Abstract
CARM1 is a cancer-relevant protein arginine methyltransferase that regulates many aspects of transcription. Its pharmacological inhibition is a promising anti-cancer strategy. Here SKI-73 is presented as a CARM1 chemical probe with pro-drug properties. SKI-73 can rapidly penetrate cell membranes and then be processed into active inhibitors, which are retained intracellularly with 10-fold enrichment for days. These compounds were characterized for their potency, selectivity, modes of action, and on-target engagement. SKI-73 recapitulates the effect of CARM1 knockout against breast cancer cell invasion. Single-cell RNA-seq analysis revealed that the SKI-73-associated reduction of invasiveness act via altering epigenetic plasticity and suppressing the invasion-prone subpopulation. Interestingly, SKI-73 and CARM1 knockout alter the epigenetic plasticity with remarkable difference, arguing distinct modes of action between the small-molecule and genetic perturbation. We therefore discovered a CARM1-addiction mechanism of cancer metastasis and developed a chemical probe to target this process.
Introduction
Numerous biological events are orchestrated epigenetically upon defining cellular fates.(Atlasi and Stunnenberg, 2017; Berdasco and Esteller, 2019) Among key epigenetic regulators are protein methyltransferases (PMTs), which can render downstream signals by modifying specific Arg or Lys residues of their substrates with S-adenosyl-L-methionine (SAM) as a methyl donor cofactor.(Luo, 2018) Significant efforts have been made to identify the PMT-dependent epigenetic cues that are dysregulated or addicted under specific disease settings such as cancer.(Berdasco and Esteller, 2019) Many PMTs are implicated as vulnerable targets against cancer malignancy.(Kaniskan et al., 2018; Luo, 2018) The pro-cancerous mechanism of these PMTs can be attributed to their methyltransferase activities via individual or combined effects of upregulating oncogenes, down-regulating tumor suppressors, and maintaining cancer-cell-addicted homeostasis.(Berdasco and Esteller, 2019; Blanc and Richard, 2017) Pharmacological inhibition of these epigenetic events thus presents promising anti-cancer strategies,(Berdasco and Esteller, 2019) as exemplified by the development of the clinical inhibitors of DOT1L,(Bernt et al., 2011; Daigle et al., 2011) EZH2(Kim et al., 2013; Konze et al., 2013; McCabe et al., 2012; Qi et al., 2012; Qi et al., 2017), and PRMT5.(Bonday et al., 2018; Chan-Penebre et al., 2015)
Protein arginine methyltransferases (PRMTs) act on their substrates to yield three different forms of methylated arginine: asymmetric dimethylarginine (ADMA), symmetric dimethylarginine (SDMA), and monomethylarginine (MMA)---the terminal products of Type I, II and III PRMTs, respectively.(Blanc and Richard, 2017; Yang and Bedford, 2013) Among the important Type I PRMTs is CARM1 (PRMT4), which regulates multiple aspects of transcription by methylating diverse targets including RNAPII, SRC3, C/EBPβ, PAX3/7, SOX2/9, RUNX1, Notch1, p300, CBP, p/CIP, Med12, and BAF155.(Blanc and Richard, 2017; Hein et al., 2015; Vu et al., 2013; Wang et al., 2015; Wang et al., 2014; Yang and Bedford, 2013) The physiological function of CARM1 has been linked to differentiation and maturation of embryonic stem cells to immune cells, adipocytes, chondrocytes, myocytes, and lung tissues.(Blanc and Richard, 2017; Yang and Bedford, 2013) The requirement of CARM1 is implicated in multiple cancers with its methyltransferase activity particularly addicted by hematopoietic malignancies and metastatic breast cancer.(Drew et al., 2017; Greenblatt et al., 2018; Nakayama et al., 2018; Wang et al., 2014) Our prior efforts using in vivo mouse and in vitro cell models uncovered the role of CARM1 in promoting breast cancer metastasis.(Wang et al., 2014) Mechanistically, CARM1 methylates Arg1064 of BAF155 and thus facilitates the recruitment of the BAF155-containing SWI/SNF complex to a specific subset of gene loci essential for breast cancer metastasis. CARM1 thus emerges as a novel anti-cancer target.(Wang et al., 2014)
While this cancer relevance inspired the development of CARM1 inhibitors,(Kaniskan et al., 2018; Scheer et al., 2019) many small-molecule CARM1 inhibitors lack target selectivity or cellular activity (Kaniskan et al., 2018)---two essential criteria of chemical probes.(Frye, 2010) To the best of our knowledge, EZM2302,(Drew et al., 2017; Greenblatt et al., 2018) TP-064(Nakayama et al., 2018) and SKI-73 (www.thesgc.org/chemical-probes/SKI-73) are the only selective and cell-active CARM1 chemical probes, which were developed by Epizyme, Takeda/SGC(Structural Genomic Consortium), and our team, respectively. EZM2302 and TP-064 were developed through conventional small-molecule scaffolds occupying the substrate-binding pocket of CARM1.(Drew et al., 2017; Greenblatt et al., 2018; Nakayama et al., 2018) The potential utility of EZM2302 and TP-064 is implicated by their selective anti-proliferative effects on hematopoietic cancer cells, in particular multiple myeloma cells.(Drew et al., 2017; Greenblatt et al., 2018; Nakayama et al., 2018) However, definitive molecular mechanisms of the CARM1 addiction in these contexts remain elusive.(Greenblatt et al., 2018)
Here we report the characterization and novel utility of SKI-73---a chemical probe of CARM1 with pro-drug properties. SKI-73 can readily penetrate cell membranes and then be processed into two active CARM1 inhibitors containing 6′-homosinefungin (HSF) as their core scaffold.(Scheer et al., 2019; Wu et al., 2016) Notably, the two inhibitors can be accumulated inside cells at remarkable high concentrations and for a prolonged period. The potency, selectivity, modes of action, on-target engagement, and off-target effects of these compounds were characterized with multiple orthogonal assays in vitro and under cellular settings. The pharmacological inhibition of CARM1 by SKI-73 recapitulates the anti-invasion effect of the genetic perturbation of CARM1. In the context of cellular heterogeneity, we developed a cell-cycle-aware algorithm for single-cell RNA-seq (scRNA-seq) analysis and dissected the invasion-prone subset of breast cancer cells that is sensitive to SKI-73 treatment. Our scRNA-seq analysis provides the unprecedented insight that pharmacological inhibition of CARM1 alters epigenetic plasticity and suppresses invasion by suppressing the most invasive subpopulation of breast cancer cells.
