Analysis of Gut Microbiota in Patients with Breast Cancer and Benign Breast Lesions

Breast cancer (BC) is the most common cancer among women worldwide. BC remains a significant cause of mortality in women, despite the use of adjuvant chemotherapeutic and hormonal agents (Braden et al. 2014). Genetic and other established risk factors such as early menarche age, high body mass index (BMI), and sedentary lifestyle have been associated with the onset and progression of BC. Benign breast lesions (BBLs), including fibroadenoma, are commonly found in young women. Estrogens and their receptors are implicated in the onset and progression of BBLs. Accumulating data have indicated that alterations in the host microbiome, primarily intestinal microbiota, may contribute to the pathogenesis of both gastrointestinal and extra-intestinal tumors (Belkaid and Hand 2014; Dzutsev et al. 2017).

The number of genes in the human intestinal microbiota, regarded as an alternative genome in humans, is nearly 150 times higher than that of the human genome (Zhu et al. 2010). This intestinal ecosystem is involved in a dynamic interaction with host cells, microbes, and food. Besides, it acts as a multi-dimensional “microbial organ” by enhancing the synthesis of essential amino acids and vitamins, producing small molecules, nutritional absorption, metabolism of bile acids, activation of immune cells, and inactivation of toxins and carcinogens (Eslami-S et al. 2020). The remarkable contribution of the gut microflora to human health and disease has been extensively recognized. It has been speculated that changes in the constitution and functions of the gut microbiome might contribute to the onset and progression of BC and BBLs.

Although some studies reported higher microbial diversity in BC patients than healthy controls (Gopalakrishnan et al. 2018; Zhu et al. 2018), other investigations found less microbial diversity in post-menopausal BC subjects (Goedert et al. 2015; 2018). In addition, gut dysbiosis in individuals with BBLs is not fully understood. Therefore, 16S ribosomal RNA (16S rRNA) gene sequencing technology was utilized to explore intestinal microbiota dysbiosis in BC and BBL patients.

Patient characteristics. All participants were Chinese women (healthy controls, n = 20; BC: n = 26; BBL: n = 20). There were no significant differences in age and BMI among the three groups (Table I).

Different groups showed different OTUs. Sequencing showed 3,567,593 effective sequences (average of 54,054 sequences per sample). In the three groups, 723 OTUs were detected. Among them, 517 OTUs were common to the three groups, but 64 were exclusive to BC patients, 77 were exclusive to individuals with BBLs, and 65 were exclusive to healthy controls (Fig. 1).

Fig. 1

Microbiota composition analysis. The same four bacteria were identified at the phylum level to be dominant in all three groups, including Firmicutes, Bacteroidetes, Proteobacteria, and Actinobacteria, of which Firmicutes and Bacteroidetes were predominant in each group (Fig. 2A). In addition, 23 bacteria genera with relative abundance higher than 0.5% were identified among the three groups, including Veillonella, Dialister, Oscillibacter, Lachnospiracea_incertae_sedis, Parasutterella, Megasphaera, Prevotella, Roseburia, Bifidobacterium, Clostridium_XlVa, Barnesiella, Eubacterium, Escherichia, Faecalibacterium, Phascolarctobacterium, Bacteroides, Blautia, Megamonas, Gemmiger, Parabacteroides, Ruminococcus, Alistipes, and Succinivibrio (Fig. 2B). The patterns of microbial composition were highly variable among these three groups.

Fig. 2

As shown in Table II, compared with the healthy control group, the relative richness of five bacterial genera was increased in the BC group (Escherichia, Peptoniphilus, Bilophila, Lactobacillus, and Porphyromonas) while the relative richness of fifteen bacterial genera was decreased (Faecalibacterium, Lachnospiracea_incertae_ sedis, Collinsella, Alistipes, Anaerofilum, Christensenella, Butyricimonas, Erysipelothrix, Acidaminococcus,

Victivallis, Eubacterium, Tissierella, Hydrogenoanaerobacterium, Cloacibacillus, and Oxalobacter). Also, compared with the healthy controls, the relative richness of five bacterial genera was increased in patients with BBLs (Escherichia, Peptoniphilus, Coprobacillus, Lactobacillus, and Porphyromonas), whereas the relative richness of eight bacterial genera was decreased (Collinsella, Alistipes, Megamonas, Butyricimonas, Acidaminococcus, Asaccharobacter, Tissierella, and Cloacibacillus).

