(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



SNF chromatin remodeling assemblies BAF and PBAF differentially regulate cell cycle exit and cellular invasion in vivo

['Jayson J. Smith', 'Department Of Biochemistry', 'Cell Biology', 'Stony Brook University', 'Stony Brook', 'New York', 'United States Of America', 'Yutong Xiao', 'Nithin Parsan', 'Massachusetts Institute Of Technology']

Date: 2022-03 

Chromatin remodelers such as the SWI/SNF complex coordinate metazoan development through broad regulation of chromatin accessibility and transcription, ensuring normal cell cycle control and cellular differentiation in a lineage-specific and temporally restricted manner. Mutations in genes encoding the structural subunits of chromatin, such as histone subunits, and chromatin regulating factors are associated with a variety of disease mechanisms including cancer metastasis, in which cancer co-opts cellular invasion programs functioning in healthy cells during development. Here we utilize Caenorhabditis elegans anchor cell (AC) invasion as an in vivo model to identify the suite of chromatin agents and chromatin regulating factors that promote cellular invasiveness. We demonstrate that the SWI/SNF ATP-dependent chromatin remodeling complex is a critical regulator of AC invasion, with pleiotropic effects on both G 0 cell cycle arrest and activation of invasive machinery. Using targeted protein degradation and enhanced RNA interference (RNAi) vectors, we show that SWI/SNF contributes to AC invasion in a dose-dependent fashion, with lower levels of activity in the AC corresponding to aberrant cell cycle entry and increased loss of invasion. Our data specifically implicate the SWI/SNF BAF assembly in the regulation of the G 0 cell cycle arrest in the AC, whereas the SWI/SNF PBAF assembly promotes AC invasion via cell cycle-independent mechanisms, including attachment to the basement membrane (BM) and activation of the pro-invasive fos-1/FOS gene. Together these findings demonstrate that the SWI/SNF complex is necessary for two essential components of AC invasion: arresting cell cycle progression and remodeling the BM. The work here provides valuable single-cell mechanistic insight into how the SWI/SNF assemblies differentially contribute to cellular invasion and how SWI/SNF subunit-specific disruptions may contribute to tumorigeneses and cancer metastasis.

Cellular invasion is required for animal development and homeostasis. Inappropriate activation of invasion however can result in cancer metastasis. Invasion programs are orchestrated by complex gene regulatory networks (GRN) that function in a coordinated fashion to turn on and off pro-invasive genes. While the core of GRNs are DNA binding transcription factors, they require aid from chromatin remodelers to access the genome. To identify the suite of pro-invasive chromatin remodelers, we paired high resolution imaging with RNA interference to individually knockdown 269 chromatin factors, identifying the evolutionarily conserved SWItching defective/Sucrose Non-Fermenting (SWI/SNF) ATP-dependent chromatin remodeling complex as a new regulator of Caenorhabditis elegans anchor cell (AC) invasion. Using a combination of CRISPR/Cas9 genome engineering and targeted protein degradation we demonstrate that the core SWI/SNF complex functions in a dose-dependent manner to control invasion. Further, we determine that the accessory SWI/SNF complexes, BAF and PBAF, contribute to invasion via distinctive mechanisms: BAF is required to prevent inappropriate proliferation while PBAF promotes AC attachment and remodeling of the basement membrane. Together, our data provide insights into how the SWI/SNF complex, which is mutated in many human cancers, can function in a dose-dependent fashion to regulate switching from invasive to proliferative fates.

Funding: D.Q.M. is a Damon Runyon-Rachleff Innovator supported by the Damon Runyon Cancer Research Foundation [DRR-47-17]. This work is also supported by research grants to D.Q.M. from the National Institute of General Medical Sciences (NIGMS) [R01GM121597]. J.J.S. received support from the W. Burghardt Turner Fellowship. J.J.S. [3R01GM121597-02S1], MA.Q.M [3R01GM121597-03S1] and F.E.Q.M. [3R01GM121597-04S1] were supported through NIGMS Diversity Supplements to R01GM121597. M.A.Q.M. was also supported by the National Cancer Institute (NCI) 1F30CA257383-01A1. A.Q.K. was supported by NIGMS [F31GM128319]. R.C.A. was supported by NIGMS [F32GM1283190]. T.N.M-K. was supported by NICHD [F31HD100091]. N.J.P. was supported by the American Cancer Society [132969-PF-18-226-01-CSM]. P.K. was supported by the National Institute of Neurological Disorders and Stroke [R01NS118078]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

