Fecal microbiome transplant from patients with lactation mastitis promotes mastitis in conventional lactating mice

Abstract

Introduction: Lactation mastitis seriously severely affects the health of lactating females and their infants, yet the underlying causes of clinical lactation mastitis remain unclear.

Methods: In this study, we used microbiota-humanized mice as a model to investigate the role of gut microbiota in lactation mastitis. We compared the fecal microbiota of lactation mastitis patients and healthy individuals and conducted fecal microbiota transplantation (FMT) experiments in an antibiotic-pretreated mouse model to test whether gut microbes contribute to human lactation mastitis.

Results: Our results showed that gut microbiota diversity was reduced and dysbiosis was present in lactating mastitis patients. FMT from lactation mastitis patients (M-FMT), but not from healthy individuals (H-FMT), to antibiotic-treated mice resulted in lactation mastitis. The inflammation in mice caused by gut microbiota from lactating mastitis patients appears to be pervasive, as hepatocytes from mice that received feces from lactating mastitis patients showed marked swelling. In addition, serum pro-inflammatory factors, including IL-4, IL-17, MPO, IL-6, IL-1β, and TNF-α, were significantly increased in the M-FMT group. The Firmicutes/Bacteroidetes ratio (F/B), a biomarker of gut dysbiosis, was significantly increased in the M-FMT group. At the phylum level, Actinobacteria were significantly increased, and Verrucomicrobia were significantly decreased in the M-FMT group. At the genus level, Ruminococcus and Faecalibacterium were significantly reduced, while Parabacteroides were significantly increased in the feces of both patients with lactation mastitis and M-FMT mice. Moreover, our study revealed an "amplification effect" on microbiota differences and mastitis disease following human-to-mouse FMT.

Conclusion: Collectively, our findings demonstrate that the gut microbiota in lactating mastitis patients is dysbiotic and contributes to the pathogenesis of mastitis.

Keywords: fecal microbiota transplantation; lactation mastitis; mammary gland; microbiome; pro-inflammatory.

Figure 1

The pattern of immune response markers and intestinal flora in healthy females and patients with lactation mastitis. (A) The number of white corpuscles, (B) the percentage of neutrophils and (C) lymphocytes, and the level of (D) CRP were counted. (E) Alpha diversity, including Chao 1 and Shannon diversity index. (F) shows the gut microbiota composition in terms of relative abundance at the phylum level. (G) shows the gut microbiota composition in terms of relative abundance at the order level. (H,I) Relative abundances of bacteroidales and Bifidobacteriales between healthy and mastitis. The data represent the means ±SEMs (n = 8–9 person per group). *p < 0.05, ***p < 0.001 by t-test.

Figure 2

Administration of fecal bacteria from patients with lactation mastitis to mice induced a series of mastitis-like phenotypes. (A) Schematic representation of FMT. Fresh fecal from five healthy and five lactation mastitis donors were mixed and used as a single source for H-FMT and M-FMT mice. Following antibiotic treatment, the recipient mice were randomly divided and orally inoculated daily for three consecutive days and two times per week for 62 days with prepared fecal contents. (B) Pathological changes in mammary gland surface, where two abdominal mammary glands were swelling in the mastitis group of mice on day 65 after FMT. The breast of the mice was highlighted by a red frame. (C) Representative photomicrographs of hematoxylin–eosin (H&E) stained mammary tissue. (D) The injury score of mammary glands (n = 8–10 mice per group). (E) CD45 immunohistochemical staining sections at ×400 magnification. (F) A bar graph showing the number of CD45⁺ cells in the H-FMT and M-FMT mice (n = 8–10 mice per group). (G) MPO immunohistochemical staining sections at ×400 magnification. (H) A bar graph showing the number of MPO⁺ cells in the H-FMT and M-FMT mice. The data represent the means ±SEMs (n = 8–10 mice per group). *p < 0.05, ***p < 0.001 by t-test. Scale bar, 100 μm.

