STAT6 Protein–Protein Interaction

Communication Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX-XXX

pubs.acs.org/JACS

Targeted Inhibition of the NCOA1/STAT6 Protein−Protein Interaction Yeongju Lee,† Heeseok Yoon,‡ Sung-Min Hwang,§ Min-Kyung Shin,† Ji Hoon Lee,‡ Misook Oh,† Sin-Hyeog Im,*,§,∥ Jaeyoung Song,*,‡ and Hyun-Suk Lim*,† †

Department of Chemistry and Division of Advanced Material Science, Pohang University of Science and Technology (POSTECH), Pohang 37673, South Korea ‡ New Drug Development Center, Daegu Gyeongbuk Medical Innovation Foundation, Daegu 41061, South Korea § Division of Integrative Biosciences & Biotechnology, POSTECH, Pohang 37673, South Korea ∥ Academy of Immunology and Microbiology, Institute for Basic Science (IBS), Pohang 37673, South Korea S Supporting Information *

NCOA1 could be a legitimate target for modulating STAT6 transcription activity. Because overly activated STAT6 is implicated in various human diseases including inflammatory allergic diseases (e.g., asthma and atopic disease) and cancers, STAT6 signaling has been proposed as a promising therapeutic target. Several STAT6 inhibitors have been developed, which include inhibitors of STAT6 phosphorylation and inhibitors of STAT6 dimerization.7 Here we describe the development of a novel class of STAT6 inhibitors acting by antagonizing the NCOA1/STAT6 interaction. We sought to develop a cellpermeable, proteolytically stable, stapled peptide inhibitor that directly target PAS-B domain of NCOA1 protein. We demonstrate that this inhibitor is able to inhibit IL-4/IL-13/ STAT6 signaling pathway by disrupting the TF/coactivator complex between STAT6 and NCOA1 in vitro and in cells. Furthermore, we also report the first crystal structure of a stapled peptide inhibitor in complex with NCOA1 protein at 2.25 Å resolution. The previously reported crystal structure of the NCOA1/ STAT6 complex revealed that the interaction is mediated by a short peptide segment of the STAT6 transactivation domain that is comprised of an α-helical LXXLL (in which L is leucine and X is any amino acid) motif.8 We sought to design stabilized helical peptides by optimizing the LXXLL-containing STAT6 peptide sequence. To this end, we selected a 15-residue peptide as a starting point, which is corresponding to amino acids 794− 808 in STAT6 (Figure 1). Robinson and co-workers reported that this synthetic peptide with a cyclohexylalanine (Cha) residue instead of a leucine, thereby having LXXLCha motif, displayed a considerably enhanced binding affinity to NCOA1 compared to the original LXXLL-containing peptide.9 Based on the structure of the peptide/NCOA1 complex,8 we identified four pairs of solvent-exposed i and i + 4 amino acid residues, which could be replaced with an olefin-containing unnatural amino acid or cysteine for incorporation of a staple (Figure 1). We employed two types of stapling methods, hydrocarbonbased cross-linking10 and phenyl-mediated cross-linking.11 For the synthesis of hydrocarbon-linked stapled peptides yl-1−4, peptide sequences with two (S)-2-(4-pentenyl)alanine residues

ABSTRACT: The complex formation between transcription factors (TFs) and coactivator proteins is required for transcriptional activity, and thus disruption of aberrantly activated TF/coactivator interactions could be an attractive therapeutic strategy. However, modulation of such protein−protein interactions (PPIs) has proven challenging. Here we report a cell-permeable, proteolytically stable, stapled helical peptide directly targeting nuclear receptor coactivator 1 (NCOA1), a coactivator required for the transcriptional activity of signal transducer and activator of transcription 6 (STAT6). We demonstrate that this stapled peptide disrupts the NCOA1/STAT6 complex, thereby repressing STAT6-mediated transcription. Furthermore, we solved the first crystal structure of a stapled peptide in complex with NCOA1. The stapled peptide therefore represents an invaluable chemical probe for understanding the precise role of the NCOA1/STAT6 interaction and an excellent starting point for the development of a novel class of therapeutic agents.

