Vanderbilt Institute of Chemical Biology



Discovery at the VICB







Targeting the WIN Site of WDR5 


By: Carol A. Rouzer, VICB Communications
Published:  March 21, 2019


New potent inhibitors of WDR5 function disrupt protein synthesis resulting in apoptosis in some cancer cells.   


The epigenetic modulation of gene transcription offers promising new targets for cancer chemotherapy. Already, efforts have focused on histone methyltransferases, histone deacetylases, and epigentic "readers" (proteins that bind to histone modifications) as proteins of particular interest. A somewhat different approach is to target proteins that facilitate epigenetic regulation. One such protein is WDR5, which serves as a scaffold for the assembly of protein complexes containing the epigenetic "writers" that catalyze histone modifications. WDR5 provides a scaffold for multiple writer complexes including those encompassing MLL/SET-type histone methyltransferases (HMTs) that are responsible for di- and tri-methylation of histone 3 at lysine-4 (H3K4). Aberrant expression of WDR5 occurs in some forms of cancer, suggesting a potential role in carcinogenesis and/or the malignant phenotype. This led Vanderbilt Institute of Chemical Biology members William Tansey, Stephen Fesik, and Scott Hiebert to discover and characterize potent inhibitors of WDR5 function [E. R. Aho, et al. (2019) Cell Reports, 26, 2916].


WDR5 contains a WIN site (WDR5 interaction site) through which it binds to an arginine-containing motif (WIN motif) on other proteins. We do not fully understand the function of the WIN site, but we do know that the HMT activity of MLL/SET proteins containing the MLL1 isoform requires that the site be intact. Thus, prior reports revealed that a WIN site inhibitor suppressed the growth of cancer cells expressing mutant forms of the CCAAT enhancer-binding protein α (C/EBPα) and/or p53. Other reports showed that higher affinity inhibitors blocked the growth of leukemia cells bearing rearrangements of the gene encoding MLL1, suggesting the possibility that the effects of WIN site inhibitors on these cells might be the result of suppressing H3K4 methylation. These early promising results, combined with the fact that the WIN site is a well-defined pocket that should provide a good binding site for small molecules, led the investigators to apply fragment-based approaches combined with structure-based design to obtain highly potent WIN site inhibitors.


The basis for fragment-based drug discovery is the screening of small molecules (~300 Da) that bind with low affinity to the target site and then use them as the foundation upon which to build larger molecules that exploit untapped regions in the site to gain higher affinity. The researchers started by expressing uniformly 15N-labeled WDR5 so that they could use 1H-15N heteronuclear multiple quantum coherence (HMQC) spectroscopy to observe peak shifts that occur when the protein binds to an unlabeled MLL1 WIN motif peptide. They then screened a 13,800 member fragment library for compounds that cause similar peak shifts upon interaction with WDR5. The screen yielded 47 hit molecules from which they initially selected compound C1 (Figure 1) for further study. An X-ray crystal structure of C1 complexed with WDR5 confirmed that it bound deep within the WIN site, occupying the S2 region where the arginine residue of the WIN motif would normally bind. Through structure-based design, the investigators altered the compound to create C2 and then C3 in order to exploit the S7 and S4 regions of the WIN site, resulting in a 50,000-fold increase in binding affinity. Similarly, starting with compound C4 from the fragment screen, the researchers introduced structural modifications to produce compounds C5 and C6 (Figure 1). This effort yielded a 1.5 million-fold increase in binding affinity that resulted from multiple interactions of C6 with the WIN site, as seen by X-ray crystallography (Figure 2). With affinities in the picomolar to nanomolar range, the researchers decided to use C3 and C6 as probe compounds to explore the effects of WIN site inhibition on cells. They also created compounds C3nc and C6nc, bearing small structural modifications of C3 and C6, respectively, that rendered them inactive. These served as control compounds.



FIGURE 1. Structures of C1, C2, and C3 (A) and their binding poses in the WIN site of WDR5 as determined by X-ray crystallography (B, C, and D, respectively). Structures of C4, C5, and C6 (E) and their binding poses in the WIN site of WDR5 as determined by X-ray crystallography (F, G, and H, respectively). Figure reproduced under the CC BY-NC-ND license from E. R. Aho, et al., (2019) Cell Reports, 26, 2916.




