The ubiquitin ligase RNF5 determines acute myeloid leukemia growth and susceptibility to histone deacetylase inhibitors

Khateb, Ali; Deshpande, Anagha; Feng, Yongmei; Finlay, Darren; Lee, Joo Sang; Lazar, Ikrame; Fabre, Bertrand; Li, Yan; Fujita, Yu; Zhang, Tongwu; Yin, Jun; Pass, Ian; Livneh, Ido; Jeremias, Irmela; Burian, Carol; Mason, James R.; Almog, Ronit; Horesh, Nurit; Ofran, Yishai; Brown, Kevin; Vuori, Kristiina; Jackson, Michael; Ruppin, Eytan; Deshpande, Aniruddha J.; Ronai, Ze’ev A.

doi:10.1038/s41467-021-25664-7

Download PDF

Article
Open access
Published: 13 September 2021

The ubiquitin ligase RNF5 determines acute myeloid leukemia growth and susceptibility to histone deacetylase inhibitors

Nature Communications volume 12, Article number: 5397 (2021) Cite this article

6890 Accesses
18 Citations
61 Altmetric
Metrics details

Subjects

Abstract

Acute myeloid leukemia (AML) remains incurable, largely due to its resistance to conventional treatments. Here, we find that increased abundance of the ubiquitin ligase RNF5 contributes to AML development and survival. High RNF5 expression in AML patient specimens correlates with poor prognosis. RNF5 inhibition decreases AML cell growth in culture, in patient-derived xenograft (PDX) samples and in vivo, and delays development of MLL-AF9–driven leukemogenesis in mice, prolonging their survival. RNF5 inhibition causes transcriptional changes that overlap with those seen upon histone deacetylase (HDAC)1 inhibition. RNF5 induces the formation of K29 ubiquitin chains on the histone-binding protein RBBP4, promoting its recruitment to and subsequent epigenetic regulation of genes involved in AML maintenance. Correspondingly, RNF5 or RBBP4 knockdown enhances AML cell sensitivity to HDAC inhibitors. Notably, low expression of both RNF5 and HDAC coincides with a favorable prognosis. Our studies identify an ERAD-independent role for RNF5, demonstrating that its control of RBBP4 constitutes an epigenetic pathway that drives AML, and highlight RNF5/RBBP4 as markers useful to stratify patients for treatment with HDAC inhibitors.

ZMYND11-MBTD1 induces leukemogenesis through hijacking NuA4/TIP60 acetyltransferase complex and a PWWP-mediated chromatin association mechanism

Article Open access 16 February 2021

KAT7 is a genetic vulnerability of acute myeloid leukemias driven by MLL rearrangements

Article 06 August 2020

Inhibition of ubiquitin-specific protease 7 sensitizes acute myeloid leukemia to chemotherapy

Article 23 May 2020

Introduction

AML is a heterogeneous hematological cancer characterized by the accumulation of somatic mutations in immature myeloid progenitor cells. Such mutations alter the self-renewal, proliferation, and differentiation capabilities of these progenitor cells^1,2. The prognosis of AML patients is strongly influenced by the type of chromosomal or genetic alterations and by changes in gene expression^1,3. Although numerous mutations and chromosomal aberrations that drive AML development have been identified^1,4, the molecular components and epigenetic modulators that contribute to AML etiology and pathophysiology are not well defined. Approximately one-third of AML patients fail to achieve complete remission in response to chemotherapy, and 40–70% of those who do enter remission relapse within 5 years. Thus, there is an urgent need to better understand the molecular mechanisms underlying AML development and progression to facilitate the development of more effective therapies.

RING finger protein 5 (RNF5) is an ER-associated E3 ubiquitin ligase and component of the UBC6e-p97 complex, which is implicated in ER-associated degradation (ERAD)^5,6, a pathway involved in maintaining protein homeostasis. RNF5 recognizes misfolded proteins and promotes their ubiquitination and proteasome-dependent degradation^5,6. RNF5 expression is relatively increased in several cancers, including breast cancer, hepatocellular carcinoma, and AML⁷. RNF5 regulates glutamine metabolism by promoting degradation of misfolded glutamine carrier proteins, a function important in the cancer cell response to ER stress-inducing chemotherapies such as paclitaxel⁸. RNF5 promotes degradation of the protease ATG4B, which limits basal amounts of autophagy⁹. RNF5 also limits intestinal inflammation by controlling S100A8 protein stability¹⁰. Given the important pathophysiological roles of RNF5, the observation that it is upregulated in AML cells and patient samples prompted us to investigate the possible contribution of RNF5 to the development and progression of this disease.

Histone modification by acetylation contributes to the dynamic regulation of chromatin structure and affects gene expression programs. Histone acetylation status is associated with transcriptional regulation of leukemic fusion proteins, such as AML1-ETO, PML-RARα, and MLL-CBP^11,12. Correspondingly, histone deacetylases (HDACs) are implicated in the etiology and progression of leukemia¹³, and HDAC inhibitors can block growth or promote differentiation or apoptosis of leukemia cells¹⁴. The retinoblastoma binding protein 4 (RBBP4) is a component of multi-protein complexes that function in nucleosome assembly and histone modifications, modulate gene transcription, and regulate the cell cycle and proliferation¹⁵. Such complexes include the nucleosome remodeling and deacetylase (NuRD) complex, polycomb repressor complex 2 (PRC2), and switch independent 3 A (SIN3A)^15,16. Overexpression of RBBP4 or HDAC1 correlates with clinicopathologic characteristics and poor prognosis in breast cancer¹⁷, and RBBP4 expression positively correlates with hepatic metastasis and poor prognosis in colon cancer patients¹⁸. RBBP4 is also implicated in the regulation of DNA repair genes, and its suppression in glioblastoma enhances tumor sensitivity to temozolomide chemotherapy¹⁹. However, the function of RBBP4 in AML has not been studied.

Here, we identified a central role for the RNF5-RBBP4 axis in AML maintenance and responsiveness to HDAC inhibitors. Our data suggest that targeting RNF5 and HDAC pathways represents a therapeutic modality for AML and that RNF5 or RBBP4 abundance could serve as a prognostic marker and means to stratify patients for treatment with HDAC inhibitors.

Results

Increased abundance of RNF5 in AML patient samples correlates with poor prognosis

Analysis of RNA-seq datasets for various cancer cells in the Cancer Cell Line Encyclopedia (CCLE) database identified a higher copy number and levels of RNF5 transcripts in AML, chronic myeloid leukemia (CML), and T-cell acute lymphoblastic leukemia (T-ALL) relative to other tumor types (Fig. 1a and Supplementary Fig. 1a). Higher levels of RNF5 protein were confirmed in AML and CML cell lines compared with other cancer lines (Supplementary Fig. 1b). To assess the clinical relevance of RNF5 expression in AML, we analyzed levels of RNF5 mRNA and protein in peripheral blood mononuclear cells (PBMCs) from independent cohorts of AML patients. Similar to results in AML lines, the average abundance of RNF5 protein was significantly higher in PBMCs from AML patients relative to control samples (CD34⁺ and PBMCs) (Fig. 1b, c, Supplementary Fig. 1c, d). The patient cohort included an equal number of females (24; median age = 59) and males (24; median age = 64.5; Supplementary Table 1). Given that RNF5 is a ubiquitin ligase, its transcript levels are not as reflective of activity as are protein levels, as self-degradation or other post-translational modifications can alter RNF5 subcellular localization, availability and/or activity. Indeed, analysis of patient cohorts revealed a significant increase in RNF5 protein but not transcript levels in patients as compared to healthy subjects (Fig. 1b, c and Supplementary Fig. 1e). Stratification of the 50 patients into two groups based on the top high (n = 8, 15%) and low (n = 42, 85%) RNF5 protein levels revealed that high abundance coincided with poor overall survival (P = 0.05, Fig. 1d). Notably, this difference was not due to blast counts, as they did not differ significantly among patients showing high or low RNF5 levels (Supplementary Fig. 1f, g). Independent analysis of AML patients (n = 154) based on The Cancer Genome Atlas (TCGA) dataset confirmed a significant positive correlation between high RNF5 expression (10%) and poor survival (P = 0.009, Fig. 1e). Notably, there was no pattern of RNF5 abundance that either positively or negatively correlated with the presence of FLT3 or NPM1 mutations (Supplementary Fig. 1h, i), suggesting that the significance of RNF5 activity to AML may not depend on any specific oncogenic driver(s) or activation of particular signaling pathways.

**Fig. 1: RNF5 expression in AML cell lines and patient samples.**

Assessment of an independent AML patient cohort (from the Rambam Health Campus Center, Haifa, Israel which included multiple samples obtained from 5 females with a median age of 62 and 6 males with a median age of 63.5, as detailed in Supplementary Table 1) confirmed higher levels of RNF5 protein in AML patient blood samples (n = 18) relative to samples taken from healthy donors (n = 5) (Fig. 1f, g). Because this cohort included samples taken from patients both prior to and following therapy, we were able to compare RNF5 abundance before and after therapy and at remission or relapse stages. Notably, RNF5 abundance markedly decreased following chemotherapy and during remission (n = 8) (Fig. 1h, i and Supplementary Fig. 1j). Conversely, RNF5 levels at diagnosis were similar to those seen in patients that either relapsed or were refractory to treatment (n = 5) (Fig. 1h, j). These results suggest that RNF5 levels in AML blasts may serve as a prognostic marker for AML.

