What this is
- This research investigates the role of (SA) in exacerbating mastitis linked to in cows.
- It highlights how SA contributes to gut microbiota disruption and systemic inflammation through the microbiota-gut-mammary axis.
- The findings suggest potential strategies for mastitis intervention by regulating gut metabolism.
Essence
- aggravates -induced mastitis by promoting microbiota disruption and systemic inflammation in cows, indicating a microbiota-gut-mammary axis involvement.
Key takeaways
- Cows with () show increased levels of sialic acids, with Neu5Gc and Neu5Ac rising 39.7× and 16.1× respectively. These metabolites correlate with mastitis parameters.
- Administration of to antibiotic-treated mice induces significant mastitis, evidenced by increased inflammatory markers and gut barrier disruption. This effect is not observed in healthy mice.
- Targeting the inhibition of opportunistic pathogens and regulating gut homeostasis through commensal bacteria can alleviate -facilitated mastitis, suggesting a potential intervention strategy.
Caveats
- The study primarily uses animal models, which may limit direct applicability to human conditions. Further research is needed to confirm findings in broader contexts.
- The specific mechanisms by which influences mastitis development remain to be fully elucidated, warranting additional investigation.
Definitions
- Sialic acid: A nine-carbon sugar molecule that plays a role in cellular interactions and is involved in various biological processes.
- Gut dysbiosis: An imbalance in the gut microbiota composition, often leading to negative health outcomes.
- Subacute ruminal acidosis (SARA): A condition in ruminants characterized by a decrease in ruminal pH due to high-grain diets, leading to metabolic disturbances.
AI simplified
Introduction
The complex intestinal mutualistic community, referred to as the gut microbiota, has been proven to be one of the most important regulators affecting physiological homeostasis and disease outcomes [1, 2]. A large number of studies indicated that the gut microbiota regulates host health through nutritional absorption, barrier maintenance, immune and metabolic regulation [3], and disruption of the fragile balance in the gut microbiota, termed gut dysbiosis [4], plays a central role in the initiation, development, and outcome of many diseases, including inflammatory bowel disease [5], fatty liver [6], obesity [7], cancer [4], and infections [8]. Commonly, metabolic alterations derived from gut microbiota are the most direct way of mediating host-microbiota interactions and influencing disease progression. For example, gut microbiota-derived short-chain fatty acids (SCFAs) contribute to maintaining intestinal anaerobic conditions [9], regulating mucosal and systemic Treg/Th17 balance [10], and are involved in the metabolic fitness of the brain innate immune system [11]. Microbiota-catabolized ligands of the aryl hydrocarbon receptor (AhR) from dietary tryptophan reduce central nervous system inflammation through regulating astrocyte activity by type I interferons and microglial-astrocyte interaction [12, 13]. On the other hand, gut dysbiosis-derived metabolites increase the risk for diseases. The microbial metabolite 4-ethylphenyl sulfate (4-EPS) can alter brain activity and anxiety behavior by impairing oligodendrocyte maturation and reducing oligodendrocyte-neuron interactions [14]. Koh et al. found that histidine-derived imidazole propionate by the gut microbiota impairs glucose tolerance and insulin signaling through p38-mediated activation of the mechanistic target of rapamycin complex 1 [15]. Moreover, gut microbiota production of trimethyl-5-aminovaleric acid (TMAVA) facilitates cardiac hypertrophy by reducing fatty acid oxidation [16]. These results suggest that exploring host or microbiota-mediated metabolic alterations is important for achieving an in-depth understanding of the pathogenesis of gut microbiota-associated distal diseases.
Mastitis is a common and severe disease for humans and animals that increases the risk for breast cancer [17, 18] and impairs milk yield and quality, making it a major threat to women's health and hindering the development of the dairy industry. Mounting evidence suggests a possible “microbiota-gut-mammary axis” in the context of mastitis as shown by gut dysbiosis initiating and facilitating the development of mastitis. In detail, studies reported that fecal microbiota transplantation (FMT) from mastitis cows to mice induces mastitis [19, 20]. Our previous studies also demonstrated that antibiotic-treated mice have increased susceptibility to Staphylococcus aureus (S. aureus)- and Escherichia coli (E. coli)-induced mastitis [21, 22], confirming the important role of the microbiota-gut-mammary axis in mastitis pathogenesis. A previous study showed that alteration of the gut microbiota in response to different dietary patterns induced distinct mammary metabolic profiles [23]. Impaired beneficial metabolite levels, including SCFAs and microbiota-metabolic AhR ligands, are reported to regulate the outcome of pathogen-induced mastitis in mice [21, 22]. However, little is known about the direct link between gut microbial metabolites and the risk of mastitis.
There is an increased feed ratio in a high-grain diet (HGD) due to the high-nutritional demand of cows during the milk period and the pursuit of milk yield and quality in dairy farms. However, excessive consumption of a HGD and low in dietary fiber will reduce the ruminal pH and disrupt the composition and functions of the ruminal microbiota, a phenomenon referred to as subacute ruminal acidosis (SARA) [24]. A previous study found that SARA cows have increased susceptibility to pathogen-induced mastitis [25]. Our previous result also supported the hypothesis that SARA may serve as a driver for mastitis, as shown by the long term consumption of a HGD increasing systemic inflammatory responses and mastitis [26, 27], which was associated with increased ruminal lipopolysaccharide (LPS) levels derived from the enriched Gram-negative bacteria in SARA cows [26, 28]. However, whether dysbiosis-induced metabolic changes promote the expansion of ruminal opportunistic pathogen and LPS release and regulate the development of mastitis is poorly understood.
In the present study, we investigated the ruminal metabolic and microbial profile of cows with SARA by using untargeted metabolomics and 16S rRNA sequencing. Based on metabolomics results and correlation analysis, we focused on the role of sialic acids (SAs, N-glycolylneuraminic acid (Neu5Gc) and N-acetylneuraminic acid (Neu5Ac)) in mastitis pathogenesis and the potential mechanism by using antibiotic-treated mouse model. Our results demonstrated that SA exacerbates gut dysbiosis-induced mastitis by fueling microbiota disruption through the microbiota-gut-mammary axis, which provides insight into the pathogenesis of mastitis and acts as the basis for finding potential strategies for mastitis intervention.
Results
Establishment of a cow SARA-associated mastitis model and experimental design
To investigate the potential metabolites involved in the pathogenesis of SARA-mediated mastitis, untargeted metabolomics profiling was performed on 12 ruminal samples from SARA (n = 6) and healthy (n = 6) cows using high-performance LC–MS (Fig. 1B). Integrative analysis was performed based on altered metabolites in the rumen and correction with mastitis indices (Fig. 1C). The biological function of the candidate metabolite was validated in specific-pathogen free (SPF) mice and antibiotic-treated mice (Fig. 1D). Moreover, the underlying mechanism of potential metabolite on mastitis pathogenesis was investigated by FMT, precision editing of gut microbiota, probiotic treatment and in vitro experiments (Fig. 1E). Finally, we confirmed the potential role of candidate metabolite in mastitis pathogenesis by performing ruminal microbiota transplantation (RMT) from cow to mouse and treating mice with targeted inhibitor simultaneously (Fig. 1F).
