Next Article in Journal
The Cholinergic System, the Adrenergic System and the Neuropathology of Alzheimer’s Disease
Next Article in Special Issue
Towards a New, Endophenotype-Based Strategy for Pathogenicity Prediction in BRCA1 and BRCA2: In Silico Modeling of the Outcome of HDR/SGE Assays for Missense Variants
Previous Article in Journal
Feasibility of Human Platelet Lysate as an Alternative to Foetal Bovine Serum for In Vitro Expansion of Chondrocytes
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Transcriptome Patterns of BRCA1- and BRCA2- Mutated Breast and Ovarian Cancers

1
Group of Bioinformatics, Institute of Molecular Biology National Academy of Sciences of Armenia, 0014 Yerevan, Armenia
2
Institute of Biomedicine and Pharmacy, Russian-Armenian University, 0051 Yerevan, Armenia
3
Laboratory of Human Genomics and Immunomics, Institute of Molecular Biology National Academy of Sciences of Armenia, 0014 Yerevan, Armenia
4
Institute of Chemical Biology and Fundamental Medicine, Siberian Branch of the Russian Academy of Sciences (SB RAS), 630090 Novosibirsk, Russia
5
Interdisciplinary Centre for Bioinformatics, University of Leipzig, D-04107 Leipzig, Germany
*
Author to whom correspondence should be addressed.
Int. J. Mol. Sci. 2021, 22(3), 1266; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22031266
Submission received: 17 December 2020 / Revised: 6 January 2021 / Accepted: 7 January 2021 / Published: 28 January 2021
(This article belongs to the Special Issue Advances in the Molecular Basis of BRCA-Associated Cancers)

Abstract

:
Mutations in the BRCA1 and BRCA2 genes are known risk factors and drivers of breast and ovarian cancers. So far, few studies have been focused on understanding the differences in transcriptome and functional landscapes associated with the disease (breast vs. ovarian cancers), gene (BRCA1 vs. BRCA2), and mutation type (germline vs. somatic). In this study, we were aimed at systemic evaluation of the association of BRCA1 and BRCA2 germline and somatic mutations with gene expression, disease clinical features, outcome, and treatment. We performed BRCA1/2 mutation centered RNA-seq data analysis of breast and ovarian cancers from the TCGA repository using transcriptome and phenotype “portrayal” with multi-layer self-organizing maps and functional annotation. The results revealed considerable differences in BRCA1- and BRCA2-dependent transcriptome landscapes in the studied cancers. Furthermore, our data indicated that somatic and germline mutations for both genes are characterized by deregulation of different biological functions and differential associations with phenotype characteristics and poly(ADP-ribose) polymerase (PARP)-inhibitor gene signatures. Overall, this study demonstrates considerable variation in transcriptomic landscapes of breast and ovarian cancers associated with the affected gene (BRCA1 vs. BRCA2), as well as the mutation type (somatic vs. germline). These results warrant further investigations with larger groups of mutation carriers aimed at refining the understanding of molecular mechanisms of breast and ovarian cancers.

1. Introduction

The BRCA1 and BRCA2 genes play an important role in maintaining genomic integrity and tumor suppression through the mediation of DNA repair by homologous recombination and reactivation of replication [1,2]. The prevalence of germline BRCA1 and BRCA2 mutations is 5–15% in breast cancer (BC) [3] and 10–25% of ovarian cancer (OV) patients [4,5,6,7]. Along with the mutation carriers, the cumulative risk is around 70% and 14–44% for breast and ovarian cancers, respectively [8]. The effects of BRCA1 and BRCA2 on survival are also being intensively studied, although with conflicting results. In breast cancers, most of the studies have reported worse clinical prognosis for patients with BRCA1 mutations compared with BRCA2 mutation carriers and non-carriers [9,10,11]. Meanwhile, in other studies, no difference in survival [12], the worse outcome for BRCA2 mutations [13], and favorable outcomes for BRCA1 mutations were found [14]. Similarly, discrepant results were also observed in ovarian cancers. While favorable prognosis was mostly observed for carriers of BRCA1 and BRCA2 mutations [6,15,16], some studies attribute better survival exclusively to BRCA2 [17], and others report no difference [18]. The mechanisms driving the differential association between BRCA1 and BRCA2 mutations with progression and survival in breast and ovarian cancers are not completely understood. On a molecular level, few studies were aimed at addressing this issue. The study by Jazaeri et al. (2002) has shown that there are considerable differences in transcriptome profiles in ovarian cancer patients bearing mutations in BRCA1 or BRCA2, which suggests activation of different mechanisms of the disease [19]. In much the same way, few studies found differences in gene expression profiles in BRCA1-linked breast cancer patients compared with non-carriers [20,21]. At the same time, little is known about transcriptome associated molecular effects of somatic and germline mutations in either of the BRCA genes. In this study, we aimed at systemic evaluation of the relationships between BRCA1 and BRCA2 germline and somatic mutations, associated gene expression, disease clinical features, outcome, and treatment.

2. Results

2.1. Self-organizing Maps (SOM) Portrayal of Transcriptome Landscape in BC and OV

RNA-seq data from 338 ovarian cancer (TCGA-OV) and 694 breast cancer (TCGA-BRCA) samples were analyzed using two-layer self-organizing maps (Table 1). Only primary tumor samples with mutually exclusive mutations were included in the analysis.
Multilayer SOM provided group-specific mean expression portraits averaged over all single-sample portraits per condition studied (Figure 1). These red and blue colored spot-like areas correspond to co-expressed gene clusters with correlated profiles that demonstrate high- or low expression in the different sample groups, respectively.

2.2. Differential Gene Expression Patterns between the BC and OV Cancer Groups

The distribution of up- and down-regulated spot areas in the transcriptome portraits shows considerable variability between studied groups in both diseases. Upregulated genes in mutation carrier vs. non-carrier groups showed mirroring localizations in both diseases, indicating differential, partly antagonistic activation of expression programs. Additionally, common patterns were observed. For example, BRCA1/2 mutations in breast cancer (BC) are uniquely associated with the upregulation of genes located in the right part of the map. For ovarian cancer (OV) one finds a similar result where upregulated genes in mutated samples were found in the lower right part. One also sees that somatic mutations of BRCA1 or BRCA2 associate with additional up-regulated spots compared to their germline counterparts (for example compare gBRCA1 and sBRCA1 portraits for breast cancer, or gBRCA2 vs. sBRCA2 portraits for ovarian cancer). To visualize these differences, we generated “difference portraits” between the groups. The analysis confirmed common gene expression deregulation in BRCA1/2 mutation carrier patients compared to non-carriers both in the breast (Figure 2A) and ovarian cancers (Figure 2B). The strongest deregulation of gene expression is observed for the gBRCA1 group in breast cancer.
Top up- and down-regulated mutation-associated genes are presented in Table 2, as well as Supplementary Data 1 and 2. The results of differential gene expression analysis show that in BC and OV cancers, there is a variation of gene expression profiles in patients with somatic or germline BRCA1/BRCA2 mutations, compared to patients without mutations, as well as between mutation types for the same gene. In the BC_gBRCA1 group, the highest expression was observed for LINC02188, PROM1, ROPN1, GABRP, and FDCSP genes. LINC02188 has been shown to be upregulated in triple-negative breast cancers [22], while PROM1, ROPN1, GABRP, and FDCSP were previously associated with a cancer stem cell signature in a basal-like breast cancer phenotype [23,24].
In BC_sBRCA1 we observed overexpression of S100A7 (psoriasin), a DNA damage-inducible gene associated with poor outcome in estrogen negative cancers [25], meanwhile indicating a good response to etoposide [26]. Breast cancer groups bearing germline and somatic BRCA2 mutations as well as non-mutated patients were characterized by the upregulation of luminal subtype signatures such as ESR1, TFF1, TFF2 [27]. In addition, the BC_gBRCA2 group showed overexpression of RIMS4 indicative of estrogen-positive cancers [28] as well as CNTFR, which is shown to be deregulated in breast cancer [29], however, with unknown clinical impact. Finally, ONECUT2 is upregulated in the BC_sBRCA2 group suggesting an association with cancer stem cell traits and expression of stemness-associated genes [30].
In the OV_BRCA1 group, we observed overexpression of RCCD1, which was previously identified as a susceptibility locus for ovarian cancer [31], and of SUSD2, which promotes cancer metastasis and associates with cisplatin resistance [32]. OV_sBRCA1 samples were, in turn, characterized by the upregulation of KRT16, which is linked to migration, invasion, metastasis, and cancer stemness in ovarian cancer cells [33]. The OV_gBRCA2 group was associated with upregulated OXTR, which is essential for oxytocin-mediated inhibition of cell growth, invasion, and migration [34] by repressing the expressions of MMP2 and VEGF [35]. Finally, OV_sBRCA2 was characterized by the up-regulation of zinc finger proteins (ZNF613, ZNF329, ZNF530, ZNF347), a gene family known to be involved in pathways of carcinogenesis, cancer progression, and metastasis formation [36,37].
Overall, more pronounced differential expression was observed in the case of breast cancer compared to ovarian cancer groups. Most activated genes associate with cancerogenic functions. Particularly, the core functionalities of differentially expressed genes in different groups often overlap and associate with cancer cell stemness, cancer progression, and metastasis development.