FIGURE 2. (A) Ribbon diagram of C6 bound to WDR5 and (B) a close-up of C6 in the WIN site illustrating binding interactions with key residues in the site. Figure reproduced under the CC BY-NC-ND license from E. R. Aho, et al., (2019) Cell Reports, 26, 2916.



As noted above, prior work had shown that the peptidomimetic WIN inhibitor MM-401 suppresses the growth of cell lines and transformed cells carrying rearrangements of the gene that encodes MLL1. Thus, the researchers investigated the effects of C3 and C6 on the growth of a range of cell lines that did or did not carry the MLL1 rearrangement (MLLr). They found that MLLr cell lines were more sensitive to the growth inhibitory effects of C3 and C6, and that expression of wild-type p53 by the cells also conveyed sensitivity. In general, both compounds were significantly less potent inhibitors of cell growth than would have been predicted by their affinities for the WIN1 site of WDR5; however, C6 was consistently more potent in cells than C3, and C6nc and C3nc were inactive. These findings supported the hypothesis that the growth suppressive effects of C6 and C3 were due to their interaction with WDR5.

When the researchers treated MV4:11 cells for 2 to 6 days with C3 or C6, they noted an increase in the number of cells containing a quantity of DNA less than would be expected for the G1 phase of the cell cycle. This finding suggested that cells were undergoing apoptosis. Consistently, they found an increase in caspase-mediated cleavage of poly (ADP-ribose) polymerase-1 and expression of annexin V, both hallmarks of the apoptotic response. Increased expression of the p53 protein and its target p21 also occurred, and the researchers found that treating cells with p53-directed shRNAs to knockdown cellular levels of the protein reduced their sensitivity to the WIN site inhibitors. The use of CRISPR-Cas9 technology to create cells in which the gene encoding p53 had been deleted also conveyed C3 and C6 resistance. These results strongly suggested that WIN site inhibition leads to p53-dependent apoptosis, particularly in cells carrying an MLL rearrangement.

To better understand the impact of C6 treatment, the investigators used RNA-seq to look at the effects of a 3 day exposure to C6 versus C6nc on gene expression in MV4:11 cells. C6nc elicited no changes in gene expression, whereas C6 caused increased expression of 72 genes and decreased expression of 462 genes. Gene ontology analysis revealed no insights about pathways affected by the induced genes, but the genes that exhibited decreased expression were strongly associated with pathways involving protein synthesis, DNA replication, and the cell cycle. Of particular interest was a subset of genes encoding small and large ribosomal proteins. Follow-up studies using chromatin immunoprecipitation coupled with next-generation sequencing (ChIP-seq) yielded 158 high confidence WDR5 chromatin binding sites. The genes identified by ChIP-seq included 40% and 70% of those encoding small and large ribosomal subunit proteins, respectively. Thus the RNA-seq and ChIP-seq data combined indicated that WDR5 is strongly associated with the regulation of ribosomal protein synthesis, and thereby protein synthesis. The data also indicated that C6 treatment leads predominantly to repression of WDR5-associated gene expression. For example, of the 158 WDR5-associated genes identified by ChIP-seq, only 1 was also identified as being induced by RNA-seq (Figure 3). In contrast, 59 of the WDR-associated genes were repressed, as indicated by RNA-seq (Figure 4). Notably, the ChIP-seq studies identified no genes associated with DNA replication or the cell cycle, suggesting that changes in these genes detected by RNA-seq are secondary rather than direct effects of WIN site inhibition.



FIGURE 3. Venn diagram showing the number of genes identified by ChiP-seq using the D9E11 antibody and those identified as induced by RNA-seq. Figure reproduced under the CC BY-NC-ND license from E. R. Aho, et al., (2019) Cell Reports, 26, 2916.