Biodiversity analysis. Alpha diversity indices (Sobs and Chao1) are shown in Fig. 3. Compared with the healthy controls, BC patients had significantly lower alpha diversity indices (Sobs index, p = 0.019; Chao1 index, p = 0.033). There were no differences in Sobs and Chao1 indices between patients with BBLs and healthy individuals (p = 0.279, p = 0.314, respectively).

Fig. 3

In addition, beta diversity assessments based on weighted UniFrac were markedly different among the three groups (both p < 0.001, Fig. 4A, B). These results suggested an altered gut microbiota composition in BC and BBL patients.

Fig. 4

The weighted UniFrac PCoA plot showed no visible separation among the three groups (Fig. 5A), but the PLS-DA analysis separated the three groups (Fig. 5B). Collectively, this observation revealed that the structure of the gut microbiota community was different among the three groups.

Fig. 5

Bacterial taxonomic differences. Prevotella, Porphyromonas, Peptoniphilus, and Megamonas were the major taxonomic groups in the BC group, whereas Lactobacillus, Escherichia, and Coprobacillus were the major taxonomic groups in the BBL group. Cloacibacillus, Asaccharobacter, Christensenella, Alistipes, Tissierella, Hydrogenoanaerobacterium, Butyricimonas, Acidaminococcus, Oxalobacter, Collinsella, and Eubacterium were the major taxonomic groups in the healthy controls (Fig. 6).

Fig. 6

This cross-sectional study reveals a decreasing trend in gut diversity of BC and BBL subjects. It was similar to a previous case-control study, which showed a lower diversity and altered composition of microbiota in the fecal samples of postmenopausal BC patients (Goedert et al. 2015). In another study, milk from mastitis patients demonstrated microbiota dysbiosis, including lower microbial diversity with increased opportunistic pathogens and reduced commensal organisms (Patel et al. 2017). The diversity of gut microbiota is essential for maintaining health (Katagiri et al. 2019). A low diversity of gut microbiota is often a hallmark of intestinal dysbiosis and has been linked to inflammatory bowel disease, obesity, allergic rhinitis, and gastric carcinoma (Ferreira et al. 2018; Watts et al. 2021). Mounting evidence has shown that growing up in microbe-rich environments, for instance, traditional farms, improves children’s health, and a high diversity has been associated with increased health in the elderly (Claesson et al. 2012; Le Chatelier et al. 2013). In addition, prebiotics, probiotics, and diverse nutrition have been shown to decrease the risk of BC (Newman et al. 2019; Goubet et al. 2021; Jiang and Fan 2021; Méndez Utz et al. 2021; Pourbaferani et al. 2021). On the other hand, decreased microbial diversity caused by long-term use of antibiotics has been shown to increase the BC risk (Sergentanis et al. 2010; Wirtz et al. 2013; Simin et al. 2020), although conflicting results have been observed (García Rodríguez and González-Pérez 2005; Sørensen et al. 2005).

Previous studies showed a direct and strong association between fecal microbiota diversity and estrogen levels in women (Flores et al. 2012). Generally, estrogens and their metabolites undergo sulfation and glucuronidation in the liver. The conjugated estrogens can then be excreted via stool and urine. Intestinal bacteria can directly affect estrogen production by secreting β-glucuronidase (GUS), an enzyme that depolymerizes estrogens into their active forms, to control the concentration of estrogens reabsorbed into the enterohepatic circulation. In addition, gut microbes synthesize estrogen-like compounds or estrogen mimics from the daily diet.

