Skip to main content
Erschienen in: Journal of Translational Medicine 1/2023

Open Access 01.12.2023 | Research

Pan-cancer onco-signatures reveal a novel mitochondrial subtype of luminal breast cancer with specific regulators

verfasst von: Ines Simeone, Michele Ceccarelli

Erschienen in: Journal of Translational Medicine | Ausgabe 1/2023

Abstract

Background

Somatic alterations in cancer cause dysregulation of signaling pathways that control cell-cycle progression, apoptosis, and cell growth. The effect of individual alterations in these pathways differs between individual tumors and tumor types. Recognizing driver events is a complex task requiring integrating multiple molecular data, including genomics, epigenomics, and functional genomics. A common hypothesis is that these driver events share similar effects on the hallmarks of cancer. The availability of large-scale multi-omics studies allows for inferring these common effects from data. Once these effects are known, one can then deconvolve in every individual patient whether a given genomics alteration is a driver event.

Methods

Here, we develop a novel data-driven approach to identify shared oncogenic expression signatures among tumors. We aim to identify gene onco-signature for classifying tumor patients in homogeneous subclasses with distinct prognoses and specific genomic alterations. We derive expression pan-cancer onco-signatures from TCGA gene expression data using a discovery set of 9107 primary pan-tumor samples together with respective matched mutational data and a list of known cancer-related genes from COSMIC database.

Results

We use the derived ono-signatures to state their prognostic significance and apply them to the TCGA breast cancer dataset as proof of principle of our approach. We uncover a “mitochondrial” sub-group of Luminal patients characterized by its biological features and regulated by specific genetic modulators. Collectively, our results demonstrate the effectiveness of onco-signatures-based methodologies, and they also contribute to a comprehensive understanding of the metabolic heterogeneity of Luminal tumors.

Conclusions

These findings provide novel genomics evidence for developing personalized breast cancer patient treatments. The onco-signature approach, demonstrated here on breast cancer, is general and can be applied to other cancer types.
Hinweise

Supplementary Information

The online version contains supplementary material available at https://​doi.​org/​10.​1186/​s12967-023-03907-z.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Abkürzungen
TCGA
The Cancer Genome Atlas
COSMIC
The Catalogue of Somatic Mutations in Cancer
ACC
Adrenocortical carcinoma
BLCA
Bladder urothelial carcinoma
BRCA
Breast invasive carcinoma
CESC
Cervical squamous cell carcinoma and endocervical adenocarcinoma
CHOL
Cholangiocarcinoma
COAD
Colon adenocarcinoma
DLBC
Lymphoid neoplasm diffuse large B-cell lymphoma
ESCA
Esophageal carcinoma
GBM
Glioblastoma multiforme
HNSC
Head and neck squamous cell carcinoma
KICH
Kidney chromophobe
KIRC
Kidney renal clear cell carcinoma
KIRP
Kidney renal papillary cell carcinoma
LGG
Brain lower grade glioma
LIHC
Liver hepatocellular carcinoma
LUAD
Lung adenocarcinoma
LUSC
Lung squamous cell carcinoma
MESO
Mesothelioma
OV
Ovarian serous cystadenocarcinoma
PAAD
Pancreatic adenocarcinoma
PCPG
Pheochromocytoma and paraganglioma
PRAD
Prostate adenocarcinoma
READ
Rectum adenocarcinoma
SARC
Sarcoma
SKCM
Skin cutaneous melanoma
STAD
Stomach adenocarcinoma
TCGT
Testicular germ cell tumors
THCA
Thyroid carcinoma
THYM
Thymoma
UCEC
Uterine corpus endometrial carcinoma
UCS
Uterine carcinosarcoma
UVM
Uveal melanoma
GDC
Genomic Data Commons
MiRNA
MicroRNA
DEG
Differentially expressed genes
MUT
Mutated
WT
Wild-type
NES
Normalized enrichment score
GO
Gene ontology
ORA
Over-representation analysis
DEmiR
Differentially expressed microRNAs
TSS
Transcription start site
IC50
Half-maximal inhibitory concentration
OS
Overall survival
FDR
False discovery rate
CNV
Copy number variation
TF
Transcription factor
TME
Tumor microenvironment

Background

The integration of next-generation sequencing together with other high-throughput techniques has provided an excellent opportunity for the study of molecular alterations occurring in cancer [1, 2]. In particular, the platforms for gene expression profiling have been widely used to identify cancer biomarkers. Over the last few decades, it has been recognized the general idea that a singular alteration can not cause cancer, rather than it was recognized as the result of a wider sequence of genetic and genomic events occurring during the progression from normal epithelial tissue to metastatic disease [3, 4]. For this reason, methodologies based on the use of gene signatures, i.e., lists of genes sharing a common pattern of expression among multiple tumor types, is currently recognized as a more biologically significant approach to understanding the biology of cancer [5].
In the present study, we identified 105 onco-signatures associated with the more frequent mutational events shared among the various cancer types. As proof of principle, we evaluate the power of the derived onco-signatures to classify TCGA breast cancer patients in relevant groups with distinct biology and clinical outcome. In addition, we have successfully identified two different metabolic subtypes of Luminal tumors based on 28 specific breast cancer prognostic onco-signatures.
Here, we propose a novel methodological framework to identify commonly shared onco-signatures in cancer that can contribute to the understanding of the role of alterations in tumor disease and the identification of novel molecular mechanisms useful for developing precision therapeutic strategies.

Methods

Data collection

RNA-seq gene expression data, somatic mutation information, and clinical patient annotation of 32 TCGA solid primary tumor type datasets (ACC, BLCA, BRCA, CESC, CHOL, COAD, DLBC, ESCA, GBM, HNSC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, MESO, OV, PAAD, PCPG, PRAD, READ, SARC, SKCM, STAD, TCGT, THCA, THYM, UCEC, UCS, UVM) were retrieved from the Genomic Data Commons (GDC) Data Portal (https://​portal.​gdc.​cancer.​gov) using the R/Bioconductor package TCGAbiolinks [6]. Gene count RNA-seq data, including 9107 primary tumor samples, was normalized by using both within-lane gene-level GC-content and full/upper-quantile normalization methods [7]. The Catalogue of Somatic Mutations in Cancer (COSMIC) database (https://​cancer.​sanger.​ac.​uk/​cosmic/​curation) was used to obtain the list of genes containing mutations implicated in cancer disease (N = 277). The public BRCA loci-based isoform.quantification.txt file, which reports raw and normalized counts of every distinct small RNA-seq observed, was downloaded from TCGA by using TCGAbiolinks [6]. The expression level of each mature miRNA was then calculated as the sum of all isoforms corresponding to the same unique miRBase (https://​www.​mirbase.​org) MIMAT identifier (MIMAT-id). Next, each MIMAT-id was translated into a miRBase name using the miRNA Converter tool of MiRandola database [8]. TCGA BRCA GISTIC2.0 [9] all_thresholded_by_genes.txt output data file, which reports a score to indicate if a gene is considered undergoing homozygous deletion, copy number loss, copy number gain and/or amplification (respectively scores equal to − 2, − 1, 0, 1, 2), was downloaded from the GDAC firehose web portal (https://​gdac.​broadinstitute.​org) by using the Bioconductor RTCGA Toolbox package [10]. Whereas TCGA DNA methylation beta values always related to breast cancer disease were retrieved by using the R/Bioconductor package TCGAbiolinks [6].