Experimental design Cows were treated with a standard or high-grain diet for 8 weeks () and ruminal samples were collected for metabolomic analysis ().Untargeted metabolomics was performed by LC–MS and metabolites that were significantly different between healthy and SARA cows were identified.The biological functions of sialic acid were validated in mice.The underlying mechanism of potential metabolites on mastitis pathogenesis was investigated by fecal microbiota transplantation, precision editing of gut microbiota, in vitro experiments and probiotic treatment.Ruminal microbiota transplantation and pharmacological sialidase inhibition were performed to confirm the role of sialic acid. LC–MS, liquid chromatography-mass spectrometry; Neu5Ac, N-acetylneuraminic acid; SARA, subacute ruminal acidosis; SCC, somatic cell count; SPF, specific pathogen free; SRMT, SARA ruminal microbiota transplantation A B C D E F
SARA cows have altered metabolite levels in the rumen
We identified 83 significantly differential metabolites in SARA cows compared with healthy cows (49 upregulated and 34 downregulated, Table S1, S2, and S3) by considering variable importance in the projection (VIP), fold change (FC) and P value (VIP > 1, FC ≥ 2, or FC ≤ 0.5 and P value < 0.05). Moreover, the metabolite levels in SARA cow rumens were significantly different from those in healthy cows through differential metabolite cluster analysis by using a heatmap (Fig. 2B). KEGG pathway enrichment analysis indicated that pathways enriched in SARA cows were mainly protein digestion and absorption, serotonergic synapse, pentose phosphate pathway and arachidonic acid metabolism (Figure S4C). The global distribution of different metabolites was shown by volcano plots, which indicated distinct metabolite levels between SARA and healthy ruminal samples (Fig. 2C). We further performed correlation analysis between significant metabolites and clinical inflammatory markers. The results showed that 22 metabolites were positively correlated with inflammatory markers, while 8 metabolites were negatively correlated with inflammatory markers (Figure S5). A total of 6 metabolites were upregulated in SARA cows with more than 16-fold FC (Fig. 2C), and 5 of them were positively correlated with inflammatory markers (Figure S5), including tretinoin, formononetin, Neu5Gc, 8Z,11Z,14Z-eicosatrienoic acid, and Neu5Ac. Among these, Neu5Gc and Neu5Ac had higher P values by a -log10 p value of more than 8 (Fig. 2C, Table S2). Collectively, these results indicated that SARA cows have altered metabolite levels, some of which were correlated with SARA-associated mastitis parameters.
Metabolic profiles of health and SARA cows.PCA score plots for ruminal samples. Theaxis andaxis represent the percentages of contribution the first two principal components (PC1 and PC2) (= 6).Hierarchical cluster analysis of ruminal metabolites that were different between health and SARA cows (= 6).Volcano plots indicating the results of pairwise comparisons of metabolites in health and SARA cows (= 6). The vertical and horizontal dashed lines indicate the threshold for the twofold abundance difference and the= 0.05 threshold, respectively. Student’stest was performed for the comparison. Significant metabolites are presented in red (upregulated) or green (downregulated) (VIP > 1,value < 0.05 and FC ≥ 2 or FC ≤ 0.5). The relative expression levels of Neu5Gc () and Neu5Ac () in health and SARA samples were determined (= 6). Each dot represents an individual and Student’stest was performed, followed by Tukey test. ***< 0.001. Neu5Gc, N-Glycolylneuraminic acid; VIP, variable importance in the projection A B C D E X Y n n n p t P n t p
Consumption of SA exacerbates gut dysbiosis-induced mastitis in mice
Neu5Gc and Neu5Ac are nine-carbon backbone monosaccharides that are commonly known as SAs, and are abundantly displayed on all mucosal surfaces [32]. Neu5Gc is the modified form of Neu5Ac produced by cytidine monophosphate N-acetylneuraminic acid hydroxylase [32]. Levels of SA are very low in the gastrointestinal tract under normal conditions, while adverse factors, such as gut dysbiosis or mucosal inflammation, can increase free SA levels by increasing the expression of bacterial sialidases or host sialyltransferase, respectively [33, 34]. We showed a 39.7-fold FC and 16.1-fold FC increase in Neu5Gc and Neu5Ac in SARA cows respectively (Fig. 2D, E, Table S2). Different metabolite correlation analyses also found that Neu5Gc and Neu5Ac were positively correlated with tretinoin (Figure S6), the metabolite with the greatest increase in SARA cows according to the FC that has been reported at inflammation sites and occurs concurrently with the development of inflammation [35]. We next focused on the potential effects of SA on SARA-associated mastitis pathogenesis.
Sialic acid exacerbates gut dysbiosis-induced mastitis in mice.Experimental protocol of the sialic acid treatment. Mice were treated with free water, Neu5Ac, ampicillin or ampicillin + Neu5Ac for the indicated times. The mammary tissues were harvested 14 days after parturition.Histological analysis of mammary tissues using H&E-stained sections (scale bar, 20 μm). Red arrows indicate leukocyte infiltration.Histological scores of mammary gland in different groups based on H&E-stained sections (= 7–8). Inflammatory parameters of mammary glands from different groups, including TNF-α () and IL-1β () concentrations and MPO activity (), were measured (= 7–8). Representative protein levels of ZO-1, Occludin and Claudin-3 () and intensity analysis of ZO-1, Occludin, and Claudin-3 from different treated mice () were determined (= 4). Data are expressed as the mean ± SD (and). One-way ANOVA was performed, followed by Tukey test (and). *< 0.05, **< 0.01, ***< 0.001 indicate significant difference. A, ampicillin; C, control; N, Neu5Ac; AN, ampicillin + Neu5Ac A B C D E F G H-J C–F H–J C–F H–J n n n p p p
SA promotes mucosal and systemic immune imbalance in antibiotic-treated mice
