Exosome-Transmitted PSMA3 and PSMA3-AS1 Promote Proteasome Inhibitor Resistance in Multiple Myeloma

Published OnlineFirst January 4, 2019; DOI: 10.1158/1078-0432.CCR-18-2363

Translational Cancer Mechanisms and Therapy Clinical Cancer Research Exosome-Transmitted PSMA3 and PSMA3-AS1 Promote Proteasome Inhibitor Resistance in Multiple Myeloma Hongxia Xu1,2, Huiying Han1, Sha Song1, Nengjun Yi3, Chen'ao Qian4, Yingchun Qiu1, Wenqi Zhou1, Yating Hong5, Wenyue Zhuang6, Zhengyi Li7, Bingzong Li5, and Wenzhuo Zhuang1

Abstract

Purpose: How exosomal RNAs released within the bone Results: We identified that PSMA3 and PSMA3-AS1 in MSCs marrow microenvironment affect proteasome inhibitors' (PI) could be packaged into exosomes and transferred to myeloma sensitivity of multiple myeloma is currently unknown. This cells, thus promoting PI resistance. PSMA3-AS1 could form an study aims to evaluate which exosomal RNAs are involved and RNA duplex with pre-PSMA3, which transcriptionally promot- by which molecular mechanisms they exert this function. ed PSMA3 expression by increasing its stability. In xenograft Experimental Design: Exosomes were characterized by models, intravenously administered siPSMA3-AS1 was found dynamic light scattering, transmission electron microscopy, and to be effective in increasing carfilzomib sensitivity. Moreover, Western blot analysis. Coculture experiments were performed to plasma circulating exosomal PSMA3 and PSMA3-AS1 derived assess exosomal RNAs transferring from mesenchymal stem from patients with multiple myeloma were significantly asso- cells (MSC) to multiple myeloma cells. The role of PSMA3-AS1 ciated with PFS and OS. in PI sensitivity was further evaluated in vivo. To determine the Conclusions: This study suggested a unique role of exoso- prognostic significance of circulating exosomal PSMA3 and mal PSMA3 and PSMA3-AS1 in transmitting PI resistance from PSMA3-AS1, a cohort of patients with newly diagnosed multiple MSCs to multiple myeloma cells, through a novel exosomal myeloma was enrolled to study. Cox regression models and PSMA3-AS1/PSMA3 signaling pathway. Exosomal PSMA3 and Kaplan–Meier curves were used to analyze progression-free PSMA3-AS1 might act as promising therapeutic targets for PI survival (PFS) and overall survival (OS). resistance and prognostic predictors for clinical response.

mechanisms underlying proteasome inhibitor (PI) resistance Introduction have been studied: genetic mutations, gene expression signatures, copy number abnormalities, and bone marrow microenviron- Multiple myeloma is a malignancy of the plasma cell charac- ment (4). Nevertheless, multiple myeloma is a multifaceted terized by proliferation of plasma cell clones (1). The proteasome disease related to genetic, epigenetic, and chromosomal altera- inhibitor bortezomib has shown promise in the treatment of tions, and the mechanisms underlying PI resistance in multiple multiple myeloma (2). However, its therapeutic activity was myeloma remain elusive. severely impeded by bortezomib resistance (3). Multifactorial The proteasome is a 26S enzyme complex that consists of a 19S regulatory complex and a core 20S catalytic complex. The 20S complex is composed of two rings of seven a subunits and two 1 Department of Cell Biology, School of Biology & Basic Medical Sciences, rings of seven b subunits (5). PIs could slightly inhibit the peptidyl Soochow University, Suzhou, China. 2Xiangyang No.1 People's Hospital, Hubei 3 glutamyl–like activity at the b1 subunit (PSMB6) and mainly University of Medicine, Xiangyang, China. Department of Biostatistics, Univer- b sity of Alabama at Birmingham, Birmingham, Alabama. 4Department of Bioin- target the chymotrypsin-like (ChT-L) activity of 5 subunit formatics, School of Biology & Basic Medical Sciences, Soochow University, (PSMB5; refs. 6, 7). The mutations in b5 subunit or the increased Suzhou, China. 5Department of Haematology, the Second Affiliated Hospital of expression of b5 subunit was detected in bortezomib-resistant Soochow University, Suzhou, China. 6Department of Molecular Biology, School hematologic tumor cell lines (8). However, the b5 mutations and 7 of Laboratory Medicine of Beihua University, Jilin, China. Department of Clinical overexpression are absent in clinical samples derived from borte- Examination Basis, Laboratory Academy of Jilin Medical College, Jilin, China. zomib-resistant patients (9, 10). Therefore, the b5 subunit dysre- Note: Supplementary data for this article are available at Clinical Cancer gulation may not be uniquely responsible for PI resistance. It is Research Online (http://clincancerres.aacrjournals.org/). conceivable that other mechanisms could be associated with PI H. Xu and H. Han contributed equally to this article. resistance. Corresponding Authors: Wenzhuo Zhuang, Soochow University, Ren Ai Road The bone marrow microenvironment activates many pathways 199, Suzhou 215123, China. Phone: 8651-2627-23369; Fax: 8651-2658-80103; leading to disease progression. MSCs support tumor cell growth, E-mail: [email protected]; and Bingzong Li, The Second Affiliated metastasis, and evasion of the immune system (11). The inter- Hospital of Soochow University, San Xiang Road 1055, Suzhou 215006, China. action between MSCs and myeloma cells performs a critical role in Phone: 8651-2677-84069; E-mail: [email protected] multiple myeloma pathogenesis, progress, and chemotherapy doi: 10.1158/1078-0432.CCR-18-2363 resistance (12, 13). These findings indicate a need for developing 2019 American Association for Cancer Research. novel drugs to counteract these cancer–stroma interactions.

www.aacrjournals.org OF1

Xu et al.