Onco-signatures derivation

For each cancer-related gene annotated in COSMIC data annotation, differentially expressed genes (DEGs) between the mutated (MUT) and wild-type (WT) samples were identified using the bi.deg function of DEComplexDisease R package [10]. DeComplexDisease is a tool built to find the differential expressed genes for phenotypes characterized by heterogeneous genomic expression profiles like complex diseases. DeComplexDisease applies a bi-clustering algorithm to find the genes shared by patients associated with complex phenotypes, which is why they are affected by the identical altered molecular mechanism. As the first step DEComplexDisease applies the bi.deg function, which transforms the RNA-seq counts or normalized expression matrix into binary differential expression matrix of − 1, 0, and 1, which indicates respectively the down-regulation, no change, and up-regulation. There are three main steps: (i) the normal samples are used to construct the expression references and estimate two parameters of the distribution, which are mean and dispersion for RNA-seq counts (modeled by a negative-binomial distribution) or mean and standard deviation for normalized or microarray gene expression data; (ii) for every gene i in every sample j of the disease matrix xij, it computes the probability that a random value of the estimated distribution gets a value equal or greater than xij; (iii) using the p-value cutoff defined by the users (0.05 in our experiments), the bi.deg function assigns 1 or − 1 to indicate the up- or down-regulated genes, where 1 is the up-regulated genes and − 1 is the down-regulated genes. The other genes are assigned with 0. This is the final output DEGs matrix of the For each derived binary DEGs matrix, the total number of mutated samples showing − 1 or 1 was determined, gene by gene, to estimate the quantile distribution of counts and select the DEGs most commonly shared across the mutated samples, and so more associated to the mutated phenotype, that for us are the DEGs falling above the 98th percentile of the quintile distribution of counts. Next, unique marker genes were identified for each derived onco-signature and only the signatures with at least five marker genes were selected for further analyses (N = 105).

Normalized enrichment score (NES) estimation and NES clustering analysis of breast cancer survival-associated onco-signatures

Cox’s proportional hazards regression analysis was performed to assess the prognostic ability of our collection of 105 onco-signatures in primary TCGA breast cancer samples to identify the most significant (p-value < 0.05) onco-signatures associated with survival in breast cancer disease. Next, gene-sets enrichment analysis was performed to estimate the normalized enrichment score (NES) of the more relevant breast cancer survival-associated onco-signatures. In particular, the analysis was carried out by interrogating the gene expression values profiled by RNA sequencing of 1093 primary breast cancer tumor samples from TCGA using the mwwGST function of yaGST R package, which runs a competitive single-sample Mann–Whitney–Wilcoxon gene set test [11]. NES is an estimate of the probability that the expression of a gene in the geneset is greater than the expression of a gene outside this set:
\({NES}= 1-\frac{U}{mn}\) where m is the number of genes in a gene set, n is the number of those outside the gene set, \({U}={mn}+{m(m+1)} -{T}\) and T is the sum of the ranks of the genes in the geneset. Then, the matrix of NES concerning the 1093 TCGA breast cancer samples was used to calculate distances between onco-signatures’ gene-sets to build a hierarchical clustering, using as clustering parameters ward criterion (ward.D2 method) and the number of clusters (k) equal to 4.

Differential expression analysis and gene ontology enrichment analysis

The TCGAanalyze_DEA function of TCGAbiolinks R/Bioconductor package [6] was used to perform differential expression analysis by applying the edgeR method [12]. The clusterProfiler R package [13] was used for Gene Ontology (GO) enrichment analysis by its gseGO function. The dotplot function of enrichplot R package [14] was used to visualize functional enrichment results. Normalization of the gene expression matrix was performed using the EDASeq approach [7]. The miRNA Enrichment Analysis and Annotation Tool (miEAA, https://​ccb-compute2.​cs.​uni-saarland.​de/​mieaa2/​) was used to perform over-representation analysis (ORA) of differentially expressed microRNAs (DEmiRs).

Transcription factors motif discovery

Promoter sequences (1500 nucleotides upstream of gene transcription start site (TSS) and 500 nucleotides downstream of TSS) of MIR135A1 and MIR135A2 genes were retrieved from UCSC (https://​genome.​ucsc.​edu) by using R/Bioconductor packages BSgenome.Hsapiens.UCSC.hg19 [15] and Biostrings [16]. MotifDb package available in Bioconductor [17] was used to search and retrieve in JASPAR database (https://​jaspar.​genereg.​net/​) DNA-binding motifs of known transcription factors. Then, for each JASPAR transcription factor motif, the possible matches in the promoters of the two genes coding for hsa-miR-135a-5p were identified by using the matchPWM function of the Biostrings package [16].

Tumor microenvironment infiltration estimation and chemotherapeutic sensitivity prediction

The abundance of six tumor-infiltrating immune cells subsets (B cells, CD4 T cells, CD8 T cells, macrophages, neutrophils, and dendritic cells) was estimated by using TIMER2.0 webserver (http://​timer.​cistrome.​org) [18]. TIMER2.0 provides immune infiltrates’ abundances estimation by multiple immune deconvolution algorithms (TIMER [19, 20], CIBERSORT [21], quanTIseq [22], xCell [23], MCP-counter [24], and EPIC [25]). In addition, pRRophetic R package [26] was used to predict the clinical chemotherapeutic response of the two different metabolic Luminal breast cancer groups to 138 known anticancer drugs. Wilcoxon rank sum test was then used to compare the half-maximal inhibitory concentration (IC50) differences between the two groups in comparison.

Results

Onco-signatures identify four distinct phenotypes of breast cancer

Cancer is caused by an accumulation of somatic mutations in genes involved in important biological processes like cellular growth or DNA repair activity [3, 4]. Due to the specificity of the mutational event, its penetrance, the compensation mechanisms deployed by the cell to reduce the impact of the injury, and to the cancer environment, the molecular lesion can generate new patterns of gene expression that can be shared among cancer types and driver events [27]. To identify common cancer-relevant molecular tissue-agnostic patterns induced by recurrent mutations, we developed a novel integrative procedure summarized in Fig. 1. A total of 105 onco-signatures were derived by using a pan-cancer approach (Additional file 1: Table S1). We evaluated the impact of derived onco-signatures only in breast cancer type. We applied the Cox proportional hazard model using the survival data of 1093 primary breast cancer patients from TCGA to identify the onco-signatures whose activities are significantly (p < 0.05) associated with the survival, selecting in this way a total of 28 breast cancer survival associated onco-signatures. Next, we tested the ability of the selected 28 breast cancer prognostic onco-signatures to stratify the 1093 TCGA primary breast cancer samples by clustering the normalized enrichment score values obtained for each onco-signature in each tumor sample.
The clustering analysis identified four groups that show a different magnitude of onco-signatures activation (Fig. 2A).

