RBM3: A Prognostic and Treatment Predictive

Introduction it has also been shown that RBM3 attenu- Cancer is continuously a major health ates stem cell-like properties in prostate problem throughout the world. In 2008, cancer cells6. there were an estimated 12.7 million new cancer cases world-wide, and 7.6 RBM3 monoclonal million deaths that could be attributed to The monoclonal Anti-RBM3 antibody cancer1. Due to a world population with AMAb90655, used in all cases presented an increased life-expectancy, there is no here (available for purchase from Atlas reason to expect a decline in cancer inci- , Stockholm, Sweden, atlasan- dence in the near future2. tibodies.com), has shown excellent spe- cificity in Western Blot analysis of human Cancer, though often denoted as a sin- cell lines, and is routinely used for stain- gular disease, is truly a multitude of dis- ing of formalin fixed paraffin embedded eases. This understanding has evolved tissue in IHC. A representative image of over the years, and though common immunohistochemical staining using the knowledge today, many patients are still RBM3 monoclonal antibody AMAb90655 not receiving optimal treatment for their can be seen in Figure 1. disease. For cancer patients to receive a more individualized treatment, there is RBM3 as a prognostic biomarker still a need for new and better ways to After identification of RBM3 as a poten- stratify patients. The classical, pathol- tial prognostic biomarker, RBM3 ogy based, prognostic factors such as expression has been analyzed using stage and grade of the tumor are insuf- AMAb90655 in many different patient co- ficient for a correct estimation of patient horts from various forms of cancer. Levels prognosis. Additional information from of RBM3 expression was found to have immunopathology promise to a significant connection to patient survival substantially improve this estimation, ul- in breast7, colon8, ovarian9,10, testicular, timately leading to a more individualized urothelial11, and prostate12 cancer as well treatment, thus avoiding both under- and as in malignant melanoma13. over treatment of patients.

RBM3 and the Human Protein Atlas The RNA-binding motif protein 3 (RBM3) was recently identified via the Human Figure 1 Protein Atlas (HPA) as a potential oncol- Figure 1 shows representative nuclear staining of ogy biomarker through the differential RBM3 in urothelial bladder cancer, with no RBM3 ex- expression pattern present in several pression (A), intermediate RBM3 expression (B) and high RBM3 expression (C) respectively. cancers investigated as part of the HPA project (proteinatlas.org)3,4.

RBM3 is an RNA- and DNA-binding pro- tein, whose function has not been fully elucidated. It has been shown that the protein is expressed as an early event in mild hypothermia, and also in other con- ditions relating to cellular stress, such as glucose deprivation and hypoxia5. During stress, RBM3 is thought to protect the cells by aiding in maintenance of protein synthesis needed for survival5. Recently,

Page 1 (4) Urothelial cancer A 1,0 B 1,0 A major clinical problem in urothelial can- cer is the identification of high-risk- pa 0,8 0,8 tients among those diagnosed with non- invasive disease. For patients with stage 0,6 T1 disease (where the tumor has invaded 0,6 connective tissue but not yet muscle), the prognosis is generally good, but nearly 0,4 0,4 one third of these patients will still even- Disease specific survival tually require cystectomy when other 0,2 Five year overall survival 0,2 RBM3 negative; n=67, p<0.001 RBM3 negative; n=78, p<0.001 therapies have failed. The identification of RBM3 intermediate; n=188, p=0.016 RBM3 intermediate; n=215, p=0.035 RBM3 high; n=46 (reference) RBM3 high; n=53 (reference) these patients at an early stage would be 0,0 0,0 a great improvement for patients as well 0 50 100 150 200 250 0 10 20 30 40 50 60 Months from diagnosis as society. Due to high recurrence rates, Months from diagnosis and need for invasive monitoring of the Figure 2 disease, the health care cost per patient Figure 2 shows Kaplan-Meier analysis of cancer-specific (A) and 5-year overall (B) survival of 343 bladder can- cer patients stratified according to RBM3 expression. is higher for urothelial cancer than for any other cancer in the USA11. For patients with muscle-invasive disease, prognos- tic markers that could guide in the choice A 1,0 B 1,0 of treatment and/or treatment intensity would also mean an improvement in pa- tient care. 0,8 0,8

