NEW BIOMARKERS AND BIOTARGETS IN RENAL CELL CARCINOMA

Abstract

Renal Cell Carcinoma (RCC) encompasses a heterogeneous group of cancers derived from renal tubular epithelial cells and has a worldwide mortality. However, mortality rates have barely improved over the last 20 years. Novel biomarkers and biomarkers are thus urgently required for this cancer. The inventors have devised a strategy to produce mouse cancer cell lines of progressively enhanced aggressiveness and specialization. The mouse renal cancer cell line RENCA was serially passaged in vivo using multiple implantation strategies designed to replicate different aspects of primary tumour growth and metastasis. Transcriptomic and epigenomic data has been acquired for the derived cell lines and primary analyses have been performed. The inventors then selected plurality of genes with no reported role in RCC which were upregulated in their specialized cell lines, and checked their relevance in patient data and clinical samples. This approach contributes to identify 4 serum biomarkers, namely IL-34, SAA2, PONL1 and CFB that are suitable for predicting survival time in patients suffering from RCC. The inventors also validated that the 4 proteins are also biotargets for the treatment of RCC.

Claims

1. A method for predicting the survival time of a patient suffering from a renal cell carcinoma (RCC) comprising i) determining the expression level of at least one biomarker selected from the group consisting of IL-34, SAA2, PONL1 and CFB in a sample obtained from the patient, ii) comparing the expression level determined at step i) with a predetermined reference value and wherein a difference between the determined expression level and said predetermined reference value is indicative whether the patient will have a long or short survival time.

2. The method of claim 1 wherein the sample is a blood sample (e.g. serum sample) or a tumor tissue sample.

3. The method of claim 1 wherein the expression levels of 2 biomarkers are determined in the sample.

4. The method of claim 1 wherein the expression levels of 3 biomarkers are determined in the sample.

5. The method of claim 1 wherein the expression levels of the 4 biomarkers (i.e. IL-34, SAA2, PONL1 and CFB) are determined in the sample.

6. The method of claim 1 wherein a score which is a composite of the expression levels of the different biomarkers is determined and compared to the predetermined reference value wherein a difference between said score and said predetermined reference value is indicative whether the patient will have a long or short survival time.

7. Use of the method of claim 1 for selecting a therapeutic regimen or determining if a certain therapeutic regimen is more appropriate for a patient identified as having a poor prognosis.

8. A method of treating RCC in a patient comprising identifying the patient as having a poor prognosis by testing the patient according to the method of claim 1, administering an anti-cancer therapy to the patient when the patient is identified as having a poor prognosis, retesting the patient after the step of administering, and administering the anti-cancer therapy to the patient at a maintenance dose when the patient is identified as having a good prognosis.

9. Use of the method of claim 1 for determining whether the patient will be responsive or experience a positive treatment outcome to a treatment.

10. A method of treating renal cell carcinoma (RCC) in a patient in need thereof comprising administering to the subject a therapeutically effective amount of an inhibitor of IL-34, SAA2, PONL1 or CFB.

11. The method of claim 10 wherein the treatment comprises administering to the patient an anti-VEGF agent.

12. The method of claim 10 wherein the patient was previously predicted as having a poor prognosis by the method of claim 1.

13. The method of claim 10 wherein the inhibitor is an antibody having specificity for IL-34, SAA2, PONL1 or CFB.

14. The method of claim 13 wherein the antibody is a chimeric antibody, a humanized antibody of a human antibody.

15. The method of claim 10 wherein the inhibitor is an inhibitor of expression.

16. The method of claim 15, wherein the inhibitor of expression is a siRNA or an antisense oligonucleotide.

Description

FIGURES

[0057] FIG. 1. Interleukin-34 (IL34) expression in mouse and human samples. (A) IL34 expression increases in mouse cell lines rendered increasingly aggressive by in vivo passages. (B) IL34 protein secretion in conditioned media of Passage 6 cell lines (Kidney, Lung and Tail). (c-d) High versus Low IL34 expression predicts Overall (C) and Progression-Free (D) survival in TCGA ccRCC patient cohort. (E-G) In UroCCR patient tissue samples, IL34 RNA is overexpressed in tumour versus healthy kidney (e), increases with Fuhrman grade (F) and correlates with reduced overall patient survival. (H-I) IL34 staining score correlates with Fuhrman Grade (i) and is predictive of Progression Free Survival (j).

