-CT fractions,
FMNL1ɣ-CT has a significantly higher amount of actin distributed in the pellet fraction versus the supernatant fraction (Figure 2B). We also determined the amount of actin bundling specific to each FMNL1 isoform. We observed a 50% increase in pelleted actin by FMNL1ɣ, when compared with the alpha or beta isoforms (Figure 2C).
FMNL1ɣFH2ᴓ-CT was able to efficiently bundle actin equally well, indicating that FH2 domain interaction with barbed ends is dispensable for this process (Figure 2C). We
181 also utilized a dose-response assay to determine whether FMNL1ɣ-CT bundling of actin exhibits a saturation point. We found that FMNL1ɣ-CT concentrations exceeding 300nM showed no further increase in pelleted actin filaments, indicating a saturable event
(Figure 2D). The bundling activity of FMNL1ɣ-CT was also visually confirmed with fluorescence microscopy. Post-bundled actin filaments were stained with rhodamine phalloidin and transferred to coverslips for analysis. Actin filaments bundled by
FMNL1ɣ-CT were aggregated together forming a distinct morphology in contrast to control filaments which did not demonstrate this phenotype (Figure 2E). From this data, we conclude that the 58 amino acids distinguishing FMNL1ɣ have an added ability to bundle actin filaments, exceeding that of other isoforms, which may be responsible for its requirement in macrophage migration.
FMNL1 contributes to breast cancer migration.
FMNL1 is expressed in multiple cell types of all embryonic origin (Krainer et al.,
2013). If actin bundling served a mechanistic role in macrophage migration, we hypothesized that it should have a similar effect upon the migration of other cells expressing this formin and tested this in MDA-MB-231 cells, a well characterized migratory epithelial line. We initially depleted cells of FMNL1 expression via siRNA transduction and subjected these cells to a random motility assay to measure migration as described in Materials and Methods. Three siRNA oligonucleotides were successful in reducing FMNL1 expression in these cells and s2228 was used for remaining studies due to its superior efficiency (Figure 3E). Following time-lapse deconvolution microscopy with single-cell tracking, rose plots were analyzed to determine the accumulated distance and velocity of individual cells (Figure 3A, B). Cells lacking
FMNL1 expression traveled an average accumulated distance of 269.54µm at an average velocity of 0.28µm/min., a significant reduction compared to control cells which
182 traveled an average accumulated distance of 742.06µm at an average velocity of
0.78µm/min. (Figure 3C, D).
We employed indirect immunofluorescence microscopy to observe the cellular distribution of endogenous FMNL1 expression in MDA-MB-231 cells (Figure 3F), as well as to confirm reduction of FMNL1 expression following siRNA transduction (Figure 3G).
Localization patterns of endogenous FMNL1 expression were similar to those which previously reported FMNL1 dispersed throughout the cytoplasm, with some enrichment at the nucleus, the perinuclear region, and the cell periphery where the protein co- localized with F-actin (Figure 3G) (Gomez et al., 2007, Han et al., 2009, Colon-Franco et al., 2011). Cells transduced with FMNL1 siRNA displayed a lack of any FMNL1 endogenous protein expression (Figure 3G). We also observed that the number of
FMNL1-depleted cells appeared reduced when compared to cell cultures transduced with control siRNA. We have previously reported that FMNL1 is essential for proper adhesion of macrophages, and, as such, attributed this reduced cell number with an analogous adhesion defect in MDA-MB-231 cells (Miller and Blystone, 2015,A). These same cells were observed to be much smaller in size when compared to cells transduced with control siRNA (Figure 3F, G).
The three alternative splice isoforms of FMNL1 expressed in MDA-MB-231 cells display unique localization patterns.
The nondescript localization of endogenous FMNL1 led us to question whether any of the alternative splice isoforms of FMNL1 have more specific localization patterns.
In order to accomplish this, we infected MDA-MB-231 cells with lentivirus coding for
GFP-tagged, full-length FMNL1 alternative splice isoforms, as well as for a mutant truncated at the alternative splice site T1069 (FMNL1Δ), a mutant of FMNL1ɣ with a
183 three codon non-targeting siRNA site mutation (FMNL1ɣR) masking it from s2227, and
GFP alone, followed by antibiotic selection of stable populations.
The three alternative splice isoforms of FMNL1 each exhibited a discrete localization to different regions of MDA-MB-231 cells (Figure 4A), together which combined to mimic the localization patterning seen for endogenous FMNL1 (Figure 3F).
FMNL1Δ localized to the perinuclear region of the cell and the cell periphery, similar to endogenous FMNL1 expression and logical given FMNL1Δ lacks a C-terminal region following the splice site. FMNL1α and FMNL1β both localized predominantly to the perinuclear region of the cell, with alpha appearing in a compact region while beta was punctate and limited primarily to one side (Figure 4A). FMNL1ɣ localized to small regions of the cell periphery and near the lamella-lamellipodia interface (Figure 4A).
In order to confirm that all three alternative splice isoforms of FMNL1 were normally expressed in MDA-MB-231 cells, we used site-specific primers in conjunction with reverse-transcriptase PCR (RT-PCR) as described in Materials and Methods
(Figure 4B). Products representing all three isoforms were readily identified, however attempts to design quantitative primers that could provide relative expression levels were unsuccessful, as were the use of splice-specific antibodies (Colon-Franco et al., 2011).
These results confirmed expression of all three alternative splice isoforms of FMNL1 at the message level.
Exogenous protein expression was confirmed via Western blot analysis. Using antibody that targets a region N-terminal to the isoform splice site, we observed endogenous FMNL1 expression at ~120kDa and exogenous GFP-FMNL1 alternative splice isoform expression at ~27kDa higher (Figure 4C). In order to quantify GFP expression, we lysed cells expressing these constructs with saturated 2M KOH and measured fluorescence at 488nm (Figure 4E). MDA-MB-231 cells expressed FMNL1Δ at an average level of 20.8% when compared to GFP alone, the lowest level of
184 expression in this analysis (Figure 4E). FMNL1α and FMNL1β were expressed at an average of 78.03% and 82.15%, respectively, with FMNL1β expression being the highest of all cells expressing these constructs (Figure 4E). FMNL1ɣ, FMNL1ɣFH2ᴓ, and FMNL1ɣR, the non-targeted rescue construct, were expressed at a similar level of
~50% (Figure 4E). To confirm that FMNL1ɣR was not targeted by FMNL1 siRNA, we transduced cells expressing FMNL1ɣ and FMNL1ɣR with FMNL1 siRNA. Both endogenous and exogenous FMNL1 expression was significantly diminished in cells expressing FMNL1ɣ (Figure 4D). However, cells expressing FMNL1ɣR displayed a reduction in endogenous FMNL1, while exogenous FMNL1ɣR expression was unaffected, confirming the specificity of the non-targeting siRNA mutation (Figure 4D).
This rescue construct also exhibited localization identical to unmodified FMNL1ɣ (Figure
4A).
FMNL1ɣ is essential for the regulation of breast cancer cell morphology and adhesion.
We have previously demonstrated that FMNL1ɣ is crucial for proper macrophage adhesion and migration (Mersich et al., 2010, Miller and Blystone, 2015,A, Miller et al.,
2017). This led us to question whether this may be true for cancer cells as well. We observed that siRNA depletion of FMNL1 appeared to decrease MDA-MB-231 cell size and deplete cultures of adherent cells (Figure 2G). Cells expressing GFP-FMNL1ɣ and
GFP-FMNL1ɣR displayed a significantly larger perimeter and surface area, demonstrating that FMNL1ɣ, but not FMNL1α or FMNL1β, may play an important role in maintaining cell morphology (Figure 5A, B). Similar increased perimeter and area were seen in cells expressing FMNL1ɣ with an inactivated FH2 domain (Figure 5A, B). These results are consistent with our results showing specific behavior attributable to the alternatively spliced region defining FMNL1ɣ and its bundling of actin filaments.
185
To directly examine FMNL1ɣdependent adhesion, MDA-MB-231 cells were transduced with FMNL1-targeting siRNA and subjected to a cell adhesion assay. The number of adherent cells were significantly reduced upon FMNL1 expression depletion when compared with control cells (Figure 5C). However, cells expressing GFP-
FMNL1ɣR were protected from inhibition of adhesion by avoiding targeting by siRNA and actually exhibited an increase in the number of adherent cells when compared to control cells (Figure 5C). This striking observation led us to investigate whether cells expressing exogenous FMNL1 alternative splice isoforms exhibited any difference in adhesive ability. Cells expressing GFP-FMNL1α or GFP-FMNL1β did not display any difference in adhesion when compared to control cells (Figure 5D). However, cells expressing GFP-FMNL1ɣ showed a significant increase in adhesive abilities when compared to control cells (Figure 5D). This gain-of-function is especially interesting as we can conclude it is most certainly due to the amino acid residue composition in the C- terminal region of FMNL1. Taken together with our data demonstrating the unique localization patterning of each alternative splice isoform, we can further propose that these alternative splice isoforms have distinct functions within the cell that are ordained by their C-termini.
We next measured the transmigration and invasion abilities of cells expressing each of the GFP-FMNL1 mutants. Following siRNA-mediated depletion of endogenous
FMNL1 expression, cells expressing GFP, GFP-FMNL1Δ, GFP-FMNL1α, GFP-FMNL1β,
GFP-FMNL1ɣ, and GFP-FMNL1ɣFH2ᴓ exhibited a significant, ~50-60% reduction in transmigration through an uncoated porous filter barrier compared to controls (Figure
5E). Importantly, transmigration of cells expressing GFP-FMNL1ɣR did not differ from that of control cells (Figure 5E). This confirms that expression of non-targeted, exogenous FMNL1ɣ can rescue the transmigratory abilities of cells depleted of FMNL1 expression. Analysis of invasion through a Matrigel-coated porous filter barrier showed
186 similar results to the transmigration assays (Figure 5F). Invasion of WT MDA-MB-231 cells was significantly reduced to 37.6% and cells expressing GFP-FMNL1Δ, GFP-
FMNL1α, and GFP-FMNL1β were similarly affected following siRNA removal of endogenous FMNL1 expression. Invasion of cells expressing GFP-FMNL1ɣ and GFP-
FMNL1ɣFH2ᴓ were both significantly reduced when compared to control cells at 51.2% and 37.6%, respectively, but invasion by cells expressing GFP-FMNL1ɣR did not differ from control cells, confirming that expression of this non-targeted, exogenous FMNL1 alternative splice isoform can rescue the invasive abilities of FMNL1-depleted MDA-MB-
231 cells. These data, along with our own previous studies, reinforces that FMNL1ɣ is essential for efficient cell migration and invasion, an observation that is true for cell types as varied as adenocarcinoma and macrophages.
Regulation of invasion by FMNL1 is distinct among diaphanous formins.
Studies have implicated several formin family members in cancer metastasis
(Han et al., 2009, Lizarraga et al., 2009, Kitzing et al., 2010, Randall and Ehler, 2013,
Kage et al., 2017,B). We compared the relative contribution of these to MDA-MB-231 cell invasion to evaluate the potency of FMNL1ɣ. Reported quantitative real-time reverse transcriptase PCR (qRT-PCR) results indicated that the Dia, FHOD, and FMNL subfamily of formins were all expressed at a substantially higher level than other formin subfamily members in the human breast adenocarcinoma cell line MDA-MB-231 and were chosen for further study (Krainer et al., 2013). Expression of Dia1, Dia2, FHOD1, and FMNL1 in MDA-MB-231 cells was suppressed by siRNA-mediated protein depletion and verified by Western blot analysis (Figure 6A). Cells depleted of formin protein expression were challenged to cross Matrigel-coated, 8.0µm porous filter barriers as a measure of invasive capacity using fetal bovine serum (FBS) as a chemoattractant. We observed that cells with reduced expression levels of FMNL1 and Dia1, but not Dia2 and
187
FHOD1, demonstrated a significant reduction in successful Matrigel invasion when compared to cells transduced with control siRNA (Figure 6B). We also performed experiments evaluating the effect of protein expression silencing on two formins simultaneously. Knockdown of both FMNL1 and Dia1 simultaneously exhibited a cumulative effect, resulting in a significant reduction in the invasive abilities of these siRNA-transduced cells when compared to both knockdown of FMNL1 only or Dia1 only
(Figure 6C), suggesting that Dia1 and FMNL1 contribute to invasion through separate mechanisms. Upon co-depletion of FMNL1 and FHOD1, as well as FMNL1 and Dia2, there is no further decrease in invasion when compared to knockdown of FMNL1 only, confirming that FHOD1 and Dia2 may not be essential for cell invasion (Figure 6C).