RBM3 has been shown to be a prognostic 0,6 0,6 marker in urothelial cancer, with a signifi- cantly improved survival for patients with 0,4 0,4 Overall survival high levels of RBM3 expression in their Overall survival tumors11 (see Figure 2). This means that 0,2 RBM3 negative; n=33, p=0.005 0,2 RBM3 negative; n=33, p=0.006 analysis of RBM3 expression could be a RBM3 intermediate; n=151, p=0.231 RBM3 positive; n=198 (reference) RBM3 high; n=47 (reference) prognostic factor for better stratification 0,0 of patients, leading to a more individual- 0,0 ized treatment of patients with urothelial 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Months from diagnosis Months from diagnosis cancer. Figure 3 Figure 3 shows Kaplan-Meier analysis of 5-year overall survival of the 231 patients in the cohort diagnosed with Figure 3 shows overall survival of the pa- stage Ta or T1 disease. Patients were stratified according to RBM3 expression, either according to negative, tients in the cohort with Ta (non invasive) intermediate, and high expression (A), or according to negative vs. positive expression (B). and T1 tumors, and the same results can be seen as for the full cohort. For the sub- group of patients with T1 tumors, 5-year overall survival was significantly reduced for patients with tumors not expressing RBM3, indicating that IHC analysis of RBM3 expression may indeed be used to identify stage T1 patients in need of cys- tectomy. The prognostic significance of RBM3 in urothelial cancer has been con- firmed in several independent cohorts.

Page 2 (4) A 1,0 B 1,0 RBM3 low; n=123, p<0.001 In colorectal cancer, a major clinical prob- RBM3 intermediate; n=60, p=0.064 RBM3 low; n=123, p<0.001 RBM3 high; n=62 (reference) RBM3 high; n=122 (reference) lem is to identify patients in need of adju- 0,8 0,8 vant therapy. The only curative treatment today is surgery, but adjuvant treatment 0,6 0,6 may prolong patient survival, and is rec- ommended for patients with stage III and 0,4 0,4 Overall survival high-risk stage II disease. However, not Overall survival all high-risk stage II patients benefit from 0,2 0,2 adjuvant treatment, thus to categorize these patients is of outmost importance. 0,0 0,0 A biomarker that provides additional in- 0 5 10 15 20 0 5 10 15 20 Years from diagnosis Years from diagnosis formation on risk assessment would be a Figure 4 great step to improve patient care. Figure 4 shows Kaplan-Meier analysis of overall survival of 305 colorectal cancer patients. Patients were strati- fied according to RBM3 expression, either according to low, intermediate or high expression (A), or according to low versus high expression (B). RBM3 has been shown to be a prognostic marker in colorectal cancer in two inde- pendent patient cohorts, with a significant- RBM3 as a treatment predictive in the cisplatin sensitive A2780 cells, they ly improved survival for patients with high biomarker became significantly less sensitive to cis- levels of RBM3 expression in their tumors8 During analysis of RBM3 as a prognos- platin treatment as can be seen in Figure (see Figure 4). This means that analysis of tic biomarker in ovarian cancer, it was 7, where the viability of RBM3-high cells RBM3 expression could be a a prognostic speculated about the potential connection is lower than that of RBM3-low cells. In factor for better stratification of patients, between sensitivity to platinum-based a clinical setting, this finding would corre- leading to more appropriate treatment for therapies and RBM3 expression9. Cis- spond to improved survival for cisplatin- patients with colorectal cancer. When ana- platin treatment is a cornerstone in ovar- treated patients with tumors expressing lyzing patients with stage II disease only, ian cancer treatment and most patients high levels of RBM3 as compared to cis- similar results can be seen as for the full in the cohorts analyzed were thus likely palatin-treated patients with low levels of cohort (Figure 5), indicating that analysis treated with cisplatin. The ovarian can- RBM3 expression in their tumors. of RBM3 expression may be a valuable cer cell lines A2780 and A2780-Cp70 (a tool in deciding whether or not stage II cisplatin resistant cell line derived from colorectal cancer patients should receive A2780) can be used as a model system adjuvant treatment for their disease. for studies related to cisplatin treatment and response. Studies of these two cell It could be concluded that RBM3 is a lines revealed that RBM3 expression (as- prognostic marker that may aid in deter- sessed by Western Blot as well as IHC mining if patients should receive adjuvant analysis) was significantly higher in the treatment, and also the intensity of the cisplatin sensitive A2780 cell line than treatment. in the cisplatin resistant cells (Figure 6). This finding was also confirmed by mRNA 1,0 analysis of RBM3 mRNA levels in the two RBM3 low; n=48, p=0,05 cell lines. When the RBM3 was si- RBM3 high; n=54 (reference) 0,8 lenced by siRNA-mediated knock-down