[0058] FIG. 2 IL34 Crispr-Cas9 deletion and response to Sutent treatment. (A) IL34 deletion via three different Crispr-Cas9 constructs (CrisprIL34-1a, 1b, 1c) in RENCA cells strongly inhibits primary tumour formation versus control (Crispr-LacZ), leading to reduced tumour weight. (B-C) IL34 deletion via three different Crispr-Cas9 constructs in RENCA cells strongly inhibits experimental metastasis formation (Tail Vein injection, b). Metastases account for a reduced % of lung tissue area (B) and have reduced numbers of MMR+ cells (type 2 macrophages, C). (D) Plasma IL34 levels rise markedly in a subset of patients given first-cycle Sutent therapy for metastases. (E) In a mouse xenograft model, subcutaneous tumours from mice treated with Sutent (sunitinib) versus vehicle control (CT) showed increased levels of IL34 RNA. Species specific qPCR primers show upregulation in both human (tumour cell) and mouse (stromal cell) compartments (E-F).

[0059] FIG. 3. Serum Amyloid Protein A2 (SAA2) expression in mouse and human RCC samples. ( ) SAA2 mRNA expression is upregulated with passage in Lung cell lines (transcriptomic data). (B-C) High versus Low SAA2 expression predicts Overall (b) and Progression-Free (c) survival in TCGA ccRCC cohort. (D-E) SAA2 mRNA is highly expressed in both healthy kidney and tumour tissue (D) however expression in tumour samples increases with grade (E) in UroCCR patient tissue samples. (F-G) mRNA expression is predictive of shortened Overall (F) and Progression Free (G) survival in UroCCR patient cohort.

[0060] FIG. 4. Serum Amyloid Protein A2 (SAA2) in patient plasma/serum samples. (A-C) UroCCR patients, plasma collected from 47 patients prior to surgical resection of primary tumour. Mean plasma SAA2 level is increased in patients with concomitant metastases (M1) or patients without metastases who later develop them (M0 progressors) versus patients without metastases who do not progress during follow-up (M0 non progressors). Higher plasma level predicts shortened Progression Free (B) and Overall (C) survival. (D-F) SUVEGIL 11 patients, following surgical resection of primary tumour. Serum samples collected at diagnosis with metastases, and following one cycle of Sutent treatment. Patients who did not show progression following treatment tended to show a decrease in SAA2 level following treatment (D), whereas patients who's metastases progressed (nonresponders) showed increased levels (E, F).

[0061] FIG. 5. Complement Factor B (CFb) expression in mouse and human RCC samples. (A) CFb mRNA expression is upregulated with passage in Lung and Tail cell lines (qPCR data). (B) CFb mRNA expression is increased in RENCA isolated from metastases versus their primary tumours of origin (paired samples). (c, d) High versus Low CFb expression predicts Overall (C) and Progression-Free (D) survival in TCGA ccRCC cohort. (E-G) CFb mRNA is overexpressed in tumour versus healthy UroCCR tissue samples, and predicts shortened Overall (F) and Progression-Free (G) survival.