T

he recruitment of coactivator proteins to TFs is essentially required for their transcriptional activity.1 Since dysregulated TFs are linked to many diseases, disruption of TF/ coactivator complexes, thereby repressing aberrantly activated TFs, could be an attractive therapeutic strategy.2 In general, however, development of such TF inhibitors has proven very challenging because modulation of TFs generally requires the targeting of PPIs.3 Since PPI interfaces are often relatively large and flat, they are not amenable to small-molecule modulation. Here we report the successful identification of inhibitors of the TF/coactivator interaction between STAT6 and NCOA1. STAT6 is a TF that plays a central role in regulating interleukin-4 (IL-4) and IL-13 mediated signaling pathways.4 For STAT6 transcriptional activity, recruitment of transcriptional coactivators such as NCOA1 is needed. NCOA1 (also known as steroid receptor coactivator-1) has intrinsic histone acetyl-transferase activity that makes DNA accessible to transcription machinery by acetylating histone.5 Studies showed that NCOA1 is overexpressed in many human cancers and plays an important role in tumor progression and metastasis.6 Due to its key role in activating STAT6-mediated transcription, © XXXX American Chemical Society

Received: August 22, 2017

A

DOI: 10.1021/jacs.7b08972 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

a fluorescein-labeled STAT6 peptide (Figure 2B). In the hydrocarbon-stapled peptide series, three stapled peptides (yl1, yl-2, and yl-4) showed enhanced inhibitory activity compared to cp-1. In particular, yl-2 was the most effective at disassociating the NCOA1/STAT6 interaction with the Ki value of 20 nM, while yl-3, a peptide with cross-link at positions E799 and T803, had no measurable inhibitory effect. Among the phenyl-mediated stapled peptide series, yl-6, a peptide cross-linked at positions T803 and L807, exerted the most potent activity, whereas placement of a cross-link at E799 and T803 positions diminished binding activity. These results are consistent with the trend observed in the hydrocarbon-stapled peptide series. It is noteworthy that yl-2 did not bind to the eTAFH domain of AML1-ETO protein, which also has an LXXLL-binding site (Figure S2),12 suggesting the possibility of selectivity for NCOA1. To gain structural insight into how the stapled peptides bind to NCOA1, we solved the crystal structure of PAS-B domain of NCOA1 in complex with the most active compound, yl-2, at 2.25 Å resolution by molecular replacement (Table S2). Analysis of the structure confirmed that yl-2 adopts an αhelix (Figure 3 and Figure S3). Hydrophobic residues of yl-2

Figure 1. Design of stapled peptides. Fifteen-residue linear peptide cp1 containing the LXXLCha motif was modified by incorporating (S)-2(4-pentenyl)alanine or cysteine at positions X in the sequence and subsequent cross-linking by ring-closing metathesis (for yl-1−4) or bromophenyl-mediated macrocyclization (for yl-5−8). Cha = cyclohexylalanine.

at positions i and i + 4 were synthesized by Fmoc solid-phase peptide synthesis (SPPS). The two unnatural side chains within the peptide sequences were cross-linked through on-resin ringclosing olefin metathesis (Scheme S1). Likewise, peptides cross-linked with the phenyl derivative yl-5−8 were synthesized by bromophenyl-mediated macrocyclization of peptides containing i and i + 4 cysteine residues (Scheme S2). The resultant stapled peptides were cleaved from resin and purified by HPLC (Figure S1). Initially, we examined the helicity of the synthesized stapled peptides by circular dichroism (CD) experiments. As seen in Figure 2A and Table S1, most synthesized stapled peptides

Figure 3. Crystal structure of the yl-2 and NCOA1 complex (PDB 5Y7W).

consisting of Leu794, Leu795, Pro796, Pro797, and Leu782 make contacts with the hydrophobic groove in PAS-B domain of NCOA1. The hydrophobic interaction was reinforced by a salt bridge between NCOA1 Arg311 and yl-2 backbone oxygen of Leu794 in close proximity (Figure S3C). In addition, a hydrogen bond is formed between NCOA1 Tyr297 and yl-2 Glu799 in 2.55 Å (Figure S3D). The hydrocarbon staple of yl-2 did not directly participate in hydrophobic interaction with NCOA1. The Cha residue was found to bind to the leucinebinding pocket in NCOA1 as anticipated in the previous study (Figure S3E).9 This structural data will provide valuable information for designing more potent and specific ligands for NCOA1. To explore the cell-penetrating ability of the stapled peptides, we prepared a fluorescently labeled derivative of the stapled peptide, yl-2-FL, together with the linear peptide with fluorescein tag, cp-1-FL (Scheme S1 and Figure S1). HeLa cells or Th2 cells were treated with yl-2-FL or cp-1-FL, and their cellular uptake was assessed by confocal fluorescence microscope and fluorescence-activated cell sorting experiments (Figure S4). Not surprisingly, yl-2 exhibited excellent cellular uptake comparable to (Arg)8, a cell-penetrating peptide,