Simultaneous knockdown of FMNL1 and Dia1 resulted in a similar level of invasion inhibition as that of SMIFH2, a pan-formin FH2 domain pharmacological inhibitor (Figure
6D). This compound has previously been shown to pharmacologically inhibit formin
FH2-mediated actin filament assembly in a concentration-dependent manner (Rizvi et al., 2009). We previously reported an insensitivity of FMNL1 to SMIFH2 in macrophage adhesion (Miller and Blystone, 2015,A). In studies presented here, we illustrate that
FMNL1ɣ modulates adhesion and migration of MDA-MB-231 cells independent of the
FH2 domain. These data would suggest that SMIFH2 is targeting the Dia1 contribution to the invasion of these cells.
To determine why the 58 amino acid splice in FMNL1ɣ has unique effects on actin bundling and cell migration, we subjected this region to structural analysis; two distinctive features are evident. First, underlined in Figure 1A by dotted and dashed lines respectively, are two sequences meeting the definition of a WH2 motif. While variant in length, this includes the presence of three aliphatic residues, two paired and a third trailing by 7-10 residues (Paunola et al., 2002, Chereau et al., 2005, Aguda et al.,
2006, Husson et al., 2011, Dominguez, 2016). In WASp (Wiskott-Aldrich Syndrome
188 protein) and WAVE2 (WASp Family Verprolin-Homologous Protein-2), these paired residues are leucines and sequence alignments between FMNL1ɣand WASp and
WAVE2 are shown with paired aliphatic residues in yellow and trailing residues in pink
(Figure 6E) (Modified from Vaillant et al., 2008 and Heimsath and Higgs, 2012). WH2 motifs typically occur within an amphiphilic helical structure that is also seen when the
FMNL1ɣ-specific spliced region is objectively predicted (Figure 6F). It should be noted that, in contrast with other reported tandem WH2 motifs, those of the FMNL1ɣ sequence are overlapping with both putative WH2 motifs appearing on the same helical face. A second identified sequence feature in FMNL1ɣ is the presence of five phenylalanines, alternating with unremarkable residues. The number of phenylalanine residues and their arrangement was found as unique in human sequence databases searches.
Interestingly, these residues also appear in a separate but adjoining predicted helix to that bearing the WH2 motif (Figure 1A, 6F). These unique features and arrangements may be responsible for the unique behaviors we describe for FMNL1ɣ and merit further investigation.
189
Figure 1
190
Figure 1: FMNL1ɣ and FMNL1ɣFH2ᴓ inhibit actin assembly. (A) Schematic representing domain structure of FMNL1, including isoform sequences, and GST-tagged
FMNL1 construct design. Lines emitting from the FH2 domain represent I720A and
K871A mutations and are labeled “ᴓ.” Following the splice site at T1069, amino acid residue sequences are listed through to the C-terminus. The 30 common amino acid residues shared by FMNL1α and FMNL1ɣ are in red. Black dotted line shows location of the paired leucines of a first putative WH2 domain and its trailing aliphatic residues.
Black dashed line shows location of the paired leucines of a second putative WH2 domain and its trailing aliphatic residues. Numbers listed just prior to the C-terminus are the final amino acid residue count. (B) SDS-PAGE resolution of FMNL1ɣ-CT and
FMNL1ɣFH2ᴓ-CT fusion proteins, purified as described in Materials and Methods.
Lanes labeled “Lysate” contain lysate of Escherichia coli following induction. Lanes labeled “Sepharose” contain GST-tagged protein bound to glutathione Sepharose 4B.
Lanes labeled “FMNL1ɣ-CT” and “FMNL1ɣFH2ᴓ-CT” contain purified fusion protein following thrombin cleavage. (C) FMNL1ɣ-CT inhibits actin filament assembly. 3.0µM
G-actin (20% pyrene-labeled) was incubated with 100nM CYK1-CT or 500nM FMNL1ɣ-
CT and fluorescence was measured as described in Materials and Methods. Triangles are raw data and curve is a 2nd-order polynomial best-fit curve. Data is representative of three separate experiments. (D) Least-squares regression analysis was performed on
(C). Data is representative of three separate experiments. (E) Both FMNL1ɣ-CT and
FMNL1ɣFH2ᴓ-CT inhibit actin filament assembly. Experiments were performed as in (C) except 100nM CYK1-CT was replaced with 500nM FMNL1ɣFH2ᴓ-CT. Triangles are raw data. Curve is a 2nd-order polynomial best-fit curve. Data is representative of three separate experiments. (F) Least-squares regression analysis was performed on (E).
Data is representative of three separate experiments. (G) Actin assembly rates for
CYK1-CT and FMNL1ɣ-CT were determined as described in Materials and Methods.
191
Bars represent the average assembly rate in nmol/s ± SD (n = 9). * p < 0.05 compared to control assembly rate. (H) Actin assembly rates for FMNL1ɣ-CT and FMNL1ɣFH2ᴓ-
CT were determined as in (G). Bars represent average assembly rate in nmol/s ± SD (n
= 9). * p < 0.05 compared to control assembly rates.
192
Figure 2
193
Figure 2: FMNL1ɣ is an efficient actin-bundling protein. (A) The C-terminus of
FMNL1 is required for proper actin bundling. Low-speed co-sedimentation assays were performed using 10µM actin and 500nM indicated purified fusion protein as described in
Materials and Methods, control lane contains buffer only. Data is representative of multiple experiments. (B) FMNL1ɣ-CT bundles more actin than control or FMNL1Δ-CT.
The percentage of actin distribution was determined from four separate experiments.
Shaded and open bars represent the average percent actin distribution in the supernatant and pellet fraction, respectively ± SD. * p < 0.05 compared to the percentage of actin distributed in the control and FMNL1Δ-CT pelleted fractions. (C)
FMNL1 alternative splice isoforms vary in their ability to bundle actin. Actin distribution in the pellet fraction was measured for FMNL1α, FMNL1β, FMNL1ɣ, and FMNL1ɣFH2ᴓ and the average ± SD of four experiments is displayed. * p < 0.05 compared to percentage of actin pelleted by FMNL1α-CT. (D) Actin bundled by FMNL1ɣ-CT is saturable. A dose-response low-speed co-sedimentation assay was performed with indicated increasing concentrations of FMNL1ɣ-CT, as performed in (A). Bands are actin pelleted by FMNL1ɣ-CT. Representative image of three multiple experiments is shown. (E) Indirect fluorescence microscopy analysis visually demonstrates the ability of FMNL1ɣ-CT to bundle actin. Control image provides visual confirmation of fixed F- actin. F-actin incubated with FMNL1ɣ-CT is bundled more tightly together and not well distributed. Images are representative of three multiple experiments. Bar = 5µm.
194
Figure 3
195
Figure 3: FMNL1 regulates breast cancer cell 2D motility. Representative rose plot of 20 different 16hr. tracks of MDA-MB-231 cells transduced with 400nM control siRNA
(A) or 400nM FMNL1 siRNA (B). Different colors represent different cells with cell tracks beginning at the origin and ending at circles corresponding to track color. (C) The average accumulated distance traveled by individual cells is significantly reduced upon silencing of FMNL1 expression. Cells were transduced with siRNA and tracked as described in Materials and Methods. Data is from multiple experiments and is presented as the average ± SD (n = 24). * p < 0.05 compared to cells transduced with control siRNA. (D) The average velocity of individual cells is significantly reduced upon silencing of FMNL1 expression. Average velocity was measured as described in
Materials and Methods. Data from multiple experiments is presented as the average accumulated distance ± SD (n = 24). Two-sided Student’s t-test was used for statistical analysis. * p < 0.05 compared to cells transduced with control siRNA. (E)
Representative Western blots demonstrating depletion of FMNL1 expression. Three separate FMNL1 siRNA oligonucleotides exhibit successful protein expression silencing.
Transaldolase (TALDO1) was probed as a loading control. (F) FMNL1 displays a unique localization pattern in MDA-MB-231 cells. Endogenous FMNL1 expression was visualized via immunofluorescence microscopy with FMNL1 1° antibody conjugated to
DyLight 488 (green). F-actin was visualized with rhodamine phalloidin (red) and DAPI was used for nuclear staining (blue). Bar = 5µm. (G) FMNL1 expression is reduced upon transduction with siRNA. MDA-MB-231 cells were transduced with 400nM FMNL1- targeting siRNA for 48hrs. after which they were fixed, stained, and observed via immunofluorescence microscopy as in (F).
196
Figure 4
197
Figure 4: Expression of GFP-tagged, full-length FMNL1 alternative splice isoforms. (A) Cells expressing indicated FMNL1 constructs were co-stained with rhodamine phalloidin to visualize F-actin and DAPI for nuclear staining. Unconjugated
GFP is primarily expressed in the nucleus while FMNL1Δ has distributed expression similar to that of endogenous FMNL1. FMNL1α and FMNL1β both localize to the nuclear and perinuclear region while FMNL1ɣ localizes to the lamella-lamellipodia interface and the cell periphery. Localization of FMNL1ɣFH2ᴓ and FMNL1ɣR resemble
FMNL1ɣ localization. Bar = 5µm. (B) FMNL1 alternative splice isoforms are expressed at the mRNA level in MDA-MB-231 cells. Image is representative of RT-PCR experiments performed with MDA-MB-231 cDNA generated from whole-cell RNA.
Arrows indicate PCR products specific for FMNL1Δ and the alternative splice isoforms.
Two bands are seen in the FMNL1α and FMNL1ɣ lane as the same reverse primer was used since they share the same common end sequence. (C) Expression of endogenous
FMNL1 and exogenous, GFP-tagged FMNL1 alternative splice isoforms in transfected
MDA-MB-231 cell populations. Representative blot shows lower bands of endogenous
FMNL1 in all populations and upper bands, ~27kDa higher, indicate the expression of exogenous, GFP-tagged mutants and alternative splice isoforms in transfected cells.
Transaldolase (TALDO1) is ~37.5 kDa and used as a loading control for each lane.
Numbers indicate approximate molecular weight in kDa. (D) Cells expressing GFP-
FMNL1ɣR are depleted of endogenous, but not exogenous, FMNL1 expression. Cells expressing GFP-FMNL1ɣ and GFP-FMNL1ɣR were transduced with FMNL1 siRNA and probed for FMNL1 as described in Materials and Methods. Representative Western blot shows depletion of endogenous and exogenous FMNL1 expression except in cells expressing GFP-FMNL1ɣR. Transaldolase (TALDO1) is used as a loading control for each experiment. (E) GFP expression levels of FMNL1 constructs in MDA-MB-231 cell populations were determined spectrophotometrically as described in Materials and
198
Methods. Data is presented as the percentage of unconjugated GFP expression and represents an average ± SD (n = 12).
199
Figure 5
200
Figure 5: FMNL1 mediates cellular morphology, adhesion, and locomotion. (A)
MDA-MB-231 cells expressing GFP-FMNL1ɣ, GFP-FMNL1ɣFH2ᴓ, or GFP-FMNL1ɣR display a significant increase in cell perimeter. Cells expressing indicated proteins were fixed, stained, and observed via indirect fluorescence microscopy as described in
Materials and Methods. The perimeter of 20 cells for each condition from multiple experiments was measured using NIS Elements. Data is the average perimeter in µm ±
SD (n = 3). * p < 0.05 compared to cells expressing GFP only. (B) MDA-MB-231 cells expressing GFP-FMNL1ɣ, GFP-FMNL1ɣFH2ᴓ, and GFP-FMNL1ɣR display a significant increase in cell surface area when compared to cells expressing GFP only. Data was acquired and measured as described in (A) and shown as the average surface area in
µm2 ± SD (n = 3). * p < 0.05 compared to cells expressing GFP only. (C) Silencing of
FMNL1 expression results in a decrease in cellular adhesive abilities. WT and GFP-
FMNL1ɣR expressing MDA-MB-231 cells were transduced with control or FMNL1- targeting siRNA and subjected to adhesion assays as described in Materials and
Methods. Adhesion was measured as a percentage of WT cell adhesion. Data displayed are the average of multiple experiments ± SD (n = 12). * p < 0.05 compared to cells transduced with control siRNA. (D) Cells expressing GFP-FMNL1ɣ or GFP-
FMNL1ɣFH2ᴓ demonstrate an adhesive gain-of-function. Cell adhesion experiments were performed under the same conditions as (C) but without siRNA. Adhesion was measured as a percentage of WT cell adhesion. Data displayed are the average of multiple experiments ± SD (n = 12). * p < 0.05 compared to WT adhesion. (E)
Transmigration of cells expressing exogenous, GFP-tagged FMNL1 alternative splice isoforms is significantly inhibited following silencing of FMNL1 expression. Cells expressing indicated proteins were challenged to cross an uncoated porous filter membrane following FMNL1 depletion, as described in Materials and Methods.