The four identified breast cancer clusters show a different outcomes in terms of overall survival

Clustering analysis divided the TCGA primary breast cancer cohort into four groups (Fig. 2A), showing different outcomes in terms of overall survival (OS) (Fig. 2B). Cluster 4, enriched in Luminal A breast cancer subtypes (see Additional file 2: Fig. S1), showed a better OS with respect to the other groups (Fig. 2B). Conversely, Cluster 2 was characterized by the worse prognosis (Fig. 2B), although it was also mainly composed of Luminal tumors (Luminal A and Luminal B together; Additional file 2: Fig. S1). Cluster 1, enriched by Luminal subtypes, and Cluster 3, composed almost exclusively of Basal-like cancers, which are classically defined as the more aggressive subtypes of breast cancer, instead, showed medium survival with respect to Cluster 2 and the Cluster 4 (Fig. 2B). To verify whether the prognostic difference between clusters is specifically driven by the amount of Luminal tumors in the various groups, we decided to re-run the survival analysis only using the Luminal breast cancer cohort. As expected and according to the previous analysis depicted in Fig. 2A, B, the cancers belonging to the Cluster 4 remained associated with prolonged survival using only the BRCA-TCGA Luminal type dataset (Fig. 2C). The analysis also confirmed the worse survival for Luminal samples enriched in Cluster 2, while Cluster 1’s Luminal tumors showed an intermediate prognosis compared to Cluster 4 and Cluster 2 (Fig. 2C). Of note, due to the low number of Luminal cases in cluster 3 (N = 6), it was excluded both from this and the further analyses.

A subset of luminal tumors shows a OXPHOS phenotype

To evaluate the biological functions and the molecular features that could potentially explain the observed difference in survival outcomes across the Luminal breast cancer patients belonging to the identified clusters, we performed a comparative analysis between the samples that make-up to more extreme phenotypes, i.e. Cluster 4, characterized by the better survival, and Cluster 2, with poorer outcome (Fig. 2D). We performed a differential gene expression analysis between Cluster 4 versus Cluster 2 and, according to the cutoff value of |Log2FC| ≥ 1.5 and false discovery rate (FDR) < 0.05, we found 645 differentially expressed genes (DEGs; 243 up- and 402 down-regulated genes. DEGs’ list is provided as Additional file 3: Table S2). Gene ontology (GO) enrichment analysis showed that the biological process enrichment terms of up-regulated DEGs were mainly associated with mitochondrial respiration functions, electron transport chain, and oxidation–reduction processes (Fig. 2E). Conversely, the most significantly down-regulated genes were involved in GO terms related to brain-tissue functions, extracellular matrix organization, and cell adhesion (Fig. 2E). MiRNA differential expression analysis identified 60 deregulated miRNAs (16 up- and 44 down-regulated microRNAs; |Log2FC| ≥ 1 and FDR < 0.05). Over-representation analysis of the 16 up-regulated miRNAs captured the enrichment categories associated with cancer, including brain cancer, and neurodegenerative disorders (Fig. 3A). MiRNA set enrichment analysis of the 44 down-regulated microRNAs, instead, revealed enrichment substantially linked to cancer, and in particular with lung squamous cell carcinoma (adjusted p-value = 1.94e−15; Fig. 3B).

The tumor-suppressive miR-135a-5p is up-regulated in mitochindiral luminal samples, and it could partially explain its favorable phenotype

Luminal breast cancers belonging to Cluster 4 showed a significant up-regulation in miR-135a-5p expression (log2FC = 2.9 and FDR = 3e−99) when compared with Cluster 2’s Luminal tumors (the list of DEmiRs is provided as Additional file 4: Table S3). Since it is well known that miR-135a-5p is a regulator of breast cancer epithelial–mesenchymal transition (EMT) acting by the Wnt/β-catenin signaling pathway [28, 29], we compared the transcriptional levels of both EMT and canonical and non-canonical WNT signaling pathway markers in Cluster 4 versus Cluster 2. As it is depicted in Fig. 3C, the EMT and the WNT signaling pathway members perfectly segregate the Luminal samples according to the cluster classification. Moreover, the two clusters show opposite profiles of expression of particular markers, with a heightened expression of epithelial members in Cluster 4 and elevated expression of mesenchymal markers and WNT members in Cluster 2 (Fig. 3C).

“Mitochondrial” luminal A tumors are characterized by a more favorable outcome independently of their intrinsic molecular classification

As the percentage of Luminal B samples, known to be more aggressive than Luminal A tumors, is higher in Cluster 2 compared to Cluster 4 (respectively 29% and 12%), we re-run all analyses removing Luminal B participants in order to reduce the potential bias arising from molecular subtyping. Surprisingly, superimposable results were obtained by running the analysis only on the Lumina A cohort as depicted in Fig. 3D–F.

The downregulation of miR-135a-5p in Cluster 2 could be, partially, explained in terms of TDMD

In order to try to find molecular features related to such a different survival outcome between Luminal tumors of Cluster 2 and Cluster 4, additional analyses were performed. When we analyzed GISTIC processed SNPs data, we found that although the proportion of cases showing gains and/or losses was comparable across the two analyzed groups, Cluster 2 globally showed a more stable pattern of copy number alterations compared to Cluster 4 (see Additional file 5: Fig. S2). Instead, analyzing the frequencies of losses and gains between the two clusters at the single-gene level we found 3118 amplified and 1757 deleted genes significantly enriched in Cluster 2 versus Cluster 4, and just 130 amplified genes significantly enriched in Cluster 4 versus Cluster 2 (Fig. 4A). Interestingly, GO enrichment analysis of genes more frequently altered in Cluster 2 revealed that the 569 amplified genes were involved in functions and processes linked to response to drug and organic anion transport (Fig. 4B), whereas the 292 deleted genes showed enrichment in GO related to immune response and regulation of receptor signaling pathway via JAK-STAT (Fig. 4C). Conversely, we did not find enriched GO terms for the 23 frequently amplified genes identified in Cluster 4. Furthermore, the CNVs of the two genes coding for miR-135a-5p transcript, respectively MIR135A1 and MIR135A2 genes, were compared by chi-squared test between the two clusters, but neither gene showed significantly (p-value < 0.05) higher frequencies of alteration in Cluster 2 (MIR135A1: deleted and amplified, respectively, in the 32.1% and in the 6.4% of samples; MIR135A2: deleted and amplified, respectively, in the 12.2% and in the 21.0% of samples) or Cluster 4 (MIR135A1: deleted and amplified, respectively, in the 23.4% and in the 7.8% of samples; MIR135A2: deleted and amplified, respectively, in the 9.4% and in the 23.4% of samples). Similarly, on analysis of methylation beta values, we found that the promoter regions of MIR135A1 and MIR135A2 genes did not show different methylation status between the two clusters (Fig. 5A, B). Next, in order to understand if the difference of expression of miR-135a-5p between Cluster 2 and Cluster 4 could be explained at the genomic level in terms of transcriptional regulation, we tried to find candidate transcription factors (TFs) binding sites in DNA promoter sequences of MIR135A1 and MIR135A2 genes using an approach based on sequence matching. From this analysis, we found 122 potential transcription factors binding sites enriched in MIR135A promoter genes (for more details related to the list of putative transcription factor binding sites in both human MIR135A promoter elements see Additional file 6: Table S4). The intersection between the 122 candidate TFs and the DEGs list revealed an overlap of just 4 genes (ZNF354C, NFATC2, NFIC, and EHF), all significantly down-regulated in Cluster 4 versus Cluster 2. We next investigated whether the difference between the expression levels of miR-135a-5p in Cluster 4 and Cluster 2 could be induced by a post-transcriptional event. Since it is currently known that many RNA transcripts are able to trigger the degradation of microRNAs via the mechanism known as target-directed miRNA degradation (TDMD) [3033], we tried to query TDMDfinder database (http://​213.​82.​215.​117:​9999/​TDMDfinder/​index.​php), the first and unique catalog of miRNA:TDMD-target predictions [34], to search potential TDMD target degraders of miR-135a-5p. An overview of the TDMD mechanism is depicted in Fig. 6A. TDMDfinder tool predicted two highly confident TDMD targets for miR-135a-5p, i.e., the Zinc Finger Protein 135 (ZNF135), involved in cytoskeleton organization and cell morphogenesis processes [35], and HMG20A (High Mobility Group 20A) gene, which is known both to play a role in neuronal differentiation and to be an essential factor for the development of the mesenchymal phenotype [36]. When we evaluated the expression levels of the two potential TDMD inducers in Cluster 2 and Cluster 4, we found that they were significantly higher in Cluster 2 compared to Cluster 4 (Wilcoxon test p-value < 0.05, Fig. 6B, C), according to a potential TDMD effect.