Results
Development of 6′-homosinefungin derivatives as potent and selective CARM1 inhibitors
Upon developing cofactor-competitive PMT inhibitors,(Wu et al., 2016; Zheng et al., 2012) we identified 6′-homosinefungin (HSF, 1) for its general high affinity to Type I PRMTs (Fig. 1a,b, S1). As a SAM mimic, 1 binds to the Type I PRMTs---PRMT1, CARM1, PRMT6 and PRMT8--- with IC50 of 13~300 nM (Fig. 1a,c, Table S1). Its relative affinity to Type I PRMTs aligns with that of the SAM mimics SAH and SNF (around 20-fold lower IC50 of 1 versus SAH and SNF, Fig. 1a,c, Table S1). This observation argues that 1 retains the structural features of SAH and SNF to engage PRMTs and meanwhile leverages its 6′-methyleneamine group for additional interaction. Strikingly, the HSF derivative 2a, which was synthesized via the same precursor 3 (Fig. 1b, S1), preferentially binds to CARM1 with IC50 = 30 ± 3 nM and > 10-fold selectivity over other 7 human PRMTs and 26 methyltransferases of other classes (Fig. 1c, Table S1). The structural difference between 2a and 1 (Fig. 1b) suggests that the N-benzyl substituent enables 2a to engage CARM1 via a distinct mechanism (see results below). This engagement is expected to be maintained by 5a, an amide derivative of 2a prepared from the common precursor 3 and then the intermediate 4 (Fig. 1b, S2). Here 5a shows an IC50 of 43 ± 7 nM against CARM1 and a >10-fold selectivity over the panel of 33 diverse methyltransferases (Fig. 1c, Table S1). In comparison, the negative control compounds 2b (Bn-SNF)(Zheng et al., 2012) and 5b (Figure 1b, S3), which differ from 2a and 5a only by the 6′-methylene group, poorly inhibit CARM1 (IC50 > 25 μM and 1.91 ± 0.03 μM) (Fig. 1c, Table S1). The dramatic increase of the potency of 2a and 5a in contrast to 2b and 5b supports an essential role of the 6′-methylene moiety on binding CARM1. Distinguished from SAM mimics SAH, SNF and 1 as nonspecific PMT inhibitors, 2a and 5a were developed as potent and selective SAM analog inhibitors of CARM1 (Fig. 1c, Table S1).
Modes of interaction of 6′-homosinefungin derivatives as CARM1 inhibitors
With 2a and 5a characterized as CARM1 inhibitors, we leveraged orthogonal in vitro assays to explore their modes of interaction (Fig. 2a). CARM1 inhibition by 2a and 5a was assessed in the presence of various concentrations of SAM cofactor and H3 peptide substrate (Fig. 2b,c). IC50 values of 2a and 5a showed a linear positive correlation with SAM concentrations, as expected for SAM-competitive inhibitors.(Daigle et al., 2011; Luo, 2018; Zheng et al., 2012) The Kd values of 2a and 5a (Kd,2a = 17 ± 8 nM; Kd,5a = 9 ± 5 nM) were extrapolated from the y-axis intercepts upon fitting the equation IC50 = [SAM]×Kd/Km,SAM +Kd (Fig. 2b).(Segel, 1993) Km,SAM of 0.21 ± 0.09 μM and 0.28 ± 0.14 μM (an averaged Km,SAM = 0.25 μM) for competition with 2a and 5a can also be derived through the ratio of the y-axis intercepts to the slopes (Fig. 2b and Supplementary Methods).(Segel, 1993) In contrast, the presence of the H3 peptide substrate had negligible effect on the binding of 2a and 5a, indicating their substrate-noncompetitive character (Fig. 2c). The SAM analogs 2a and 5a were thus characterized as SAM-competitive, substrate-noncompetitive inhibitors of CARM1.