FIGURE 4. Venn diagram showing the number of genes identified by ChiP-seq using the D9E11 antibody and those identified as repressed by RNA-seq. Figure reproduced under the CC BY-NC-ND license from E. R. Aho, et al., (2019) Cell Reports, 26, 2916.



The suppression of genes encoding ribosomal proteins suggested that a primary mechanism by which WIN site inhibitors exert their toxicity is through disruption of protein synthesis. To further test this hypothesis, the investigators pulsed MV4:11 cells with O-propargyl-puromycin, which binds to newly synthesized peptides and enables fluorescent labeling of those peptides using click chemistry. Following this treatment, the researchers could employ flow cytometry to measure changes in protein synthesis as a result of C6 treatment. The data clearly confirmed that C6 leads to a suppression of protein synthesis that, after 6 days of exposure, was as severe as that observed in cells treated with cycloheximide. Consistently, the researchers also observed a redistribution of nucleophosmin from the nucleolus to the nucleoplasm, a marker of nucleolar stress that can be triggered by disruptions in protein synthesis.

Nucleolar stress is a primary trigger of p53-dependent signaling, often through stabilization of the p53 protein. However, the researchers found no evidence that the rate of p53 degradation was affected by WIN site inhibition. Instead, C6 exposure elicited an increase in mRNA encoding p53 on polysomes, indicating that at a time when the cells were experiencing an overall disruption in protein synthesis, the synthesis of p53 was upregulated.

As noted above, early studies of WIN site inhibitors demonstrated that they blocked MLL1-dependent methylation of H3K4 sites. This led the researchers to hypothesize that their inhibitors could act via changes in epigenetic modulation of gene expression. They first tested this hypothesis by showing that both C3 and C6 suppressed the HMT activity of MLL/SET complexes containing MLL1 but not other MLL family proteins. They then used precision nuclear run-on sequencing (PRO-seq) to examine the location of active RNA polymerases across the genome of cells treated with C3. They found decreases in gene body polymerase occupancy in 47 transcription units. No increases in transcription were observed for any genes. Almost all of the genes exhibiting transcription suppression had been identified in ChIP-seq studies as sites of WDR5 binding, and 70% of them encoded ribosomal subunit proteins. However, the investigators found no evidence for suppression of H3K4 methylation of WDR5-bound genes. These results were all consistent with the prior findings that WIN site inhibitors block the synthesis of ribosomal proteins, but they did not support the hypothesis that the inhibitors act through modification of epigenetic regulation, at least not at the earlier time points. Of particular note was the finding that WIN site inhibition reduced the binding of WDR5 to its target genes by greater than 10-fold. Thus, it appears that the primary function of the WIN site is to enable WDR5 to associate with binding sites on chromatin, and WIN site inhibitors work primarily by disrupting this association.

In conclusion, two new high affinity probes that bind to the WIN site of WDR5 are now available. Both are efficacious in intact cells, though at much higher concentrations than would be predicted based on in vitro binding affinity, suggesting that improvements in parameters affecting compound uptake and disposition within the cell are in order. Nevertheless, the current studies demonstrate that these compounds displace WDR5 from its binding sites in chromatin resulting in a downregulation of transcription of the associated genes, many of which are involved in ribosomal protein synthesis. The result is dysfunctional translation leading to nucleolar stress and ultimately p53-dependent apoptosis (Figure 5). Why MLL rearrangements convey higher sensitivity to these inhibitors remains a topic of future research, as does their potential application in cancer chemotherapy.




FIGURE 5. Proposed mechanism of WIN site inhibition. The inhibitor (i) binds to the WIN site of WDR5, displacing it from its binding site on chromatin. A major result is a decrease in transcription of genes encoding ribosomal proteins (RPGs), leading to a disruption of protein synthesis, the nucleosomal stress response, and ultimately, p53-dependent apoptosis. Figure reproduced under the CC BY-NC-ND license from E. R. Aho, et al., (2019) Cell Reports, 26, 2916.



ViewCell Reports, article: Displacement of WDR5 from Chromatin by a WIN Site Inhibitor with Picomolar Affinity







The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2014 by Vanderbilt University Medical Center