The two different metabolic luminal breast cancer groups also show marked differences in terms of tumor microenvironment (TME) state

The abundance of different tumor-infiltrating immune and cancer cell types was evaluated for the Luminal tumors of the two groups of interest, i.e., Cluster 2 and Cluster 4, by using the six immune deconvolution methods provided by TIMER2.0 webtool. As depicted in Fig. 7, the Luminal tumors enriched in Cluster 4 showed higher infiltration of immune effectors like γδ T cells, T follicular helper cells (Tfh), Macrophages M2, natural killer T (NKT) cells, Eosinophils, Neutrophils, plasma B cells, and non-regulatory CD4 T cells.
Conversely, the tumor microenvironment associated with Luminal cancers of Cluster 2 showed a more high infiltration of regulatory T cells (Tregs), Plasmacytoid Dendritic Cells (PDCs), Common Lymphoid Progenitors, Macrophages M0 and M1, CD4+ Th1 cells, and CD4+ Th2 cells, as well as, a higher cytotoxicity score. In addition, we also used pRRophetic algorithm to estimate the IC50 values of 138 anticancer drugs in the Luminal samples enriched in Cluster 2 and Cluster 4. Interestingly, Cluster 2’s Luminal tumors demonstrated much higher sensitivity to treatment with three anticancer drugs, i.e. Entinostat (or MS-S75), Olaparib (or AZD2281), and BI-D1870 (Wilcoxon test, p < 0.05; Fig. 7B–D), respect to Cluster 4 Luminal samples.

Discussion

In this study, we presented a novel pipeline of analysis to identify robust onco-signatures potentially able to predict disease outcome in cancer patients. Using a pan-cancer discovery set of 9107 primary tumor samples together with respective matched mutational data, and a list of known cancer-related genes, we identified 105 onco-signatures, each one composed by a group of distinct marker genes. Aiming to investigate the predictive power of the 105 onco-signatures in breast cancer disease, the Cox proportional hazard regression model was constructed using the TCGA BRCA gene expression dataset, identifying 28 BRCA survival-associated onco-signatures. Next, by performing a gene-set enrichment analysis followed by an unsupervised hierarchical cluster analysis of NESs we identified four discrete breast cancer groups of clinical relevance. Our approach has successfully stratified the Basal-like breast tumors but not the Luminal tumors, who showed high diversity in terms of overall survival across the different clusters. Confirmation of the prognostic difference observed for Luminal cancers enriched in the four identified groups encouraged in-silico molecular analyses to discover the associated genetic variables, which showed profound differences between the more extreme Luminal phenotypes (i.e., Cluster 2 and Cluster 4) with respect to differential gene expression, CNV status, and activation of oncogenic signatures.
Differential gene expression analysis between the two groups of interest provided additional details on their molecular status. Cluster 4 Luminal tumors showed up-regulation of genes linked to mitochondrial respiration and oxidative phosphorylation. In contrast, Cluster 2 Luminal tumors displayed enrichment of genes involved in the development of central nervous system components, and extracellular matrix organization. Looking at genomic imbalances related to the two clusters we also noted that at the genomic level Cluster 4 tumors exhibited a higher frequency of amplifications and deletions as compared to Cluster 2 samples, although at genic-level these imbalances affected a higher number of genes in Cluster 2. Interestingly, we found that the Cluster 2 tumors were enriched in amplifications borne by genes involved in several metabolic processes. On the other hand, the deleted genes in Cluster 2 are involved in immune-cell suicide mechanisms. Although it is not possible to define the impact that such molecular events might have caused, some findings captured our attention. Currently, cancer is considered both a proliferative disorder and a metabolic disease [37, 38]. It is well known that different breast cancer subtypes have distinct bioenergetic and metabolic phenotypes which are associated with different survival outcomes [38]. For example, the Luminal-like tumors present a higher mitochondrial respiratory rate compared to the more metastatic Basal-like cancers that require, instead, an intensive glycolytic flux together with the reduction in OXPHOS processes [38]. In line with our findings, it is possible to hypothesize that an altered mitochondrial metabolism in Cluster 2 may be linked to an unfavorable survival outcome for the Luminal tumors enriched in this group, conferring them a growth advantage. In addition, since recent studies have demonstrated that mitochondrial metabolic processes are modulated by tumor cell-microenvironment [3942], the specific deletion of genes involved in immune-cell death pathways that we found in Cluster 2 could reveal a potential cross-talk between mitochondrial dysfunction and the cancer immune microenvironment of this cluster.
Differential miRNA expression analysis provided additional details on the biology of the two subgroups of Luminal tumors in comparison, identifying the hsa-miR-135a-5p as the most up-regulated microRNA enriched in the cluster with the longer overall survival, i.e., Cluster 4. It has been demonstrated that mammary tumors display an altered expression of the microRNAs, many of whom function as oncogenes or tumor suppressors and modulate a variety of biological processes such as cell proliferation, migration, invasion, metastasis, apoptosis, differentiation, and cellular metabolism [43, 44]. Dysregulation of miR-135a-5p has been described in several cancer types [4547]. Studies on the biological function of miR-135a in cancer have shown that it can play both oncogenic and antitumor roles depending on the cancer type, although it has been described that in breast cancer miR-135a-5p overexpression is able to inhibit EMT by acting through Wnt/β-catenin signaling pathway [28, 48]. In addition, other investigations have also linked mitochondrial activity to epithelial–mesenchymal transition in breast cancer, suggesting that the down-regulation of CDH1 and CTNNB1 in triple-negative breast tumors is correlated to a significant decrease in mitochondrial respiration [49]. Interestingly, our findings confirm the notion that miR-135a-5p is a potential tumor suppressor in breast cancer disease and could be used as a potential prognostic marker for this pathology.
Recent studies have described a special avenue for the downregulation of miRNAs, named target-directed microRNA degradation (TDMD), which induces the direct degradation of miRNAs [3033, 5052]. Although so far there are few studies on the TDMD mechanism, Simeone et al. [34] have performed the first computationally prediction of TDMD inducers in mammalian genomes making available their prediction in TDMDfinder webtool (http://​213.​82.​215.​117:​9999/​TDMDfinder/​index.​php). Investigating the possible cause that could explain the up-regulation of miR135a-5p in Cluster 4, or alternatively its down-regulation in Cluster 2, we queried TDMDfinder tool, which has predicted two high confident TDMD targets for hsa-miR-135a-5p. When we evaluated the expression levels of two potential TDMD inducers in our cohorts, we found that they are predominantly higher in Cluster 2 compared to Cluster 4. In addition, we also found that the biological functions of the two predicted TDMD-genes were associated to ontologies linked to neuronal tissues, where this mechanism was originally described as being particularly active, and pathways frequently altered in human tumors [30, 33]. Taken together these results suggest that the TDMD mechanism may be operative in Cluster 2.
To obtain better insights into the functional roles of the two different metabolic groups of Luminal breast tumors identified in this study, we also conducted analyses to evaluate both their immune infiltrating cell composition and the chemotherapeutic sensitivity to several drugs. Our findings showed that the tumor microenvironment of the Cluster 4 was characterized by a higher infiltration of anti-cancer effector cells, like the γδ T cells, T follicular helper cells, Macrophages M2, and natural killer cells, which are known to contribute to a good prognosis. Conversely, in the Luminal Cluster 2 samples, there was obvious immunosuppressive cells (e.g., the regulatory T cells) infiltration. This result may, at least partially, explain the favorable survival outcome observed in Cluster 4. In addition, in the present study we also showed that the Cluster 2's Luminal tumors were more sensitive to three chemotherapeutic compounds, i.e. Entinostat, Olaparib, and BI-D1870. Entinostat is an oral inhibitor of class I histone deacetylases (HDAC1) that shows a potent antiproliferative effect in breast cancer. Infact, mounting preclinical evidence suggests that it may have a role in immunogenic modulation inhibiting regulatory T cells and promoting tumor infiltration of lytic CD8+ T cells [53]. BI-D1870, instead, is a potent small molecule inhibitor of p90 ribosomal S6 kinases (RSKs) and it is widely used experimentally to revert the EMT phenotype in breast cancer cell lines since it can powerfully inhibit the growth of breast cancer cell lines [54, 55]. Olaparib is an oral poly(ADP ribose) polymerase (PARP) inhibitor that has promising antitumor activity in patients with aggressive forms of breast cancer disease. It, in fact, is the first treatment FDA (Food and Drug Administration)-approved specifically for BRCA mutation carriers with HER2-negative metastatic breast cancer [5657]. Thus, the Luminal cancers characterized both by a low metabolic state and the Cluster 2-like genetic features group may also be valuable for clinical treatment, since our results demonstrated that this group of tumors was more sensitive to the different chemotherapeutic agents.
However, there are also limitations in our study. As in many studies on cancers, all omics data used in this work (CNV data, methylation data, mutational data, and miRNA expression data) could not be retrieved for additional datasets in order to perform a punctual in-silico validation. In addition, one of the more used breast cancer datasets has been created using microarray technology, making reproducibility of the onco-signatures very difficult. The lack of external independent validation results in a limit for our study. For this reason, future computational studies aiming to examine additional datasets, followed by biological validations useful to consolidate our findings are desirable.