The CARM1-binding kinetics of 2a and 5a were also examined using surface plasmon resonance (SPR) (Fig. 2d). The SPR signal progression of 2a and 5a fits with a biphasic rather mono-phasic binding mode with the lower Kd1,2a = 0.06 ± 0.02 μM, Kd1,5a = 0.10 ± 0.01 μM, and the higher Kd2,5b = 0.54 ± 0.07 μM, Kd2,2a = 0.4 ± 0.1 μM, likely due to multi-phase binding kinetics of 2a and 5a (Figure 2d). In vitro thermal shift assay(Blum et al., 2014) further showed that the binding of 2a and 5a increases the melting temperature (Tm) of CARM1 by 4.4 °C and 6.5 °C, respectively (Fig. 2e, Tm,2a = 44.2 ± 0.4 °C and Tm,5a = 46.3 ± 0.3 °C versus Tm,DMSO = 39.8 ± 0.3 °C as control). In contrast, the binding of SAM and 1 shows much less effects on Tm of CARM1 (Fig. 2e, Tm,SAM = 40.1 ± 0.3 °C and Tm,1 = 42.8 ± 0.4 °C versus Tm,DMSO = 39.8 ± 0.3 °C). Therefore, albeit comparable affinity of 1, 2a and 5a to CARM1 (IC50 = 13~43 nM, Fig. 1c), their well-separated effects on Tm suggest that these inhibitors engage CARM1 differentially (see results below). Multiple orthogonal biochemical assays thus verified tight binding of 2a and 5a with CARM1.
Structural rationale of 6′-homosinefungin derivatives as CARM1 inhibitors
To further seek structural rationale of 5a and 2a for CARM1 inhibition, we solved the X-ray structure of CARM1 in complex with 5a and modeled the CARM1 binding of 2a (Fig. 3, Supplementary Results and Methods). The overall topology of the CARM1-5a complex is indistinguishable with a V-shape subunit of CARM1 dimer in complex with SNF and 1 (Figure S4, S5 and Table S2-4)---the Rossmann fold of Class I methyltransferases (Fig. 3a).(Luo, 2018) However, 5a adopts a noncanonical pose with its 6′-N-benzyl moiety in a binding pocket that used to be occupied by the α-amino carboxylate moiety of canonical ligands such as SAH, SNF and 1 (Fig. 3b, S4 and Table S2-4), while the α-amino methoxyphenethyl amide moiety of 5a protrudes into the substrate-binding pocket.(Boriack-Sjodin et al., 2016; Sack et al., 2011) This noncanonical mode is consistent with the SAM-competitive character of 5a (Fig. 2b). Under the canonical setting, the guanidinium moiety of Arg168 forms a salt bridge with the carboxylic moiety of canonical ligands (Fig. 3c). In contrast, Arg168 in the CARM1-5a complex has to adopt an alternative orientation (two possible configurations), accompanied by an altered conformation of Glu257, to accommodate the 6′-N-benzyl moiety of 5a (Fig. 3c). The α-amino amide moiety of 5a also engages CARM1 through the combined outcomes of a hydrogen-bond network with Glu266 and His414 and hydrophobic interactions with Phe152 and Tyr261 (Fig. 3d). Interestingly, the overlaid structures of CARM1 in complex with 5a and a substrate peptide implicate a steric clash and thus a potential competitive-binding mode between 5a and a CARM1 substrate (Fig. 3e). However, the apparent substrate-noncompetitive character of 5a (Figure 2c) suggests that this steric clash might be avoided if there is no significant energy penalty for the substrate Arg to adopt alternative conformation(s).
The binding mode of the CARM1-2a complex was modeled via molecular docking followed by molecular dynamics (MD) simulation (Supplementary Methods). Here we uncovered two distinct poses of 2a (Binding Pose 1/2 or BP1/2) with the C4'-C5'-C6'-C7' dihedral angle of −50° and −170°, respectively (Fig. 3f). BP1 was characterized by the direct interaction between the α-amino carboxylate moiety of 2a with the guanidinium of Arg168, while BP2 features a titled orientation of Arg168 to accommodate the 6′-N-benzyl moiety of 2a (the Cβ-Cγ-Cδ-Nε dihedral angle χ3 = 180° for BP1 versus χ3 = −65° of BP2) (Fig. 3f, S7). The BP1 and BP2 of 2a closely resemble those of 1 and 5a, respectively, in terms of the orientations of Arg168 and the α-amino carboxylate moiety of ligands. When the same modeling protocol was applied to the CARM1-SNF complex, only the canonical pose was identified (Fig. S7). Energy calculation indicated that both BP1 and BP2 are stable with comparable binding free energies. Interestingly, the side chain configurations of His414 in both BP1 and BP2 (the C-Cα-Cβ-Cγ dihedral angle χ1 = −48° and −66°) are different from those in the CARM1-5a complex and the CARM1-SNF complex (the C-Cα-Cβ-Cγ dihedral angle χ1 = 81°) (Fig. S7). Collectively, 5a and 2a, though structurally related to the SAM analogs 1 and SNF, engage CARM1 via distinct modes of interaction.
A pro-drug-like 6′-homosinefungin derivative as a cell-active CARM1 inhibitor
While the in vitro characterization demonstrated the potency and selectivity of 2a and 5a against CARM1, we anticipated their poor membrane permeability as observed for structurally-related analogs such as SAH and SNF (Fig. 1a).(Boriack-Sjodin et al., 2016; Sack et al., 2011) The lack of membrane penetration is likely due to their primary amine moiety, which has pKa of ~ 10 and is fully protonated at a physiological pH of 7.4. Given the essential roles of the 9′-amine moiety of 2a and 5a in CARM1 binding (Fig. 3d), we envisioned overcoming the membrane permeability issue via a pro-drug strategy by cloaking this amine moiety with a redox-triggered trimethyl-locked quinone propionate moiety (TML, Fig. 4a).(Levine and Raines, 2012) We thus prepared 6a as well as its control compound 6b by derivatizing 5a and 5b with the TML moiety (Fig. S2, S3). To assess the cellular activity of 6a, we relied on our prior knowledge that CARM1 methylates the Arg1064 of BAF155, a core component of the SWI/SNF chromatin remodeling complex, and CARM1 knockout abolishes this posttranslational modification in MCF-7 cells.(Wang et al., 2014) Treatment of MCF-7 cells with 10 μM of 6a fully suppressed this methylation mark, whereas treatment with 2a and 5a did not affect this mark (Fig. 4b). We thus demonstrated the prodrug-like cellular activity of 6a.