Translational Relevance Materials and Methods Patient samples PI resistance is a major challenge for multiple myeloma. Samples were taken from patients with newly diagnosed mul- The bone marrow microenvironment promotes myeloma cell tiple myeloma who were enrolled to the study. The study was survival and proliferation through the interaction between conducted in accordance with the Declaration of Helsinki. the MSCs and multiple myeloma cells. The bone marrow Informed written consent was obtained from each subject or each microenvironment promotes the interactions between the subject's guardian. The human investigations were performed MSCs and multiple myeloma cells that permit multiple mye- after approval by an institutional review board and in accordance loma to survive and proliferate. Here we discovered that with an assurance filed with. The detailed clinical features of these exosome-mediated transfer of PSMA3 (encodes proteasome patients are listed in Supplementary Table S1. Progression-free subunit a7) and lncPSMA3-AS1 from MSCs to multiple mye- survival (PFS) was defined as the time between initiation of loma cells contributed to PI resistance. PSMA3-AS1, which bortezomib therapy and the date of first evidence of progressive. arises from the antisense strand of PSMA3, was highly Patients who were progression-free at the time of analysis were expressed in myeloma cells (r-MM) and MSCs (r-MSCs) censored using the time between initial treatment and last follow- derived from bortezomib-resistant patients. As a pair of pro- up. More details of the patient samples are available in Supple- tein-coding/noncoding antisense transcripts, PSMA3 and mentary Data. PSMA3-AS1 were disordered concurrently and correlated positively in multiple myeloma cells, driving PI sensitivity in multiple myeloma. These results provided in vitro and in vivo Plasmid construction evidence that interference with exosomal RNAs served as a Full-length cDNA (PMSA3 or PSMA3-AS1) was synthesized Eco Bam promising approach to overcome PI resistance in multiple and subcloned into the RI and HI sites of Lenti-CMV-puro fi myeloma. Moreover, our data indicated that circulating exo- vector. The constructs were veri ed by DNA sequencing. somal PSMA3 and PSMA3-AS1 could develop the prognostic fi stratification of patients with multiple myeloma, in addition Exosome puri cation to the international staging system. Exosomes secreted by cultured cell lines and MSCs were isolated using ExoQuick solution (System Biosciences). Exosomes in the plasma were isolated using ExoQuick-LP solution (System Biosciences). The detailed procedures of exosome isolation are Exosomes act as key communicators between tumor microen- available in the Supplementary Data. The exosome pellet was vironment and cancer cells (14, 15). The interaction between isolated and the protein content of the exosome suspension was MSCs and multiple myeloma cells plays a crucial role in multiple analyzed by BCA Protein Assay Kit (Beyotime Biotechnology). myeloma pathogenesis and drug resistance by exosomes (16, 17). Recently, lncRNAs have been reported to exist in exosomes Dynamic light scattering analysis (18, 19). LncRNAs are transcribed from thousands of loci in Dynamic light scattering (DLS) analysis was performed to mammalian genomes and function in a wide range of biological characterize and quantify the particle size distribution of exo- processes (20, 21). In cancers, lncRNAs are emerging important somes (Zetasizer Nano S90). regulators in oncogenic pathways (22, 23). However, the role of lncRNAs in the multiple myeloma pathogenesis and progression In vivo xenograft studies has not been fully elucidated. In particular, how exosomal The animal experiments were performed after acquired per- lncRNAs derived from the bone marrow microenvironment con- mission from the local ethical committee according to the insti- tribute to PI resistance is unexplored in multiple myeloma. tutional guidelines for the use of laboratory animals. More details Abundant bidirectional transcription of eukaryotic genes has of in vivo xenograft studies are available in the Supplementary been explored across genomes (24). Emerging evidence suggests Data. that antisense lncRNAs play vital roles in regulating their sense transcript partners, in particular, affecting the expression of their Statistical analysis associated protein-coding genes (24, 25). However, we know little Survival analysis was performed to investigate the associations about the precise function of natural antisense transcripts and the between the censored outcomes and the clinical variables and the molecular mechanisms of their biogenesis in multiple myeloma. expression measures of PSMA3 and PSMA3-AS1 using the R This study explored how the MSC-derived exosomal lncRNAs package survival. PFS and OS were the primary outcomes of conferred PI resistance to multiple myeloma cells and through interest. The clinical stage (ISS) was treated as a three-level factor which molecular mechanisms they elicited this function. Our data and thus coded as two dummy variables (ISS2 and ISS3) to underscored that exosomal lncPSMA3-AS1 was the key determi- compare stage I with stages II and III. All other clinical variables nant in mediating resistance to PIs by regulating the stability of and the expression measures of PSMA3 and PSMA3-AS1 were PSMA3, which encodes proteasome subunit a7. Upregulation of numeric and were standardized to facilitate the comparison of PSMA3 led to the a-subunit–related increase in ChT-L protea- HRs. some activity. These combined approaches targeting PMSA3 and We first used univariate Cox survival regression to separately PSMA3-AS1 in conjunction with PIs could enhance therapeutic analyze the predictor variables. To illustrate the impact of the efficacy. Likewise, our findings suggest that the circulating exoso- PSMA3 and PSMA3-AS1, we partitioned the patients to two mal PSMA3 and PSMA3-AS1 from the plasma of patients with groups based on the medians of the PSMA3 and PSMA3-AS1 multiple myeloma could improve to identify the newly diagnosed expressions, and estimated the survival curves of these two groups patients with poor outcomes. of patients using the Kaplan–Meier method and compared their

OF2 Clin Cancer Res; 2019 Clinical Cancer Research

Exosomal PSMA3 and PSMA3-AS1 in PI Resistance

difference using the log-rank test. We then used multivariate Cox A volcano plotshowed that three transcripts(PSMA3: fold change ¼ survival regression to jointly fit all the clinical variables with either 1.29/P ¼ 0.0014, PSMA3-AS1: fold change ¼ 1.27/P ¼ 0.000595, PSMA3 or PSMA3-AS1, which adjusted for the potential con- USP13: fold change ¼ 1.257/P ¼ 0.0013) were significantly upre- founding effects of the clinical variables on PSMA3 or PSMA3- gulated abundances in bortezomib-resistant samples compared AS1. By fitting these Cox models, we obtained the estimates of with bortezomib-sensitive samples (Fig. 2B). Subsequently, we HRs, their 95% confidence intervals (CI), the corresponding P identified that PSMA3 and PSMA3-AS1, not USP13, were present values, and Harrell concordance (C-index). We also validated the in MSC-derived exosomes (Fig. 2C). Then we focused on PSMA3 prognostic values of these models using the leave-one-out cross- and PSMA3-AS1 transcripts. PSMA3 and PSMA3-AS1 expression validation. were higher in bortezomib-resistant samples compared with bor- Pearson correlation analysis was employed to determine the tezomib-sensitive samples (Fig. 2D). In addition, Kaplan–Meier þ correlation between the expressions of PSMA3 or PSMA3-AS1. analysis showed that high PSMA3 expression in CD138 multiple Student t test was performed to analyze the difference of mean myeloma cells was associated with decreased PFS (P ¼ 0.0307) and values between two groups for other continuous outcomes. poor OS (P ¼ 0.0328; Fig. 2E). Cox proportional hazards regression Other detailed assays are available in Supplementary Data. analysis further confirmed that high PSMA3 could act as an independent prognostic factor for patients with multiple myeloma with bortezomib therapy in a multivariate analysis (HR ¼ 1.3104; 95% Results CI ¼ 1.1113–1.545; P ¼ 0.0013). MSC-derived exosomes were transferred to myeloma cells and To investigate whether the PSMA3 content of multiple myelo- conferred proteasome inhibitor resistance to myeloma cells ma was involved in facilitating multiple myeloma progression, we We first examined whether MSCs secreted exosomes. Primary analyzed the datasets obtained from the Oncomine database þ r-MSCs obtained from bortezomib-resistant patients, or s-MSCs to determine the PSMA3 expression alterations in CD138 cells obtained from bortezomib-sensitive patients, were cultured in in monoclonal gammopathy of undetermined significance serum-free medium for 48 hours. Then we isolated exosomes (MGUS), smoldering myeloma, multiple myeloma, and plasma using standard methods and confirmed their identity by trans- cell leukemia (PCL). As shown in Fig. 2F, PSMA3 levels were þ mission electron microscope (Fig. 1A). DLS analysis was per- increased in CD138 cells from smoldering myeloma compared formed to quantify the exosome size distribution (Fig. 1B). West- with those from MGUS (26), and the PSMA3 expressions were þ ern blot analysis was used to confirm the expression of Flotillin-1 upregulated in CD138 cells from multiple myeloma compared and HSP70, which were generally used as exosome markers with those from MGUS (Fig. 2G; ref. 27). Moreover, the PSMA3 (Fig. 1C). levels showed a progressive increase in MGUS, multiple myeloma, Next we investigated whether MSC-derived exosomes could be and PCL (Fig. 2H; ref. 28). These analyses consistently suggested þ transmitted into myeloma cells. Exosomes derived from MSCs that the expression levels of PSMA3 in CD138 cells were corre- were fluorescently labeled with PKH67, and then cultured with lated with multiple myeloma progression. MM.1S or U266. The ability of myeloma cells to take MSC-derived Consistent with this previously published study, our clinical exosomes was confirmed using confocal microscope (Fig. 1D; data showed that the mRNA levels of PSMA3 and PSMA3-AS1 þ Supplementary Fig. S1A) and flow cytometric assays (Fig. 1E; were upregulated in CD138 multiple myeloma cells derived Supplementary Fig. S1B). As the recipient cells, myeloma cells from bortezomib-resistant patients relative to those from borte- showed the same uptake efficiency of exosomes from r-MSCs and zomib-sensitive patients (Fig. 2I). To further explore the mechan- those from s-MSCs (Fig. 1F). These results suggested that exo- isms of PI resistance, multiple myeloma cell lines with a 4-fold or somes derived from MSCs could be transmitted to myeloma cells, greater resistance to bortezomib, carfilzomib, or ixazomib were indicating a promising role in regulating biologic functions of developed. The PI-resistant (bortezomib resistance, BR; carfilzo- myeloma cells. mib resistance, CR; ixazomib resistance, IR) models consistently To further investigate the effect of MSC-derived exosomes on displayed upregulation of PSMA3 and PSMA3-AS1 transcript the PI resistance, U266 or MM.1S cells were treated with or levels (Fig. 2J; Supplementary Fig. S2). In addition, r-MSCs had without different amounts of exosomes in the presence of borte- increased expression of PSMA3 and PSMA3-AS1 compared with zomib or carfilzomib. R-MSC-derived exosomes significantly s-MSCs (Fig. 2K). Moreover, the expressions of PSMA3 and reduced PI sensitivity (Fig. 1G; Supplementary Fig. S1C). How- PSMA3-AS1 in MSCs were positively correlated with those in þ ever, no difference in the antiproliferative activities of PIs was CD138 myeloma cells (Fig. 2L). These results suggested that high observed between treatment with or without different amounts of levels of PSMA3 and PSMA3-AS1 were associated with PI resis- exosomes obtained from s-MSCs. These data demonstrated that tance of myeloma cells. exosomes derived from r-MSCs, but not from s-MSCs, reduced the PI sensitivity in MM cells. Exosome-mediated transfer of PSMA3 and lncPSMA3-AS1 from MSCs to multiple myeloma cells contributed to PI Identifying mRNAs involved in bortezomib-resistant patients resistance To explore the possible mechanisms of PI resistance in multiple Next we investigated whether exosome-transferred PSMA3 myeloma, gene expression profiling of bortezomib-resistant andPSMA3-AS1couldconferproteasomeinhibitorsresistance patients (NC/PD) and bortezomib-sensitive patients (CR/PR/ to multiple myeloma cells. We found that PSMA3 and PSMA3- MR) were analyzed using GEO database (GSE9782). GSEA AS1 levels statistically increased in multiple myeloma cells showed that the positive regulation of proteasome gene signatures incubated with r-MSCs–derived exosomes. This change did not (the gene sets: BIOCARTA_PROTEASOME_PATHWAY, WONG_ occur when multiple myeloma cells were incubated with PROTEASOME_GENE_MODULE, KEGG_PROTEASOME) were s-MSC–derived exosomes (Fig. 3A; Supplementary Fig. S3A). significantly enriched in bortezomib-resistant samples (Fig. 2A). Moreover, U266 or MM.1S incubated with exosomes from

