Downloaded by guest on September 30, 2021 www.pnas.org/cgi/doi/10.1073/pnas.1818798116 a Arora response Reety Park Esther the Donglim enhances virus-positive carcinoma in cell MDM4 Merkel and of inhibition Dual h rsneo idtp 5 nvrspstv C,w sus- we MCC, virus-positive Given in 8). (7, p53 mutations wild-type UV-induced of for presence In evidence the and 8). no RB (7, wild-type and signature contains p53 mutational usually UV MCC a virus-positive contrast, with and along p53 of harbors mutations sequencing typically RB Next-generation MCC virus-negative (6). MCV that p53 inacti- reveals not contrast, and MCC but In RB, (TP53) to (6). p53 binds functions LT antigen tumor-suppressive tumor RB their cellular vates the retinoblastoma-associated the and to (RB1) binds LT (SV40) (5). MCC types, can in cell roles LT oncogenic several truncated their in with and consistent transformation ST and MCV proliferation viral of promote for C- Expression required the (3). domains deletes replication helicase but and motif, DNA-binding RB-binding terminal retains LXCXE, LT N-terminal truncated large tumor-associated the of MCC MCV- form genome (4). truncated (LT) tumors. a MCV antigen and the T (ST) MCC antigen of T 10 small copies expresses of integrated and 8 contains In MCC 2). in positive (1, integrated (MCV; decades polyomavirus clonally two cell Merkel last MCPyV) discovered the (3) al. in et tripled Feng 2008, has that States United M lenalidomide p53 p53 other and MCC virus-positive tar- in tumors. dual MDM4 wild-type support and and results MDM2 cells of These MCC geting mice. MCV-positive in in xenografts response MCC-derived apoptotic an to leading targeting lenalidomide CK1α with of inhibitors levels an MDM2 high com- expressed Combining ST–MYCL–EP400 CK1α, cells MDM4. the MCC as of MCV-positive addition, well In target plex. as as MDM4, MDM2 of identified activator depletion fol- EP400 analysis RNA-sequencing lowing and mechanism ChIP underlying Integrated the known. target not but was p53 p53, inactivate complex increased in ST–MYCL–EP400 functionally MCV to EP400 could the of that led suspected depletion lines We expression. that transactivates cell observed and MCC We complex MCV-positive (L-Myc) genes. remodeler MYCL target chromatin homologue specific MDM2, MYC EP400 of the p53 the to reduced recruits inhibitor to ST binds ST an LT of ARF, MCV coexpression MCV However, activation. of that p53. here levels of report activation increased and We inhibit to p53. and leading bind bind can RB, not LT does MCV-truncated antigen In it the T in (RB1). RB, While mutations RB large RB1. inactivating and and of contains (TP53) TP53 form MCC p53 truncated virus-negative wild-type a contrast, contains and usually and (ST) expresses (LT) antigen MCC MCV-positive T skin. small the neu- approximately aggressive of highly to a carcinoma (MCCs), roendocrine carcinomas 2018) contributes cell 2, Merkel November (MCV) all review of for 80% polyomavirus (received 2018 3, cell December approved Merkel and NY, York, New University, Columbia Prives, Carol by Edited Switzerland; g Basel, CH-4056 Research, 02215; MA d Boston, Institute, Cancer rga nVrlg,Gaut colo rsadSine,HradUiest,Cmrde A02138; MA Cambridge, University, Harvard Sciences, and Arts of School Graduate Virology, in Program xeietlTeaetc oe efrCne o ple acrSine aaFre acrIsiue otn A02215 MA Boston, Institute, Cancer Dana-Farber Science, Cancer Applied for Center Belfer Core, Therapeutics Spain; Experimental Barcelona, 08028 Catalonia, of Bioengineering for Institute Group, Nanobioengineering h rttpcploaiu iinvcoaigvrs40 virus vacuolating Simian polyomavirus prototypic The | eklcl carcinoma cell Merkel ra D4ihbtrcue yegsi ciaino p53, of activation synergistic caused inhibitor MDM4 an or orn acnm ftesi iha niec nthe in incidence an neuroen- with aggressive skin an the is of carcinoma (MCC) docrine carcinoma cell erkel b,f ihleL Tillgren L. Michelle , a,b | ige Cheng Jingwei , MDM2–MDM4 c eateto eiie rga n oe’ optl avr eia col otn A02115; MA Boston, School, Medical Harvard Hospital, Women’s and Brigham Medicine, of Department f ainlCne o ilgclSine,Tt nttt fFnaetlRsac,506 aglr,Ida and India; Bangalore, 560065 Research, Fundamental of Institute Tata Sciences, Biological for Center National g rflaC Gokhale C. Prafulla , | aenkns alpha 1 kinase casein b,c hita Berrios Christian , | g n ae .DeCaprio A. James and , a,b onMontero Joan , 1073/pnas.1818798116/-/DCSupplemental. y at online information supporting contains article This 1 the under Published Submission.y Early Direct Alliance PNAS work.y a India related is the for DBT article with This Institute Trust filed was Novartis Wellcome patent by by A IA/E/14/1/501773. employed supported Fellowship are Career was S.F. R.A. and Research. M.C.-C. Biomedical statement: interest of new Conflict contributed wrote R.A. J.A.D. and and D.E.P. S.F., and paper.y data; M.C.-C., the J.C., analyzed M.L.T. D.E.P., C.B., and J.M., research; J.C., D.E.P., tools; research; designed reagents/analytic performed J.A.D. M.L.T. and and P.C.G., J.M., M.C.-C., D.E.P., contributions: Author diploid doxycycline- IMR90 a into introduced cells, truncated was normal LT tumor-derived of in or forms p53 GFP full-length or on expressing antigens vector inducible T MCV Response. p53 of the Dampens ST and Activates LT Discussion and Results MCC. and in MDM2 We both MDM4 MCC. targeting of in efficacy function CK1α synergistic p53 the and inhibit demonstrate to MDM2 MDM4 of with levels cooperate turn, increase to activator be tional (13). can CSNK1A1) interaction (CK1α, bind- alpha autoinhibitory 1 reduces MDM4 kinase casein The that by (12). relieved domain p53 autoinhibitory (11). to MDM2 an ing to has ubiquitin RING also of its MDM4 (11). recruitment p53, the ubiquitinate domains with facilitates directly RING domain structures not C-terminal does similar MDM4 and have Although both binding (MDMX) MDM4 MDM4 p53 of or and N-terminal overexpression p53, MDM2 targeting by 10). ligase (9, inactivated ubiquitin functionally a be MDM2, can p53 type p53 inactivate functionally could antigens T activity. MCV that pected owo orsodnesol eadesd mi:james Email: addressed. be edu. should correspondence whom To eidMVcnrlo 5 nMCaddmntae the demonstrates p53 in and combinatorially MCC MDM4 tumors. and in wild-type MDM2 p53 targeting of of utility control mechanism the MCV uncovers work behind Our . activate induce to and inhibitors p53 MDM2 Targeted with MDM4. synergistically increasing of acts inhibitor by activator CK1α an activation of CK1α, p53 degradation and reduces while MDM2 activation, of antigen p53 levels in T results small RB to MCV binding MCV that antigen demonstrate T We large p53. usu- and MCC RB MCV-positive wild-type contains p53, ally and RB in mutations contains inactivating MCC (MCV)] polyomavirus While cell cancer. [Merkel skin virus-negative aggressive an is (MCC) carcinoma cell Merkel Significance ee edmntaeta C Tfntosa transcrip- a as functions ST MCV that demonstrate we Here, wild- Alternatively, cancers. of variety wide a in mutated is p53 y e ies raOclg,Nvri ntttsfrBiomedical for Institutes Novartis Oncology, Area Disease NSlicense.y PNAS b,c,d at Cort Marta , b a,b,c,1 eateto eia nooy Dana-Farber Oncology, Medical of Department ylnldmd raseicMDM4 specific a or lenalidomide by es-Cros ´ www.pnas.org/lookup/suppl/doi:10. e NSLts Articles Latest PNAS St , paeFerretti ephane ´ osuyteeffect the study To [email protected]. ht in that, | f6 of 1 e ,