RNF5 is required for AML cell proliferation and survival

Next, we asked how RNF5 knockdown (RNF5-KD) would impact leukemia cell growth in vitro. Interestingly, KD using RNF5-targeting short hairpin RNAs (shRNF5) decreased viability and attenuated growth of MOLM-13 and U937 AML lines (Fig. 2a and Supplementary Fig. 2a) but not of CML (K-562) or T-ALL (Jurkat) lines (Supplementary Fig. 2b, c). RNF5 KD in MOLM-13 or U937 AML cells also promoted accumulation of cells in the G1 phase of the cell cycle (Fig. 2b), an effect accompanied by increase in levels of the cell cycle regulatory proteins p27 and p21 (Fig. 2c). Moreover, AML cells MOLM-13 and U937 with RNF5-KD showed reduced colony formation in soft agar relative to controls (Fig. 2d) and increased abundance of proteins associated with apoptosis, as reflected by levels of cleaved forms of caspase-3 (Fig. 2e) and poly ADP ribose polymerase (PARP) (Supplementary Fig. 2d). Effects of RNF5-KD on U937 and MOLM-13 cells were confirmed in two additional AML cell lines HL-60 and THP-1 (Supplementary Fig. 2e–h). Importantly, re-expression of RNF5 WT, but not the catalytically inactive RING mutant (RNF5 RM), restored cell proliferation (Fig. 2f, g), confirming the specificity of these phenotypes and suggesting that RNF5 catalytic activity is required for AML cell proliferation.

To verify changes seen upon KD, we used CRISPR-Cas9 gene-editing technology to deplete RNF5 in MOLM-13 cells stably expressing Cas9 using RNF5-targeting guide RNAs (sgRNAs). Relative to control cells transduced with Renilla luciferase-targeting sgRNAs, cells transduced with RNF5-targeting sgRNAs showed impaired growth based on CellTiter-Glo luminescence assay (Fig. 2h, i).

To further assess RNF5 function in AML, we monitored the viability of xenografted patient-derived AML cells (PDX, AML-669²⁰) transduced with shRNF5 or control constructs. As expected, using two independent shRNF5s (albeit limited KD efficiency), we observed decreased viability of xenografted RNF5-KD cells relative to controls (Fig. 2j–i). These findings confirm our observations in AML cell lines and support the notion that RNF5 downregulation impairs the proliferation of AML blasts.

RNF5 inhibition enhances ER stress-induced apoptosis of AML cells

As RNF5 functions as part of ERAD and the ER stress response, we asked whether changing RNF5 abundance would alter the ER stress response in AML cells. To do so, we exposed RNF5-KD or control MOLM-13 cells to thapsigargin or tunicamycin to inhibit the ER Ca²⁺-ATPase (SERCA) or protein glycosylation, respectively²¹, as a mean to induce ER stress. Thapsigargin treatment of MOLM-13 RNF5-KD cells increased apoptotic markers to levels higher than those seen in control cells (Fig. 3a, b). Thapsigargin treatment also decreased the viability of MOLM-13 RNF5-KD cells to a greater extent than seen in control MOLM-13 WT RNF5 cells (Fig. 3c). Tunicamycin treatment also decreased the viability of RNF5-KD HL-60 cells compared to tunicamycin-treated control HL-60 cells (Supplementary Fig. 3a). Consistent with a function in ER stress, RNF5-KD increased levels of transcripts encoding key UPR components, including CHOP, ATF3, and sXBP1, in thapsigargin-treated MOLM-13 (Fig. 3d) and HL-60 cells (Supplementary Fig. 3b), relative to mock-transduced controls.

**Fig. 3: RNF5 inhibition sensitizes AML cells to ER stress-induced apoptosis.**

Given the link between ER stress and proteasomal degradation, we assessed potential synergy between RNF5 KD and proteasomal inhibition. Indeed, RNF5-KD MOLM-13 cells treated with the proteasome inhibitor bortezomib (BTZ) showed increased levels of apoptotic markers such as cleaved forms of caspase-3 and PARP (Fig. 3e) and decreased viability (Fig. 3f) relative to control-treated cells. Using annexin V and propidium iodide staining, which monitor the degree of programmed cell death, we showed that RNF5-KD also enhanced apoptosis of BTZ-treated HL-60 cells (Fig. 3f), decreasing the BTZ IC₅₀ from 9.6 nM in controls to 5.4 nM in RNF5-KD cells (Fig. 3g). These data suggest that RNF5 plays a role in the response of AML cells to proteotoxic stress.

RNF5 loss delays leukemia establishment and progression

We next asked whether RNF5 activity modulates leukemia growth in vivo. To do so, we used a human AML xenograft model in which luciferase-expressing U937 cells (U937-pGFL) were transduced with doxycycline-inducible shRNF5 or control shRNA before being injected intravenously into NOD/SCID mice (Supplementary Fig. 4a). Following leukemia establishment, as confirmed by bioluminescence, mice were fed a doxycycline-containing diet and monitored for disease progression and overall survival. Interestingly, animals injected with RNF5-KD cells exhibited a markedly decreased leukemia burden and prolonged survival relative to control mice (Fig. 4a, b and Supplementary Fig. 4b). RT-PCR analysis of splenocytes isolated from mice transplanted with RNF5- KD cells confirmed expression of shRNF5 (Supplementary Fig. 4c). Western blot analysis of splenocyte lysates revealed more abundant expression of the cell cycle regulatory protein p27 in shRNF5 relative to control cells (Supplementary Fig. 4d), consistent with our in vitro data and with the delayed leukemia progression observed following RNF5 KD (Fig. 4a, b). Collectively, these data indicate that RNF5 is required for AML cell proliferation in vivo.

**Fig. 4: RNF5 suppression impairs leukemia establishment and progression in vivo.**

We also asked whether RNF5 functions in AML initiation using the MLL-AF9 model²² for in vitro and in vivo studies. For in vitro analysis, we purified hematopoietic stem and progenitor (Lin-depleted) cells (HSPCs) from the bone marrow of Rnf5^−/−, which exhibit normal development and hematopoiesis²³, and wild-type (WT) C57/BL6 mice. HSPCs from these mice were retrovirally transduced with a bicistronic construct harboring MLL-AF9 linked to a green fluorescent protein (GFP) marker. When we assessed colony-forming capacity (CFC), we found that compared to WT GFP-MLL-AF9 cells, Rnf5^−/− GFP-MLL-AF9 cells exhibited markedly reduced CFC in methylcellulose after 7, 14, and 21 days in culture and observed a striking reduction in the number of blast-like colonies (Fig. 4d, e). These phenotypes are consistent with apparent terminal differentiation of Rnf5^−/− cells, as reflected by a greater cytoplasm/nucleus ratio and more vacuolated cytoplasm (Fig. 4e, f).

To assess leukemogenesis in vivo, we injected sublethally-irradiated WT C57/BL6 recipient mice with GFP-MLL-AF9–transduced Rnf5^WT or Rnf5^−/− cells and monitored cell engraftment by flow cytometry for GFP-positive (GFP+) cells in peripheral blood (Fig. 4c). Analysis on days 15 and 28 post-injection identified fewer GFP+ cells in mice injected with GFP-MLL-AF9 Rnf5^−/− cells than in mice injected with GFP-MLL-AF9 Rnf5^WT cells, indicating a delay in leukemia development (Fig. 4g). Moreover, mice harboring GFP-MLL-AF9 Rnf5^–/– cells exhibited prolonged survival relative to mice injected with GFP-MLL-AF9 Rnf5^WT cells (Fig. 4h). Collectively, these data show that RNF5 loss decreases the colony-forming capacity of MLL-AF9–transformed pre-leukemic cells in vitro and delays leukemia progression in vivo.

RNF5 activity modulates transcription in AML cells

To identify pathways modulated by RNF5 activity in AML cells, we first monitored transcriptional changes in MOLM-13, U937, and HL-60 AML lines expressing either RNF5-KD or control constructs. RNA sequencing (RNA-seq) analysis identified a total of 237, 814, and 1380 dysregulated genes in MOLM-13, U937, and HL-60, respectively, following RNF5 KD relative to control (RNF5-WT) cells (Fig. 5a, b and Supplementary Data file 1). Ingenuity Pathway Analysis identified selective enrichment of genes implicated in myeloid cell function such as NF-κB signaling, IL-8 signaling, reactive oxygen species, and several pathways related to cell migration such as Rho GTPase and Tec kinase signaling (Supplementary Fig. 5a). Expression of a total of 59 genes (35 up- and 24 downregulated) were significantly altered by RN5F KD in all three AML lines (Fig. 5a, b). Among upregulated genes were CDKN1A and CDKN2D, which encode cell cycle inhibitors; LIMK1, which encodes a kinase functioning in regulation of the actin cytoskeleton; ANXA1, which encodes a calcium-binding protein functioning in metabolism, EGFR signaling, and cell death programs; and NCF1, which encodes a subunit of NADPH oxidase (Fig. 5c and Supplementary Fig. 5b). Downregulated genes included antiapoptotic BCL2A1, and SAP18, which encodes a histone deacetylase complex subunit functioning in transcriptional repression (Supplementary Fig. 5b). Moreover, such changes were consistent with phenotypic changes seen in RNF5-KD AML cell lines, such as reduced proliferation and increased apoptosis. Interestingly, analysis of the Library of Integrated Network-Based Cellular Signatures (LINCS) drug screening database identified a notable overlap between transcriptomic changes induced by the HDAC1 inhibitor mocetinostat in various cancer cells and those seen in shRNF5-expressing MOLM-13 and HL-60 cells (Fig. 5d). Five and one out of the top ten transcriptional changes identified in LINCS following HDAC inhibition overlapped with those seen following RNF5 KD in MOLM-13 and HL-60 cells, respectively (Fig. 5d, Supplementary Fig. 5c and Supplementary Data file 2). Among commonly affected pathways were activation of GP6 and Rho GTPase signaling and repression of the nucleotide excision repair (NER) pathway (Fig. 5e). These observations suggest that RNF5 may regulate HDAC activity in AML cells.

**Fig. 5: Transcriptional and survival analysis of RNF5- or RBBP4-deficient AML cells.**

RNF5 interacts with and ubiquitinates the retinoblastoma binding protein 4

We hypothesized that RNF5 elicits transcriptional changes through intermediate regulatory component(s). Thus, to identify RNF5-interacting proteins or substrates we performed liquid chromatography-tandem mass spectrometry (LC-MS/MS) and compared proteins immunoprecipitated from lysates of MOLM-13 cells expressing inducible Flag-tagged RNF5 versus those expressing empty vector. Among 65 RNF5-interacting proteins identified were previously reported substrates, such as 26 S proteasome components, VCP and S100A8^5,10, as well as proteins implicated in AML development, such as DHX15²⁴ and gelsolin^25,26 (Supplementary Data file 3). Among the more abundant RNF5-bound proteins were components of ERAD, translation initiation, proteolytic and mRNA catabolic processes (Fig. 5f, Supplementary Fig. 5d, e and Supplementary Data file 3).