Furthermore, multiple bacterial metabolites (e.g., short-chain fatty acids, acetate, butyrate, pyruvate, formate, active amines, bile acids and derivatives, indole derivatives, etc.) can be involved in cancer cell growth, apoptosis, and invasion, epithelial-to-mesenchymal transition, and antitumor immune activity (Kovács et al. 2021). Changes in microbiome composition will lead to changes in the profiles of metabolites (Kovács et al. 2021). We, therefore, speculate that the proportion of microbiota-encoded GUS enzymes changed, thus affecting the metabolism of steroid hormones, metabolite profiles, and alpha diversity of intestinal microorganisms in BC and BBL patients.

Furthermore, microbial diversity can affect the efficacy of anticancer therapy. Fecal samples from melanoma patients receiving anti-PD-1 treatment exhibited a more diverse microbiome, and patients had significantly longer progression-free survival. The microbiota of immune therapy responders may upregulate the immune response by enhancing antigen presentation or increasing T cell recruitment in the local tumor environment (Gopalakrishnan et al. 2018).

Similarly, gut microbiota conditions the metastasis and therapeutic efficacy of trastuzumab in HER2-positive BC (Ingman 2019; Di Modica et al. 2021). Probiotic administration can significantly increase the number of bacterial species and the bacterial diversity assessed with the Chao1 index in overweight BC survivors (Pellegrini et al. 2020). Therefore, we hypothesize that the reduced microbial diversity may affect the treatment efficacy of BC patients.

According to the LEfSe analysis, Prevotella, Porphyromonas, Peptoniphilus, and Megamonas were indicator bacterial species in BC patients. Prevotella and Porphyromonas were also identified as potential microbial markers for postmenopausal BC patients (Amanatullah et al. 2017; Zhu et al. 2018). The two genera are also associated with colorectal cancer and precancerous adenomas (Warren et al. 2013; Lasry et al. 2016). Notably, Prevotella has been found on breast skin and mammary tissue (Urbaniak et al. 2014; Hieken et al. 2016; Urbaniak et al. 2016). Transferring microorganisms from the intestine to the breast tissue leads to increased systemic inflammation in BC and is therefore considered a cause of BC (Rao et al. 2007). Previous studies have shown that inflammatory indicators, such as platelet/lymphocyte ratio and lymphocyte/monocyte ratio, significantly influence the prognosis of various cancers, and neutrophilia is associated with a poor prognosis of BC (Lakritz et al. 2015). Systemic inverse interactions among microbes, interleukin-6 (IL-6), and neutrophils have been noted in BC (Rutkowski et al. 2015). A high lymphocyte/neutrophil ratio increases the risk of relapse in BC patients (Margolis et al. 2007). Therefore, we hypothesize that Prevotella was involved in the inflammatory response in BC patients. In addition, Prevotella can activate Toll-like receptor 2, leading to the production of Th17-polarizing cytokines by antigen-presenting cells, including IL-23 and IL-1. Prevotella can also stimulate epithelial cells to produce IL-8, IL-6, and CCL20, which promote mucosal Th17 immune responses (Larsen 2017). Porphyromonas uenonis showed a weak positive correlation with CD19 in BC patients (Zhu et al. 2018). Peptoniphilus were abundant in endocrine receptor-positive, human epidermal growth factor receptor 2-positive, and triple-negative BC types (Banerjee et al. 2018). Megamonas decreased significantly in patients with Bechet’s disease, and this alteration may be associated with immune aberration (Shimizu et al. 2019). Thus, we infer that the bacterial abundance changes were involved in the disruption of immune homeostasis in BC patients.

Among the genera with a decreased abundance in patients with BC, Collinsella has been associated with a cancer-free status and a better prognosis in BC patients (Terrisse et al. 2021). In the present study, Alistipes was decreased in the gut of patients with BC but was increased in the nipple aspiration fluid of patients with BC in a previous study (Laborda-Illanes et al. 2020). In BC, the number of Anaerofilum in the gut appears to be associated with the number of tumor-infiltrating lymphocytes (Shi et al. 2019). A lower number of gut Butyricimonas have been reported before in BC (Bobin-Dubigeon et al. 2021), supporting the present study. Acidaminococcus and Cloacibacillus have been observed with different gut abundance among BC subtypes associated with prognoses (Wu et al. 2020; Yang et al. 2021).

