Abstract
Platinum-based neoadjuvant chemotherapy prior to radical cystectomy is the preferred treatment for muscle-invasive bladder cancer despite modest survival benefit and significant associated toxicities. Here, we profile the global proteome of muscle-invasive bladder cancers pre- and post-neoadjuvant chemotherapy treatment using archival formalin-fixed paraffin-embedded tissue. We identify four pre-neoadjuvant chemotherapy proteomic clusters with distinct biology and response to therapy and integrate these with transcriptomic subtypes and immunohistochemistry. We observe proteomic plasticity post-neoadjuvant chemotherapy that is associated with increased extracellular matrix and reduced keratinisation compared to pre-neoadjuvant chemotherapy. Post-neoadjuvant chemotherapy clusters appear to be differentially enriched for druggable proteins. For example, MTOR and PARP are over-expressed at the protein level in tumours identified as neuronal-like. In addition, we determine that high intra-tumoural proteome heterogeneity in pre-neoadjuvant chemotherapy tissue is associated with worse prognosis. Our work highlights aspects of muscle-invasive bladder cancer biology associated with clinical outcomes and suggests biomarkers and therapeutic targets based on proteomic clusters.
Similar content being viewed by others
Introduction
Bladder cancer (BC) is one of the most frequently diagnosed cancers1. It is responsible for over 570,000 new diagnoses and 212,000 deaths per year worldwide2. BC can be categorised as non-muscle-invasive (NMIBC; 75%) or muscle-invasive (MIBC; 25%). MIBC is an aggressive neoplasm that progresses to metastatic urothelial carcinoma (mUC) and death in approximately half of cases3.
MIBC is optimally treated with cisplatin-based neoadjuvant chemotherapy (NAC) followed by radical cystectomy (RC) in patients who are cisplatin-eligible, with an overall survival (OS) benefit of 5–7% at 5 years. Approximately 40% of patients treated with NAC exhibit a major pathologic response (defined as absence of muscle-invasive disease and nodal invasion) at the time of RC, and those with ypT0 showed 85% 5-year survival3,4. Patients with residual MIBC are exposed to a risk of adverse effects from NAC and a delay in definitive local therapy without therapeutic benefit5. To date, no validated molecular markers exist to guide therapeutic choices in MIBC patients. Identifying who will benefit from NAC and subsequent RC therefore remains an active field of research.
Transcriptomic characterisation of MIBC has defined biological subgroups that are associated with distinct clinical outcomes6,7, as well as potential benefit to NAC6,8,9,10,11,12 and/or immunotherapy13,14,15. Choi et al.12 reported on the ‘p53-like’ subtype associated resistance to NAC. Seiler et al.8 subsequently showed improved survival but no difference in pathologic response in MIBC patients with ‘basal’ tumours but not for non-basal subtypes who received NAC compared to those who did not receive NAC. Other studies10,11,16 reported heterogeneous results, owing to differences in endpoints, patient populations, and the transcriptomic subtype classifiers used. Mutations in certain DNA damage repair genes have also been associated with response to chemotherapy and improved survival in both mUC and MIBC17. Examples include ERCC218,19,20,21, ATM, RB1, and FANCC22,23. Implementation of these markers into clinical practice requires prospective validation24,25.
Despite this progress, transcriptomic profiles and DNA alterations often do not directly reflect protein expression, making analysis of the proteome necessary to facilitate identification of therapeutic targets and clinically useful biomarkers26. Technologic advances have enabled proteomic analysis of easy-to-access formalin-fixed paraffin-embedded (FFPE) archival specimens by mass spectrometry (MS)27,28,29. Recent studies extensively investigated BC proteogenomics30,31,32. Investigations included whole exome, transcriptome, proteome and phosphoproteome for a mixed cohort of NMIBC, chemo-naïve MIBC samples31,32, and normal and hyperplastic lesions24,30 with detailed clinical annotations.
These studies provided tumour proteome data, but there is a gap in proteomic characterisation of MIBC regarding mechanisms of response and resistance to NAC. Here, we use an advanced MS-based approach (SP3-CTP33) on FFPE specimens to analyze the proteome of 107 MIBC patients undergoing NAC. Integrating proteomic and transcriptomic data before and after chemotherapy reveals four distinct proteomic clusters pre-NAC and four post-NAC, each with varied biology and clinical outcomes. Our analysis identifies potential prognostic biomarkers and proteins relevant to chemoresistance, offering a platform for further exploration.
Results
Patient characteristics
We analyzed the proteome of 107 MIBC patients undergoing chemotherapy (neoadjuvant or induction, collectively referred to as ‘NAC’ in this study), and N = 25 NMIBC patients. All MIBC patients received transurethral resection (TUR) followed by RC, with 44% showing nodal metastasis and 49% being cT3-4 pre-NAC, with a complete response of 30% defined at time of RC as ypT0N0M0 (Table 1).
Pre-NAC TUR tissue was harvested from all 107 MIBC patients, with an additional 36 duplicate TUR samples collected to study intra-tumoural heterogeneity (ITH) for a total of 143 TUR specimens. 75 of 107 patients were partial (pTis/a/1N0M0) or non-responders (≥pT2NXM0) after NAC and RC. From these, we collected tissue from 55 post-NAC RC specimens (20 had no available material), and 14 additional duplicate RC specimens to study ITH, for a total of 69 post-NAC RC specimens. We also pooled N = 37 benign ureter (BU) samples collected at time of RC. The NMIBC cohort comprised 29 TUR tissue specimens from 25 NMIBC patients (4 duplicate samples) with high-grade Ta (Ta-HG) tumours for comparison (Supplementary Table 1). The total number of samples was 278 (107 + 36 = 143 TUR; 55 + 14 = 69 RC; 25 + 4 = 29 NMIBC; 37 BU). An overview of all cohorts, samples, study design, and disease history is provided in Fig. 1a, b, and Supplementary Fig. 1a.
a Summary schematic describing the different sample types and analyses performed in this study. Bi-directional arrows indicate comparison analyses performed. Additional sample information can be found in Supplementary Fig. 1a. b Sankey plot showing an overview of the study and disease history, progressing from left to right. Response is defined as ypT≤1N0M0 and non-response as ≥ypT2N0-3M0. c Heatmap showing k-means consensus clustering of Euclidean distance between pre-NAC TUR samples (N = 107), based on the relative abundance of 1170 proteins (the 25% most variable proteins), defines four clusters with different biological characteristics (CC1-Luminal, N = 42; CC2-Nuclear, N = 7; CC3-Basal, N = 20; CC4-Stroma-rich, N = 38). The heatmap displays the most significant terms selected through g: profiler enrichment analysis. BU and NMIBC samples are shown as median pooled expression. Rows are clustered by Ward’s hierarchical agglomerative clustering method based on Euclidean distance. d Heatmap showing normalised enrichment score of selected biological processes across the four pre-NAC TUR proteomic clusters. CC1-Luminal, N = 42; CC2-Nuclear, N = 7; CC3-Basal, N = 20; CC4-Stroma-rich, N = 38. The full list of processes is listed in the Source Data. e Boxplots show relative protein abundance of key subtype specific MIBC proteins across the four pre-NAC TUR proteomic clusters. Boxplots include the median (centre line in bar) and the third and first quartiles (upper and lower edges, respectively) of protein expression and the whiskers extend to the most extreme data point which is no more than 1.5 times the IQR from the box. Horizontal dotted line represents the base mean for all samples. Asterisks show the pairwise significance of the mean in each group against the mean for all samples as a reference: (*p < 0.05), (**p < 0.01), (***p < 0.001), (****p < 0.0001). Results are derived from a two-sided t-test of the means of each cluster compared to the mean for all samples. Protein abundance values are based on log2 ratio for PSMs abundances divided by the relative PIS value in each TMT plex. For each protein, the median ratio of the five most abundant PSMs was used as relative abundance. CC1-Luminal, N = 42; CC2-Nuclear, N = 7; CC3-Basal, N = 20; CC4-Stroma-rich, N = 38. Source data are provided as a Source Data file. Abbreviations: CC consensus cluster, GCis gemcitabine/cisplatin, GCarbo gemcitabine/carboplatin, NAC neoadjuvant chemotherapy, IQR interquartile range, ITH intra-tumour heterogeneity, TUR transurethral resection, RC radical cystectomy, MIBC muscle-invasive bladder cancer, NMIBC non-muscle invasive bladder cancer, GU genitourinary, BU benign ureter, GSEA gene set enrichment analysis, PIS pooled internal standard, PSM peptide spectrum mass.
All FFPE samples underwent SP3-CTP 11-multiplex Tandem Mass Tag (TMT)-MS analysis, as described previously33,34. The 278 samples were distributed across 31 TMT 11-plexes (Methods and Supplementary Fig. 1b). Results were high-quality, with the MS spectra mapping to 182,899 total peptides and 9769 total proteins, of which 5823 were quantified across all samples submitted (Supplementary Fig. 1c–e). BU technical replicates showed robust quantification across plexes with a median coefficient of variation (CV) < 10% for overall and ubiquitous proteins, while technical replicates (N = 37) of the cell line mixture standard (SuperMix, not used in the analysis in this manuscript) showed higher CV (Supplementary Fig. 1f) probably due to differences in nature (in vitro/in vivo) and preservation (frozen/FFPE) of the cell line mix versus tumour samples, respectively. Principal component analysis (PCA) on the proteome of all MIBC samples revealed no segregation by institution or material age (Supplementary Fig. 1g, h). Thus, we merged the Vancouver and Bern MIBC cohorts into a single MIBC ‘NAC’ cohort (including TUR and RC samples) for further analysis. Samples from TUR and RC clustered separately, suggesting proteomic heterogeneity according to sample type and/or intervening systemic therapy and not collection site (Supplementary Fig. 1i).
The pre-NAC proteome reveals four different MIBC clusters with distinct biology, clinical features, pathological response and survival
We performed consensus clustering on 107 pre-NAC MIBC samples derived from TURs using the 25% most variable proteins across tumours (N = 1170) with robust quantification (CV < 10% across BU technical replicate samples). We chose these thresholds based on previous studies6 and to focus on the most dynamic proteins that are reliably identified by MS. Based on examination of the consensus matrix and delta plots focusing on the variation in consensus cumulative distribution function (CDF) area (Supplementary Fig. 1i–k), four strongly segregated Consensus Clusters (CCs; CC1, CC2, CC3, CC4) with unique biological characteristics, response to NAC, and prognostic relevance were identified (Fig. 1c). We named the clusters based on their protein expression profiles in relation to other known subtype classifiers8,9, and their enrichment of biological signatures.