Although none of the interacting proteins identified here were transcription factors, epigenetic modifications initiated by changes in RNF5 expression could also underlie changes in gene expression. In fact, we identified one RNF5-interacting protein as the epigenetic regulator histone-binding protein RBBP4 (Fig. 5f and Supplementary Data file 3). Analysis of transcriptome data from TCGA revealed an inverse correlation of RBBP4 expression with expression of genes upregulated in RNF5-KD cells (Fig. 5g and Supplementary Fig. 5f), suggesting that RNF5 positively controls RBBP4 transcriptional regulatory function. RBBP4 is a component of several chromatin assembly, remodeling, and nucleosome modification complexes, including PRC2²⁷ and the NuRD co-repressor complex, which contains HDAC1 and HDAC2²⁸. Indeed, the inverse correlation between RBBP4 expression and RNF5-upregulated genes was mirrored when we analyzed HDAC1, HDAC2, and EZH2 expression relative to RNF5-upregulated genes (Supplementary Fig. 5g–i). Increased RBBP4 expression was also correlated with malignant phenotypes of several human tumors including AML²⁹. Analysis of tumor data in TCGA revealed high RBBP4 expression in AML, compared with other tumor types (Supplementary Fig. 5j). Assessment of an AML patient cohort confirmed higher RBBP4 expression in samples from AML patients compared to healthy donors (Fig. 5h and Supplementary Fig. 5k). Stratification of AML patients based on RBBP4 expression indicated that high expression (the top 30%) correlated with poor overall survival (Fig. 5i).

If RNF5 positively regulates RBBP4, RBBP4 KD should promote phenotypic changes in AML cells similar to RNF5 KD. Indeed, shRNA-based RBBP4 KD in MOLM-13 and U937 cells impaired their growth (Fig. 5j and Supplementary Fig. 5l), promoted PARP cleavage indicative of apoptosis (Fig. 5k and Supplementary Fig. 5m), and induced expression of genes that were also induced by RNF5-KD (Fig. 5l). Furthermore, when we examined RBBP4 function in vivo using the U937 xenograft model, mice harboring RBBP4 KD in xenografted cells showed delayed AML development and prolonged survival relative to WT-RBBP4 controls, phenocopying changes seen upon RNF5 KD (Fig. 5m, n and Supplementary Fig. 5n, o). Notably, western blot analysis of splenocytes from mice transplanted with RBBP4 KD cells showed that these cells retained RBBP4 expression (Supplementary Fig. 5p). Since RBBP4 KD was confirmed prior to injection of these cells, it is likely that these cells emerged as escapers during in vivo selection (Supplementary Fig. 5n, o). The latter explains the shorter survival observed in mice that harbored the escapers, compared with mice that retained RBBP4 KD (Fig. 5n). Interestingly, similar to outcomes seen in RNF5-KD AML cells, RBBP4 KD blocked the growth of AML, but not CML and T-ALL, cell lines (Supplementary Fig. 6a, b), confirming a link between RNF5 and RBBP4 in the context of AML.

RNF5 is a transmembrane protein primarily associated with the ER, and its ubiquitin ligase domain is located in the cytosol^6,30. We assessed potential interaction between RNF5 and RBBP4 in the HEK293T line by coimmunoprecipitation of ectopically expressed WT RNF5, a catalytically inactive RING mutant (RNF5 RM), or a C-terminal transmembrane domain deletion mutant (RNF5 ∆CT) (Fig. 6a). Endogenous RBBP4 coimmunoprecipitated with all RNF5 constructs, suggesting that both the RING and transmembrane domains are dispensable for protein-protein interaction (Fig. 6b). Reciprocal IP using RBBP4 as bait confirmed interaction with RNF5 (Fig. 6c). Interaction between endogenous RBBP4 and ectopically expressed RNF5 was also confirmed in MOLM-13 cells (Fig. 6d). Next, we assessed potential effect of RNF5 on RBBP4 ubiquitination. Co-expression of HA-tagged ubiquitin, Myc-tagged RBBP4, and Flag-tagged RNF5 constructs in HEK293T cells revealed that RBBP4 was ubiquitinated by WT RNF5, but not by RNF5 RM or RNF5 ΔCT (Fig. 6e), indicating that ubiquitin ligase activity (RING domain-dependent) and membrane association are both required for an RNF5-mediated increase in RBBP4 ubiquitination. Correspondingly, RNF5-KD in HEK293T or MOLM-13 cells decreased RBBP4 ubiquitination relative to controls (Supplementary Fig. 6c, d).

**Fig. 6: RBBP4 is ubiquitinated by RNF5 and regulates RNF5 target genes.**

Notably, neither RNF5 overexpression nor RNF5 KD altered the abundance of RBBP4 protein, suggesting that RBBP4 ubiquitination by RNF5 does not occur via the formation of proteasome-targeting K48 ubiquitin chains and does not alter RBBP4 stability (Fig. 6f, g and Supplementary Fig. 6e). Immunoprecipitation of RBBP4 and immunoblot using an antibody specific for the K63 chain topology revealed no notable differences in cells overexpressing any form of RNF5, suggesting that RNF5 does not induce K63 ubiquitination of RBBP4 (Supplementary Fig. 6f). To assess the linkage-specific polyubiquitin induced by RNF5 on RBBP4, we used mutant HA-ubiquitin constructs with only one lysine available for linkage (K-only mutants: K29, K11, K27, K6, and K33) in which all lysine residues except that indicated are mutated to arginine allowing a single type homotypic chain. We then monitored changes in Myc-tagged RBBP4 ubiquitination in cells overexpressing Flag-tagged RNF5. Poly-ubiquitination of RBBP4 was enhanced by RNF5 only in the presence of K29 ubiquitin (Fig. 6h), strongly suggesting that RNF5 induces K29-topology polyubiquitination of RBBP4.

RNF5 promotes recruitment of RBBP4 to gene promoters

Because RNF5 activity does not alter RBBP4 stability, we asked whether RNF5 affects RBBP4 localization or interactions with other proteins. Subcellular fractionation in MOLM-13 cells and immunofluorescent analyses of nuclear and chromatin-bound RBBP4 did not identify changes in RBBP4 localization following RNF5 KD (Supplementary Fig. 6g and h). Next, since RBBP4 is a component of PRC2 and complexes containing HDACs^31,32, we asked whether RNF5 activity alters the formation of these complexes or their recruitment to target gene promoters. Neither overexpression nor KD of RNF5 affected RBBP4 interaction with HDAC1, HDAC2, or EZH2 (Fig. 6i, j and Supplementary Fig. 6i), suggesting that RBBP4 ubiquitination by RNF5 is not required for assembly of RBBP4-containing these complexes.

We then used chromatin immunoprecipitation (ChIP) and quantitative PCR (qPCR) to investigate RBBP4 recruitment to promoters of genes regulated by either RNF5 or RBBP4. RNF5 KD decreased RBBP4 recruitment to ANXA1, NCF1, and CDKN1A promotors (Fig. 6k). Examination of histone modifications at promoters of these genes identified that RNF5 KD increased H3K9 and H3K27 acetylation (Fig. 6l, m) and reduced H3K27 methylation (Fig. 6n), changes indicative of increased gene expression^33,34. These changes are consistent with increased expression of these genes following RNF5 KD (Fig. 5c) and suggest that RNF5 control of gene expression in AML cells is mediated by RBBP4.

RNF5 inhibition sensitizes AML cells to HDAC inhibitors

As an independent support for the function of the RNF5-RBBP4 regulatory axis in promoting AML cell growth, we screened for synergistic interactions between RNF5 and epigenetic modulators. To do so, we assessed the effect of 261 epigenetic inhibitors at two concentrations (Supplementary Table 2) on the growth of U937 cells that stably express inducible shRNF5 (Fig. 7a, b). Of epigenetic inhibitors tested, 49 decreased viability of shRNF5-expressing cells relative to control WT RNF5 AML cells (Fig. 7b). Among the 49 inhibitors were several hypomethylation agents, including several histone methyltransferases (such as G9a), histone demethylases (such as Jumonji histone demethylases), and HDAC inhibitors (such as TMP269, pimelic diphenylamide 106, and N-acetyldinaline [CI-994]). Because RBBP4 is a key component of the HDAC complex and given that RNF5 KD induces transcriptional changes comparable to HDAC1 inhibition (Fig. 5d, e), we assessed possible synergy between RNF5 inhibition and HDAC inhibitors. To do so, we selected the HDAC inhibitor CI-994, which is in clinical trials against several cancers (https://www.drugbank.ca/drugs/DB12291), for additional validation. Indeed, U937 and HL-60 cells subjected to RNF5-KD were more sensitive to CI-994 relative to control cells (Fig. 7c and Supplementary Fig. 7a), suggesting that RNF5 KD sensitizes AML cells to HDAC inhibition.

We note that the HDAC inhibitor romidepsin (also known as FK228) did not score positively in our screen. This was likely due to the relatively high concentrations tested, which were lethal to both RNF5-KD and control U937 cells. FK228 has been approved by the Food and Drug Administration (FDA) to treat peripheral T-cell lymphoma³⁵ and has been investigated in preclinical studies as a potential treatment for AML^25,36. Therefore, we re-assessed FK228 at non-lethal concentrations (up to 6 nM for 24 h) using multiple AML cell lines. Notably, when combined with RNF5-KD, FK228 had an additive effect in decreasing cell viability (Fig. 7d, e and Supplementary Fig. 7b–d) and inducing apoptosis (Fig. 7f). We also confirmed the additive effect of RNF5 loss plus treatment with HDACi in MOLM-13 cells (made RNF5 deficient using the CRISPR/Cas9 system; Supplementary Fig. 7e). The additive effect on AML cell death of RNF5-KD plus FK228 treatment was lost upon RNF5 re-expression (Fig. 7g), confirming a specific role for RNF5 in sensitizing AML cells to HDAC inhibition.