2.3. Alteration of Expression of BRCA1/2 and PARP (poly(ADP-ribose) polymerase) Genes between the Groups

Next, we compared the expression of BRCA1 and BRCA2 genes between the different groups. In breast cancer, their lowest expression was observed in patients bearing germline mutations followed by somatic mutation carriers and non-carriers (Figure 3A). On the contrary, in ovarian cancer, the expression of BRCA1 was the lowest in samples with somatic BRCA1 mutations compared to other groups, while the expression of BRCA2 expression showed a similar pattern as in breast cancer (Figure 3B). These results show that mutation-associated decrease of BRCA2 expression is consistently observed in both cancer types, while the expression of BRCA1 varies in a cancer-specific fashion.
In addition to BRCA1 and BRCA2, we were also interested in evaluating the expression of poly(ADP-ribose) polymerase (PARP) genes, representing another key gene family in BC and OV pathophysiology and treatment [38,39,40]. PARP family genes have enzymatic and scaffolding activities and are implicated in DNA repair properties [41]. So, we performed a cluster analysis using expression values of 13 PARP genes available in our datasets (Figure 4). Of three PARP genes (PARP1–3) implicated in DNA repair [38], the highest expression of PARP1 was observed in germline BRCA1 carriers in both breast and ovarian cancers. Moreover, in any group, the expression of at least one PARP1–3 gene was increased. Expression of other PARP genes was higher in somatic BRCA1 and BRCA2 carriers in breast cancer, and in gBRCA1, sBRCA1, and gBRCA2 in ovarian cancers.
Taken together, these results indicate significant variability of transcriptomic programs in breast and ovarian cancers associated with germline and somatic mutations in BRCA1 and BRCA2 genes. Moreover, our results show that the patterns of expression of BRCA genes as well as PARP family genes vary in groups stratified by mutations as well as cancer types.

2.4. Functional Context of Gene Expression Deregulation Associated with BRCA Mutations

Differential gene expression analysis showed only subtle variation between germline and somatic mutations for BRCA1 and BRCA2 genes in ovarian and breast cancers. However, it is well known that even weak alterations of the expression of multiple genes can cause dramatic activity changes of biological pathways if they act in a concerted fashion [42,43,44]. To evaluate the functional context of such concerted changes in gene expression on the level of biological processes and pathways, we used Gene Set Z-score (GSZ) analysis [45].
In breast cancer, germline BRCA1, as well as somatic BRCA1 and BRCA2 mutations, were markedly associated with elevated immune system signatures, while cell proliferation/mitotic cell cycle and DNA repair were exclusively linked with germline mutations in BRCA1. On the other hand, germline BRCA2 mutations showed upregulation in functional categories of protein transport and nucleosome assembly. Consistent with previous reports, germline and, to a lesser extent, somatic BRCA1 mutations were associated with basal breast cancer phenotype-related functional gene sets, while germline and somatic BRCA2 mutations were associated with luminal phenotype-related gene sets [46]. Simultaneously, all mutation-associated groups demonstrate upregulation in epithelial-mesenchymal transition (EMT)-related processes and suggest more aggressive and metastatic cancer subtypes compared with non-mutated breast cancer (Figure 5A).
In contrast, in the ovarian cancer dataset, most of the deregulations in functional gene sets were associated with somatic BRCA1 and germline BRCA2 mutations (Figure 5B). sBRCA1 and gBRCA2 mutation carriers were characterized by the upregulation of inflammatory/immune response, cytokine signaling, and T cell activation, as well as EMT and KRAS signaling. These stromal and inflammatory phenotypes were opposed by more proliferative ones in the case of sBRCA2 mutations, which is strongly associated with cell cycle, cell proliferation, and telomere maintenance functionalities.
Next, we compared to what extent functional gene set deregulation associated with mutation type and mutated gene were shared across the breast and ovarian cancers (Figure 6). The results showed that there is a little overlap of upregulated functional gene sets associated with germline mutations both for BRCA1 and BRCA2 genes. In contrast, the upregulated gene sets considerably overlap in somatic mutation groups for both genes. In breast cancer, the germlines BRCA1, BRCA2, as well as somatic BRCA1 were associated with DNA damage/repair, cell cycle, chromosome maintenance, and transcription. On the other hand, the same mutations in ovarian cancer were associated with adaptive and innate immunity as well as inflammatory gene sets. Interestingly, the functional associations were reversed in somatic BRCA2 groups: In breast cancer, they were associated with immune response, while in ovarian cancer these mutations were associated with chromosome organization and maintenance.
Overall, our results indicate that the transcriptome landscape in breast and ovarian cancers are linked to a range of deregulated biological functions, mainly centered around DNA damage repair/cancer expansion as well as immune/inflammatory response. However, the context of the deregulation of functional processes largely depends on the disease, gene, and mutation types.

2.5. Phenotype and Survival Associations of BRCA1/2 Mutations in Cancers

Next, we created phenotype maps based on regression coefficients between clinical data and transcriptome metagene profiles in studied groups. Phenotype portraits reflect the mutual association between deregulated gene clusters and the respective phenotype characteristics. The overlap of spot areas on phenotype and transcriptome maps indicate the mutual correlation between these parameters. For example, the upregulated gene spot on the right part of the BC_gBRCA1 portrait well overlaps with the corresponding spot on the aneuploidy score portrait (Figure 7A), indicating the positive correlation between gene expression and aneuploidy score.
In breast cancers, mutation-associated transcriptome portraits were associated with increased aneuploidy, the fraction of genome alterations, and microsatellite instability, consistent with previous results demonstrating the causal link between mutations in DNA damage response genes genome instability [47,48]. Meanwhile, both germline and somatic BRCA1 mutations were strongly correlated with basal cancer subtype. The strongest association between transcriptome portraits and advanced neoplasm stage, as well as T (size and extent of the main tumor) and M (degree of metastasis) stages, were observed for gBRCA2. Furthermore, the non-carrier group showed the strongest association with N stage (regional lymph node infiltration) (Figure 7A), which is a negative prognostic factor in non-carrier breast cancer patients [49].
Ovarian cancers associate with advanced histologic grade and disease stage (BRCA1 and BRCA2 mutation groups, respectively), similarly to the results reported by Lakhani et al. (2004) [50]. No positive association was observed for aneuploidy score, microsatellite instability, overall genome alterations, and platinum sensitivity in any of the groups (Figure 7B) consistent with previous reports [51]. Overall, our results suggest that in breast cancers, mutations in BRCA genes stronger associate with phenotypes, while ovarian cancers are characterized by higher heterogeneity.
The analysis of the association of treatment regimens with transcriptome profiles indicated that the patients bearing different types of mutations in BRCA1 and BRCA2 genes have received different treatments (Supplementary Figure S6A,B). The prescribed drugs partially overlapped between breast and ovarian cancer patients; however, no consistency in gene/mutation and drug maps has been observed in studied datasets. For example, the paclitaxel map correlated BC_gBRCA1 and BC_sBRCA1 transcriptome portraits, while in the ovarian cancer dataset, it correlated with OV_sBRCA1 and OV_noBRCA.
Finally, we were interested in how PARP inhibitor treatment-related genes are mapped on transcriptome portraits of studied groups. We populated PARP inhibitor gene signatures from previously published articles [52,53,54] and projected them as white color dots onto transcriptome maps. The results showed that the majority of PARP inhibitor signature genes were located in or near the deregulated gene clusters across the transcriptome landscapes of studied disease groups, implying that signature genes (most affected by the drug or indicative for drug efficacy) are among the most upregulated ones in the respective groups. It has also become apparent that different PARP inhibitor-associated genes map to different deregulated spots depending on the disease, mutated BRCA1 or BRCA2 gene, and mutation type (Figure 8A,B).
We have also analyzed the survival data of BC and OV associated with BRCA mutations. In breast cancer, no differences were observed in overall, disease-free, disease-specific, and progression-free survival in studied groups (Figure 9A). In contrast, overall survival, as well as disease-specific survival in OV, were slightly better in gBRCA1 and sBRCA2 groups, respectively (Figure 9B).