A comprehensive investigation of the regulatory mechanism(s) governing AC invasion should include a thorough description of the suite of chromatin agents and chromatin regulating factors that are required for G 0 /G 1 cell cycle arrest and invasive differentiation in the AC. In this study we perform an RNA interference (RNAi) screen in C. elegans, specifically focusing on genes involved in chromatin structure and remodeling or histone modification (collectively called “chromatin factors”). We identify 82 chromatin factors whose transcriptional depletion resulted in significant AC invasion defects. Among the 82 hits recovered in the screen, the SWI/SNF complex emerged as the most well-represented single chromatin remodeling complex. RNAi knockdown of subunits specific to the SWI/SNF core (swsn-1 and snfc-5/swsn-5), and both BAF (BRG/BRM-Associated Factors; swsn-8/let-526) and PBAF (Polybromo Associated BAF; pbrm-1 and swsn-7) assemblies resulted in penetrant loss of AC invasion. We generated fluorescent reporter knock-in alleles of subunits of the core (GFP::swsn-4) and BAF (swsn-8::GFP) assembly of the SWI/SNF complex using CRISPR/Cas9-mediated genome engineering. These alleles, used in conjunction with an endogenously labeled PBAF (pbrm-1::eGFP) assembly subunit, enabled us to determine the developmental dynamics of the SWI/SNF ATPase and assembly-specific subunits, gauge the efficiency of various SWI/SNF knockdown strategies, and assess intra-complex and inter-assembly regulation. Using improved RNAi constructs and an anti-GFP nanobody degradation strategy [ 31 ], we demonstrated that the cell autonomous contribution of the SWI/SNF complex to AC invasion is dose dependent. This finding parallels similar studies in cancer [ 32 – 35 ] and C. elegans mesoblast development [ 36 ]. Surprisingly, examination using a CDK activity sensor [ 37 ] revealed assembly-specific contributions to AC invasion: whereas BAF promotes AC invasion in a cell cycle-dependent manner, PBAF contributes to invasion in a cell cycle-independent manner. Finally, we utilized the auxin-inducible degron (AID) system combined with PBAF RNAi to achieve strong combinatorial PBAF subunit depletion in the AC, which resulted in loss of both AC invasion and adhesion to the BM. Together, these findings provide insight into how the SWI/SNF complex assemblies may contribute to distinct aspects of proliferation and metastasis in human malignancies.

Cell-extrinsic and cell-intrinsic factors, such as chromatin remodeling complexes and TFs, control many aspects of cell fate from plasticity to terminal differentiation and cell cycle arrest. This decision between plasticity and specification is in part the consequence of a complex, genome-wide antagonism between Polycomb group (PcG) transcriptional repression and Trithorax group (TrxG) transcriptional activation [ 25 – 27 ]. One example of this is the binding of pioneer TFs OCT4 and SOX2 to target DNA in order to retain pluripotency in murine embryonic stem cells; the association of these TFs with their targets has been characterized as an indirect consequence of chromatin accessibility at these target regions [ 28 ]. A recent study has shown that chromatin accessibility of enhancers in crucial cell cycle genes which promote the G 1 /S transition, including Cyclin E and E2F transcription factor 1, is developmentally restricted to reinforce terminal differentiation and cell cycle exit during Drosophila melanogaster pupal wing morphogenesis [ 29 ]. In C. elegans myogenesis, the SWItching defective/Sucrose Non-Fermenting (SWI/SNF) ATP-dependent chromatin remodeling complex, a member of the TrxG family of complexes, both regulates the expression of the MyoD transcription factor (hlh-1) and acts redundantly to promote differentiation and G 0 cell cycle arrest with several core cell cycle regulators including cullin 1 (CUL1/cul-1), cyclin-dependent kinase inhibitor 1 (cki-1), FZR1 (fzr-1), and the RB transcriptional corepressor (RBL1/lin-35) [ 30 ]. The importance of the dynamic regulation of chromatin states for the acquisition and implementation of differentiated behaviors is also reflected in the C. elegans AC, as previous work has shown that the histone deacetylase hda-1 (HDAC1/2) is required for pro-invasive gene expression and therefore the differentiated behavior of cellular invasion [ 24 ] ( Fig 1B ).

