mRNA expression levels of FA pathway genes in breast cancer
Based on the ONCOMINE data, the transcriptional levels of various FA pathway genes were found to be elevated in various types of cancers, such as colorectal cancer, cervical cancer and breast cancer (Fig. 1). Moreover, in breast cancer, the transcriptional levels of FANCA, FANCB, UBE2T (FANCT), FANCD2, FANCI, BRCA2 (FANCD1), BRIP1 (FANCJ), RAD51 (FANCR), BRCA1 (FANCS), MAD2L2 (FANCV) and RFWD3 (FANCW) were significantly elevated, while the transcriptional levels of FANCC and XRCC2 (FANCU) were significantly suppressed in some specific datasets. Then, the expression levels of all the FA pathway genes in breast cancer were determined using the UCSC Xena browser. It was found that all the FA pathway genes were highly expressed in BRCA (Fig. 2). Notably, FANCI and RFWD3 (FANCW) exhibited the highest mRNA levels while FANCB had the lowest. We further compared the mRNA expression levels of FA pathway genes in breast cancer and their corresponding normal tissues. Results from the UALCAN database revealed that the genes were significantly upregulated in breast cancer compared to their corresponding normal tissues, except for FANCE and FANCM which were downregulated in tumors (Fig. 3).
The prognostic value of FA pathway genes in breast cancer
The Kaplan-Meier plotter was used to determine the potential prognostic value of FA pathway genes in breast cancer. About half of the genes that were highly expressed were shown to exhibit a significant positive correlation with worse recurrent free survival (RFS) of breast cancer patients (Fig. 4). These genes include; FANCB, FANCG, FANCL, UBE2T (FANCT), FANCI, BRIP1 (FANCJ), BRCC5 (FANCR), BRCA1 (FANCS), MAD2L2 (FANCV) and RFWD3 (FANCW). In contrast, elevated mRNA expression levels of FANCC, SLX4 (FANCP), PALB2 (FANCN), XRCC2 (FANCU) and ERCC4 (FANCQ) were associated with favorable RFS. The other genes did not exhibit a significant correlation with the RFS of breast cancer patients.
We further used the Kaplan-Meier plotter to investigate the correlations between the expression levels of FA pathway genes and patients’ overall survival (OS), distant metastasis free survival (DMFS) and post progression survival (PPS). The result forest plots are presented in Fig. 5.
Association between the prognostic value of FA pathway genes in breast cancer with different clinicopathological features
The relationship between FA pathway genes and different clinicopathological features was investigated to elucidate on the roles of these genes in breast cancer prognosis. The clinicopathological features included cancer grade, ER status, PR status, HER2 status and TP53 status. It was revealed that elevated mRNA expression levels of FANCI were associated with poor RFS in grade 1 breast cancer. Additionally, elevated mRNA expression levels of FANCA, FANCE, FANCL, FANCI, BRCC5 and RFWD3 were correlated to worse RFS in grade 2 breast cancer. However, FANCF and PALB2 were associated with better RFS in grade 2 breast cancer. Moreover, low mRNA expression levels of FANCI and BRCA1 were associated with better RFS in grade 3 breast cancer, while BRIP1 and RAD51C were identified as good prognostic factors for grade 3 breast cancer. These results are presented in Table 1.
Moreover, BRCA1 was found to be a promising marker for unfavorable prognosis in both ER positive and negative patients, while FANCA, FANCB, FANCE, FANCG, FANCL, UBE2T, FANCI, BRIP1, RAD51C, BRCC5 and RFWD3 were significantly associated with unfavorable RFS in ER positive patients. However, FANCC and FANCD1 were associated with favorable RFS in ER negative patients (Table 2).
Furthermore, elevated mRNA expression levels of FANCA, FANCG, UBE2T, FANCI, BRIP1, RAD51C, BRCC5 and BRCA1 were shown to contribute to unfavorable RFS in PR positive patients while elevated expression levels of SLX4 and ERCC4 were associated with shorter RFS in PR negative patients (Table 3).
Table 4 shows that PALB2 exhibited a significant association with unfavorable RFS in HER2 positive patients. However, FANCA, FANCB, FANCC, FANCG, UBE2T, SLX4, FANCI, BRIP1, RAD51C, BRCA1 and ERCC4 were associated with poor RFS while FANCF was associated with better RFS in HER2 negative patients. Nevertheless, FANCD2 was a good prognostic factor in both HER2 positive and negative patients.
Based on the TP53 status, elevated expression levels of UBE2T were strongly associated with worse RFS while FANCL, FAND2 and RAD51C were correlated with better RFS in TP53-mutated patients. In patients with wild type TP53, FANCA, FANCI, BRCA1 and RFWD3 were associated with unfavorable RFS (Table 5).
Genetic alterations and interaction analysis of FA pathway genes in breast cancer
We further performed a comprehensive analysis of the molecular characteristics of genes in the FA pathway. The frequency of genetic alterations in these genes among breast cancer patients was determined using the cBioPortal database. It was found that mRNA deregulation was one of the most important single factors for genetic alterations in different kinds of breast cancers (Fig. 6a). Mutation and amplification were the most common alterations in these samples. In addition, OncoPrint was used to show a visual summary of alterations in the FA pathway genes across a set of breast cancer samples (Fig. 6b). Moreover, expression correlations were determined using GEPIA to further define the relationships among the FA pathway genes. There was a low to high positive correlation among most FA pathway genes (Fig. 7a). Additionally, a network of FA pathway genes and their functionally related genes was constructed using GeneMANIA (Fig. 7b). Twenty genes were found to be closely associated with the regulatory functions of differentially expressed FA pathway genes. These genes were FAAP24, FAAP100, RAD51B, XRCC3, RAD52, BARD1, BLM, RMI1, TOP3A, ATR, FAN1, APITD1, TOPBP1, USP1, G2E3, RAD51D, RPA1, RPA2, ERCC1 and ATRIP.
Functional enrichment analysis of FA-related genes in breast cancer
We used UALCAN to isolate the top 50 genes that were positively and negatively correlated with individual genes of the FA pathway in breast cancer. This was done to explore the underlying mechanisms of FA pathway genes in cancer. In addition, DAVID was used to perform GO and KEGG pathway enrichment analysis of the FA-associated genes in breast cancer. Fig. 8 shows the most highly enriched GO items that were positively and negatively correlated with FA pathway genes in breast cancer. Among the positively correlated GO items, the most enriched biological process (BP) term was cell division, the most enriched cellular component (CC) term was nucleoplasm, and the most enriched molecular function (MF) term was protein binding (Fig. 8a). In the negatively correlated GO items, the FA-associated genes were shown to participate in various functions, especially SRP-dependent co-translational protein targeting to membrane as well as ribosome and structural constituent of ribosome (Fig. 8b). Moreover, KEGG pathway enrichment analysis revealed that the positively correlated pathways were mainly involved in cell cycle, oocyte meiosis and viral carcinogenesis (Fig. 8c). Besides, ribosomes, along with their metabolic pathways, were the most enriched pathways that were found to be negatively correlated with FA pathway genes in breast cancer (Fig. 8d).
Correlation between immune infiltration and FA pathway genes in patients with breast cancer
Given that inflammatory responses and infiltrating immune cells can affect breast cancer prognosis, we evaluated the association between differentially expressed FA pathway genes and immune cell infiltration using the TIMER database. It was found that the mRNA expression levels of FA pathway genes were positively associated with tumor purity, while FANCE, FANCM, BRCA2 (FANCD1) and MAD2L2 (FANCV) had no significant correlation with the tumor purity of patients with breast cancer (Fig. 9). Furthermore, most FA pathway genes were positively associated with immune infiltration levels of B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils and dendritic cells, except for FANCE of macrophages, FANCF of dendritic cells, FANCG of macrophages, FANCL of macrophages, UBE2T of macrophages, RAD51C of CD4+ T cells, MAD2L2 of CD8+ T cells and macrophages, which showed significant negative associations.