www.aacrjournals.org Clin Cancer Res; 2019 OF3

Xu et al.

Figure 1. MSC-derived exosomes were transferred to myeloma cells and conferred PI resistance to myeloma cells. A, Representative transmission electron microscopy imaging of exosomes derived from r-MSCs and s-MSCs (scale bar, 100 nm). B, Exosomes isolated from r-MSCs and s-MSCs were measured by DLS. C, Representative Western blot analysis of HSP70 and flotillin-1 proteins in r-MSC–derived exosomes and s-MSC–derived exosomes. D, Myeloma cells were cultured in the presence of MSC-derived PKH67-labeled exosomes for 24 hours. Exosome uptakes by myeloma cells were shown using a confocal microscope (original magnification, 400). Myeloma cells were stained using DAPI (nuclei) and 594-conjugated anti-actin antibody. Scale bars, 10 mm. E, Flow cytometric analysis of U266 after incubation with fluorescently labeled exosomes. FL1 fluorescence indicates exosome uptake. F, Flow cytometric analysis of myeloma cells after incubation with fluorescently labeled exosomes from r-MSCs and s-MSCs for the indicated time. G, U266 and MM.1S cells were planted in 96-well plates with or without different amounts of exosomes (5 mg, 20 mg, 40 mg) and bortezomib or carfilzomib for 72 hours. The cell viability was measured using CCK-8. Error bars, mean SD of three independent experiments (, P < 0.01).

r-MSCs but not those from s-MSCs, showed an increase in the To further evaluate the possibility that exosome-mediated ChT-L activity of the proteasome (Supplementary Fig. S3B). transmission of PSMA3 and PSMA3-AS1 from MSCs to multiple Treatment directly with r-MSC–derived exosomes or s-MSC– myeloma cells contributed to PI resistance, we examined whether derived exosomes had no effects on cell viability and apoptosis the effect could be reversed when PSMA3 or PSMA3-AS1 expres- (Supplementary Fig. S3C and S3D). sion was downregulated in r-MSCs. As expected, the levels of

OF4 Clin Cancer Res; 2019 Clinical Cancer Research

Exosomal PSMA3 and PSMA3-AS1 in PI Resistance

www.aacrjournals.org Clin Cancer Res; 2019 OF5

Xu et al.