Transmigration is expressed as a percentage of unconjugated GFP-expressing cells
201 transduced with siRNA. Data shows the average ± SD (n = 9). * p < 0.05 compared to
GFP only cells transduced with control siRNA. NS = not significantly different. (F)
FMNL1ɣ is necessary for invasion. Cells expressing indicated proteins were challenged to a Matrigel invasion assay as described in Materials and Methods. Invasion is expressed as a percentage of WT MDA-MB-231 cells transduced with control siRNA.
Shaded and open bars represent cells transduced with control or FMNL1 siRNA, respectively, and are shown as the average ± SD (n = 3). * p < 0.05 compared to control siRNA transduced cells expressing their respective GFP-FMNL1 alternative splice isoform. NS = not significantly different.
202
Figure 6
203
Figure 6: The formin FMNL1 is necessary for efficient invasion and displays a unique WH2 domain structure. (A) Representative Western blots demonstrating depletion of formins Dia1, Dia2, and FHOD1 by siRNA. Transaldolase (TALDO1) is shown as a loading control for each experiment. Numbers indicate approximate molecular weights in kDa. (B) MDA-MB-231 cells were transduced with 400nM indicated siRNA and allowed to migrate through a Matrigel-coated membrane for as described in Materials and Methods. Matrigel invasion is presented as the average of untreated cells ± SD (n = 9). * p < 0.05 compared to invasion by cells transduced with control siRNA. (C) MDA-MB-231 cells were transduced with combinations of 400nM indicated siRNAs and allowed to migrate through a Matrigel-coated membrane. Matrigel invasion is presented as the average of untreated cells ± SD (n = 9). * p < 0.05 compared to invasion of cells transduced with control siRNA. ** p < 0.05 compared to invasion by cells transduced with FMNL1 siRNA. (D) MDA-MB-231 cells were treated with SMIFH2 or DMSO as a vehicle control. Matrigel invasion is presented as the average of untreated cells ± SD (n = 9). *** p < 0.05 compared to cells treated with
DMSO. (E) Amino acid sequence alignments of WH2 domain-containing proteins
FMNL1ɣ, WASP, and WAVE2 (Modified from Vaillant et al., 2008 and Heimsath and
Higgs, 2012). Tandem leucine repeats are shown in yellow and trailing aliphatic residues in pink. (F) Predicted molecular structure of FMNL1ɣ (D1066-H1123) visualized with PyMOL. Tandem leucine repeats are shown in yellow and phenylalanine residues are highlighted red in ribbon.
204
Discussion
In this study, we have described a novel actin-regulatory function independent of the FH2 domain and unique to FMNL1ɣ. We have identified localization patterns exclusive to individual alternative splice isoforms of FMNL1 and shown that FMNL1ɣ specifically plays a role in maintaining cell morphology. We demonstrated that loss of
FMNL1 expression results in both inhibition of cellular adhesion and reduction of invasion and migration. Notably, we have established that FMNL1 is a proficient actin bundling protein, independent of interactions between the FH2 domain and actin filament barbed ends, and requires the entirety of its C-terminus to efficiently bundle actin filaments.
Several formins have been implicated in cellular motility. FMNL1 and Dia1, but not Dia2 and FHOD1, play an essential role in cell invasion in this model system. Dia1 has been shown to be essential for the nucleation and assembly of F-actin rich structures such as filopodia, lamellipodia, membrane ruffles, focal adhesions, and invadopodia (Watanabe et al., 1999, Sarmiento et al., 2008, Lizarraga et al., 2009, Tanji et al., 2010, Isogai et al., 2015, Young et al., 2015). Furthermore, Dia1 has been shown to stabilize microtubules and regulate endosomal trafficking of matrix metalloproteases
(MMPs), essential for proteolytic degradation of the ECM (Tominaga et al., 2000,
Gasman et al., 2003, Pan et al., 2014, Kim et al., 2016). Inhibition of Dia1 expression could prevent the formation of the necessary cellular structures and the regulation of cytoskeletal remodeling, thereby impeding invasion.
It was originally speculated that FMNL1 was predominately expressed in cells of the hematopoietic lineage (Favaro et al., 2013, Gomez et al., 2007, Han et al., 2013).
However, we have shown that FMNL1 is expressed in a variety of other cell and tissue types, including cerebral tissue, skeletal muscle, and some epithelial cells (Krainer et al.,
2013). Others have confirmed FMNL1 expression in both malignant T- and B-cells, as
205 well as cancer cell lines of an epithelial origin (Gomez et al., 2007, Schuster et al., 2007,
Han et al., 2009, Colon-Franco et al., 2011). In past studies, we have shown that
FMNL1-based actin regulation is essential for both macrophage adhesion and migration
(Mersich et al., 2010, Miller and Blystone, 2015,A, Miller et al., 2017). Therefore, it is not surprising that silencing of FMNL1 expression in MDA-MB-231 cells results in a reduction of invasion. Simultaneous silencing of FMNL1 and Dia1 results in a cumulative effect and reduction of invasion to a level similar to that of cells treated with the pharmacological inhibitor SMIFH2, targeting FH2 domain activity. This leads us to speculate that FMNL1 and Dia1 may each play a substantial role in invasion through two different mechanisms of actin regulation: Dia1 has been shown to be an efficient actin nucleator with the ability to rapidly assemble filaments whereas FMNL1 has the ability to both sever and/or bundle actin filaments (Li and Higgs, 2003, Harris et al., 2004, Harris et al., 2006, Esue et al., 2008). Simultaneous silencing of these proteins could result in exacerbated actin-modifying and cytoskeletal defects, limiting the motile and invasive behavior of cells.
RT-PCR experiments confirmed expression of all three FMNL1 alternative splice isoforms at the mRNA level. We further confirmed protein expression of FMNL1 via
Western blotting and immunofluorescence microscopy. Analysis of localization patterning confirmed expression of FMNL1 to be at the cell periphery, the nucleus, and the perinuclear region. Localization patterns of exogenous, GFP-tagged, full-length
FMNL1 alternative splice isoforms match the localization patterns of both endogenous
FMNL1 and GFP-FMNL1Δ, the mutant truncated at the splice site. This finding indicates that the C-terminal region of FMNL1 defines localization of the alternative splice isoforms within the cell and possibly their function. FMNL1α and FMNL1β localized to the perinuclear region, while FMNL1ɣ expression was enriched at the cell periphery. This differs from previous reports implicating FMNL1ɣ as a Golgi complex component,
206 however those studies utilized indirect immunofluorescence with peptide-directed antibodies as opposed to the direct visualization of exogenously expressed, full-length,
GFP-tagged alternative splice isoforms used here (Gomez et al., 2007, Han et al., 2009,
Colon-Franco et al., 2011).
Cells expressing exogenous FMNL1ɣ, but not other isoforms, in conjunction with endogenous FMNL1, displayed a significantly higher surface area and perimeter. To our knowledge, this is the first time an alternative splice isoform of FMNL1 has been shown to alter the morphological status of a cell. FMNL1ɣ expression at the cell periphery could result in a modification of filamentous actin in the lamella. An increase in linear actin bundles could result in the growth of this region, broadening the cell itself, permitting more adhesive contacts. Additionally, FMNL1ɣ has been previously shown to co-localize with myosin IIb, an actin-binding and cross-linking protein essential for adhesion and migration (Han et al., 2009). This suggests FMNL1ɣ may modify actin where the lamella and lamellipodia meet to permit myosin behaviors. This region of the cell periphery contains linear actin bundles actively undergoing assembly and disassembly, potentially explaining the brevity of visualization of FMNL1ɣ accumulations within the cells (Ponti et al., 2004, Vicente-Manzanares et al., 2009).
This interpretation correlates with our experiments detailing the role of FMNL1 in migration. Depletion of FMNL1 expression significantly reduced the average velocity and average accumulated distance of MDA-MB-231 cells. Migration and invasion were also significantly reduced in cells expressing exogenous GFP-FMNL1 mutants except for those expressing the non-targeted GFP-FMNL1ɣR construct. A loss of FMNL1 could result in disruption of proper actin filament bundling, a key actin regulatory event essential for promoting cell migration through cytoskeletal remodeling. Rescue by expression of GFP-FMNL1ɣR confirms that FMNL1ɣ is required for efficient migration and invasion of MDA-MB-231 cells. Expanding the importance of this formin, this data is
207 also in agreement with our past studies indicating FMNL1 is essential for proper macrophage migration and with deletion resulting in embryonic lethality in mice (Mersich et al., 2010, Miller and Blystone, 2015,A, Miller et al., 2017).
With our findings that FMNL1ɣ contributed substantively to the migration of diverse cells, we sought to identify a mechanism of action, presuming it related to the modification of actin dynamics. In agreement with previous reports assessing FMNL1β, we find that FMNL1ɣ shows no ability to enhance the polymerization of actin; rather its presence suppresses the inherent rate of actin assembly, independent of a fully functional FH2 domain (Harris et al., 2004, Harris et al., 2006). Inclusion of profilin in these studies did not result in FMNL1-enhancement of actin assembly (data not shown).
Our earlier reports on macrophages showed that FMNL1 stabilized podosomes; increasing their half-life and maintaining compactness while associating with linear actin filaments of the podosome core (Mersich et al., 2010). We hypothesized that FMNL1ɣ may be contributing to cell motility by stabilizing linear actin filaments through bundling activity. Indeed, low-speed co-sedimentation assays demonstrated that all three alternative splice isoforms of FMNL1 can act as bundlers. However, there is a significant increase in FMNL1ɣ bundling activity compared to FMNL1α, an isoform that shares 30 common amino acids with FMNL1ɣ within the C-terminal splice region. This leads us to believe that the addtional 58 amino acid retention found in the FMNL1ɣ alternative splice isoform may play a key role in this bundling activity. We propose that predicted WH2 motifs found in this region of FMNL1ɣ (Figure 1A, 6E, F) may be responsible for this FMNL1ɣ activity. This could occur independent of the FH2 domain and outside of DID-DAD regulation by Rho-family GTPases, in agreement with reports showing neither change in GTPase levels upon FMNL1 deletion, nor effects on FMNL1 localization following manipulation of GTPase activity (Mersich et al., 2010, Miller et al.,
2017). An alternative explanation may be that the effects of FMNL1ɣ are due to cellular
208 localization signals specific to this alternative splice sequence, however the correlating in vitro actin biochemistry assays argue against this hypothesis.
Aside from FMNL1, WH2 motifs have been identified in the C-terminal region of other mammalian formins including Dia1, DAAM1, FMNL2, FMNL3, and INF2 (Chhabra and Higgs, 2006, Vaillant et al., 2008, Gould et al., 2011, Heimsath and Higgs, 2012).
These motifs have also been identified in numerous other actin-associated proteins
(Dominguez, 2016). While isolated WH2 domains can only prevent G-actin pointed-end
(- end) assembly via monomer subunit binding, their functions are exceedingly varied when interacting with other proteins or other regions of the molecule (Carlier et al., 2013,
Dominguez, 2016). The domain itself varies from ~17-35 amino acid residues in length and primarily consists of an amphiphilic α-helix containing at least three aliphatic amino acid residues. These aliphatic residues are found as a pair and the third is located ~7-
10 amino acid residues distally (Paunola et al., 2002, Chereau et al., 2005, Aguda et al.,
2006, Husson et al., 2011, Dominguez, 2016). Most often, these paired amino acid residues are leucines, however, in FMNL2 and FMNL3 this pair consists of one leucine residue and one isoleucine residue (Vaillant et al., 2008, Dominguez, 2016).
Interestingly, FMNL1ɣ contains two tandem leucine residue pairs, a unique sequence among WH2 domain-containing proteins (Figure 1A). Structural predictions indicate that both pairs reside on the same face of the amphiphilic α-helix and, following the subsequent aliphatic residues, contain a region of five phenylalanine residues, each separated by only one amino acid residue. This unique sequence of phenylalanine residues is uncommon and could potentially interact with the WH2 domain. This theory is supported by previous studies demonstrating that WASp interacts not only with actin and Arp2/3, but also different regions of itself (Kim et al., 2000, Chereau et al., 2005, Ti et al., 2011).
209
This distinct amino acid residue sequence of two overlapping WH2 domains is exclusive to FMNL1ɣ and could explain the actin-modifying functions we have observed.
Much like Spire and FMNL3, WH2 motifs in FMNL1ɣ could be capping actin filaments and preventing assembly (Bosch et al., 2007, Carlier et al., 2013, Heimsath and Higgs,
2012). This could also explain why FMNL1ɣFH2ᴓ-CT, which has a dysfunctional FH2 domain, was able to lower actin assembly rates to levels similar to FMNL1ɣ-CT.