CC1-Luminal, which encompassed 39% (42/107) of tumours, was enriched for luminal markers (e.g. PPARG, FOXA1) consistent with urothelial differentiation. CC1-Luminal displayed high metabolic and mitochondrial activity by gene set enrichment analysis (GSEA; Fig. 1d, e). CC1-Luminal also had a greater proportion of micropapillary secondary histology, and the highest pathologic response rate to NAC (50%, Fig. 2a). In agreement with other MIBC studies using transcriptomic-based classifiers8,9,11, patients in CC1-Luminal had the best OS (Fig. 2b).
a Stacked barplot showing summary of clinical characteristics of patients in each pre-NAC TUR proteomic CC. CC1-Luminal, N = 42; CC2-Nuclear, N = 7; CC3-Basal, N = 20; CC4-Stroma-rich, N = 38. A detailed description of the cohort is provided in Table 1. b Kaplan-Meier plots show RFS and OS for the four pre-NAC TUR proteomic CCs. CC1-Luminal, N = 41; CC2-Nuclear, N = 7; CC3-Basal, N = 20; CC4-Stroma-rich, N = 38. Note that one CC1-Luminal patient did not have survival data available. Abbreviations: CC consensus cluster, cN clinical node stage, cT clinical tumour stage, CR complete responder, GCarbo gemcitabine carboplatin, GemCis gemcitabine cisplatin, NAC neoadjuvant chemotherapy, OS overall survival, PR partial responder, RFS recurrence-free survival, TUR transurethral resection, ypT pathologic stage following NAC. Source data are provided as a Source Data file.
CC2-Nuclear (7/107 = 7%) was enriched for nuclear features including cellular processes involving mRNA binding, RNA metabolism, and histone lysine N-methyltransferase activity indicative of hypermethylated tumours (Fig. 1d, e). Approximately half of the most common neuroendocrine differentiation markers35 were enriched in CC2-Nuclear (e.g. PSP1, TUBB2B; Supplementary Fig. 2a). CC2-Nuclear captured three out of five histologic small cell/neuroendocrine tumours (Fig. 2a). Patients in CC2-Nuclear had the most favourable oncologic outcomes (Fig. 2b), which differs from prior reports using transcriptomic classifiers6,9. This is probably related to the small sample size of this subgroup and their unexpectedly high CR rate after NAC (80%). One of these patients had only T1N0M0 disease prior to treatment.
CC3-Basal (20/107 = 19%) was characterised by markers of basal cell differentiation (e.g. CD44), keratinisation (e.g. KRT5, CD44) and immune pathway enrichment (e.g. IFN-γ production; Fig. 1d, e). This cluster was 44% female sex (p = 0.1577, Fisher test), a trend consistent with prior reports9,36. These tumours correlated with higher clinical T-stage, squamous differentiation (Fig. 2a), and the worst RFS and OS among clusters (Fig. 2b). These findings align with reports in the context of NAC10 or other neoadjuvant/adjuvant therapy11, where basal tumours were highly immune-infiltrated (particularly with cytotoxic T cells and natural killer cells) and were associated with worse treatment response and survival compared to other subtypes.
Tumours in CC4-Stroma-rich (38/107 = 35%) exhibited high expression of stroma-related proteins, including structural proteins (e.g. COL1A1, DES), extracellular matrix (ECM) signatures, and high immune-infiltration (Fig. 1c–e) in line with previous studies9,11. These patients showed similar OS outcomes (p > 0.05) compared to patients of CC1-Luminal and CC2-Nuclear (Fig. 2b), and more favourable RFS compared to CC3-Basal (Fig. 2b, p < 0.01). Response rates trended to be higher in CC4-Stroma-rich tumours in our study. CC1-Luminal and CC4-Stroma-rich shared the most cluster-defining proteins with BU, including luminal markers like SPINK1, SNCG, and KRT20 for CC1-Luminal, and stromal markers such as COL2A1, COL1A1, and CPA3 for CC4-Stroma-rich, compared to other clusters (Supplementary Fig. 3a). By Euclidean distance their profiles were the closest to the BU, indicating greater urothelial differentiation (CC1-Luminal) and stromal content (CC4-Stroma-rich) of these two clusters than CC2-Nuclear and CC3-Basal (Supplementary Fig. 3e). The existence of a stroma-rich group has also been described by others using transcriptomics, and identifies tumours with intermediate levels of urothelial differentiation, a higher proportion of non-tumour cells and distinct immune infiltrates (e.g. T and B cell populations)9,10. Stroma-rich tumours were reported by others to have a trend toward better pathologic responses to NAC than their basal counterparts11. Their response to immunotherapy appears to be relatively poor in the pembrolizumab-treated PURE-01 trial patients37, as well as the atezolizumab-treated ABACUS trial patients14, which the authors attributed to an immune-excluded phenotype.
We compared our proteomics cluster centroids to previously published global proteomic data generated from FFPE bladder tissue (Supplementary Fig. 2b)31. Xu et al. described three distinct clusters (U.I, U.II, and U.III) in a mixed cohort of MIBC RC and NMIBC TUR samples without chemotherapy history. We found that CC1-Luminal and U.II cluster centroids shared similar profiles, although U.II had notable differences, including low metabolic processes and high activity of RNA and DNA nuclear processes, aligning more closely with CC2-Nuclear. U.III showed centroid similarity with both CC3-Basal and CC4-Stroma-rich, while U.I (enriched for NMIBC) did not align with any of our clusters (Supplementary Fig. 2b). Comparison of our proteomic cluster centroids with those of the consensus RNA-based subtypes9 showed high concordance. CC2-Nuclear aligned with the consensus NE-like subtype, CC3-Basal with Ba/Sq, and CC4-Stroma-rich with Stroma-rich. As expected, CC1-Luminal showed similarity with all three luminal subtypes, with Luminal-Non-Specified (LumNS) being the closest centroid (Supplementary Fig. 2c). Collectively, despite notable cohort differences, our centroids were largely concordant with others, and deviations are probably attributed to cohort composition, sample type (RC versus TUR; Supplementary Fig. 1h), and differences in comparing RNA to protein expression.
The proteome of non-muscle invasive bladder cancer is similar to benign urothelium and distinct from muscle-invasive bladder cancer
The NMIBC cohort was independent of the MIBC cohort and only included patients with high-grade Ta tumours. Specific differences in the proteome of these NMIBC tumours compared to MIBC tumours could reflect global features of tumour aggressiveness crucial in BC progression. Comparing the most differentially expressed proteins (N = 26) of each pre-NAC CC with the NMIBC cohort showed that the CC1-Luminal proteome had the highest overlap with the NMIBC proteome (Supplementary Fig. 3a). This is consistent with other multi-omic studies showing that the majority of NMIBC, and virtually all Ta tumours, express a luminal proteomic and transcriptomic profile38,39. Further analysis revealed proteins (e.g. CD44, MT2A) that are enriched in MIBC compared to NMIBC (Supplementary Fig. 3b). Direct comparison of CC1-Luminal with NMIBC revealed additional unique markers (e.g. FUCA1, PLS1) exclusively upregulated in CC1-Luminal compared to NMIBC (Supplementary Fig. 3b). GSEA showed that three of the five most upregulated processes in MIBC versus NMIBC involved antibody-mediated immune responses. Three of the five most upregulated pathways in NMIBC were associated with metabolic processes (Supplementary Fig. 3c). Despite some discrepancies, likely due to fixation methods, this comparison between NMIBC and MIBC generally agreed with a similar analysis by Stroggilos et al.40 in 117 tumour samples (98 NMIBC and 19 MIBC, Supplementary Fig. 3d).
We analyzed the pooled BU proteome and observed a high degree of similarity of BU with both NMIBC and CC1-Luminal (Supplementary Fig. 3a). Proteins implicated in urothelial differentiation, such as PPARG and FOXA1, were similarly expressed in BU, NMIBC, and CC1-Luminal. Conversely, proteins commonly expressed by stromal cell populations (e.g. COL1A1, DES) were highest in BU and CC4-Stroma-rich, suggesting BU probably contains a mix of urothelium and underlying layers of the ureter (Fig. 1e). Comparing the BU pool with all pre-NAC tumours revealed the proteomic signals of CC1-Luminal and CC4-Stroma-rich were closer to that of the BU than CC2-Nuclear and CC3-Basal (Supplementary Fig. 3e).
Immunohistochemical validation of cluster-defining and prognostic proteins
Based on the availability of high-quality antibodies for immunohistochemistry (IHC), as well as high number of peptides, high protein abundance, cross-sample abundance in the pre-NAC TUR sample, and no isoDoping (Supplementary Fig. 4a), we selected one cluster-defining protein from each CC for IHC validation using a matched TMA (TMA-internal; Supplementary Table 2) with pre-NAC samples. We observed positive correlation values (all p < 0.01) between IHC H-score and MS-derived protein abundances (log2FC) for SNCG (CC1-Luminal, R = 0.79), PSP1 (CC2-Nuclear, R = 0.82), SERPINB3 (CC3-Basal, R = 0.51) and COL1A1 (CC4-Stroma-rich, R = 0.51; Supplementary Fig. 4b, c). COL1A1, specific to stroma cells, was most abundant in CC1-Luminal and CC4-Stroma-rich (R = 0.51). We confirmed the relative selectivity of each protein for its corresponding CC by IHC (Supplementary Fig. 4c). Representative IHC profiles for each CC are shown in Supplementary Fig. 5a–d.
We identified 84 proteins from the MS-derived proteomic data in pre-NAC TUR specimens with the greatest prognostic value. For IHC validation, we selected candidate proteins based on the same criteria as above (Supplementary Fig. 6a, b). We then further selected proteins with clinical and transcriptomic/proteomic clustering relevance, assessing their prognostic value in both pre- and post-NAC setting following uni- and multivariable analysis (Supplementary Fig. 6c). Several luminal markers showed trends toward favourable prognostic value (FGFR3, FOXA1, GATA3, with HRs <1), whereas markers of basal subtypes (KRT5, KRT6A, KRT6B) had unfavourable prognostic value (HRs >1, Supplementary Fig. 6c41,42).
We selected two prognostic markers for further validation that were not associated with any CCs (Supplementary Fig. 6c), based on criteria described above for cluster-defining proteins: one with favourable (MAPK9/JNK2) and one with unfavourable (NES) prognostic value in both the pre- and post-NAC setting. We stained them in the matched TMA-internal (Supplementary Fig. 7a, b; Supplementary Table 2) and the unmatched TMA-validation (Supplementary Fig. 8a, b; Supplementary Table 3). High expression of MAPK9/JNK2 tended to identify better OS in both the TMA-internal (p = 0.14, Supplementary Fig. 7a) and the TMA-external (p = 0.075, Supplementary Fig. 8a), but did not reach statistical significance. Likewise, high expression of NES in tissue was not associated with worse OS in either of the TMAs (p = 0.27, p = 0.38, Supplementary Figs. 7b and 8b).