Conclusions

In conclusion, our study presents a valid and novel method based on gene set enrichment analysis to construct prognostic gene signatures in cancer. Analyzing in detail the TCGA breast cancer dataset, we demonstrated that our model is useful in predicting the prognosis of patients, and it is also able to stratify the cancer samples into more relevant subgroups. In particular, our onco-signatures were able to distinguish Luminal tumors characterized by different clinical and genetic features, as well as by a different metabolic state and a different tumor microenvironment. In addition, our analysis also provided potential therapeutic targets as well as candidate chemotherapeutic drugs for the improvement in treatment in Luminal patients with a lower mitochondrial activity which, of course, require further clinical confirmation.

Acknowledgements

We are thankful all members of the Ceccarelli’s lab for invaluable discussion. This study was supported by Associazione Italiana per la Ricerca sul Cancro (AIRC) IG 2018 ID 21846 to M.C.

Declarations

Ethical is not applicable because these data are from public database.
All authors have provided their consent for publication.

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, 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 changes were made. 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/​4.​0/​. The Creative Commons Public Domain Dedication waiver (http://​creativecommons.​org/​publicdomain/​zero/​1.​0/​) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Literatur
1.
Zurück zum Zitat Roychowdhury S, Iyer MK, Robinson DR, Lonigro RJ, Wu Y-M, Cao X, et al. Personalized oncology through integrative high-throughput sequencing: a pilot study. Sci Transl Med. 2011;3:111ra121.CrossRef Roychowdhury S, Iyer MK, Robinson DR, Lonigro RJ, Wu Y-M, Cao X, et al. Personalized oncology through integrative high-throughput sequencing: a pilot study. Sci Transl Med. 2011;3:111ra121.CrossRef
2.
Zurück zum Zitat Horak P, Fröhling S, Glimm H. Integrating next-generation sequencing into clinical oncology: strategies, promises and pitfalls. ESMO Open. 2016;1: e000094.CrossRef Horak P, Fröhling S, Glimm H. Integrating next-generation sequencing into clinical oncology: strategies, promises and pitfalls. ESMO Open. 2016;1: e000094.CrossRef
3.
Zurück zum Zitat Jolly C, Van Loo P. Timing somatic events in the evolution of cancer. Genome Biol. 2018;19:95.CrossRef Jolly C, Van Loo P. Timing somatic events in the evolution of cancer. Genome Biol. 2018;19:95.CrossRef
4.
Zurück zum Zitat Sugimura T, Terada M, Yokota J, Hirohashi S, Wakabayashi K. Multiple genetic alterations in human carcinogenesis. Environ Health Perspect. 1992;98:5–12.CrossRef Sugimura T, Terada M, Yokota J, Hirohashi S, Wakabayashi K. Multiple genetic alterations in human carcinogenesis. Environ Health Perspect. 1992;98:5–12.CrossRef
5.
Zurück zum Zitat Qian Y, Daza J, Itzel T, Betge J, Zhan T, Marmé F, et al. Prognostic cancer gene expression signatures: current status and challenges. Cells. 2021;10:648.CrossRef Qian Y, Daza J, Itzel T, Betge J, Zhan T, Marmé F, et al. Prognostic cancer gene expression signatures: current status and challenges. Cells. 2021;10:648.CrossRef
6.
Zurück zum Zitat Colaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D, et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 2016;44: e71.CrossRef Colaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D, et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 2016;44: e71.CrossRef
7.
Zurück zum Zitat Risso D, Schwartz K, Sherlock G, Dudoit S. GC-content normalization for RNA-Seq data. BMC Bioinform. 2011;12:480.CrossRef Risso D, Schwartz K, Sherlock G, Dudoit S. GC-content normalization for RNA-Seq data. BMC Bioinform. 2011;12:480.CrossRef
8.
Zurück zum Zitat Russo F, Di Bella S, Vannini F, Berti G, Scoyni F, Cook HV, et al. miRandola 2017: a curated knowledge base of non-invasive biomarkers. Nucleic Acids Res. 2018;46:D354–9.CrossRef Russo F, Di Bella S, Vannini F, Berti G, Scoyni F, Cook HV, et al. miRandola 2017: a curated knowledge base of non-invasive biomarkers. Nucleic Acids Res. 2018;46:D354–9.CrossRef
9.
Zurück zum Zitat Mermel CH, Schumacher SE, Hill B, Meyerson ML, Beroukhim R, Getz G. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. 2011;12:R41.CrossRef Mermel CH, Schumacher SE, Hill B, Meyerson ML, Beroukhim R, Getz G. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. 2011;12:R41.CrossRef
10.
Zurück zum Zitat Samur MK. RTCGAToolbox: a new tool for exporting TCGA Firehose data. PLoS ONE. 2014;9: e106397.CrossRef Samur MK. RTCGAToolbox: a new tool for exporting TCGA Firehose data. PLoS ONE. 2014;9: e106397.CrossRef
11.
Zurück zum Zitat Frattini V, Pagnotta SM, Tala, Fan JJ, Russo MV, Lee SB, et al. A metabolic function of FGFR3-TACC3 gene fusions in cancer. Nature. 2018;553:222–7.CrossRef Frattini V, Pagnotta SM, Tala, Fan JJ, Russo MV, Lee SB, et al. A metabolic function of FGFR3-TACC3 gene fusions in cancer. Nature. 2018;553:222–7.CrossRef
12.
Zurück zum Zitat Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26:139–40.CrossRef Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26:139–40.CrossRef
13.
Zurück zum Zitat Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, et al. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021;2: 100141. Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, et al. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021;2: 100141.
14.
Zurück zum Zitat Guangchuang Y. enrichplot: visualization of functional enrichment result. Computer software. Bioconductor; 2021. Guangchuang Y. enrichplot: visualization of functional enrichment result. Computer software. Bioconductor; 2021.
15.
Zurück zum Zitat The Bioconductor Dev Team TBDT. BSgenome.Hsapiens.UCSC.hg19: full genome sequences for Homo sapiens (UCSC version hg19, based on GRCh37.p13). Computer software. Bioconductor; 2020. The Bioconductor Dev Team TBDT. BSgenome.Hsapiens.UCSC.hg19: full genome sequences for Homo sapiens (UCSC version hg19, based on GRCh37.p13). Computer software. Bioconductor; 2020.
16.
Zurück zum Zitat Pagès H, Aboyoun P, Gentleman R, DebRoy S. Biostrings: efficient manipulation of biological strings. Computer software. Bioconductor; 2021. Pagès H, Aboyoun P, Gentleman R, DebRoy S. Biostrings: efficient manipulation of biological strings. Computer software. Bioconductor; 2021.
17.
Zurück zum Zitat Shannon P, Richards M. MotifDb: an annotated collection of protein-DNA binding sequence motifs. Computer software. Bioconductor; 2021. Shannon P, Richards M. MotifDb: an annotated collection of protein-DNA binding sequence motifs. Computer software. Bioconductor; 2021.
18.
Zurück zum Zitat Li T, Fu J, Zeng Z, Cohen D, Li J, Chen Q, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48:W509–14.CrossRef Li T, Fu J, Zeng Z, Cohen D, Li J, Chen Q, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48:W509–14.CrossRef
19.
Zurück zum Zitat Li B, Severson E, Pignon J-C, Zhao H, Li T, Novak J, et al. Comprehensive analyses of tumor immunity: implications for cancer immunotherapy. Genome Biol. 2016;17:174.CrossRef Li B, Severson E, Pignon J-C, Zhao H, Li T, Novak J, et al. Comprehensive analyses of tumor immunity: implications for cancer immunotherapy. Genome Biol. 2016;17:174.CrossRef
20.