Of AML cell lines tested, the MV-4-11 line showed very low levels of RNF5 protein (Supplementary Fig. 1b). MV-4-11 cells were also most sensitive to FK228 treatment (Supplementary Fig. 7f), and RNF5 KD did not increase their sensitivity (Supplementary Fig. 7g). These observations further support the importance of RNF5 abundance in the response of AML cells to HDAC inhibitors. Moreover, RBBP4 KD sensitized AML cells to FK228 treatment (Fig. 7h and Supplementary Fig. 7h), consistent with our findings that RNF5 positively regulates RBBP4. Notably, elevated H3K9 acetylation at the promotors of RNF5- or RBBP4-regulated genes, such as ANXA1 and CDKN1A, was seen following FK228 treatment and further increased upon RNF5 KD (Fig. 7i). The latter finding is consistent with increased ANXA1 and CDKN1A expression seen after treatment with FK228 alone or in combination with RNF5 KD (Fig. 7j and Supplementary Fig. 7i). Notably, RNF5 KD in the K562 (CML) line did not sensitize cells to FK228, and RNF5 KD in Jurkat cells (T-ALL) only slightly enhanced cell sensitivity to FK228 (Supplementary Fig. 7j and k).

Next, to corroborate these findings in primary AML blasts, we performed an ex-vivo analysis of AML patients’ samples (n = 4) and assessed their sensitivity to FK228 treatment. These samples were selected based on RNF5 and RBBP4 protein levels (2 high, 2 low). Interestingly, and similar to phenotypes seen in our KD experiments, samples expressing low RNF5 and RBBP4 were more sensitive (AML-075B logIC₅₀ = −10.7 M, AML-034 logIC₅₀ =−10.4 M) to FK228 treatment, compared with the high-expressing group (AML-013 logIC₅₀ = −9.9 M, AML-072B logIC₅₀ = −9.6 M) (Fig. 7k–m). Finally, to assess the relevance of combined RNF5 loss and HDAC treatment in patients, we analyzed data from a bioinformatic pipeline that identifies clinically relevant synthetic lethal interactions³⁷. This analysis revealed a more favorable prognosis in patients with concomitant downregulation of both HDAC and RNF5 (Fig. 7n), substantiating the sensitization to HDAC in RNF5 low expressing specimens. Collectively, these findings suggest that RNF5 signaling is a critical determinant of AML cell sensitivity to HDAC inhibitors.

Discussion

High mortality seen in patients with AML predominantly results from failure to achieve complete remission following chemotherapy, coupled with a high relapse rate. Here we identify an important role for the ubiquitin ligase RNF5 in AML and demonstrate how RNF5 contributes to this form of leukemia. Our studies establish a function for RNF5 beyond its previously characterized activity in ERAD and proteostasis^6,30 and reveal mechanisms underlying its regulation of gene expression programs governing AML development and response to HDAC inhibitors. The clinical relevance of RNF5 and RBBP4 to AML is supported by our studies of patient samples and genetic mouse models. In mice, RNF5 or RBBP4 depletion inhibited AML progression and prolonged mouse survival (Fig. 4). In human, analysis of AML samples from two independent clinical cohorts revealed that a high abundance of RNF5 protein, which is commonly seen in AML patient samples, correlates with poor prognosis. Those phenotypes are mediated via RNF5 interaction with the chromatin remodeling protein RBBP4, which results in its non-canonical (K29 topology) ubiquitination that promotes RBBP4 recruitment to specific gene promoters (among them, ANXA1, NCF1, and CDKN1A), and a concomitant regulation of genes implicated in AML maintenance. Our finding that RNF5 modifies RBBP4 in a way that alters the expression of AML-related genes is confirmed by our ChIP analysis showing that RNF5 promotes recruitment of RBBP4 to gene promoters. Moreover, an inverse correlation between RBBP4 expression and the expression of genes that were upregulated in RNF5 KD cells was also found in AML samples from the TCGA database (Fig. 5). Future genome-wide assessment of promoter-bound RBBP4 will likely identify additional genes whose transcription is regulated by RNF5-modified RBBP4.

Independent support for a function for RNF5 in recruiting RBBP4 to transcriptional regulatory complexes comes from our finding that RNF5/RBBP4 abundance governs the sensitivity of AML cells to HDAC inhibitors. Correspondingly, transcriptional changes induced by RNF5 KD overlapped with those seen following treatment with HDAC1 inhibitors. Furthermore, AML primary blasts expressing low RNF5/RBBP4 levels were more sensitive to FK228 compared to high-expressing blasts. Along these lines, synthetic lethal analysis identified a favorable prognosis in a cohort of AML patients with low expression of both HDAC and RNF5 (Fig. 7). These findings suggest that RNF5 or RBBP4 abundance may serve as useful markers for the stratification of AML patients for treatment with HDAC inhibitors.

Notably, although RNF5 is expressed at high levels in AML, CML, and T-ALL cell lines²⁰ it is critical for cell survival only in AML cells. In fact, the CCLE database reveals that CML and T-ALL lines express on average higher levels of RNF5 than do AML lines. Nonetheless, K-562 (CML) and Jurkat (T-ALL) lines subjected to RNF5 KD do not exhibit growth inhibition or undergo cell death, while similarly treated AML lines do. Likewise, inhibition of RBBP4 does not impact CML or T-ALL cell but rather inhibits AML cell growth in a manner similar to that seen after RNF5 inhibition. Along these lines, RNF5 regulation and function are likely to be cell type and tissue-dependent. Our earlier studies in breast cancer and melanoma further support such context-dependent functions: RNF5 promotes melanoma growth via changes in immune and intestinal epithelial cells, while inhibits breast cancer growth through the tumor-intrinsic expression of glutamine carrier proteins^7,8,38.

It is important to note that relatively high RNF5 expression in AML cell lines is likely due to a high copy number, as shown by analysis of copy number alterations in various cancer cells²⁰). Analysis of the TCGA database reveals increases in RNF5 mRNA levels in 3% of patients. Our patient cohorts revealed a significant increase in RNF5 abundance but not transcription levels (Fig. 1b, c and Supplementary Fig. 1e). We therefore believe that RNF5 overexpression (at the protein level) is attributable to relative increases in RNF5 protein stability, as supported by our assessments of two independent AML cohorts (Fig. 1). As a ubiquitin ligase, RNF5 activity is regulated primarily by post translational modifications affecting its stability, rather than transcription. However, that these increases may be linked to a pre-existing mutation that could increase RNF5 abundance in AML patient PMBCs or to micro-vesicle-based cell-cell communication cannot be ruled out.

Given that RNF5 protein is ER-anchored, its interaction with a chromatin regulatory protein such as RBBP4 is unanticipated. Among that possibilities that may explain such interaction are: (i) The interaction may occur as the RBBP4 gene is translated, prior to nuclear translocation, a mechanism reported for other ubiquitin ligases³⁹, (ii) An as yet unidentified post-translational modification may promote nuclear localization of RNF5. For example, the possibility that RNF5 undergoes sumoylation should be considered given the high probability predicted using GPS-SUMO tool⁴⁰; Finally, (iii) RBBP4/RNF5 interaction may occur at specific phases of the cell cycle, for example, at the entry to mitosis, when the nuclear envelop breaks down and nuclear contents are released to the cytoplasm. Future studies are needed to further examine these possibilities.

In summary, using genetic mouse models and clinical data, our findings establish a central role for the RNF5-RBBP4 axis in AML maintenance and responsiveness to HDAC inhibitors. Our identification of crosstalk between ubiquitination and epigenetic regulation offers a new paradigm for ERAD-independent RNF5 function in controlling RBBP4 activity and subsequent transcriptional networks implicated in AML. Our findings also demonstrate the ability of HDAC inhibitors to treat AML, particularly AML expressing low levels of RNF5, and provides a method to stratify AML patients for treatment with HDAC inhibitors.

Methods

Animal studies

All animal experiments were approved by the Sanford Burnham Prebys Medical Discovery Institute’s Institutional Animal Care and Use Committee (approval AUF 16-028). Animal care followed institutional guidelines. Rnf5^−/− mice were generated on a C57BL/6 background as described²³. C57BL/6 WT mice were obtained by crossing Rnf5^−/− mice. Female mice were maintained under controlled temperature (22.5 °C) and illumination (12 h dark/light cycle) conditions and were used in experiments at 6–10 weeks of age.

The xenograft model was established using U937 cells expressing the p-GreenFire1 Lenti-Reporter Vector (pGFL). NOD/SCID (NOD.CB17-Prkdcscid/J) mice were obtained from the SBP Animal Facility. Mice were irradiated (2.5 Gy), and U937-pGFL cells (2 × 10⁴ per mouse) were injected intravenously. Leukemia burden was serially assessed using noninvasive bioluminescence imaging by injecting mice intraperitoneally (i.p.) with 150 mg/kg D-Luciferin (PerkinElmer, 122799) in phosphate-buffered saline (PBS, pH 7.4), anesthetizing them with 2–3% isoflurane, and imaging them on an IVIS Spectrum (PerkinElmer). For in vivo RNF5 KD experiments, at disease onset (day 15, as measured by bioluminescent imaging), mice were fed rodent chow containing 200 mg/kg doxycycline (Dox diet, Bio-Serv) to induce RNF5-KD. Mice were sacrificed upon signs of morbidity resulting from leukemic engraftment (>10% weight loss, lethargy, and ruffled fur).

Cell culture

Human HEK293T and A375 cells were obtained from the American Type Culture Collection (ATCC). U937 and K562 cells were kindly provided by Prof. Yuval Shaked; Kasumi-1 cells were from Prof. Tsila Zuckerman; and MV-4-11, GRANTA, THP-1, and MEC-1 cells were from Dr. Netanel Horowitz. MOLM-13 cells were kindly provided by Dr. Ani Deshpande (SBP Discovery Institute, USA), KG-1a, HL-60, Jurkat, RPMI 8226, and HAP-1 cells were a kind gift from Prof. Ciechanover (Technion, Israel). MOLM-13, U937, THP-1, Kasumi, Jurkat, and RPMI-8226 cells were cultured in RPMI medium; HL-60, MV-4-11, K-562, MEC-1, HAP-1, and KG-1α cells were cultured in IMDM; and GRANTA, A375, and HEK293T cells were cultured in DMEM. All media were supplemented with 10% fetal bovine serum (FBS), 1% l-glutamine, penicillin (83 U/mL), and streptomycin (83 µg/mL) (Gibco). Cells were regularly checked for mycoplasma contamination using a luminescence-based kit (Lonza).