Figure 2. (a) CD spectra. (b) Inhibition curves of stapled peptides yl1−8 and cp-1 for fluorescein-labeled STAT6 peptide binding to NCOA1 determined by FP assays. Error bars represent standard deviation from three independent experiments.

exhibited significantly increased α-helicity over unmodified linear peptide cp-1. Because stapling influenced the helical content of the peptides, it was anticipated that these conformationally stabilized peptides could bind more tightly to the target protein, NCOA1, without major entropy penalty. We conducted competitive fluorescence polarization (FP) assay to evaluate the ability of the stapled peptides to disrupt the interaction between recombinant PAS-B domain in NCOA1 (amino acids 257−385) containing STAT6-binding pocket and B

DOI: 10.1021/jacs.7b08972 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

reaction (RT-qPCR) experiment. A549 human lung carcinoma cells were treated with DMSO, yl-2, or leflunomide (a known STAT6 inhibitor as a positive control)14 and then stimulated with IL-4 or IL-13. The expression level of eotaxin-3, a STAT6 target gene, which was normalized to β-actin levels, was quantified by RT-qPCR. The treatment of yl-2 resulted in dosedependent decrease of eotaxin-3 expression (Figure 4B). For comparison, we used yl-3 as a negative control, a stapled peptide with no measurable binding affinity to NCOA1 (Figure 2B). As expected, yl-3 was inactive at concentrations up to 10 μM (Figure S6), indicating that the decreased eotaxin-3 expression was due to the specific activity of yl-2. The inhibitory activity of yl-2 on the expression of eotaxin-3 was confirmed using an eotaxin-3 secretion assay, which measured eotaxin-3 protein levels by enzyme-linked immunosorbent assay (ELISA). In a good agreement with the RT-qPCR result, the eotaxin-3 protein level, elevated by IL-4 or IL-13, was markedly reduced by yl-2 in a dose-dependent fashion (Figure 4C). To further probe the effect of yl-2 on transcriptional activity, a dual luciferase reporter assay was conducted using HEK293T cells or Jurkat T cells transfected with STAT6-dependent Firef ly luciferase and Renilla luciferase genes (Figure S7). In HEK293T cells, the treatment of yl-2 led to dose-dependent inhibition of STAT6-responsive luciferase activity, while yl-3 had little effect. This inhibitory activity of yl-2 was more significant in Jurkat T cells. Importantly, yl-2 showed better activity compared to leflunomide, a known STAT6 inhibitor as a positive control, suggesting that inhibition of STAT6/NCOA1 complex would be an effective means to repress STAT6 transcription activity. It is of note that STAT6-dependent luciferase activity was normalized to Renilla luciferase control, suggesting that repressed STAT6 transcription activity was not due to the general toxicity of yl-2. Indeed, yl-2 did not affect cell viability (Figure S8). Overall, the results from RT-qPCR and luciferase report assays clearly demonstrated that inhibition of the STAT6/NCOA1 interaction could effectively suppress STAT6 transcriptional activity. In summary, we have identified a cell-permeable, proteolytically stable, stapled peptide inhibitor directly targeting NCOA1. We have demonstrated that this stapled peptide disrupts the NCOA1/STAT6 complex, leading to inhibition of IL-4/IL-13/ STAT6 signaling. To the best of our knowledge, our stapled peptide represents the first cell permeable inhibitor of the NCOA1/STAT6 interaction. Furthermore, we have solved the first crystal structure of a stapled peptide in complex with NCOA1. This molecule will block the interaction between NCOA1 and STAT6, without affecting the some other interactions (e.g., STAT6/p300 interaction and STAT6/DNA interaction), whereas current STAT6 inhibitors7 and biological methods (e.g., siRNA) suppress the entire functions of STAT6 or affect the other STAT family as well. Taken together, yl-2 will serve as a highly useful chemical tool to probe the functions of the NCOA1 PAS-B domain. Further, it also provides a promising starting point for the development of a novel class of therapeutic agents for the treatment of various human diseases including allergic diseases and cancers.