Alternatively or additionally, the presence of this WH2 domain may result in the enhanced bundling activity demonstrated by FMNL1ɣ-CT. This also supports our theory that this bundling activity is occurring regardless of FH2 domain activity since FMNL1ɣ-
CT and FMNL1ɣFH2ᴓ-CT are able to bundle actin at similar levels. While the question of how this unique WH2 domain contributes to the bundling process remains to be determined, the work here advances understanding of this process and the mechanism behind WH2 domain function in formins.
This study provides evidence that regions of formin proteins other than the FH2 domain may be critical for regulating actin-modifying events. This study is also the first to show that not only are the actin regulatory events of FMNL1 determined by the C- terminal region of the formin, but also that expression of FMNL1ɣ, but not FMNL1α or
FMNL1β, plays an essential role in both cell migration and morphology in a variety of cell types of different embryonic origin. Future work elucidating the mechanism behind how the C-terminal region of FMNL1ɣ operates within the cell may provide a potential therapeutic target for regulating cell migration.
210
References
Aguda AH, Xue B, Irobi E, Preat T, Robinson RC (2006). The structural basis of actin interaction with multiple WH2/β-thymosin motif-containing proteins. Structure 14, 469- 476.
Block J, Breitsprecher D, Kuhn S, Winterhoff M, Kage F, Geffers R, Duwe P, Rohn JL, Baum B, Brakebusch C, Geyer M, Stradal TE, Faix J, Rottner K (2012). FMNL2 drive actin-based protrusion and migration downstream of Cdc42. Curr Biol 22, 1005-1012.
Bosch M, Le KH, Bugyi B, Correia JJ, Renault L, Carlier MF (2007). Analysis of the function of Spire in actin assembly and its synergy with formin and profilin. Mol Cell 28, 555-568.
Bravo-Cordero JJ, Hodgson L, Condeelis J (2012). Directed cell invasion and migration during metastasis. Curr Opin Cell Biol 24, 277-283.
Carlier MF, Pernier J, Avvaru BS (2013). Control of actin filament dynamics at barbed ends by WH2 domains: from capping to permissive and processive assembly. Cytoskeleton 70, 540-549.
Chereau D, Kerff F, Graceffa P, Grabarek Z, Langsetmo K, Dominguez R (2005). Actin- bound structures of Wiskott-Aldrich syndrome protein (WASP)-homology domain 2 and the implications for filament assembly. Proc Natl Acad Sci 102, 16644-16649.
Chhabra ES and Higgs HN (2006). INF2 is a WASP homology 2 motif-containing formin that severs actin filaments and accelerates both polymerization and depolymerization. J Biol Chem 281, 26754-26767.
Colon-Franco JM, Gomez TS, Billideau DD (2011). Dynamic remodeling of the actin cytoskeleton by FMNL1ɣ is required for structural maintenance of the Golgi complex. J Cell Sci 124, 3118-3126.
Dominguez R (2016). The WH2 Domain and Actin Nucleation: Necessary but Insufficient. Trends Biochem Sci 41, 478-490.
Esue O, Harris ES, Higgs HN, Wirtz D (2008). The filamentous actin cross- linking/bundling activity of mammalian formins. J Mol Biol 384, 324-334.
Favaro P, Traina F, Machado-Neto JA, Lazarini M, Lopes MR, Pereira JK, Costa FF, Infante E, Ridley AJ, Saad ST (2013). FMNL1 promotes proliferation and migration of leukemia cells. J Leukoc Biol 94, 503-512.
Fife CM, McCarroll JA, Kavallaris M (2014). Movers and shakers: cell cytoskeleton in cancer metastasis. Br J Pharmacol 171, 5507-5523.
Gasman S, Kalaidzidis Y, Zerial M (2003). RhoD regulates endosome dynamics through Diaphanous-related Formin and Src tyrosine kinase. Nat Cell Biol 5, 195-204.
211
Gaucher JF, Mauge C, Didry D, Guichard B, Renault L, Carlier MF (2012). Interactions of isolated C-terminal fragments of neural Wiskott-Aldrich syndrome protein (N-WASP) with actin and Arp2/3 complex. J Biol Chem 287, 34646-34659.
Gomez TS, Kumar K, Medeiros RB, Shimizu Y, Leibson PJ, Billideau DD (2007) Formins regulate the actin-related protein 2/3 complex-independent polarization of the centrosome to the immunological synapse. Immunity 26, 177-190.
Goode BL and Eck MJ (2007). Mechanism and function of formins in control of actin assembly. Annu Rev Biochem 76, 593-627.
Gould CJ, Maiti S, Michelot A, Graziano BR, Blanchoin L, Goode BL (2011). The formin DAD domain plays dual roles in autoinhibition and actin nucleation. Curr Biol 21, 384- 390.
Haglund CM, Choe JE, Skau CT, Kovar DR, Welch MD (2010). Rickettsia Sca2 is a bacterial formin-like mediator of actin-based motility. Nat Cell Biol 12, 1057-1063.
Han Y, Eppinger E, Schuster IG, Weigand LU, Liang X, Kremmer E, Peschel C, Krackhardt AM (2009). Formin-like 1 (FMNL1) is regulated by N-terminal myristoylation and induces polarized membrane blebbing. J Biol Chem 284, 33409-33417.
Han Y, Yu G, Sarioglu H, Caballero-Martinez A, Schlott F, Ueffing M, Haase H, Peschel C, Krackhardt AM (2013). Proteomic investigation of the interactome of FMNL1 in hematopoietic cells unveils a role in calcium-dependent membrane plasticity. J Proteomics 78, 72-82.
Harris ES and Higgs HN (2006). Biochemical analysis of mammalian formin effects on actin dynamics. Methods Enzymol 406, 190-214.
Harris ES, Li F, Higgs HN (2004). The mouse formin, FRLα, slows actin filament barbed end elongation, competes with capping protein, accelerates polymerization from monomers, and severs filaments. J Biol Chem 279, 20076-20087.
Harris ES, Rouiller I, Hanein D, Higgs HN (2006). Mechanistic differences in actin bundling actibity of two mammalian formins, FRL1 and mDia2. J Biol Chem 281, 14383- 92.
Heimsath EG and Higgs HN (2012). The C terminus of formin FMNL3 accelerates actin polymerization and contains a WH2 domain-like sequence that binds both monomers and filament barbed ends. J Biol Chem 287, 3087-3098.
Husson C, Renault L, Didry D, Pantaloni D, Carlier MF (2011). Cordon-Bleu uses WH2 domains as multifunctional dynamizers of actin filament assembly. Mol Cell 43, 464-477.
Isogai T, van der Kammen R, Leyton-Puig D, Kedziora KM, Jalink K, Innocenti M (2015). Initiation of lamellipodia and ruffles involves cooperation between mDia1 and Arp2/3 complex. J Cell Sci 128, 3796-3810.
212
Kage F, Steffen A, Ellinger A, Ranftler C, Gehre C, Brakebusch C, Pavelka M, Stradal T, Rottner K (2017,A) FMNL2 and -3 regulate Golgi architecture and anterograde transport downstream of Cdc42. Sci Rep 7, 9791.
Kage F, Winterhoff M, Dimchev V, Mueller J, Thalheim T, Freise A, Bruhmann S, Kollasser J, Block J, Dimchev G, Geyer M, Schnittler HJ, Brakebusch C, Stradal TE, Carlier MF, Sixt M, Faix J, Rottner K (2017,B). FMNL formins boost lamellipodial force generation. Nat Commun 8, 14832.
Katoh M and Katoh M (2003). Identification and characterization of human FMNL1, FMNL2, and FMNL3 genes in silico. Int J Oncol 22, 1161-1168.
Kim AS, Kakalis LT, Abdul-Manan N, Liu GA, Rosen MK (2000) Autoinhibition and activation mechanisms of the Wiskott-Aldrich syndrome protein. Nature 404, 151-158.
Kim D, Jung J, You E, Ko P, Oh S, Rhee S (2016). mDia1 regulates breast cancer invasion by controlling membrane type 1-matrix metalloproteinase localization. Oncotarget 7, 17829-17843.
Kitzing TM, Wang Y, Pertz O, Copeland JW, Grosse R (2010). Formin-like 2 drives amoeboid invasive cell motility downstream of RhoC. Oncogene 29, 2441-2448.
Kovar DR (2006). Molecular details of formin-mediated actin assembly. Curr Opin Cell Biol 18, 11-17.
Krainer EC, Ouderkirk JL, Miller EW, Miller MR, Mersich AT, Blystone SD (2013). The multiplicity of human formins: expression patterns in cells and tissues. Cytoskeleton 70, 424-438.
Lammers M, Meyer S, Kuhlmann D, Wittinghofer A (2008). Specificity of interactions between mDia isoforms and Rho proteins. J Biol Chem 283, 35236-35246.
Li F and Higgs HN (2003). The mouse Formin mDia1 is a potent actin nucleation factor regulated by autoinhibition. Curr Biol 13, 1335-1340.
Lizarraga F, Poincloux R, Romao M, Montagnac G, Le Dez G, Bonne I, Rigaill G, Raposo G, Chavrier P (2009). Diaphanous-related formins are require for invadopodia formation and invasion of breast tumor cells. Cancer Res 69, 2792-2800.
Mersich AT, Miller MR, Chkourko H, Blystone SD (2010). The formin FRL1 (FMNL1) is an essential component of macrophage podosomes. Cytoskeleton 67, 573-585.
Miller MR and Blystone SD (2015,A). Human macrophages utilize the podosome formin FMNL1 for adhesion and migration. Cellbio 4, 1-11.
Miller MR and Blystone SD (2015,B). Reliable and inexpensive expression of large, tagged, exogenous proteins in murine bone marrow-derived macrophages using a second generation lentiviral system. J Biol Methods 2, e23.
Miller MR, Miller EW, Blystone SD (2017). Non-canonical activity of the podosomal formin FMNL1ɣ supports immune cell migration. J Cell Sci 130, 1730-1739.
213
Moseley JB, Sagot I, Manning AL, Xu Y, Eck MJ, Pellman D, Goode BL (2004). A conserved mechanism for Bni1- and mDia1-induced actin assembly and dual regulation of Bni1 by Bud6 and profilin. Mol Biol Cell 15, 896-907.
Neidt EM, Skau CT, Kovar DR (2008). The cytokinesis formins from the nematode worm and fission yeast differentially mediate actin filament assembly. J Biol Chem 283, 23872- 23883.
Nezami AG, Poy F, Eck MJ (2006). Structure of the autoinhibitory switch in formin mDia1. Structure 14, 257-263.
Otomo T, Tomchick DR, Otomo C, Panchal SC, Machius M, Rosen MK (2005,B). Structural basis of actin filament nucleation and processive capping by a formin homology 2 domain. Nature 433, 488-494.
Pan J, Lordier L, Meyran D, Rameau P, Lecluse Y, Kitchen-Goosen S, Badirou I, Mokrani H, Narumiya S, Alberts AS, Vainchenker W, Chang Y (2014) The formin DIAPH1 (mDia1) regulates megakaryocyte proplatelet formation by remodeling the actin and microtubule cytoskeletons. Blood 124, 3967-3977.
Paul AS and Pollard TD (2009,B). Review of the mechanism of processive actin filament elongation by formins. Cell Motil Cytoskeleton 66, 606-617.
Paunola E, Mattila PK, Lappalainen P (2002). WH2 domain: a small, versatile adapter for actin monomers. FEBS Lett 513, 92-97.
Pollard TD (1986). Rate constants for the reactions of ATP- and ADP-actin with the ends of actin filaments. J Cell Biol 103, 2747-2754.
Pollard TD and Cooper JA (1986). Actin and actin-binding proteins. A critical evaluation of mechanisms and functions. Annu Rev Biochem 55, 987-1035.
Ponti A, Machacek M, Gupton SL, Waterman-Storer CM, Danuser G (2004). Two distinct actin networks drive the protrusion of migrating cells. Science 305, 1782-1786.
Pruyne D, Evangelista M, Yang C, Bi E, Zigmond S, Bretscher A, Boone C (2002) Role of formins in actin assembly: nucleation and barbed-end association. Science 297, 612- 615.
Quinlan ME, Heuser JE, Kerkhoff E, Mullins RD (2005). Drosophila Spire is an actin nucleation factor. Nature 433, 382-388.
Randall TS and Ehler E (2013). A formin-g role during development and disease. Eur J Cell Biol 93, 205-211.
Rizvi SA, Neidt EM, Cui J, Feiger Z, Skau CT, Gardel ML, Kozmin SA, Kovar DR (2009). Identification and characterization of a small molecule inhibitor of formin-mediated actin assembly. Chem Biol 16, 1158-1168.