Even if the association between gut dysbiosis and BC has been extensively studied (Kovács et al. 2021), the association between gut dysbiosis and BBLs has not been extensively explored before. The present study suggests that women with BBLs display changes in the gut microbiome compared with healthy women. Many BBLs are precursor lesions in a spectrum of lesions leading to BC or to be markers of increased risk of breast cancer (Hartmann et al. 2005; Worsham et al. 2009; Johansson et al. 2021). Some of the bacteria found to be increased or decreased in patients with BBLs were also observed in patients with BC (increased Escherichia, Peptoniphilus, Lactobacillus, and Porphyromonas; decreased: Collinsella, Alistipes, Butyricimonas, Acidaminococcus, Tissierella, and Cloacibacillus). It is supported by Yang et al. (2021), who showed that among 31 genera of gut microbiota, only one (Citrobacter) was different between patients with BC and BBL. Still, Meng et al. (2018) reported differences in gut microbiota between BBL and BC. The roles of the various bacteria in immunity and cancer development discussed above might also apply to BBLs. Therefore, it could be hypothesized that a gut dysbiosis is an early event in the development of BBLs and BC and that the changes in gut microbiota are an early event in the spectrum of events from normal breast tissue to BBLs to BC. Ideally, longitudinal studies should be performed to examine this point.

In this study, the populations of Escherichia and Lactobacillus were significantly upregulated in BBL patients. The impaired barrier function allows bacterial access to the intestinal epithelium, enabling the delivery of toxins. Escherichia coli can putatively induce tumorigenesis by generating DNA mutagens such as genotoxin colibactin (Arthur et al. 2012). Staphylococcus aureus is an important factor inducing mutation of the MED12 gene, which may contribute to uterine leiomyomas and breast fibroadenomas (Bullerdiek and Rommel 2018). However, no significant changes in S. aureus were detected. Surprisingly, Lactobacillus was upregulated in BBL patients. Lactobacillus is usually considered to be a beneficial bacterium. Oral consumption of Lactobacillus acidophilus can decrease in fecal enzyme activity of GUS (Kwa et al. 2016), thereby reducing the estrogen burden in the body. Lactobacillus reuteri was found to be helpful in suppressing mammary tumorigenesis in genetically susceptible Her2 mutant mice (Lakritz et al. 2014). In addition, Lactobacillus exhibited anti-inflammatory properties in E. coli-stimulated bovine mammary epithelial cells (Bouchard et al. 2015). Hence, the upregulation of Lactobacillus in BBL patients may be due to the presence of the tumor, which allows the intestine to attract more beneficial bacteria to fight it.

In conclusion, non-malignant breast diseases have been far less studied. However, the great potential of intestinal microbiota in the development and treatment of benign breast diseases cannot be overlooked. The use of probiotics to treat mastitis in breastfeeding women has been reported. Probiotics are potentially effective at eliminating chronic subclinical infections as antibiotic treatment (Arroyo et al. 2010). Therefore, more related studies are required in the future.

Herein, we performed 16S rRNA gene sequencing of fecal samples collected from BC and BBL patients and healthy controls matched by gender, age, and BMI. Compared with healthy controls, BC and BBL patients showed a decreasing trend in intestinal microbiota diversity, which may be associated with their pathogenesis. The up- or down-regulated strains may be an essential indicator of the initiation of BC and BBLs. These results may provide a valuable reference for future related studies. However, several limitations must be addressed in future studies. First, species-level differences were not captured due to the limitations of 16S rRNA sequencing. More studies with whole-genome sequencing are needed. Second, this study was a single-center study with relatively small sample size. Third, the dietary structures differed among individuals, which might have influenced the results. Additional studies should be conducted with larger samples to explore the functions of intestinal flora in BC and BBLs.