Characterization of 6a (SKI-73) as a chemical probe of CARM1
To further evaluate 6a as a chemical probe of CARM1, we assessed the efficiency of 6a to suppress CARM1-dependent invasion of breast cancer cells. Because of the pro-drug character of 6a and its control compound 6b, we first developed quantitative LC-MS/MS methods to examine their cellular fates (Supplementary Methods). Upon the treatment of MDA-MB-231 cells with 6a, we observed its time- and dose-dependent intracellular accumulation (Fig. 4c). While we anticipated the conversion of the pro-drug 6a into 5a, a striking finding is that 6a can also be readily processed into 2a inside cells (Fig. 4c). Remarkably, > 100 μM of 2a can be accumulated inside cells for 2 days after 6-h treatment with a single dose of 5~10 μM 6a. This observation likely reflects a slow efflux and thus effective intracellular retention of 2a due to its polar α-amino acid zwitterion moiety. Given that cellular CARM1 inhibition is involved with multiple species (2a, 5a and 6a) in competition with SAM, we modeled the ligand occupancy of cellular CARM1 on the basis of their Kd values (Kd,2a =17 nM, Kd,5a =9 nM, Kd,6a =0.28 μM and Kd,SAM ≈Km,SAM=0.25 μM) and MS-quantified intracellular concentrations (Fig. 4d, eqs. S5-S7, Supplementary Methods). The SAM cofactor, whose intracellular concentration was determined to be 89 ± 16 μM (Fig. 4c,d), is expected to occupy > 99.5% CARM1 with residual < 0.5% as the apo-enzyme under a native setting. With single doses of 6a of 2.5~10 μM, the combined CARM1 occupancy by 2a, 5a and their pro-drug precursor 6a rapidly reached the plateaus of >95% within 6 h, and was maintained at this level for at least 48 h (Fig. 4e). Notably, the treatment of 6a as low as 0.5 μM is sufficient to reach 60% target engagement within 10 h and maintain this occupancy for 48 h (Fig. 4e). The time- and dose-dependent progression of the CARM1 occupancy by these ligands thus provides quantitative guidance upon the treatment of MDA-MB-231 cells with 6a.
With a cellular thermal shift assay (CETSA),(Jafari et al., 2014) we further observed that the treatment of MDA-MB-231 cells with 6a but not the control compound 6b increases cellular Tm and thus thermal stability of CARM1 by 4.3 ± 0.6 °C (Fig. 4f). The distinct effect of 6a in contrast to 6b on the cellular Tm of CARM1 aligns well with the 4.1~6.2 °C difference of in vitro Tm of CARM1 upon binding 2a and 5a versus SAM (Figs. 2e). Here 6b can penetrate cell membrane and be processed into 5b and 2b in a similar manner as 6a (Figure S11). These observations thus present the cellular evidence of CARM1 engagement of 2a and 5a.
To further characterize 6a as a CARM1 chemical probe, we treated MDA-MB-231 cells with 6a and examined the Arg1064 methylation of BAF155 and the Arg455/Arg460 methylation of PABP1, two well-characterized cellular methylation marks of CARM1.(Lee and Bedford, 2002; Wang et al., 2014) These methylation marks can be fully suppressed by 6a in a dose-dependent manner (Fig. 5a). The resultant EC50 values of 0.45~0.75 μM (Fig. 5b) are well correlated with the modeled 60% cellular occupancy of CARM1 upon the treatment of 0.5 μM 6a for 48 h (Fig. 4e). In contrast, the treatment of the negative control compound 6b showed no effect on these methylation marks (Fig. 5a). We therefore demonstrated the robust use of 6a (SKI-73) as a CARM1 chemical probe and 6b (SKI-73N) as its control compound.
Inhibition of in vitro invasion but not proliferation of breast cancer cells by SKI-73
After demonstrating 6a (SKI-73) as a chemical probe of CARM1, we examined whether chemical inhibition of CARM1 can recapitulate biological outcomes associated with CARM1 knockout (CARM1-KO).(Wang et al., 2014) Our prior work showed that CARM1’s methyltransferase activity is required for invasion of MDA-MB-231 cells.(Wang et al., 2014) We thus conducted a matrigel invasion assay with MDA-MB-231 cells in the presence of 6a. Relative to the control treatment with DMSO, the treatment of 6a (SKI-73) but not its negative control compound 6b (SKI-73N) suppressed the invasion of MDA-MB-231 cells in a dose-dependent manner (EC50 = 1.3 μM) (Fig. 5c,d). The treatment with ≥ 10 μM 6a reached the maximal 80% suppression on the invasion of MDA-MB-231 relative to the DMSO control, which is comparable with the phenotype of CARM1-KO (Fig. 5e). Critically, no further inhibition by 6a on the invasiveness was observed upon its treatment of MDA-MB-231 CARM1-KO cells in comparison with the treatment with DMSO or 6b (Fig. 5e). Notably, the treatment with 6a and 6b under the current condition has no apparent impact on the proliferation of parental and CARM1-KO MDA-MB-231 cells (Figure S12), consistent with the intact proliferation upon the treatment with other CARM1 chemical probes.(Drew et al., 2017; Greenblatt et al., 2018; Nakayama et al., 2018) These results suggest that 6a (SKI-73) and CARM1 knockout perturb the common, proliferation-independent biological process and then suppresses 80% of the invasiveness of MDA-MB-231 cells. We thus characterized 6a (SKI-73) as a chemical probe to interrogate CARM1-dependent invasion of breast cancer cells.