whereas cp-1 was impermeable. Cell penetration of both yl-2 and (Arg)8 was dose-dependently increased at the concentration range of 1−10 μM like other stapled peptides.13 Another important feature of stapled peptides is their increased proteolytic stability over unmodified peptides. To test the proteolytic resistance, yl-2 or cp-1 was incubated with fetal bovine serum at 37 °C. The serum stability of the peptides was analyzed by LC/MS. As expected, the stapled peptide yl-2 exhibited considerably enhanced proteolytic stability, whereas cp-1 was almost completely degraded within 4 h (Figure S5). Because the ability of yl-2 to antagonize the NCOA1/STAT6 interaction was examined in vitro using a recombinant protein and a short peptide, we sought to determine whether the stapled peptides could inhibit the interaction between endogenous proteins in living cells. For this work, we performed coimmunoprecipitation (Co-IP) assays. Cultured human embryonic kidney (HEK) 293T cells were treated with increasing concentrations of yl-2, cp-1, or DMSO as a vehicle control. Cell lysates were immunoprecipitated with a STAT6specific antibody, and the precipitated proteins were analyzed by Western blot probing. In yl-2-treated cells, immunoprecipitated NCOA1 was dose-dependently decreased, while cp-1 had no effect (Figure 4A). This result verified that yl-2 is cell-

Figure 4. Cellular activities of yl-2. (a) Co-IP assay. HEK293T cells were treated with yl-2 or cp-1 for 6 h. The effect of yl-2 on the NCOA1/STAT6 interaction was analyzed by STAT6 immunoprecipitation and NCOA1 western analysis. The results are representative of three independent experiments. (b) RT-qPCR experiment. Eotaxin3 and β-actin mRNAs (as a reference gene) were measured in A549 cells treated with yl-2 or leflunomide (a known STAT6 inhibitor). (c) yl-2 inhibits IL-4 or IL-13-dependent eotaxin-3 protein secretion in A549 cells in a dose-dependent manner as determined by ELISA experiments.

permeable and able to effectively target endogenous NCOA1 in a complex protein mixture in living cells, leading to the inhibition of the native PPI. Note that yl-2 showed micromolar range activity, whereas it has far more potent in vitro activity (Ki = 20 nM). This discrepancy in its activity between in vitro and cellular inhibition could be due to its somewhat limited cell-penetration ability. Next, we determined whether yl-2 as an inhibitor of STAT6/ NCOA1 had the ability to repress STAT6 transcription activity. First, we performed a quantitative real time polymerase chain

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b08972. Experimental details and characterization data (PDF) C

DOI: 10.1021/jacs.7b08972 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

■

(12) Park, S.; Chen, W.; Cierpicki, T.; Tonelli, M.; Cai, X.; Speck, N. A.; Bushweller, J. H. Blood 2009, 113, 3558−3567. (13) Chu, Q.; Moellering, R. E.; Hilinski, G. J.; Kim, Y.-W.; Grossmann, T. N.; Yeh, J. T. H.; Verdine, G. L. MedChemComm 2015, 6, 111−119. (14) Siemasko, K.; Chong, A. S.-F.; Jäck, H.-M.; Gong, H.; Williams, J. W.; Finnegan, A. J. Immunol. 1998, 160, 1581−1588.

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] *[email protected] ORCID

Hyun-Suk Lim: 0000-0003-4083-2998 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Prof. John A. Robinson (University of Zürich) for the plasmid expressing NCOA1 and Prof. Won Jong Kim (POSTECH) for providing access to a CD spectrometer and a RT-PCR system. This work was supported by the National Research Foundation of Korea (NRF-2017R1A2B3004941 and NRF-2017M3A9G405 2952).