MICROBIOLOGY lung fibroblasts (Fig. 1A and SI Appendix, Fig. S1A). The trun- changes in gene expression of known p53 target genes (9). cated forms of LT include L21 (encoding residues 1–292), 162 EP400 depletion led to increased levels of many p53 target (residues 1–320), and 168 (residues 1–275), each containing an genes, including p21 (CDKN1A) (Fig. 2A). By using the fold intact LXCXE motif (6). LT-L21 or -162 expression in IMR90 change cutoff of 1.5, a total of 59 genes were up-regulated significantly increased levels of ARF and several p53 target and 17 were down-regulated of 198 total p53 target genes genes, including GDF15 and p21 (CDKN1A), as assessed by (SI Appendix, Table S1) (9). EP400 depletion also led to a quantitative RT-PCR (RT-qPCR) (Fig. 1A). Inhibition of RB decrease in MDM2 E3 ligase and CK1α levels (Fig. 2B and activates the E2F transcription factors, leading to increased SI Appendix, Fig. S2 A–C). MDM2 and MDM4 contain an levels of ARF (10). ARF is a potent inhibitor of the major N-terminal p53-binding domain that binds directly to the trans- p53-degrading E3 ligase MDM2 (14). LT expression increased activation domain of p53 to block p53 activation (11). MDM4 protein levels of p53 as well as phospho-serine 15 p53 (P-p53) can regulate its own activity toward p53. The p53-binding domain and p21, indicative of p53 activation. Expression of LT-162 of MDM4 forms an intramolecular interaction with its central also led to increased levels of cleaved PARP(∗∗), indicating an W motif region and thereby reduces binding to p53 (12). CK1α apoptotic response. (CSNK1A) is a serine/threonine kinase that binds and phos- LT and ST are coexpressed as splice variants from the inte- phorylates MDM4, which in turn prevents this autoinhibitory grated MCPyV viral DNA in MCC tumors. To mimic this in interaction and activates MDM4 (13). RT-qPCR and Western IMR90 cell lines, we introduced a genomic version of LT-L21 blotting confirmed that p21 levels increased and MDM2 and that coexpresses ST. When ST was coexpressed with LT-L21, CK1α levels decreased upon EP400 knockdown in MKL-1 cells lower levels of p53 activation (p21, P-p53, and Ac-p53) were (Fig. 2B and SI Appendix, Fig. S2C). observed compared with the response to LT-L21 only (Fig. 1B ChIP-sequencing (ChIP-seq) with antibodies to ST, MAX and SI Appendix, Fig. S1B). The results indicate that truncated (which dimerizes with MYCL), and EP400 revealed enrichment LT can activate p53, while ST can reduce this response when for the MDM2 and CK1α promoters (SI Appendix, Fig. S2E) coexpressed. (15). We performed ChIP-qPCR for MAX, ST, and EP400 and To determine whether increased levels of ARF required LT observed specific enrichment for these promoters (Fig. 2C). We binding to RB, we introduced a point substitution mutation in depleted ST or ST and LT using specific shRNAs in MKL-1 the LXCXE motif (E216K) of LT-L21. When stably expressed cells and found that the levels of MDM2 and CK1α decreased in HCT116 cells, L21-E216K was unable to coprecipitate RB, (Fig. 2D) (16). This result together with the RT-qPCR and ChIP but retained binding to VPS39 that binds to a different region data indicates that MDM2 and CK1α are direct transcriptional of LT (SI Appendix, Fig. S1C) (5). When expressed in IMR90 targets of the ST–MYCL–EP400 complex. Of note, MDM4 lev- cells, L21-E216K did not increase levels of ARF, p21, p53, or els decreased upon depletion of ST, although we did not find Ac-p53 (Fig. 1C). These results indicate that LT binding to RB evidence for direct activation of MDM4 by ST. ST increases contributes to increased levels of ARF and activation of the p53 levels of CK1α that could serve to activate MDM4 activity response. toward p53. Since MDM2 is a p53 target gene, it is possible that the MDM2 and CK1α Are Transcriptional Targets of the ST–MYCL–EP400 MCV T antigens indirectly increase MDM2 levels by activating Complex. We recently reported that MCV ST recruits the MYC p53 (9). To exclude this possibility, we introduced a dominant- homologue MYCL (L-Myc) to the EP400 chromatin remod- negative p53 (p53DD) that binds and inactivates the endogenous eler complex to bind specific gene promoters and activate their p53 into IMR90 cells (17). The IMR90-p53DD cells were fur- expression (15). To identify genes regulated by the ST–MYCL– ther transduced with MYCL and MCV LT-L21 with ST (17). EP400 complex in MKL-1 cells, RNA-sequencing (RNA-seq) We detected ST binding to the MDM2 and CK1α promot- was performed after depleting EP400 by using three different ers by ChIP-qPCR and observed that EP400 enrichment to the shRNAs (15). Using the reported RNA-seq results, we assessed MDM2 promoter increased in the presence of MCV T antigens