Sialic acid facilitates gut dysbiosis-induced mucosal and systemic immune imbalance in mice.Representative H&E-stained colon sections from the indicated mice (scale bar, 50 μm).Histological score of colons from different treatment groups (= 78).Representative H&E-stained liver sections from different groups (scale bar, 50 μm).Histological scores of the livers based on H&E staining (= 78).Colon mRNA levels from the indicated mice as assessed by qPCR (= 78).Fecal lipocalin-2 levels were assessed to validate mucosal inflammatory responses (= 78). Serum ALT () and AST () levels were determined to assess liver injury (= 78). Levels of serum TNF-α (), IL-1β (), and LPS () showed the systemic inflammatory responses caused by gut dysbiosis and sialic acid treatment (= 78). Data are expressed as the mean ± SD (,, and) and one-way ANOVA was performed, followed by Tukey test (and). *< 0.05, **< 0.01, ***< 0.001 indicate significant difference A B – C D – E – F – G H – I J K – B D F–K B, D, F–K n n n n n n p p p
SA aggravates gut barrier disruption in antibiotic-treated mice
Sialic acid promotes gut dysbiosis-induced intestinal barrier disruption. The goblet cells of colons from different treated mice are shown by AB-PAS staining (scale bar, 50 μm) (,) (= 78). Immunohistochemistry staining was performed to assess the mucin-2 level in the colon using mucin-2 antibody (scale bar, 50 μm). The positive-stained cells are shown in brown (,) (= 78).Relative mRNA levels of colon tight junction proteins, including ZO-1, Occludin, and Claudin-3, from the indicated mice were assessed by qPCR (= 78). Protein levels of colon ZO-1 (), Occludin (), and Claudin-3 () were measured by western blotting (= 4). Data are expressed as the mean ± SD (–) and one-way ANOVA was performed, followed by Tukey test (–). *< 0.05, **< 0.01, ***< 0.001 indicate significance. AB-PAS, Alcian Blue Periodic acid Schiff; MUC2, mucin-2 A C – B D – E – F G H C H C H n n n n p p p
SA potentiates inflammation through the TLR4-NF-κB/NLRP3 signatures
Sialic acid potentiates the activation of the TLR4-NF-κB/NLRP3 pathways caused by antibiotic in the mammary gland and colon. The protein levels of TLR4-NF-κB/NLRP3 pathways in mammary () and colonic () tissues from the indicated groups were determined by western blotting. The relative intensities of TLR4, p-p65, p-IκB, NLRP3, ASC, and IL-1β in the mammary gland (–) and colon (–) were determined (= 4). Data are expressed as the mean ± SD (–and–) and one-way ANOVA was performed, followed by Tukey test (–and–). *< 0.05, **< 0.01, ***< 0.001 indicate significance A H B G I N B G I N B G I N n p p p
SA facilitates antibiotic-induced gut dysbiosis
Sialic acid facilitates antibiotic-induced gut dysbiosis in mice. Mice were treated as described above and the fecal microbiota was assessed using 16S rDNA sequencing on day 14.Shannon index showed that the N, A, and AN groups had a reduced alpha diversity (= 7–8).Venn analysis indicated the common and different OTUs in the different treatment groups (= 7–8).PCoA score plots for mouse fecal samples indicating the distinct intestinal microbiota structure (= 0.5357,= 0.001) from the different treated groups based on unweighted UniFrac distance (= 7–8).Bacterial composition at the phylum level from indicated groups (= 7–8).The relative abundances of,, andin the different groups (= 7–8). Data are expressed as the mean ± SD and one-way ANOVA was performed, followed by Tukey test. *< 0.05, ***< 0.001 indicate significance. ns, no significance.Bacterial composition in genus level from indicated groups (= 7–8).LEfSe was performed to indicate the different bacterial taxa enriched in different treated groups (logLDA score > 3).Spearman correlation between gut microbiota and inflammatory parameters from the different groups. The red color denotes a positive correlation, while the blue color denotes a negative correlation. The intensity of the color is proportional to the strength of the Spearman correlation. *< 0.05, **< 0.01, ***< 0.001 indicate significance. LEfSe, Linear discriminant analysis effect size A B C D E F G H n n R P n n Firmicutes Proteobacteria Bacteroidetes n p p n p p p 10
FMT from the A or AN group mimics mastitis in recipient mice
FMT from AN group mice induces mastitis in recipient mice.Representative images of H&E-stained sections from different FMT groups (scale bar, 50 μm). Arrows indicate inflammatory infiltration.The histological scores of mammary glands from the different FMT groups (= 6). The proinflammatory markers TNF-α () and IL-1β () and MPO activity () were measured (= 6).Representative images of ZO-1, Occludin, and Claudin-3 in mammary glands from the indicated mice. The intensities of ZO-1, Occludin, and Claudin-3 were determined (–) (= 4).The protein levels of the TLR4-NF-κB/NLRP3 pathways in the mammary gland were assessed by western blotting. The relative intensities of TLR4, p-p65, p-IκB, NLRP3, ASC, and IL-1β in the mammary gland were determined (–) (= 4). Serum LPS (), TNF-α (), IL-1β (), and fecal lipocalin-2 () were determined (= 6). Data are expressed as the mean ± SD (–,–,–and–) and one-way ANOVA was performed, followed by Tukey test (–,–,–and–). *< 0.05, **< 0.01, ***< 0.001 indicate significant difference A B C D E F G I J K P Q R S T B E G I K P, Q T B E G I K P, Q T n n n n n p p p
SA boosts pathogenic bacterial growth and virulence gene expression, aggravating pathogen-induced proinflammatory cytokine production and mastitis
To confirm the role of Enterobacteriaceae in SA-mediated mastitis, sodium tungstate, a specific inhibitor of the expansion of Enterobacteriaceae during episodes of inflammation [46, 47], was performed in AN-treated mice. We first confirmed that sodium tungstate treatment (AN + T) reduced intestinal Enterobacteriaceae but had little influence on other genera (Figure S11A-B). Tungstate treatment ameliorated mastitis compared with that in the AN group, as evidenced by improved mammary injury and reduced levels of proinflammatory parameters (Fig. 9G–K). In addition, the AN + T mice had a restored blood-milk barrier integrity with increased TJ proteins compared with that of the AN group (Fig. 9L). Moreover, sodium tungstate treatment also alleviated AN-triggered mucosal and systemic inflammation in mice (Fig. 9M, N). Altogether, these results indicated that SA serves as a preferential carbon source for pathogens to promote pathogen expansion and virulence gene expression and subsequently trigger a more severe inflammatory response.