Primary AML cells

AML patient samples were obtained from Scripps MD Anderson, La Jolla, CA (IRB-approved protocol 13-6180), and written informed consent was obtained from each participant. Samples were also obtained from the Rambam Health Campus Center, Haifa, Israel (IRB-approved protocol 0372-17). Fresh blood samples were obtained by a peripheral blood draw, PICC line, or central catheter. Filgrastim-mobilized peripheral blood cells were collected from healthy donors and cryopreserved in DMSO. PBMCs were isolated by centrifugation through Ficoll-Paque^TM PLUS (17-1440-02, GE Healthcare). Residual red blood cells were removed using RBC Lysis Buffer for humans (Alfa Aesar, cat. # J62990) according to the manufacturer’s instructions. The final PBMC pellets were resuspended in Bambanker serum-free freezing medium (Wako Pure Chemical Industries, Ltd.) and stored under liquid N₂. Patients’ information is provided in Supplementary Table 1.

MLL-AF9 patient-derived xenograft (PDX) samples (from the Jeremias Lab, Munich, Germany) were cultured in IMDM medium with 20% BIT (Stemcell Technologies), human cytokines, and StemRegenin 1 (SR1) and UM171, as described⁴¹. Cells were transduced with empty vector or different shRNF5 constructs as described below (see transfections and transduction section) and plated in 100 µL per well of a complete medium in 96-well plates. Growth was monitored every 24 h using the CellTiter Glo reagent.

For the drug dose responses, FK228 was diluted in DMSO at 10 mM and serially diluted (1/3, x13 concentrations) in a Labcyte Echo Low Dead Volume (LDV) plate. 25nLs of drugs at 1000x concentration were spotted in quadruplicate in 384-well plates (Greiner #781098) using a Labcyte Echo 550 acoustic dispenser, and patient AML cells (described above) were seeded (2.5 k cells/well in 25 µLs) onto 3 plates with a Multidrop Combi Reagent Dispenser (Thermo). After 2 days, cell viability was assessed by adding 10 µLs/well of CellTiter-Glo reagent (Promega #G7572) using a Multidrop Combi, and luminescence was read on an Envision plate reader (Perkin Elmer). Raw data were processed in Microsoft Excel, with cell viability values normalized to percentages relative to vehicle (0.1% DMSO) controls. Data were graphed and subjected to statistical analyses using GraphPad Prism software (v.9.1.1).

Antibodies and reagents

The RNF5 antibody was produced as described previously (1:1000)^7,23. Other antibodies used were: rabbit anti-cleaved caspase-3 (#9661, Cell Signaling Technology, 1:1000), rabbit anti-PARP (#9532, Cell Signaling Technology, 1:1000), mouse anti-RBBP4 (NBP1-41201, Novus Biologicals, 1:5000), mouse anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH; ab8245, Abcam, 1:10000), mouse anti-Tubulin (T9026, Sigma, 1:5000), mouse anti-Flag (F1804, Sigma, 1:2000), mouse anti-Myc-Tag (#2276, Cell Signaling Technology, 1:1000), mouse anti-HA (901501, Biolegend, 1:2000), rabbit anti-HDAC1 (#2062, Cell Signaling Technology, 1:1000), rabbit anti-HDAC2 (57156, Cell Signaling Technology, 1:1000), rabbit anti-Ezh2 (5246, Cell Signaling Technology, 1:1000), mouse anti-HSP90 (sc-13119, Santa Cruz Biotechnology, 1:1000), rabbit anti-p27, (#3688, Cell Signaling Technology, 1:1000), rabbit anti-p21 (#2947, Cell Signaling Technology, 1:1000), mouse anti-Ubiquitin (#3939, Cell Signaling Technology, 1:1000), rabbit anti-K63-linkage Specific Polyubiquitin (#5621, Cell Signaling Technology, 1:1000), rabbit anti-Actin (#4970, Cell Signaling Technology, 1:1000), rabbit anti-Histone H3 (#9717, Cell Signaling Technology, 1:1000), mouse anti-Caspase-3 (sc-56053, Santa Cruz Biotechnology, 1:1000), and mouse anti- Calregulin (sc-166837, Santa Cruz Biotechnology, 1:1000). HRP-conjugated secondary antibodies were from Jackson ImmunoResearch (goat-anti-mouse-HRP (AB_2338504) and goat-anti-rabbit-HRP (AB_2337938) and diluted 1:10000.

Romidepsin and N-acetyldinaline were purchased from Cayman Chemicals. Thapsigargin and tunicamycin were purchased from Sigma-Aldrich. MG132 was obtained from Selleckchem. Puromycin was purchased from Merck. Annexin V-APC and propidium iodide were from BioLegend.

Plasmids and constructs

Plasmids expressing Flag-RNF5-WT, Flag-RNF5-RM, and Flag-RNF5-∆CT were described previously^5,7. To generate doxycycline-inducible RNF5-WT, RNF5-RM, and RNF5-∆CT overexpression vectors, coding sequences were amplified by PCR from pCDNA3.1-RNF5-WT, pCDNA3.1-RNF5-RM, and pCDNA3.1-RNF5-∆CT, respectively, and the product was inserted into EcoRI-linearized pLVX TetOne-puro plasmid (Clontech) using the NEBuilder HiFi Assembly kit (New England BioLabs). Expression vectors encoding Myc-RBBP4 (#20715), HA-Ubiquitin (#18712), and HA-ubiquitin mutants (including K6 (#22900), K11 (#22901), K27 (#22902), K29 (#22903), and K33 (#17607)) were obtained from Addgene.

Gene silencing

Lentiviral pLKO.1 vectors expressing RNF5 or RBBP4-specific shRNAs were obtained from the La Jolla Institute for Immunology RNAi Center (La Jolla, CA, USA). Sequences were: shRNF5 #1 (TRCN0000004785) GAGTGTCCAGTATGTAAAGCT, shRNF5 #2 (TRCN0000004788) CGGCAAGAGTGTCCAGTATGT, shRNF5 #3 GAGGATGGATTGAGAGAAT, and inducible shRNF5, which has the same sequence as shRNF5#1. Sequences for RBBP4-specific shRNAs were: shRBBP4#1 (TRCN0000286103) GCCTTTCTTTCAATCCTTATA, shRBBP4#2 (TRCN0000293556) TGGTCATACTGCCAAGATATC, shRBBP4#3 (TRCN0000293554) ATGCGTCACACTACGACAGTG.

Transfections and transduction

Transient transfections were carried out using CalFectin (SignaGen) according to the manufacturer’s recommendations. Lentiviral particles were prepared using standard protocols. In brief, HEK293T cells were transfected with relevant vectors and the second-generation packaging plasmids ΔR8.2 and Vsv-G (Addgene). Virus-containing supernatants were collected 48 h later and then added in the presence of Polybrene to AML cells pre-seeded at ~5 × 10⁵/well in 24-well plates (Sigma-Aldrich). After 8 h, cells were transferred to 10 cm culture dishes for an additional 24 h prior to experiments.

Western blotting

Cells were washed twice with cold PBS and lysed by addition of Tris-buffered saline (TBS)-lysis buffer (TBS [50 mM Tris-HCl pH 7.5, 150 mM NaCl], 0.5% Nonidet P-40, 1× protease inhibitor cocktail [Merck], and 1× phosphatase inhibitor cocktail⁴² followed by incubation on ice for 20 min. Blood cells from healthy control subjects and AML patients were lysed using hot lysis buffer [100 mM Tris-HCl pH 7.5, 5% sodium dodecyl sulfate (SDS)] followed by incubation 5 min at 95 °C and sonication. Some samples were subjected to fractionation using a subcellular protein fractionation kit (Thermo Scientific Pierce), as indicated. Samples were resolved on SDS-PAGE and transferred to nitrocellulose membranes. Membranes were incubated for 1 h at room temperature with blocking solution (0.1% Tween 20/5% non-fat milk in TBS) and then overnight at 4 °C with primary antibodies. Membranes were washed with TBS and incubated for 1 h at room temperature with appropriate secondary antibodies (Jackson ImmunoResearch). Finally, proteins were visualized using a chemiluminescence method (Image-Quant LAS400, GE Healthcare, or ChemiDoc MP imaging system, Bio-Rad). The uncropped scans for all western blot are provided in the Source Data file.

Immunoprecipitation

Cells were lysed in TBS-lysis buffer as described above, centrifuged for 10 min at 17,000 g, and incubated overnight at 4 °C with appropriate antibodies. Protein A/G agarose beads (Santa Cruz Biotechnology) were then added for 2 h at 4 °C. Beads were pelleted by centrifugation, washed five times with TBS-lysis buffer, and boiled in Laemmli buffer to elute proteins. Finally, proteins were resolved on SDS-PAGE and subjected to Western blotting as described above.

LC-MS/MS

MOLM-13 cells were infected with doxycycline-inducible Flag-tagged RNF5-encoding or empty plasmids and expression were induced by the addition of doxycycline (1 µg/mL) for 48 h. The proteasome inhibitor MG132 (10 μM) was added for 4 h prior to harvest. Cells were lysed in TBS-lysis buffer, and total cell lysates were incubated with anti-Flag-M2-agarose beads (Sigma-Aldrich) overnight at 4 °C. Beads were washed with TBS-lysis buffer, and proteins were eluted from beads by addition of 3×Flag peptides (150 μg/mL, Sigma) for 1 h at 4 °C and then subjected to tryptic digestion followed by LC-MS/MS, as described⁴³.