Previous work demonstrated a high degree of evolutionary conservation in the cell-autonomous mechanisms underlying BM invasion [ 3 , 15 ], including basolateral polarization of the F-actin cytoskeleton/cytoskeletal regulators and the expression of matrix metalloproteinases (MMPs) [ 16 – 21 ]. Moreover, in order to breach the BM, the AC requires the expression of transcription factors (TFs), whose human homologs are common to metastatic cancers, including egl-43 (EVI1/MEL), fos-1 (FOS), hlh-2 (E/Daughterless), and nhr-67 (TLX/Tailless) [ 22 ] ( Fig 1B ). In addition to the expression of pro-invasive genes, there is increasing evidence that cells must also arrest in the cell cycle to adopt an invasive phenotype [ 23 ]. Our previous work has demonstrated that the AC must terminally differentiate and arrest in the G 0 /G 1 phase of the cell cycle to invade the BM and make contact with the underlying primary vulval precursor cells (1° VPCs) [ 22 , 24 ]. The regulatory mechanisms that couple G 0 /G 1 cell cycle arrest with the ability of a cell to invade the BM remain unclear.

(A) Schematic depicting AC invasion in the mid-L3 stage of C. elegans development (left) and micrographs demonstrating the coordination of AC (magenta, cdh-3p::PH::mCherry) invasion through the BM (green, laminin::GFP) with primary vulval development in the uterine-specific RNAi hypersensitive background used in the chromatin factor RNAi screen. The fluorescent AC-specific membrane marker and BM marker are overlaid on DIC in each image. White arrowheads indicate ACs, yellow arrowheads indicate boundaries of breach in the BM, and white brackets indicate 1° VPCs. Scale bar, 5 μm. (B) Overview of the transcription factor GRN governing AC invasion [ 22 , 24 ], which consists of cell cycle-independent (fos-1) and dependent (egl-43, hlh-2, and nhr-67) subcircuits, which together with hda-1 promote pro-invasive gene expression and maintain cell cycle arrest in the AC.

A variety of in vitro and in vivo models have been developed to study the process of cellular invasion at the genetic and cellular levels. In vitro invasion assays typically involve 3D hydrogel lattices, such as Matrigel, through which cultured metastatic cancer cells will invade in response to chemo-attractants [ 7 ]. Recently, microfluidic systems have been integrated with collagen matrices to improve these in vitro investigations of cellular invasion [ 8 ]. While in vitro invasion models provide an efficient means to study the mechanical aspects of cellular invasion, they are currently unable to replicate the complex microenvironment in which cells must invade during animal development and disease. A variety of in vivo invasion models have been studied, including cancer xenograft models in mouse [ 9 – 11 ] and zebrafish [ 12 , 13 ], each having their respective benefits and drawbacks. Over the past ~15 years, Caenorhabditis elegans anchor cell (AC) invasion has emerged as a powerful alternative model due to its visually tractable single-cell nature ( Fig 1A ) [ 14 ].

Cellular invasion through basement membranes (BMs) is a critical step in metazoan development and is important for human health and fitness. Early in hominid development, trophoblasts must invade into the maternal endometrium for proper blastocyst implantation [ 1 ]. In the context of immunity, leukocytes become invasive upon injury or infection to travel between the bloodstream and interstitial tissues [ 2 , 3 ]. Atypical activation of invasive behavior is associated with a variety of diseases, including rheumatoid arthritis wherein fibroblast-like synoviocytes adopt invasive cellular behavior, leading to the expansion of arthritic damage to previously unaffected joints [ 4 , 5 ]. Aberrant activation of cell invasion is also one of the hallmarks of cancer metastasis [ 6 ].