214
Sarmiento C, Wang W, Dovas A, Yamaguchi H, Sidani M, El-Sibai M, Desmarais V, Holman HA, Kitchen S, Backer JM, Alberts A, Condeelis J (2008). WASP family members and formin proteins coordinate regulation of cell protrusions in carcinoma cells. J Cell Biol 180, 1245-1260.
Schonichen A and Geyer M (2010). Fifteen formins for an actin filament: a molecular view on the regulation of human formins. Biochim Biophys Acta 2, 152-163.
Schuster IG, Busch DH, Eppinger E, Kremmer E, Milosevic S, Hennard C, Kuttler C, Ellwart JW, Frankenberger B, Nossner E, Salat C, Bogner C, Borkhardt A, Kolb HJ, Krackhardt AM (2007). Allorestricted T cells with specificity for the FMNL1-derived peptide PP2 have potent antitumor activity against hematologic and other malignancies. Blood 110, 2931-2939.
Seth A, Otomo C, Rosen MK (2006). Autoinhibition regulates cellular localization and actin assembly activity of the diaphanous-related formins FRLα and mDia1. J Cell Biol 174, 701-713.
Spudich JA and Watt S (1971). The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the tropomyosin-troponin complex with actin and proteolytic fragments of myosin. J Biol Chem 246, 4866-4871.
Tanji M, Ishizaki T, Ebrahimi S, Tsuboguchi Y, Sukezane T, Akagi T, Frame MC, Hashimoto N, Miyamoto S, Narumiya S (2010) mDia1 targets v-Src to the cell periphery and facilitates cell transformation, tumorigenesis, and invasion. Mol Cell Biol 30, 4606- 4615.
Ti SC, Jurgenson CT, Nolen BJ, Pollard TD (2011). Structural and biochemical characterization of two binding sites for nucleation-promoting factor WASp-VCA on Arp2/3 complex. Proc Natl Acad Sci 108, 463-471.
Tominaga T, Sahai E, Chardin P, McCormick F, Courtneidge SA, Alberts AS (2000). Diaphanous-related formins bridge Rho GTPase and Src tyrosine kinase signaling. Mol Cell 5, 13-25.
Vaillant DC, Copeland SJ, Davis C, Thurston SF, Abdennur N, Copeland JW (2008). Interaction of the N- and C-terminal autoregulatory domains of FRL2 does not inhibit FRL2 activity. J Biol Chem 283, 33750-33762.
Vega FM, Fruwirth G, Ng T, Ridley AJ (2011). RhoA and RhoC have distinct roles in migration and invasion by acting through different targets. J Cell Biol 193, 655-665.
Vicente-Manzanares M, Ma X, Adelstein RS, Horwitz AR (2009). Non-muscle myosin II takes centre stage in cell adhesion and migration. Nat Rev Mol Cell Biol 10, 778-790.
Vizcarra CL, Bor B, Quinlan ME (2014). The role of formin tails in actin nucleation, processive elongation, and filament bundling. J Biol Chem 289, 30602-30613.
Wang F, Zhang L, Duan X, Zhang GL, Wang ZB, Wang Q, Xiong B, Sun SC (2015). RhoA-mediated FMNL1 regulates GM130 for actin assembly and phosphorylates MAPK for spindle formation in mouse oocyte meiosis. Cell Cycle 14, 2835-2843.
215
Watanabe N, Kato T, Fujita A, Ishizaki T, Narumiya S (1999). Cooperation between mDia1 and ROCK in Rho-induced actin reorginzation. Nat Cell Biol 1, 136-143.
Xu Y, Moseley JB, Sagot I, Poy F, Pellman D, Goode BL, Eck MJ (2004). Crystal structures of a Formin Homology-2 domain reveal a tethered dimer architecture. Cell 116, 711-723.
Yayoshi-Yamamoto S, Taniuchi I, Watanabe T (2000). FRL, a novel formin-related protein, binds to Rac and regulates cell motility and survival of macrophages. Mol Cell Biol 20, 6872-6881.
Young LE, Heimsath EG, Higgs HN (2015). Cell type-dependent mechanisms for formin- mediated assembly of filopodia. Mol Biol Cell 26, 4646-4659.
Zigmond SH, Evangelista M, Boone C, Yang C, Dar AC, Sicheri F, Forkey J, Pring M (2003). Formin leaky cap allows elongation in the presence of tight capping proteins. Curr Biol 13, 1820-1823.
216
Chapter 5
General Discussion
217
Introduction
Cancer cell invasion is an essential component of the metastatic cascade, necessary for cells to travel through the ECM and eventually intravasate through the endothelial cell lining of the circulatory system or the lymphatics (Yamaguchi and
Condeelis, 2007, Leber and Efferth, 2009). In order for this to be accomplished, the actin cytoskeleton must undergo various spatiotemporal rearrangements. The primary aim of this dissertation was to determine if the alternative splice isoform of FMNL1,
FMNL1ɣ, contributes to this remodulation of the actin cytoskeleton in a breast adenocarcinoma model and, if so, to discern the mechanism by which this process occurs. This was determined using a combination of molecular, biochemical, and cellular techniques, allowing us to answer this question on three separate levels. As a result, this dissertation as a whole provides a comprehensive study on the role of this alternatively spliced formin in mediating actin cytoskeletal changes in a breast adenocarcinoma model through a unique F-actin bundling mechanism. Furthermore, these data may eventually contribute to the identification of a unique target for cancer therapeutics, as well as other pathologies that may utilize FMNL1ɣ for different adhesive and migratory abilities, such as atherosclerosis and other inflammation disorders (Singh et al., 2002). By performing an mRNA expression analysis, we have been able to identify unique formin expression patterns as well as the expression of all three alternative splice isoforms of FMNL1 in a breast adenocarcinoma model. Using biochemical techniques and this same model system, we successfully demonstrated that two specific residues of a highly conserved domain in FMNL1 are essential for F-actin binding. Based on these two studies, we were able to further characterize and identify a unique region of the FMNL1ɣ alternative splice isoform that is essential for bundling F- actin and mediating actin cytoskeletal alterations imperative for cell adhesion and migration.
218
FMNL1 and FMNL1 Alternative Splice Isoform Expression
Past studies on formin expression levels have provided conflicting results due to different techniques used to measure expression (Gardberg et al., 2010, Kitzing et al.,
2010, Schonichen and Geyer, 2010, Favaro et al., 2013, Gardberg et al., 2014). By performing a complete study using standard qRT-PCR techniques, we were able to provide a comprehensive mRNA expression analysis on all 15 human formins in seven different cell types. This allowed us to subsequently identify unique patterns in formin expression across a variety of cell types. Additionally, we were able to identify mRNA expression of all three FMNL1 alternative splice isoforms in four different cell types. The results of this study will enhance our knowledge of formins, allowing us to efficiently study and advance the formin field of research at both the cellular and molecular level.
Using site-specific primers and seven different cell types, we analyzed the mRNA expression levels of all 15 human formins. HeLa cells are one of the most well- characterized cell types used for research purposes (Masters, 2002). These cervical adenocarcinoma cells displayed the second-highest average total formin expression level across all seven cell types. While FMNL1 was the lowest FMNL subfamily member expressed in these cells, when compared across all seven cell types, the average
FMNL1 expression level observed in HeLa cells was the second highest. This high
FMNL1 expression has been confirmed at the protein level in past studies via Western blot and immunofluorescence. This same study also showed that the FMNL1ɣ alternative splice isoform has been implicated in maintaining the structural actin architecture of the Golgi complex (Colón-Franco et al., 2011). FMNL1 protein expression in HeLa cells was also confirmed in a previous study where Rac-activated
FMNL1 has been shown to mediate membrane ruffling via a complex it forms with srGAP2 in these cells (Mason et al., 2010). These data demonstrate two very different
219 activities in this cell type which could explain the high average FMNL1 expression level we observed here.
HeLa cells were not the only epithelial adenocarcinoma cell type which exhibited high average FMNL1 mRNA expression. We also observed this in MDA-MB-231 cells where average FMNL1 mRNA expression was the third highest amongst all seven cell types studied. FMNL1 mRNA and protein expression in MDA-MB-231 cells has been observed before (Kitzing et al., 2010). Additionally, high expression of FMNL1 has also been confirmed in the basal subtype of human breast cancers, of which MDA-MB-231 cells belong (Gardberg et al., 2014). We also detected expression of all three FMNL1 alternative splice isoforms at the mRNA level in MDA-MB-231 cells. Past studies have suggested FMNL1ɣ could be required for membrane blebbing in these cells (Han et al.,
2009). Furthermore, siRNA-mediated depletion of FMNL1 protein expression reduced invasion of MDA-MB-231 cells (Kitzing et al., 2010). Taken together, these data suggest that the high mRNA expression level we observed with FMNL1 could correspond to the different locomotive functions of these cells and it may be specifically mediated by the
FMNL1ɣ alternative splice isoform.
The highest level of average FMNL1 mRNA expression was also observed in an epithelial cell type. HK-2 cells are kidney cortex epithelial cells which, in addition to the high average FMNL1 mRNA expression, also displayed the highest average total formin mRNA expression level. A past study also confirmed the mRNA expression for all three alternative splice isoforms of FMNL1 in epithelial cells of the embryonic kidney
(HEK293T cells) (Han et al., 2009). Furthermore, qRT-PCR analysis also demonstrated high FMNL1ɣ mRNA expression compared to FMNL1α and FMNL1β as well as localization of FMNL1ɣ to membrane blebs (Han et al., 2009). These data correlate well with our observation of high FMNL1 mRNA expression, which could be explained by the
220 requirement of FMNL1ɣ specifically for regulation of membrane bleb formation (Han et al., 2009).
While we observed the highest levels of average FMNL1 mRNA expression levels in epithelial cell types, expression was still observed in cells of neuronal and myeloid lineage. The cortical neuronal cell line HCN exhibited FMNL1 mRNA expression levels similar to that of macrophages, where FMNL1 function has been well- characterized (Mersich et al., 2010, Naj et al., 2013, Miller and Blystone, 2015, Miller et al., 2017). A previous immunohistochemical analysis on human brain sections did not detect any FMNL1 protein expression in neurons or glial cells (Gardberg et al., 2014).
This could correspond with our results demonstrating low mRNA expression as well.
While Western blot analysis has confirmed FMNL1 protein expression in the cerebellum and whole brain, this was actually quite low compared to the mouse macrophage
RAW264.7 cell line (Mason et al., 2010).
While we did not measure FMNL1 mRNA expression in RAW264.7 cells, we did determine expression in primary human macrophages. We have previously shown that upon differentiation from monocytes, FMNL1 mRNA expression in macrophages is increased six-fold and protein expression is increased by 108% (Mersich et al., 2010).
Additionally, past work from our lab has demonstrated that FMNL is an important component of macrophage podosomes, necessary for proper regulation of this F-actin rich complex (Miller and Blystone, 2015, Miller et al., 2017). FMNL1 mRNA expression levels for macrophages were lower compared to some of the other cell and tissue types but this may be attributed to the limited, albeit essential, functions of macrophages. We further analyzed this cell type and demonstrated that it expresses all three FMNL1 alternative splice isoforms. We have previously shown that macrophage podosome formation and function requires the ɣ alternative splice isoform of FMNL1 (Miller et al.,
2017). Additionally, while this may seem like a lower level of expression amongst these
221 seven cell types, other have demonstrated that amongst over double that number,
FMNL1 is highly expressed in cells of the myeloid lineage (Gardberg et al., 2014). Our findings presented here demonstrate that all three FMNL1 alternative splice isoforms are expressed in primary human macrophages and that while FMNL1 mRNA expression levels may be lower compared to some cell types, its expression and function at the protein level in these cells has been previously confirmed (Mersich et al., 2010, Miller et al., 2017).
We further examined the expression of FMNL1 in myeloid lineage cells by determining expression in Meg-01 cells and platelets. While FMNL1 expression in these two cell types was lower compared to FMNL2 and FMNL3, we were able to detect expression of all three FMNL1 alternative splice isoforms in Meg-01 cells. While we were not able to confirm FMNL1 alternative splice isoform expression in platelets, previous work has confirmed high expression of FMNL1 at the mRNA level in these cells, correlating well with what we have demonstrated here (Zuidscherwoude et al.,
2018). Further, a previous mRNA screen of FMNL1 expression in different cell and tissue types identified high expression of FMNL1 in some myeloid cells, although this was not confirmed at the protein level (Gardberg et al., 2014).