PSMA3 or PSMA3-AS1 were decreased in multiple myeloma cells results further underscored the involvement of PSMA3 and upon incubation with exosomes from PSMA3 or PSMA3-AS1- PSMA3-AS1 overexpression in conferring PI resistance and pro- knockdown r-MSCs compared with siRNA control cells (Fig. 3B; vided a targeted strategy to reverse PI resistance. Supplementary Fig. S3E). Moreover, U266 or MM.1S incubated with exosomes from r-MSCs displayed decreased PIs sensitivity, PSMA3-AS1 reduced PI sensitivity by regulating PSMA3 which could be abolished by downregulating PSMA3 or PSMA3- expression in multiple myeloma cells AS1 in r-MSCs (Fig. 3C; Supplementary Fig. S3F). Consistent with To identify lncRNA PSMA3-AS1 involved in PI resistance in the loss-of-function experiments, when PSMA3 or PSMA3-AS1 multiple myeloma, we characterized the PSMA3-AS1. RNA FISH was upregulated in s-MSCs, levels of PSMA3 or PSMA3-AS1 were and nuclear and cytoplasmic fractions assays indicated that increased in multiple myeloma cells incubated with exosomes PSMA3-AS1 existed both in the nucleus and in the cytoplasm from PSMA3 or PSMA3-AS1–overexpressing s-MSCs compared (Fig. 4A and B). Rapid amplification of cDNA ends (RACE) was with vector control cells (Fig. 3B; Supplementary Fig. S3E). In performed for directional sequencing of 50 and 30 ends and two concordance with upregulation of PSMA3 and PSMA3-AS1, cells splice variants, 2531 nt and 2440 nt were identified (Fig. 4C). We showed decreased PI sensitivity (Fig. 3C; Supplementary Fig. S3F). found the presence of a poly-A tail, indicating that PSMA3-AS1 is a Similar results were obtained when directly knocking down fully processed transcript of RNA polymerase II. Coding Potential PSMA3 or PSMA3-AS1 in U266 or MM.1S (Fig. 3D; Supplemen- Calculator (CPC, http://cpc.cbi.pku.edu.cn/; ref. 30) and Coding tary Fig. S3G). These data suggested that PSMA3- or PSMA3-AS1– Potential Assessment Tool (CPAT, http://lilab.research.bcm.edu/ level differences in MSC-derived exosome content could mediate cpat/; ref. 31) analyses displayed that PSMA3-AS1 has no coding PI resistance in multiple myeloma cells. potentiality (Supplementary Tables S4 and S5). Northern blotting Previous study has showed that PSMA3, which encodes the assay further confirmed the existence of full-length PSMA3-AS1 in constitutive proteasome component, a7 subunit, performed an MM.1S (Fig. 4D). important role in proteasome formation and function (29). To It is reported that the antisense transcripts can commonly further determine the role of PSMA3 and PSMA3-AS1 in PI regulate the sense transcripts expressions in two ways (32). In a sensitivity in multiple myeloma cells, we examine whether concordant way, the antisense transcripts have positive effects on PSMA3 and PSMA3-AS1 levels were associated with proteasome sense transcripts (33). Conversely, the level of the sense RNA, or activity. As expected, knockdown of PSMA3 or PSMA3-AS1 caused corresponding protein level is negatively regulated by the anti- a decrease in the ChT-L activity of the proteasome in U266 or sense transcripts in discordant way (34). To explore the biology of MM.1S, whereas upregulated PSMA3 or PSMA3-AS1 enhanced PSMA3-AS1 and consider its potential significance, the correla- the ChT-L proteasome activity (Fig. 3E; Supplementary Fig. S3H). tion between PSMA3-AS1 and PSMA3 was examined in multiple These findings demonstrated that knockdown of either PSMA3 or myeloma cells derived from 57 patients with multiple myeloma PSMA3-AS1 increased the PI sensitivity. Consistent with this by Pearson correlation analysis. The results showed that PSMA3- finding, overexpression of either PSMA3 or PSMA3-AS1 reduced AS1 was positively correlated with PSMA3 in multiple myeloma the PI sensitivity. Taken together, these data indicate that the MSC- cells (Fig. 4E). derived exosomal PSMA3 and PSMA3-AS1 conferred PI resistance To examine whether the PSMA3-AS1 transcript regulated the to the multiple myeloma cells. Higher levels of PSMA3 and PSMA3 expression, we transiently transfected siRNA PSMA3-AS1 PSMA3-AS1 were less sensitive to the PIs, which produced a to U266 or MM.1S. We found that knockdown PSMA3-AS1 greater increase in cell proliferation and ChT-L proteasome significantly decreased PSMA3 level in U266 or MM.1S activity. (Fig. 4F; Supplementary Fig. S4A). Next, we determined whether downregulation of PSMA3 and Given that PSMA3-AS1 knockdown decreased the proteasome PSMA3-AS1 would lead to restoration of PI sensitivity in PI- activity and subsequently reduced the cell viability (Fig. 3D and E; resistant myeloma cells. Either PSMA3-siRNA–dependent preven- Supplementary Fig. S3G and S3F), we sought to determine tion or PSMA3-AS1-siRNA–dependent prevention after PI expo- whether reconstituting the expression of PSMA3 in PSMA3- sure was accompanied by a significantly lower cell proliferation AS1–downregulated cells could reverse the altered biological (Fig. 3G), higher fraction of apoptotic cells (Fig. 3H), as well as activity. We found that restoring the expression of PSMA3 in decreased ChT-L proteasome activity (Fig. 3F). Collectively, these PSMA3-AS1 knockdown cells significantly abrogated the reduced

Figure 2. Identifying mRNAs involved in bortezomib (BTZ)-resistant patients. A, Speciﬁc baseline gene expression signatures are associated with proteasome activity. Representative GSEA plots illustrating the enrichment of genes involved in positive regulation of proteasome among the transcripts in bortezomib-resistant patients. FDR, false discovery rate; NES, normalized enrichment score. B, Volcano plot showing the gene expression changes in the proteasome activity gene sets between bortezomib-resistant samples and bortezomib-sensitive samples. PSMA3, PSMA3-AS1, and USP13, which had a signiﬁcantly altered expression,

are highlighted in red. Vertical and horizontal dashed lines indicate cutoff values (P ¼ 0.05 and absolute logarithmic fold change of 1.25). C, 1/Ct values from qPCR analysis to determine the loading efﬁciency of transcripts. The mean 1/Ct and SD of the three independent experiments are shown. D, Analysis of PSMA3 and PSMA3-AS1 expression in bortezomib-resistant multiple myeloma (NC/PD) and bortezomib-sensitive multiple myeloma (CR/PR/MR) using the GEO database (GSE9782). E, Kaplan–Meier analysis of PFS and OS in the high and low PSMA3 groups according to the median PSMA3 level in CD138þ multiple myeloma cells. F–H, PSMA3 mRNA expression in CD138þ cells from MGUS, smoldering myeloma (SM), multiple myeloma (MM), or PCL was analyzed by using

Oncomine database. Data were presented as log2 median-centered intensity. I, Analysis of PSMA3 and PSMA3-AS1 expression in bortezomib-resistant multiple myeloma (SC/PD) and bortezomib-sensitive multiple myeloma (sCR/CR/VGPR/PR/MR) using our clinical data. RR, relapsed/refractory patients; ND, newly diagnosed patients. J, PSMA3 and PSMA3-AS1 expression were determined in U266 drug-na€ve, bortezomib-resistant (B20R), carﬁlzomib-resistant (C40R), ixazomib-resistant (I20R) cells using qPCR. PI-resistant cells had increased expression of PSMA3 and PSMA3-AS. K, Analysis of PSMA3 and PSMA3-AS1 expression in r-MSCs and s-MSCs. L, Pearson correlation analysis between PSMA3 and PSMA3-AS1 levels in CD138þ myeloma cells and that in MSCs. Error bars, mean SD of three independent experiments (, P < 0.05; , P < 0.01; , P < 0.001).

OF6 Clin Cancer Res; 2019 Clinical Cancer Research

Exosomal PSMA3 and PSMA3-AS1 in PI Resistance

Figure 3. Exosome-mediated transfer of PSMA3 and PSMA3-AS1 from MSCs to multiple myeloma cells contributed to PI resistance. A, qPCR analysis of PSMA3 and PSMA3-AS1 expression in U266 after incubation with indicated exosomes. R-exo, R-MSC–derived exosomes; S-exo, S-MSC–derived exosomes. B, qPCR analysis of PSMA3 in U266 after incubation with indicated exosomes. oePSMA3, PSMA3-overexpressing; oePSMA3-AS1, PSMA3-AS1–overexpressing; oeVec, empty vector. C, MSCs were transfected with indicated siRNAs or overexpressing vectors for 24 hours, and exosomes were isolated. These exosomes were then treated with RNase to remove unincorporated RNAs. Cell viability of U266 was assessed 48 hours after incubation with or without indicated exosomes and bortezomib (BTZ) using CCK-8 assay. D, U266 cells were transfected with PSMA3 siRNA, PSMA3-AS1 siRNA, control siRNA, PSMA3-overexpressing vector, PSMA3-AS1– overexpressing vector or empty vector and treated with or without bortezomib. The cell viability was measured using CCK-8. E, U266 cells were transfected with PSMA3 siRNA, PSMA3-AS1 siRNA, control siRNA, PSMA3-overexpressing vector, PSMA3-AS1–overexpressing vector or empty vector. The cell numbers were counted and a proteasome activity assay was done. Fold changes of activity against no-treatment control was calculated. F–H, After transfection with PSMA3 siRNA, PSMA3-AS1 siRNA, control siRNA, PI-resistant cells were incubated (24–48 hours) with bortezomib, carﬁlzomib (CFZ), or ixazomib, respectively. The proteasome activity (F) and cell viability (G) were measured as described previously. The apoptosis was measured by Annexin V/7-AAD apoptosis detection kit (H). Error bars represent the mean SD of three independent experiments (, P < 0.05; , P < 0.01; , P < 0.001; , P < 0.0001).

www.aacrjournals.org Clin Cancer Res; 2019 OF7

Xu et al.

Figure 4. PSMA3-AS1 reduced PI sensitivity by regulating PSMA3 expression in multiple myeloma cells. Characterization of lncRNA PSMA3-AS1. The nuclear and cytoplasmic location of PSMA3-AS1 in MSCs. A, RNA FISH assay of PSMA3-AS1 in U266, MM.1S, and MSCs. Scale bars, 10 mm. B, qPCR detection of PSMA3-AS1 in the cytoplasmic (C) and nuclear (N) fractions. GAPDH served as an internal normalization control. C, Gel electrophoresis of nested PCR products from 50-RACE and 30-RACE. D, Northern blot analysis with in vitro–transcribed strand-speciﬁc RNA probes for PSMA3-AS1 in MM.1S. b-Actin serves as a loading control. E, Correlation between PSMA3 and PSMA3-AS1 in CD138þ myeloma cells from myeloma patients. r ¼ 0.538, P < 0.001. F, U266 or MM.1S was transiently transfected with PSMA3-AS1siRNA or a control siRNA. qPCR was performed to determine gene levels, with GAPDH used as an internal normalization control. PSMA3 protein expression was determined by Western blot analysis. G, U266 cells were transfected with PSMA3-AS1 siRNA, control siRNA with or without PSMA3-overexpressing vector or empty vector. The cell numbers were counted and the proteasome activity assay was done. Fold changes of activity against no-treatment control were calculated. H, U266 cells were treated with or without bortezomib and transfected with PSMA3-AS1 siRNA or control siRNA, with or without PSMA3-overexpressing vector or empty vector. The cell viability was measured using CCK-8. Error bars represent the mean SD of three independent experiments (, P < 0.05; , P < 0.01; , P < 0.001; , P < 0.0001).

ChT-L activity and the increased PI sensitivity conferred by in opposite directions. PSMA3-AS1 overlaps the 2029 nucleotides PSMA3-AS1 downregulation in multiple myeloma cells within the 7th intron of PSMA3 (Supplementary Fig. S5A). To (Fig. 4G and H; Supplementary Fig. S4B and S4C). These data verify the existence of RNA duplex structure between PSMA3-AS1 demonstrated that PSMA3-AS1 had effects on proteasome activity and pre-PSMA3, RNase protection assay was performed on RNAs and PI sensitivity by regulating PSMA3 expression in multiple from U266 and MM.1S. RNase A digested the single-stranded myeloma cells. RNA and the remaining double-stranded RNA was protected from degradation. According to the PCR assays using indicated primers, PSMA3-AS1 formed an RNA duplex with PSMA3-AS1 pre-mRNA the overlapping part of both transcripts existed, which suggested and increased its stability that PSMA3-AS1 and pre-PSMA3 indeed form a RNA duplex Antisense transcripts commonly interact with their sense tran- (Supplementary Fig. S5B). RNA–RNA duplex could change the scripts in a double-stranded RNA structures form, which may secondary or tertiary structure of RNA and protect RNA from potentially regulate the transport, splicing, and stability of the RNase degradation, thereby increase its stability (33, 35). To sense transcripts (24). PSMA3 and PSMA3-AS1 genes are over- investigate whether the duplex form between PSMA3-AS1 and lapped and located at chromosome 14p23.1. They are transcribed pre-PSMA3 affected the PSMA3 mRNA stability, we assessed the

OF8 Clin Cancer Res; 2019 Clinical Cancer Research

Exosomal PSMA3 and PSMA3-AS1 in PI Resistance

stability of PSMA3 transcripts by blocking new RNA synthesis and ISS with PSMA3-AS1 to show the predictive value of the with a-amanitin (an RNA polymerase II inhibitor) and measuring PSMA3 and PSMA3-AS1 expressions (Supplementary Table S7). the loss of PSMA3, PSMA3-AS1, GAPDH, and 18s RNA over a 24- To validate the prognostic values of PSMA3 and PSMA3-AS1, the hour period (Supplementary Fig. S5C). The results showed that leave-one-out cross-validation was performed (Supplementary knocked down PSMA3-AS1 decreased the stability of PSMA3. Table S7). The results showed that both exosomal PSMA3 and Conversely, PSMA3-AS1 overexpression increased the stability of PSMA3-AS1 were significant predictive biomarkers on PFS and PSMA3 (Supplementary Fig. S5D). Together, these results dem- OS. These data indicated that the exosomal PSMA3 or PSMA3-AS1 onstrated that PSMA3-AS1 formed an RNA duplex with PSMA3- could be crucial in identifying worse prognosis in patients with AS1 pre-mRNA, which led to enhanced stability of PSMA3 by newly diagnosed multiple myeloma. Therefore, it is advisable to reducing its decay. evaluate PSMA3 and PSMA3-AS1 expression in multiple myeloma to identify patients who might benefit from bortezomib Targeting PSMA3-AS1 increased PI sensitivity in vivo therapy before deciding on a course of treatment. To evaluate the therapeutic potential of PSMA3-AS1 in multiple myeloma in vivo, we established bioluminescent multiple myeloma models (U266-luc), which recapitulates the bone patho- Discussion physiology and the clinical sequelae observed in patients with Resistance to the PIs is a major clinical problem in multiple myeloma (ref. 36; Fig. 5A). The mRNA levels of PSMA3-AS1 and myeloma and the mechanisms have not been completely þ PSMA3 in the mouse CD138 cells were performed by qPCR to revealed. The bone marrow microenvironment could promote confirm the injected siRNA efficiency in vivo (Fig. 5C). Biolumi- the survival and drug resistance of myeloma cells. It has been nescent imaging revealed that therapeutic PSMA3-AS1 siRNA proposed that MSCs contribute to multiple myeloma microen- efficiently increased the sensitivity of U266 xenografts to carfil- vironment where they can promote or suppress multiple myelo- zomib treatment (Fig. 5B). Moreover, the combination treatment ma progression (11, 39). We aimed to find a way to interfere these with PSMA3-AS1 siRNA and carfilzomib significantly prolonged networks to stop tumor growth and drug resistance. the overall survival compared with carfilzomib combination with Many types of cells could release exosomes (70–120 nm) into siRNA control or siPSMA3-AS1–treated alone (Fig. 5D). No the extracellular environment (38, 40). Exosomes are involved in significant weight loss or unexpected deaths was found during intercellular communication, and in this study we investigated treatment. Collectively, downregulation of PSMA3-AS1 in vivo how the transfer of exosomal PMSA3 and PSMA3-AS1 from MSCs significantly enhanced the sensitivity of multiple myeloma xeno- to multiple myeloma cells affected PI resistance (Fig. 6H). grafts to carfilzomib. These data revealed that PSMA3-AS1 was PI resistance is a vital impediment to the clinical treatment. essential for PI resistance in multiple myeloma cells. Numerous underlying mechanisms have been identified (41–43), but the primary cause is still unknown. Initial studies in hema- Prognostic role of circulating exosomal PSMA3 and PSMA3-AS1 tologic tumor cell line models demonstrated that overexpression in multiple myeloma of the b5 proteasome subunit or mutations in the b5 subunit's Several cell types, including cancer cells, secrete exosomes. bortezomib-binding pocket were implicated in bortezomib resis- Exosomes have been reported to be isolated from the peripheral tance (8, 44). However, the b5 mutations and overexpression are blood and act as powerful markers for disease progression in absent in clinical specimen obtained from bortezomib-resistant cancer (37, 38). To determine the prognostic significance of patients (9, 10). To further elucidate mechanisms of PI resistance, circulating exosomal PSMA3 and PSMA3-AS1 in multiple mye- we retrieved a database containing gene expression profile of 169 loma, we isolated and characterized exosomes from the plasma of myeloma cases with clinical response and disease prognosis (45). patients with multiple myeloma. A cohort of 57 patients with The analysis of this dataset showed that the mRNA levels of þ newly diagnosed multiple myeloma was uniformly treated and PSMA3 and PSMA3-AS1 in CD138 cells are upregulated in followed. qPCR assays were performed to analyze circulating bortezomib-resistant patients. Moreover, increased PSMA3 exosomal RNAs derived from the plasma of patients with multiple and PSMA3-AS1 mRNA expression correlated with proteasome myeloma. inhibitors resistance and poor outcome for patients with myelo- First, we isolated plasma-circulating exosomes from patients ma treated with bortezomib. Further analyses of Oncomine with multiple myeloma and confirmed them by transmission data showed that the PSMA3 levels appeared a progressive electron microscope (Fig. 6A) and dynamic light scattering assays increase in MGUS, smoldering myeloma, multiple myeloma, and (Fig. 6B). Similar numbers of exosomes were isolated from PCL (26, 28). Similarly, our PI-resistant models (U266BR, bortezomib-resistant patients and bortezomib-sensitive patients U266CR, U266IR, MM.1SBR, MM.1SCR, MM.1SIR) consistently (Fig. 6C). By comparing PSMA3 and PSMA3-AS1 levels in circu- displayed upregulation of PSMA3 and PSMA3-AS1 expression. lating exosomes and the corresponding exosome-depleted plas- These results suggested that the PI insensitivity in multiple mye- ma, we found that circulating PSMA3 and PSMA3-AS1 predom- loma was due to upregulated PSMA3 and PSMA3-AS1. inantly existed in exosomes (Fig. 6D). Moreover, expressions of LncRNAs are extensively involved in cellular processes, includ- PSMA3 and PSMA3-AS1 in circulating exosomes were positively ing cell cycle, apoptosis, and immunity (23, 46). LncRNAs fulfill þ correlated with that in CD138 myeloma cells (Fig. 6E). their functions through multiple mechanisms, such as scaffolding The results from the univariate and multivariate Cox survival of nuclear or cytoplasmic complexes, regulation of gene expres- regressions were present in Supplementary Table S6. We can see sion, and pairing with DNA or RNAs. Recently, several protein- that exosomal PSMA3 and PSMA3-AS1 were significantly associ- coding mRNAs have been reported to have natural antisense ated with both PFS and OS in the univariate and multivariate transcript partners, most of which are noncoding RNAs (20, 47). analyses (Fig. 6F and G; Supplementary Table S6). We compared Natural antisense lncRNA transcripts could serve as therapeutic the C-indices of the models fitting the ISS alone, ISS with PSMA3, targets through regulating the sense gene expression (25, 48). In

www.aacrjournals.org Clin Cancer Res; 2019 OF9

Xu et al.

Figure 5. Targeting PSMA3-AS1 increased PI sensitivity in vivo. A, NSG mice were transplanted with U266-lucþ cells, and the combined chemotherapy regimens were started on day 15 after xenotransplantation as indicated in the top panel. B, Representative whole-body bioluminescence images of NSG mice. Therapy of carﬁlzomib in combination with PSMA3-AS1 siRNA or siRNA control or PSMA3-AS1 siRNA alone in vivo was performed. Tumor development was monitored by noninvasive bioluminescence imaging. (n ¼ 5/group). Error bars represent the mean SD of three independent experiments (, P < 0.05; , P < 0.01). C, qPCR was performed to determine gene levels in mouse CD138þ cells, with GAPDH used as an internal normalization control. Error bars represent the mean SD of three independent experiments (, P < 0.01; , P < 0.001 versus vehicle control). D, Survival curves (Kaplan–Meier) of mice showed prolongation of overall survival with indicated treatments (log-rank test, P < 0.01).

OF10 Clin Cancer Res; 2019 Clinical Cancer Research

Exosomal PSMA3 and PSMA3-AS1 in PI Resistance

Figure 6. Exosomal PSMA3 and PSMA3-AS1 levels in plasma and CD138þ multiple myeloma cells correlate with bortezomib (BTZ) response in patients with multiple myeloma. A and B, Characterization of circulating exosomes from the plasma of human subjects. A, Representative TEM imaging of exosomes from the plasma of human subjects (scale bar ¼ 100 nm). B, Exosomes from the plasma of human subjects were measured by DLS. C, The number of exosomes isolated from bortezomib-resistant patients and bortezomib-sensitive patients was determined. SP, bortezomib-sensitive patients; RP, bortezomib-resistant patients. D, Circulating PSMA3 and PSMA3-AS1 predominantly exist in exosomes. Exosomes isolated from three cases of multiple myeloma patient plasma as well as the corresponding exosome-depleted plasma were subjected to RNA extraction and qRT-PCR. The ratio of mRNAs level in the exosomes to those in exosome-free fraction is shown. Relative gene expression values in the same volume of plasma were normalized to exogenous reference l polyA. E, Pearson correlation analysis between PSMA3 and PSMA3-AS1 levels in CD138þ myeloma cells and that in circulating exosomes of patients with multiple myeloma (n ¼ 57). Relative gene expression values in the same volume of plasma were normalized to exogenous reference l polyA. Results are presented as mean SD. F, The prognostic impact of exosomal PSMA3-AS1 on overall progression-free survival (PFS) and survival (OS). G, The prognostic impact of exosomal PSMA3 on overall progression-free survival (PFS) and survival (OS). H, A schematic diagram depicting the mechanism of transferring of PSMA3 and PSMA3-AS1 by exosomes conferring PI resistance to MSCs.

www.aacrjournals.org Clin Cancer Res; 2019 OF11

Xu et al.

this study, we identified a natural antisense transcript, PSMA3- PSMA3 and PSMA3-AS1 could serve as prognostic factors for AS1, arises from intergenic regions of the PSMA3 gene. Our data multiple myeloma in a multivariate Cox regression analysis. Our indicated that as a pair of sense protein-coding/antisense non- study provided evidence of the association between circulating coding transcripts, PSMA3-AS1 could form a duplex structure with exosomal lncRNAs and outcomes in patients with newly diag- PSMA3 pre-mRNA, thereby increasing its stability by reducing its nosed multiple myeloma. decay, and by enhancing the stability of PSMA3 pre-mRNA, Here, we revealed that the informative lncRNAs derived from PSMA3-AS1 upregulated the PSMA3 expression. In addition, MSC exosomes involved in the communication between the downregulation of PSMA3-AS1 decreased PSMA3 expression, and MSCs and myeloma cells thus resulted in PI resistance in multiple subsequently reducing the ChT-L proteasome activity and increas- myeloma. Our findings confirmed that PI resistance was conferred ing the PI sensitivity in myeloma cells. Importantly, therapeutic by increased expression of PMSA3 and PSMA3-AS1. The data also lncPSMA3-AS1 siRNA efficiently increased the sensitivity of U266 suggested that low expression of PSMA3 and PSMA3-AS1 in xenografts to carfilzomib treatment, suggesting the significance of pretherapy multiple myeloma exhibited a significant improve- lncPSMA3-AS1 in multiple myeloma. Collectively, our data ment in prognosis after receiving bortezomib, whereas patients showed that disturbance of an antisense RNA PSMA3-AS1 could with high PSMA3 and PSMA3-AS1 expressions were correlated alter the expression of sense mRNA PSMA3, suggesting that with a poor bortezomib response. antisense transcript PSMA3-AS1 contributed to the regulation of PSMA3 in multiple myeloma cells. These results suggested that Disclosure of Potential Conflicts of Interest many lncRNAs might exert their functions by RNA–RNA direct No potential conflicts of interest were disclosed. interactions and that widely studying these interactions may further explore lncRNA function and mechanism. Authors' Contributions Exosomal RNAs have been reported to contribute to some Conception and design: H. Han, S. Song, Wenzhuo Zhuang aspects of oncology, including tumorigenesis, metastasis, and Development of methodology: H. Han chemoresistance (15, 38). Recent evidence suggests that more Acquisition of data (provided animals, acquired and managed patients, than 20% of exosomal RNAs are lncRNAs (49). Our in vitro provided facilities, etc.): H. Xu, H. Han, Y. Hong, Z. Li, B. Li Analysis and interpretation of data (e.g., statistical analysis, biostatistics, data indicated that lncPSMA3-AS1 could be packaged into computational analysis): H. Han, N. Yi, C. Qian, Y. Qiu, W. Zhou exosomes and transmitted to myeloma cells, conferring PI Writing, review, and/or revision of the manuscript: H. Han, N. Yi, Wenyue resistance to multiple myeloma cells by increasing the PSMA3 Zhuang, Wenzhuo Zhuang expression. Administrative, technical, or material support (i.e., reporting or organizing Circulating exosomal RNAs have been implicated to act as data, constructing databases): C. Qian, Y. Qiu, W. Zhou potential noninvasive molecular predictors of prognosis and drug Study supervision: B. Li, Wenzhuo Zhuang response (50). However, the discovery of specific exosomal RNAs Acknowledgments as disease-specific biomarkers depends on validated methods for fi This study was supported by Natural Science Foundation of Jiangsu sample isolation and analyses. Here, we identi ed circulating Province China (BK20161218, BK20161223), the Science and Technology exosomal PSMA3 and PSMA3-AS1 as clinically relevant biomar- Development Project of Suzhou City (SS201856), National Natural Science kers using standardized techniques for sample collection and Foundation of China (81670191, 81673448), The Applied Basic Research exosomal RNA measurement. Another challenge is that RNA copy Programs of Suzhou City (SYS201546), and Natural Science Foundation of numbers are generally low in circulating exosomes, which makes Jilin Province (20160101234JC). it difficult to obtain sufficient RNA for analyses (50). In this study, we did observe that PSMA3 and PSMA3-AS1 were relatively The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked enriched in exosomes derived from the plasma of patients with advertisement fi in accordance with 18 U.S.C. Section 1734 solely to indicate multiple myeloma by qPCR. Collectively, we identi ed that this fact. circulating exosomal PSMA3 and PSMA3-AS1 derived from the plasma of patients with multiple myeloma were associated with Received July 23, 2018; revised October 16, 2018; accepted January 2, 2019; poor outcomes with regard to PFS and OS. The levels of exosomal published first January 4, 2019.

References 1. Kyle RA, Gertz MA, Witzig TE, Lust JA, Lacy MQ, Dispenzieri A, et al. Review 6. Kisselev AF, Callard A, Goldberg AL. Importance of the different proteolytic of 1027 patients with newly diagnosed multiple myeloma. Mayo Clin Proc sites of the proteasome and the efficacy of inhibitors varies with the protein 2003;78:21–33. substrate. J Biol Chem 2006;281:8582–90. 2. Hideshima T, Chauhan D, Richardson P, Anderson KC. Identification and 7. CrawfordLJ,WalkerB,OvaaH,ChauhanD,AndersonKC,MorrisTC, validation of novel therapeutic targets for multiple myeloma. J Clin Oncol et al. Comparative selectivity and specificity of the proteasome inhi- 2005;23:6345–50. bitors BzLLLCOCHO, PS-341, and MG-132. Cancer Res 2006;66: 3. Ahn JS, Jung SH, Yang DH, Bae SY, Kim YK, Kim HJ, et al. Patterns of relapse 6379–86. or progression after bortezomib-based salvage therapy in patients with 8. Oerlemans R, Franke NE, Assaraf YG, Cloos J, van Zantwijk I, Berkers CR, relapsed/refractory multiple myeloma. Clin Lymphoma Myeloma Leuk et al. Molecular basis of bortezomib resistance: proteasome subunit beta5 2014;14:389–94. (PSMB5) gene mutation and overexpression of PSMB5 protein. Blood 4. Nikesitch N, Ling SC. Molecular mechanisms in multiple myeloma drug 2008;112:2489–99. resistance. J Clin Pathol 2016;69:97–101. 9. Franke NE, Niewerth D, Assaraf YG, van Meerloo J, Vojtekova K, van 5. Chen P, Hochstrasser M. Autocatalytic subunit processing couples active Zantwijk CH, et al. Impaired bortezomib binding to mutant beta5 subunit site formation in the 20S proteasome to completion of assembly. Cell of the proteasome is the underlying basis for bortezomib resistance in 1996;86:961–72. leukemia cells. Leukemia 2012;26:757–68.

OF12 Clin Cancer Res; 2019 Clinical Cancer Research

Exosomal PSMA3 and PSMA3-AS1 in PI Resistance

10. Lichter DI, Danaee H, Pickard MD, Tayber O, Sintchak M, Shi H, et al. 31. Wang L, Park HJ, Dasari S, Wang S, Kocher JP, Li W. CPAT: Coding-Potential Sequence analysis of beta-subunit genes of the 20S proteasome in patients Assessment Tool using an alignment-free logistic regression model. with relapsed multiple myeloma treated with bortezomib or dexameth- Nucleic Acids Res 2013;41:e74. asone. Blood 2012;120:4513–6. 32. Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, Nakamura M, 11. Kuehl WM, Bergsagel PL. Molecular pathogenesis of multiple myeloma et al. Antisense transcription in the mammalian transcriptome. Science and its premalignant precursor. J Clin Invest 2012;122:3456–63. 2005;309:1564–6. 12. McDonald MM, Fairfield H, Falank C, Reagan MR. Adipose, bone, and 33. Faghihi MA, Modarresi F, Khalil AM, Wood DE, Sahagan BG, Morgan TE, myeloma: contributions from the microenvironment. Calcif Tissue Int et al. Expression of a noncoding RNA is elevated in Alzheimer's disease and 2017;100:433–48. drives rapid feed-forward regulation of beta-secretase. Nat Med 2008;14: 13. Tassone P, Neri P, Carrasco DR, Burger R, Goldmacher VS, Fram R, et al. A 723–30. clinically relevant SCID-hu in vivo model of human multiple myeloma. 34. Wahlestedt C. Natural antisense and noncoding RNA transcripts as poten- Blood 2005;106:713–6. tial drug targets. Drug Discov Today 2006;11:503–8. 14. Hsu YL, Hung JY, Chang WA, Lin YS, Pan YC, Tsai PH, et al. Hypoxic lung 35. Su WY, Li JT, Cui Y, Hong J, Du W, Wang YC, et al. Bidirectional regulation cancer-secreted exosomal miR-23a increased angiogenesis and vascular between WDR83 and its natural antisense transcript DHPS in gastric permeability by targeting prolyl hydroxylase and tight junction protein cancer. Cell Res 2012;22:1374–89. ZO-1. Oncogene 2017;36:4929–42. 36. Mitsiades CS, Mitsiades NS, McMullan CJ, Poulaki V, Shringarpure R, 15. Shao H, Chung J, Lee K, Balaj L, Min C, Carter BS, et al. Chip-based analysis Akiyama M, et al. Inhibition of the insulin-like growth factor receptor-1 of exosomal mRNA mediating drug resistance in glioblastoma. tyrosine kinase activity as a therapeutic strategy for multiple myeloma, Nat Commun 2015;6:6999. other hematologic malignancies, and solid tumors. Cancer Cell 2004;5: 16. Roccaro AM, Sacco A, Maiso P, Azab AK, Tai YT, Reagan M, et al. BM 221–30. mesenchymal stromal cell-derived exosomes facilitate multiple myeloma 37. Manier S, Liu CJ, Avet-Loiseau H, Park J, Shi J, Campigotto F, et al. progression. J Clin Invest 2013;123:1542–55. Prognostic role of circulating exosomal miRNAs in multiple myeloma. 17. Wang J, Hendrix A, Hernot S, Lemaire M, De Bruyne E, Van Valckenborgh E, Blood 2017;129:2429–36. et al. Bone marrow stromal cell-derived exosomes as communicators in 38. Zhou W, Fong MY, Min Y, Somlo G, Liu L, Palomares MR, et al. Cancer- drug resistance in multiple myeloma cells. Blood 2014;124:555–66. secreted miR-105 destroys vascular endothelial barriers to promote metas- 18. Qu L, Ding J, Chen C, Wu ZJ, Liu B, Gao Y, et al. Exosome-transmitted tasis. Cancer Cell 2014;25:501–15. lncARSR promotes sunitinib resistance in renal cancer by acting as a 39. Reagan MR, Ghobrial IM. Multiple myeloma mesenchymal stem cells: competing endogenous RNA. Cancer Cell 2016;29:653–68. characterization, origin, and tumor-promoting effects. Clin Cancer Res 19. Li Q, Shao Y, Zhang X, Zheng T, Miao M, Qin L, et al. Plasma long 2012;18:342–9. noncoding RNA protected by exosomes as a potential stable biomarker 40. Boelens MC, Wu TJ, Nabet BY, Xu B, Qiu Y, Yoon T, et al. Exosome transfer for gastric cancer. Tumour Biol 2015;36:2007–12. from stromal to breast cancer cells regulates therapy resistance pathways. 20. Wang P, Xue Y, Han Y, Lin L, Wu C, Xu S, et al. The STAT3-binding long Cell 2014;159:499–513. noncoding RNA lnc-DC controls human dendritic cell differentiation. 41. Tomlin FM, Gerling-Driessen UIM, Liu YC, Flynn RA, Vangala JR, Lentz CS, Science 2014;344:310–3. et al. Inhibition of NGLY1 inactivates the transcription factor Nrf1 and 21. Scarola M, Comisso E, Pascolo R, Chiaradia R, Marion RM, Schneider C, potentiates proteasome inhibitor cytotoxicity. ACS Cent Sci 2017;3: et al. Epigenetic silencing of Oct4 by a complex containing SUV39H1 and 1143–55. Oct4 pseudogene lncRNA. Nat Commun 2015;6:7631. 42. Bianchi G, Oliva L, Cascio P, Pengo N, Fontana F, Cerruti F, et al. The 22. Liu B, Sun L, Liu Q, Gong C, Yao Y, Lv X, et al. A cytoplasmic NF-kappaB proteasome load versus capacity balance determines apoptotic sensitivity interacting long noncoding RNA blocks IkappaB phosphorylation and of multiple myeloma cells to proteasome inhibition. Blood 2009;113: suppresses breast cancer metastasis. Cancer Cell 2015;27:370–81. 3040–9. 23. Wang Y, He L, Du Y, Zhu P, Huang G, Luo J, et al. The long noncoding RNA 43. Shabaneh TB, Downey SL, Goddard AL, Screen M, Lucas MM, Eastman lncTCF7 promotes self-renewal of human liver cancer stem cells through A, et al. Molecular basis of differential sensitivity of myeloma cells to activation of Wnt signaling. Cell Stem Cell 2015;16:413–25. clinically relevant bolus treatment with bortezomib. PLoS One 2013;8: 24. Faghihi MA, Wahlestedt C. Regulatory roles of natural antisense transcripts. e56132. Nat Rev Mol Cell Biol 2009;10:637–43. 44. Lu S, Yang J, Chen Z, Gong S, Zhou H, Xu X, et al. Different mutants of 25. Carrieri C, Cimatti L, Biagioli M, Beugnet A, Zucchelli S, Fedele S, et al. Long PSMB5 confer varying bortezomib resistance in T lymphoblastic lympho- non-coding antisense RNA controls Uchl1 translation through an embed- ma/leukemia cells derived from the Jurkat cell line. Exp Hematol 2009;37: ded SINEB2 repeat. Nature 2012;491:454–7. 831–7. 26. Zhan F, Barlogie B, Arzoumanian V, Huang Y, Williams DR, Hollmig K, 45. Mulligan G, Mitsiades C, Bryant B, Zhan F, Chng WJ, Roels S, et al. Gene et al. Gene-expression signature of benign monoclonal gammopathy expression profiling and correlation with outcome in clinical trials of the evident in multiple myeloma is linked to good prognosis. Blood 2007; proteasome inhibitor bortezomib. Blood 2007;109:3177–88. 109:1692–700. 46. Hudson WH, Pickard MR, de Vera IM, Kuiper EG, Mourtada-Maarabouni 27. Zhan F, Hardin J, Kordsmeier B, Bumm K, Zheng M, Tian E, et al. Global M, Conn GL, et al. Conserved sequence-specific lincRNA-steroid receptor gene expression profiling of multiple myeloma, monoclonal gammopathy interactions drive transcriptional repression and direct cell fate. of undetermined significance, and normal bone marrow plasma cells. Nat Commun 2014;5:5395. Blood 2002;99:1745–57. 47. He Y, Vogelstein B, Velculescu VE, Papadopoulos N, Kinzler KW. The 28. Agnelli L, Mosca L, Fabris S, Lionetti M, Andronache A, Kwee I, et al. A SNP antisense transcriptomes of human cells. Science 2008;322:1855–7. microarray and FISH-based procedure to detect allelic imbalances in 48. Saldana-Meyer R, Gonzalez-Buendia E, Guerrero G, Narendra V, Bonasio R, multiple myeloma: an integrated genomics approach reveals a wide gene Recillas-Targa F, et al. CTCF regulates the human p53 gene through direct dosage effect. Genes Chromosomes Cancer 2009;48:603–14. interaction with its natural antisense transcript, Wrap53. Genes Dev 2014; 29. Brehm A, Liu Y, Sheikh A, Marrero B, Omoyinmi E, Zhou Q, et al. Additive 28:723–34. loss-of-function proteasome subunit mutations in CANDLE/PRAAS 49. Huang X, Yuan T, Tschannen M, Sun Z, Jacob H, Du M, et al. Character- patients promote type I IFN production. J Clin Invest 2015;125:4196–211. ization of human plasma-derived exosomal RNAs by deep sequencing. 30. Kong L, Zhang Y, Ye ZQ, Liu XQ, Zhao SQ, Wei L, et al. CPC: assess the BMC Genomics 2013;14:319. protein-coding potential of transcripts using sequence features and support 50. Bardelli A, Pantel K. Liquid biopsies, what we do not know (Yet). vector machine. Nucleic Acids Res 2007;35:W345–9. Cancer Cell 2017;31:172–9.

www.aacrjournals.org Clin Cancer Res; 2019 OF13

Exosome-Transmitted PSMA3 and PSMA3-AS1 Promote Proteasome Inhibitor Resistance in Multiple Myeloma

Hongxia Xu, Huiying Han, Sha Song, et al.

Clin Cancer Res Published OnlineFirst January 4, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-18-2363

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2019/01/04/1078-0432.CCR-18-2363.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2019/02/03/1078-0432.CCR-18-2363. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.