Zurück zum Zitat Li T, Fan J, Wang B, Traugh N, Chen Q, Liu JS, et al. TIMER: a web server for comprehensive analysis of tumor-infiltrating immune cells. Cancer Res. 2017;77:e108–10.CrossRef Li T, Fan J, Wang B, Traugh N, Chen Q, Liu JS, et al. TIMER: a web server for comprehensive analysis of tumor-infiltrating immune cells. Cancer Res. 2017;77:e108–10.CrossRef
21.
Zurück zum Zitat Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12:453–7.CrossRef Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12:453–7.CrossRef
22.
Zurück zum Zitat Finotello F, Mayer C, Plattner C, Laschober G, Rieder D, Hackl H, et al. Molecular and pharmacological modulators of the tumor immune contexture revealed by deconvolution of RNA-seq data. Genome Med. 2019;11:34.CrossRef Finotello F, Mayer C, Plattner C, Laschober G, Rieder D, Hackl H, et al. Molecular and pharmacological modulators of the tumor immune contexture revealed by deconvolution of RNA-seq data. Genome Med. 2019;11:34.CrossRef
23.
Zurück zum Zitat Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol. 2017;18:220.CrossRef Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol. 2017;18:220.CrossRef
24.
Zurück zum Zitat Becht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, et al. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol. 2016;17:218.CrossRef Becht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, et al. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol. 2016;17:218.CrossRef
25.
Zurück zum Zitat Racle J, Gfeller D. EPIC: a tool to estimate the proportions of different cell types from bulk gene expression data. Methods Mol Biol. 2020;2120:233–48.CrossRef Racle J, Gfeller D. EPIC: a tool to estimate the proportions of different cell types from bulk gene expression data. Methods Mol Biol. 2020;2120:233–48.CrossRef
26.
Zurück zum Zitat Geeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS ONE. 2014;9: e107468.CrossRef Geeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS ONE. 2014;9: e107468.CrossRef
27.
Zurück zum Zitat Cantini L, Calzone L, Martignetti L, Rydenfelt M, Blüthgen N, Barillot E, et al. Classification of gene signatures for their information value and functional redundancy. NPJ Syst Biol Appl. 2018;4:2.CrossRef Cantini L, Calzone L, Martignetti L, Rydenfelt M, Blüthgen N, Barillot E, et al. Classification of gene signatures for their information value and functional redundancy. NPJ Syst Biol Appl. 2018;4:2.CrossRef
28.
Zurück zum Zitat Jiang D, Zhou B, Xiong Y, Cai H. miR-135 regulated breast cancer proliferation and epithelial–mesenchymal transition acts by the Wnt/β-catenin signaling pathway. Int J Mol Med. 2019;43:1623–34. Jiang D, Zhou B, Xiong Y, Cai H. miR-135 regulated breast cancer proliferation and epithelial–mesenchymal transition acts by the Wnt/β-catenin signaling pathway. Int J Mol Med. 2019;43:1623–34.
29.
Zurück zum Zitat Ribatti D, Tamma R, Annese T. Epithelial–mesenchymal transition in cancer: a historical overview. Transl Oncol. 2020;13: 100773.CrossRef Ribatti D, Tamma R, Annese T. Epithelial–mesenchymal transition in cancer: a historical overview. Transl Oncol. 2020;13: 100773.CrossRef
30.
Zurück zum Zitat de la Mata M, Gaidatzis D, Vitanescu M, Stadler MB, Wentzel C, Scheiffele P, et al. Potent degradation of neuronal miRNAs induced by highly complementary targets. EMBO Rep. 2015;16:500–11.CrossRef de la Mata M, Gaidatzis D, Vitanescu M, Stadler MB, Wentzel C, Scheiffele P, et al. Potent degradation of neuronal miRNAs induced by highly complementary targets. EMBO Rep. 2015;16:500–11.CrossRef
31.
Zurück zum Zitat Ghini F, Rubolino C, Climent M, Simeone I, Marzi MJ, Nicassio F. Endogenous transcripts control miRNA levels and activity in mammalian cells by target-directed miRNA degradation. Nat Commun. 2018;9:3119.CrossRef Ghini F, Rubolino C, Climent M, Simeone I, Marzi MJ, Nicassio F. Endogenous transcripts control miRNA levels and activity in mammalian cells by target-directed miRNA degradation. Nat Commun. 2018;9:3119.CrossRef
32.
Zurück zum Zitat Bitetti A, Mallory AC, Golini E, Carrieri C, Carreño Gutiérrez H, Perlas E, et al. MicroRNA degradation by a conserved target RNA regulates animal behavior. Nat Struct Mol Biol. 2018;25:244–51.CrossRef Bitetti A, Mallory AC, Golini E, Carrieri C, Carreño Gutiérrez H, Perlas E, et al. MicroRNA degradation by a conserved target RNA regulates animal behavior. Nat Struct Mol Biol. 2018;25:244–51.CrossRef
33.
Zurück zum Zitat Kleaveland B, Shi CY, Stefano J, Bartel DP. A network of noncoding regulatory RNAs acts in the mammalian brain. Cell. 2018;174:350-362.e17.CrossRef Kleaveland B, Shi CY, Stefano J, Bartel DP. A network of noncoding regulatory RNAs acts in the mammalian brain. Cell. 2018;174:350-362.e17.CrossRef
34.
Zurück zum Zitat Simeone I, Rubolino C, Noviello TMR, Farinello D, Cerulo L, Marzi MJ, et al. Prediction and pan-cancer analysis of mammalian transcripts involved in target directed miRNA degradation. Nucleic Acids Res. 2022;50:2019–35.CrossRef Simeone I, Rubolino C, Noviello TMR, Farinello D, Cerulo L, Marzi MJ, et al. Prediction and pan-cancer analysis of mammalian transcripts involved in target directed miRNA degradation. Nucleic Acids Res. 2022;50:2019–35.CrossRef
35.
Zurück zum Zitat Bai SW, Herrera-Abreu MT, Rohn JL, Racine V, Tajadura V, Suryavanshi N, et al. Identification and characterization of a set of conserved and new regulators of cytoskeletal organization, cell morphology and migration. BMC Biol. 2011;9:54.CrossRef Bai SW, Herrera-Abreu MT, Rohn JL, Racine V, Tajadura V, Suryavanshi N, et al. Identification and characterization of a set of conserved and new regulators of cytoskeletal organization, cell morphology and migration. BMC Biol. 2011;9:54.CrossRef
36.
Zurück zum Zitat Rivero S, Ceballos-Chávez M, Bhattacharya SS, Reyes JC. HMG20A is required for SNAI1-mediated epithelial to mesenchymal transition. Oncogene. 2015;34:5264–76.CrossRef Rivero S, Ceballos-Chávez M, Bhattacharya SS, Reyes JC. HMG20A is required for SNAI1-mediated epithelial to mesenchymal transition. Oncogene. 2015;34:5264–76.CrossRef
37.
Zurück zum Zitat Wei Q, Qian Y, Yu J, Wong CC. Metabolic rewiring in the promotion of cancer metastasis: mechanisms and therapeutic implications. Oncogene. 2020;39:6139–56.CrossRef Wei Q, Qian Y, Yu J, Wong CC. Metabolic rewiring in the promotion of cancer metastasis: mechanisms and therapeutic implications. Oncogene. 2020;39:6139–56.CrossRef
38.
Zurück zum Zitat Avagliano A, Ruocco MR, Aliotta F, Belviso I, Accurso A, Masone S, et al. Mitochondrial flexibility of breast cancers: a growth advantage and a therapeutic opportunity. Cells. 2019;8:401.CrossRef Avagliano A, Ruocco MR, Aliotta F, Belviso I, Accurso A, Masone S, et al. Mitochondrial flexibility of breast cancers: a growth advantage and a therapeutic opportunity. Cells. 2019;8:401.CrossRef
39.
Zurück zum Zitat Neagu M, Constantin C, Popescu ID, Zipeto D, Tzanakakis G, Nikitovic D, et al. Inflammation and metabolism in cancer cell-mitochondria key player. Front Oncol. 2019;9:348.CrossRef Neagu M, Constantin C, Popescu ID, Zipeto D, Tzanakakis G, Nikitovic D, et al. Inflammation and metabolism in cancer cell-mitochondria key player. Front Oncol. 2019;9:348.CrossRef
41.