3. Discussion

The prevalence of germline and somatic BRCA1/2 mutations are highest in breast and ovarian cancers and their presence is associated with an indication for PARP-inhibitor therapy. However, there are growing reports indicating differences in clinical outcomes, chemotherapy sensitivity, as well as variability of BRCA gene expression depending on the mutation types, at least in ovarian cancers [55,56]. Less clear information is available for breast cancers [57,58]. Moreover, there are virtually no studies aimed at understanding the influence of mutation types on the molecular mechanisms associated with breast and ovarian cancers. Usually, germline mutation phenotypes are being compared with the sporadic cases, however, without specifically focusing on the somatic mutations affecting the same gene [19,21,59].
In this study, we evaluated perturbations in transcriptome landscapes as a function of the disease (breast vs. ovarian cancers), the gene (BRCA1 vs. BRCA2), and the mutation type (somatic vs. germline) by applying a multilayer self-organizing maps approach on the next generation RNA-sequencing data from TCGA-OC and TCGA-BRCA projects. Our results clearly showed the “multivariate” character of these perturbations.
The most notable changes in our study refer to the mutation types in breast and ovarian cancers. So far, the differential effects of germline and somatic mutations in BRCA1 or BRCA2 genes on the transcriptome and associated functional processes have not been studied to a great extent as opposed to the clinical effects. A recent transcriptomic study reported similarity between germline and somatic mutations of BRCA1/2 genes in breast cancer [60] in agreement with subtle differences in gene expression observed in this study (Figure 1). However, our functional analysis revealed a series of novel details. In ovarian cancers, we find an enhanced immune signature in somatic BRCA1 and germline BRCA2 carriers in agreement with previous reports [61], in which, however, differentiation between mutation types has not been explored. The most profound differences in ovarian cancers were observed between the transcriptomes of germline and somatic BRCA2 mutated cases. While the former upregulated functions were related to the immune response, the latter was associated with chromatin silencing, telomere organization, and cell cycle checkpoints. On the other side, both, germline BRCA1 and BRCA2 mutations in breast cancer were associated with the latter functions, namely DNA damage, proliferation and chromosomal organization, and telomere maintenance, while somatic mutations in those genes were mostly linked to the immune and inflammatory response. Indeed, it has been previously shown that germline BRCA1/2 cancers are less immunologically active, which could be attributed to the compromised immune system because of the mutations [62]. In addition, there is evidence that germline and somatic mutations may have different effects on the structure and function of the encoded protein as well as be linked to different classes of diseases [63]. Finally, we also observed differential expression of BRCA1, BRCA2, and PARP family genes depending on the mutation type, the gene, and the disease. These observations are of special importance since PARP inhibitors are thought to be equally effective in treating BRCA1- or BRCA2-linked ovarian or breast cancers [64].
Recent studies indicate that the function of PARP genes extends beyond maintaining genome stability and is gene-specific. For example, PARP-1, but not PARP-2, is involved in the formation of immunosuppressive macrophage phenotypes in the tumor microenvironment after olaparib treatment and further modulate immunosuppression by enhancing PD-1 expression [65]. Furthermore, PARP-2 is essential for thymocyte development, while PARP-1 regulated Treg development [66]. Even less is known about the biological functions of other PARP genes in the context of regulation of tumor microenvironment and other extra functions. Collectively, this and other data suggest that targeted therapies with PARP inhibitors should consider the intended action on aberrant pathways (e.g., directly activated/deactivated by mutations as discussed), but also the accompanying effects such as modulation of the tumor microenvironment (as indicated by changed “inflammatory signatures”), and provide indications for research into combinations of cytotoxic and immunotherapies to increase treatment efficacy.
Besides differential transcriptome response to germline and somatic mutations, we also observed disparity in deregulation of functional modules associated with the BRCA1 and BRCA2 genes. Both BRCA1 and BRCA2 are crucial in maintaining genomic stability through double-strand DNA repair by homologous recombination [1,2]. However, both being tumor suppressors, BRCA1 and BRCA2 seem to be involved in different stages of DNA damage response and repair [67,68]. Moreover, BRCA1 mutations are usually associated with estrogen receptor deficiency, which is not the case of BRCA2 mutations [69]. Furthermore, BRCA1 serves as a co-transcription factor for OCT-1, c-Myc, ERα, p53, Smad3, and others [70]. Our results also indicate that there are considerable differences in deregulation for gene sets either associated with BRCA1 or BRCA2 genes. Thus, our results agree with previous reports on distinctive mechanisms associated with the dysfunction of BRCA1 or BRCA2 in cancers [19,62].
Finally, we observed considerable differences in the deregulation of transcriptome for the same mutation when comparing breast and ovarian cancers. Consistent with previous reports BRCA mutations in breast cancer were mostly associated with regulation of cell cycle, DNA damage, and cell proliferation in breast cancer [71], and with immune system-related processes in ovarian cancer [61], which may be an indicator of differential role of BRCA1 and BRCA2 in the pathogenesis of these diseases. Previous studies have already suggested mechanisms of how DNA damage may trigger immune response [72,73]; however, the question of why its intensity is higher in ovarian rather than in breast cancer remains open.
The principal limitation of our study is linked to the sample size. Even in large datasets such as TCGA-OV and TCGA-BRCA, there are still a small number of mutation carriers for BRCA1 and BRCA2 genes, especially when they are stratified into groups by the mutation type. This was also a reason for preventing further stratification of biallelic and monoallelic mutations. However, a small sample size is compensated to a certain degree by the exploitation of the multi-layer SOM approach. The SOM-based training ensures that the obtained metagene clusters incorporate the whole variety of the expression profiles existing in the high-dimensional data [74,75]. As a result, we were able to take advantage of various features of the TCGA-OV and TCGA-BRCA datasets regardless of the uneven distribution of the features among the sample groups. Larger samples of mutation carriers will enable a more accurate exploration of the expression profiles and metagene-linked molecular mechanisms associated with the different types of BRCA mutations in breast and ovarian cancers in future investigations.

4. Materials and Methods

4.1. Data Sources and Preprocessing

RNA-seq data for breast (BC, TCGA-BRCA project) and ovarian (OV, TCGA-OV project) cancer samples were obtained from The Cancer Genome Atlas Program (TCGA) repository [76]. OV dataset contains RNA-seq gene expression profiles from 338 cases with serous cystadenocarcinoma; BC dataset contains RNA-seq data from 694 cases. Only RNA-seq data from primary tumor samples from the first vial (“-01A”, refer to the TCGA barcode page) were selected, to ensure that the earliest time point samples were included in further analyses.
Raw RNA-seq counts were filtered to remove transcripts with zero 0 counts across all samples, then were log10 transformed, and centralized against global mean expression values.
Mutation status of BRCA1 and BRCA2 were obtained from the cBio genomics portal [76], which contains four TCGA-BRCA and three TCGA-OV partially overlapping datasets. Information about the germline and somatic status of BRCA1 and BRCA2 mutations was obtained from accompanying publications ([46] and [77] for BC and OV, respectively) as well as from GDAC Firehose data (https://gdac.broadinstitute.org/). As samples without BRCA mutations, we selected those samples that did not have reported BRCA mutations in any of the mentioned cBio portal datasets.
Clinical, survival, and treatment-related information was obtained from the TCGA and cBio data portals.

4.2. Transcriptome Portrayal with Self-Organizing Maps

Transcriptome analysis was performed using a multi-layer self-organizing maps (SOM) machine learning approach described in detail previously [74,75,78,79]. In the present study, we performed two-layer training, each containing the transcriptomic dataset from one TCGA project. The SOM approach represents dimension reduction that translates M = 32,039 gene expression profiles in each layer into K = 2025 (45 × 45) metagenes, each representing a cluster of genes with similar profiles of expression across samples. The SOM training algorithm distributes the N genes over the K metagenes using the minimal Euclidean distance of the expression profiles within and between layers as a similarity measure. It clusters genes with similar profiles in the same or closely located metagenes. Each metagene profile can be interpreted as the mean profile averaged over all gene profiles referring to the respective metagene cluster. The metagene expression values of each sample are visualized (expression portrayal) by arranging them into a two-dimensional 45 × 45 grid and by using maroon to blue colors for maximum to minimum expression values in each of the portraits. Multi-layer SOM ensures that each of the layers is projected into identical SOM-space formed of metagenes that contain the same single genes at the same position of the metagene-grid in each of the layers, which made them directly comparable across the layers [79].

4.3. Gene Set Z-Score Analysis

Functional analysis of co-regulated genes in spot modules was performed using Fisher Exact test and Gene Set Z-score algorithm [46] based on gene set collection available in oposSOM package [78], which includes gene sets obtained from GSEA-repository, “hallmarks of cancer”, NIH Roadmap Epigenomics Consortium for chromatin state-related gene sets, as well as from the retrieved from various publications.

4.4. Phenotype Mapping to Co-Expressed Gene Modules

Phenotype information such as medication, disease stage, and grade was obtained from the TCGA data portal. Phenotype maps were constructed based on coefficients of logistic (categorical variables) or linear regression (numerical variables) between phenotype categories and metagene expression. Correlation between phenotype and transcriptomic maps was assessed using Pearson’s correlation coefficient.