[0062] FIG. 6. Complement Factor B (CFb) in patient plasma/serum samples. (A) UroCCR patients, plasma collected from 47 patients prior to surgical resection of primary tumour. Mean plasma CFb level is increased in patients with concomitant metastases (M1) or who patients without metastases who later progress (M0 progressors) versus patients without metastases who do not progress during followup (M0 non progressors). (B) UroCCR patients. Plasma collected from N=20 patients approximately 1 month following surgery for primary tumour removal. The plasma CFb level is again higher in patients with metastases (M1) versus those without (M0). (C) SUVEGIL 11 patients, following surgical resection of primary tumour. Serum samples collected at diagnosis with metastases, and following one cycle of Sutent treatment. Patients who's CFb serum level increased following treatment had faster progression than those who's level decreased. (D) SUVEGIL 11 patients, divided into three groups according to increase/decrease in both SAA2 and CFb as combinatorial analysis. Patients with decrease in neither protein show best outcome, and the patient with increase in both the worst outcome.

[0063] FIG. 7 Podocan-like1 (Podnl1) expression in mouse and human RCC samples. (A, B) Podnl mRNA expression is upregulated with passage in Kidney cell lines (microarray, b qPCR data). (C) Podnl mRNA expression is increased in RENCA isolated from primary tumours versus in vitro culture but is not further increased in metastatic tumour cells (paired samples). (D, E) High versus Low Podnl1 expression predicts Overall (D) and Progression-Free (E) survival in TCGA ccRCC cohort. (F, G) Podnl1 mRNA is overexpressed in tumour versus healthy UroCCR tissue samples, and predicts shortened Overall (F) and Progression-Free (G) survival.

[0064] FIG. 8. Clinical relevance of SAA2 and CFB after anti-angiogenic treatment (SUVEGILTORAVA cohorts). (A and B) Correlation between plasmatic SAA2 levels at diagnosis and survival (OS and PFS) in patients after sunitinib treatment (plasmatic level at the diagnosis less or greater than a cut-off for SAA2 (269 μg/ml) [OS: HR(log-rank)=5.557; PFS: HR(logrank)=7.669. (C) Correlation between plasmatic SAA2 levels at diagnosis and PFS in patients after sunitinib or bevacizumab treatment (plasmatic level at the diagnosis less or greater than a third quartile cut-off for SAA2 (269 μg/ml; HR(log-rank)=1.987). (D) Correlation between plasmatic CFB levels at diagnosis and PFS in patients after sunitinib or bevacizumab treatment (plasmatic level at the diagnosis less or greater than a third quartile cut-off for CFB (310 μg/ml; HR(log-rank)=3.113) (E and F) PFS (E) and OS (F) patients treated with either Sunitinib of bevacizumab and stratified according to plasma levels of both SAA2 and CFB. Three subgroups were identified i) CFB low and SAA2 low, ii) CFB low and SAA2 high or CFB high and SAA2 low, iii) CFB high and SAA2 high (Low-low vs high-high: OS HR(log-rank)=5.086; PFS HR(log-rank)=4.196).

EXAMPLE

[0065] Methods

[0066] Mice and Cell Lines

[0067] Female BALB/c mice 6-8 weeks of age were purchased from Charles River Laboratories. Mice were housed in the animal facility of Bordeaux University (Animalerie Mutualisée Bordeaux, France). The GFP expressing Renca murine renal cancer cell line (RENCA-GFP) and sub-cell lines generated (Kidney, Tail, Lung) were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% foetal bovine serum (FBS) and 1% penicillin/streptomycin and were incubated at 37° C., 5% CO.sub.2 in an incubator. Crispr-Cas9_IL34 and CrisprCas9_LacZ cell lines were generated using standard protocols.

[0068] Mouse Orthtotopic Subcapsular and Experimental Metastasis (Tail Vein Injection) Models.

[0069] Tumours were implanted by sub-capsular injections of 1×10.sup.5 RENCA-GFP cells into the left kidney of wild type BALB/c mice. For the intravenous injections, 5×10.sup.5 RENCA-GFP cells were injected into the caudal vein of wild type BALB/c mice. When the endpoints defined by the approved protocols were reached, mice were sacrificed, and tumour tissues and lungs were collected. For immunochemistry, tissue were fixed in paraformaldehyde 4% (PFA 4%, Santa Cruz Biotechnology, sc-281692) for 2 hours and then incubated for 72 hours in 30% sucrose. Tissues were frozen in OCT Compound (Tissue-Tek OCT compound, Sakura, 4583). Prior to embedding, lungs were inflated with 1 mL of diluted OCT (1:1 PBS/OCT dilution). Frozen tissues were preserved at −80° C. For protein, DNA and RNA analysis, tissues were snap-frozen in liquid nitrogen and preserved at −80° C.