Raw data were analyzed using MaxQuant (v1.5.5.0)⁴⁴ with default settings. Protein intensities were normalized using the median centering method. Fold-changes were calculated by dividing the protein intensity of Flag immunoprecipitates from RNF5-overexpressing cells by the protein intensity of Flag immunoprecipitates from control cells. Thresholds were set at 2 for fold-change and 0.05 for p value obtained using a two-sided Welch’s t-test. Proteins identified in all RNF5 immunoprecipitation replicates but in one or no control IP replicates were considered potential RNF5 interactors if their corresponding fold-change was at least 2. Data from the Crapome (version 2.0)⁴² repository were downloaded to filter potential contaminants. Cytoscape (version 3.8.1)⁴⁵ was used to generate the RNF5 interaction network and pathway enrichment analysis. Raw MS data were deposited in the MassIVE repository under the accession code MSV000083160.

Immunofluorescence microscopy

Cells were placed on coverslips on glass slides using a StatSpin cytofuge and fixed with 4% paraformaldehyde for 20 min at room temperature. Slides were then rinsed three times in PBS, and cells were permeabilized in 0.2% Triton X-100 in PBS for 5 min and blocked with 0.2% TX-100/10% FBS in PBS for 30 min. Primary antibodies were diluted in staining buffer (0.2% Triton X-100/2% FBS in PBS) and added to cells, and the slides were then incubated overnight at 4 °C in a humidified chamber. Slides were then washed three times in staining buffer, and secondary antibodies (Life Technologies) were diluted in staining buffer and added to slides for 1 h at room temperature in a humidified chamber shielded from light. Finally, slides were washed three times in staining buffer and mounted with Fluoroshield Mounting Medium containing 4′, 6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich). Cells were analyzed using a fluorescence microscope (DMi8; Leica) with a 60× oil immersion objective. Images were processed using the 3D deconvolution tool from LASX software (Leica), and the same parameters were used to analyze all images.

Cell viability assay

Cell viability and growth were assayed using the CellTiter Glo kit (Promega) according to the manufacturer’s recommendations. Cell lines were plated in white 96-well clear-bottomed plates (Corning) at a density of 7 × 10³ cells/well, and growth was monitored every 24 h using CellTiter Glo reagent. Viability was quantified by measuring luminescence intensity with an Infinite 2000 Pro reader (Tecan).

Cell cycle analysis

The distribution of cells in each phase of the cell cycle was analyzed by propidium iodide staining (Merck). Briefly, 1 × 10⁶ cells were washed twice with cold PBS and fixed in 70% ethanol in PBS at 4 °C overnight. Cells were washed, pelleted by centrifugation, and treated with RNase A (100 μg/mL) and propidium iodide (40 μg/mL) at room temperature for 30 min. Cell cycle distribution was assessed by flow cytometry (BD LSRFortessa™, BD Biosciences), and data were analyzed using FlowJo software.

Annexin V and propidium iodide staining

Cells were collected in FACS tubes, washed twice with ice-cold PBS, and resuspended in 100 µL PBS. Annexin V-APC (1.4 µg/mL) was added for 15 min at room temperature in the dark. Then, cells were washed in PBS and resuspended in 200 µL PBS, and then propidium iodide (50 µg/mL) was added. Finally, samples were analyzed by flow cytometry (BD LSRFortessa™, BD Biosciences). The gating strategy is provided in Supplementary Fig. 8c.

Colony-forming assays

For the soft agar assay, a base layer was formed by mixing 1.5% soft agar (low-melting-point agarose, Bio-Rad) and culture medium at a 1:1 ratio and placing the mixture in 6-well plates. Cells were resuspended in a medium containing 0.3% soft agar and added to the base layer at 1 × 10⁴ (MOLM-13) or 5 × 10³ (U937) cells/well. Agar was solidified by incubation at 4 °C for 10 min before incubation at 37 °C. Plates were incubated at 37 °C in a humidified atmosphere for 12–18 days. Cells were then fixed overnight with 4% paraformaldehyde, washed with PBS, and stained with 0.05% crystal violet (Merck) for 20 min at room temperature, and washed again with PBS. Plates were photographed and colonies were counted on the captured images.

For the methylcellulose assay, WT or Rnf5^–/– Lin⁻Sca1⁺c-Kit⁺ cells transformed with GFP-MLL-AF9 were resuspended in methylcellulose M3234 (Stem Cell Technologies) supplemented with 6 ng/mL IL-3, 10 ng/mL IL-6, and 20 ng/mL stem cell factor (PeproTech). Cells were then added to 35 mm dishes at 10³ cells/well and incubated for 6–7 days. Colonies were classified as compact and hypercellular (blast-like) or small and diffuse (differentiated). Virtually all colonies fell into one of these two categories.

RT-qPCR analysis

RNA was extracted using a GenElute Mammalian Total RNA Purification Kit (Sigma) according to standard protocols. RNA concentration was measured using a NanoDrop spectrophotometer (ThermoFisher). cDNA was synthesized from aliquots of 1 μg total RNA using a qScript cDNA Synthesis Kit (Quanta). Quantitative PCR was performed with SYBR Green I dye master mix (Roche) and a CFX connect Real-Time PCR System (Bio-Rad). Primer sequences are listed in Supplementary Table 3. Primer efficiency was measured in preliminary experiments, and amplification specificity was confirmed by dissociation curve analysis.

Gene targeting using CRISPR/Cas9

RNF5 sgRNAs were cloned into the pKLV2-U6gRNA-(BbsI)-PGKpuro2ABFP-W lentiviral expression vector and transduced into Cas9-expressing cell lines. All gRNAs were cloned into the BbsI site of the gRNA expression vector as previously described⁴⁶. Briefly, HEK293T cells were co-transfected with pKLV2-U6gRNA-(BbsI)-PGKpuro2ABFP-W and ectopic packaging plasmids using CalFectin transfection reagent (SignaGen). Virus-containing supernatants were collected 48 h later. MOLM-13 cells were infected by the addition of supernatants for 48 h. Cells were then selected with puromycin (0.5 µg/mL) for 48 h and viability was measured. The RNF5-targeting sgRNA sequences were: sgRNF5 #3 F-GCACCTGTACCCCGGCGGAA and R-TTCCGCCGGGGTACAGGTGC, and sgRNF5 #4 F-GTTCCGCCGGGGTACAGGTG and R-CACCTGTACCCCGGCGGAAC.

RNA-seq analysis

PolyA RNA was isolated using the NEBNext Poly(A) mRNA Magnetic Isolation Module, and bar-coded libraries were constructed using the NEBNext Ultra™ Directional RNA Library Prep Kit for Illumina (NEB, Ipswich, MA). Libraries were pooled and single end-sequenced (1 × 75) on the Illumina NextSeq 500 using the High output V2 kit (Illumina, San Diego, CA). Quality control was performed using Fastqc (v0.11.5, Andrews S. 2010), reads were trimmed for adapters, low quality 5′ bases, and the minimum length of 20 using CUTADAPT (v1.1). The number of reads per sample and the number of aligned reads suggested that read quality and number were good and that the data were valid for analysis. High-quality data were then mapped to a human reference genome (hg19) using the STAR mapping algorithm (version 2.5.2a)⁴⁷. Feature Counts implemented in Subread (v1.50)⁴⁸ were used to count the sequencing reads from mapped BAM files. Analyses of differentially expressed genes were subsequently performed using a negative binomial test method (edgeR, v3.34.0)⁴⁹ implemented using SARTools R Package (v1.2.0)⁵⁰. A list of the differentially expressed genes was exported into an excel file (Supplementary Data file 1), and pathway analysis was performed by uploading the lists of differentially expressed genes to IPA (http://www.ingenuity.com) using the following criteria: |log2(fold-change) | > 0.4 and P value < 0.05. P values were determined using “Negative Binomial Generalized Linear Model (two-sided)” to generate the DEGs list. Multiple comparisons were also applied based on the “Benjamini & Hochberg” method. LINCS database⁵¹ and other public data sets were processed by IPA. Molecular signatures for canonical pathways, upstream regulators, and causal networks were generated for each data set by IPA. Enrichment results in this study were compared to the LINCS molecular signatures by Analysis Match using z-scores developed by IPA. The z-scores represent how well activated or inhibited entities match data sets (% similarity). The top matched experiments in LINCS were selected by ranking the overall z-scores.

Chromatin immunoprecipitation (ChIP)

ChIP analysis was performed using a ChIP Assay Kit (Millipore) following the manufacturer’s instructions. Briefly, 1 × 10⁶ cells were used for each reaction. Cells were fixed in 1% formaldehyde at 37 °C for10 min, and nuclei were isolated with nuclear lysis buffer (Millipore) supplemented with a protease inhibitor cocktail (Millipore). Chromatin DNA was sonicated and sheared to a length between 200 bp and 1000 bp. Sheared chromatin was immunoprecipitated at 4 °C overnight with anti-H3K9ac (9649, Cell Signaling Technology), anti-H3K27ac (ab3594, Abcam), anti-H3K27me3 (9733, Cell Signaling Technology), or anti-RBBP4 (NBP1-41201, Novus). IgG was used as a negative control and anti-RNA polII (Millipore) served as a positive control antibody. Protein A/G bead-antibody/chromatin complexes were washed with low salt buffer, high salt buffer, LiCl buffer, and then TE buffer to eliminate nonspecific binding. Protein/DNA complexes were reverse cross-linked, and DNA was purified using NucleoSpin^®. Levels of ChIP-purified DNA were determined by qPCR (see Supplementary Table 4 for primer sequences). Relative enrichment of the indicated DNA regions were calculated using the Percent Input Method according to the manufacturer’s instructions and are presented as % input.