4.5. Survival Analysis

Survival analysis (overall survival, disease-free survival, disease-specific survival, and progression-free survival) depending on the status of BRCA1/2 gene mutations was performed using the Cox proportional hazards regression using survival and survminer R packages.

4.6. Data Availability

The complete analysis results were deposited as supplementary datasets in the open-access repository Zenodo (https://zenodo.org/) [80].

5. Conclusions

Transcriptomic landscapes of breast and ovarian cancers show considerable variation depending on the affected gene (BRCA1 or BRCA2) as well as the mutation type (somatic or germline). Our results warrant further investigations with larger groups of mutation carriers that could pave a way for a better understanding of the fine molecular mechanisms of breast and ovarian cancers.

Supplementary Materials

The following are available online at https://0-www-mdpi-com.brum.beds.ac.uk/1422-0067/22/3/1266/s1, Supplementary Data 1: Pairwise differential expression of genes across groups in breast cancer, Supplementary Data 2: Pairwise differential expression of genes across groups in ovarian cancer, Figure S1: Clustering of deregulated BRCA1- and BRCA2-associated GO Biological Process functional categories in the breast (A) and ovarian (B) cancers, Figure S2: Clustering of deregulated BRCA1- and BRCA2-associated GO Molecular Function functional categories in the breast (A) and ovarian (B) cancers, Figure S3: Clustering of deregulated BRCA1- and BRCA2-associated transcription factor targets in the breast (A) and ovarian (B) cancers, Figure S4: Clustering of deregulated BRCA1- and BRCA2-associated “cancer hallmarks” in the breast (A) and ovarian (B) cancers, Figure S5: Clustering of deregulated BRCA1- and BRCA2-associated GSEA C2 functional categories in the breast (A) and ovarian (B) cancers, Figure S6: Drug transcriptome map relations for breast(A) and ovarian(B) cancers.

Author Contributions

A.A., U.B., M.F., and H.B. contributed to the study design, algorithm development, data analysis, and interpretation; S.H., A.M., A.S., L.N., and M.N. contributed to algorithm development and data analysis; A.A., H.B. drafted the manuscript, M.N. and L.N. contributed to visualization, A.A., U.B., and M.N. responsible for funding acquisition. All authors contributed to manuscript revision, read and approved the submitted version. All authors have read and agreed to the published version of the manuscript.

Funding

The study was supported by the grants of the Russian Foundation for Basic Research (18-515-05004) and the Science Committee of the MESCS of Armenia (18RF-112) as well as the Armenian National Science and Education Fund (ANSEF) (compsci-2324). Computational resources were provided by the Institute of Informatics and Automation Problems NAS RA in the frame of targeted grant 1-8/20TB from the Science Committee of the MESCS of Armenia.

Institutional Review Board Statement

Not applicable since this is a secondary data analysis study.

Informed Consent Statement

Not applicable since this is a secondary data analysis study.

Data Availability Statement

The complete analysis results are deposited as supplementary datasets in the open-access repository Zenodo (https://zenodo.org/record/4326452) as well as supplementary materials of this article.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