Zurück zum Zitat Romero-Garcia S, Moreno-Altamirano MMB, Prado-Garcia H, Sánchez-García FJ. Lactate contribution to the tumor microenvironment: mechanisms, effects on immune cells and therapeutic relevance. Front Immunol. 2016;7:52.CrossRef Romero-Garcia S, Moreno-Altamirano MMB, Prado-Garcia H, Sánchez-García FJ. Lactate contribution to the tumor microenvironment: mechanisms, effects on immune cells and therapeutic relevance. Front Immunol. 2016;7:52.CrossRef
42.
Zurück zum Zitat Ippolito L, Morandi A, Giannoni E, Chiarugi P. Lactate: a metabolic driver in the tumour landscape. Trends Biochem Sci. 2019;44:153–66.CrossRef Ippolito L, Morandi A, Giannoni E, Chiarugi P. Lactate: a metabolic driver in the tumour landscape. Trends Biochem Sci. 2019;44:153–66.CrossRef
43.
Zurück zum Zitat Palmero EI, de Campos SGP, Campos M, de Souza NCN, Guerreiro IDC, Carvalho AL, et al. Mechanisms and role of microRNA deregulation in cancer onset and progression. Genet Mol Biol. 2011;34:363–70.CrossRef Palmero EI, de Campos SGP, Campos M, de Souza NCN, Guerreiro IDC, Carvalho AL, et al. Mechanisms and role of microRNA deregulation in cancer onset and progression. Genet Mol Biol. 2011;34:363–70.CrossRef
44.
Zurück zum Zitat Iorio MV, Ferracin M, Liu C-G, Veronese A, Spizzo R, Sabbioni S, et al. MicroRNA gene expression deregulation in human breast cancer. Cancer Res. 2005;65:7065–70.CrossRef Iorio MV, Ferracin M, Liu C-G, Veronese A, Spizzo R, Sabbioni S, et al. MicroRNA gene expression deregulation in human breast cancer. Cancer Res. 2005;65:7065–70.CrossRef
45.
Zurück zum Zitat Nagel R, le Sage C, Diosdado B, van der Waal M, Oude Vrielink JAF, Bolijn A, et al. Regulation of the adenomatous polyposis coli gene by the miR-135 family in colorectal cancer. Cancer Res. 2008;68:5795–802.CrossRef Nagel R, le Sage C, Diosdado B, van der Waal M, Oude Vrielink JAF, Bolijn A, et al. Regulation of the adenomatous polyposis coli gene by the miR-135 family in colorectal cancer. Cancer Res. 2008;68:5795–802.CrossRef
46.
Zurück zum Zitat Mao XP, Zhang LS, Huang B, Zhou SY, Liao J, Chen LW, et al. Mir-135a enhances cellular proliferation through post-transcriptionally regulating PHLPP2 and FOXO1 in human bladder cancer. J Transl Med. 2015;13:86.CrossRef Mao XP, Zhang LS, Huang B, Zhou SY, Liao J, Chen LW, et al. Mir-135a enhances cellular proliferation through post-transcriptionally regulating PHLPP2 and FOXO1 in human bladder cancer. J Transl Med. 2015;13:86.CrossRef
47.
Zurück zum Zitat Ren J-W, Li Z-J, Tu C. MiR-135 post-transcriptionally regulates FOXO1 expression and promotes cell proliferation in human malignant melanoma cells. Int J Clin Exp Pathol. 2015;8:6356–66. Ren J-W, Li Z-J, Tu C. MiR-135 post-transcriptionally regulates FOXO1 expression and promotes cell proliferation in human malignant melanoma cells. Int J Clin Exp Pathol. 2015;8:6356–66.
48.
Zurück zum Zitat Cao Z, Qiu J, Yang G, Liu Y, Luo W, You L, et al. MiR-135a biogenesis and regulation in malignancy: a new hope for cancer research and therapy. Cancer Biol Med. 2020;17:569–82.CrossRef Cao Z, Qiu J, Yang G, Liu Y, Luo W, You L, et al. MiR-135a biogenesis and regulation in malignancy: a new hope for cancer research and therapy. Cancer Biol Med. 2020;17:569–82.CrossRef
49.
Zurück zum Zitat Lunetti P, Di Giacomo M, Vergara D, De Domenico S, Maffia M, Zara V, et al. Metabolic reprogramming in breast cancer results in distinct mitochondrial bioenergetics between luminal and basal subtypes. FEBS J. 2019;286:688–709.CrossRef Lunetti P, Di Giacomo M, Vergara D, De Domenico S, Maffia M, Zara V, et al. Metabolic reprogramming in breast cancer results in distinct mitochondrial bioenergetics between luminal and basal subtypes. FEBS J. 2019;286:688–709.CrossRef
50.
Zurück zum Zitat Ameres SL, Horwich MD, Hung J-H, Xu J, Ghildiyal M, Weng Z, et al. Target RNA-directed trimming and tailing of small silencing RNAs. Science. 2010;328:1534–9.CrossRef Ameres SL, Horwich MD, Hung J-H, Xu J, Ghildiyal M, Weng Z, et al. Target RNA-directed trimming and tailing of small silencing RNAs. Science. 2010;328:1534–9.CrossRef
51.
Zurück zum Zitat Marcinowski L, Tanguy M, Krmpotic A, Rädle B, Lisnić VJ, Tuddenham L, et al. Degradation of cellular mir-27 by a novel, highly abundant viral transcript is important for efficient virus replication in vivo. PLoS Pathog. 2012;8: e1002510.CrossRef Marcinowski L, Tanguy M, Krmpotic A, Rädle B, Lisnić VJ, Tuddenham L, et al. Degradation of cellular mir-27 by a novel, highly abundant viral transcript is important for efficient virus replication in vivo. PLoS Pathog. 2012;8: e1002510.CrossRef
52.
Zurück zum Zitat Cazalla D, Steitz JA. Down-regulation of a host microRNA by a viral noncoding RNA. Cold Spring Harb Symp Quant Biol. 2010;75:321–4.CrossRef Cazalla D, Steitz JA. Down-regulation of a host microRNA by a viral noncoding RNA. Cold Spring Harb Symp Quant Biol. 2010;75:321–4.CrossRef
53.
Zurück zum Zitat Hicks KC, Chariou PL, Ozawa Y, Minnar CM, Knudson KM, Meyer TJ, et al. Tumour-targeted interleukin-12 and entinostat combination therapy improves cancer survival by reprogramming the tumour immune cell landscape. Nat Commun. 2021;12:5151.CrossRef Hicks KC, Chariou PL, Ozawa Y, Minnar CM, Knudson KM, Meyer TJ, et al. Tumour-targeted interleukin-12 and entinostat combination therapy improves cancer survival by reprogramming the tumour immune cell landscape. Nat Commun. 2021;12:5151.CrossRef
54.
Zurück zum Zitat Senigagliesi B, Penzo C, Severino LU, Maraspini R, Petrosino S, Morales-Navarrete H, et al. The high mobility group A1 (HMGA1) chromatin architectural factor modulates nuclear stiffness in breast cancer cells. Int J Mol Sci. 2019;20:2733.CrossRef Senigagliesi B, Penzo C, Severino LU, Maraspini R, Petrosino S, Morales-Navarrete H, et al. The high mobility group A1 (HMGA1) chromatin architectural factor modulates nuclear stiffness in breast cancer cells. Int J Mol Sci. 2019;20:2733.CrossRef
55.
Zurück zum Zitat Stratford AL, Reipas K, Hu K, Fotovati A, Brough R, Frankum J, et al. Targeting p90 ribosomal S6 kinase eliminates tumor-initiating cells by inactivating Y-box binding protein-1 in triple-negative breast cancers. Stem Cells. 2012;30:1338–48.CrossRef Stratford AL, Reipas K, Hu K, Fotovati A, Brough R, Frankum J, et al. Targeting p90 ribosomal S6 kinase eliminates tumor-initiating cells by inactivating Y-box binding protein-1 in triple-negative breast cancers. Stem Cells. 2012;30:1338–48.CrossRef
56.
Zurück zum Zitat Robson M, Im SA, Senkus E, Xu B, Domchek SM, Masuda N, et al. Olaparib for metastatic breast cancer in patients with a germline BRCA mutation. N Engl J Med. 2017;377:523–33.CrossRef Robson M, Im SA, Senkus E, Xu B, Domchek SM, Masuda N, et al. Olaparib for metastatic breast cancer in patients with a germline BRCA mutation. N Engl J Med. 2017;377:523–33.CrossRef
57.
Zurück zum Zitat US Food & Drug Administration. FDA approves olaparib for germline BRCA-mutated metastatic breast cancer. cmr. 2018. US Food & Drug Administration. FDA approves olaparib for germline BRCA-mutated metastatic breast cancer. cmr. 2018.
Metadaten
Titel
Pan-cancer onco-signatures reveal a novel mitochondrial subtype of luminal breast cancer with specific regulators
verfasst von
Ines Simeone
Michele Ceccarelli
Publikationsdatum
01.12.2023
Verlag
BioMed Central
Erschienen in
Journal of Translational Medicine / Ausgabe 1/2023
Elektronische ISSN: 1479-5876
DOI
https://doi.org/10.1186/s12967-023-03907-z