LT LT+ST A B GFP ST LT LT+ST C GFP ST LT LT+ST E216K E216K DOX - + - + - + - + DOX - + - + - + - + - + - + LT ST

ST

P-p53 LT

Ac-p53 Ac- p53

p53 MDM2 p21 p21

PARP ** ARF Vinculin Vinculin

Fig. 1. MCPyV LT activates and ST dampens the p53 response. (A) Inducible expression of truncated tumor isoforms of MCV LT increases ARF and p53 target genes in IMR90 cells. Expression of GFP or LT-L21 and -162 truncated LT was induced with doxycycline (DOX) treatment for 24 h. The LT, ARF, and p53 target gene RNA levels were normalized to those of the GFP-induced cells, whereas the GFP levels were normalized to the LT-L21 samples. Data are shown as mean ± SD. *P < 0.05; **P < 0.005; ***P < 0.0005; ****P < 0.00005 (Student t test). (B) IMR90 cells were induced to express GFP, ST, LT-L21, or LT-L21 with ST for 40 h. Lysates were prepared before (−) or after (+) DOX. Activation of p53 response is reflected by increased levels of p53, P-p53, acetyl-lysine 382 p53 (Ac-p53), p21, and cleaved PARP (**). (C) Expression of L21, but not an LT mutant in the LXCXE motif (E216K), activates p53 through inhibiting RB and inducing ARF.