[0070] Tissue Dissociation and Tumour Cell Purification

[0071] For tumour cell purification, tissues were cut into small pieces with a scalpel and digested with Collagenase I and Collagenase II (Liberase TL, Roche, 05401020001) for 1 hour at 37° C. To further improve the dissociation, digested tissues were filtered in cell strainers (100 μm, 70 μm and 40 μm) and seeded in complete medium, and incubated at 37° C., 5% CO.sub.2. Cell cultures were checked daily and passaged as necessary. Tumour cell outgrowth and primary cell death resulted in tumour cell only cultures, verified by visualisation of GFP using fluorescence microscopy When no GFP-negative cells could be visually detected, cell cultures were considered sufficiently pure. RENCA-GFP cells were collected for analysis or re-implanted into mice for further in-vivo passage. In some cases, cells were cultured in serum free media for 24 hrs to generate conditioned medium.

[0072] Xenograft mouse experiments were done with subcaneously injected 786-0 human RCC cells in immunodeficient mice and treated with sunitnib (40 mg/kg) according to published protocols (Dufies et al, Cancer Res, March 2017 DOI: 10.1158/0008-5472.CAN-16-3088)

[0073] Gene Expression Analysis

[0074] Total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen, #74134), according to the manufacturer's instructions. Agilent mouse full Genomic Array was used for transcriptomic analysis.

[0075] Quantitative PCR (qPCR) analyses: 1 μg of total RNA was reverse-transcribed into complementary DNA (cDNA) using the high-capacity cDNA reverse transcription kit (Applied Biosystems, 4368814). The resulting cDNA were amplified using specific primers for the genes of interest. HPRT was used as internal control.

[0076] Enzyme-linked immunosorbent assays (ELISA) were performed according to the manufacturer's instructions on conditioned media or human plasma or serum samples.

[0077] Patient Samples

[0078] UroCCR Tissue Bank.

[0079] Clinical data and biological samples (frozen/paraffin-embedded tissue, plasma and urine samples) were obtained from the French research network on kidney cancer www.uroCCR.fr funded by INCa and localised in Bordeaux. ClinicalTrials.gov identifier: NCT03293563. These samples are referred to as UroCCR cohort. Tissue samples were obtained from patients on the day of surgery for removal of the primary tumour. Plasma and urine samples were obtained either on the day of surgery for the primary tumour or at a time point approximately one month following surgery.

[0080] SUVEGIL Serum Samples.

[0081] Serum samples from the SUVEGIL clinical trial (Sunitinib Malate in Treating Patients With Kidney Cancer, ClinicalTrials.gov identifier NCT00943839). Patients receive oral sunitinib malate once daily on days 1-28. Courses repeat every 6 weeks in the absence of disease progression or unacceptable toxicity. Blood samples are collected at baseline and then every 6 weeks for pharmacokinetic analysis. In this case, samples tested were obtained at the point of diagnosis of metastases and following the first cycle of treatment.