Results

An RNAi screen of 269 chromatin factors identifies SWI/SNF as a key regulator of AC invasion To identify the suite of chromatin factors that, along with hda-1, contribute to AC invasion, we generated an RNAi sub-library of 269 RNAi clones from the complete Vidal RNAi library and a subset of the Ahringer RNAi library [38, 39] targeting genes implicated in chromatin state, chromatin remodeling, or histone modification (Fig 1B and S1 Table). Because chromatin regulatory factors act globally to control gene expression, we screened each RNAi clone by high-resolution differential interference contrast (DIC) and epifluorescence microscopy in a uterine-specific RNAi hypersensitive background containing labeled BM (laminin::GFP) and an AC reporter (cdh-3p::PH::mCherry) (Fig 1A and S1 Table) [14, 22, 24, 40]. This genetic background allowed us to limit the effect of RNAi transcriptional knockdown of chromatin factors to the AC and the neighboring uterine tissue, and only for a time period following the specification of the AC [40]. As the neighboring uterine cells do not contribute to the invasion program [14], AC invasion defects following RNAi treatments in this background are indicative of cell autonomous pro-invasive gene function [24, 40]. In wild-type animals, by the time the 1° fated P6.p vulval precursor cell has divided twice (P6.p 4-cell stage), 100% of ACs have successfully breached the underlying BMs and made contact with the P6.p grand-daughters [14]. Similarly, we found that all ACs invaded in the uterine-specific RNAi hypersensitive strain used in our RNAi screen, though we observed a low penetrance of ACs with a delay in the timing of invasion, such that at the P6.p 4-cell stage, when we scored invasion, 2% (2/100 animals) still had an intact BM. Thus, we used this baseline defect as a statistical reference point for this genetic background. We defined the cut-off threshold for significant defects in invasion following RNAi treatment as those RNAi clones that resulted in loss of invasion in at least ~13% of treated animals (4/30 animals, Fisher’s exact test = 0.0252). By this threshold, we recovered 82 chromatin factors (30.5% of total screened) that significantly regulate AC invasion (S2 Table). The finding that loss of nearly a third of the chromatin factors included in the RNAi screen results in significant AC invasion defects suggests a general requirement for regulation of chromatin states in the acquisition of invasive behavior. Interestingly, five subunits of the broadly conserved SWI/SNF chromatin remodeling complex were recovered as significant regulators of AC invasion: swsn-1(SMARCC1/SMARCC2; 23% AC invasion defect), swsn-5/snfc-5 (SMARCB1; 20% AC invasion defect), swsn-7 (ARID2; 23% AC invasion defect), swsn-8/let-526 (ARID1A/ARID1B; 23% AC invasion defect), and pbrm-1 (PBRM1; 20% AC invasion defect) (S2 Table). As such, SWI/SNF is well-represented among the roster of significant regulators of AC invasion identified in the screen, with representation of the core (swsn-1 and swsn-5), BAF (swsn-8) and PBAF (pbrm-1 and swsn-7) assemblies. Given the prevalence of SWI/SNF subunits recovered as significant regulators of AC invasion in our RNAi screen and the crucial role SWI/SNF plays in the regulation of animal development [41–46], tumorigenesis [33, 47–49], and cell cycle control [30, 36, 50–52], we chose to focus our efforts on characterizing the role of the SWI/SNF complex in promoting AC invasion. To confirm our RNAi results implicating the SWI/SNF complex as an activator of AC invasion, we obtained two temperature sensitive hypomorphic alleles, swsn-1(os22) and swsn-4(os13) [42], and scored for defects in AC invasion in a genetic background containing both BM (laminin::GFP) and AC (cdh-3p::mCherry::moeABD) reporters. While we observed no defects in AC invasion in animals grown at the permissive temperature (15°C) (S1A Fig), animals containing hypomorphic alleles for core subunits swsn-1 and swsn-4 cultured at the restrictive temperature (25°C) displayed defects in 20% (10/50) and 24% (12/50) of animals, respectively (S1B Fig). These data with the swsn-1(os22) allele corroborated our swsn-1(RNAi) data from the chromatin factor RNAi screen. Additionally, since neither of the RNAi libraries used to compose the chromatin factor screen in this study (see above) contained a swsn-4(RNAi) clone, results with the swsn-4(os13) allele also complement data from our RNAi screen by suggesting that AC invasion depends on the expression of the sole C. elegans SWI/SNF ATPase subunit in addition to the 5 subunits identified in the screen.