Small molecule epigenetic regulator screen

Aliquots of compounds (10 mM in DMSO) from a library of 261 epigenetic regulators were dispensed at final concentrations of 0.5 µM or 5 µM into the wells of a Greiner (Monroe, NC, Cat #781080) 384-well TC-treated black plate using a Labcyte Echo 555 acoustic pipette (Labcyte, San Jose, CA). U937 cells expressing an inducible shRNF5 vector were induced with doxycycline for 72 h and dispensed into the prepared plates at a density of 5 × 10²/well in 50 µL RPMI-based culture medium (described above) using a Multidrop Combi (Thermo Fisher Scientific, Pittsburgh, PA). Plates were briefly centrifuged at ~100 g and incubated at 37 °C with 5% CO₂ for 6 more days using MicroClime Environmental lids (Labcyte, San Jose, CA). Plates were placed at room temperature for 30 min to equilibrate, 20 µL/well CellTiter-Glo Luminescent Cell Viability Assay reagent (Promega, Madison, WI) was added using a Multidrop Combi, and plates were analyzed with an EnVision multimode plate reader (PerkinElmer, Waltham, MA).

For the analysis, the intensity of induced shRNF5-expressing cells was divided by the intensity of uninduced cells. Ratios were log₂ transformed and thresholds were calculated based on the distribution of the log₂ ratios. The upper threshold was calculated as the Q3 + 1.5xQ, where Q3 is the third quartile and IQ is interquartile. The lower threshold was calculated as the Q1 − 1.5xIQ, where Q1 is the first quartile. Ratios outside these thresholds were considered outliers from the global ratio distribution and thus were potential candidates for having a differential effect on RNF5-KD or control cells.

MLL-AF9–mediated transformation of bone marrow cells and generation of MLL-AF9–leukemic mice

HEK293T cells were co-transfected with Murine Stem Cell Virus (MSCV)-based MLL-AF9 IRES-GFP²² and ectopic packaging plasmids. Viral supernatants were collected 48 h later and added to Lin⁻Sca-1⁺c-Kit⁺ cells isolated from the bone marrow of WT or Rnf5^−/− C57BL/6 mice. Transduced cells were maintained in DMEM supplemented with 15% FBS, 6 ng/mL IL-3, 10 ng/mL IL-6, and 20 ng/mL stem cell factor, and transformed cells were selected by sorting for GFP⁺cells. To generate “primary AML mice,” GFP-MLL-AF9–transduced cells were resuspended in PBS at 1 × 10⁶ cells/200 µL and injected intravenously into sublethally irradiated (650 Rad) 6- to 8-week-old C57BL/6 female mice.

Statistical analysis

Differences between two groups were assessed using two-tailed unpaired or paired t-tests or Wilcoxon rank-sum test, and differences between group means were evaluated using t-tests or ANOVA. Two-way ANOVA with Tukey’s multiple comparison test was used to evaluate experiments involving multiple groups. Survival was analyzed by the Kaplan–Meier method and evaluated with a log-rank test. All data were analyzed using GraphPad Prism version 8 or 9 (GraphPad, La Jolla, CA, USA) and expressed as means ± SD or SEM. P < 0.05 was considered significant. NS stands for not statistically significant.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The RAW MS data have been deposited in the MassIVE repository under accession code MSV000083160 (https://massive.ucsd.edu/ProteoSAFe/dataset.jsp?task=321eefef71fe4baa8900da284d5f66f3). RNA-Seq RAW data in FASTQ format from RNF5 knockdown experiments have been deposited in the NCBI Gene Expression Omnibus (GEO) database under accession code GSE155929. The raw data of TCGA database are available through GDC Data Portal (https://portal.gdc.cancer.gov/). The raw data of LINCS are available through NIH LINCS (https://lincsproject.org/). All other data are available within the article and its Supplementary Information. Source data are provided with this paper.

References

Cancer Genome Atlas Research, N. et al. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N. Engl. J. Med 368, 2059–2074 (2013).
Article CAS Google Scholar
Mrozek, K., Heerema, N. A. & Bloomfield, C. D. Cytogenetics in acute leukemia. Blood Rev. 18, 115–136 (2004).
Article PubMed Google Scholar
Shen, Y. et al. Gene mutation patterns and their prognostic impact in a cohort of 1185 patients with acute myeloid leukemia. Blood 118, 5593–5603 (2011).
Article CAS PubMed Google Scholar
Dohner, H. et al. Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood 115, 453–474 (2010).
Article PubMed CAS Google Scholar
Tcherpakov, M. et al. Regulation of endoplasmic reticulum-associated degradation by RNF5-dependent ubiquitination of JNK-associated membrane protein (JAMP). J. Biol. Chem. 284, 12099–12109 (2009).
Article CAS PubMed PubMed Central Google Scholar
Younger, J. M. et al. Sequential quality-control checkpoints triage misfolded cystic fibrosis transmembrane conductance regulator. Cell 126, 571–582 (2006).
Article CAS PubMed Google Scholar
Bromberg, K. D. et al. Increased expression of the E3 ubiquitin ligase RNF5 is associated with decreased survival in breast cancer. Cancer Res 67, 8172–8179 (2007).
Article CAS PubMed PubMed Central Google Scholar
Jeon, Y. J. et al. Regulation of glutamine carrier proteins by RNF5 determines breast cancer response to ER stress-inducing chemotherapies. Cancer Cell 27, 354–369 (2015).
Article CAS PubMed PubMed Central Google Scholar
Kuang, E. et al. Regulation of ATG4B stability by RNF5 limits basal levels of autophagy and influences susceptibility to bacterial infection. PLoS Genet 8, e1003007 (2012).
Article CAS PubMed PubMed Central Google Scholar
Fujita, Y. et al. Regulation of S100A8 stability by RNF5 in intestinal epithelial cells determines intestinal inflammation and severity of colitis. Cell Rep. 24, 3296–3311 e3296 (2018).
Article CAS PubMed PubMed Central Google Scholar
Garcia-Manero, G. et al. Phase 1 study of the oral isotype specific histone deacetylase inhibitor MGCD0103 in leukemia. Blood 112, 981–989 (2008).
Article CAS PubMed PubMed Central Google Scholar
Masetti, R., Serravalle, S., Biagi, C. & Pession, A. The role of HDACs inhibitors in childhood and adolescence acute leukemias. J. Biomed. Biotechnol. 2011, 148046 (2011).
Redner, R. L., Wang, J. & Liu, J. M. Chromatin remodeling and leukemia: new therapeutic paradigms. Blood 94, 417–428 (1999).
Article CAS PubMed Google Scholar
Bhalla, K. N. Epigenetic and chromatin modifiers as targeted therapy of hematologic malignancies. J. Clin. Oncol. 23, 3971–3993 (2005).
Article CAS PubMed Google Scholar
Millard, C. J. et al. The structure of the core NuRD repression complex provides insights into its interaction with chromatin. Elife 5, e13941 (2016).
Article PubMed PubMed Central CAS Google Scholar
Zhang, Y., Iratni, R., Erdjument-Bromage, H., Tempst, P. & Reinberg, D. Histone deacetylases and SAP18, a novel polypeptide, are components of a human Sin3 complex. Cell 89, 357–364 (1997).
Article CAS PubMed Google Scholar
Guo, Q. et al. Expression of HDAC1 and RBBP4 correlate with clinicopathologic characteristics and prognosis in breast cancer. Int J. Clin. Exp. Pathol. 13, 563–572 (2020).
PubMed PubMed Central Google Scholar
Li, Y. D., Lv, Z., Xie, H. Y. & Zheng, S. S. Retinoblastoma binding protein 4 up-regulation is correlated with hepatic metastasis and poor prognosis in colon cancer patients. Hepatobiliary Pancreat. Dis. Int 18, 446–451 (2019).
Article PubMed Google Scholar
Kitange, G. J. et al. Retinoblastoma Binding Protein 4 Modulates Temozolomide Sensitivity in Glioblastoma by Regulating DNA Repair Proteins. Cell Rep. 14, 2587–2598 (2016).
Article CAS PubMed PubMed Central Google Scholar
Heckl, B. C. et al. Frequent and reliable engraftment of certain adult primary acute lymphoblastic leukemias in mice. Leuk. Lymphoma Mar. 60, 848–851 (2019).
Article Google Scholar
Breckenridge, D. G., Germain, M., Mathai, J. P., Nguyen, M. & Shore, G. C. Regulation of apoptosis by endoplasmic reticulum pathways. Oncogene 22, 8608–8618 (2003).
Article CAS PubMed Google Scholar
Krivtsov, A. V. et al. Transformation from committed progenitor to leukaemia stem cell initiated by MLL-AF9. Nature 442, 818–822 (2006).
Article ADS CAS PubMed Google Scholar
Delaunay, A. et al. The ER-bound RING finger protein 5 (RNF5/RMA1) causes degenerative myopathy in transgenic mice and is deregulated in inclusion body myositis. PLoS One 3, e1609 (2008).
Article ADS PubMed PubMed Central CAS Google Scholar
Pan, L. et al. DHX15 is associated with poor prognosis in acute myeloid leukemia (AML) and regulates cell apoptosis via the NF-kB signaling pathway. Oncotarget 8, 89643–89654 (2017).
Article PubMed PubMed Central Google Scholar
Tabe, Y. et al. Novel role of HDAC inhibitors in AML1/ETO AML cells: activation of apoptosis and phagocytosis through induction of annexin A1. Cell Death Differ. 14, 1443–1456 (2007).
Article CAS PubMed Google Scholar
Watek, M. et al. Hypogelsolinemia in patients diagnosed with acute myeloid leukemia at initial stage of sepsis. Med Sci. Monit. 25, 1452–1458 (2019).
Article CAS PubMed PubMed Central Google Scholar
Kuzmichev, A., Nishioka, K., Erdjument-Bromage, H., Tempst, P. & Reinberg, D. Histone methyltransferase activity associated with a human multiprotein complex containing the enhancer of Zeste protein. Genes Dev. 16, 2893–2905 (2002).
Article CAS PubMed PubMed Central Google Scholar
Allen, H. F., Wade, P. A. & Kutateladze, T. G. The NuRD architecture. Cell Mol. Life Sci. 70, 3513–3524 (2013).
Article CAS PubMed PubMed Central Google Scholar
Casas, S. et al. Changes in apoptosis-related pathways in acute myelocytic leukemia. Cancer Genet. Cytogenet 146, 89–101 (2003).
Article CAS PubMed Google Scholar
Morito, D. et al. Gp78 cooperates with RMA1 in endoplasmic reticulum-associated degradation of CFTRDeltaF508. Mol. Biol. Cell 19, 1328–1336 (2008).
Article CAS PubMed PubMed Central Google Scholar
Xu, C. et al. Binding of different histone marks differentially regulates the activity and specificity of polycomb repressive complex 2 (PRC2). Proc. Natl Acad. Sci. USA 107, 19266–19271 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Liu, Z. et al. Structural basis of plant homeodomain finger 6 (PHF6) recognition by the retinoblastoma binding protein 4 (RBBP4) component of the nucleosome remodeling and deacetylase (NuRD) complex. J. Biol. Chem. 290, 6630–6638 (2015).
Article CAS PubMed PubMed Central Google Scholar
Duque-Afonso, J. et al. The HDAC class I-specific inhibitor entinostat (MS-275) effectively relieves epigenetic silencing of the LAT2 gene mediated by AML1/ETO. Oncogene 30, 3062–3072 (2011).
Article CAS PubMed Google Scholar
Clarke, K. et al. The histone deacetylase inhibitor Romidepsin induces as a cascade of differential gene expression and altered histone H3K9 marks in myeloid leukaemia cells. Oncotarget 10, 3462–3471 (2019).
Article PubMed PubMed Central Google Scholar
Argnani, L., Broccoli, A. & Zinzani, P. L. Cutaneous T-cell lymphomas: Focusing on novel agents in relapsed and refractory disease. Cancer Treat. Rev. 61, 61–69 (2017).
Article CAS PubMed Google Scholar
Byrd, J. C. et al. A phase 1 and pharmacodynamic study of depsipeptide (FK228) in chronic lymphocytic leukemia and acute myeloid leukemia. Blood 105, 959–967 (2005).
Article CAS PubMed Google Scholar
Lee, J. S. et al. Harnessing synthetic lethality to predict the response to cancer treatment. Nat. Commun. 9, 2546 (2018).
Article ADS PubMed PubMed Central CAS Google Scholar
Li, Y. et al. Gut microbiota dependent anti-tumor immunity restricts melanoma growth in Rnf5(−/−) mice. Nat. Commun. 10, 1492 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Foot, N., Henshall, T. & Kumar, S. Ubiquitination and the regulation of membrane proteins. Physiol. Rev. 97, 253–281 (2017).
Article PubMed Google Scholar
Zhao, Q. et al. GPS-SUMO: a tool for the prediction of sumoylation sites and SUMO-interaction motifs. Nucleic Acids Res 42, W325–W330 (2014).
Article CAS PubMed PubMed Central Google Scholar
Pabst, C. et al. Identification of small molecules that support human leukemia stem cell activity ex vivo. Nat. Methods 11, 436–442 (2014).
Article CAS PubMed Google Scholar
Mellacheruvu, D. et al. The CRAPome: a contaminant repository for affinity purification-mass spectrometry data. Nat. Methods 10, 730–736 (2013).
Article CAS PubMed PubMed Central Google Scholar
Braten, O. et al. Numerous proteins with unique characteristics are degraded by the 26S proteasome following monoubiquitination. Proc. Natl Acad. Sci. USA 113, E4639–E4647 (2016).
Article CAS PubMed PubMed Central Google Scholar
Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367–1372 (2008).
Article CAS PubMed Google Scholar
Shannon, P. et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13, 2498–2504 (2003).
Article CAS PubMed PubMed Central Google Scholar
Tzelepis, K. et al. A CRISPR dropout screen identifies genetic vulnerabilities and therapeutic targets in acute myeloid leukemia. Cell Rep. 17, 1193–1205 (2016).
Article CAS PubMed PubMed Central Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014).
Article CAS PubMed Google Scholar
Robinson, M. D. & Oshlack, A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 11, R25 (2010).
Article PubMed PubMed Central CAS Google Scholar
Varet, H., Brillet-Gueguen, L., Coppee, J. Y. & Dillies, M. A. SARTools: a DESeq2- and EdgeR-based R pipeline for comprehensive differential analysis of RNA-Seq data. PLoS One 11, e0157022 (2016).
Article PubMed PubMed Central CAS Google Scholar
Subramanian, A. et al. A next generation connectivity map: L1000 platform and the first 1,000,000 profiles. Cell 171, 1437–1452 e1417 (2017).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank Drs. Yuval Shaked, Tsila Zuckerman, and Netanel Horowitz (Faculty of Medicine, Technion) for providing leukemic cell lines, and members of the Deshpande and Ronai labs for technical support and discussions. We thank SBP and Technion Core facilities for help along with the different phases of this study. We thank Nancy R. Gough (BioSerendipity, LLC) for editorial assistance. Z.A.R. gratefully acknowledges support from the National Cancer Institute grant (R35CA197465) and the Technion. A.K. was supported by a Faculty of Medicine fellowship at the Technion. Sanford Burnham Prebys Shared Resources are supported by an NCI Cancer Center Support Grant (P30 CA030199).