Sialic acid facilitates mastitis by promoting the growth and virulence of.A Wilcoxon rank-sum test was performed to identify the differential bacterial taxa in the A and AN groups (FDR < 0.05) (= 8).The effect of different monosaccharides on the growth ofwas determined in M9 medium (= 3).,.was cultured with 2.5, 5, or 10 mM Neu5Ac or glucose for 6 h, and the relative mRNA levels of() and() were measured using qPCR (= 3).,.was cultured with 10 mM Neu5Ac or glucose for 6 h, and then macrophages (10cells/mL in 6-well plates) were treated with different concentrations offor 24 h. The proinflammatory cytokines TNF-α () and IL-1β () were measured by ELISA (= 3).Representative images of H&E-stained mammary glands from the indicated groups (scale bar, 50 μm).Histological scores of the mammary gland. TNF-α (), IL-1β (), MPO (), fecal lipocalin-2 (), and serum LPS () levels were determined (= 6).The TJ proteins ZO-1, Occludin, and Claudin-3 were assessed by western blotting (n = 4). Data are expressed as the mean ± SD (–,–, and–). Statistical significance in the same time was determined by Student’stest (). For–, two-way ANOVA was performed for statistical analysis (–). For–and–, one-way ANOVA was performed, followed by Tukey test. *< 0.05, **< 0.01, ***< 0.001 indicate significance. ns, no significance. FDR, false discovery rate; T, sodium tungstate Enterobacteriaceae n E. coli n E. coli ler tir n E. coli E. coli n n t p p p A B C D C D E F E F G H I J K M N L C F H K M N B C F C F H K M N 6
Commensal L. reuteri alleviates SA-aggravated mastitis in mice
ameliorates sialic acid-facilitated mastitis in antibiotic-treated mice.Schematic representation ofsupplementation. Mice were treated with ampicillin and Neu5Ac for 14 days before parturition, followed by the removal of ampicillin and treatment with(10CFU/mouse per day) for the next 14 days.Representative H&E-stained sections of mammary glands from different groups (scale bar, 50 μm).Histological scores of the mammary sections (= 6).–. Mammary MPO (), TNF-α (), IL-1β (), serum LPS (), and fecal lipocalin-2 () levels were assessed (= 6).. Representative images of ZO-1, Occludin, and Claudin-3.–. The relative intensities of mammary ZO-1 (), Occludin (), and Claudin-3 () were determined (= 4).The protein levels of TLR4-NF-κB/NLRP3 pathways were determined by western blotting.–Relative intensities of mammary TLR4 (), p-p65 (), NLRP3 (), ASC (), and IL-1β () were determined (= 4). Data are expressed as the mean ± SD (–,–, and–) and one-way ANOVA was performed, followed by Tukey test (–,–, and–). *< 0.05, **< 0.01, ***< 0.001 indicate significance L. reuteri L. reuteri L. reuteri n n n n p p p A B C D H D E F G H I J L J K L M N R N O P Q R C H J L N R C H J L N R 9
SARA cows have different ruminal microbial compositions and sialidase inhibitor alleviates mastitis in mice caused by RMT for SARA cows (SRMT)
Sialidase inhibitor alleviates SRMT-induced mastitis and systemic inflammatory responses in mice.The ruminal sialidase levels from indicated groups (= 6).. PCoA showed distinct ruminal microbial structure (= 0.1833,= 0.009) between health and SARA groups based on unweighted UniFrac distance (= 6).Microbial compositions at the phylum level from indicated groups (= 6).LEfSe analysis identified the different taxa in different treated groups (logLDA score > 3) (= 6).RMT experimental protocol. Mice were treated with antibiotics (200 mg/kg ampicillin, metronidazole and neomycin and 100 mg/kg vancomycin) daily through oral gavage for 5 days. Furthermore, mice were orally gavaged with 300 μL of ruminal supernatant for three consecutive days after antibiotics removal and once every 2 days thereafter for 19 days. The control mice were orally gavaged with an equal volume of sterile PBS. SRMT + Z group was treated with zanamivir (0.5 mg/mL in drinking water) simultaneously for 21 days (= 7).PCoA showed successful ruminal microbial colonization after RMT based on unweighted UniFrac distance (= 0.6720,= 0.001,= 7).PCoA showed distinct ruminal microbial structure between different RMT groups based on unweighted UniFrac distance (= 0.6085,= 0.001,= 7).Microbial compositions at the phylum level in different treated groups (= 7).. LEfSe analysis identified the different taxa in different RMT groups (logLDA score > 3.5) (= 7).,The intestinal sialidase and Neu5Ac levels from different treated groups (= 7).Representative H&E-stained sections of mammary glands from the indicated mice (scale bar, 50 μm).Histological scores of mammary tissues from the different treatment groups (= 7). Inflammatory markers of mammary TNF-α (), IL-1β () and MPO activity (), and fecal lipocalin-2 () and serum LPS (), and were assessed by ELISA (= 7). Data are expressed as boxplot () or the mean ± SD (–and–). A Student’stest () or one-way ANOVA was performed, followed by Tukey test (–and–). *< 0.05, **< 0.01, ***< 0.001 indicate significance. Z, zanamivir A B C D E F G H I J K L M N O P Q R A J K M R A J K M R n R P n n n n R P n R P n n n n n n t p p p 10 10
Discussions
Milk is essential for people’s normal lives because it provides abundant nutrients and trace elements, which has led to milk yield and quality serving as markers of national agricultural modernization [51]. Mastitis is the most severe disease that decreases milk yield and quality, which results in huge economic losses every year throughout the world [52]. Although pathogen invasion of the mammary gland, especially by bacteria including Gram-positive Streptococcus and S. aureus and Gram-negative E. coli and Klebsiella, has been thought to be the major cause of mastitis [53, 54], an increasing number of studies have suggested that gut microbiota play a crucial role in the pathogenesis of mastitis. Our previous results suggested that gut dysbiosis caused by antibiotics can initiate mastitis and promote its development upon pathogen invasion in mice [21, 22]. Moreover, cows with mastitis showed distinct gut microbiota profiles, and FMT from cow to mouse induced mastitis in mice [19, 20]. In dairy cows, SARA induced by the excessive consumption of a HGD increased the risk for mastitis [27, 55], but little is known about how SARA-mediated metabolic changes contribute to mastitis pathogenesis and its influence on microbiota in turn. In the present study, we revealed that SARA induced marked metabolic changes in the rumen. Among the most enriched metabolites in SARA cows was SA, which was correlated with mastitis parameters. Treatment with SA exacerbates mastitis in antibiotic-treated mice through the microbiota-gut-mammary axis by facilitating the growth of opportunistic pathogen and activating TLR4-NF-κB/NLRP3 signatures, while limitation of SA production, regulation of gut homeostasis by commensal microbes and targeted inhibition of opportunistic pathogen expansion attenuated gut dysbiosis-associated mastitis in mice.
Evidence supporting a causal role for host or microbiota-associated metabolites in the pathogenesis of distal diseases has been reported. For example, microbial 4-EPS causes anxiety behavior by impairing oligodendrocyte maturation and reducing oligodendrocyte-neuron interactions [14]. Trimethylamine derived from the gut microbiota was found to be elevated in alcohol-associated hepatitis and contributes to ethanol-induced liver injury in mice [56]. Moreover, gut microbiota produced delta-valerobetaine facilitates Western diet-induced obesity by depressing mitochondrial fatty acid oxidation [57]. Gao et al. showed that SARA cows demonstrate alternations in plasma metabolites, including β-hydroxybutyrate, non-esterified fatty acids, cholesterol and triglyceride, which are associated with a change in oxidative stress levels [58]. Another study found that a grain-based SARA increases serum LPS levels and triggers systemic inflammation [59]. Although SARA has been reported to shape the systemic metabolic profile and is associated with the development of mastitis, whether a ruminal dysbiosis-derived metabolite, such as SA, can lead to mastitis has not been investigated.
SAs exist widely on the surface of vertebrate cells, especially on mucosal surfaces [32], where SAs directly regulate the interaction between microbes and the host. Hence, the elevated concentration of SAs in SARA cows potentially does not originate directly from foods and agrees with previous findings that there is a low free SA level in the gut under physiological conditions [34]; however, increased SA levels were observed in intestinal inflammation and gut dysbiosis by increasing levels of sialidase [33, 34]. Indeed, increased gastrointestinal inflammation by enriched LPS levels and distinct microbiota compositions have been reported in SARA cows [27, 55]. In the current study, we confirmed that the increased SAs contributed to gut dysbiosis-associated mastitis, as evidenced by SA supplementation in conventional mice showing no obvious effect on mastitis in mice. In antibiotic-treated mice, however, SA consumption aggravated antibiotic-induced mammary inflammatory responses, indicating that SA facilitated the development of mastitis and that this effect depended on the gut microbiota. Previous studies have shown that SA can trigger an inflammatory response when it encounters circulating anti-SA antibodies [60]. Moreover, increased serum Neu5Ac levels were associated with myocardial injury [61]. The gap between our findings and previous studies may be attributed to distinct experimental conditions or animal models. In addition, SA-antibiotic-treated mice had increased mucosal and systemic inflammation, accompanied by damaged gut barrier functions and increased serum LPS levels, which has been associated with remote diseases through TLR4 [41]. Increased TLR4 levels were detected in the colon and mammary gland of AN mice, accompanied by enhanced downstream NF-κB and NLRP3 pathways, which are involved in mastitis pathogenesis and other gut dysbiosis-associated diseases [22, 42, 62, 63]. These results suggest that SA facilitates the production of gut dysbiosis-derived LPS, enabling the development of mastitis through the microbiota-gut-mammary axis by TLR4-mediated signatures [63].
FMT has been widely used to verify the causative relationship between gut microbiota changes and disease outcomes [19, 21, 64]. We showed that FMT from the A or AN group induced mastitis in recipient mice, suggesting that gut dysbiosis is attributed to the development of mastitis [19, 21, 22]. Our results agree with previous findings that microbiota-liberated host SA facilitates enteric pathogen invasion and promotes host gene mutant-induced opportunistic pathogen expansion [34, 65, 66]. Similar findings have also been reported in other microbiota-liberated host sugars, including succinate and fucose [45, 67], indicating that changed carbon metabolism may confer an expansion advantage for pathogens. It has been reported that intestinal inflammation can induce gut dysbiosis and promote the overgrowth of Enterobacteriaceae [68], which was also observed in dysbiosis-induced mastitis mice treated with antibiotics [21]. Our findings can be extended to a previous study of the inhibition of SA release by the knockout of the α2,3 sialyltransferase St3gal4, which reduced Enterobacteriaceae abundance and alleviated colitis in DSS-treated mice [69]. In vitro experiments showed that SA served as a preferential energy source for pathogens, which further promoted bacterial growth, enhanced the virulence and triggered inflammation. E. coli contains a SA transporter (nanT) and aldolase/lyase (nanA), which transports free SA and subsequently catabolizes SA into pyruvate and N-acetylmannosamine respectively [70]. Similar results have been reported that host-derived sugars promote pathogen growth and virulence [45, 71]. Inhibition of Enterobacteriaceae by sodium tungstate improved mastitis, which indicates that target depletion of opportunistic pathogen during gut dysbiosis is a potential strategy to alleviate gut and parenteral disease [47, 72]. Interestingly, although the N group also had a changed microbiota composition including enriched Lactobacillus, Faecalibaculum, Enterorhabdus, Bacteroides, Blautia and Roseburia, no inflammatory changes were detected, which implies the beneficial role of commensal microbiota on the metabolism of SA [34]. Among these, Lactobacillus has been identified as an effective SA-utilizing bacterium [73]. Other microbes enriched in SA-treated conventional mice including Faecalibaculum, Bacteroides, Blautia and Roseburia, are well known as SCFA producers [74–77], which suggest that SA may act as a nutrient to promote commensal microbes to produce SCFA, but further verification is needed. Interestingly, mice supplemented with L. reuteri, a probiotic that can improve E. coli-induced mastitis [22], alleviated SA-facilitated mastitis in antibiotic-treated mice. However, whether the beneficial role of L. reuteri is attributed to the effective capacity of SA metabolism or its immune regulation ability remains unknown. Together, these results suggest that different gut microbial contents may confer distinct results on host SA metabolism and highlight the importance of maintaining gut homeostasis for the prevention of mastitis.
SAs are rarely synthesized by microbes and are commonly derived from host glycans by sialidase to become freely available for commensal or pathogenic bacteria [70, 78]. In this study, we showed that SARA cows had increased ruminal microbial sialidase, accompanied by a distinct ruminal microbiota structure. A previous study showed that SARA cows had increased Ruminococcus levels [48], which was reported to release SA from sialyllactose and promote bacterial adaptation to the intestine [79]. Commensal Bacteroides encode sialidase but lack the catabolic pathway [34, 70]. Thus, this genus commonly facilitated other pathogens expansion in the gut, including Salmonella typhimurium, Clostridioides difficile and E. coli [34, 69]. In the present study, ruminal SA-utilizing commensal Prevotellaceae is depleted [73], while the SA-utilizing opportunistic pathogenic Moraxellaceae is enriched in SARA cows [50], which supports that abnormal SA metabolism facilitates microbial dysbiosis. Notably, Moraxellaceae also enriched in the rumen of clinical and subclinical mastitis cows [49, 80], implying its pathogenic effect in mastitis pathogenesis but further research is needed. Consistent with our results that zanamivir treatment reduced Moraxellaceae abundance and SRMT-mediated mastitis, Ahmed et al. found that inhibition of sialidase can limit Moraxellaceae-induced inflammation [81]. However, the origin of microbial sialidase and how it is elevated are unclear in the present study. We confirmed that inhibition of SA release alleviated SARA-induced mastitis by zanamivir, which could be extended to previous studies showing that inhibition of sialidase is a potential strategy to improve gut dysbiosis and subsequent inflammatory responses [66, 69]. Another potential reason for the increased SA is mucosal inflammation, in which LPS triggered host neuraminidase release through TLR4 and treatment with sialidase inhibitor attenuated inflammation caused by LPS or pathogen invasion [33, 82]. Although the specific mechanism for the pathogenesis of SARA-induced mastitis has not been proven, ruminal microbiota disruption and inflammation may act synergistically. Hence, the protective effects of sialidase inhibition may work in a variety of ways, and sialidase-deficient mice will be used in the future to determine the role of host sialidase in mastitis pathogenesis and gut microbiota. Our findings did not allow us to rule out the pathological roles of other metabolites. Likewise, it is still unclear which certain species of Enterobacteriaceae are the key taxa responsible for mastitis pathogenesis. Moreover, further studies are needed to gain greater insight into other potentially harmful metabolites on the pathogenesis of mastitis and the protective effects of metabolites depleted in SARA cows.
Schematic illustration of the mechanism by which SA facilitates gut dysbiosis-induced mastitis through the microbiota-gut-mammary axis.Under gut homeostasis conditions, sialic acids bind to mucus, with few free SAs present in the intestinal lumen, which are then degraded by polysaccharide-degrading commensal bacteria (e.g.,and).Gut dysbiosis caused by a changed dietary pattern (e.g., high-grain diet-induced SARA) or antibiotic reduces polysaccharide-degrading commensal bacteria and prompts free SAs production by increasing sialidase, causing abundant free SAs in the intestinal lumen which fuels polysaccharide-degrading opportunistic pathogen (e.g.,and).Polysaccharide-degrading opportunistic pathogen, such as, proliferate by using SAs as an energy source, which facilitates their expansion and virulence enhancement, prompting the degradation of the mucus layer and gut barrier damage.A leaky gut barrier allows toxic factors (e.g., LPS) to enter intestinal tissues and induces immune cell activation through TLR4-NF-κB/NLRP3 signatures, causing intestinal inflammation.The increased LPS derived from expansion of gut opportunistic pathogen subsequently enters the blood and, together with proinflammatory cytokines, induces systemic inflammation.Persistent chronic systemic low-grade inflammation disrupts the blood-milk barrier, causing cytokines and LPS to enter and accumulate in the mammary gland, inducing mastitis through TLR4-NF-κB/NLRP3 pathways ① ② ③ ④ ⑤ ⑥ Lactobacillaceae Prevotellaceae Enterobacteriaceae Moraxellaceae Enterobacteriaceae
Conclusions
The current study revealed marked metabolic changes in SARA cows and identified that increased SA metabolism facilitated the development of gut dysbiosis-induced mastitis in mice. The underlying mechanism is involved in the overexpansion of Enterobacteriaceae and subsequent gut barrier damage and systemic inflammation by TLR4-NF-κB/NLRP3 signatures (Fig. 12). In contrast, depletion of Enterobacteriaceae, treatment with commensal bacteria and reduction of free SAs improved gut dysbiosis-associated mastitis in mice. Our findings constitute the first demonstration, to our knowledge, that SARA-derived metabolic changes exacerbate systemic inflammation and mastitis by facilitating gut opportunistic pathogen expansion through the microbiota-gut-mammary axis. Likewise, in addition to highlighting the adverse impact of host-derived sugar on mastitis in the context of gut dysbiosis, the findings of this study also indicate that inhibition of SA release and maintenance of gut homeostasis might be an alternative strategy to control mastitis by blocking the gut-mammary axis in the context of gut dysbiosis.
Material and methods
Cows treatment and samples collection
Twelve Holstein cows (4–6 years, averaging ~ 600 kg of weight) were obtained from a farm in Qingzhou, Shangdong Province, China, and none of the animals had diseases or treated with antibiotics and drugs within 6 months. Cows separated into health groups and SARA group randomly. All cows meet the daily nutrient requirements for lactating. The health group was treated with a standard diet of grass-legume hay and SARA group was treated with a HGD (70% grain and 30 forage diet) as previous described [27, 83]. After 8 weeks treatment, the rumen fluid from different treated cows were harvested according to previous study [27]. Briefly, each ruminal sample was harvested by rumenocentesis using a nonpyrogenic needles (2.0 × 120 mm) and 50 mL syringes. All samples were stored at liquid nitrogen until metabolomics analysis. Rumen pH was used to diagnosis SARS as previously [29].
Untargeted metabolomics
Rumen fluid (100 mL) was mixed with prechilled methanol (400 μL) by well vortexing and then incubated on ice for 5 min, followed by centrifuging at 15,000 rpm, 4 °C for 5 min. LC–MS grade water was used to dilute the supernatant to a final concentration containing 53% methanol. After transferring to a fresh Eppendorf tube, the samples were centrifuged at 15,000 × g, 4 °C for 10 min and injected into the LC–MS/MS system analysis.
LC–MS/MS analyses were performed using a Vanquish UHPLC system (Thermo Fisher) coupled with an Orbitrap Q Exactive series mass spectrometer (Thermo Fisher). Samples were injected onto a Hyperil Gold column (100 × 2.1 mm, 1.9 μm) using a 16-min linear gradient at a flow rate of 0.2 mL/min. The raw data files generated by UHPLC-MS/MS were processed using the Compound Discoverer 3.1 (CD3.1, Thermo Fisher) to perform peak alignment, peak picking, and quantitation for each metabolite. The main parameters were set as follows: retention time tolerance, 0.2 min; actual mass tolerance, 5 ppm; signal intensity tolerance, 30%; signal/noise ratio, 3; and minimum intensity, 100,000. After that, peak intensities were normalized to the total spectral intensity. The normalized data was used to predict the molecular formula based on additive ions, molecular ion peaks and fragment ions. And then, peaks were matched with the mzCloud (https://www.mzcloud.org/↗), mzVault and MassList database to obtain the accurate qualitative and relative quantitative results. Statistical analyses were performed using the statistical software R (R version R-3.4.3), Python (Python 2.7.6 version), and CentOS (CentOS release 6.6), When data were not normally distributed, normal transformations were attempted using of area normalization method.
These metabolites were annotated using the KEGG database (http://www.genome.jp/kegg/↗), HMDB database (http://www.hmdb.ca/↗) and Lipidmaps database (http://www.lipidmaps.org/↗). PCA and PLS‐DA were performed at metaX. Univariate analysis (t test) was applied to calculate the statistical significance (P value). The metabolites with VIP > 1 and P value < 0.05 and FC ≥ 2 or FC ≤ 0.5 were considered to be differential metabolites. Volcano plots were used to filter metabolites of interest which based on Log2 (FC) and -log10 (P value) of metabolites. For clustering heat maps, the data were normalized using z-scores of the intensity areas of differential metabolites and were ploted by Pheatmap package in R language. The functions of these metabolites and metabolic pathways were studied using the KEGG database. The metabolic pathway enrichment of differential metabolites were performed, when ratio were satisfied by x/n > y/N, metabolic pathway were considered as enrichment, when P value of metabolic pathway < 0.05, metabolic pathway were considered as statistically significant enrichment.
Mouse husbandry condition and treatments
All SPF grade BABL/c mice (22–24 g) used in the present study were obtained from Liaoning Changsheng Biotecnology Co., Ltd. (Benxi, China). Mice were raised in SPF grade feeding conditions with 12-h light daily for a week. These mice were then mixed at a ratio of three females to one male in separated cages and the male mice were removed after confirming pregnancy by vaginal spermatozoa.
For the SA supplementation experiment, 14 days of pregnancy mice were separated into four groups randomly, referred to as (1) Control (C) group: feeding with standard diet (AIN93G) and free water and treated with 200 μL PBS orally twice a day for 21 days; (2) Neu5Ac (N) group: treating with 1% Neu5Ac (Sigma Aldrich, MO, USA) in the water and orally gavaged 1 mg/200 μL of Neu5Ac twice a day for 21 days [34]; (3) Ampicillin (A) group: mice were treated with 1 g/L ampicillin (Sigma Aldrich, MO, USA) for 21 days modified according to previous study [36] and orally gavaged with 200 μL PBS twice a day; (4) Ampicillin + Neu5Ac (AN) group: mice were treated with antibiotics for 21 days accompanied with 1% Neu5Ac treatment and orally gavaged 1 mg/200 μL of Neu5Ac twice a day as described above. At the day 14 after delivery, mice were sacrificed and feces, mammary gland, ileum, colon and liver were harvested and stored at − 80 ℃ until detection.
For sodium tungstate treatment experiment, 0.8% sodium tungstate (RHAWN, Shanghai, China) was added in drinking water along with ampicillin and Neu5Ac treatments [47, 72]. For L. reuteri supplementation experiment, mice were treated ampicillin and Neu5Ac for 14 days before parturition, and then supplemented with L. reuteri CNCM I-5022 (109 CFU) daily for 14 days through oral gavage [22].
For FMT or RMT experiments, feces or ruminal samples from different treated groups were harvested and prepared as previously described [19, 20, 46, 48, 49]. Briefly, fresh feces from each group were pooled and homogenized using sterile saline (50 mg feces/mL) [64]. In addition, a total of 6 ruminal samples (0.5 g/sample) were mixed and dispensed according to the experimental demand. These samples were centrifuged at 100 × g, 4 ℃ for 2 min, and then the supernatant were collected. For the transplantation, the mice (pregnancy for a week) were treated with antibiotic (200 mg/kg ampicillin, metronidazole (Sigma-Aldrich, MO, USA) and neomycin (Sigma-Aldrich, MO, USA) and 100 mg/kg vancomycin (Sigma-Aldrich, MO, USA)) daily through oral gavage for 5 days to eliminate commensal microbiota [46]. Furthermore, the antibiotics were removed for 1 day and mice were colonized with ruminal microbiota. In data, mice were oral gavaged with 300 μL fecal or ruminal supernatant for three consecutive days after antibiotics removal, and once every 2 days thereafter for 19 days with prepared microbiota samples [46]. For RMT supplemented with a HSD, mice were fed with a high-starch (5% cellulose) diet during RMT based on previously described [48]. For the sialidase inhibition experiment, mice were performed RMT as mentioned above and treated with zanamivir (0.5 mg/mL in drinking water, Yuanyebio, China) for 21 days [33].
Bacteria culture and growth assay
To investigate the effect of SA on bacterial growth, E. coli AIEC strain LF82 was incubated in M9 medium (Sigma-Aldrich, MO, USA) supplemented with 10 mM Neu5Ac or glucose and the bacteria intensities were assessed at 600 nm optical density (OD600) [34]. To investigate the influence of SA on the virulent gene expressions of E. coli, E. coli LF82 were incubated in M9 medium with 2.5, 5, or 10 mM Neu5Ac or glucose, and then the virulent genes were determined using qPCR [45]. L. reuteri CNCM I-5022 was grown in in MRS (Haibo, Qingdao, China) as previously described [22].
Cell culture and inflammatory responses assay
Macrophages (RAW264.7) were performed to confirm the effect of SA primed E. coli on inflammatory responses. Macrophages were cultured in 1640 medium (Sigma-Aldrich, MO, USA) suppled with 10% fetal bovine serum (BI, Israel) and 1% ampicillin and streptomycin at 37 ℃ with 5% CO2 For the E. coli stimulation experiment, cells (106 cells/mL) were incubated in 24-well plates and removed the ampicillin and streptomycin (Sigma Aldrich, MO, USA), and then cells were treated with Neu5Ac or glucose-cultured E. coli (106, 107, and 108 CFU) for 24 h. Finally, the supernatants were harvested and proinflammatory cytokines were determined by ELISA.
Histological analysis
All samples used for histological analysis, including mammary glands, ileums, colons, and livers, were fixed with 4% paraformaldehyde for 48 h and embedded in paraffin to prepare 5-μm paraffin sections. All the sections were stained with hematoxylin and eosin (H&E) after dewaxing and hydration. The histological changes of different tissues were performed by blinded pathologists using an optical microscope (Olympus, Tokyo, Japan) and the histological scores were assessed as previously described [22].
Cytokines assay
ELISA assay was performed to determine the predominant proinflammatory cytokines of TNF-α and IL-1β in the serums, mammary glands and cultured cell supernatants. For the mammary samples, 10% tissue homogenates were prepared using PBS and supernatants were collected by centrifuging at 12,000 rpm for 10 min. The concentrations of TNF-α and IL-1β in serums, mammary glands and cell supernatants were assessed and calculated according to the manufacturer’s instructions (Biolegend, CA, USA).
MPO activity, ALT, and AST assay
To determine the MPO levels, 10% tissue homogenates were prepared using MPO buffer and detected by MPO assay kit according to the manufacturer’s instructions (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). To investigate liver injury, the serum ALT and AST were determined using ALT and AST assay kits according to the manufacturer’s instructions (Nanjing Jiancheng Bioengineering Institute, Nanjing, China).
LPS concentrations determination
LPS assay kit was performed to measure the LPS level in the serum or ruminal samples according to the manufacturer’s instructions (Lanpaibio, Shanghai, China). In brief, blood were collected and centrifuged at 3000 rpm for 10 min to separate the serum carefully. Rumen fluids were collected by centrifuge without heat source, precipitation was removed by centrifuging at 3000 rpm for 10 min, and supernatants were harvested for LPS determination.
Lipocalin-2 determination
The concentrations of fecal lipocalin-2 were detected using lipocalin-2 ELISA kit according to manufacturer’s instructions (Lanpaibio, Shanghai, China).
Sialidase activity and sialic acid assays
Sialidase activity in ruminal and intestinal samples and intestinal Neu5Ac levels were assessed by using the Sialidase Activity Assay Kit (MAK121, Sigma-Aldrich, MO, USA) and the Sialic Acid Assay Kit (MAK314, Sigma-Aldrich, MO, USA) following the manufacturer’s protocols, respectively.
Goblet assessment by AB-PAS staining
The goblet cell and intestinal morphological analysis were performed by AB-PAS staining (Solarbio, Beijing, China) as previous described [40]. In brief, paraffin sections were dewaxed washed with water for 2 min, and then stained with AB for 15 min followed by washing three times per 2 min with water. Furthermore, the sections were oxidized using oxidant and washed for twice, and then the sections were stained by Schiff regent for 15 min followed by washing for 10 min. The nuclear was stained by hematoxylin solution for 2 min. Finally, the sections were treated by Scott Bluing Solution for 3 min and washed for 3 min. After dehydrating by series of ethanol and transparent by xylene, the sections were sealed with resinene and determined by blinded pathologists using an optical microscope (Olympus, Tokyo, Japan).
Immunochemistry
Immunochemistry staining was performed to measure the Mucin 2 expression in the colon using SAP (Mouse/Rabbit) IHC Kit (MXB, China) as previous described [40]. In brief, colon paraffin sections were dewaxed by xylene and series of alcohol. The prepared sections were subjected to for antigen retrieval using sodium citrate following phosphate buffer (PBS) wash. Prepared sections were treated with endogenous peroxidase blockers for 40 min at room temperature. After washing by PBS for 3 times per 5 min, the sections were treated with normal nonimmune goat serum for 40 min at room temperature and then incubated with mucin-2 antibody (1:200, diluted with 5% goat serum) overnight at 4 °C. Furthermore, the sections were incubated with the secondary antibody (goat-anti rabbit IgG) for 30 min at room temperature after being washed with PBS. The sections were then incubated with horseradish peroxidase for 20 min at room temperature. After being washed with PBS, the sections were developed for 5 min using a color developing agent under the microscope and terminated by water. The nuclei were staining with hematoxylin for 2 min followed by 1% muriatic acid alcohol differentiation and ammonium hydroxide treatment. Following dehydration, the sections were mounted utilizing neutral resins and determined by blinded pathologists using an optical microscope (Olympus, Tokyo, Japan).