scRNA-seq and cell-cycle-aware algorithm reveals CARM1-dependent epigenetic plasticity
Because of the advancement of scRNA-seq technology, stunning subpopulation heterogeneity has been uncovered even for well-defined cellular types.(Tanay and Regev, 2017) In the context of tumor metastasis including its initial step---invasion, epigenetic plasticity is required to offer a small subset of tumor cells to adapt distinct transcriptional cues for neo-properties.(Chatterjee et al., 2017; Flavahan et al., 2017; Wu et al., 2018) To explore the feasibility of dissecting the CARM1-dependent, invasion-prone subset of MDA-MB-231 breast cancer cells, we formulated a cell-cycle-aware algorithm of scRNA-seq analysis and dissected those subpopulations sensitive to CARM1 perturbation (Figure 6a, Supplementary Methods). Here we conducted 10× Genomics droplet-based scRNA-seq of 3,232, 3,583 and 4,099 individual cells (the total of 10,914 cells) exposed to 48-hour treatment with SKI-73 (6a), SKI-73N (6b) and DMSO, respectively. Guided by Silhouette analysis of each treatment condition as well as their combination for the modularity-based shared-nearest-neighbor(SNN) graph clustering, cell-cycle-associated transcripts were identified as dominant signatures to define subpopulations (Figure S13-30). These signatures naturally exist for proliferative cells and are not expected to be specific for the invasive phenotype. To dissect subpopulation-associated transcriptomic signatures of invasive cells, we included one additional layer for hierarchical clustering by first classifying the individual 10,914 cells into G0/G1, S, and G2/M stages (6,885, 1,520 and 2,509 cells, respectively) (Figure S18, Table S5), and then conducted the unsupervised clustering within each cell-cycle-aware subset (Figure S18, S31-38, S42-45, Table S5). To resolve efficiently the subpopulations associated with the three treatment conditions (6a, 6b and DMSO) without redundant clustering, we developed an entropy analysis method and relied on the Fisher Exact test (Supplementary Methods). The optimal scores of the combined methods were implemented for the modularity-based SNN graph clustering and to determine the numbers of cluster for each subset (Figure S32, S36, S43).(Butler et al., 2018) The cell-cycle-aware algorithm allowed the clustering of the 10,914 cancer cells according to the three cell cycle stages under the three treatment conditions (6a, 6b and DMSO) and resulted in 21, 7 and 6 subpopulations in G0/G1, S, and G2/M phases, respectively (Figure 6b, S33, S37, S44, Table S6-8). Notably, the 48-hour treatment with SKI-73 (6a) or SKI-73N (6b) had no effect on the cell cycle, as indicated by the comparable cell-cycle distribution patterns between SKI-73, SKI-73N, and DMSO treatment (Figure S18, Table S5). This result is also consistent with the intact proliferation upon the treatment with SKI-73 and SKI-73N (Figure S12).
CARM1-associated epigenetic plasticity of breast cancer cells with single-cell resolution
With the 21, 7 and 6 subpopulations clustered into the G0/G1, S, and G2/M stages, respectively, we then conducted population analysis between the three treatment conditions (SKI-73 and SKI-73N versus DMSO) (Figure 6c, S39, S46 and Table S6-8). These subpopulations can be readily classified into five distinct categories according to how the cells respond to SKI-73 and SKI-73N treatment in each cell cycle stage: commonly resistant/emerging/depleted versus differentially depleted/emerging (SKI-73/SKI-73N-specific) (Figure 6c, S39, S46 and Table S6-8). Here we are particularly interested in the SKI-73-specific depleted subpopulations (0/2/8/11/13/14/17/19 of G0/G1-phase cells and 3 of S-phase cells) as the potential invasion-associated subpopulations, given their sensitivity to SKI-73 but not its control compound SKI-73N. The subpopulations that remain unchanged after the treatment of SKI-73 and SKI-73N (1/7/12/15 of G0/G1-phase cells; 2/4/5 of S-phase cells; 1 of G2/M-phase cells) were defined as the common resistant subset. SKI-73-specific emerging subpopulations (3/4/5/6/16 of G0/G1-phase cells; 6 of S-phase cells; 4 of G2/M-phase cells) are expected to be suppressed by CARM1 but emerge upon its inhibition. Other subpopulations are either associated with effects of the small-molecule scaffold of SKI-73/SKI-73N (commonly emerging Subpopulation-9 of G0/G1-phase cells, 0/5 of G2/M-phase cells; commonly depleted Subpopulation-2/3 of G2/M-phase cells) or SKI-73N-specific effects (differentially depleted Subpopulation-10/18 of G0/G1-phase cells, 0 of S-phase cells; differentially emerging Subpopulation-20 of G0/G1-phase cells). Interestingly, in comparison with SKI-73 treatment, scRNA-seq analysis of 3,291 CARM1-KO cells suggests that CARM1 knockout has more profound effects on the overall landscape of the epigenetic plasticity (Figure S49). Collectively, the chemical probe SKI-73 alters the epigenetic plasticity of MDA-MB-231 breast cancer cells via the combined effects of SKI-73’s molecular scaffold and specific inhibition of CARM1’s methyltransferase activity.