2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1818798116 Park et al. Downloaded by guest on September 30, 2021 Downloaded by guest on September 30, 2021 IAppendix, (SI lines al. et virus-positive Park in higher MDM4- was the ratio Furthermore, 3A). -S MCC13, (Fig. to MDM4 (UISO, lines FL total virus-negative MCC of the MCC26) levels to lines and higher relative cell significantly MDM4-FL MCC had and lines in BroLi) and cell PeTa, levels WaGa, MCC MS-1, domain) -2, (MKL-1, RING (short Virus-positive (18). the MDM4-S and (all missing MDM4 (18). (full-length), total variant p53 assess MDM4-FL wild-type to variants), primers with MDM4 splice cancers of sets some three in used We found be MCC. CK1α can Virus-Positive MDM4 in and Overexpressed not MDM2 Is did MDM4 the depletion (SI to p53 ST enrichment that ST and found promoters. and EP400 We EP400 of with S2G). affect MKL-1 ChIP-qPCR Fig. in pro- performed p53 Appendix, these depleted and to we p53, shRNA and 2F binds of (Fig. complex P-p53, independently activated moters p53DD ST–MYCL–EP400 become p53, of to the increase failed whether presence p53 the not that in did indicating levels, nutlin-3 PUMA with treatment 2E (Fig. (1 nutlin-3 with treated were IMR90-p53DD times and three MKL-1 performed p53DD. was expressing ChIP-qPCR (F cells. (E times. LT. MKL-1 and five in ST promoters performed or indicated ST of targeting (qPCR) shRNAs PCR ( with quantitative shown. by input followed percent antibodies average IgG with and ST, EP400, MAX, with TqC a efre ihML1clsatrsRAwsidcdfr8d ed eenraie oRL0aduidcdsmls h xeietwas experiment The against samples. plotted uninduced was and (FC) RPLP0 change to normalized fold were mean log2 Reads as gene d. shown Each 8 are shRNA. for Data induced control averaged. was to and shRNA relative times after three shRNA cells performed inducible MKL-1 with with performed EP400 was of RT-qPCR depletion after lines the cell MCC MKL-1 in 2. Fig. C A −log10 D2adCK1α and MDM2 and P au o ttsia infiac.Gendt niaegnsta ettetoodcag uof n e ossgiyadjusted signify dots red and cutoff, change twofold the meet that genes indicate dots Green significance. statistical for value i.S2F Fig. Appendix, SI uln3tetetde o lctp3rsos,btMVTatgn nraelvl fMM,MM,adCK1α and MDM4, MDM2, of levels increase antigens T MCV but response, p53 elicit not does treatment Nutlin-3 ) ocn ltilsrtn ifrnilyepesdp3tre genes target p53 expressed differentially illustrating plot Volcano (A) complex. ST–MYCL–EP400 the of targets transcriptional are targets Non-ST n D4lvl.ML1clswr transduced were cells MKL-1 levels. MDM4 and CK1α, MDM2, the in reduction a causes antigens T MCV of Depletion D) .Hwvr D2inhibitor MDM2 However, ). hPqC o h D2pooe nIR0clsi h rsne()o bec ( absence or (+) presence the in cells IMR90 in promoter MDM2 the for ChIP-qPCR ) Vinculin .T ute determine further To ). D shRNA MDM4 MDM2 CK1α ST LT vrxrsinof Overexpression ± D *P SD. p53 Target Genes < )fr2 h. 24 for µM) E .5 **P 0.05; h rsneo eaioie(1 2.W sesdwhether assessed We 22). (21, CK1α lenalidomide target of specifically presence (Cereblon) inactivat- the can CRBN an the ligase that harboring E3 demonstrated cells reports Recent MS-1 2F (20). (Figs. containing in mutation lines not p53 cell ing but MCC BroLi p53, PUMA and and wild-type p21, PeTa, Ac-p53, WaGa, P-p53, MKL-1, p53, in shown total as of p53 levels activated increased by treatment MCC. inhibitor in MDM2 p53 nutlin-3 Activates that MDM4 and MDM2 of Inhibition Fig. , Appendix (SI MCV of (19). presence copies) S3B) the eight of to regardless (three MDM4 gains of low-copy MCC frequent virus-positive 101 revealed of tumors in (Oncopanel) a p53 sequencing expression. lev- MDM4 next-generation of had and abundant Targeted MDM2 activation line on express dependency LT’s create not cell may protein. did MCC UISO MDM4 but of ratio, virus-negative -S els the lev- to note, lines high MDM4-FL cell high Of expressed MCC virus-negative 3B). lines with (Fig. cell compared MDM4-FL MCC MCC of these els Virus-positive in levels lines. protein MDM4-FL cell assessed We S3A). Fig. < .0;***P 0.005; < .05 ****P 0.0005; B Vinculin Nutlin-3 F MDM4 MDM2 PUMA P-p53 n 4A and CK1α p53 ST LT < .00 (Student 0.00005 and fMVTatgn.CI was ChIP antigens. T MCV of −) + + - + - - MKL-1 NSLts Articles Latest PNAS i.S2B) Fig. Appendix, SI o bqiiainin ubiquitination for MCV T- IMR90-p53DD IMR90-p53DD eobserved We t nIR0cells IMR90 in et.(C test). P MCV T+ < | ..(B) 0.1. ChIP ) f6 of 3