0,6 A B A278 A2780/C A2780 A2780/Cp70 20 30 50 20 30 50 ug

0,4 50 Overall survival 37 0,2

25

0,0 20

0 5 10 15 20 RBM3 Years from diagnosis Figure 6 Figure 5 Figure 6 shows RBM3 expression in cisplatin sensi- Figure 5 shows Kaplan-Meier analysis of overall sur- tive A2780 ovarian cancer cells and cisplatin resist- vival of the 102 patients in the cohort diagnosed with ant A2780/Cp70 cells as determined by immunocyto- chemistry (A) and Western Blot (B). stage II disease, stratified according to RBM3 expres- actin sion.

Page 3 (4) The results from the in vitro studies above Figure 7 Figure 7 shows viability of the ovarian cancer cell line were further verified in a colorectal cancer 140 sictrl (RBM3 high) A2780 when exposed to increasing concentrations of cohort, where patient survival according to si58 (RBM3 low) cisplatin. The control cells (labeled sictrl) show a sig- 120 si59 (RBM3 low) nificantly reduced viability compared to cells whose RBM3 expression was compared between si60 (RBM3 low) 100 RBM3 gene was silenced by siRNA transfection us- patients receiving no adjuvant treatment, ing three different constructs (labeled si58, si59, and non platinum based adjuvant treatment, 80 si60 respectively). and platinum based adjuvant treatment. 60

Comparing the patients with tumors ex- (%) Viability pressing high levels of RBM3 in the three 40

different groups, it can be seen that pa- 20 tients receiving platinum based adjuvant 0 treatment had a markedly increased sur- 0 20 40 60 80 100 vival compared to patients receiving no- Cisplatin (uM) or non platinum based adjuvant treatment (Figure 8).

Figure 8 1,0 1,0 A B Figure 8 shows Kaplan-Meier analysis of colorectal cancer specific survival of 113 curatively treated pa- tients with stage III-IV disease stratified according 0,8 0,8 to RBM3 expression (low, intermediate, or high ex- pression). In Figure 8A, all patients are represented. For further analysis, patients were divided into three 0,6 0,6 groups by adjuvant treatment modality; no adjuvant treatment (B), non platinum based adjuvant treat- ment (C), and platinum based adjuvant treatment (D). 0,4 0,4

