Exploring breast tissue microbial composition and the association with breast cancer risk factors

Abstract

Background: Microbial dysbiosis has emerged as an important element in the development and progression of various cancers, including breast cancer. However, the microbial composition of the breast from healthy individuals, even relative to risk of developing breast cancer, remains unclear. Here, we performed a comprehensive analysis of the microbiota of the normal breast tissue, which was analyzed in relation to the microbial composition of the tumor and adjacent normal tissue.

Methods: The study cohorts included 403 cancer-free women (who donated normal breast tissue cores) and 76 breast cancer patients (who donated tumor and/or adjacent normal tissue samples). Microbiome profiling was obtained by sequencing the nine hypervariable regions of the 16S rRNA gene (V1V2, V2V3, V3V4, V4V5, V5V7, and V7V9). Transcriptome analysis was also performed on 190 normal breast tissue samples. Breast cancer risk score was assessed using the Tyrer-Cuzick risk model.

Results: The V1V2 amplicon sequencing resulted more suitable for the analysis of the normal breast microbiome and identified Lactobacillaceae (Firmicutes phylum), Acetobacterraceae, and Xanthomonadaceae (both Proteobacteria phylum) as the most abundant families in the normal breast. However, Ralstonia (Proteobacteria phylum) was more abundant in both breast tumors and histologically normal tissues adjacent to malignant tumors. We also conducted a correlation analysis between the microbiome and known breast cancer risk factors. Abundances of the bacterial taxa Acetotobacter aceti, Lactobacillus vini, Lactobacillus paracasei, and Xanthonomas sp. were associated with age (p < 0.0001), racial background (p < 0.0001), and parity (p < 0.0001). Finally, transcriptome analysis of normal breast tissues showed an enrichment in metabolism- and immune-related genes in the tissues with abundant Acetotobacter aceti, Lactobacillus vini, Lactobacillus paracasei, and Xanthonomas sp., whereas the presence of Ralstonia in the normal tissue was linked to dysregulation of genes involved in the carbohydrate metabolic pathway.

Conclusions: This study defines the microbial features of normal breast tissue, thus providing a basis to understand cancer-related dysbiosis. Moreover, the findings reveal that lifestyle factors can significantly affect the normal breast microbial composition.

Keywords: Acetobacter aceti; Breast cancer risk factors; Lactobacillus; Normal breast; Ralstonia; Xanthomonas sp..

Fig. 1

Variable regions analysis of the breast tissue. A Microbiome analysis of the breast was performed using primers covering all nine hypervariable regions of the 16S rRNA (V1V2, V2V3, V3V4, V4V5, V5V7, and V7V9). B Principal component analysis based on Bray–Curtis distance illustrating the differences between bacterial communities among 40 representative Normal samples via the indicated hypervariable regions at both family and genus level. C Shannon diversity index for each of the hypervariable region was determined at both family and genus level. D Total read values from the sequencing of each indicated hypervariable region for normal (Normal), normal adjacent to the tumor (NAT), tumor breast tissues (Tumor), and breast cancer metastasis (Met) samples. Nonparametric two-tailed Mann–Whitney U test was used for statistical analysis. *p < 0.05; **p < 0.001; ***p < 0.0001

Fig. 2

Bacterial abundance in breast tissues. Bacteria taxonomy was examined in the normal (Normal), adjacent normal (NAT), tumor breast tissues (Tumor) and breast cancer metastases (Met) at either family (A) and genus (B) level. Data from the V1V2 amplified region are shown. Most abundant bacteria (≥ 2% abundance) are displayed in the histobar graph on the left, whereas OTU count of the three most abundant bacteria is shown on the right as the scatter plot, where each dot represents a sample (value of 0 are not included because of the logarithmic scale). Nonparametric two-tailed Mann–Whitney U test was used for statistical analysis. *p < 0.05; **p < 0.001

Fig. 3

Bacterial abundance in breast tissues at species level. Principal component analysis of Weighted UniFrac (A) and Unweighted UniFrac (B) plots showing microbiota beta diversity clustering, heat map (C) and bar chart (D) demonstrating relative abundance of species after 16S rRNA sequencing of healthy (Normal), adjacent normal (NAT), and tumor (Tumor) breast tissues as well as breast cancer metastases (Met). (E) Bacterial abundance in Normal, NAT, Tumor and Met samples, each dot represents a sample. Because of the use of the log10 scale values of 0 are not shown. Nonparametric two-tailed Mann–Whitney U test was used for statistical analysis. *p < 0.05; **p < 0.001; ***p < 0.0001

Fig. 4

Association of microbial abundance in normal breast tissues with breast cancer risk factors. Pearson’s correlation analysis was performed to examine the link of the abundance of Ralstonia (Ral), Acetobacter aceti (Acet), Lactobacillus vini (L.vini), Lactobacillus paracasei (L.par), and Xanthomonas sp. (Xantho) with breast cancer susceptibility (A), racial background (B), smoking (C), menopausal status (D), parity (E) and recent and past breastfeeding (F). *p < 0.05; **p < 0.001; ***p < 0.0001

Fig. 5

Link between microbial composition and gene enrichment in the breast. Transcriptomic analysis of the normal breast tissues (N = 190) was performed. (A) Venn diagram shows the common differentially expressed genes linked to the microbial abundance of Acetobacter aceti (A. aceti), Lactobacillus vini (L. vini), Lactobacillus paracasei (L. paracasei), Xanthomonas sp, and Ralstonia. B) STRING-generated molecular pathway of the genes enriched in both Lactobacillus vini and Xanthomonas sp. analyses. KEGG pathway analysis was performed for genes enriched in tissues abundant with Acetobacter aceti (C), Lactobacillus vini (D), Lactobacillus paracasei (E), Xanthomonas sp. (F) and Ralstonia (G)

Fig. 6

Major microbiome changes in the breast: from healthy to tumor. The normal breast tissue (Normal) shows low to undetectable levels of Ralstonia, whereas the tissue adjacent to tumor (NAT), which may represent an intermediate stage between normal and cancer status, and the tumor contain increasing level of this bacterial species. The opposite trends is observed for Lactobacillus, Acetobacter aceti, and Xanthomonas sp., which appear highly abundant in normal breast tissue and scarce in NAT and tumor