[0082] Immunochemistry and Immunofluorescence

[0083] Mouse tissues: For frozen mouse tissues obtained from experiments using CrisprCas9_IL34 and CrisprCas9_LacZ cell lines 10 μm sections were performed with a cryostat (Leica CM1900). For frozen tissue immunofluorescence, sections were incubated 1 hour with a blocking buffer (5% BSA in PBS). Slides were incubated overnight with primary antibody (MMR: R&D Systems, AF2535; GFP: Torrey Pines Biolabs, TP401=>table), and then with secondary fluorescent antibody (REFERENCE=>table) and DAPI (Roche, 10236276001). Images were obtained using a slide scanner (Hamamatsu, Nanozoomer 2.0HT), and processed using NDP.scan software (Hamamatsu). Image analysis using Fiji software (Schindelin, J.; Arganda-Carreras, I. & Frise, E. et al. (2012) Nature methods 9(7): 676-682) was used to calculate the area of tumour tissue as percentage of total tissue section area (%) based on GFP staining. Type 2 macrophage density in tumour tissue was calculated by counting number of MMR-positive cells/pixel area using the “Cell Counter” plugin (Kurt de Vos). Mean areas/cell counts are expressed normalised to those obtained from control tumours.

[0084] Human tissues: For paraffined tissues sections were prepared with a microtome. For paraffin tissue sections slides were deparaffinised, re-hydrated and heated in Antigen Retrieval Solution pH6 (HIER Sodium Citrate Buffer, pH6; 10 mM Sodium Citrate, 0.05% Tween 20, pH 6,0). To block endogenous peroxidase activity, slices were treated with 0,3% hydrogen peroxide. After 1 hour of blocking in PBS 5% BSA, slides were incubated overnight with primary antibody (see table), and then incubated with biotinylated secondary antibody for 1 h (see table). Secondary antibodies were HRP-conjugated using the “ABC” technique (Vectastain PK-6100) and then revealed with a peroxidase substrate kit (DAB, Vector Laboratories, SK-4100).

[0085] In Silico Analyses

[0086] Transcriptional and clinical patient data was obtained from The Human Genome Atlas via the BioPortal website. using the Kidney Renal Cell Carcinoma (KIRC) database. Kaplan Meier graphs representing Overall Survival (OS) and Progression Free Survival (PFS) and all statistical analyses were performed using GraphPad Prism software. For Kaplan Meier analyses, where patient numbers per high/low group are not stated, the cut point is the median value.

[0087] Results:

[0088] Results are depicted in FIG. 1-7.

[0089] FIGS. 1A to 1I show the interleukin-34 (IL34) expression in mouse and human samples. FIGS. 2A to 2F show the IL34 Crispr-Cas9 deletion and response to Sutent treatment. FIG. 3A to 3G show the Serum Amyloid Protein A2 (SAA2) expression in mouse and human RCC samples. FIG. 4A to 4F show the Serum Amyloid Protein A2 (SAA2) in patient plasma/serum samples. FIG. 5A to 5G show the Complement Factor B (CFb) expression in mouse and human RCC samples. FIG. 6A to 6D show the Complement Factor B (CFb) in patient plasma/serum samples. FIG. 7A to 7G show the Podocan-like1 (Podnl1) expression in mouse and human RCC samples.

[0090] Serum Amyloid A2 (SAA2)

[0091] SAA2 is an acute phase protein related to SAA1, which was previously linked to metastasis. Its expression was strongly upregulated with passage in the Lung cell lines (data not shown). In silico analysis of the TCGA KIRC database SAA2 was a very strong predictor of OS (FIG. 3B) and DFS (FIG. 3C). Furthermore, the analysis was also done for the M0 and M1 subgroups (FIG. 3E). Analysis of the UroCCR patient cohort confirmed the effect on OS and DFS (FIG. 3G). Tumors from patients with the highest Fuhrman Grade, had a significantly increased SAA2 expression compared to all other grades (data not shown). We used grade-matched plasma samples from patients with and without metastases, collected before primary tumor surgery (data not shown). Patients with metastases had higher plasma levels of SAA2. When patients were divided into two groups of equivalent size, the group with higher SAA2 levels had a significantly shorter DFS (Supplementary FIG. 8l). A second set of plasma samples, collected in the weeks following surgery for removal of the primary tumor, was tested for SAA2 (Supplementary FIG. 8m). In this case, patients with higher expression had shorter OS. Hence, circulating SAA2 levels appear as an indicator of metastatic progression that deserves to be evaluated at diagnosis. We next used plasma samples from metastatic patients before receiving a first cycle of sunitinib or bevacizumab (SUVEGIL and TORAVA clinical trials). Patients treated with sunitinib only and stratified according to low and high SAA2 levels, had a spectacular better OS and progression-free survival (PFS) when belonging to the SAA2 low group (cut-off of 269 μg/ml) (FIG. 8A and 8B). When patients treated with sunitinib and bevacizumab were analyzed together, the PFS was of limited significance (borderline p-value of 0.0507) (FIG. 8C). The median of PFS for SAA2high patients was of 5.35 month versus 16.17 month for the SAA2low group. Thus, determining SAA2 plasma levels could be a useful measure for deciding a treatment strategy in RCC.