  1. Jiang, Q.; Greenberg, R.A. Deciphering the BRCA1 tumor suppressor network. J. Biol. Chem. 2015, 290, 17724–17732. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  2. Lord, C.J.; Ashworth, A. The DNA damage response and cancer therapy. Nature 2012, 481, 287–294. [Google Scholar] [CrossRef] [PubMed]
  3. Armstrong, N.; Ryder, S.; Forbes, C.; Ross, J.; Quek, R.G.W. A systematic review of the international prevalence of BRCA mutation in breast cancer. Clin. Epidemiol. 2019, 11, 543–561. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  4. Ledermann, J.A.; Drew, Y.; Kristeleit, R.S. Homologous recombination deficiency and ovarian cancer. Eur. J. Cancer 2016, 60, 49–58. [Google Scholar] [CrossRef]
  5. Hennessy, B.T.J.; Timms, K.M.; Carey, M.S.; Gutin, A.; Meyer, L.A.; Flake, D.D.; Abkevich, V.; Potter, J.; Pruss, D.; Glenn, P.; et al. Somatic mutations in BRCA1 and BRCA2 could expand the number of patients that benefit from poly (ADP ribose) polymerase inhibitors in ovarian cancer. J. Clin. Oncol. 2010, 28, 3570–3576. [Google Scholar] [CrossRef] [Green Version]
  6. Alsop, K.; Fereday, S.; Meldrum, C.; DeFazio, A.; Emmanuel, C.; George, J.; Dobrovic, A.; Birrer, M.J.; Webb, P.M.; Stewart, C.; et al. BRCA mutation frequency and patterns of treatment response in BRCA mutation-positive women with ovarian cancer: A report from the Australian ovarian cancer study group. J. Clin. Oncol. 2012, 30, 2654–2663. [Google Scholar] [CrossRef] [Green Version]
  7. Zhang, S.; Royer, R.; Li, S.; McLaughlin, J.R.; Rosen, B.; Risch, H.A.; Fan, I.; Bradley, L.; Shaw, P.A.; Narod, S.A. Frequencies of BRCA1 and BRCA2 mutations among 1342 unselected patients with invasive ovarian cancer. Gynecol. Oncol. 2011, 121, 353–357. [Google Scholar] [CrossRef]
  8. Kuchenbaecker, K.B.; Hopper, J.L.; Barnes, D.R.; Phillips, K.A.; Mooij, T.M.; Roos-Blom, M.J.; Jervis, S.; Van Leeuwen, F.E.; Milne, R.L.; Andrieu, N.; et al. Risks of breast, ovarian, and contralateral breast cancer for BRCA1 and BRCA2 mutation carriers. JAMA-J. Am. Med. Assoc. 2017, 317, 2402–2416. [Google Scholar] [CrossRef] [Green Version]
  9. Zhong, Q.; Peng, H.L.; Zhao, X.; Zhang, L.; Hwang, W.T. Effects of BRCA1- And BRCA2-related mutations on ovarian and breast cancer survival: A meta-analysis. Clin. Cancer Res. 2015, 21, 211–220. [Google Scholar] [CrossRef] [Green Version]
  10. Moller, P.; Evans, D.G.; Reis, M.M.; Gregory, H.; Anderson, E.; Maehle, L.; Lalloo, F.; Howell, A.; Apold, J.; Clark, N.; et al. Surveillance for familial breast cancer: Differences in outcome according to BRCA mutation status. Int. J. Cancer 2007, 121, 1017–1020. [Google Scholar] [CrossRef]
  11. Huzarski, T.; Byrski, T.; Gronwald, J.; Górski, B.; Domagała, P.; Cybulski, C.; Oszurek, O.; Szwiec, M.; Gugała, K.; Stawicka, M.; et al. Ten-year survival in patients with BRCA1-negative and BRCA1-positive breast cancer. J. Clin. Oncol. 2013, 31, 3191–3196. [Google Scholar] [CrossRef] [PubMed]
  12. Rennert, G.; Bisland-Naggan, S.; Barnett-Griness, O.; Bar-Joseph, N.; Zhang, S.; Rennert, H.S.; Narod, S.A. Clinical Outcomes of Breast Cancer in Carriers of BRCA1 and BRCA2 Mutations. N. Engl. J. Med. 2007, 357, 115–123. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  13. Goodwin, P.J.; Phillips, K.A.; West, D.W.; Ennis, M.; Hopper, J.L.; John, E.M.; O’Malley, F.P.; Milne, R.L.; Andrulis, I.L.; Friedlander, M.L.; et al. Breast cancer prognosis in BRCA1 and BRCA2 mutation carriers: An international prospective breast cancer family registry population-based cohort study. J. Clin. Oncol. 2012, 30, 19–26. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  14. Cortesi, L.; Masini, C.; Cirilli, C.; Medici, V.; Marchi, I.; Cavazzini, G.; Pasini, G.; Turchetti, D.; Federico, M. Favourable ten-year overall survival in a Caucasian population with high probability of hereditary breast cancer. BMC Cancer 2010, 10, 90. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  15. Safra, T.; Lai, W.C.; Borgato, L.; Nicoletto, M.O.; Berman, T.; Reich, E.; Alvear, M.; Haviv, I.; Muggia, F.M. BRCA mutations and outcome in epithelial ovarian cancer (EOC): Experience in ethnically diverse groups. Ann. Oncol. 2013, 24 (Suppl. S8), viii63–viii68. [Google Scholar] [CrossRef]
  16. Bolton, K.L.; Chenevix-Trench, G.; Goh, C.; Sadetzki, S.; Ramus, S.J.; Karlan, B.Y.; Lambrechts, D.; Despierre, E.; Barrowdale, D.; McGuffog, L.; et al. Association between BRCA1 and BRCA2 Mutations and Survival in Women with Invasive Epithelial Ovarian Cancer. JAMA J. Am. Med. Assoc. 2012, 307, 382–390. [Google Scholar] [CrossRef] [Green Version]
  17. Hyman, D.M.; Zhou, Q.; Iasonos, A.; Grisham, R.N.; Arnold, A.G.; Phillips, M.F.; Bhatia, J.; Levine, D.A.; Aghajanian, C.; Offit, K.; et al. Improved survival for BRCA2-associated serous ovarian cancer compared with both BRCA-negative and BRCA1-associated serous ovarian cancer. Cancer 2012, 118, 3703–3709. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  18. Cunningham, J.M.; Cicek, M.S.; Larson, N.B.; Davila, J.; Wang, C.; Larson, M.C.; Song, H.; Dicks, E.M.; Harrington, P.; Wick, M.; et al. Clinical characteristics of ovarian cancer classified by BRCA1, BRCA2, and RAD51C status. Sci. Rep. 2014, 4, 4026. [Google Scholar] [CrossRef]
  19. Jazaeri, A.A.; Yee, C.J.; Sotiriou, C.; Brantley, K.R.; Boyd, J.; Liu, E.T. Gene expression profiles of BRCA1-linked, BRCA2-linked, and sporadic ovarian cancers. J. Natl. Cancer Inst. 2002, 94, 990–1000. [Google Scholar] [CrossRef]
  20. Hedenfalk, I.A. Gene expression profiling of hereditary and sporadic ovarian cancers reveals unique BRCA1 and BRCA2 signatures. J. Natl. Cancer Inst. 2002, 94, 960–961. [Google Scholar] [CrossRef] [Green Version]
  21. Dudaladava, V.; Jarząb, M.; Stobiecka, E.; Chmielik, E.; Simek, K.; Huzarski, T.; Lubiński, J.; Pamuła, J.; Pękala, W.; Grzybowska, E.; et al. Gene Expression Profiling in Hereditary, BRCA1-linked Breast Cancer: Preliminary Report. Hered. Cancer Clin. Pract. 2006, 4, 28. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  22. Bernstein, B.E.; Stamatoyannopoulos, J.A.; Costello, J.F.; Ren, B.; Milosavljevic, A.; Meissner, A.; Kellis, M.; Marra, M.A.; Beaudet, A.L.; Ecker, J.R.; et al. The NIH roadmap epigenomics mapping consortium. Nat. Biotechnol. 2010, 28, 1045–1048. [Google Scholar] [CrossRef] [Green Version]
  23. Park, S.Y.; Choi, J.H.; Nam, J.S. Targeting cancer stem cells in triple-negative breast cancer. Cancers 2019, 11, 965. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  24. Panaccione, A.; Guo, Y.; Yarbrough, W.G.; Ivanov, S.V. Expression Profiling of Clinical Specimens Supports the Existence of Neural Progenitor-Like Stem Cells in Basal Breast Cancers. Clin. Breast Cancer 2017, 17, 298–306.e7. [Google Scholar] [CrossRef]
  25. Emberley, E.D.; Alowami, S.; Snell, L.; Murphy, L.C.; Watson, P.H. S100A7 (psoriasin) expression is associated with aggressive features and alteration of Jab1 in ductal carcinoma in situ of the breast. Breast Cancer Res. 2004, 6, R308–R315. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  26. Kennedy, R.D.; Gorski, J.J.; Quinn, J.E.; Stewart, G.E.; James, C.R.; Moore, S.; Mulligan, K.; Emberley, E.D.; Lioe, T.F.; Morrison, P.J.; et al. BRCA1 and c-Myc associate to transcriptionally repress psoriasin, a DNA damage-inducible gene. Cancer Res. 2005, 65, 10265–10272. [Google Scholar] [CrossRef] [Green Version]
  27. Smid, M.; Wang, Y.; Zhang, Y.; Sieuwerts, A.M.; Yu, J.; Klijn, J.G.M.; Foekens, J.A.; Martens, J.W.M. Subtypes of breast cancer show preferential site of relapse. Cancer Res. 2008, 68, 3108–3114. [Google Scholar] [CrossRef] [Green Version]
  28. Abba, M.C.; Hu, Y.; Sun, H.; Drake, J.A.; Gaddis, S.; Baggerly, K.; Sahin, A.; Aldaz, C.M. Gene expression signature of estrogen receptor α status in breast cancer. BMC Genom. 2005, 6, 37. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  29. Omokehinde, T.; Johnson, R.W. Gp130 cytokines in breast cancer and bone. Cancers 2020, 12, 326. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  30. Shen, M.; Dong, C.; Ruan, X.; Yan, W.; Cao, M.; Pizzo, D.; Wu, X.; Yang, L.; Liu, L.; Ren, X.; et al. Chemotherapy-induced extracellular vesicle miRNAs promote breast cancer stemness by targeting OneCUT2. Cancer Res. 2019, 79, 3608–3621. [Google Scholar] [CrossRef]
  31. Kar, S.P.; Beesley, J.; Al Olama, A.A.; Michailidou, K.; Tyrer, J.; Kote-Jarai, Z.S.; Lawrenson, K.; Lindstrom, S.; Ramus, S.J.; Thompson, D.J.; et al. Genome-wide meta-analyses of breast, ovarian, and prostate cancer association studies identify multiple new susceptibility loci shared by at least two cancer types. Cancer Discov. 2016, 6, 1052–1067. [Google Scholar] [CrossRef] [Green Version]
  32. Xu, Y.; Miao, C.; Jin, C.; Qiu, C.; Li, Y.; Sun, X.; Gao, M.; Lu, N.; Kong, B. SUSD2 promotes cancer metastasis and confers cisplatin resistance in high grade serous ovarian cancer. Exp. Cell Res. 2018, 363, 160–170. [Google Scholar] [CrossRef] [PubMed]
  33. Huang, W.C.; Jang, T.H.; Tung, S.L.; Yen, T.C.; Chan, S.H.; Wang, L.H. A novel miR-365-3p/EHF/keratin 16 axis promotes oral squamous cell carcinoma metastasis, cancer stemness and drug resistance via enhancing β5-integrin/c-met signaling pathway. J. Exp. Clin. Cancer Res. 2019, 38, 89. [Google Scholar] [CrossRef] [Green Version]
  34. Morita, T.; Shibata, K.; Kikkawa, F.; Kajiyama, H.; Ino, K.; Mizutani, S. Oxytocin inhibits the progression of human ovarian carcinoma cells in vitro and in vivo. Int. J. Cancer 2004, 109, 525–532. [Google Scholar] [CrossRef] [PubMed]
  35. Ji, H.; Liu, N.; Yin, Y.; Wang, X.; Chen, X.; Li, J.; Li, J. Oxytocin inhibits ovarian cancer metastasis by repressing the expression of MMP-2 and VEGF. J. Cancer 2018, 9, 1379–1384. [Google Scholar] [CrossRef] [Green Version]
  36. Cui, X.-X.; Zhou, C.; Lu, H.; Han, Y.-L.; Wang, F.-M.; Fan, W.-R.; Ren, Y.; Zhang, R. High Expression of ZNF93 Promotes Proliferation and Migration of Ovarian Cancer Cells and Relates to Poor Prognosis. Int. J. Clin. Exp. Pathol. 2020, 13, 944–953. [Google Scholar] [PubMed]