Future work analyzing FMNL1 alternative splice isoform mRNA expression level in all cell types will be necessary to better characterize the function of the FMNL1 gene as a whole. While we were able to accomplish this in four different cell types, this work proved to be difficult as we often had to design different primers, both forward and reverse, to obtain a qualitative PCR product. Primer design was limited as there is only a select amount of sequence in the C-terminal regions of FMNL1 to use for PCR. While we attempted this on several occasions, we were not able to obtain any insightful data.
Additionally, since FMNL1α and FMNL1ɣ share a same common end sequence, specifically determining the mRNA expression level of FMNL1α could prove difficult.
222
We could use other methods to determine mRNA expression levels as well. For example, Northern blotting would allow us to determine expression of these three alternative splice isoforms, although quantitation may not be as accurate as what we could obtain with successful qRT-PCR. Additionally, Northern blotting carries a high risk for RNAse-mediated RNA degradation. RNAseq would also be a useful tool for this analysis, although this can be expensive. However, it may provide us with the most accurate results. Regardless, identification of these mRNA expression levels will further help us characterize the function of the different FMNL1 alternative splice isoforms in different cell types.
While determining FMNL1 alternative splice isoform mRNA expression levels will be useful, ideally we would like to determine their expression at the protein level.
Currently, all commercially available FMNL1 primary antibodies only target an interior region of the molecule and not the C-terminal region. As a result, we would most likely have to design our own primary antibodies. However, this can prove costly and we are unable to know how successful these antibodies are until we procure them following synthesis. Further, much like designing primers for qRT-PCR, there is only a small sequence to work with. Others have generated primary antibodies against the FMNL1 alternative splice isoforms, and we have used some of these, but in our hands we were unable to verify protein expression in cells (Han et al., 2009, Colón-Franco et al., 2011).
Additionally, the similarity between FMNL1α and FMNL1ɣ could prove to be an issue as well. Nonetheless, we could attempt using other methods, such as mass spectrometry or some type of isotope-labeling.
Past studies, including our own, have shown that the different alternative splice isoforms of FMNL1 have specific functions, especially FMNL1ɣ, which has been shown to regulate actin architecture surrounding the Golgi complex, contribute to membrane blebbing, and mediate podosome formation and function (Han et al., 2009, Colón-Franco
223 et al., 2011, Miller et al., 2017). This has been observed with other formins as well.
Isoform IV of FMN1 nucleates actin, mediates cellular protrusion and is essential for adhesion complex formation (Dettenhofer et al., 2008). FHOD3 has a splice site in the
C-terminus, similar to FMNL1, and a specific alternative splice isoform of FHOD3 localizes to striated muscle tissue. Further, phosphorylation of a specific site in this alternative splice isoform results in inhibition of FHOD3 interaction with p62/SQSTM1, a protein essential for autophagy (Iskratsch et al., 2010). Additionally, a specific FMNL2 alternative splice isoform is upregulated in colorectal cancer and melanoma cells and contributes their invasive abilities (Péladeau et al., 2016). These studies demonstrate that characterization and identification of formin function should also focus on the alternative splice isoforms of these proteins. As each alternative splice isoform has varying functions, they should almost be treated as if they are entirely different proteins.
Future work measuring expression patterns at both the mRNA and protein level, as well as subsequent functional assays, should help identify specific traits of these alternative splice isoforms which contribute to different cellular processes.
Two conserved amino acid residues of FMNL1 contribute to actin binding.
FH2 domain-mediated interactions of the FMNL subfamily with actin have been well-characterized in previous studies, implicating this domain as one of the main functional regions of the protein (Harris et al., 2006, Harris et al., 2010, Colon-Franco et al., 2011, Heimsath and Higgs, 2012, Thompson et al., 2013, Kage et al., 2017).
Furthermore, the two highly-conserved amino acid residues of the FH2 domain (I720 and K871 in FMNL1) have been shown to be imperative for this interaction (Shimada et al., 2004, Xu et al., 2004, Otomo et al., 2005b, Harris et al., 2006). Our data demonstrating that inactivation of this domain via mutation of these two conserved residues (I720A and K871A) is the first time that a direct relationship between the FH2
224 domain and cytoplasmic actin has been shown. Moreover, the results of this study as a whole provide evidence that formin function is not specifically reliant upon FH2 domain- mediated barbed end interactions, but other parts of the protein as well.
Interestingly, along with FMNL1β and FMNL1ɣ, the truncation mutant FMNLΔ also exhibited actin-binding abilities. This is intriguing as FMNL1Δ lacks part of the DAD domain and the remaining C-terminal region of this protein. It would be expected that the lack of this region of FMNL1 would contribute to some loss of actin binding, as
FMNL1β contains a unique amino acid sequence and FMNL1ɣ contains the two putative, tandem, overlapping WH2 domains. However, since the truncation site is located within the DAD domain, this may affect activation of the formin. Without an intact DAD domain, the formin may remain in an open, active conformation by default, thereby allowing the
FH2 domain to proceed to interact with actin. This would not be surprising as previous studies have shown that removal or mutation of the DAD domain can result in a constitutively active formin (Han et al., 2009, Gould et al., 2011). Indeed, it has previously been postulated that FMNL1ɣ is itself constitutively active as a result of the sequence retention in its C-terminus (Han et al., 2009). This retention could potentially affect binding of the DAD domain to the DID domain as hydrophobic interactions have been implicated in regulating the binding of the DAD and DID domains in Dia1 (Nezami et al., 2006, Lammers et al., 2008, Han et al., 2009). This could also explain the limited interactions we have observed between FMNL1ɣ and Rho-family GTPases. If this protein is constitutively active, Rho-family GTPase activation may not be required, which would also explain why the ROCK inhibitor Y-27632 did not affect FMNL1ɣ-induced membrane blebbing in the human embryonic kidney cell line 293T (Han et al., 2009).
Future work mutating C-terminal amino acid residues and observing their effect on actin binding could help pinpoint precise points that are imperative to FMNL1 function.
Additionally, this could help reconcile what we observed with FMNL1β, as perhaps this
225 unique sequence contains specific amino acid residues that do impede binding of the
DID and DAD domains, resulting in subsequent activation.
The FMNL1ɣ C-terminus bundles actin to potentiate cancer migration.
The requirement of the entirety of the FMNL1 C-terminus for efficient actin bundling demonstrates that regions of formins other than the FH2 domain are critical for different actin-modifying functions. Truncation of the FMNL1 C-terminu eliminated F- actin bundling activity compared to actin alone. However, there is an increase in F-actin bundling activity when the entire C-terminal region of any alternative splice isoform is present. Further, this bundling activity is significantly enhanced when mediated by the
FMNL1ɣ alternative splice isoform. This is especially interesting as even when the conserved Ile and Lys residues of the FH2 domain are mutated and barbed end binding is inhibited, this alternative splice isoform is still able to bundle F-actin as efficiently as
WT FMNL1ɣ. These data clearly identify the 58 amino acid retention of the ɣ alternative splice isoform as the region responsible for regulating this augmented F-actin bundling activity.
This region of the ɣ alternative splice isoform bears several unique features compared to FMNL1α and FMNL1β. Not only does this retention bear two putative, tandem, overlapping WH2 domains, but it also contains five Phe amino acid residues, repeating one after the other shortly after the WH2 domains. Each of these different amino acid residue sequences could affect the bundling activity of FMNL1ɣ in a variety of ways. In other proteins, WH2 domains have been identified as regulators of actin nucleation and assembly, however, two overlapping WH2 domains have not previously been described (Quinlan et al., 2005, Chhabra and Higgs, 2006, Ahuja et al., 2007,
Liverman et al., 2007, Chereau et al., 2008, Haglund et al., 2010, Gould et al., 2011,
Gaucher et al., 2012, Heimsath and Higgs, 2012). Specifically in formins, WH2 domains
226 have been identified in the C-terminal region of Dia1, DAAM1, FMNL2, FMNL3, and
INF2 (Chhabra and Higgs, 2006, Vaillant et al., 2008, Gould et al., 2011, Heimsath and
Higgs, 2012). These motifs have also been identified in numerous other actin- associated proteins (Dominguez, 2016). While isolated WH2 domains can only prevent
G-actin pointed-end (- end) assembly via monomer subunit binding, their functions are exceedingly varied when interacting with other proteins or other regions of the molecule
(Chaudhry et al., 2010, Carlier et al., 2013, Dominguez, 2016). The WH2 domain itself varies in amino acid residue length (~17-35 amino acid residues) and primarily consists of an amphiphilic α-helix containing at least three aliphatic amino acid residues. These aliphatic residues are found as a pair and then the third is located ~7-10 amino acid residues following them (Paunola et al., 2002, Chereau et al., 2005, Aguda et al., 2006,
Husson et al., 2011, Dominguez, 2016). These paired amino acid residues often consist of leucines, however, in FMNL2 and FMNL3 this pair substitues one isoleucine residue for a leucine residue (Vaillant et al., 2008, Dominguez, 2016). The two tandem leucine residue pairs of FMNL1ɣ are unique among WH2 domain-containing proteins, in addition to the five Phe residues following the trailing aliphatic residues. This unique sequence of
Phe residues is uncommon and could potentially interact with the WH2 domain.
Two overlapping WH2 domains is a unique feature of FMNL1ɣ and could explain the actin-modifying functions we have observed. Much like Spire and FMNL3, these
WH2 domains in FMNL1ɣ could be capping actin filaments, resulting in an inhibition of assembly, which we observed in the pyrene-actin fluorescence assays. Both Spire and
FMNL3 exhibit this function (Bosch et al., 2007, Carlier et al., 2011, Heimsath and Higgs,
2012). FMNL1ɣ may also be nucleating actin filaments via its WH2 domains in a manner similar to COBL which has three WH2 domains (Ahuja et al., 2007).
Furthermore, this could potentially explain the increase in actin we observe in the bundling assays. Severing of actin filaments could also contribute to the increase of
227
FMNL1ɣ-bundled actin. Spire has four WH2 domains through which it not only inhibits profilin-actin binding but also severs filaments (Bosch et al., 2007). How these tandem, overlapping WH2 domains contribute to the bundling process remains to be determined, but we have some hypotheses based on the data included in this dissertation.
It has previously been demonstrated that mouse FMNL1β actually slows elongation from the barbed ends of muscle actin (Harris et al., 2004, Harris et al., 2006).
However, one of these studies also demonstrated that actin assembly from muscle actin monomer subunits is enhanced by FMNL1β (Harris et al., 2006). The data we have presented here clearly show that human FMNL1ɣ inhibits actin filament assembly regardless of FH2 domain interactions. Mouse FMNL1β and human FMNL1ɣ do share some homology, however, the C-terminal regions, which in FMNL1ɣ contains two WH2 domains, vary greatly. We propose that this C-terminal region is an essential component for the formin to interact with actin. We also know that inhibition of barbed end binding by FMNL1ɣ via mutation of I720 and K871 in the FH2 domain inhibits barbed end binding, but yet this mutant inhibits actin assembly at the same rate as WT
FMNL1ɣ (Miller et al., 2017). This suggests that I720 and K871 are not necessary for some formin-actin interactions and that amino acid residues on the “outside” of the FH2 domain, along with the WH2 domains in the tail, are necessary for these interactions.
This presents somewhat of a conundrum in that our actin pull-down assays indicate that binding of FMNL1ɣ to actin requires the FH2 domain-mediated barbed end interactions. This could be explained by the use of two different sources of actin in these experiments. In the pull-down assays, the actin is from whole cell lysate, not purified rabbit muscle actin. While actin isoforms are >90% homologous to each other, the differences in these isoforms could be why we see a difference in these two sets of experiments (Perrin and Ervasti, 2010). Furthermore, there are two isoforms of actin found in the cytoplasm, and since the actin pull-down assays were performed with whole
228 cell lysate, we cannot be certain about the ratios of ɣcyto-actin to βcyto-actin (Perrin and
Ervasti, 2010). How these proteins are modified in the cytoplasm could also affect their binding to actin as when we lysed these cells, this was done using a non-denaturing detergent.