Given the prognostic relevance of the pre-NAC CCs themselves, we assessed whether the four single cluster-defining markers described above retained prognostic value when split into high versus low H-score (Supplementary Figs. 7c–f and 8c–f). High SNCG H-score was robustly associated with favourable OS in both TMAs (p < 0.005 and p < 0.018). The high PSPI1 H-score approached significance (p = 0.061) in the matched TMA-internal but not in the TMA-validation TMA (p = 0.36). SERPINB3 and COL1A1 did not correlate with prognosis in either TMA. We subsequently integrated the four cluster-defining markers (SNCG, SERPINB3, PSIP1, COL1A) into a combined score (i.e. high above the median, or low below the median, see ‘Methods’). In the TMA-internal data, the combined scores trended towards the identification of worse survival in the ‘high’ group (Supplementary Fig. 7g, p = 0.071) but did not in the TMA-validation (Supplementary Fig. 8g, p = 0.28). Future validation studies will require additional IHC markers in larger independent MIBC cohorts.
Comparison of the proteome before and after neoadjuvant chemotherapy
To study tissue plasticity and biological changes associated with resistance to NAC, we profiled residual tumour tissues from 55 patients with paired pre- and post-NAC tumours (Fig. 1a, b; Supplementary Fig. 1a). 93% were non-responders (≥ypT2NXM0) and 7% were partial responders (ypTis/a/1N0M0), with 58% experiencing a local or distant recurrence within 6 months (Supplementary Table 4).
We performed consensus clustering on 1170 proteins (25% most variable proteins) that exhibited robust quantification across BU technical replicates (CV < 10% in BU technical replicate samples). We identified four post-NAC clusters with different biology and clinical outcomes (Fig. 3a–c, and Supplementary Fig. 9a–c); their correlation with the pre-NAC CCs is shown in Fig. 3d, with proteomic centroid analysis between pre- and post-NAC tissue showing good alignment among clusters as depicted in Supplementary Fig. 9d. The post-NAC (PN)-CC1 proteome (N = 3) resembled that of pre-NAC CC2-Nuclear. Like CC2-Nuclear, PN-CC1 tumours showed upregulated nuclear processes and this cluster captured both pre-NAC CC2-Nuclear cases with residual tumour. Two of three PN-CC1 patients subsequently recurred. The PN-CC2 tumours (N = 15) were characterised by high immune infiltration indicative of adaptive immune responses, and abundant ECM protein expression. Eleven of 13 pre-NAC CC3-Basal cases with residual tumour were classified as PN-CC2. Recurrence was subsequently observed in 10 of 15 cases. PN-CC3 (N = 15) tumours were characterised by moderate immune- and stroma-cell infiltration and were enriched for antibody-mediated immune responses and wound healing pathways. PN-CC3 captured a mix of pre-NAC CC1-Luminal and CC4-Stroma-rich tumours. Nine of 15 PN-CC3 patients recurred. Lastly, PN-CC4 (N = 22) was characterised by high metabolic pathway activity and was derived primarily of pre-NAC CC1-Luminal tumours and, to a lesser extent, pre-NAC CC4-Stroma-rich tumours. Recurrence was observed in 11 of 21 PN-CC4 cases (one patient had missing clinical information at time of RC). Although limited by small sample size, PN-CC1 showed poor RFS and OS, while PN-CC4 showed a trend toward longer RFS (Fig. 3b, c).
a Heatmap showing k-means consensus clustering of Euclidean distance between 55 post-NAC samples from RC (i.e. resistant tumours) based on the 25% most variable proteins (N = 1170) reveals four distinct proteomic clusters (PN-CC1, N = 3; PN-CC2, N = 15; PN-CC3, N = 15; PN-CC4, N = 22) with biological differences and divergence from the original pre-NAC TUR CC. Rows are clustered by Ward’s hierarchical agglomerative clustering method based on Euclidean distance. b Kaplan-Meir plot showing RFS, from time of RC, specific to patients with post-NAC RC tissues. PN-CC1, N = 3; PN-CC2, N = 15; PN-CC3, N = 15; PN-CC4, N = 21. Note that one patient from CC4/PN-CC4 did not have survival data available. c Kaplan-Meir plot showing OS, from time of RC, specific to patients with post-NAC RC tissues. PN-CC1, N = 3; PN-CC2, N = 15; PN-CC3, N = 15; PN-CC4, N = 21. Note that one patient from CC4/PN-CC4 did not have survival data available. d Sankey plot showing CC switch from the TUR specimen and their corresponding post-NAC RC consensus cluster used in Fig. 3a (N = 54). The number of tumour samples included in this analysis is indicated in parenthesis for both pre-NAC and post-NAC specimens. Note that one patient from CC4/PN-CC4 did not have survival data available. e Scatterplot showing differential expression analysis of paired post-NAC RC and pre-NAC TUR on the x-axis, and pooled benign ureter versus TUR on the y-axis. Proteins that are exclusively upregulated in the RC versus the TUR specimen are highlighted. f Lollipop plot showing GSEA analysis enrichment scores of RC versus TUR shows up (positive NES values) and downregulated (negative NES values) pathways with potential relevance in chemoresistance. Pathway size refers to the number of proteins present in the pathway. g Scatter plot showing proteins enriched in the RC samples (N = 55) versus matched TUR (N = 107). Highlighted are proteins which are known to be mediators of platinum-based chemotherapy resistance, druggable targets, or both according to Huang et al. 202174. All p-values are adjusted by the Benjamini-Hochberg method. Abbreviations: CC consensus cluster in pre-NAC TUR tissue, NAC neoadjuvant chemotherapy, NES normalised enrichment score, OS overall survival, PN-CC post-NAC consensus cluster in post-NAC RC tissue, RC radical cystectomy, RFS recurrence-free survival, TUR transurethral resection. Source data are provided as a Source Data file.
Paired analysis of cluster centroids between pre-NAC CCs and PN-CCs showed convergent proteomic profiles, particularly regarding the cluster-defining proteins identified in the pre-NAC samples (Supplementary Fig. 9d, e). However, some differences in biological pathways were noted between the convergent cluster centroids. Supplementary Table 5 highlights key differences between the clusters. For instance, CC1 exhibited higher mitochondrial activity and lower immune response compared to PN-CC4. CC2 had a stronger immune response and lower mitochondrial activity than PN-CC1. CC3 showed increased immune pathway activity compared to PN-CC2 and, CC4 while similar to PN-CC3, also had higher transcription and splicing activity. Pre-NAC CC1-Luminal and CC4-Stroma-rich showed the most plasticity after NAC, with CC1-Luminal patients aligning with PN-CC3 or PN-CC4, while CC4-Stroma-rich patients associated with PN-CC2, PN-CC3, or PN-CC4. Pre-NAC CC2-Nuclear and CC3-Basal were most concordant with the respective post-NAC CC (Fig. 3d; Supplementary Fig. 9f). It is important to acknowledge that differences in proteome between TUR and RC were likely influenced by NAC, the surgical procedure and associated wound healing process (Supplementary Fig. 1h) and it is not possible to distinguish the effect of each of these in isolation. Moreover, inherent differences may exist in the tumour and tumour immune microenvironment when comparing tumour samples obtained from TUR and RC specimens. Additionally, post-NAC samples were closer in proteome composition to the BU than the pre-NAC specimens, likely due to a consequence of their enrichment in stromal proteins (Supplementary Fig. 9g). Therefore, we compared the proteins enriched in post- versus pre-NAC tissue to those enriched in BU versus pre-NAC tissue. This analysis highlights differences in individual proteins, such as CTHRC1 and NNMT that are involved in ECM and epithelial-to-mesenchymal transition, and which are exclusively upregulated post-NAC (Fig. 3e; Supplementary Fig. 9h). At the pathway level, RC specimens were upregulated for ECM and antibody-mediated immune responses, whereas TURs were enriched for epigenomic processes and keratinisation (Fig. 3f).
Tumours resistant to NAC have a high rate of recurrence and progression, leading to poor OS5. Analysis of differential proteome expression between post- and pre-NAC tissue may reveal proteins related to platinum chemoresistance. Proteins enriched in post-NAC residual tumours could represent therapeutic targets to circumvent resistance (Fig. 3g). Using interaction scores >40 from the Drug Gene Interaction Database43, we identified post-NAC enrichment of proteins with roles in platinum-resistance in other cancers, including SOD1 and FN1, both of which are targets of FDA-approved agents. PN-CCs showed differential expression of druggable proteins across clusters for which there are agents that are FDA-approved (Supplementary Fig. 10a) or under investigation (Supplementary Fig. 10b). For example, PN-CC1 tumours showed homogeneous enrichment for MTOR, PARP1, and PARP2, while PN-CC4 showed enrichment of NECTIN4 in about 50% of cases (Supplementary Fig. 10a). These are targets of clinically approved drugs44,45,46. This analysis identifies cluster-specific therapies that could be investigated for clinical utility to treat recurrences.
Integrating the transcriptomic and proteomic landscape of pre-NAC MIBC
The proteomic CCs largely overlapped with previously described RNA subtypes9 based on the 25% most variable proteins and their corresponding transcriptome values (Fig. 4a). Integrating the two transcriptomic platforms used in the study revealed a modest positive correlation (R = 0.34) between protein and RNA expression across the entire MIBC cohort. Notably, the correlation of transcriptome and protein was higher in microarray data compared to RNA-seq (R = 0.41 versus 0.29; Fig. 4b; Supplementary Fig. 11a). CC1-Luminal captured tumours of the Luminal papillary (LumP), Luminal non-specified (LumNS) and Luminal unstable (LumU) classes according to the transcriptomic consensus classifier9 with highly concordant expression of proteins and RNA (Fig. 4a). CC2-Nuclear encompassed all tumours classified as NE-like by RNA expression, in addition to two tumours classified as Basal/Squamous (Ba/Sq) and one as LumU. CC3-Basal was the most concordant with RNA classifiers and had the highest level of RNA cluster separation. All but one CC3-Basal tumours classified as Ba/Sq. CC4-Stroma-rich showed a mix of all RNA subtypes, indicating highly heterogeneous tumours with a large stromal component. This may partly be explained by the loss of the stromal component in the RNA signal (lower right quadrant, Fig. 4a) and the extended half-life of extracellular collagens at the protein level47.
a Heatmap displays proteomic clustering of the pre-NAC TUR specimens (CC1-Luminal, N = 36; CC2-Nuclear, N = 7; CC3-Basal, N = 17; CC4-Stroma-rich N = 35) showing the top 25% most variable proteins (top) and the corresponding mRNA expression when available (RNA-seq in middle row, or microarray in bottom row). RNA data were scaled (Z-score) and those data are shown separately. Samples with no paired transcriptome are not shown. Transcriptomic molecular subtypes as defined by the MIBC consensus cluster classifier (Kamoun et al. 20209) are shown in the top bar alongside cluster separation levels. Values close to 1 indicate high fidelity with the consensus cluster classifier, whereas values close to 0 indicate low confidence. Columns and row orders are based on Fig. 1c. b Boxplot showing mRNA versus protein Spearman’s correlation (median values in red) according to gene expression microarray, RNA-seq, or combined RNA platforms. The box shows the interquartile range containing all points from 1st to 3rd quartile and the whiskers extend to the most extreme data point which is no more than 1.5 times the IQR from the box. Microarray samples, N = 48; RNA-seq samples, N = 47; Combined samples, N = 95. c Scatterplot showing Log2FC of protein and mRNA expression (RNA-seq) in pre-NAC TUR tissue samples from responders (N = 32) versus non-responders (N = 75) to NAC (Pearson’s correlation). Highlighted are proteins and/or genes that are significantly differentially expressed (two-sided adjusted p-val <2.2e−16) at the RNA level and are consistently enriched or depleted at both the protein and RNA level in responders compared to non-responders. Abbreviations: Ba/Sq basal-squamous, IQR interquartile range, LumNS luminal non-specified, LumP luminal papillary, LumU luminal unstable, MIBC muscle-invasive bladder cancer, NAC neoadjuvant chemotherapy, NE-like neuroendocrine like, RC radical cystectomy, RNA-seq RNA sequencing, TUR transurethral resection. Source data are provided as a Source Data file.
Further exploration of the differences in RNA and proteome expression between responders and non-responders to NAC revealed a positive correlation between data types (R = 0.45, p < 0.001, Fig. 4c). While no significant differences were observed between responders and non-responders at the protein level, several transcripts were significantly up- or down-regulated, indicating disparity in expression of RNA and the corresponding protein48.
To delineate the immune component across pre-NAC CCs, we performed ESTIMATE49 and CIBERSORTx50 analyses on both the proteomic and transcriptomic (RNA-seq and microarray) datasets. We confirmed that CC3-Basal and CC4-Stroma-rich had the highest immune and stroma scores compared to CC1-Luminal and CC2-Nuclear (Supplementary Fig. 11b, c). Although this matrix was derived from single cell RNA-seq and its validity is limited when applied to different data modalities, even with batch correction enabled, this constitutes to the best of our knowledge the best approach to deconvolution in BC. Deconvolution of specific immune and other cell populations using CIBERSORTx was most effective with transcriptomic data, as endothelial and mast cells were not delineated using the proteomic data (Supplementary Fig. 11d). CC3-Basal and CC4-Stroma-rich were most enriched for monocytes, and B cells, suggesting these are immune ‘hot’ tumours. Conversely, CC1-Luminal and CC2-Nuclear showed immune ‘cold’ features, except for T cell populations, which appeared equally enriched across all CCs. These results align with those of Xu et al., where U.III (basal) tumours were defined as immune ‘hot’ and U.I (luminal) tumours were classed as ‘cold’31.
To further study the ECM in CC4-Stroma-rich tumours, we visualised the matched expression of RNAs and proteins across the CCs for collagens, proteoglycans, and secreted proteins, collectively defined as the ‘matrisome’51 (Supplementary Fig. 12a, b). In line with Fig. 1d, we confirmed CC4-Stroma-rich to be enriched for collagens and proteoglycans at the protein, and to a lesser extent RNA, level. Notably, CC4-Stroma-rich contained a mix of transcriptome classes, with tumours of luminal subtypes showing the most enrichment for collagens and proteoglycans. Analysis across the cohort identified high levels of calprotectin (S100A8/A9) in CC3-Basal tumours.
Analysis of intra-tumoural proteomic heterogeneity and its association with clinical outcomes
We performed multiregional sampling of morphologically similar pre-NAC tissue sections in 36 of 107 MIBC patients. Unsupervised clustering using all 5828 proteins identified across all pre-NAC TUR duplicate samples showed high inter-tumour heterogeneity, but low ITH across the cohort (Fig. 5a). However, four of 36 sample pairs clustered separately, suggesting some samples had higher ITH despite having similar histologic features (Fig. 5a).
a Heatmap showing hierarchical clustering using all 5828 proteins identified across all pre-NAC TUR samples by Ward’s hierarchical agglomerative clustering method based on Euclidean distance. Duplicate samples (N = 36) are displayed with same colour coding and identified by their TUR_ID. Single samples (N = 72) have no colour coding or ´TUR_ID´ shown in the heatmap. b Dot plot shows intra-tumoural heterogeneity (ITH) scores of all patients with duplicate pre-NAC TUR samples. ITH scores were split according to the median (ITH = 21, horizontal dashed lined) into distant (i.e. high ITH) and close (i.e. low ITH). Pre-NAC TUR proteomic cluster calls were made for the combined proteomic signal of both paired samples. c Representative H&E images of two sets of duplicate pre-NAC TUR samples: one with high ITH and one with low ITH as shown in (b). Scale bars represent 100 µm. d Boxplot showing ITH scores between NAC non-responder (≥ypT2N0-3M0) and partial and complete responder (≤ypTa/1/isN0M0) patients. Boxplots include the median (centre line in bar) and the third and first quartiles (upper and lower edges, respectively) of intra-tumour heterogeneity score and the whiskers extend to the most extreme data point which is no more than 1.5 times the IQR from the box. P value from a one-tailed t-test is indicated. Non-responder, N = 18; partial- and complete-responder, N = 18. e Kaplan–Meier plots show RFS and OS stratified between distant and close (split by median) ITH patient tumours in pre-NAC TUR samples. Close, N = 17; distant, N = 18. Note that one CC1-Luminal (classed as ´close´ in this analysis) patient did not have survival data available. f Sankey plot showing consensus clustering of average protein expression from each pair of the pre-NAC TUR samples (N = 72 total samples from N = 36 patients) compared to that of the individual pre-NAC TUR samples consensus centroids used in Fig. 1c. Number of samples in each cluster are indicated in parenthesis. g PCA plot highlighting morphologically similar matched duplicate pre-NAC TUR samples and their corresponding duplicate post-NAC RC counterparts. Only patients with two pre-NAC TUR and two post-NAC RC samples are included (N = 8). Ellipsoids are based on Khachiyan’s algorithm for minimum-volume enclosing. Samples and ellipses representing specimens from the same patient are assigned the same colour. Samples with ID #641 and #689 show distinct proteomic profiles in the TUR or RC specimens, respectively. Abbreviations: H&E hematoxylin and eosin, IQR interquartile range, ITH intra-tumour heterogeneity, NAC neoadjuvant chemotherapy, RC radical cystectomy, TUR transurethral resection. Source data are provided as Source Data file.
To quantify the level of ITH between paired samples, we assigned each pair an ITH score by Euclidian distance. Despite similar clinical profiles, patients whose tumours had above-median ITH (median = 21, range 12–37) correlated with poorer NAC response rates (t-test, p = 0.029) and survival outcomes than those with below-median ITH (RFS p = 0.0056 and OS p = 0.046; Fig. 5b, d, e). Analysis of the distance for individual pre-NAC TUR samples (N = 72) to the CC centroids confirmed individual sample pre-NAC clusters were robust among pairs, with only four specimens having a CC switch within the same sampled tumour (two from CC1-Luminal to CC4-Stroma-rich and two from CC4-Stroma-rich to CC1-Luminal, Fig. 5f). Three of these four belonged to the high-ITH group (Fig. 5b).
PCA analysis of paired pre- and post-NAC multiregional samples (N = 8) showed high level of heterogeneity between the TUR and RC specimen (Fig. 5g), which agrees with data shown in Supplementary Fig. 1h. Despite the low number of samples included in this part of the study, our data suggest that ITH is a potential biomarker to identify NAC-resistant tumours with poor prognosis. Prior studies using DNA sequencing on pre- and post-NAC-treated tissue have similarly reported that high ITH is associated with worse outcomes52.
Discussion
To assess determinants of NAC response in MIBC, we used pre-NAC specimens to compare the biology of tumours that subsequently responded to systemic therapy to those that did not. Additional analysis of the residual post-NAC chemoresistant tissue enabled the investigation of tumour evolution after treatment, providing insight into the biology of NAC-resistant tumours and potential protein targets for those that recur. Factors associated with response to NAC were a non-basal proteomic subtype with low immune and stromal infiltration. We identified several druggable proteins that are enriched in individual post-NAC proteomic subtypes and may inform subsequent personalised lines of therapy. High proteomic ITH in the pre-NAC specimen was associated with worse outcomes. Table 2 summarises the main biological and clinical features distinguishing MIBC tumours based on their proteome.
We observed that CC3-Basal tumours were associated with poor outcomes and worse pathologic response to NAC. These tumours were immune-infiltrated and overlapped with the Ba/Sq RNA Consensus Cluster subtype. In contrast, CC1-Luminal tumours, which overlapped with luminal RNA subtypes9, had the best outcomes and highest rate of response to NAC. Previous studies reported variable results for NAC response according to transcriptomic subtype. Lotan et al. applied inverse probability weighting to a cohort of 601 patients to facilitate comparison between patients treated with NAC or RC alone16. They found non-luminal tumours by RNA benefited the most from NAC. Luminal tumours were more likely organ-confined and patients with this subtype did well irrespective of treatment. Our study lacks a comparison cohort without NAC treatment so we cannot make similar comparisons. Nonetheless, our findings are consistent with those reported in two large series of platinum-treated patients10,11.
Given established differences between the transcriptome and proteome in cancer28,31,53, there is reason to speculate that RNA-based classification may not robustly align with proteomic subtyping28,31,53. We observed an expected degree of discordance between the transcriptome and proteome, but centroid-based analysis of the proteins driving the pre-NAC clusters showed good correlation with the RNA-based consensus classifier for MIBC9. Our data shows CC4-Stroma-rich shares similarities with that of the Lund SCCL/Mes-Inf subtype, as it presents with histologic, proteomic and transcriptomic evidence of stroma enrichment7. This cluster is highly heterogeneous, capturing different intrinsic cell cancer subtypes of luminal and basal class as highlighted by the mix of RNA-based subtype calls with relatively low RNA cluster separation values (Fig. 4). These proteome-transcriptome discrepancies may be relevant for future use of RNA classifiers in clinical decision making, as our data support the hypothesis that some tumours may be classified as luminal or basal by RNA although the molecular characteristics are not entirely consistent with either. Continuous class scores may provide more granularity to aid in risk stratification and identification of true tumour biology.
As anticipated, the NMIBC proteome was enriched with luminal proteins and was most similar to that of the BU and CC1-Luminal samples. This agrees with NMIBC being luminal at the RNA and protein levels30,39,54. Sjödahl et al.55 described a series of patient-matched NMIBC tumours that progressed to MIBC and found that these tumours were predominantly luminal and rarely shifted in molecular subtype with progression. Exceptions to this observation showed abrupt molecular changes with upregulation of invasion-related pathways and acquisition of TP53 mutations. By comparing NMIBC and CC1-Luminal, we identified biological processes and markers that may play a role in progression, including increased antibody-mediated immunity and changes in the ECM, coupled with marked decreases in metabolic processes. Changes in ECM by way of deregulated proteoglycan and glyocosaminoglycan function are well-documented in the progression of cancers, including BC56.
Our results are in alignment with the dynamic changes shown in other studies using transcriptomic profiling of pre- and post-treated MIBC tissues treated with NAC or single-agent neoadjuvant immunotherapy (NIT). Seiler et al.57 showed that the luminal and luminal-infiltrated classes (most similar to CC1-Luminal and CC4-Stroma-rich in this study) were the most plastic following NAC. They defined a class of post-NAC tumours that they named scar-like due to enrichment of wound healing/scarring processes. The ABACUS (neoadjuvant atezolizumab) and PURE-01 (neoadjuvant pembrolizumab) trials similarly observed that ECM and stromal pathways were upregulated in post-NIT tissues14,37. In agreement with these prior reports14,37, we showed an enrichment in ECM and antibody-mediated immunity, coupled with a decrease in nuclear metabolism, loss of keratinisation, and lower tumour purity in the post-treatment tissue. It is not possible to discern whether some of these changes are intrinsic to sample type (TUR versus RC), and/or the effects of the TUR and subsequent wound healing on the tissue rather than exposure to systemic therapy, since all these reports lack a cohort treated without neoadjuvant therapy. However, different tumours appear to experience differential effects in response to treatment as reflected in the four distinct post-NAC clusters identified here. Importantly, in our analyses, patients whose post-NAC tumours exhibited increased stroma content had more favourable outcomes compared to those without57, suggesting that this phenotype might represent a partial response to therapy. Our proteomic analysis provides further insight into the biology of post-NAC tissue, with individual tumour profiles revealing actionable proteins to inform potential subsequent lines of therapy. For example, MTOR and PARP inhibitors may be relevant in the treatment of the highly aggressive PN-CC1 cluster, with features indicative of NE-differentiation58,59.
Clinical implementation of advanced multi-omics may be limited by its complexity, cost, and equipment availability. While targeted proteomics may offer benefits compared to IHC (e.g. limited IHC antibody availability) and may be implementable in the future26, global proteomics currently primarily aids the prioritised development of candidate IHC markers for clinical testing. In our study we assessed the prognostic value of using IHC to measure cluster-defining markers, as well as two independent proteins (MAPK9, NES) that identified favourable or poor prognosis. When combined into composite IHC-signatures, these markers were able to suggest trends in risk stratification of patients independent of clinical variables. These findings need further validation in larger independent cohorts. In addition, we found that high pre-NAC proteomic ITH was a prognostic factor in our cohort, identifying a subset of patients with worse response to NAC. Others have found similar associations in the context of high ITH by whole-exome sequencing52. These results provide further prognostic information in risk stratification for future work focused on predicting response to NAC.
This study has some limitations. First, we lack a cohort without intervening chemotherapy (RC-only) to account for the impact of the TUR procedure itself on the biology of the residual tumour in the RC specimen. A RC-only control group would allow us to distinguish the prognostic versus predictive value of some of the findings presented in this study. Furthermore, our results may be confounded by the inclusion of patients with nodal metastasis, differences in demographics between the Bern and Vancouver cohorts, clinical variables, RNA platforms, and differences in NAC regimens. We also acknowledge spatial limitations when comparing TMA and bulk protein/RNA, as TMA only accounts for the uppermost layer of the tumour composition. In addition, our comparison between NMIBC and MIBC is confounded by the lack of patient-matched tumours. Overall, our results highlight additional aspects of tumour biology in the context of NAC, demonstrating the added value of BC proteome. We have identified prognostic proteomic markers that need further validation in larger independent cohorts, with a direct comparison to RC-only treated patients.
Methods
Ethics approval and consent
This study complies with all relevant ethical regulations, and obtained approval from the research ethics boards of the University of British Columbia (H09-01628, H16-01490, H20-02726), University of Bern (KEK-Be 219/2015), and Queen’s University (PATH-167-17). Compliance with ethical standards, including the Declaration of Helsinki and the Canadian Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans-TCPS2, was ensured, and all patients provided written informed consent and were not compensated for participating.
Patient samples and study datasets
This study included 107 patients with MIBC (cT2-4N0-3M0) and one patient with cT1N0M0 neuroendocrine BC treated with NAC. We obtained bladder tumour tissue in all patients from: (1) TUR, N = 107 of the bladder tumour prior to neoadjuvant platinum-based chemotherapy, and (2) RC, N = 55 following NAC. This ‘MIBC cohort’ was compiled from two institutions (Vancouver, Canada and Bern, Switzerland). For simplicity, we also included induction chemotherapy under the term NAC for the subset of patients with cN1-3 disease (43% of patients). A detailed diagram highlighting all the samples included in this study can be found in Supplementary Fig. 1a. Comparison statistics between the Vancouver and Bern cohorts were performed with the R-package ‘gtsummary’; Fisher’s exact and Pearson’s Chi-squared tests were used.
An additional NMIBC cohort exclusively containing high-grade Ta specimens60 (N = 25) was included from Queen’s University, Kingston, Canada. A detailed overview of the study design is shown in Fig. 1a and Supplementary Fig. 1a. Detailed clinicopathologic characteristics for the MIBC and NMIBC cohorts are listed in Table 1 and Supplementary Table 1, respectively.
The Vancouver cohort included 49 patients who were diagnosed in 2014–2016. One patient had cT1N0M0 neuroendocrine urothelial carcinoma treated with NAC. Neuroendocrine and squamous differentiation was present as primary histologic subtypes in two additional patients, and secondary subtypes were present in six patients. TUR samples were collected from multiple hospitals across British Columbia, whereas all 55 RC tissues were obtained from Vancouver General Hospital. The Bern cohort included 58 patients between 2000–20138,57. TUR samples were collected from different sites across Switzerland and all RCs were performed at the University of Bern. Secondary histologic subtypes were present in 19 (32.7%) Bern cases. Comprehensive clinicopathologic data are shown in Table 1. Samples of BU; 2 cores per case were harvested from 37 RC specimens matched to the Vancouver cohort and pooled to represent normal urothelium.
Tissue collection and processing were carried out at the Vancouver Prostate Centre (VPC). Pathology review was conducted by two investigators experienced in genitourinary pathology (HZO and RS). Up to 10 tissue cores from tumour-rich areas (>50% tumour content) were obtained for proteomic, transcriptomic, and histologic analysis. Whole transcriptome analysis including assignment to RNA-subtypes8,57 and whole exome analysis has been published elsewhere for the Bern cohort and is not included in this study21.
The NMIBC cohort included 29 tumour samples (four were duplicates) from 25 patients exclusively with high-grade61 Ta urothelial carcinoma from 2008–2016. Ta tumours are, by definition, non-invasive papillary tumours confined to the urothelium. Carcinoma in situ was present in 6 samples but was not profiled. Although this cohort did not include any low-grade or T1 tumours, for simplicity we refer to it as the ‘NMIBC cohort’. A genitourinary pathologist reviewed all cases (DB).
Two tissue microarrays (TMAs) with pre-NAC TUR specimens from MIBC patients were available and used to validate the proteomics findings in this study. The ‘TMA-validation’ was part of a multi-institutional, independent published cohort (N = 80 patients)8 with no sample overlap with this study. The ‘TMA-internal’ included 40 of 49 matched samples from the Vancouver cohort of this study. The Bern cohort did not have a matched TMA available. Both TMAs were made in-house using duplicate 1 mm cores. TMAs were used for IHC to confirm main cluster-defining and prognostic protein markers. The clinicopathologic characteristics of both TMA cohorts are shown in Supplementary Tables 3 and 4, respectively.
Additional proteomics cohorts for validation
Two independent, multi-institutional studies with MS proteomic data from NMIBC and MIBC using frozen40 or FFPE specimens31 were used for validation of the findings of this study.
SP3-clinical tissue LC-MS/MS-based proteomics
Tissue sample acquisition and preparation
At least two adjacent 1.5 mm diameter cores were obtained from each MIBC specimen for proteomic analyses. Histologically similar, but distant (actual distances were not captured in this study), regions of a single tumour block were sampled in duplicate for the specimens included in the analysis of heterogeneity. FFPE cores of the NMIBC cohort were arranged horizontally inside a histology cassette between two sponges. The cassettes were then paraffin-embedded in a horizontal orientation to generate a block of cores. Each block of cores was sectioned, H&E stained, and reviewed a second time for tumour content. Tumour regions across the length of each core were circled on the H&E slide, matched to the block of cores, and cored a second time with a 1.0 mm punch or 0.5 mm punch. The ‘cored cores’ with enriched tumour were placed in their respective microcentrifuge tubes for each sample. Additional details on preparation of FFPE samples, SP3 processing, clean-up, digestion, tandem mass tag labelling of peptides, control samples (SuperMix cell line standard consisting of 13 cancer cell lines and the isoDoped peptides as below), and the composition of the pooled sample internal standard (PIS) made up of aliquots from the tumour and BU samples, high-pH reversed-phase fractionation, and MS data acquisition have been described in detail elsewhere,28,33,34 and detailed methods are included in the Supplementary Methods.
Design of the Tandem Mass Tag (TMT) 11-plex batches, isoDoping libraries and design of synthetic peptides
31 TMT 11-plex batches were included in this study, each 11-plex included a mix of PIS standard, TUR, RC, NMIBC, BU, and SuperMix samples (Supplementary Fig. 1b). BC samples were distributed such that all the samples from the same patient were in the same TMT 11-plex when possible. Due to stochastic sampling in data-dependent acquisition, some peptides of interest may be detected only in some samples due to their low abundance. Consequently, we designed a list of 179 proteins, including those important for transcriptome subtyping, immune markers, therapy resistance, and immune components. A pool of 857 synthetic peptides corresponding to a total of these 179 proteins were spiked into the PIS standard channel of each 11-plex set to boost their MS1 response and ensure that the combined MS1 signal for these peptide ions were above the selection threshold for MS2. Peptides were selected with PeptideRanger62. These spiked-in proteins represent only 1.83% of the total detected (see ‘Results’) and did not affect the overall clustering and sample characterisation. Additional details are reported elsewhere28 and a full list of the isobaric-doped peptides/proteins are included in the Supplementary Files.
Bioinformatics, statistical analysis and reproducibility
No statistical method was used to predetermine sample size in this study. Instead, we included as many patients and samples as were feasible during the initial phase of this study. Source data to generate figures of this manuscript is provided. Our studies make use of well-established computational and statistical analysis software, and these are fully referenced in the main text and below in this section.
Proteomic data analysis
Thermo RAW files were converted to mzML by ThermoRawFileParser (v1.3)63. Spectra were searched using the MSFragger search engine (v3.3)64 in the FragPipe computational platform (v16.0) against the UniProt Human proteome (2021/07/16, 20,371 sequences) database appended to a list of common contaminants. Identification parameters in MSFragger were specified as trypsin digestion, maximum of two missed cleavages allowed, minimum peptide length of 6, precursor mass tolerance of 20 ppm, and a fragment mass tolerance of 20 ppm. MS and MS/MS mass calibration, MS/MS spectral deisotoping, and parameter optimisation were enabled41. Cysteine carbamidomethylation (+57.0215), lysine TMT labelling (+229.1629), and peptide N-terminal TMT labelling (+229.1629) were included as fixed modifications. Methionine oxidation (+15.9949) and serine TMT labelling (+229.1629) were included as variable modifications. The search output was processed by Philosopher workflow42 and Percolator65. Proteins were filtered to 1% protein-level False Discovery Rate (FDR) using the best peptide approach and picked FDR target-decoy strategy. Data from multiple TMT plexes were summarised using TMT-Integrator (max_pep_prob_thres = 0.9, min_pep_prob = 0.9, min_purity = 0.5, min_percent = 0.05 and median centreing normalisation)66. For patients where multiregional samples were profiled the median protein abundance was used for inter-patient analysis. Differential protein expression analysis was performed with the Differential Expression analysis of quantitative MS data (DEqMS) R package67. Gene set enrichment analysis on differential expression was performed on the pre-ranked t-statistic with the R package FGSEA68 (minSize = 2,maxSize = 500), using the GO term signature derived from the Molecular Signature Database (MSigDB)69. Samples were clustered by only selecting proteins with a CV lower than 10% across BU technical replicates and that showed the highest variability (top 25% median absolute deviation) across tumour samples. The ConsensusClusterPlus R package70,71 was used with the following parameters (maxK = 7, clusterAlg = ‘km’, distance = ‘euclidean’, reps = 2500). The number of final clusters used was determined based on visual inspection of consensus matrix and examining the change in consensus CDF area with delta plots.
For each heatmap, protein clusters were analyzed by g:profiler72 enrichment analysis (parameters: organism = ‘hsapiens’, ordered_query = FALSE, multi_query = FALSE, significant = TRUE, exclude_iea = TRUE, measure_underrepresentation = FALSE, evcodes = TRUE, user_threshold = 0.05, correction_method = ‘g_SCS’, domain_scope = ‘annotated’, custom_bg = NULL, numeric_ns = “”, sources = NULL, and source in GO: MF, GO: BP or REACTOME’), and the most relevant terms were selected for visualisation.
Immune cell fractions for each data type were imputed with CIBERSORTx50 using a custom signature matrix (see below) with the following parameters: RNAseq(Batch correction: disabled, Disable quantile normalisation: true, Run mode (relative or absolute): relative), microarray (Batch correction: enabled, Batch correction mode: B-mode, Disable quantile normalisation: true, Run mode (relative or absolute): relative), Protein(Batch correction: enabled, Batch correction mode: B-mode, Disable quantile normalisation: true, Run mode (relative or absolute): relative).
Signature matrix generation for CIBERSORTx
We downloaded publicly available single-cell RNA-seq data from BC tissue73. Genes having summed counts per million (CPM) < 10 across all samples or CV of log2(CPM) < 0.01 were removed, and cells for each cell type were split into training, validation, and test sets in a 70:15:15 ratio. We then generated various signature matrices by sampling different numbers of cells from the training set and using different CIBERSORTx parameter settings. To assess signature matrices, we simulated pseudo-bulk RNA-seq data with various mixtures of cell types from the validation set of cells and then applied CIBERSORTx, using the different signature matrices, to these pseudo-bulk mixtures. We evaluated the errors in the cell fractions estimated by CIBERSORTx and based on these selected a signature matrix to use for analyses. The final signature matrix was generated from a maximum 1000 cells per cell type sampled from the training set and the following CIBERSORTx parameter settings: replicates = 20, sampling = 0.5, fraction = 0, k.max = 50, q.value = 0.01, G.min = 300, G.max = 1000, filter = FALSE, QN = FALSE. The final matrix is available as part of the Source Data file.
Other analyses
Protein association with platinum-based chemotherapy resistance was annotated based on a manually curated database of 937 proteins that have been associated with platinum resistance in the literature74. Protein annotation as a ‘Drug target’ was based on the Drug Gene Interaction Database (DGIdb, v.4.2) database43. Protein classification as ‘Proteoglycans’, ‘Secreted’ and ‘Collagen’ was obtained from MatrisomeDB (v2.0)51. Intra-tumoural heterogeneity was defined as the Euclidean distance between two duplicate pre-NAC samples in the high-dimensional data space of protein expression.
Survival analysis and Kaplan–Meier plots
Hazard ratios were calculated for the univariate and multivariable survival analyses using stratified log-rank tests and Cox proportional hazards regression model with the endpoints of recurrence-free (RFS) and overall survival (OS). The RFS was specified as the time between the date of initial MIBC diagnosis by TUR and the detection of any disease recurrence (local, regional, or distant). The OS was defined as the time between the date of initial MIBC diagnosis and death from any cause. Survival results were shown using Kaplan-Meier curves stratified by several risk factors, including proteomic cluster, intra-tumoural heterogeneity score, or protein expression status.
Multivariable analysis
Prognostic proteomic data arising from pre-NAC samples were adjusted for sex, cT-stage (cT1-4, computed as individual variables), cN-stage (cN1-3, computed as individual variables), and NAC-regimen (gemcitabine/cisplatin versus gemcitabine/carboplatin). For post-NAC RC-samples the multivariable analysis included sex, ypT stage (ypT1-4, computed as individual variables), ypN-stage (ypN0-3, computed as individual variables), and NAC regimen. The stratified log-rank tests and Cox regression models both examined the relationship between the expression of a single protein with RFS and OS as continuous variables. The Benjamini–Hochberg FDR approach was used to account for multiple testing in analyses involving multiple comparisons. The ‘survminer’ and ‘survival’ packages of the statistical software R were used to conduct statistical survival studies.
Transcriptomic analysis and subtyping
Out of 54 pre-NAC samples of the Bern cohort, 45 were profiled by exon array as previously published (GSE87304)8. For the Vancouver cohort, RNA was extracted from at least two 1.5 mm cores using the Promega Maxwell RSC RNA FFPE kit following the manufacturer’s instructions and quantified with a Qubit fluorometer. Total RNA was assessed for fragment size using the Agilent Bioanalyzer 2100 (Psomagen). The TruSeq RNA exome kit (Illumina) was used to prepare sequencing libraries, following the manufacturer’s instructions. Libraries were pooled and sequenced on Illumina NovaSeq6000 S4 v1.5 (paired-end, 150 bp) at a desired output of 50 million total reads per sample (Psomagen).
The final base plus any bases with PHRED quality score <10 and potential adaptor sequences (AGATCGGAAGAGCACACGTCTGAAC for Read 1 and AGATCGGAAGAGCGTCGTGTAGGGA for Read 2) were trimmed from the 3′ end of paired RNA-seq reads using CutAdapt (version 3.5)75. Reads shorter than 24 bases after trimming were discarded. Trimmed reads were aligned to the hg38 human reference genome (GENCODE GRCh38 v39 primary assembly) with added External RNA Control Consortium (ERCC) spike-in control sequences using Spliced Transcripts Alignment to a Reference (STAR; version 2.7.9a)76 with the -peOverlapNbasesMin option set to 6. Duplicate fragments were marked (though still retained) using Samtools markdup (Samtools version 1.14)77 with the optical distance (-d) option set to value 2500, as recommended for patterned flow cells, and the --no-multi-dup option applied. The numbers of aligned fragments per gene were counted using featureCounts (Subread version 2.0.1)78 with the following options: -F GTF -t exon -g gene_id -p -C -d 24. The -p option specifies that fragments (which generally correspond to read pairs) should be counted rather than individual reads. The -C option specifies that read pairs whose ends map to different chromosomes or to different strands of the same chromosome should not be counted. The ‘-d 24’ indicates that only fragments at least 24 bases long should be counted. FastQC (version 0.11.9)79 and QualiMap (version 2.2.2-dev)80 were used for quality control (QC), and various QC metrics for all samples were aggregated into a single report using MultiQC (version 1.11)81. Differential gene expression analysis between patient groups was performed using DESeq282 with raw gene counts. For other analyses, gene-expression data were normalised per sample to transcripts per million (TPM; i.e. divided by gene length in kilobases, then normalised to sum to 1 million).
Correlations of the RNA-seq profiles of individual samples with consensus molecular subtypes of MIBC were computed using the single-sample classifier from the ‘consensusMIBC’ R package9.
Immunohistochemistry (IHC) and scoring
The TMA-internal and TMA-validation were used for testing and validation of cluster-defining and prognostic proteins. Freshly cut 4 µm serial TMA sections were used for IHC, using the Ventana Discovery Ultra autostainer (Ventana Medical Systems, Tucson, Arizona). Antibody quality was assessed using Protein Atlas, and only enhanced antibodies were used for further IHC testing. Details can be found in the link: https://www.proteinatlas.org/about/antibody+validation. Tissue sections underwent antigen retrieval with standard Cell Conditioning 1 (95 °C for 64 min, Ventana Medical Systems) followed by primary antibody incubation at room temperature: 16 min for the anti-NES antibody (Neuromics, #MO15012), or 60 min for anti-SNCG (Abcam #ab52366, dilution 1:200), anti-PSIP1 (Abcam #ab177159, dilution 1:250), anti-SERPINB3 (Abcam #ab180396, dilution 1:100), anti-COL1A1 (Abcam #ab138492, dilution 1:1000), and anti-MAPK9/JNK2 (Abcam #ab76125, dilution 1:400), and detected using a UltraMap DAB anti-Rb or anti-Ms Detection Kit (Ventana Medical Systems). All stained slides were digitised with Leica scanner (Aperio AT2, Leica Microsystems; Concord, Ontario, Canada) at magnification equivalent to 40×. The images were subsequently stored in the Aperio eSlide Manager (Leica Microsystems) at the Vancouver Prostate Centre.
Scoring of all markers was reported using the H-scoring system (intensity × positivity) for the staining observed in the invasive BC cells, and in the case of COL1A1, in all cells (cancer and stroma). Intensity H-scores were reported as 0-none, 1-weak, 2-moderate, 3-strong, and the percentage of positive cells were reported as 0–100% for each core. The averaged H-score between the duplicate cores per case was used for the IHC protein expression data analysis. For analysis of these IHC markers as categorical variables, H-scores were dichotomised using cut-points optimised for best Cox model fit (assessed by Akaike Information Criterion; internally validated on 500 bootstrap samples).
To assess whether key cluster-defining proteins had prognostic relevance we used the TMA-internal and TMA-validation. Here, markers were binarized into high or low score according to a sliding window for H-scores by using maximally selected rank statistics from ‘maxstat’ R package. For individual tumours with multiple cores, the average H-score was used. SNCG high (H-score > 110) versus low expression (H-score 110); PSP1 high (H-score > 130) versus low (H-score ≤ 130); SERPINB3 high (H-score > 1) versus low (H-score ≤ 1); and COL1A1 high (H-score > 127.5) versus low (H-score ≤ 127.5). Similarly, we also tested 2 proteins that showed prognostic value in the Cox proportional hazards regression analysis: NES high (H-score > 0.5) versus low (H-score ≤ 0.5) and MAPK2/JNK2 (H-score > 175) versus low (H-score ≤ 175). To assess the prognostic value of the aggregated 4 cluster identifying proteins (SNCG, PSP1, SERPINB3 and COL1A1) we calculated a risk score defined as the sum of the individual H-scores multiplied by each protein Cox proportional hazard coefficient. This combined score was used to divide the cohort in high risk (score above median) or low risk (score below or equal to the median risk score). Clinicopathologic details for each TMA are described in Supplementary Tables 2 and 3. All biomarkers were independently scored by an investigator experienced in genitourinary pathology (HZO) blinded to clinical variables and outcome data.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The raw RNA sequencing data from the Vancouver cohort is available indefinitely for research use only and under controlled access due to ethics board requirements via the corresponding authors through the European Genome Archive under the study ID EGAS50000000741. Timeframe for data access will be subject to EGA policy and process, and will be dealt with within 1 month. Availability of the data once access has been granted will be one year. The MS proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE83 partner repository with the dataset identifier PXD049435. The following publicly available transcriptomic and proteomic datasets used in this study include: the Bern cohort8 microarray data under accession code GSE87304, the Stroggilos cohort40 supplementary Appendix S1, and the Xu cohort31 via their Table S1. All other data supporting the findings of this study are available within the article, including its Supplementary Data and Source Data files. Source data are provided with this paper.
Code availability
Code used to process RNA-seq data (Vancouver cohort) can be found at https://github.com/joshscurll/Black_Lab/tree/main/Projects/NAC2020/RNAseq/bioinfo_pipeline. All software used to perform these studies is publicly available.
Change history
21 February 2025
A Correction to this paper has been published: https://doi.org/10.1038/s41467-025-57187-w
References
Ferlay, J. et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int. J. Cancer 136, E359–E386 (2015).
Sung, H. et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 71, 209–249 (2021).
Kamat, A. M. et al. Bladder cancer. Lancet 388, 2796–2810 (2016).
Grossman, H. B. et al. Neoadjuvant chemotherapy plus cystectomy compared with cystectomy alone for locally advanced bladder cancer. N. Engl. J. Med. 349, 859–866 (2003).
Bhindi, B. et al. Oncologic outcomes for patients with residual cancer at cystectomy following neoadjuvant chemotherapy: a pathologic stage-matched analysis. Eur. Urol. 72, 660–664 (2017).
Robertson, A. G. et al. Comprehensive molecular characterization of muscle-invasive bladder cancer. Cell 171, 540–556 e525 (2017).
Sjodahl, G., Eriksson, P., Liedberg, F. & Hoglund, M. Molecular classification of urothelial carcinoma: global mRNA classification versus tumour-cell phenotype classification. J. Pathol. 242, 113–125 (2017).
Seiler, R. et al. Impact of molecular subtypes in muscle-invasive bladder cancer on predicting response and survival after neoadjuvant chemotherapy. Eur. Urol. 72, 544–554 (2017).
Kamoun, A. et al. A consensus molecular classification of muscle-invasive bladder cancer. Eur. Urol. 77, 420–433 (2020).
Sjödahl, G. et al. Different responses to neoadjuvant chemotherapy in urothelial carcinoma molecular subtypes. Eur. Urol. https://doi.org/10.1016/j.eururo.2021.10.035 (2021).
Taber, A. et al. Molecular correlates of cisplatin-based chemotherapy response in muscle invasive bladder cancer by integrated multi-omics analysis. Nat. Commun. 11, 4858 (2020).
Choi, W. et al. Identification of distinct basal and luminal subtypes of muscle-invasive bladder cancer with different sensitivities to frontline chemotherapy. Cancer Cell 25, 152–165 (2014).
Necchi, A. et al. Impact of molecular subtyping and immune infiltration on pathological response and outcome following neoadjuvant pembrolizumab in muscle-invasive bladder cancer. Eur. Urol. 77, 701–710 (2020).
Powles, T. et al. Clinical efficacy and biomarker analysis of neoadjuvant atezolizumab in operable urothelial carcinoma in the ABACUS trial. Nat. Med. 25, 1706–1714 (2019).
McConkey, D. J. et al. A prognostic gene expression signature in the molecular classification of chemotherapy-naive urothelial cancer is predictive of clinical outcomes from neoadjuvant chemotherapy: a phase 2 trial of dose-dense methotrexate, vinblastine, doxorubicin, and cisplatin with bevacizumab in urothelial cancer. Eur. Urol. 69, 855–862 (2016).
Lotan, Y. et al. Patients with muscle-invasive bladder cancer with nonluminal subtype derive greatest benefit from platinum based neoadjuvant chemotherapy. J. Urol. 207, 541–550 (2022).
Teo, M. Y. et al. DNA damage response and repair gene alterations are associated with improved survival in patients with platinum-treated advanced urothelial carcinoma. Clin. Cancer Res. 23, 3610–3618 (2017).
Van Allen, E. M. et al. Somatic ERCC2 mutations correlate with cisplatin sensitivity in muscle-invasive urothelial carcinoma. Cancer Discov. 4, 1140–1153 (2014).
Liu, D. et al. Clinical validation of chemotherapy response biomarker ERCC2 in muscle-invasive urothelial bladder carcinoma. JAMA Oncol. 2, 1094–1096, (2016).
Li, Q. et al. ERCC2 helicase domain mutations confer nucleotide excision repair deficiency and drive cisplatin sensitivity in muscle-invasive bladder cancer. Clin. Cancer Res. 25, 977–988 (2019).
Gil-Jimenez, A. et al. Assessment of predictive genomic biomarkers for response to cisplatin-based neoadjuvant chemotherapy in bladder cancer. Eur. Urol. https://doi.org/10.1016/j.eururo.2022.07.023 (2022).
Plimack, E. R. et al. Defects in DNA repair genes predict response to neoadjuvant cisplatin-based chemotherapy in muscle-invasive bladder cancer. Eur. Urol. 68, 959–967 (2015).
Miron, B. et al. Defects in DNA repair genes confer improved long-term survival after cisplatin-based neoadjuvant chemotherapy for muscle-invasive bladder cancer. Eur. Urol. Oncol. 3, 544–547 (2020).
Geynisman, D. M. et al. A phase II trial of risk-enabled therapy after initiating neoadjuvant chemotherapy for bladder cancer (RETAIN). J. Clin. Oncol. 41, 438–438 (2023).
Griffin, J. et al. Verification of molecular subtyping of bladder cancer in the GUSTO clinical trial. J. Pathol. Clin. Res. 10, e12363 (2024).
Doll, S., Gnad, F. & Mann, M. The Case for proteomics and phospho-proteomics in personalized cancer medicine. Proteom. Clin. Appl. 13, e1800113 (2019).
Mertins, P. et al. Proteogenomics connects somatic mutations to signalling in breast cancer. Nature 534, 55–62 (2016).
Asleh, K. et al. Proteomic analysis of archival breast cancer clinical specimens identifies biological subtypes with distinct survival outcomes. Nat. Commun. 13, 896 (2022).
Dong, L. et al. Proteogenomic characterization identifies clinically relevant subgroups of intrahepatic cholangiocarcinoma. Cancer Cell 40, 70–87 e15 (2022).
Yao, Z. et al. Proteogenomics of different urothelial bladder cancer stages reveals distinct molecular features for papillary cancer and carcinoma in situ. Nat. Commun. 14, 5670 (2023).
Xu, N. et al. Integrated proteogenomic characterization of urothelial carcinoma of the bladder. J. Hematol. Oncol. 15, 76 (2022).
Groeneveld, C. S. et al. Proteogenomic characterization of bladder cancer reveals sensitivity to apoptosis induced by tumor necrosis factor-related apoptosis-inducing ligand in FGFR3-mutated tumors. Eur. Urol. https://doi.org/10.1016/j.eururo.2023.05.037 (2023).
Hughes, C. S. et al. Single-pot, solid-phase-enhanced sample preparation for proteomics experiments. Nat. Protoc. 14, 68–85 (2019).
Hughes, C. S. et al. Quantitative profiling of single formalin fixed tumour sections: proteomics for translational research. Sci. Rep. 6, 34949 (2016).
Batista da Costa, J. et al. Molecular characterization of neuroendocrine-like bladder cancer. Clin. Cancer Res. 25, 3908–3920 (2019).
Choi, W. et al. Intrinsic basal and luminal subtypes of muscle-invasive bladder cancer. Nat. Rev. Urol. 11, 400–410 (2014).
Robertson, A. G. et al. Expression-based subtypes define pathologic response to neoadjuvant immune-checkpoint inhibitors in muscle-invasive bladder cancer. Nat. Commun. 14, 2126 (2023).
Hurst, C. D. et al. Stage-stratified molecular profiling of non-muscle-invasive bladder cancer enhances biological, clinical, and therapeutic insight. Cell Rep. Med. 2, 100472 (2021).