MICROBIOLOGY ABMCV+ MCV-

MDM4-FL

p53

ST

Vinculin

Fig. 3. MDM4 is overexpressed in MCV-positive MCC. (A) RNA from MCC cell lines and human foreskin fibroblasts (HFF) was harvested for RT-qPCR for MDM4 (total), MDM4-FL (full-length variant), and MDM4-S (short splice variant). MDM4 levels were normalized with the geomean of RPLP0, 18s rRNA, and beta-actin RNA controls. Data are shown as mean ± SD. *P < 0.05 for MDM4-FL (Student t test). (B) Western blot of MCC cell lines and HFF with indicated antibodies. MS-1 and MCC13 overexpress p53 due to inactivating mutations, and MKL-2 and MCC26 do not express detectable levels of p53.

lenalidomide could decrease CK1α levels and activate p53 in Lenalidomide treatment alone led to a modest increase in MCC cells. MKL-1 cells were treated with lenalidomide with p53 levels. However, when nutlin-3 and lenalidomide were or without cycloheximide to block protein synthesis. Although combined, larger increases in p53 and p53 target genes were CK1α levels did not change appreciably with cycloheximide observed in the cell lines with wild-type p53 (Fig. 4A and SI treatment alone for 6 h, levels rapidly decreased following the Appendix, Fig. S4B). We assessed p53 stability in MKL-1 cells addition of lenalidomide (SI Appendix, Fig. S4A). treated with nutlin-3, lenalidomide, or both in the presence of 2 A Nutlin-3 - - + + C Len -- + - - E - Len - + - + Pom -- -+ - CK1α-1 CK1α-2 Control CK1α Control Control CK1α-1 CK1α-2 -- - - + CK1α-1 P-p53 Thal sgRNA Nutlin-3 - + + + + CK1α Ac-p53 P-p53 P-MDM2 p53 p53 P-p53 p21 P-MDM2 p53 PUMA p21 p21 Vinculin MDM4 MDM4 Vinculin B MKL-1 UISO CK1α Nutlin-3 - + - - + + Nutlin-3 - - - + + + + + + Len - - - + - + Vinculin Len - - --+ -+ -+ p21 F DMSO Nutlin-3 HDM201 D Input MDM4 p53 CK1α Len -+ - - + - - + -

IgG -- +- - + - - + P-p53 Nutlin-3 - + + - - + + - + + - + + UR99 Len --+ ---+ - -+ - - + P-p53 p53 MDM4 p53 P-MDM2 P53 p21 MDM4 PUMA CK1α CK1α MDM4 ST P-MDM2 LT Vinculin Vinculin Vinculin

Fig. 4. Inhibition of MDM2 and MDM4 enhances p53 activation in MCC cell lines. (A) Lenalidomide enhances p53 activation by nutlin-3 in MKL-1. MKL-1 cells were treated with nutlin-3 (5 µM), lenalidomide (Len; 10 µM), or both for 40 h. (B) MKL-1 (MCV+ MCC) or UISO (MCV− MCC) cells were treated with nutlin-3 (1 µM) with or without lenalidomide (10 µM) for 24 h. Of note, UISO has less MDM2 and MDM4 than MKL-1. (C) Lenalidomide and, to a lesser degree, pomalidomide, but not thalidomide, cooperated with nutlin-3 to activate p53. MKL-1 cells were treated with nutlin-3 with lenalidomide, pomalidomide (10 µM), or thalidomide (10 µM) for 24 h. (D) Lenalidomide treatment reduced MDM4 binding to p53 and activated MDM2. MKL-1 cells were treated with nutlin-3 (5 µM), lenalidomide (10 µM), or both for 40 h and harvested for immunoprecipitation (IP) with antibodies to MDM4, p53, and CK1α followed by Western blotting. (E) Depletion of CK1α by CRISPR transduction enhanced p53 activation by nutlin-3. MKL-1 cells stably expressing each of two CK1α single-guide RNAs (sgRNAs) were treated with nutlin-3 (1 µM) with or without lenalidomide for 24 h. Lenalidomide further decreased CK1α that sgRNAs did not completely deplete. (F) MDM4 inhibitor SC-24-UR99 (UR99) cooperated with MDM2 inhibitors in activating p53. MKL-1 cells were treated with nutlin-3 (1 µM) or HDM201 (0.1 µM) with or without lenalidomide (1 µM) or UR99 (0.1 µM).