Figure 3

The murine inflammation induced by lactation mastitis patient’s intestinal microbiota seemed pervasive. (A) Representative photomicrographs of the hematoxylin–eosin-stained liver (×200) (left). The injury score of the liver (right). (B) Representative photomicrographs of hematoxylin–eosin-stained colon (×100) (left). The injury score of colons (right). Quantification of inflammatory cytokines including IL-4 (C), IL-17 (D), MPO (E), IL-6 (F), IL-1β (G), and TNF-α (H) in serum using ELISA. The assays were all performed for the two groups of mice on Day 65 after FMT. The data represent the means ±SEMs (n = 8–10 mice per group). *p < 0.05, **p < 0.01, ***p < 0.001 by t-test. Scale bar, 100 μm.

Figure 4

The distinction between healthy and lactation mastitis intestinal microbiota in the mice after FMT. (A) Alpha diversity, including Chao 1 and Shannon diversity index. (B) Venn diagram. (C) Unweighted UniFrac distances between healthy and lactation mastitis groups in a PCoA. (D) Hierarchical clustering tree on genus level using the Bray–Curtis distance (left)and the component proportion of bacteria at the genus level in each group (right). (E) Relative abundance of the phyla in the feces. (F) The Firmicutes: bacteroidetes ratio(F/B) was calculated as a biomarker for gut dysbiosis. (G,H) Relative abundances of Actinobacteria and Verrucomicrobia between H-FMT and M-FMT groups. n = 8–10 mice per group. *p < 0.05, ***p < 0.001 by t-test.

Figure 5

Comparison of mastitis-associated microbiota in humans and those in the mouse. PCoA clustering (A) and relative similarity (B) of person and mice microbiota based on organismal structure (via Meta-Storm distance) were shown. (C) Unique and shared OTUs before and after FMT in person and mice. (D) Random Forest (RF) classification of genus-level gut microbiota in two groups of volunteers. (E) Random Forest (RF) classification of gut microbiota in two groups of recipient mice. (F–H) Relative abundances of Ruminococcus, Faecalibacterium, and Parabacteroides between the human and mouse groups. n = 8–15 person or mice per group. *p < 0.05 by t-test.

Figure 6

Univariate linear regression between Ruminococcus, Faecalibacterium and Parabacteroides relative abundance in fecal samples and mammary tissue injury score, CD45⁺ cells and MPO⁺ cells of the matched mice. (A) Spearman correlation between Ruminococcus and mammary tissue injury score. (B) Spearman correlation between Ruminococcus and CD45⁺cells. (C) Spearman correlation between Ruminococcus and MPO⁺ cells. (D) Spearman correlation between Faecalibacterium and mammary tissue injury score. (E) Spearman correlation between Faecalibacterium and CD45⁺cells. (F) Spearman correlation between Faecalibacterium and MPO⁺ cells. (G) Spearman correlation between Parabacteroides and mammary tissue injury score. (H) Spearman correlation between Parabacteroides and CD45⁺ cells. (I) Spearman correlation between Parabacteroides and MPO⁺ cells.

Figure 7

Univariate linear regression between Ruminococcus, Faecalibacterium and Parabacteroides relative abundance in fecal samples and the white blood cell count, the percentage of neutrophils and the percentage of lymphocytes of the matched mice. (A) Spearman correlation between Ruminococcus and the white blood cell count. (B) Spearman correlation between Ruminococcus and the percentage of neutrophils. (C) Spearman correlation between Ruminococcus and the percentage of lymphocyte cells. (D) Spearman correlation between Faecalibacterium and the white blood cell count. (E) Spearman correlation between Faecalibacterium and the percentage of neutrophils. (F) Spearman correlation between Faecalibacterium and the percentage of lymphocyte cells. (G) Spearman correlation between Parabacteroides and the white blood cell count. (H) Spearman correlation between Parabacteroides and the percentage of neutrophils. (I) Spearman correlation between Parabacteroides and the percentage of lymphocyte cells.