Nonetheless, this inhibition of actin filament assembly clearly demonstrates that
FMNL1 is having an effect on actin, which we were able to demonstrate using subsequent downstream experiments. FMNL1 has been shown to bundle before, but with mouse FMNL1β and not any of the human alternative splice isoforms (Harris et al.,
2006). Our data here demonstrates that FMNL1ɣ bundles actin significantly more efficiently than the other alternative splice isoforms. Moreover, this does not require FH2 domain-mediated barbed end interactions as FMNL1ɣFH2ø can bundle actin at a level similar to WT FMNL1ɣ. Taken together, this evidence suggests that the bundling activity of FMNL1 is mediated by the FH2 domain and further enhanced by the C-terminal region of the FMNL1 alternative splice isoforms. Since FMNL1α and FMNL1β both have the ability to bundle F-actin, in addition to FMNL1ɣ and FMNL1ɣFH2ø, the FH2 domain is implicated in this process, independent of barbed end binding by the conserved Ile and
Lys residues. Most likely, amino acid residues on the “outside” of the FH2 domain, as opposed to the “inside” residues like I720 and K871, are important for the F-actin bundling we have observed. This is further supported by past work demonstrating a similar effect in mouse FMNL1β where the authors hypothesize that hydrophobic patches on the “outer” surface of the FH2 domain mediate bundling activity by the formin
(Harris et al., 2006). While a FH2 dimer dissociation model has been proposed for this bundling activity, the precise mechanism remains to be resolved (Harris et al., 2006). In order to clarify this, future work should focus on mutating the amino acid residues located in the hydrophobic patches of the FH2 domain and observing what, if any, effect
229 occurs on the bundling activity. Additionally, TIRF microscopy and direct visualization of this biochemical function could help further confirm the bundling mechanism.
An explanation for the increased bundling activity observed with FMNL1ɣ lies in the C-terminal region, where two WH2 domains are located. This enhanced bundling activity could be similar to what has been observed with the espin family of proteins, where the presence of WH2 domains greatly increases the observed bundling activity, as opposed to those who lack it (Loomis et al., 2006). This has previously been observed in the FMNL subfamily member FMNL2 as well (Vaillant et al., 2008). How the
WH2 domains are affecting bundling remains to be determined, however, we propose a model where the WH2 domains may contribute to an actin nucleation activity that results in bundling enhancements (Figure 1). Mouse FMNL1β has been shown to be a poor nucleator, however, this protein lacks the WH2 domains present in the FMNL1ɣ alternative splice isoform (Harris et al., 2004, Harris et al., 2006). Therefore, the nucleation activity could be quite different between these two alternative splice isoforms, as WH2 domains are quite proficient at monomer sequestration and nucleation
(Campellone and Welch, 2010, Dominguez, 2016). Another mechanism involving F- actin severing could also be proposed (Figure 2). While mouse FMNL1β has exhibited
FH2 domain-mediated severing, we propose that the WH2 domains of FMNL1ɣ actually contribute to this function (Harris et al., 2004). By using the WH2 domains in the C- terminal region to sever actin filaments, this could generate new filaments of a specific polarity. These new filaments of a preferential polarity could then be bundled by the amino acid residues on the “outside” of the FH2 domain. Furthermore, FMNL1ɣ may prefer to bundle filaments of a specific length, using the WH2 domain to sever filaments to the correct size. While mouse FMNL1β has previously been shown to bundle both parallel and anti-parallel filaments in a mixed orientation, it remains to be determined if
FMNL1ɣ shares the same preference (Harris et al., 2006).
230
Figure 1
Figure 1: Proposed model of bundling mechanism incorporating WH2-mediated actin nucleation. (1) FH2 domains of FMNL1ɣ could initially bundle F-actin via electrostatic interactions regulated by basic patches of amino acid residues on the outside of the FH2 dimer. (2) Following intital bundle formation, FMNL1ɣ could use the
WH2 domain located in the C-terminal region to initate the formation of new actin filaments via a WH2-mediated actin nucleation event. (3) Following nucleation and the assembly of new actin filaments, other FMNL1ɣ FH2 dimers could initiate FH2-mediated bundling, resulting in enhanced bundling formation.
231
Figure 2
Figure 2: Proposed model of bundling mechanism incorporating WH2-mediated actin filament severing. (1) WH2 domains of FMNL1ɣ could interact with actin filaments, possibly via side-binding. (2) Actin filament severing could then occur via the
WH2 domain, resulting in the generation of two actin filaments. (3) FH2-mediated bundling could occur using electrostatic interactions via the basic patches on the outside of the FH2 domain. Due to the new filaments generated via WH2-mediated severing,
FMNL1ɣ may then bundle actin filaments, possibly in a preferred orientation.
Several methods could be utilized in order to determine if the WH2 domains are indeed severing filaments. Severing assays using fluorescence microscopy could be used to compare any activity of FMNL1ɣ to FMNL1α or FMNL1β. Ideally, TIRF microscopy could be used in order to observe this activity in real-time. Moreover, mutations rendering the WH2 domains of FMNL1ɣ inactive could be used and interactions with this mutant and F-actin could be directly observed via TIRF microscopy to determine if severing is indeed inhibited. These assays could be done in conjunction
232 with control assays using well-characterized severing proteins, such as ADF/cofilin or twinfilin, in order to properly evaluate this function (Andrianantoandro and Pollard, 2006).
With FMNL1 having several potential actin-modifying functions, it is not surprising that depletion of FMNL1 expression in MDA-MB-231 cells results in a significant inhibition of cell migration. This is interesting as we have previously observed that knockdown of FMNL1 in primary human macrophages did not result in a significant decrease in average accumulated distance traveled or average velocity in a similar 2D random cell motility assay (Data not shown). However, this might be due more to cellular function, as depletion of FMNL1 expression in primary human macrophages does inhibit macrophage migration across a chemotactic gradient (Mersich et al., 2010,
Miller et al., 2015, Miller et al., 2017). It seems likely that highly motile cancer cells would not require such a gradient in order to induce locomotion as fluctuations in oncogene expression could very well lead to unwarranted cellular motility.
The localization patterns observed with the different FMNL1 alternative splice isoforms can also provide clues to their function. While we observed similar localization patterns with FMNL1α and FMNL1β, localization patterns exhibited by FMNL1ɣ were unique. We were not surprised with the intense localization of FMNL1α or FMNL1β in the perinuclear region. Past studies have shown that this localization is not uncommon as FMNL1 expression has been linked to regulation of actin architecture surrounding the
Golgi complex in HeLa cells (Han et al., 2009, Colon-Franco et al., 2011). However, this was attributed more to FMNL1ɣ and not FMNL1α or FMNL1β (Colon-Franco et al.,
2011). Nonetheless, FMNL1ɣ localization has been observed in the cell membrane in
293T cells while FMNL1α and FMNL1β display more of a cytoplasmic dispersion (Han et al., 2009). This is in agreement with our results, demonstrating enhanced FMNL1ɣ localization at the cell periphery while FMNL1α and FMNL1β display more of a perinuclear-cytoplasmic localization.
233
We propose that FMNL1ɣ mediates actin filament bundling at the lamella- lamellipodia interface (Figure 3). This is supported by past studies demonstrating that
FMNL1ɣ co-localizes with myosin IIb in this same region (Han et al., 2009). This region of the cell is frequently undergoing rapid actin cytoskeletal changes and contains different variations of actin architecture as the lamella is composed of thick actin bundles while the lamellopodia is composed more of a dendritic actin network (Vincente-
Manzanares et al., 2009, Ridley, 2011). It is in this region where myosin IIb could be regulating actomyosin contractility via interaction with FMNL1ɣ-generated actin bundles.
Further, these FMNL1ɣ-generated actin filament bundles may be necessary for adhesion formation, which could explain the defects of adhesion we observe with FMNL1 depletion, as well as the significant increase in adhesion when exogenous FMNL1ɣ is expressed. Additionally, this would also correlate with our past data demonstrating that
FMNL1 co-localizes and co-IPs with the β3 integrin (Mersich et al., 2010). This could also help explain the results of our cellular morphology studies demonstrating that expression of FMNL1ɣ significantly enhances both the perimeter and surface area of
MDA-MB-231 cells. Enhancement of expression of FMNL1ɣ could lead to alterations in actomyosin stress fiber formation and, ultimately, changes in cell shape. Indeed, some formins have exhibited the ability to regulate myosin function in cells (Homem and
Peifer, 2008, Vogler et al., 2014).
234
Figure 3
Figure 3: Proposed functions of FMNL1ɣ in the cell. The WH2 domains of FMNL1ɣ could be enhancing bundling activity at the lamella-lamellipodia interface. This could be important for dorsal stress fiber and transverse arc formation, as well as actomyosin contractility via myo IIb-force generation. Regulation of actomyosin contractility could be important for several functions, including cell migration, adhesion complex assembly, intracellular transport, and cell morphology regulation. FMNL1 may also be regulating bundle formation at the perinuclear region, possibly contributing to polarization events, as well as at the Golgi complex, where it may be important for Golgi organization and localization.
In order to evaluate the relationship between FMNL1ɣ and myosin IIb, a variety of experiments could be performed. Fluorescence microscopy could be utilized to verify co-localization of FMNL1ɣ with myosin IIb in MDA-MB-231 cells. Additionally, if these
235 proteins do interact with each other, co-IPs could help evaluate these protein-protein interactions. It would also be interesting to see if exogenous expression of FMNL1ɣ could increase myosin IIb accumulation in lamella-lamellipodia region of the cell, as the increase of FMNL1ɣ-generated bundles could subsequently result in the requirement of more myosin IIb for the cell to properly mediate the force required for efficient mechanotransduction.
The localization patterns we observed with FMNL1α and FMNL1β clearly indicate that future work should focus on the perinuclear region of the cell. Since FMNL1 has been previously implicated in regulating actin architecture surrounding the Golgi complex, it would make sense to explore what, if any, interactions these two alternative splice isoforms have with this organelle (Colon-Franco et al., 2011). This could be accomplished by using fluorescence microscopy along with Golgi markers like GM130
(cis-Golgi) or Golgin-97 (trans-Golgi). It would also be interesting to see if these two alternative splice isoforms regulate MTOC polarity as FMNL1 has been shown to regulate centrosome polarization in T-cells (Gomez et al., 2007). This could be done by depleting FMNL1 expression in cells and observing any subsequent effects via staining with α-tubulin.
Our evaluation of the role of formins in invasion shows that, when compared to
Dia1, FMNL1 contributes to this process in a unique way. Therefore, FMNL1 may be a potential therapeutic target for inhibition of cancer cell metastasis. Knockdown of
FMNL1 results in inhibition of invasion similar to the pan-formin inhibitor SMIFH2. As we have previously demonstrated that FMNL1 has an insensitivity to SMIFH2, this indicates that perhaps FMNL1 function is more reliant upon the C-terminal region of this protein
(Miller et al., 2015). Therefore, since this formin has a unique region that is responsible for its interactions with actin, this could prove to be a good target for therapeutic intervention. Additionally, the two tandem, overlapping WH2 domains are also unique,
236 further promoting this sequence as an excellent candidate target for different inhibitors.
As a result, future translational research could focus on identifying inhibitors through drug library screens or drug design specifically targeting the two unique WH2 domains in the C-terminal region of FMNL1ɣ.
With the possession of a conditional, FMNL1 knockout mouse, our lab is well- equipped to further study the effects of FMNL1 depletion on breast cancer metastasis in vivo. The p53fl/flMMTV-Cre+/+ mouse model results in mouse mammary tumorigenesis that occurs similarly to humans. Mouse mammary epithelial cells from this model are p53-inactive, resulting in primary mammary tumors and subsequent metastases to the lungs and liver (Lin et al., 2004). Since Cre recombinase activity is under control of the
MMTV promoter, we could potentially delete FMNL1 expression in mouse mammary glands which are also deficient of p53, a powerful tumor suppressor gene. This would allow us to observe the in vivo effects of FMNL1 depletion on the formation of secondary metastatic sites.
Concluding Remarks
The valuable insight provided by this study should greatly help advance the formin field of research. We have provided evidence suggesting that formin protein function is not entirely defined by the characteristic FH2 domain and that other regions of the protein should be thoroughly dissected in order to evaluate all possible actin- modifying abiltiies. This work also demonstrates that the alternative splice isoforms of these proteins be assessed as well. Since all formin genes code for some number of isoforms, it would behoove us to determine which of these isoforms contribute to different cellular processes. Unique isoform sequences, like what we have observed with FMNL1ɣ, could provide putative therapeutic targets for treating different disorders.
237
The major hallmarks of this work demonstrate that an actin-modifying function unique to a specific alternative splice isoform of the formin FMNL1 contributes to breast adenocarcinoma adhesion and migration. We have established a role for FMNL1ɣ distinctive amongst other proteins of this family. FMNL1ɣ regulates cellular locomotion during development, homeostasis, and cancer cell migration. Additionally, the F-actin bundling activity of FMNL1ɣ does not rely on FH2-mediated barbed end binding but on
WH2 domain interactions. Resolution of the precise mechanism driving FMNL1ɣ functions will reveal novel insight into a powerful family of proteins, a highy conserved domain found in numerous other proteins, and potential therapeutic targets for pathologies characterized by cellular locomotion.