[0092] Complement Factor-B (CFB)

[0093] CFB was most strongly upregulated in the “Lung” and to a lesser extent in the “Tail” group, both considered to recapitulate features of metastasis (data not shown). TCGA analysis in ccRCC showed that CFB expression is correlated in primary tumors with shortened DFS and OS (FIG. 5D). We also performed the analysis in the M0 and M1 subgroups (data not shown). Using samples and data from the UroCCR cohort, we demonstrated that CFB was overexpressed in the tumor tissue versus the adjacent kidney at the mRNA level (FIG. 5E), and that increased expression correlated with reduced DFS and OS, consistent with the results obtained with the TCGA cohort (FIGS. 5F and 5G). As for SAA2, CFB can be measured in the blood. For this purpose, we used UroCCR plasma samples collected from patients either before surgery (primary tumor intact) or in the following weeks after surgery (no primary tumor present but metastases in situ possible). Before surgery, a trend was observed without reaching significance whereas after surgery patients with metastases had higher plasma CFB levels compared to patients without metastases (data not shown). This suggests that circulating CFB measurement may be useful as a blood-born marker of metastasis in the follow-up after surgical tumor removal. As for SAA2, CFB plasma levels were tested in patients with metastases before the first cycle treatment with sunitinib or bevacizumab (SUVEGIL and TORAVA clinical trials). Patients whose levels were high (cut-off 310 μg/ml) had faster disease progression compared to patients whose levels were low (high CFB, 3.58 month; low CFB, 18.7 month, p=0.0004) (FIG. 8D). We then grouped the significance of testing SAA2 and CFB plasmatic levels (FIGS. 8E and 8F). Three different groups with different survival can be identified: group 1 (CFB low+SAA2 low, PFS: 19.37 months, OS: NR), group 2 (CFB high SAA2 low or CFB low SAA2 high, PFS: 9.87 months, OS: 20.9 months), group 3 (CFB high Saa2 high, PFS: 2.8 month, OS:8.33 months). Group 1 had the best survival rate while group 3 had the worst. Group 2 had intermediate survival outcome. Thus, the combined analysis of these two markers is a powerful predictor of patient outcome following anti-angiogenic treatment with sunitinib or bevacizumab.

[0094] Podocan Like Protein-1 (PODNL1)

[0095] PODNL1 is a member of the small leucine-rich proteoglycan (SLRP) family of 17 genes. It is secreted extracellularly and its function is currently unknown. High expression has previously been linked with poor outcome in ovarian cancer and glioblastoma. PODNL1 expression was upregulated in our mouse cell lines in the “Kidney” subgroup (data not shown), although the increase was relatively modest. However, this gene showed a very strong link with reduced DFS and OS in the TCGA KIRC database (FIGS. 7D and 7E). We have also performed this analysis in M0 and M1 patients (data not shown). When using UroCCR samples, PODNL1 was overexpressed at the mRNA level in the tumor versus healthy tissue (FIG. 7G). In the UroCCR biobank, DFS also showed different trends depending on PODNL1 expression albeit statistically not significant, the latter was also the case for OS (FIG. 7F). This largely unknown and interesting gene may play a key role in RCC and further studies are required to investigate this possibility.

REFERENCES

[0096] Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.