Identification of CARM1-dependent, invasion-prone subpopulations of breast cancer cells
Given that SKI-73 has no effect on cell cycle and proliferation of MDA-MB-231 cells under the current treatment dose and duration, we envision that the invasion capability of MDA-MB-231 cells mainly arises from an invasion-prone subset, 80% of which is depleted by SKI-73 treatment (Figure 5c-e). We thus focused on Subpopulation-0/2/8/11/13/14/17/19 of G0/G1-phase cells and Subpopulation-3 of S-phase cells---in total nine depleted subpopulations specific for SKI-73 (Figure 6c, S39, S46 and Table S6-8). To identify invasion-prone subpopulation(s) among these candidates, we compared their transcriptional signature(s) with those that freshly invaded through Matrigel within 16 hours. Strikingly, in comparison with the highly heterogenous scRNA-seq signature of the parental MDA-MB-231 cells, the freshly-harvested invasive cells (3,793 cells for scRNA-seq) are relatively homogeneous with their subpopulations mainly determined by the cell-cycle-related transcriptomic signatures (Figure S49-53). Like the cells treated with DMSO, SKI-73 and SKI-73N, we classified the freshly-harvested invasive cells into G0/G1, S and G2/M stages (Figure S51, Table S5). Through the correlation analysis between the invasion cells and the subpopulations within each cell-cycle stage (Figure 6d, S40, S41, S47, S48, S54, S55), we readily revealed the subsets whose transcriptional signatures closely relate to those of the invasion cells including Subpopulation-6/7/8/9/14 in G0/G1-phase cells, 0/3 in S-phase cells and 1/2 of G2/M-phase cells (Table S6-11). In the context of population analysis for the nine SKI-73-specific depleted subpopulations, Subpopulation-8/14 of G0/G1-phase cells and Subpopulation-3 in S-phase are putative invasion-prone candidates. Subpopulation 8 of G0/G1-phase cells is the most sensitive and the only subpopulation that can be depleted by around 80% with SKI-73 treatment (Figure 6c). Given the ~80% suppression and ~20% residual invasion capability upon SKI-73 treatment, we argue that the invasive phenotype of MDA-MB-231 cells predominantly arises from the Subpopulation-8 of G0/G1-phase cells, which only accounts for ~8% of the parental cells in G0/G1 phase or ~5% without cell-cycle awareness. Differential expression analysis further revealed the single-cell transcriptional signatures of metastasis-implicated genes (e.g. MORC4, S100A2, RPL39, IFI27, ARF6, CHD11, SDPR and KRT18) that are specific for the G0/G1-phase Subpopulation-8 and invasion cells but not other G0/G1-phase invasion-prone candidates such as Subpopulation-6/7/9/14 (Fig. 6e, S55 and Table S12). The remaining cells of G0/G1-phase Subpopulation-8 after SKI-73 treatment (Fig. 6c,d) together with others (subpopulation-6/7/9/14 in G0/G1-phase cells, 0/3 in S-phase cells and 1/2 of G2/M-phase cells, Figs. S39, S41 S46 S48) may account for the 20% residual invasion capacity. Collectively, either CARM1 knockout or CARM1 inhibition with SKI-73 alters the epigenetic plasticity in a proliferation-independent manner, depleting the most invasion-prone subpopulation and thus suppressing the invasive phenotype.
Discussion
Chemical probes of CARM1
Based on a novel small-molecule scaffold 6′-homosinefungin (HSF), SKI-73 was developed as a pro-drug-like chemical probe of CARM1 by cloaking the 9′-amine moiety of 5a with the TML moiety. SKI-73N was developed as a control compound of SKI-73. The inhibitory activity of SKI-73 against CARM1 was demonstrated by the ability of SKI-73 but not SKI-73N to abolish the cellular methylation marks of CARM1---the Arg1064 methylation of BAF155 and the Arg455/Arg460 methylation of PABP1.(Lee and Bedford, 2002; Wang et al., 2014) While the ready intracellular cleavage of TML is expected for the conversion of SKI-73 and SKI-73N into 5a and 5b, respectively, it is remarkable that SKI-73 and SKI-73N can also be efficiently processed into 2a and 2b inside cells. Here 2a and 5a are presented as potent and selective CARM1 inhibitors, while their control compounds 2b and 5b poorly interact with CARM1. Competitive assays with SAM cofactor and peptide substrate showed that 2a and 5a act on CARM1 in a SAM-competitive and substrate-noncompetitive manner. The SAM-competitive mode is consistent with the ligand-complex structures of CARM1, in which the SAM binding site is occupied by 2a and 5a. Strikingly, as revealed by their ligand-CARM1 complex structures, 2a and 5a engage CARM1 via noncanonical modes with their 6′-N-benzyl moiety in the binding pocket that is otherwise occupied by the α-amino carboxylic moiety of the conventional SAM analogs such as SAH, SNF and 1. This observation is consistent with the 4.1~6.5 °C increase in in vitro and cellular Tm of CARM1 upon binding 2a and 5a in contrast to the less Tm changes with SAM as a ligand. The distinct modes of interaction of CARM1 with 2a and 5a (Figure 3b,3f) also rationalize the CARM1 selectivity of the two SAM analogs over other methyltransferases including closely related PRMT homologs. Through mathematic modeling using the inputs of the LC-MS/MS-quantified intracellular concentrations and CARM1-binding constants of relevant HSF derivatives and SAM cofactor, we concluded that high intracellular concentrations of 5a and 2a and thus efficient CARM1 occupancy can be achieved rapidly and maintained for several days with a single low dose of SKI-73. The polar α-amino acid zwitterion moiety of 2a and the polar α-amino moiety of 5a likely account for their accumulation and long-time retention inside cells.