238
References
Aguda AH, Xue B, Irobi E, Preat T, Robinson RC (2006). The structural basis of actin interaction with multiple WH2/β-thymosin motif-containing proteins. Structure 14, 469- 476.
Ahuja R, Pinyol R, Reichenbach N, Custer L, Klingensmith J, Kessels MM, Qualmann B (2007). Cordon-bleu is an actin nucleation factor and controls neuronal morphology. Cell 131, 337-350.
Andrianatoandro E and Pollard TD (2006). Mechanism of actin filament turnover by severing and nucleation at different concentrations of ADF/cofilin. Mol Cell 24, 13-23.
Bosch M, Le KH, Bugyi B, Correia JJ, Renault L, Carlier MF (2007). Analysis of the function of Spire in actin assembly and its synergy with formin and profiling. Mol Cell 28, 555-568.
Campellone KG and Welch MD (2010). A nucleator arms race: cellular control of actin assembly. Nat Rev Mol Cell Biol 11, 237-251.
Carlier MF, Husson C, Renault L, Didry D (2011). Control of actin assembly by the WH2 domains and their multifunctional tandem repeats in Spire and Cordon-Bleu. Int Rev Cell Mol Biol 290, 55-85.
Carlier MF, Pernier J, Avvaru BS (2013). Control of actin filament dynamics at barbed ends by WH2 domains: from capping to permissive and processive assembly. Cytoskeleton 70, 540-549.
Chaudhry F, Little K, Talarico L, Quintero-Monzon O, Goode BL (2010). A central role for the WH2 domain of Srv2/CAP in recharging actin monomers to drive actin turnover in vitro and in vivo. Cytoskeleton 67, 120-133.
Chereau D, Boczkowska M, Skwarek-Maruszewska A, Fujiwara I, Hayes DB, Rebowski G, Lappalainen P, Pollard TD, Dominguez R (2008). Leiomodin is an actin filament nucleator in muscle cells. Science 320, 239-243.
Chereau D, Kerff F, Graceffa P, Grabarek Z, Langsetmo K, Dominguez R (2005). Actin- bound structures of Wiskott-Aldrich syndrome protein (WASP)-homology domain 2 and the implications for filament assembly. Proc Natl Acad Sci 102, 16644-16649.
Chhabra ES and Higgs HN (2006). INF2 is a WASP homology 2 motif-containing formin that severs actin filaments and accelerates both polymerization and depolymerization. J Biol Chem 281, 26754-26767.
Colón-Franco JM, Gomez TS, Billadeau DD (2011). Dynamic remodeling of the actin cytoskeleton by FMNL1ɣ is required for structural maintenance of the Golgi complex. J Cell Sci 124, 3118-3126.
239
Dettenhofer M, Zhou F, Leder P (2008). Formin 1-isoform IV deficient cells exhibit defects in cell spreading and focal adhesion formation. PLoS One 3, e2497.
Dominguez R (2016). The WH2 Domain and Actin Nucleation: Necessary but Insufficient. Trends Biochem Sci 41, 478-490.
Favaro P, Traina F, Machado-Neto JA, Lazarini M, Lopes MR, Pereira JK, Costa FF, Infante E, Ridley AJ, Saad ST (2013). FMNL1 promotes proliferation and migration of leukemia cells. J Leukoc Biol 94, 503-512.
Franz CM, Jones GE, Ridley AJ (2002). Cell migration in development and disease. Dev Cell 2, 153-158.
Gardberg M, Heuser VD, Iljin K, Kampf C, Uhlen M, Carpén O (2014). Characterization of leukocyte formin FMNL1 expression in human tissues. J Histochem Cytochem 62, 460-470.
Gardberg M, Talvinen K, Kaipio K, Iljin K, Kampf C, Uhlen M, Carpén O (2010). Characterization of Diaphanous-related formin FMNL2 in human tissue. BMC Cell Biol 11, 55.
Gaucher JF, Mauge C, Didry D, Guichard B, Renault L, Carlier MF (2012). Interactions of isolated C-terminal fragments of neural Wiskott-Aldrich syndrome protein (N-WASP) with actin and Arp2/3 complex. J Biol Chem 287, 34646-34659.
Gomez TS, Kumar K, Medeiros RB, Shimizu Y, Leibson PJ, Billadeau DD (2007). Formins regulate the actin-related protein 2/3 complex-independent polarization of the centrosome to the immunological synapse. Immunity 26, 177-190.
Gould CJ, Maiti S, Michelot A, Graziano BR, Blanchoin L, Goode BL (2011). The formin DAD domain plays dual roles in autoinhibition and actin nucleation. Curr Biol 21, 384- 390.
Haglund CM, Choe JE, Skau CT, Kovar DR, Welch MD (2010). Rickettsia Sca2 is a bacterial formin-like mediator of actin-based motility. Nat Cell Biol 12, 1057-1063.
Han Y, Eppinger E, Schuster IG, Weigand LU, Liang X, Kremmer E, Peschel C, Krackhardt AM (2009). Formin-like 1 (FMNL1) is regulated by N-terminal myristoylation and induces polarized membrane blebbing. J Biol Chem 284, 33409-33417.
Han Y, Yu G, Sarioglu H, Caballero-Martinez A, Schlott F, Ueffing M, Haase H, Peschel C, Krackhardt AM (2013). Proteomic investigation of the interactome of FMNL1 in heamtopoietc cells unveils a role in calcium-dependent membrane plasticity. J Proteomics 78, 72-82.
Harris ES, Gauvin TJ, Heimsath EG, Higgs HN (2010). Assembly of filopodia by the formin FRL2 (FMNL3). Cytoskeleton 67, 755-772.
240
Harris ES, Rouiller I, Hanein D, Higgs HN (2006). Mechanistic differences in actin bundling activity of two mammalian formins, FRL1 and mDia2. J Biol Chem 281, 14383- 14392.
Heimsath EG and Higgs HN (2012). The C terminus of formin FMNL3 accelerates actin polymerization and contains a WH2 domain-like sequence that binds both monomers and filament barbed ends. J Biol Chem 287, 3087-3098.
Homem CC and Peifer M (2008). Diaphanous regulates myosin and adherens junctions to control cell contractility and protrusive behavior during morphogenesis. Development 135, 1005-1018.
Husson C, Renault L, Didry D, Pantaloni D, Carlier MF (2011). Cordon-Bleu uses WH2 domains as multifunctional dynamizers of actin filament assembly. Mol Cell 43, 464-477.
Iskratsch T, Lange S, Dwyer J, Kho AL, dos Remedios C, Heler E (2010). Formin follows function: a muscle-specific isoform of FHOD3 is regulated by CK2 phsophorylation and promotes myofibril maintenance. J Cell Biol 191, 1159- 1172.
Kage F, Winterhoff M, Dimchev V, Mueller J, Thalheim T, Freise A, Bruhmann S, Kollasser J, Block J, Dimchev G, Geyer M, Schnittler HJ, Brakebusch C, Stradal TE, Carlier MF, Sixt M, Faix J, Rottner K (2017). FMNL formins boost lamellipodial force generation. Nat Commun 8, 14832.
Katoh M and Katoh M (2003). Identification and characterization of human FMNL1, FMNL2, and FMNL3 genes in silico. Int J Oncol 22, 1161-1168.
Kim AS, Kakalis LT, Abdul-Manan N, Liu GA, Rosen MK (2000) Autoinhibition and activation mechanisms of the Wiskott-Aldrich syndrome protein. Nature 404, 151-158.
Lammers M, Meyer S, Kuhlmann D, Wittinghofer A (2008). Specificity of interactions between mDia isoforms and Rho proteins. J Biol Chem 283, 35236-35246.
Leber MF and Efferth T (2009). Molecular principles of cancer invasion and metastasis. Int J Oncol 34, 881-895.
Lin SC, Lee KF, Nikitin AY, Hilsenbeck SG, Cardiff RD, Li A, Kang KW, Frank SA, Lee WH, Lee EY (2004). Somatic mutation of p53 leads to estrogen receptor alpha-positive and –negative mouse mammary tumors with high frequency metastasis. Cancer Res 64, 3525-3532.
Liverman AD, Cheng HC, Trosky JE, Leung DW, Yarbrough ML, Burdette DL, Rosen MK, Orth K (2007). Arp2/3-independent assembly of actin by Vibrio type III effector VopL. Proc Natl Acad Sci USA 104, 17117-17122.
Loomis PA, Kelly AE, Zheng L, Changyaleket B, Sekerková G, Mugnaini E, Ferreira A, Mullins RD, Bartles JR (2006). Targeted wild-type and jerker espins reveal a novel, WH2-domain-dependent way to make actin bundles in cells. J Cell Sci 119, 1655-1665.
241
Mersich AT, Miller MR, Chkourko H, Blystone SD (2010). The formin FRL1 (FMNL1) is an essential component of macrophage podosomes. Cytoskeleton 67, 573-585.
Miller MR, Miller EW, Blystone SD (2017). Non-canonical activity of the podosomal formin FMNL1ɣ supports immune cell migration. J Cell Sci 130, 1730-1739.
Naj X, Hoffmann AK, Himmel M, Linder S (2013). The formins FMNL1 and mDia1 regulate coiling phagocytosis of Borrelia burgdorferi by primary human macrophages. Infect Immun 81, 1683-1695.
Nezami AG, Poy F, Eck MJ (2006). Structure of the autoinhibitory switch in formin mDia1. Structure 14, 257-263.
Otomo T, Tomchick DR, Otomo C, Panchal SC, Machius M, Rosen MK (2005). Structural basis of actin filament nucleation and processive capping by a formin homology 2 domain. Nature 433, 488-494.
Paunola E, Mattila PK, Lappalainen P (2002). WH2 domain: a small, versatile adapter for actin monomers. FEBS Lett 513, 92-97.
Péladeau C, Heibein A, Maltez MT, Copeland SJ, Copeland JW (2016). A specific FMNL2 isoform is up-regulated in invasive cells. BMC Cell Biol 17, 32.
Perrin BJ and Ervasti JM (2010). The actin gene family: function follows isoform. Cytoskeleton 67, 630-634.
Quinlan ME, Heuser JE, Kerkhoff E, Mullins RD (2005). Drosophila Spire is an actin nucleation factor. Nature 433, 382-388.
Ridley AJ (2011). Life at the leading edge. Cell 145, 1012-1022.
Schönichen A and Geyer M (2010). Fifteen formins for an acti filament: a molecular view on the regulation of human formins. Biochim Biophys Acta 1803, 152-163.
Shimada A, Nyitrai M, Vetter IR, Kühlmann D, Bugyi B, Narumiya S, Geeves MA, Wittinghofer A (2004). The core FH2 domain of diaphanous-related formins in an elongated actin binding protein that inhibits polymerization. Mol Cell 13, 511-522.
Singh RB, Mengi SA, Xu YJ, Arneja AS, Dhalla NS (2002). Pathogenesis of atherosclerosis: A multifactorial process. Exp Clin Cardiol 7, 40-53.
Thompson ME, Heimsath EG, Gauvin TJ, Higgs HN, Kull FJ (2013). FMNL3 FH2-actin structure gives insight into formin-mediated actin nucleation and elongation. Nat Struct Mol Biol 20, 111-118.
Ti SC, Jurgenson CT, Nolen BJ, Pollard TD (2011). Structural and biochemical characterization of two binding sites for nucleation-promoting factor WASp-VCA on Arp2/3 complex. Proc Natl Acad Sci 108, 463-471.
242
Vaillant DC, Copeland SJ, Davis C, Thurston SF, Abdennur N, Copeland JW (2008). Interaction of the N- and C-terminal autoregulatory domains of FRL2 does not inhibit FRL2 activity. J Biol Chem 283, 33750-33762.
Vicente-Manzanares M, Ma X, Adelstein RS, Horwitz AR (2009). Non-muscle myosin II takes center stage in cell adhesion and migration. Nat Rev Mol Cell Biol 10, 778-790.
Vogler G, Liu J, Iafe TW, Migh E, Miháky J, Bodmer R (2014). Cdc42 and formin activity control non-muscle myosin dynamics during Drosophila heart morphogenesis. J Cell Biol 206, 909-922.
Xu Y, Moseley JB, Sagot I, Poy F, Pellman D, Goode BJ, Eck MJ (2004). Crystal structures of a Formin-Homology 2 domain reveal a tethered dimer architecture. Cell 116, 711-723.
Yamaguchi H and Condeelis J (2007). Regulation of the actin cytoskeleton in cancer cell migration and invasion. Biochim Biophys Acta 1773, 642-652.
Zuidscherwoude M, Green HLH, Tomas SG (2018). Formin proteins in megakaryoctyes and platelets: regulation of actin and microtubule dynamics. Platelets 18, 1-8.
243
244