  37. Sadłecki, P.; Grabiec, M.; Grzanka, D.; Jóźwicki, J.; Antosik, P.; Walentowicz-Sadłecka, M. Expression of zinc finger transcription factors (ZNF143 and ZNF281) in serous borderline ovarian tumors and low-grade ovarian cancers. J. Ovarian Res. 2019, 12, 23. [Google Scholar] [CrossRef] [Green Version]
  38. Steffen, J.D.; Brody, J.R.; Armen, R.S.; Pascal, J.M. Structural implications for selective targeting of PARPs. Front. Oncol. 2013, 3, 301. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  39. Min, A.; Im, S.A. PARP inhibitors as therapeutics: Beyond modulation of parylation. Cancers 2020, 12, 394. [Google Scholar] [CrossRef] [Green Version]
  40. Weil, M.K.; Chen, A.P. PARP Inhibitor Treatment in Ovarian and Breast Cancer. Curr. Probl. Cancer 2011, 35, 7–50. [Google Scholar] [CrossRef] [Green Version]
  41. Rouleau, M.; Patel, A.; Hendzel, M.J.; Kaufmann, S.H.; Poirier, G.G. PARP inhibition: PARP1 and beyond. Nat. Rev. Cancer 2010, 10, 293–301. [Google Scholar] [CrossRef] [Green Version]
  42. Subramanian, A.; Tamayo, P.; Mootha, V.K.; Mukherjee, S.; Ebert, B.L.; Gillette, M.A.; Paulovich, A.; Pomeroy, S.L.; Golub, T.R.; Lander, E.S.; et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA 2005, 102, 15545–15550. [Google Scholar] [CrossRef] [Green Version]
  43. Arakelyan, A.; Nersisyan, L.; Petrek, M.; Löffler-Wirth, H.; Binder, H. Cartography of pathway signal perturbations identifies distinct molecular pathomechanisms in malignant and chronic lung diseases. Front. Genet. 2016, 7, 79. [Google Scholar] [CrossRef] [Green Version]
  44. Arakelyan, A.; Nersisyan, L.; Poghosyan, D.; Khondkaryan, L.; Hakobyan, A.; Löffler-Wirth, H.; Melanitou, E.; Binder, H. Autoimmunity and autoinflammation: A systems view on signaling pathway dysregulation profiles. PLoS ONE 2017, 12, e0187572. [Google Scholar] [CrossRef]
  45. Mishra, P.P.; Medlar, A.; Holm, L.; Törönen, P. Robust multi-group gene set analysis with few replicates. BMC Bioinform. 2016, 17, 526. [Google Scholar] [CrossRef] [Green Version]
  46. Koboldt, D.C.; Fulton, R.S.; McLellan, M.D.; Schmidt, H.; Kalicki-Veizer, J.; McMichael, J.F.; Fulton, L.L.; Dooling, D.J.; Ding, L.; Mardis, E.R.; et al. Comprehensive molecular portraits of human breast tumours. Nature 2012, 490, 61–70. [Google Scholar]
  47. Tutt, A.; Gabriel, A.; Bertwistle, D.; Connor, F.; Paterson, H.; Peacock, J.; Ross, G.; Ashworth, A. Absence of Brca2 causes genome instability by chromosome breakage and loss associated with centrosome amplification. Curr. Biol. 1999, 9, 1107–1110. [Google Scholar] [CrossRef] [Green Version]
  48. Xu, X.; Weaver, Z.; Linke, S.P.; Li, C.; Gotay, J.; Wang, X.W.; Harris, C.C.; Ried, T.; Deng, C.X. Centrosome amplification and a defective G2-M cell cycle checkpoint induce genetic instability in BRCA1 exon 11 isoform-deficient cells. Mol. Cell 1999, 3, 389–395. [Google Scholar] [CrossRef]
  49. Huszno, J.; Kołosza, Z.; Grzybowska, E.W.A. BRCA1 mutation in breast cancer patients: Analysis of prognostic factors and survival. Oncol. Lett. 2019, 17, 1986–1995. [Google Scholar] [CrossRef]
  50. Lakhani, S.R.; Manek, S.; Penault-Llorca, F.; Flanagan, A.; Arnout, L.; Merrett, S.; McGuffog, L.; Steele, D.; Devilee, P.; Klijn, J.G.M.; et al. Pathology of Ovarian Cancers in BRCA1 and BRCA2 Carriers. Clin. Cancer Res. 2004, 10, 2473–2481. [Google Scholar] [CrossRef] [Green Version]
  51. Testa, U.; Petrucci, E.; Pasquini, L.; Castelli, G.; Pelosi, E. Ovarian Cancers: Genetic Abnormalities, Tumor Heterogeneity and Progression, Clonal Evolution and Cancer Stem Cells. Medicines 2018, 5, 16. [Google Scholar] [CrossRef] [Green Version]
  52. Severson, T.M.; Wolf, D.M.; Yau, C.; Peeters, J.; Wehkam, D.; Schouten, P.C.; Chin, S.F.; Majewski, I.J.; Michaut, M.; Bosma, A.; et al. The BRCA1ness signature is associated significantly with response to PARP inhibitor treatment versus control in the I-SPY 2 randomized neoadjuvant setting. Breast Cancer Res. 2017, 19. [Google Scholar] [CrossRef]
  53. Hill, K.; Rattani, A.; Lietz, C.; Garbutt, C.; Choy, E.; Cote, G.; Culhane, A.; Kelly, A.; Haibe-Kains, B.; Spentzos, D. Machine learning-based prediction of response to PARP inhibition across cancer types. medRxiv 2019, 19007757. [Google Scholar] [CrossRef]
  54. Hu, H.M.; Zhao, X.; Kaushik, S.; Robillard, L.; Barthelet, A.; Lin, K.K.; Shah, K.N.; Simmons, A.D.; Raponi, M.; Harding, T.C.; et al. A Quantitative Chemotherapy Genetic Interaction Map Reveals Factors Associated with PARP Inhibitor Resistance. Cell Rep. 2018, 23, 918–929. [Google Scholar] [CrossRef] [Green Version]
  55. Yang, D.; Khan, S.; Sun, Y.; Hess, K.; Shmulevich, I.; Sood, A.K.; Zhang, W. Association of BRCA1 and BRCA2 mutations with survival, chemotherapy sensitivity, and gene mutator phenotype in patients with ovarian cancer. JAMA J. Am. Med. Assoc. 2011, 306, 1557–1565. [Google Scholar] [CrossRef] [Green Version]
  56. Liu, G.; Yang, D.; Sun, Y.; Shmulevich, I.; Xue, F.; Sood, A.K.; Zhang, W. Differing clinical impact of BRCA1 and BRCA2 mutations in serous ovarian cancer. Pharmacogenomics 2012, 13, 1523–1535. [Google Scholar] [CrossRef] [Green Version]
  57. Sambiasi, D.; Lambo, R.; Pilato, B.; Tommasi, S.; Trojano, G.; Kardhashi, A.; Digennaro, M.; Trojano, V.; Simone, G.; Paradiso, A. BRCA1/2 and clinical outcome in a monoinstitutional cohort of women with hereditary breast cancer. Oncol. Rep. 2014, 31, 365–369. [Google Scholar] [CrossRef] [Green Version]
  58. Garber, J.E. BRCA1/2-associated and sporadic breast cancers: Fellow travelers or not? Cancer Prev. Res. 2009, 2, 100–103. [Google Scholar] [CrossRef] [Green Version]
  59. Hedenfalk, I.; Duggan, D.; Chen, Y.; Radmacher, M.; Bittner, M.; Simon, R.; Meltzer, P.; Gusterson, B.; Esteller, M.; Raffeld, M.; et al. Gene-Expression Profiles in Hereditary Breast Cancer. N. Engl. J. Med. 2001, 344, 539–548. [Google Scholar] [CrossRef]
  60. Bodily, W.R.; Shirts, B.H.; Walsh, T.; Gulsuner, S.; King, M.C.; Parker, A.; Roosan, M.; Piccolo, S.R. Effects of germline and somatic events in candidate BRCA-like genes on breast-tumor signatures. PLoS ONE 2020, 15, e0239197. [Google Scholar] [CrossRef]
  61. Przybytkowski, E.; Davis, T.; Hosny, A.; Eismann, J.; Matulonis, U.A.; Wulf, G.M.; Nabavi, S. An immune-centric exploration of BRCA1 and BRCA2 germline mutation related breast and ovarian cancers. BMC Cancer 2020, 20, 197. [Google Scholar] [CrossRef] [PubMed]
  62. Tartaglia, M.; Martinelli, S.; Stella, L.; Bocchinfuso, G.; Flex, E.; Cordeddu, V.; Zampino, G.; Van Der Burgt, I.; Palleschi, A.; Petrucci, T.C.; et al. Diversity and functional consequences of germline and somatic PTPN11 mutations in human disease. Am. J. Hum. Genet. 2006, 78, 279–290. [Google Scholar] [CrossRef]
  63. Lim, D.; Ngeow, J. Evaluation of the methods to identify patients who may benefit from PARP inhibitor use. Endocr. Relat. Cancer 2016, 23, R267–R285. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  64. Mehta, A.; Cheney, E.; Hartl, C.; Pantelidou, C.; Oliwa, M.; Castrillon, J.; Lin, J.-R.; Hurst, K.; Taveira, M.; Johnson, N.T.; et al. Targeting immunosuppressive macrophages overcomes PARP-inhibitor resistance in BRCA1-associated triple-negative breast cancer. J. Immunother. Cancer 2020, 8, A913. [Google Scholar] [CrossRef]
  65. Yélamos, J.; Moreno-Lama, L.; Jimeno, J.; Ali, S.O. Immunomodulatory roles of PARP-1 and PARP-2: Impact on PARP-centered cancer therapies. Cancers 2020, 12, 392. [Google Scholar] [CrossRef] [Green Version]
  66. Stoppa-Lyonnet, D. The biological effects and clinical implications of BRCA mutations: Where do we go from here? Eur. J. Hum. Genet. 2016, 24, S3–S9. [Google Scholar] [CrossRef]
  67. Roy, R.; Chun, J.; Powell, S.N. BRCA1 and BRCA2: Different roles in a common pathway of genome protection. Nat. Rev. Cancer 2012, 12, 68–78. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  68. Foulkes, W.D.; Metcalfe, K.; Sun, P.; Hanna, W.M.; Lynch, H.T.; Ghadirian, P.; Tung, N.; Olopade, O.I.; Weber, B.L.; McLennan, J.; et al. Estrogen Receptor Status in BRCA1- and BRCA2-Related Breast Cancer: The Influence of Age, Grade, and Histological Type. Clin. Cancer Res. 2004, 10, 2029–2034. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  69. Yi, Y.W.; Kang, H.J.; Bae, I. BRCA1 and oxidative stress. Cancers 2014, 6, 771–795. [Google Scholar] [CrossRef] [Green Version]
  70. Li, Y.; Zhou, X.; Liu, J.; Yin, Y.; Yuan, X.; Yang, R.; Wang, Q.; Ji, J.; He, Q. Differentially expressed genes and key molecules of BRCA1/2-mutant breast cancer: Evidence from bioinformatics analyses. PeerJ 2020, 2020, e8403. [Google Scholar] [CrossRef] [Green Version]
  71. Tian, W.; Shan, B.; Zhang, Y.; Ren, Y.; Liang, S.; Zhao, J.; Zhao, Z.; Wang, G.; Zhao, X.; Peng, D.; et al. Association between DNA damage repair gene somatic mutations and immune-related gene expression in ovarian cancer. Cancer Med. 2020, 9, 2190–2200. [Google Scholar] [CrossRef] [Green Version]
  72. Nastasi, C.; Mannarino, L.; D’incalci, M. DNA damage response and immune defense. Int. J. Mol. Sci. 2020, 21, 7504. [Google Scholar] [CrossRef]
  73. Nakad, R.; Schumacher, B. DNA damage response and immune defense: Links and mechanisms. Front. Genet. 2016, 7, 147. [Google Scholar] [CrossRef] [Green Version]
  74. Wirth, H.; Löffler, M.; von Bergen, M.; Binder, H. Expression cartography of human tissues using self organizing maps. BMC Bioinform. 2011, 12, 306. [Google Scholar] [CrossRef] [Green Version]
  75. Wirth, H.; von Bergen, M.; Binder, H. Mining SOM expression portraits: Feature selection and integrating concepts of molecular function. BioData Min. 2012, 5, 18. [Google Scholar] [CrossRef] [Green Version]
  76. Weinstein, J.N.; Collisson, E.A.; Mills, G.B.; Shaw, K.R.M.; Ozenberger, B.A.; Ellrott, K.; Sander, C.; Stuart, J.M.; Chang, K.; Creighton, C.J.; et al. The cancer genome atlas pan-cancer analysis project. Nat. Genet. 2013, 45, 1113–1120. [Google Scholar] [CrossRef]