4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1818798116 Park et al. Downloaded by guest on September 30, 2021 Downloaded by guest on September 30, 2021 ake al. et Park pathway. mg/kg), p53–MDM2–MDM4 (50 the lenalidomide 200 targeting mg/kg), were for (40 mean tumors MCC HDM201 as sensitize xenograft lenalidomide. treatments). shown antigens when and combination are starting HDM201 and daily, Data of HDM201 orally times. treatment administered three between combinational the were performed to (C drugs was respond both experiment thalidomide. mice The or with SCID h. in not 16 xenografts but for MCC drugs MKL-1 SC-24–UR99, both or or lenalidomide nutlin-3, lenalidomide, and nutlin-3 between synergy CK1α of of levels reduced that CK1α suggests This untreated controls. with nutlin-3–treated compared MDM2 or activated p53 or of CK1α, lenalidomide coprecipitation MDM4, p53, increase by not of did levels nutlin-3 with increased cotreatment the Despite only. nutlin-3 CK1α reduced lenalidomide nutlin-3 4D and (Fig. nutlin-3 h with 40 CK1α treated for and cells p53, MKL-1 MDM4, from pared for IP performed we 4C). (Fig. cells MKL-1 in nutlin-3 p53 the to enhance response not did reduce thalidomide or pomalidomide not thermore, did thalidomide less CK1α and much pomalidomide CK1α were lenalidomide, recruiting related pomalidomide of the and capable while CK1α, thalidomide with compounds interaction the promote strongly 4B). (Fig. cells UISO in nutlin-3 CK1α by activation the decreased mide uln3 G38 rAG3,adXTasywspromdatr9 ftetet **P treatment. of h 96 after performed was assay XTT and AMG232, or RG7388, nutlin-3, 5. Fig. little express which cells, UISO 3 (Fig. virus- wild-type MDM4 in and p53 MDM2 CK1 activation and p53 of We enhances Depletion negative p53. lenalidomide (23). toward whether activities nutlin-3 tested MDM4’s of decrease of may presence stability lenalidomide the the , Appendix in increased (SI significantly p53 blotting Lenalidomide S4D). Western Fig. quantitative by cycloheximide fCK1α If CK1α of contribution the test To could lenalidomide that revealed CRBN of analysis Structural .Fur- S4C). Fig. Appendix, (SI cells MKL-1 in levels protein erae D4bnigt p53. to binding MDM4 decreased K- n S1clswr rae ihMM inhibitors, MDM2 with treated were cells MS-1 and MKL-1 (A) apoptosis. by death cell induces synergistically CK1α-MDM4 and MDM2 of Inhibition ol eueMM idn op3adehnep53 enhance and p53 to binding MDM4 reduce would nbe D4bnigt 5,w xetdta loss that expected we p53, to binding MDM4 enables and .Lnldmd and Lenalidomide S4E). Fig. Appendix, SI eesadicesdp3lvl eaieto relative levels p53 increased and levels A eesbtfie oehnep53 enhance to failed but levels and oCB 2) oprdwith Compared (22). CRBN to .W bevdta lenalido- that observed We B). oMM idn op53, to binding MDM4 to # h td a emntdbcuetetmrvlm ece aiu emsil ie (E size. permissible maximum reached volume tumor the because terminated was study The ihlstspre- lysates with α by uln3 G38 rAG3 o 6hadpromda2, inhibitors a MDM2 performed and type) the h wild with 96 (p53 for cells MKL-1 AMG232 mutant) or treated RG7388, (p53 we nutlin-3, Apo- viability, MS-1 Induce cell and to MCC Lines. Inhibitors on Cell MCC MDM2 in with ptosis Synergizes Lenalidomide increased UR99 inhibitors. MDM2 by the MDM4 inhi- can by of growth activation it inhibition (50% p53 that nM MDM4, 21 found of of We potency bition). levels a nM high with 13 proliferation express inhibit was which potency cells, its OCI-AML2 p53, the and measuring (IC MDM4 assay (BRET) between transfer interaction energy resonance cence (IC was nM FRET 1.4 MDM2-p53 was time-resolved UR99 p53, biochemical of and potency MDM4 a between the interaction In the MDM2 4F p53. measuring assay of to (Fig. FRET binding activities HDM201 MDM4 the or blocks enhance nutlin-3 inhibitor can inhibitors MDM4 specific UR99) a whether (SC-24-UR99, tested We activation. marks the p53 increased of by further and either CK1α CRISPR by CK1α, depleted reduced completely of lenalidomide Notably, of reduction (25). activation addition a p53 enhances that sgRNAs, or indicating lenalidomide P-p53, p21, P-MDM2), and 166; (phospho-serine p53, MDM2 activated of of knockout els CRISPR (24). 4E) fol- CK1α (Fig. activation treatment p53 nutlin-3 assessed lowing and sgRNAs CK1α CRISPR depleted we independent this, test To (13). activation 50 ,adi was it and ), e oicesdp3atvt sesdb nrae lev- increased by assessed activity p53 increased to led mm < H rfiigwspromdwt K- el rae with treated cells MKL-1 with performed was profiling BH3 ) 3 .0 (multiple 0.005 aaaesona mean as shown are Data . 000n nteMM-5 RT In BRET. MDM2-p53 the in nM >10,000 ots h feto D2inhibition MDM2 of effect the test To 000n.I ellrbiolumines- cellular a In nM. >10,000 t ls yeg etdsly strong a displays test synergy Bliss (B) test). ± D *P SD. ± 50 E.*P SEM. ,wieisptnyi the in potency its while ), NSLts Articles Latest PNAS < rti htwsnot was that protein .U9 specifically UR99 ). .5(Student 0.05 < nML1b two by MKL-1 in .5(multiple 0.05 Cy T MCPyV ) t et.(D) test). | f6 of 5 t test

MICROBIOLOGY 3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5- myeloma (21). It was reported that mutant CSNK1A1 predicts carboxanilide (XTT) viability assay (26). MKL-1, but not MS-1, for a poor prognosis in lenalidomide-treated MDS (32). Further- cells were sensitive to all three MDM2 inhibitors (Fig. 5A). more, MDS with mutant p53 responds less well to lenalidomide We tested the effect of combining lenalidomide with nutlin-3, and is more likely to progress to acute myeloid leukemia com- RG7388, and AMG232 and observed significantly improved pared with MDS with wild-type p53, suggesting that lenalido- cytotoxicity of all three MDM2 inhibitors when used with mide’s effects may be partially dependent on inactivating MDM4 lenalidomide (SI Appendix, Fig. S5 A–D). Synergy testing using (33). Our work provides the rationale for the combination of the Compusyn (SI Appendix, Fig. S6A) and Bliss (Fig. 5B) lenalidomide with MDM2 inhibitors in MCC and perhaps in methodologies revealed synergistic activity for the combination other solid tumors and hematologic malignancies containing of nutlin-3 or RG7388 with lenalidomide or UR99 (SI Appendix, wild-type p53. Fig. S6 B–F) (27–30). In contrast, thalidomide had no evidence for synergy when used in combination with nutlin-3, consistent Materials and Methods with its relatively reduced effects on CK1α levels (Fig. 5B). Gene Expression Analysis. The p53 targetome analysis is described in Allen To determine whether lenalidomide affected the ability of et al. (9). RNA-seq data reported in Cheng et al. (15) were reanalyzed by nutlin-3 to cause cell death by apoptosis in MCC cells, we used using the DESeq2 and EnhancedVolcano R packages (34, 35). dynamic BH3 profiling to measure changes in apoptotic prim- Plasmids. The GFP and T antigen cDNAs were Gateway (Invitrogen) cloned ing upon treatment (31). Addition of lenalidomide to nutlin-3 into pLIX 402 inducible empty or pLenti CMV Blast empty vector (w263- enhanced the priming effect for apoptosis (Fig. 5C). To deter- 1), gifts from David Root, Broad Institute of MIT and Harvard, Cam- mine the efficacy of dual inhibition of MDM2 and MDM4 in bridge, MA (Addgene 41394) and Eric Campeau, Zenith Epigenetics, Cal- vivo, we treated MKL-1 MCC xenografts with HDM201 (suitable gary, Canada (Addgene 17486), respectively. The sgRNA clones for CK1α for in vivo efficacy studies) with or without lenalidomide (Fig. 5D (BRDN0001149315, BRDN0001145680) were gifts from John Doench and and SI Appendix, Fig. S5F and Table S3). We found that the addi- David Root (Addgene plasmid 76188, 76189), both at Broad Institute of tion of lenalidomide greatly enhanced the efficacy of HDM201, MIT and Harvard, Cambridge, MA). pLKO-p53-shRNA-941 was a gift from suggesting a potential for clinical utility of the combinational Todd Waldman, Georgetown University School of Medicine, Washington, DC (Addgene 25637). therapy for p53 wild-type tumors expressing MDM2 and MDM4. We propose a model where MCV T antigens increase the depen- ACKNOWLEDGMENTS. We thank Amy E. Shade (Harvard University); dence on MDM2, MDM4, and CK1α to suppress p53 activity Anthony G. Letai [Dana-Farber Cancer Institute (DFCI)]; and Ensar Halilovic, E Jutta Beyer, Bing Yu, and Claire Fabre (Novartis) for reagents and advice; with therapeutic potential in virus-positive MCC (Fig. 5 ). Peter Howley (Harvard Medical School); Adam Bass and Thomas M. Roberts Lenalidomide has been used to treat the hematologic (DFCI); and Mitch Biermann (University of Wisconsin–Madison) for scientific malignancies, myelodysplastic syndrome (MDS), and multiple advice.