Author information

Yu Fujita
Present address: Division of Respiratory Medicine, Department of Internal Medicine, Jikei University School of Medicine, Tokyo, Japan

Authors and Affiliations

Technion Integrated Cancer Center, Faculty of Medicine, Technion Israel Institute of Technology, Haifa, Israel
Ali Khateb, Ikrame Lazar, Bertrand Fabre, Ido Livneh & Yishai Ofran
Cancer Center, Sanford Burnham Prebys Medical Discovery Institute, La Jolla, CA, USA
Ali Khateb, Anagha Deshpande, Yongmei Feng, Darren Finlay, Ikrame Lazar, Bertrand Fabre, Yan Li, Yu Fujita, Jun Yin, Ian Pass, Kristiina Vuori, Michael Jackson, Aniruddha J. Deshpande & Ze’ev A. Ronai
Cancer Data Science Lab (CDSL), National Cancer Institute, National Institute of Health, Bethesda, MD, USA
Joo Sang Lee & Eytan Ruppin
Laboratory of Translational Genomics, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD, USA
Tongwu Zhang & Kevin Brown
Research Unit Apoptosis in Hematopoietic Stem Cells, Helmholtz Center Munich, German Center for Environmental Health, Munich, Germany
Irmela Jeremias
Scripps MD Anderson Cancer Center, La Jolla, CA, USA
Carol Burian & James R. Mason
Rambam Health Care Campus, Epidemiology Department and Biobank, Haifa, Israel
Ronit Almog
Rambam Health Care Campus, Hematology and Bone marrow Transplantation Department, Haifa, Israel
Nurit Horesh & Yishai Ofran

Authors

Ali Khateb
View author publications
You can also search for this author in PubMed Google Scholar
Anagha Deshpande
View author publications
You can also search for this author in PubMed Google Scholar
Yongmei Feng
View author publications
You can also search for this author in PubMed Google Scholar
Darren Finlay
View author publications
You can also search for this author in PubMed Google Scholar
Joo Sang Lee
View author publications
You can also search for this author in PubMed Google Scholar
Ikrame Lazar
View author publications
You can also search for this author in PubMed Google Scholar
Bertrand Fabre
View author publications
You can also search for this author in PubMed Google Scholar
Yan Li
View author publications
You can also search for this author in PubMed Google Scholar
Yu Fujita
View author publications
You can also search for this author in PubMed Google Scholar
Tongwu Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Jun Yin
View author publications
You can also search for this author in PubMed Google Scholar
Ian Pass
View author publications
You can also search for this author in PubMed Google Scholar
Ido Livneh
View author publications
You can also search for this author in PubMed Google Scholar
Irmela Jeremias
View author publications
You can also search for this author in PubMed Google Scholar
Carol Burian
View author publications
You can also search for this author in PubMed Google Scholar
James R. Mason
View author publications
You can also search for this author in PubMed Google Scholar
Ronit Almog
View author publications
You can also search for this author in PubMed Google Scholar
Nurit Horesh
View author publications
You can also search for this author in PubMed Google Scholar
Yishai Ofran
View author publications
You can also search for this author in PubMed Google Scholar
Kevin Brown
View author publications
You can also search for this author in PubMed Google Scholar
Kristiina Vuori
View author publications
You can also search for this author in PubMed Google Scholar
Michael Jackson
View author publications
You can also search for this author in PubMed Google Scholar
Eytan Ruppin
View author publications
You can also search for this author in PubMed Google Scholar
Aniruddha J. Deshpande
View author publications
You can also search for this author in PubMed Google Scholar
Ze’ev A. Ronai
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

A.K. and Z.A.R. designed the experiments; A.K., A.D., Yo.F., D.F, I.L., Y.L., and Id.L. performed experiments; A.K., J.S.L., B.F., D.F., Y.F., T.Z., J.Y., K.B., E.R., A.J.D, and Z.A.R. analyzed the data; I.P. and M.J. performed the epigenetic screen; D.F., I.J, R.A., N.H., Y.O., C.B., J.R.M., and K.V. provided access to patient samples and data; and A.K. and Z.A.R. wrote the paper.

Corresponding author

Correspondence to Ze’ev A. Ronai.

Ethics declarations

Competing interests

ZAR is co-founder and serves as scientific advisor to Pangea Therapeutics. All other authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Khateb, A., Deshpande, A., Feng, Y. et al. The ubiquitin ligase RNF5 determines acute myeloid leukemia growth and susceptibility to histone deacetylase inhibitors. Nat Commun 12, 5397 (2021). https://doi.org/10.1038/s41467-021-25664-7

Download citation

Received: 23 May 2021
Accepted: 18 August 2021
Published: 13 September 2021
DOI: https://doi.org/10.1038/s41467-021-25664-7

This article is cited by

Downregulation of EphA2 stability by RNF5 limits its tumor-suppressive function in HER2-negative breast cancers
- Xiaojuan Li
- Fan Wang
- Ersheng Kuang
Cell Death & Disease (2023)
The role of E3 ubiquitin ligase WWP2 and the regulation of PARP1 by ubiquitinated degradation in acute lymphoblastic leukemia
- Xinxin Lu
- Xinyue Huang
- Lijun Zhang
Cell Death Discovery (2022)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.