  77. Bell, D.; Berchuck, A.; Birrer, M.; Chien, J.; Cramer, D.W.; Dao, F.; Dhir, R.; Disaia, P.; Gabra, H.; Glenn, P.; et al. Integrated genomic analyses of ovarian carcinoma. Nature 2011, 474, 609–615. [Google Scholar]
  78. Löffler-Wirth, H.; Kalcher, M.; Binder, H. OposSOM: R-package for high-dimensional portraying of genome-wide expression landscapes on bioconductor. Bioinformatics 2015, 31, 3225–3227. [Google Scholar] [CrossRef]
  79. Arakelyan, A.; Nersisyan, L.; Nikoghosyan, M.; Hakobyan, S.; Simonyan, A.; Hopp, L.; Loeffler-Wirth, H.; Binder, H. Transcriptome-Guided Drug Repositioning. Pharmaceutics 2019, 11, 677. [Google Scholar] [CrossRef] [Green Version]
  80. Arakelyan, A.; Nikoghosyan, M.; Hakobyan, S.; Simonyan, A.; Nersisyan, L.; Binder, H. BRCA1- and BRCA2- mutation associated transcriptome landscapes in breast and ovarian cancers: Ml-SOM results [Data set]. Zenodo 2020. [Google Scholar] [CrossRef]
Figure 1. Group-specific mean expression portraits of breast (A) and ovarian (B) cancers (see Table 1). Spot-like red and blue areas indicate clusters of genes that are concertedly up- or down-regulated in each of the group portraits. Notably, genes are position-invariant in all portraits, meaning that all portraits can be compared each with another.
Figure 1. Group-specific mean expression portraits of breast (A) and ovarian (B) cancers (see Table 1). Spot-like red and blue areas indicate clusters of genes that are concertedly up- or down-regulated in each of the group portraits. Notably, genes are position-invariant in all portraits, meaning that all portraits can be compared each with another.
Ijms 22 01266 g001
Figure 2. Pairwise differential expression portraits in the breast (A) and ovarian (B) cancers. The red-to-blue color gradient on maps is scaled to indicate up- to down-regulation of gene expression values on a given map, respectively. The number of significantly deregulated genes at different FDR thresholds are given above each map. The largest numbers of de-regulated genes (n > 100 and n > 1000 at Fdr < 0.05) are indicated by * and **, respectively. See also Table 2.
Figure 2. Pairwise differential expression portraits in the breast (A) and ovarian (B) cancers. The red-to-blue color gradient on maps is scaled to indicate up- to down-regulation of gene expression values on a given map, respectively. The number of significantly deregulated genes at different FDR thresholds are given above each map. The largest numbers of de-regulated genes (n > 100 and n > 1000 at Fdr < 0.05) are indicated by * and **, respectively. See also Table 2.
Ijms 22 01266 g002
Figure 3. Mutation-associated BRCA1 and BRCA2 expression profiles in breast (A) and ovarian (B) cancers. gBRCA1 - germline BRCA1 mutations, sBRCA1 - somatic BRCA1 mutations, gBRCA2 - germline BRCA2 mutations, sBRCA1 - somatic BRCA1 mutations, nBRCA1/2 - no BRCA mutations. Significance was calculated using a one-way ANOVA test. Overall, results show the lowest expression levels of BRCA1 observed in germline mutation carrier breast cancer groups and somatic BRCA1 mutation carriers in ovarian cancers. Meanwhile, the lowest BRCA2 expression is observed in germline mutation carriers both in breast and ovarian cancers, with an increase towards non-carriers.
Figure 3. Mutation-associated BRCA1 and BRCA2 expression profiles in breast (A) and ovarian (B) cancers. gBRCA1 - germline BRCA1 mutations, sBRCA1 - somatic BRCA1 mutations, gBRCA2 - germline BRCA2 mutations, sBRCA1 - somatic BRCA1 mutations, nBRCA1/2 - no BRCA mutations. Significance was calculated using a one-way ANOVA test. Overall, results show the lowest expression levels of BRCA1 observed in germline mutation carrier breast cancer groups and somatic BRCA1 mutation carriers in ovarian cancers. Meanwhile, the lowest BRCA2 expression is observed in germline mutation carriers both in breast and ovarian cancers, with an increase towards non-carriers.
Ijms 22 01266 g003
Figure 4. Heatmaps of poly(ADP-ribose) polymerase (PARP) family genes expression in breast (A) and ovarian cancer (B) subgroups stratified by BRCA1 and BRCA2 mutations. The results show that the highest expression of PARP1 is associated with germline BRCA1 mutations, while the expression of other PARP genes varies depending on the disease, gene, and mutation types.
Figure 4. Heatmaps of poly(ADP-ribose) polymerase (PARP) family genes expression in breast (A) and ovarian cancer (B) subgroups stratified by BRCA1 and BRCA2 mutations. The results show that the highest expression of PARP1 is associated with germline BRCA1 mutations, while the expression of other PARP genes varies depending on the disease, gene, and mutation types.
Ijms 22 01266 g004
Figure 5. Clustering of deregulated BRCA1- and BRCA2-associated functional categories in the breast (A) and ovarian (B) cancers. One sees that the most prominent deregulations in breast cancer are linked to gBRCA1 in breast and sBRCA1/gBRCA2 in ovarian cancers, respectively. The gene sets upregulated in the gBRCA1 group include basal-phenotype signatures, immune/inflammatory response. In ovarian cancers, the upregulated functional categories are associated with immunity response (sBRCA1/gBRCA2) and chromosome/telomere maintenance (sBRCA2). More detailed clustering according to gene set types available in Supplementary Figures S1–S5.
Figure 5. Clustering of deregulated BRCA1- and BRCA2-associated functional categories in the breast (A) and ovarian (B) cancers. One sees that the most prominent deregulations in breast cancer are linked to gBRCA1 in breast and sBRCA1/gBRCA2 in ovarian cancers, respectively. The gene sets upregulated in the gBRCA1 group include basal-phenotype signatures, immune/inflammatory response. In ovarian cancers, the upregulated functional categories are associated with immunity response (sBRCA1/gBRCA2) and chromosome/telomere maintenance (sBRCA2). More detailed clustering according to gene set types available in Supplementary Figures S1–S5.
Ijms 22 01266 g005
Figure 6. Overlap of functional gene sets upregulated in mutation associated group transcriptome landscapes in breast (BC) and ovarian (OV) cancers.
Figure 6. Overlap of functional gene sets upregulated in mutation associated group transcriptome landscapes in breast (BC) and ovarian (OV) cancers.
Ijms 22 01266 g006
Figure 7. Phenotype-transcriptome associations in the breast (A) and ovarian (B) cancers. Phenotype maps were generated based on linear regression coefficients between corresponding characteristics and metagene expression profiles across all samples in a given dataset (See Materials and methods section). The similarity between phenotype and transcriptome was calculated for each studied group separately using Pearson’s correlation coefficient.
Figure 7. Phenotype-transcriptome associations in the breast (A) and ovarian (B) cancers. Phenotype maps were generated based on linear regression coefficients between corresponding characteristics and metagene expression profiles across all samples in a given dataset (See Materials and methods section). The similarity between phenotype and transcriptome was calculated for each studied group separately using Pearson’s correlation coefficient.
Ijms 22 01266 g007
Figure 8. Olaparib gene signature distribution on the transcriptome landscapes in the breast (A) and ovarian (B) cancers. White dots represent Olaparib-related genes mapped to the corresponding metagenes on a given transcriptome landscape. It can be noted that signature genes are predominantly mapped on or around upregulated (red) spots on each portrait. This suggests different action effects of Olaparib depending on the disease, mutation type, and gene affected.
Figure 8. Olaparib gene signature distribution on the transcriptome landscapes in the breast (A) and ovarian (B) cancers. White dots represent Olaparib-related genes mapped to the corresponding metagenes on a given transcriptome landscape. It can be noted that signature genes are predominantly mapped on or around upregulated (red) spots on each portrait. This suggests different action effects of Olaparib depending on the disease, mutation type, and gene affected.
Ijms 22 01266 g008
Figure 9. Survival associated with BRCA1 and BRCA2 mutations in the breast (A) and ovarian cancers (B). Significance was calculated using a Cox proportional hazards regression model.
Figure 9. Survival associated with BRCA1 and BRCA2 mutations in the breast (A) and ovarian cancers (B). Significance was calculated using a Cox proportional hazards regression model.
Ijms 22 01266 g009
Table 1. Sample groups in the breast (TCGA-BRCA) and ovarian (TCGA-OV) cancer datasets were used as separate layers in self-organizing maps (SOM) analysis.
Table 1. Sample groups in the breast (TCGA-BRCA) and ovarian (TCGA-OV) cancer datasets were used as separate layers in self-organizing maps (SOM) analysis.
TCGA-OV RNA-Seq Dataset
n (%)
TCGA-BRCA RNA-Seq Dataset
n (%)
Germline BRCA119 (5.6%)12 (1.7%)
Somatic BRCA114 (4.1%)28 (4.0%)
Germline BRCA216 (4.7%)15 (2.2%)
Somatic BRCA28 (2.4%)26 (3.7%)
No BRCA mutations281 (83.1%)613 (88.3%)
Table 2. Groupwise analysis of differential expression analysis in breast and ovarian cancers.
Table 2. Groupwise analysis of differential expression analysis in breast and ovarian cancers.
GroupsBrest Cancer
ID
SymbollogFCFdrOvarian Cancer
ID
SymbollogFCFdr
gBRCA1 vs. gBRCA2ENSG00000135069
ENSG00000102854
ENSG00000162989
ENSG00000173467
FDCSP
PSAT1
KCNJ3
AGR3
5.12
3.86
5.03
5.16
0.06
0.04 *
0.05 *
0.04 *
ENSG00000204934
ENSG00000166965
ATP6V0E2-AS1
RCCD1
1.11
1.01
0.12
0.2
gBRCA1 vs. noBRCAENSG00000181617
ENSG00000094755
ENSG00000160182
ENSG00000091831
FDCSP
GABRP
TFF1
ESR1
4.58
4.57
−4.35
−4.58
0.01 **
0 **
0 **
0 **
ENSG00000099994
ENSG00000076344
ENSG00000267327
ENSG00000259129
SUSD2
RGS11