Lindskrog, S. V. et al. An integrated multi-omics analysis identifies prognostic molecular subtypes of non-muscle-invasive bladder cancer. Nat. Commun. 12, 2301 (2021).
Stroggilos, R. et al. Proteome-based classification of non-muscle invasive bladder cancer. Int. J. Cancer https://doi.org/10.1002/ijc.32556 (2019).
Yu, F. et al. Identification of modified peptides using localization-aware open search. Nat. Commun. 11, 4065 (2020).
da Veiga Leprevost, F. et al. Philosopher: a versatile toolkit for shotgun proteomics data analysis. Nat. Methods 17, 869–870 (2020).
Cotto, K. C. et al. DGIdb 3.0: a redesign and expansion of the drug-gene interaction database. Nucleic Acids Res. 46, D1068–D1073 (2018).
Adib, E. et al. Phase II clinical trial of everolimus in a pan-cancer cohort of patients with mTOR pathway alterations. Clin. Cancer Res. https://doi.org/10.1158/1078-0432.CCR-20-4548 (2021).
Rimar, K. J., Tran, P. T., Matulewicz, R. S., Hussain, M. & Meeks, J. J. The emerging role of homologous recombination repair and PARP inhibitors in genitourinary malignancies. Cancer, https://doi.org/10.1002/cncr.30631 (2017).
Klumper, N. et al. Membranous NECTIN-4 expression frequently decreases during metastatic spread of urothelial carcinoma and is associated with enfortumab vedotin resistance. Clin. Cancer Res. 29, 1496–1505 (2023).
Verzijl, N. et al. Effect of collagen turnover on the accumulation of advanced glycation end products. J. Biol. Chem. 275, 39027–39031 (2000).
Nusinow, D. P. et al. Quantitative proteomics of the cancer cell line encyclopedia. Cell 180, 387–402 e316 (2020).
Yoshihara, K. et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat. Commun. 4, 2612 (2013).
Newman, A. M. et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat. Biotechnol. 37, 773–782 (2019).
Shao, X. et al. MatrisomeDB 2.0: 2023 updates to the ECM-protein knowledge database. Nucleic Acids Res. 51, D1519–D1530 (2023).
Liu, D. et al. Mutational patterns in chemotherapy resistant muscle-invasive bladder cancer. Nat. Commun. 8, 2193 (2017).
Gry, M. et al. Correlations between RNA and protein expression profiles in 23 human cell lines. BMC Genom. 10, 365 (2009).
Tan, T. Z., Rouanne, M., Tan, K. T., Huang, R. Y. & Thiery, J. P. Molecular subtypes of urothelial bladder cancer: results from a meta-cohort analysis of 2411 tumors. Eur. Urol. https://doi.org/10.1016/j.eururo.2018.08.027 (2018).
Sjodahl, G. et al. Molecular changes during progression from non-muscle invasive to advanced urothelial carcinoma. Int. J. Cancer https://doi.org/10.1002/ijc.32737 (2019).
Winkler, J., Abisoye-Ogunniyan, A., Metcalf, K. J. & Werb, Z. Concepts of extracellular matrix remodelling in tumour progression and metastasis. Nat. Commun. 11, 5120 (2020).
Seiler, R. et al. Divergent biological response to neoadjuvant chemotherapy in muscle-invasive bladder cancer. Clin. Cancer Res. 25, 5082–5093 (2019).
Huan, J., Grivas, P., Birch, J. & Hansel, D. E. Emerging roles for mammalian target of rapamycin (mTOR) complexes in bladder cancer progression and therapy. Cancers 14, https://doi.org/10.3390/cancers14061555 (2022).
Bhattacharjee, S. et al. PARP inhibitors chemopotentiate and synergize with cisplatin to inhibit bladder cancer cell survival and tumor growth. BMC Cancer 22, 312 (2022).
Jackson, C. L. et al. Diagnostic and prognostic implications of a three-antibody molecular subtyping algorithm for non-muscle invasive bladder cancer. J. Pathol. Clin. Res. 8, 143–154 (2022).
Humphrey, P. A., Moch, H., Cubilla, A. L., Ulbright, T. M. & Reuter, V. E. The 2016 WHO classification of tumours of the urinary system and male genital organs-part b: prostate and bladder tumours. Eur. Urol. 70, 106–119 (2016).
Riley, R. M., Spencer Miko, S. E., Morin, R. D., Morin, G. B. & Negri, G. L. PeptideRanger: an R Package to optimize synthetic peptide selection for mass spectrometry applications. J. Proteome Res. 22, 526–531 (2023).
Hulstaert, N. et al. ThermoRawFileParser: modular, scalable, and cross-platform RAW file conversion. J. Proteome Res. 19, 537–542 (2020).
Kong, A. T., Leprevost, F. V., Avtonomov, D. M., Mellacheruvu, D. & Nesvizhskii, A. I. MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry-based proteomics. Nat. Methods 14, 513–520 (2017).
The, M., MacCoss, M. J., Noble, W. S. & Kall, L. Fast and accurate protein false discovery rates on large-scale proteomics data sets with percolator 3.0. J. Am. Soc. Mass Spectrom. 27, 1719–1727 (2016).
Djomehri, S. I. et al. Quantitative proteomic landscape of metaplastic breast carcinoma pathological subtypes and their relationship to triple-negative tumors. Nat. Commun. 11, 1723 (2020).
Zhu, Y. et al. DEqMS: a method for accurate variance estimation in differential protein expression analysis. Mol. Cell Proteom. 19, 1047–1057 (2020).
Korotkevich, G. et al. Fast gene set enrichment analysis. bioRxiv, 060012, https://doi.org/10.1101/060012 (2021).
Liberzon, A. et al. The molecular signatures database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425 (2015).
Monti, S., Tamayo, P., Mesirov, J. & Golub, T. Consensus clustering: a resampling-based method for class discovery and visualization of gene expression microarray data. Mach. Learn. 52, 91–118 (2003).
Wilkerson, M. D. & Hayes, D. N. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics 26, 1572–1573 (2010).
Raudvere, U. et al. g:Profiler: a web server for functional enrichment analysis and conversions of gene lists (2019 update). Nucleic Acids Res. 47, W191–W198 (2019).
Chen, Z. et al. Single-cell RNA sequencing highlights the role of inflammatory cancer-associated fibroblasts in bladder urothelial carcinoma. Nat. Commun. 11, 5077 (2020).
Huang, D. et al. A highly annotated database of genes associated with platinum resistance in cancer. Oncogene 40, 6395–6405 (2021).
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. 2011 17, 3 (2011).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014).
Andrews, S. Babraham Bioinformatics (accessed 12 November 2021) https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
Okonechnikov, K., Conesa, A. & Garcia-Alcalde, F. Qualimap 2: advanced multi-sample quality control for high-throughput sequencing data. Bioinformatics 32, 292–294 (2016).
Ewels, P., Magnusson, M., Lundin, S. & Kaller, M. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047–3048 (2016).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Perez-Riverol, Y. et al. The PRIDE database resources in 2022: a hub for mass spectrometry-based proteomics evidences. Nucleic Acids Res. 50, D543–D552 (2022).
Acknowledgements
M.J.R. received funding from the Deutsche Forschungsgemeinschaft (RE 4782/1-1/2, DFG, German Research Foundation). P.C.B. received funding from BC Canada. Berman DM received funding from Cancer Research Society and BC Canada. S.G. and P.C.B. received funding from a DoD Innovation award (W81XWH2010877). M.E.R. received funding from the Cancer Research Society. The authors would like to thank Alireza Moeen and Sonia Kung for their help with histopathology, and Mathieu Roumiguié and Ariel Qi for their help with sample handling and processing. We thank all contributing patients and their families.
Author information
Authors and Affiliations
Contributions
Conceptualisation: A.C.-S., G.L.N., D.M.B., G.B.M., P.C.B. Methodology: A.C.-S., G.L.N, M.J.R, H.Z.O, J.M.S., K.N., S.E.S, K.N., C.L.J, R.S. Formal analysis: A.C.-S, G.L.N, H.Z.O, J.M.S. Investigation: A.C.-S, G.L.N, M.J.R., H.Z.O, P.C.B. Resources: A.C-S., S.G., G.W., D.M.B, G.B.M, P.C.B. Data curation: A.C.-S, G.L.N, M.J.R. Writing—original draft: A.C.-S, M.J.R. Writing—review & editing: A.C.-S., G.L.N, M.J.R, H.Z.O, J.M.S, S.E.S, C.L.J, M.E.R., R.S., G.B.M., P.C.B. Visualisation: A.C.-S, G.L.N., M.J.R. Supervision: G.B.M, P.C.B. Project administration: A.C.-S, S.E.S., G.L.N. Funding acquisition: K.I., S.G., P.C.B.
Corresponding authors
Ethics declarations
Competing interests
A.C.-S. has consulted for Sustained Therapeutics. D.M.B. serves on the Scientific Advisory Board and has equity in Acrivon Therapeutics. Peter Black has consulted for AbbVie, Astellas Pharma, AstraZeneca, EMD-Serono, Ferring, Pfizer, Janssen Oncology, Bayer, Merck, Sanofi Canada, Bristol-Myers Squibb, Prokarium, Combat, Nonagen, Nanobot, NanOlogy, Photocure, Sumitomo, TerSera, Tolmar, and Verity. He shares a patent with Veracyte which played no role or is linked to the work presented here (Inventors Elai Davicioni, Hussam Al-Deen Ashab, Nicholas Erho, Peter Black; EP 3 571 322 B8 date of publication and mention of the grant patent 07.06.2023). R.S. has consulted for Janssen Oncology. He shares a patent with Veracyte which played no role or is linked to the work presented here (Inventors Elai Davicioni, Hussam Al-Deen Ashab, Nicholas Erho, Peter Black; EP 3 571 322 B8 date of publication and mention of the grant patent 07.06.2023). The rest of authors had no conflict of interest.
Peer review
Peer review information
Nature Communications thanks Jianmin Wang, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Contreras-Sanz, A., Negri, G.L., Reike, M.J. et al. Proteomic profiling identifies muscle-invasive bladder cancers with distinct biology and responses to platinum-based chemotherapy. Nat Commun 16, 1240 (2025). https://doi.org/10.1038/s41467-024-55665-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-024-55665-1
This article is cited by
-
Mechanisms and implications of epithelial cell plasticity in the bladder
Nature Reviews Urology (2025)
-
A functionally tunable magnetic nanochains platform for N-glycoproteomic analysis of extracellular vesicles from ultratrace biofluids
Nature Communications (2025)
-
Emerging technologies and clinical translation of urine-based liquid biopsy in urological cancers
Genes & Genomics (2025)