To the best of our knowledge, EZM2302, TP-064, SKI-73 (www.thesgc.org/chemical-probes/SKI-73) and their derivatives are the only selective and cell-active CARM1 inhibitors.(Drew et al., 2017; Nakayama et al., 2018) While the potency, selectivity, on-target engagement and potential off-target effects associated with these compounds have been examined in vitro and in cellular contexts as chemical probes, EZM2302, TP-064, SKI-73 are distinct by their molecular scaffolds and modes of interaction with CARM1 (www.thesgc.org/chemical-probes/SKI-73).(Drew et al., 2017; Nakayama et al., 2018) SKI-73 is a cofactor analog inhibitor embedding a N6′-homosinefungin moiety to engage the SAM binding site of CARM1 in a cofactor-competitive, substrate-noncompetitive manner; EZM2302 and TP-064 occupy the substrate-binding pocket of CARM1 in a SAH-uncompetitive or SAM-noncompetitive manner.(Drew et al., 2017; Nakayama et al., 2018) In particular, the prodrug property of SKI-73 allows its ready cellular uptake, followed by rapid conversion into its active forms inside cells. The prolonged intracellular CARM1 inhibition further distinguishes SKI-73 from EZM2302 and TP-064.
Anti-cancer effects and conventional mechanisms associated with pharmacological inhibition of CARM1
With SKI-73 as a CARM1 chemical probe and SKI-73N as a control compound, we showed that pharmacological inhibition of CARM1 with SKI-73, but not SKI-73N, suppressed 80% invasion capability of MDA-MB-231 cells. In contrast, the pharmacological inhibition of CARM1 with SKI-73 had no effect on the proliferation of MDA-MB-231 cells. This result is consistent with the lack of anti-proliferation activities of the other two CARM1 chemical probes EZM2302 and TP-064 against breast cancer cell lines.(Drew et al., 2017; Nakayama et al., 2018) The anti-invasion efficiency of SKI-73 is in a good agreement with the intracellular occupancy and the resulting abolishment of several methylation marks of CARM1 upon the treatment of SKI-73. Our prior work showed that the methyltransferase activity of CARM1 is required for breast cancer metastasis.(Wang et al., 2014) Among diverse cellular substrates of CARM1,(Blanc and Richard, 2017) BAF155---a key component of the SWI/SNF chromatin-remodeling complex---is essential for invasion of MDA-MB-231 cells.(Wang et al., 2014) Mechanistically, the CARM1-mediated Arg1064 methylation of BAF155 facilitates the recruitment of the SWI/SNF chromatin-remodeling complex to a specific subset of gene loci.(Wang et al., 2014) Replacement of the native CARM1 with its catalytically dead mutant or an Arg-to-Lys point mutation at the Arg1064 methylation site of BAF155 is sufficient to abolish the invasive capability of breast cancer cells.(Wang et al., 2014) CARM1 inhibition with SKI-73, but not its control compound SKI-73N, recapitulates anti-invasion phenotype associated with the genetic perturbation of CARM1. More importantly, there is no additive effect upon combining CARM1-KO with SKI-73 treatment, underlying the fact that the two orthogonal approaches target the commonly shared pathway(s) essential for invasion of breast cancer cells. In comparison to SKI-73, the CARM1 inhibitors EZM2302 and TP-064 demonstrated the anti-proliferation effects on hematopoietic cancer cells, in particular multiple myeloma.(Drew et al., 2017; Greenblatt et al., 2018; Nakayama et al., 2018) Mechanistically, genetic perturbation of CARM1 in the context of leukemia impairs cell-cycle progression, promotes myeloid differentiation, and ultimately induces apoptosis, likely via targeting pathways of proliferation and cell-cycle progression---E2F-, MYC-, and mTOR-regulated processes.(Greenblatt et al., 2018) In comparison, CARM1 inhibition with EZM2302 led to a slightly different phenotype, including reduction of RNA stability, E2F target downregulation, and induction of a p53 response signature featured for senescence.(Greenblatt et al., 2018) Collectively, the effects of CARM1 chemical probes are highly context-dependent with the different uses of SKI-73 against invasion of breast cancer cells versus TP-064 and EZM2302 against proliferation of hematopoietic cancer cells.