LINC00648
1.37
1.3
−1.32
−1.67
0.23
0.23
0.06
0.11
gBRCA1 vs. sBRCA1ENSG00000261175
ENSG00000171243
ENSG00000091831
ENSG00000101210
LINC02188
SOSTDC1
ESR1
EEF1A2
3.3
3.06
−2.94
−3.3
0.14
0.15
0.19
0.13
----
gBRCA1 vs. sBRCA2ENSG00000094755
ENSG00000261175
ENSG00000160180
ENSG00000091831
GABRP
LINC02188
TFF3
ESR1
4.77
4.23
−3.72
−4.11
0.01 **
0.01 **
0.03 *
0.02 *
ENSG00000149527
ENSG00000275426
ENSG00000256087
ENSG00000176024
PLCH2

ZNF432
ZNF613
1.53
1.07
−0.81
−0.87
0.19
0.19
0.19
0.2
gBRCA2 vs. noBRCAENSG00000122756
ENSG00000176406
ENSG00000237940
ENSG00000159763
CNTFR
RIMS2
LINC01238
PIP
2.16
1.96
−1.94
−2.41
0.09
0.13
0.13
0.22
ENSG00000180914
ENSG00000099953
ENSG00000100473
ENSG00000255571
OXTR
MMP11
COCH
MIR9-3HG
1.84
1.81
−1.1
−1.64
0.07
0.12
0.23
0.12
gBRCA2 vs. sBRCA2ENSG00000122756
ENSG00000135097
ENSG00000139618
ENSG00000196092
CNTFR
MSI1
BRCA2
PAX5
2.36
2.22
−1.37
−1.74
0.24
0.21
0.21
0.2
ENSG00000180914
ENSG00000101445
ENSG00000147536
ENSG00000123219
OXTR
PPP1R16B
GINS4
CENPK
3.24
2.4
−1.02
−1.15
0.11
0.15
0.2
0.2
sBRCA1 vs. gBRCA2ENSG00000159184
ENSG00000188257
ENSG00000101098
ENSG00000263639
HOXB13
PLA2G2A
RIMS4
MSMB
2.81
2.49
−3.2
−3.65
0.12
0.1
0.12
0.1
----
sBRCA1 vs. noBRCAENSG00000178372
ENSG00000186832
ENSG00000256612
ENSG00000153002
CALML5
KRT16
CYP2B7P
CPB1
2.11
1.82
−2.27
−2.54
0.01 **
0.05 *
0.03 *
0 **
ENSG00000186832
ENSG00000200087
ENSG00000072041
ENSG00000130294
KRT16
SNORA73B
SLC6A15
KIF1A
1.93
1.64
−1.87
−2.39
0.17
0.18
0.13
0.16
sBRCA1 vs. sBRCA2----ENSG00000155966
ENSG00000246695
ENSG00000012048
ENSG00000180071
AFF2
RASSF8-AS1
BRCA1
ANKRD18A
2.24
1.72
−1.75
−1.91
0.18
0.22
0.17
0.09
sBRCA2 vs. noBRCAENSG00000119547
ENSG00000089692
ENSG00000213759
ENSG00000082175
ONECUT2
LAG3
UGT2B11
PGR
1.34
1.04
−1.77
−2.13
0.24
0.23
0.12
0.1
ENSG00000196787
ENSG00000204860
ENSG00000259439
ENSG00000187908
HIST1H2AG
FAM201A
LINC01833
DMBT1
1.89
1.46
−1.85
−1.92
0.15
0 **
0.12
0.04 *
*-fdr ≤ 0.05, **-fdr ≤ 0.01.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Arakelyan, A.; Melkonyan, A.; Hakobyan, S.; Boyarskih, U.; Simonyan, A.; Nersisyan, L.; Nikoghosyan, M.; Filipenko, M.; Binder, H. Transcriptome Patterns of BRCA1- and BRCA2- Mutated Breast and Ovarian Cancers. Int. J. Mol. Sci. 2021, 22, 1266. https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22031266

AMA Style

Arakelyan A, Melkonyan A, Hakobyan S, Boyarskih U, Simonyan A, Nersisyan L, Nikoghosyan M, Filipenko M, Binder H. Transcriptome Patterns of BRCA1- and BRCA2- Mutated Breast and Ovarian Cancers. International Journal of Molecular Sciences. 2021; 22(3):1266. https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22031266

Chicago/Turabian Style

Arakelyan, Arsen, Ani Melkonyan, Siras Hakobyan, Uljana Boyarskih, Arman Simonyan, Lilit Nersisyan, Maria Nikoghosyan, Maxim Filipenko, and Hans Binder. 2021. "Transcriptome Patterns of BRCA1- and BRCA2- Mutated Breast and Ovarian Cancers" International Journal of Molecular Sciences 22, no. 3: 1266. https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22031266

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop