Corynebacterium parakroppenstedtii secretes a novel glycolipid to promote the development of granulomatous lobular mastitis

Abstract

Granulomatous lobular mastitis (GLM) is a chronic idiopathic granulomatous mastitis of the mammary gland characterized by significant pain and a high propensity for recurrence, the incidence rate has gradually increased, and has become a serious breast disease that should not be ignored. GLM is highly suspected relative to microbial infections, especially those of Corynebacterium species; however, the mechanisms involved are unclear, and prevention and treatment are difficult. In this study, we demonstrated the pathogenicity of Corynebacterium parakroppenstedtii in GLM using Koch's postulates. Based on the drug sensitization results of C. parakroppenstedtii, and utilizing a retrospective study in conjunction with a comprehensive literature review, we suggested an efficacious, targeted antibiotic treatment strategy for GLM. Subsequently, we identified the pathogenic factor as a new type of glycolipid (named corynekropbactins) secreted by C. parakroppenstedtii. Corynekropbactins may chelate iron, cause the death of mammary cells and other mammary -gland-colonizing bacteria, and increase the levels of inflammatory cytokines. We further analyzed the prevalence of C. parakroppenstedtii infection in patients with GLM. Finally, we suggested that the lipophilicity of C. parakroppenstedtii may be associated with its infection route and proposed a possible model for the development of GLM. This research holds significant implications for the clinical diagnosis and therapeutic management of GLM, offering new insights into targeted treatment approaches.

Fig. 1

Exploration of possible causative pathogen of Granulomatous lobular mastitis (GLM). a Relative abundance of bacteria in different breast samples. b Comparison of the alpha-diversity and Shannon index values of the bacteria in different breast samples. Data are expressed as median + interquartile range (IQR), and statistical significance was calculated using the Wilcoxon rank sum test with the Bonferroni method for multiple testing correction. c Principal coordinate analysis of the microbiota based on Bray‒Curtis heterogeneity of bacteria from different breast samples. d Linear discriminant analysis effect size of the bacterial microbiota in different breast samples. g: genus, c: class, f: family, s: species, d: domain. e Relative abundance of bacteria in the pus and breast tissues of 11 patients with GLM. f The relative abundances of Escherichia fergusonii, Corynebacterium kroppenstedtii, and Acinetobacter lwoffi in the pus and breast tissues of 11 patients with GLM. Data are expressed as the median + IQR, and a paired Wilcoxon rank sum test was performed for comparison. g Gram staining results of tissue sections from patients with GLM

Fig. 2

C. parakroppenstedtii causes GLM in rats. a Images of C. parakroppenstedtii. Left: typical colonies of C. parakroppenstedtii in BHIY with Tween 80. Middle: microscopic view of C. parakroppenstedtii. Right: C. parakroppenstedtii observed by scanning electron microscopy. b Genome circle map of C. parakroppenstedtii P1. From outside to inside, coding genes (positive-sense strand), coding genes (negative-sense strand), tRNA (orange) and rRNA (purple), gene islands (green), GC ratio, GC skew, and sequencing depth. c C. parakroppenstedtii P1 causes changes in cytokine levels in MCF-10A cells. Three biological replicates were set up for each group. Data are presented as the mean values ± standard deviation (SD), and Kruskal–Wallis tests were performed to calculate statistical significance. d Verification of the occurrence of GLM after C. parakroppenstedtii P1 infection. The experiment consisted of a control (phosphate-buffered saline [PBS]) group (12 rats) and a C. parakroppenstedtii P1 group (12 rats). After breast fat pad injection, six rats in each group were sacrificed on day 3 and day 10. Pathologic results are typical representative images selected from the supplementary fig. 2. Scale bar represents 100 μm. e The serum cytokine levels in the rats were determined using a Luminex assay. Data were compared between the rats with obvious inflammation (day 3: n = 4; day 10: n = 5) and those in the control group (n = 6). Data are presented as the mean values ± standard error of the mean (SEM), and Kruskal–Wallis tests were performed to calculate statistical significance. f The levels of cytokines in the breast tissue of rats were determined by immunofluorescence staining. Data were compared between the rats with obvious inflammation (day 3: n = 4; day 10: n = 5) and those in the control group (n = 6). Data are presented as the mean values ± SEM, and Kruskal–Wallis tests were performed to calculate statistical significance. g Verification of the occurrence of GLM after infection with different C. parakroppenstedtii strains or the blank control (PBS). There were five rats in each group. The results were observed after sacrifice on day 10 after breast fat pad injection. h Antimicrobial susceptibility testing of C. parakroppenstedtii with different antibiotics performed via disk diffusion (Becton Dickinson Deutschland, Heidelberg, Germany) as suggested for Corynebacterium by the European Committee on Antimicrobial Susceptibility Testing (http://www.eucast.org)

Fig. 3

Novel glycolipids generated by C. parakroppenstedtii and their roles. a Concentration of serum iron in 10 patients with GLM who had a confirmed diagnosis of C. kroppenstedtii infection. b Verification of iron deficiency in rats infected with C. parakroppenstedtii P1. The experiment consisted of a control (PBS) group (7 rats) and a C. parakroppenstedtii P1 group (10 rats). After breast fat pad injection, rats in each group were sacrificed on day 10. c Iron-related indicators in the serum of rats with GLM. Fe: iron ions; TIBC: total iron binding capacity; TS: transferrin saturation; FER: ferritin. Data are presented as the mean values ± SEM, and group comparisons were calculated using Kruskal–Wallis test or Wilconxon singed rank exact test (P value was adjusted using Benjamini-Hochberg correction) as appropriate. d High-performance liquid cohrmatography analysis of culture supernatants from C. parakroppenstedtii fermentation. e Structures of corynekropbactins from C. parakroppenstedtii. f Compounds on CAS detection media, corynekropbactin I: 0.8 mg; corynekropbactin II: 0.8 mg; fraction: 50 µL fraction containing corynekropbactins during purification. g Growth curves of C. parakroppenstedtii P1 in different PGT culture media with Tween 80. Except for the Fe-free group, the corresponding compounds of the remaining groups were added at 10 µM. Three biological replicates were set up for each group. Data are presented as the mean values ± SD. h The relative ratio of corynekropbactins in different PGT culture media with soy oil. The relative ratio was calculated as the ratio of the production of corynekropbactins in each group to that in the Fe-free group. Fe²⁺ and Fe³⁺ were added at 10 µM. Three biological replicates were set up for each group. Data are presented as the mean values ± SD, and ordinary one-way analysis of variance with Bonferroni correction for multiple comparison was performed to calculate statistical significance. *P < 0.05; **P < 0.01; ***P < 0.001. i Compounds 1 and 2 caused changes in the levels of cytokines in MCF-10A cells. Four biological replicates were set up for each group. Data are presented as the mean values ± SD, and Dunn’s test with Benjamini–Hochberg correction for multiple comparisons was performed to calculate statistical significance. j Liquid chromatography (LC)-tandem mass spectrometry (MS/MS) extracted ion chromatograms of supernatants of MCF-10A or C. parakroppenstedtii-treated MCF-10A cells. m/z of extracted ions: 1-811.3582, 2-1037.5491, 3-837.3735, 4-813.3735, 5-979.4710, 6-1007.5027, 7-955.4711, 8-983.5058. (k) LC-MS/MS extracted ion chromatogram of PBS-injected or C. parakroppenstedtii-injected rats. m/z of extracted ions: 1-811.3582, 2-1037.5491, 3-837.3735, 4-813.3735, 5-979.4710, 6-1007.5027, 7-955.4711, 8-983.5058. The experiment consisted of a control (PBS) group (6 rats) and a C. parakroppenstedtii P1 group (6 rats). Breast fad pad injections were performed on day 1 and day 4, and the rats in each group were sacrificed on day 8

Fig. 4

Phylogenetic trees of C. kroppenstedtii-like strains. The phylogenetic tree was constructed using whole-genome sequences, the strains marked in red were whole-genome sequenced in this study, while the whole-genome sequences of the strains marked in black were obtained from the NCBI database. Five major clusters with different background colors indicate that the C. kroppenstedtii-like strains can be divided into five species. The strains sequenced in this study were isolated from patients with GLM, and the sources of the remaining strains were defined according to their reported genome databases

Fig. 5

Lipophilicity of C. parakroppenstedtii and its possible infection route. a Biosynthetic and oxidative pathways of fatty acids in bacteria. The green-labeled genes were predicted by a Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology search of the genome of C. parakroppenstedtii, and the alpha and beta subunits of acetyl-CoA carboxylase (gray) were predicted by KEGG orthology search of the genome of C. parakroppenstedtii; however, it is uncertain whether acetyl-CoA carboxylase can fully function. The unlabeled genes were not predicted. b Growth status of C. parakroppenstedtii and C. amycolatum in different culture media. P4: C. parakroppenstedtii P4; P3: C. parakroppenstedtii P3; P1: C. parakroppenstedtii P1; P2: C. parakroppenstedtii P2; BAA-2858: C. parakroppenstedtii BAA-2858; Control: C. amycolatum. c Growth status of C. parakroppenstedtii in BHIY medium supplemented with different washing agents at 0.1% or 1%. 1–5, 10-fold serial dilutions of sample. d The growth status of C. parakroppenstedtii after UV irradiation or drying. 1–5, 10-fold serial dilutions of sample. e Growth status of C. parakroppenstedtii after natural shade drying. Scale bar represents 5 mm

Fig. 6

Hypothesis of the possible process of GLM development. This figure illustrates that fatty acids in the breast facilitate the infection of C. parakroppenstedtii, which subsequently secretes corynekropbactins. Corynekropbactins may then act as iron chelators, leading to the death of breast cells and other mammary-gland-colonizing-bacteria, and inducing the upregulation of IL-6. These multiple roles may contribute to the development of GLM