C. elegans SWI/SNF subunits exhibit intra-complex and low levels of inter-assembly cross-regulation Work in cell culture has revealed that the mammalian SWI/SNF (mSWI/SNF) complex is assembled in a step-wise fashion, with stability of the complex as a whole and association of individual subunits depending on the prior expression and association of other subunits [64]. To date it is unknown whether in C. elegans individual SWI/SNF subunits activate other SWI/SNF subunits. It is also unclear whether subunits of the two assemblies in C. elegans–BAF and PBAF–stabilize the core protein subunits or vice-versa. Therefore, we used our endogenously labeled SWI/SNF::GFP strains to ask whether transcriptional knockdown of individual subunits of the core, BAF, or PBAF induce changes in protein expression of other subunits at the time of AC invasion (S3 Fig). First, to determine whether representative subunits of the SWI/SNF assemblies promote or stabilize the ATPase of the complex, we treated GFP::swsn-4 animals with either swsn-8(RNAi) or pbrm-1(RNAi) (S3A Fig). Quantification of fluorescence expression in AC nuclei of swsn-8(RNAi) treated animals at the P6.p 4-cell stage revealed significantly lower GFP::SWSN-4 levels relative to the control group (34% GFP::SWSN-4 depletion) (S3A and S3D Fig). RNAi knockdown of the PBAF subunit pbrm-1 also resulted in a significant but weaker loss of ATPase expression in the AC (11% GFP::SWSN-4 depletion) (S3A and S3D Fig). These results suggest that individual subunits of either SWI/SNF assembly exhibit inter-complex regulation and may contribute to the protein stability and/or expression of the SWI/SNF ATPase in the C. elegans AC, with the BAF complex playing a potentially dominant activating role with respect to the ATPase. Next, we treated animals containing either the swsn-8 or pbrm-1 endogenous GFP-reporters with enhanced RNAi to knockdown the expression of the SWI/SNF ATPase or the representative subunit of the alternative SWI/SNF assembly. Interestingly, while unaffected by knockdown of the PBAF assembly subunit pbrm-1, RNAi knockdown of the ATPase swsn-4 resulted in a 42% increase in the expression of SWSN-8::GFP in the AC (S3B, S3C and S3D Fig). Finally, relative to the expression of the endogenous PBAF subunit in the ACs of control animals, AC nuclei of PBRM-1::eGFP animals treated with swsn-4(RNAi) had significantly lower levels of protein expression (38% PBRM-1::eGFP depletion), whereas ACs in swsn-8(RNAi) treated animals expressed 13% more PBRM-1::eGFP (S3C and S3D Fig). Since knockdown of either swsn-4 or swsn-8 subunits resulted in two distinct AC phenotypes–individual animals with single non-invasive ACs and animals with multiple non-invasive cells expressing the AC-reporter—we next sought to determine whether these two phenotypes were distinct with respect to SWI/SNF subunit expression. To do this, we binned data from the intra-complex RNAi experimental series (S3A, S3B, S3C and S3D Fig) into the two non-invasive phenotypes and compared the fluorescence expression levels of the endogenous proteins within SWI/SNF RNAi conditions. Given the infrequency of the multi-AC phenotype, statistical comparisons were necessarily limited to treatments in which the population of animals contained at least 10 multi non-invasive AC events. Treatment of SWSN-8::GFP with swsn-4(RNAi) resulted in a total of 24 multi non-invasive ACs (53 ACs total; n = 41 animals) and no significant difference was detected in SWSN-8::GFP expression between the nuclei of the single non-invasive AC phenotype and the multi non-invasive AC phenotype groups (S3E Fig). The second statistical comparison was made between the two phenotypes in PBRM-1::eGFP animals treated with swsn-8(RNAi) (S3E Fig), in which 14 multi non-invasive ACs were detected (51 ACs total; n = 42 animals). Quantification of endogenous PBRM-1::eGFP fluorescence expression in this condition revealed a slight (12%) increase in expression of the PBAF subunit in the nuclei of ACs of the multi non-invasive phenotype group compared to the single non-invasive phenotype (S3E Fig), reflecting the general increase in PBRM-1 levels detected in the non-binned data (S3D Fig). Based on these results, a tentative model for epistatic interactions between the SWI/SNF ATPase, BAF, and PBAF assembly subunits can be composed for the AC (S3F Fig). Our data indicate that some degree of SWI/SNF intra-complex and inter-assembly regulation occurs in the AC. We find that the most significant aspect of SWI/SNF intra-complex regulation is exercised by the ATPase on the assembly specific subunits, where swsn-4 knockdown results in a significant increase in BAF/SWSN-8 and a significant decrease PBAF/PBRM-1. SWI/SNF inter-assembly regulation appears to be weaker in the AC as knockdown of pbrm-1 does not affect SWSN-8::GFP expression, and knockdown of swsn-8 results in a slight increase in PBRM-1::GFP expression.