1. Becker JC, et al. (2017) Merkel cell carcinoma. Nat Rev Dis Primers 3:17077. 20. Houben R, et al. (2013) Mechanisms of p53 restriction in Merkel cell carcinoma cells 2. Paulson KG, et al. (2017) Merkel cell carcinoma: Current US incidence and projected are independent of the Merkel cell polyoma virus T antigens. J Invest Dermatol increases based on changing demographics. J Am Acad Dermatol 78:457–463.e2. 133:2453–2460. 3. Feng H, Shuda M, Chang Y, Moore PS (2008) Clonal integration of a polyomavirus in 21. Kronke¨ J, et al. (2015) Lenalidomide induces ubiquitination and degradation of human Merkel cell carcinoma. Science 319:1096–1100. CK1alpha in del(5q) MDS. Nature 523:183–188. 4. Rodig SJ, et al. (2012) Improved detection suggests all Merkel cell carcinomas harbor 22. Petzold G, Fischer ES, Thoma¨ NH (2016) Structural basis of lenalidomide-induced Merkel polyomavirus. J Clin Invest 122:4645–4653. CK1α degradation by the CRL4(CRBN) ubiquitin ligase. Nature 532:127–130. 5. Shuda M, Kwun HJ, Feng H, Chang Y, Moore PS (2011) Human Merkel cell poly- 23. Spiotto MT, et al. (2010) Imaging the unfolded protein response in primary tumors omavirus small T antigen is an oncoprotein targeting the 4E-BP1 translation regulator. reveals microenvironments with metabolic variations that predict tumor growth. J Clin Invest 121:3623–3634. Cancer Res 70:78–88. 6. Cheng J, Rozenblatt-Rosen O, Paulson KG, Nghiem P, DeCaprio JA (2013) Merkel cell 24. Doench JG, et al. (2016) Optimized sgRNA design to maximize activity and minimize polyomavirus large T antigen has growth-promoting and inhibitory activities. J Virol off-target effects of CRISPR-Cas9. Nat Biotechnol 34:184–191. 87:6118–6126. 25. Zhou B, et al. (2001) HER-2/neu induces p53 ubiquitination via Akt-mediated MDM2 7. Starrett GJ, et al. (2017) Merkel cell polyomavirus exhibits dominant control of the phosphorylation. Nat Cell Biol 3:973–982. tumor genome and transcriptome in virus-associated Merkel cell carcinoma. MBio 26. Scudiero D, et al. (1988) Evaluation of a soluble tetrazolium/formazan assay for cell 8:e02079-16. growth and drug sensitivity in culture using human and other tumor cell lines. Cancer 8. Harms PW, et al. (2015) The distinctive mutational spectra of polyomavirus-negative Res 48:4827–4833. Merkel cell carcinoma. Cancer Res 75:3720–3727. 27. Chou T, Motzer R, Tong Y, Bosl G (1994) Computerized quantitation of synergism and 9. Allen MA, et al. (2014) Global analysis of p53-regulated transcription identifies its antagonism of taxol, topotecan, and cisplatin against human teratocarcinoma cell direct targets and unexpected regulatory mechanisms. eLife 3:e02200. growth: A rational approach to clinical protocol design. J Natl Cancer Inst 86:1517– 10. Bates S, et al. (1998) p14ARF links the tumour suppressors RB and p53. Nature 1524. 395:124–125. 28. Ianevski A, He L, Aittokallio T, Tang J (2017) SynergyFinder: A web applica- 11. Nomura K, et al. (2017) Structural analysis of MDM2 RING separates degradation from tion for analyzing drug combination dose-response matrix data. Bioinformatics regulation of p53 transcription activity. Nat Struct Mol Biol 24:578–587. 33:2413–2415. 12. Bista M, Petrovich M, Fersht AR (2013) MDMX contains an autoinhibitory sequence 29. Ding Q, et al. (2013) Discovery of RG7388, a potent and selective p53-MDM2 inhibitor element. Proc Natl Acad Sci USA 110:17814–17819. in clinical development. J Med Chem 56:5979–5983. 13. Chen L, et al. (2015) Autoinhibition of MDMX by intramolecular p53 mimicry. Proc 30. Sun D, et al. (2014) Discovery of AMG 232, a potent, selective, and orally bioavailable Natl Acad Sci USA 112:4624–4629. MDM2-p53 inhibitor in clinical development. J Med Chem 57:1454–1472. 14. Beliveau A, Yaswen P (2007) Soothing the watchman: Telomerase reduces the p53- 31. Montero J, et al. (2015) Drug-induced death signaling strategy rapidly predicts cancer dependent cellular stress response. Cell Cycle 6:1284–1287. response to chemotherapy. Cell 160:977–989. 15. Cheng J, et al. (2017) Merkel cell polyomavirus recruits MYCL to the EP400 complex 32. Smith AE, et al. (2015) CSNK1A1 mutations and isolated del(5q) abnormality in to promote oncogenesis. PLoS Pathog 13:1–31. myelodysplastic syndrome: A retrospective mutational analysis. Lancet Haematol 16. Arora R, et al. (2012) Survivin is a therapeutic target in Merkel cell carcinoma. Sci 2:e212–e221. Transl Med 4:133ra56. 33. Saft L, et al. (2014) p53 protein expression independently predicts outcome in 17. Berrios C, et al. (2016) Merkel cell polyomavirus small T antigen promotes patients with lower-risk myelodysplastic syndromes with del(5q). Haematologica pro-glycolytic metabolic perturbations required for transformation. PLoS Pathog 99:1041–1049. 12:e1006020. 34. Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and 18. Lenos K, et al. (2012) Alternate splicing of the p53 inhibitor HDMX offers a superior dispersion for RNA-seq data with DESeq2. Genome Biol 15:550. prognostic biomarker than p53 mutation in human cancer. Cancer Res 72:4074–4084. 35. Blighe K (2018) Enhancedvolcano: Publication-ready volcano plots with enhanced 19. Sholl LM, et al. (2016) Institutional implementation of clinical tumor profiling on an colouring and labeling. Available at https://github.com/kevinblighe. Accessed Novem- unselected cancer population. JCI Insight 1:e87062. ber 23, 2018.

6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1818798116 Park et al. Downloaded by guest on September 30, 2021