CARM1-dependent epigenetic plasticity revealed by SKI-73 with single-cell resolution
Given the increased awareness of epigenetic plasticity,(Flavahan et al., 2017) we employed the scRNA-seq approach to examine MDA-MB-231 cells and their responses to chemical and genetic perturbation with CARM1. Because of the lack of the prior reference to define subpopulations of MDA-MB-231 cells, we developed a cell-cycle-aware algorithm to cluster the subpopulations with a resolution to dissect subtle changes upon the treatment of SKI-73 versus its control compound SKI-73N in each cell cycle stage. Guided by Silhouette analysis, the population entropy analysis and the Fisher Exact test, >10,000 MDA-MB-231 breast cancer cells were classified on the basis of their cell cycle stages and then clustered into 34 subpopulations. With further annotation of these subpopulations according to their different responses to the treatment of SKI-73 versus SKI-73N, we readily dissected the subpopulations that were altered in a SKI-73-specific (CARM1-dependent) manner and then identified the subsets with the transcriptional signatures that are similar to that of the freshly-isolated invasive cells. Quantitative analysis of SKI-73-depleted subpopulations further revealed the most invasion-prone subpopulation, which accounts for only 5% of the total population but at least 80% invasive capability of the parental cells. Collectively, we propose a model that MDA-MB-231 cells consist of the subpopulations with their epigenetic plasticity determined by multiple factors including the CARM1-involved BAF155 methylation.(Wang et al., 2014) SKI-73 inhibits the methyltransferase activity of CARM1, the Arg1064 methylation of BAF155, and thus the target genes associated with the methylated BAF155. These effects alter the cellular epigenetic landscape by affecting certain subpopulations of MDA-MB-231 cells without apparent effect on cell cycle and proliferation. In the context of the invasion phenotype of MDA-MB-231 cells, the subset of invasion-prone cells is significantly suppressed upon the treatment with SKI-73. Essential components to dissect the invasion-prone population in this CARM1-dependent epigenetic plasticity model are the scRNA-seq analysis of sufficient MDA-MB-231 cells (>10,000 cells here), the utility of the freshly isolated invasive cells as the reference, the timing and duration of treatment, and the use of SKI-73N and DMSO as controls. Interestingly, although the invasion-prone subpopulation is also abolished in the CARM-KO strain, CARM-KO reshapes the epigenetic plasticity in a much more profound manner---significantly reducing the subpopulation heterogeneity of MDA-MB-231 cells. The distinct outcomes between the pharmacological and genetic perturbation can be due to their different modes of action---short-term treatment with SKI-73 versus long-term clonal expansion of CARM1-KO cells. The pharmacological inhibition captures the immediate response, while the genetic perturbation reports long-term and potential resistant outcomes. This work thus presents a new paradigm to understand cancer metastasis in the context of epigenetic plasticity and provides guidance to carry out similar analysis in broader contexts---other cell lines, patient derived xenograft samples, and in vivo mouse models of breast cancer.
Online content
Supplementary results, methods, Figures S1-55, Tables S1-13, and references are available at https://doi.org/
Author Contributions
X.C.C., K.W., W.Z., and J.P.L. developed synthetic strategies and prepared compounds; X.C.C., J.W., C.S., T.H., G.I., M.V., F.L., C.C.S., and T.H. characterized the compounds with in vitro biochemical and biophysical assays; H.W., F.L., C.C.S., H.Z., A.D., and L.D. solved the X-ray crystal structures of CARM1 in complex with ligands; V.V. and L.S. modeled the interaction of CARM1 with ligands; E.J.K., X.C.C., M.J., N.Z., Y.C. D.B. and M.S. examined on-target engagement in a cellular context; S.C. and X.C.C. modeled target occupancy; E.J.K., X.C.C., M.J., and N.Z. conducted invasion assay; X.C.C., M.J., and L.M. performed scRNA-seq; T.Z., M.L., L.X.Q., X.C.C., and X.N. analyzed scRNA-seq data; X.C.C., M.L., W.X., L.X.Q., J.X., P.J.B., M.V., L.S., C.H.A., J.M., H.D. and Z.Z. designed experiments and supervised the projects; all of the authors analyzed the data; X.C.C. and M.L. wrote the manuscript with inputs of other authors.
Competing interests
The authors declare no competing interests.
Additional information
PDB codes: 4IKP for the CARM1-1 complex and 6D2L for CARM1-5a complex
Acknowledgements
The authors thank Christina Leslie for providing suggestion of scRNA-seq analysis; the National Institutes of Health of USA (ML: R01GM096056, R01GM120570), National Cancer Institute (ML: 5P30 CA008748; WX: R01CA236356, R01CA213293), Starr Cancer Consortium (ML), MSKCC Functional Genomics Initiative (ML), the Sloan Kettering Institute (ML), Mr. William H. Goodwin and Mrs. Alice Goodwin Commonwealth Foundation for Cancer Research, and the Experimental Therapeutics Center of Memorial Sloan Kettering Cancer Center (ML), MSKCC Metastasis and Tumor Ecosystems Center (ML), the Tri-Institutional PhD Program in Chemical Biology (SC), NIH (WX), Susan G. Komen Foundation (EJK : PDF17481306), National Cancer Institute of National Institutes of Health of USA (LXQ: CA214845, CA008748), and Special Funding of Beijing Municipal Administration of Hospitals Clinical Medicine Development---YangFan Project (ZZ: ZYLX201713). The Structural Genomics Consortium is a registered charity (no. 1097737) that receives funds from AbbVie; Bayer Pharma AG; Boehringer Ingelheim; Canada Foundation for Innovation; Eshelman Institute for Innovation; Genome Canada; Innovative Medicines Initiative (EU/EFPIA) (ULTRA-DD grant no. 115766); Janssen; Merck KGaA; Darmstadt, Germany; MSD; Novartis Pharma AG; Ontario Ministry of Economic Development and Innovation; Pfizer; São Paulo Research Foundation-FAPESP; Takeda; and the Wellcome Trust. The X-ray structure results of CARM1 are derived from work performed at Argonne National Laboratory, Structural Biology Center (SBC) at the Advanced Photon Source. SBC-CAT is operated by U Chicago Argonne, LLC, for the U.S. Department of Energy, Office of Biological and Environmental Research under contract DE-AC02-06CH11357. These experiments were performed using beamline 08ID-1 at the Canadian Light Source, which is supported by the Canada Foundation for Innovation, Natural Sciences and Engineering Research Council of Canada, the University of Saskatchewan, the Government of Saskatchewan, Western Economic Diversification Canada, the National Research Council Canada, and the Canadian Institutes of Health Research.