■

REFERENCES

(1) Näar̈ , A. M.; Lemon, B. D.; Tjian, R. Annu. Rev. Biochem. 2001, 70, 475−501. (2) (a) Majmudar, C. Y.; Højfeldt, J. W.; Arevang, C. J.; Pomerantz, W. C.; Gagnon, J. K.; Schultz, P. J.; Cesa, L. C.; Doss, C. H.; Rowe, S. P.; Vásquez, V.; Tamayo-Castillo, G.; Cierpicki, T.; Brooks, C. L.; Sherman, D. H.; Mapp, A. K. Angew. Chem., Int. Ed. 2012, 51, 11258− 11262. (b) Henchey, L. K.; Kushal, S.; Dubey, R.; Chapman, R. N.; Olenyuk, B. Z.; Arora, P. S. J. Am. Chem. Soc. 2010, 132, 941−943. (c) Shimogawa, H.; Kwon, Y.; Mao, Q.; Kawazoe, Y.; Choi, Y.; Asada, S.; Kigoshi, H.; Uesugi, M. J. Am. Chem. Soc. 2004, 126, 3461−3471. (3) (a) Milroy, L.-G.; Grossmann, T. N.; Hennig, S.; Brunsveld, L.; Ottmann, C. Chem. Rev. 2014, 114, 4695−4748. (b) Arkin, M. R.; Wells, J. A. Nat. Rev. Drug Discovery 2004, 3, 301−317. (c) Berg, T. Curr. Opin. Chem. Biol. 2008, 12, 464−471. (4) (a) Hou, J.; Schindler, U.; Henzel, W.; Ho, T.; Brasseur, M.; McKnight, S. Science 1994, 265, 1701−1706. (b) Foster, P. S.; Martinez-Moczygemba, M.; Huston, D. P.; Corry, D. B. Pharmacol. Ther. 2002, 94, 253−264. (c) Takeda, K.; Tanaka, T.; Shi, W.; Matsumoto, M.; Minami, M.; Kashiwamura, S.-i.; Nakanishi, K.; Yoshida, N.; Kishimoto, T.; Akira, S. Nature 1996, 380, 627−630. (5) (a) Litterst, C. M.; Pfitzner, E. J. Biol. Chem. 2001, 276, 45713− 45721. (b) O'Malley, B. W.; Spencer, T. E.; Jenster, G.; Burcin, M. M.; Allis, C. D.; Zhou, J. X.; Mizzen, C. A.; McKenna, N. J.; Onate, S. A.; Tsai, S. Y.; Tsai, M. J. Nature 1997, 389, 194−198. (c) Korzus, E.; Torchia, J.; Rose, D. W.; Xu, L.; Kurokawa, R.; McInerney, E. M.; Mullen, T.-M.; Glass, C. K.; Rosenfeld, M. G. Science 1998, 279, 703− 707. (6) (a) Qin, L.; Liu, Z.; Chen, H.; Xu, J. Cancer Res. 2009, 69, 3819− 3827. (b) Xu, J.; Wu, R.-C.; O’Malley, B. W. Nat. Rev. Cancer 2009, 9, 615−630. (7) (a) Kuperman, D.; Schleimer, R. Curr. Mol. Med. 2008, 8, 384− 392. (b) Benekli, M.; Baer, M. R.; Baumann, H.; Wetzler, M. Blood 2003, 101, 2940−2954. (c) Bruns, H. A.; Kaplan, M. H. Crit. Rev. Oncol. Hematol. 2006, 57, 245−253. (d) Miklossy, G.; Hilliard, T. S.; Turkson, J. Nat. Rev. Drug Discovery 2013, 12, 611−629. (8) Razeto, A.; Ramakrishnan, V.; Litterst, C. M.; Giller, K.; Griesinger, C.; Carlomagno, T.; Lakomek, N.; Heimburg, T.; Lodrini, M.; Pfitzner, E.; Becker, S. J. Mol. Biol. 2004, 336, 319−329. (9) Seitz, M.; Maillard, L. T.; Obrecht, D.; Robinson, J. A. ChemBioChem 2008, 9, 1318−1322. (10) (a) Schafmeister, C. E.; Po, J.; Verdine, G. L. J. Am. Chem. Soc. 2000, 122, 5891−5892. (b) Stewart, M. L.; Fire, E.; Keating, A. E.; Walensky, L. D. Nat. Chem. Biol. 2010, 6, 595−601. (11) Jo, H.; Meinhardt, N.; Wu, Y.; Kulkarni, S.; Hu, X.; Low, K. E.; Davies, P. L.; DeGrado, W. F.; Greenbaum, D. C. J. Am. Chem. Soc. 2012, 134, 17704−17713. D

DOI: 10.1021/jacs.7b08972 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX