Chinese Chemical Letters  2015, Vol.26 Issue (04):455-458   PDF    
Structure-based design of conformationally constrained cyclic peptidomimetics to target the MLL1-WDR5 protein-protein interaction as inhibitors of the MLL1 methyltransferase activity
Hacer Karatasa,b,c, Shirley Y. Leed, Elizabeth C. Townsendd, Fang Caod, Jing Xud, Denzil Bernardb,c, Liu Liub,c, Yali Doud, Shaomeng Wanga,b,c,e     
a Department of Medicinal Chemistry, University of Michigan, Ann Arbor, MI 48109, USA;
b Department of Internal Medicine, University of Michigan, Ann Arbor, MI 48109, USA;
c Comprehensive Cancer Center, University of Michigan, Ann Arbor, MI 48109, USA;
d Department of Pathology, University of Michigan, Ann Arbor, MI 48109, USA;
e Department of Pharmacology, University of Michigan, Ann Arbor, MI 48109, USA
Abstract: We described herein structure-based design, synthesis and evaluation of conformationally constrained, cyclic peptidomimetics to block the MLL1-WDR5 protein-protein interaction as inhibitors of the MLL1 histone methyltransferase activity. Our study has yielded cyclic peptidomimetics with very high binding affinities to WDR5 (Ki values <1 nmol/L) and function as antagonists of the MLL1 histone methyltransferase activity.
Key words: Epigenetics     MLL1 histone methyltransferase     WDR5-MLL1 interaction     Small-molecule inhibitors     Cyclic peptidomemtics    
1. Introduction

Histones are the most basic units for pacKing DNA into nucleosomes and covalent modifications of histones,such as methylation,acetylation and phosphorylation,play a central role for regulation of gene transcription [1, 2]. It is now well recognized that misregulation of histone modifications plays a role in a wide range of human diseases,including but not limited to cancer [3, 4].

MLL1 (Mixed Lineage Leukemia 1) protein is a Histone H3 Lysine 4 (H3K4) methyltransferase and is frequently misregulated in a subset of acute leukemia [5, 6]. MLL1 itself has a weak H3K4 methyltransferase activity but its enzymatic activity is dramatically enhanced when MLL1 is present in a core complex,consisting of MLL1,WDR5 (WD repeat domain 5 protein),ASH2L (Absent, Small,or Homeotic-2-Like) and RbBP5 (Retinoblastoma Binding Protein 5). Our recent study has clearly shown that the interaction between MLL1 and WDR5 proteins is essential for the activity of MLL1 but dispensable for the activity of other MLL family members,including MLL2,MLL3 and MLL4 [7]. Hence,blocKing the MLL1-WDR5 protein-protein can specifically inhibit the activity of MLL1 H3K4 methyltransferase activity and such blocKing may have a potential for the treatment of human diseases such as a subset of acute leukemia,whose development and progression depend upon MLL1 activity.

The interaction between MLL1 andWDR5 is mediated by a welldefined pocket inWDR5 and a 12-residueWIN (WDR5 interacting) motif (residues 3762-3773) in MLL1. In a previous study,we have determined that a 3-residue peptide,i.e. Ac-ARA-NH2 (Ki = 0.12 μmol/L) has the same binding affinity as the 12-residue WIN peptide (Ki = 0.16 μmol/L) to WDR5 [8]. We have performed extensivemodifications on Ac-ARA-NH2 and obtained highly potent linear peptidomemtics [9]. For example,MM-101 binds to WDR5 with a Ki value of<1 μmol/L and can effectively inhibit cell growth in acute leukemia cancer cell lines harboring MLL1 fusion [9].

In this study,we present our structure-based design,synthesis and evaluation of conformationally constrained cyclic peptidomimetics, based upon our previously reported linear peptidomimetics as inhibitors of the MLL1-WDR5 interaction and MLL1 methyltransferase activity. Conformational constrain by cyclization has effectively been used as a strategy to improve binding affinity of linear peptides to their target proteins by reducing the conformational entropy loss and to enhance cell-permeability and in vivo properties of linear peptides. Our study has yielded conformationally constrained cyclic peptidomimetics that bind to WDR5 with very high affinities (Ki < 1 μmol/L) and potently inhibit the MLL1 H3K4 methyltransferase activity at low nanomolar concentrations. 2. Experimental 2.1. Chemistry

All the final compounds were characterized with 1H NMR, 13C NMR (300 MHz,Bruker),and HRMS (ESI+) (Agilent Q-TOF Electrospray). The intermediates were characterized with 1H NMR,13C NMR (300 MHz,Bruker) and MS (ESI+) (Thermo Scientific LCQ Fleet). Chemical shifts were reported in ppm relative to TMS. D2O (4.79 ppm),CD3OD (3.31 ppm),CD3CN (1.94 ppm) and DMSO-d6 (2.50 ppm) were used as the internal standard for 1H NMR. D2O (1,4-dioxane,66.7 ppm),CD3OD (49.2 ppm),CD3CN (1.4 ppm) and DMSO-d6 (39.5 ppm) were used as internal standards for 13C NMR spectra. Experimental details and full structure characterization are provided in Supporting Information. 2.2. Competitive Binding Experiments

The binding affinities of all the final compounds were tested using a fluorescence polarization (FP) based competitive binding assay described earlier [8]. Briefly,to a 5 μL solution of the tested compound in DMSO,120 μL of pre-incubated complex solution (WDR5△23 and 5-FAM labeled tracer) in assay buffer (0.1 mol/L Phosphate,25 μmol/L KCl,0.01% Triton, pH 6.5) was added,giving final concentrations ofWDR5△23 and the tracer to be 4 μmol/L and 0.6 μmol/L,respectively. The plates were incubated at room temperature on a shaker for 3 h, and then the mP values were measured using the Tecan Infinite M-1000 plate reader (Tecan U.S.,Research Triangle Park,NC). Ki values were calculated using the equation described previously [8]. 2.3. MLL1 histone methyltransferase assay

To assess the ability of our designed compound to inhibit the MLL1 histone methyltransferase activity (HMT),we have developed and further optimized an in vitro fully reconstituted MLL1 histone methyltransferase assay. All HMT assays were performed as previously described,with some modifications [10]. Briefly,human MLL1 core complex was reconstituted by mixing MLL1 SET domain,WDR5,RbBP5,and Ash2L proteins in equal molar ratio in 50 μmol/L Tris pH 8.0,50 μmol/L NaCl, 5 μmol/L MgSO4,1 μmol/L DTT,and 10% glycerol. All compoundswere dissolved in 100% DMSO to make 10 μmol/L stocks for HMT assays. Different concentrations of each compound were incubated on ice with the complex for 10 minutes prior to adding 0.3 μmol/L of 3H-S-adenosyl methionine (3H-SAM) and 625 μmol/L H3 peptides (residues 1-21). The mixtures were then incubated at room temperature for 1 h before spotting 10 μL onto Whatman P81 ion exchange filter papers. Filter papers were washed three times in 50 μmol/L NaHCO3,pH 9.0, air dried,and subsequently placed in scintillation vials containing scintillating cocktail for 3H counts acquisition. All experimentswere done on Tri-Carb 2800 TR liquid scintillation counter in the unit of counts per minute (CPM). All data points were obtained in duplicates. GraphPad Prism was used to obtain the IC50s and dose-response curve presentations. 3. Results and discussion 3.1. Structure-based design

Based upon the co-crystal structure of MM-101 in complex with WDR5 [9],we have designed cyclic compounds in which one of the ethyl groups is linked to the amino group to form a cyclic structure (Fig. 1). To facilitate our synthesis,we have changed the cyclopentyl group in MM-101 to an ethyl group and one of the ethyl groups to a methyl group. Computational modeling showed that the length of the linker is critical for maintaining the bound conformation and suggested that at least a 5 carbon linker is required for effective interactions with WDR5 (Fig. 2). To test our design ideas,we synthesized three compounds with different linker lengths containing 5-,6- and 8-carbon atoms,which yielded compounds MM-301,MM-302 and MM-303,respectively (Fig. 1).

Download:
Fig. 1. Structure-based design of conformationally,constrained cyclic peptidomimetics to block the WDR5-MLL1 protein-protein interaction.

Download:
Fig. 2. Predicted binding pose for compounds MM-301 (magenta) and MM-302 (yellow) complexed with WDR5 using the GOLD program. Crystallographic pose of MM-101 (green) shown in ball and stick. Figures were generated using Pymol.
3.2. Chemical synthesis

The synthesis of these cyclic peptidomimetics is provided in Scheme 1. First,intermediate 2 was prepared on the 2-chlorotrityl chloride resin (1) using solid phase peptide synthesis and Fmoc chemistry,and then cleaved from the resin to yield carboxylic acid 3 (Scheme 1). For the synthesis of designed peptidomimetics, different alkene amines (4a,4b and 4c) were coupled to 3 to afford 5a-5c. RCM cyclization of 5a-5c followed by catalytic hydrogenation and removal of the Pbf protecting group from arginine side chain yielded the final compounds 6a-6c (MM-301,MM-302 and MM-303) as trifluoroacetic acid salts. The synthetic process and structure determination data of representative compounds are provided in supporting information.

Download:
Scheme 1. Chemical synthesis of conformationally constrained,cyclic peptidomimetics.
3.3. Evaluation of binding affinities to WDR5

We tested the binding affinities of these cyclic peptidomimetics to WDR5 using our optimized fluorescence-polarization binding assay. The data are summarized in Table 1.

Table 1
Chemical structures and binding affinities of designed peptidomimetics to WDR5, as determined in a fluorescence-polarization assay.

Our binding data showed that MM-301,MM-302 and MM-303 bind to WDR5 with IC50 values of 147 μmol/L,1.3 μmol/L and 1.2 μmol/L,respectively. Consistent with our modeling prediction, MM-301 with a 5-carbon linker binds to WDR5 with a good affinity (IC50 = 147 μmol/L,Ki value = 31.5 μmol/L) but is not optimal. Increasing the linker length by just one extra carbon (from the 5-carbon linker in MM-301 to the 6-carbon linker in MM-302) improves the binding affinity by over 100-times (IC50 = 1.3 μmol/ L). However,increasing the 6-carbon linker length in MM-302 to the 8-carbon linker length in MM-303 does not yield further improvement in binding affinity (IC50 = 1.2 μmol/L for MM-302 vs IC50 = 1.2 μmol/L for MM-303). Our binding data demonstrates that the linker length is critical for achieving a high binding affinity to WDR5 and the 6-carbon linker appears to be the shortest linker for constraining the peptidomimetic to the desirable conformations for effective binding to WDR5. 3.4. Testing the functional antagonism of the MLL1 H3K4 methyltransferase activity

Our previous study has shown that the interaction between WDR5 and MLL1 is essential for the MLL1 H3K4 methyltransferase activity. Since MM-302 binds to WDR5 with a very high affinity,we tested its ability to inhibit the MLL1 H3K4 methyltransferase activity using our in vitro functional assay,which consists of MLL1, WDR5,RbBP5,Ash2L,histone H3 peptide,and tritium labeled cofactor 3H-SAM [9]. MM-302 potently inhibits the MLL1 methyltransferase activity in a dose-dependent manner and achieves an IC50 value of 49 μmol/L (Fig. 3).

Download:
Fig. 3. Inhibition of the MLL1 methyltransferase activity by MM-302 in a fully constituted in vitro functional assay.

Taken together,our functional data showed that MM-302 effectively and potently inhibits the MLL1 H3K4 methyltransferase activity in our in vitro fully constituted functional assay. 4. Conclusion

In summary,we have performed structure-based design of conformationally constrained,cyclic peptidomimetics to target the WDR5-MLL1 protein-protein interaction and obtained high-affinity compounds. MM-302 and MM-303 bind to WDR5 with IC50 values of 1.3 μmol/L and 1.2 μmol/L,respectively. Their estimated Ki values are lower than 1 μmol/L and are >100-times more potent than the 12-residue WIN peptide derived fromthe MLL1 protein. in vitro functional assay showed that MM-302 is a potent inhibitor of the MLL1 H3K4 methyltransferase activity.

Our recent study has demonstrated that blocKing the WDR5- MLL1 interaction by small-molecule inhibitors can potently and selectively inhibit the MLL1 activity,leading to selective and effective inhibition of cell proliferation and induction of differentiation and apoptosis of acute leukemia cells harboring MLL1 fusion gene [7]. Hence,further optimization of MM-302 and MM- 303 may ultimately yield a new class of molecularly targeted agents for the treatment of a subset of acute leukemia,whose development and progression depend upon the MLL1 methyltransferase activity.

Acknowledgments

This work is supported by grants from the National Institutes of Health,USA (No. CA177307 to SW and YD),the Leukemia and Lymphoma Society (to SW and YD) and Stand Up to Cancer (to YD).

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version,at http://dx.doi.org/10.1016/j.cclet.2015.03.030.

References
[1] R. Marmorstein, Protein modules that manipulate histone tails for chromatin regulation, Nat. Rev. Mol. Cell Biol. 2 (2001) 422-432.
[2] T. Kouzarides, Chromatin modifications and their function, Cell 128 (2007) 693-705.
[3] G.G. Wang, C.D. Allis, P. Chi, Chromatin remodeling and cancer. Part I: Covalent histone modifications, Trends Mol. Med. 13 (2007) 363-372.
[4] P. Chi, C.D. Allis, G.G. Wang, Covalent histone modifications-miswritten, misinterpreted and mis-erased in human cancers, Nat. Rev. Cancer 10 (2010) 457-469.
[5] J.L. Hess, MLL: a histone methyltransferase disrupted in leukemia, Trends Mol. Med. 10 (2004) 500-507.
[6] C.D. Jude, L. Climer, D. Xu, et al., Unique and independent roles for MLL in adult hematopoietic stem cells and progenitors, Cell Stem Cell 1 (2007) 324-337.
[7] F. Cao, E.C. Townsend, H. Karatas, et al., Targeting MLL1 H3K4 methyltransferase activity in mixed-lineage leukemia, Mol. Cell 53 (2014) 247-261.
[8] H. Karatas, E.C. Townsend, D. Bernard, Y. Dou, S. Wang, Analysis of the binding of mixed lineage leukemia 1 (MLL1) and histone 3 peptides to WD repeat domain 5 (WDR5) for the design of inhibitors of the MLL1-WDR5 interaction, J. Med. Chem. 53 (2010) 5179-5185.
[9] H. Karatas, E.C. Townsend, F. Cao, et al., High-affinity, small-molecule peptidomimetic inhibitors of MLL1/WDR5 protein-protein interaction, J. Am. Chem. Soc. 135 (2013) 669-682.
[10] F. Cao, Y. Chen, T. Cierpicki, et al., An Ash2L/RbBP5 heterodimer stimulates the MLL1 methyltransferase activity through coordinated substrate interactions with the MLL1 SET domain, PLoS ONE 5 (2010) e14102.