Improved swsn-4(RNAi) vector is sufficient to recapitulate a null phenotype in the M lineage A recent study focusing on cell cycle control of SWI/SNF throughout C. elegans muscle and epithelial differentiation demonstrated tissue and lineage-specific phenotypes following weak or strong loss of core SWI/SNF subunits [36]. Within the M lineage that gives rise to posterior body wall muscles (BWMs), coelomocytes (CCs), and reproductive muscles or sex myoblast (SMs) descendants, different cell types responded differently to loss of SWI/SNF. In the BWM, strong loss of SWI/SNF resulted in hyperproliferation, like the phenotype we detect in the AC. The opposite is true in the SM lineage, where modest knockdown of swsn-4 resulted in hyperproliferation while complete loss of swsn-4 expression resulted in a null phenotype where SMs failed to divide and arrest in S phase [36]. We next sought to validate the strength of our enhanced swsn-4(RNAi) vector by examining the SM proliferative state. To accomplish this, we treated animals containing a lineage-restricted cyclin-dependent kinase (CDK) activity sensor (unc-62p::DHB::2xmKate2) with swsn-4(RNAi) (S4A Fig). In this genetic background, we determined the number (S4B Fig) and cell cycle state (S4C Fig) of SM cells at a time when the majority of SMs in control animals had finished cycling and subsequently differentiated (late P6.p 8-cell stage; 16 SM cell stage). Animals treated with swsn-4(RNAi) had significantly fewer SM cells than controls (mean SMs/animals = 5; n = 31 animals) (S4B Fig) with many instances of SMs that failed to enter a single round of cell division (n = 20 single SMs out of 43 animals). Interestingly, 28% (12/43) of animals treated with swsn-4(RNAi) were absent of SMs on either the left or the right side, whereas 100% (30/30) control animals had SMs on both sides, which may indicate a defect in specification, early cell division, and/or migration of SMs. To quantify cell cycle state, we measured localization of an SM-specific CDK sensor, which uses a fragment of mammalian DNA Helicase B (DHB) fused to two copies of mKate2 [37, 70]. In cells with low CDK activity that are quiescent or post-mitotic, the ratiometric CDK sensor is strongly nuclear localized [37, 68, 70]. In cycling cells with increasing CDK activity, the CDK sensor progressively translocates from the nucleus to the cytosoplasm, with a ratio approaching 1.0 in S phase and >1 in cells in G 2 [37]. Thus, the cytoplasmic:nuclear (C/N) ratio of DHB::2xmKate2 can serve as a proxy to identify cell cycle state. By the time the majority of SMs in the control condition were differentiating and arrested in a G 0 cell cycle state (mean C/N ratio = 0.320; n = 90 SMs) (S4C Fig), many animals treated with swsn-4(RNAi) had single SMs that failed to divide and a mean DHB C/N ratio indicative of a long pause or arrest in S phase [37] (Avg. C/N ratio = 0.803; n = 20 SMs) (S4C Fig). Together, these results suggested that the strength of our enhanced swsn-4(RNAi) targeting vector is sufficient to recapitulate a swsn-4 null condition in the SM lineage, as we detected both the hypoproliferative phenotype and S-phase arrest that was observed using a lineage-restricted catalytically inactive SWI/SNF ATPase [36].

[END]

[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1009981

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/