Weitere Artikel der Ausgabe 1/2023

Journal of Translational Medicine 1/2023 Zur Ausgabe

Leitlinien kompakt für die Innere Medizin

Mit medbee Pocketcards sicher entscheiden.

Seit 2022 gehört die medbee GmbH zum Springer Medizin Verlag

Schadet Ärger den Gefäßen?

14.05.2024 Arteriosklerose Nachrichten

In einer Studie aus New York wirkte sich Ärger kurzfristig deutlich negativ auf die Endothelfunktion gesunder Probanden aus. Möglicherweise hat dies Einfluss auf die kardiovaskuläre Gesundheit.

Intervallfasten zur Regeneration des Herzmuskels?

14.05.2024 Herzinfarkt Nachrichten

Die Nahrungsaufnahme auf wenige Stunden am Tag zu beschränken, hat möglicherweise einen günstigen Einfluss auf die Prognose nach akutem ST-Hebungsinfarkt. Darauf deutet eine Studie an der Uniklinik in Halle an der Saale hin.

Klimaschutz beginnt bei der Wahl des Inhalators

14.05.2024 Klimawandel Podcast

Auch kleine Entscheidungen im Alltag einer Praxis können einen großen Beitrag zum Klimaschutz leisten. Die neue Leitlinie zur "klimabewussten Verordnung von Inhalativa" geht mit gutem Beispiel voran, denn der Wechsel vom klimaschädlichen Dosieraerosol zum Pulverinhalator spart viele Tonnen CO2. Leitlinienautor PD Dr. Guido Schmiemann erklärt, warum nicht nur die Umwelt, sondern auch Patientinnen und Patienten davon profitieren.

Zeitschrift für Allgemeinmedizin, DEGAM

Typ-2-Diabetes und Depression folgen oft aufeinander

14.05.2024 Typ-2-Diabetes Nachrichten

Menschen mit Typ-2-Diabetes sind überdurchschnittlich gefährdet, in den nächsten Jahren auch noch eine Depression zu entwickeln – und umgekehrt. Besonders ausgeprägt ist die Wechselbeziehung laut GKV-Daten bei jüngeren Erwachsenen.

Update Innere Medizin

Bestellen Sie unseren Fach-Newsletter und bleiben Sie gut informiert.