RBM3 nuclear staining References: 0,2 Low; n=35, 17 events 0,2 1) Ferlay J et al. (2008) Estimates of worldwide burden of can- Colorectal cancer specific survival Colorectal cancer specific survival Intermediate; n=36, 14 events RBM3 nuclear staining cer in 2008: GLOBOCAN. Int J Cancer 2010;127:2893-917 High; n=62, 18 events Low; n=12, 5 events Intermediate; n=18, 8 events High; n=28, 8 events 2) Salomon JA et al. (2012) Healthy life expectancy for 187 0,0 0,0 countries, 1990–2010: a systematic analysis for the Global Burden Disease Study. The Lancet 2012;380:2144-62 0 5 10 15 20 0 5 10 15 20 Years from diagnosis Years from diagnosis 3) Berglund L et al. (2008) A gene-centric human protein atlas for expression profiles based on antibodies. Molecular & Cel- C 1,0 D 1,0 lular Proteomics 7:2019-2027. 4) Uhlén M et al. (2010) Towards a knowledge-based Human Protein Atlas. Nat Biotechnol 28(12):1248-50. 0,8 0,8 5) Ehlén Å (2011) PhD Thesis: The role of RNA-binding motif 3 in epithelial ovarian cancer: A biomarker discovery approach. 0,6 0,6 6) Zeng Y et al. (2013) Stress response protein RBM3 at- tenuates the stem-like properties of prostate cancer cells by interfering with CD44 variant splicing. Cancer Res. May 10. 0,4 0,4 [Epub ahead of print]

7) Jögi A et al. (2009) Nuclear expression of the RNA-binding protein RBM3 is associated with an improved clinical outcome 0,2 0,2 RBM3 nuclear staining RBM3 nuclear staining in breast cancer. Mod Pathol. Dec;22(12):1564-74. Colorectal cancer specific survival Colorectal cancer specific survival Low; n=16, 9 events Low; n=5, 2 events Intermediate; n=12, 2 events Intermediate; n=6, 4 events 8) Hjelm B et al. (2011) High nuclear RBM3 expression is High; n=18, 8 events High; n=13, 1 event 0,0 0,0 associated with an improved prognosis in colorectal cancer. Proteomics Clin Appl. Dec;5(11-12):624-35. 0 2 4 6 8 10 12 0 2 4 6 8 10 12 Years from diagnosis Years from diagnosis 9) Ehlén Å et al (2010) Expression of the RNA-binding protein RBM3 is associated with a favourable prognosis and cisplatin sensitivity in epithelial ovarian cancer. J Transl Med. Aug 20; 8:78.

Summary 10) Ehlén Å et al. (2011) RBM3-regulated promote DNA integrity and affect clinical outcome in epithelial ovarian • RBM3 is a prognostic marker in several cancers, with a high RBM3 expression indicating cancer. Transl Oncol. Aug;4(4):212-21. improved patient survival. 11) Boman K et al (2013) Decreased expression of RNA- binding motif protein 3 correlates with tumour progression and poor prognosis in urothelial bladder cancer. BMC Urol. • RBM3 is a treatment predictive marker, with a high RBM3 expression indicating 2013;13:17 response to platinum-based therapies and a better overall survival. 12) Jonsson L et al. (2011) High RBM3 expression in prostate cancer independently predicts a reduced risk of biochemi- cal recurrence and disease progression. Diagn Pathol. Sep • Determining the RBM3 protein expression may improve patient care by aiding in 28;6:91. deciding on the proper treatment for cancer patients. 13) Jonsson L et al. (2011) Low RBM3 protein expression correlates with tumour progression and poor prognosis in ma- lignant melanoma: an analysis of 215 cases from the Malmö Diet and Cancer Study. J Transl Med. Jul 21;9:114.

2013-08-14 Page 4 (4)

Atlas Antibodies AB Phone +46(0)8 54 59 58 50 IBAN SE1230000000032661708553 Plusgiro 209563-6 AlbaNova University Center Fax +46(0)8 54 59 58 51 Swift code NDEASESS Bankgiro 5469-1092 SE-106 91 Stockholm [email protected] Reg. no 556682-8082 Registered Office Stockholm. Sweden Sweden atlasantibodies.com VAT ID no. SE55668208201 Innehar F-skattsedel