Total bacterial DNA extraction and Illumina MiSeq sequencing
Microbial community genomic DNA was extracted from mouse feces or ruminal samples using the FastDNA® Spin Kit for Soil (MP Biomedicals, USA) according to manufacturer’s instructions. The DNA extract was checked on 1% agarose gel, and DNA concentration and purity were determined with NanoDrop 2000 UV–vis spectrophotometer (Thermo Scientific, Wilmington, USA). The hypervariable region V3-V4 of the bacterial 16S rRNA gene were amplified with primer pairs 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) by an ABI GeneAmp® 9700 PCR thermocycler (ABI, CA, USA). The PCR amplification of 16S rRNA gene was performed as follows: initial denaturation at 95 ℃ for 3 min, followed by 27 cycles of denaturing at 95 ℃ for 30 s, annealing at 55 ℃ for 30 s and extension at 72 ℃ for 45 s, and single extension at 72 ℃ for 10 min, and end at 4 ℃. PCR reactions were performed in triplicate and the PCR product was extracted from 2% agarose gel and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to manufacturer’s instructions and quantified using Quantus™ Fluorometer (Promega, USA). Purified amplicons were pooled in equimolar and paired-end sequenced on an Illumina MiSeq PE300 platform/NovaSeq PE250 platform (Illumina, San Diego, USA) according to the standard protocols by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). OTUs with 97% similarity cutoff were clustered using UPARSE version 7.1, and chimeric sequences were identified and removed. The taxonomy of each OTU representative sequence was analyzed by RDP Classifier version 2.2 against the 16S rRNA database using confidence threshold of 0.7. PCoA based on ANOSIM was used to identify microbial structure and LEfSe was performed to identify bacterial taxa that were differentially enriched in different treatment groups. A Wilcoxon rank-sum test (FDR < 0.05) was performed to identify the differential bacterial taxa between the two groups.
Total RNA extraction and quantitative RT-PCR
The total RNA of colon and mammary tissues were extracted by Trizol (Invitrogen, CA, USA) as our previously described [22]. Briefly, 100 mg tissues were extracted by 1 mL Trizol and subjected to chloroform, isopropanol and 75% ethyl alcohol treatment under RNase-free conditions. After reverse transcription using TransStart Tip Green qPCR SuperMix (TransGen Biotech, Beijing, China), the cDNA was reacted with specific primers using a FastStart Universal SYBR Green Master Mix (ROX) (Roche, Switzerland, Basel) in a Step One Plus apparatus (Applied Biosystems, Foster City, CA, USA). The reaction conditions were performed as previously mentioned [22]. Primers used in this study are presented in Table S4, and GAPDH or rpoA served as an endogenous control for tissues or E. coli, respectively. The 2−ΔΔCt method was performed to calculate the relative expression of genes through calibrating by the control group.
Western blotting
The total protein of colon and mammary gland were harvested using a tissue protein extract (Thermo Fisher Scientific, USA) as previously described [22]. 10% or 15% SDS-PAGE were used to separate proteins and then combined with 0.45 μm PVDF membranes. After being blocked in 5% skim milk, the prepared PVDF membranes were incubated at 4 ℃ overnight with specific primary antibodies, including ZO-1, Occludin, Claudin-3, p-p65, p65, p-IκB, and IκB obtained from Affinity Biosciences (OH, USA), and TLR4, NLRP3, ASC, IL-1β and β-actin obtained from Cell Signaling Technology (CST, Boston, USA). After washing three times with TBST, the PVDF membranes were treated with Goat anti-rabbit or Rabbit anti-mouse IgG (1:20,000, Affinity Biosciences, OH, USA) for 2 h at room temperature. Ultimately, the proteins were determined by the ECL plus western blotting Detection System (Tanon, China).
Statistical analysis
All data were analyzed using GraphPad Prism 8 (San Diego, CA, USA) and expressed as the boxplot or the mean ± SD. Student’s t test was performed for the comparison of two groups. One-way analysis of variance (ANOVA) was performed for the comparison more than two groups, followed by Tukey test to determine the differences among groups. p < 0.05 indicated statistical significance. Other special analysis was stated in legends.
Supplementary Information
Additional file 1: Figure S1. SARA induces mastitis in cows. Cows were treated with a standard or high-grain diet for two months and ruminal, serum, milk and mammary gland tissues were harvested for analysis. A. Ruminal PH at day 60 from Health and SARA cows (n=6). B-C. Ruminal and serum LPS levels were detected using ELISA (n=6). D. Somatic cell count was performed in Health and SARA cows (n=6). E. Representative images of H&E-stained sections from Health and SARA samples. Red arrows indicate leukocyte infiltration. Blue arrows show the structure injury of mammary gland. Black arrows indicate edema. F. Histological score based on H&E-stained sections (n=6). Mammary TNF-α (G) and IL-1β (H) from Health and SARA groups were measured by ELISA (n=6). Each dot represents an individual cow (A-D and F-H) and Student’s t test was performed (A-D and F-H). **p < 0.01, ***p < 0.001 indicate significance. Figure S2. Data quality checks. A. The Pearson correlation of ruminal QC samples. B-C. The PCA score plots for Health and SARA samples containing QC samples (n=6). QC, quality control; PCA, Principal component analysis. Figure S3. Classification and functional annotation of metabolites. A. KEGG pathway annotation for Health and SARA ruminal samples. B. HMDB classification annotation. C. Lipid maps annotation for Health and SARA groups. HMDB, Human Metabolome Database; KEGG, Kyoto Encyclopedia of Genes and Genomes. Figure S4. SARA induced ruminal metabolic changes. A. PLS-DA score plots for ruminal samples (n=6). B. Cross-validation plot with a permutation test repeated 200 times. The intercepts of R2 (0.0, 0.57) and Q2 (0.0,–1.09) indicate that the PLS-DA model was not overfitting. C. Pathway enrichment analysis of significantly elevated metabolites in SARA sample according to the KEGG pathway. Figure S5. Spearman correlation between metabolites and inflammatory parameters. The red color denotes a positive correlation, while green color denotes a negative correlation. The intensity of the color is proportional to the strength of Spearman correlation. *p < 0.05, **p < 0.01, ***p < 0.001 indicate significance. Figure S6. Spearman correlation among metabolites. The top 20 correlated metabolites were showed based on the P value. The red color denotes a positive correlation, while blue color denotes a negative correlation. Figure S7. SA and FMT change the intestinal SA levels. A-B. The intestinal SA levels from different treatment groups (n=6-8). Data are expressed as the mean ± SD (A-B) and one-way ANOVA was performed, followed by Tukey test (A-B). *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant difference. Figure S8. Sialic acid treatment has minimum effects on the synthesis of milk proteins and ileum histology. A-D. Relative mammary gene expressions associated with the synthesis of milk proteins from indicated groups, including Csn1, Csn2, Csn3, and Wap (n=7-8). E. The average weight changes of litters from different treatment groups (n=6). F. Representative images of H&E-stained ileum sections from indicated mice. G. Histological score based on H&E-stained sections (n=7-8). Figure S9. Composition of the gut microbiota in different groups. A. Chao1 index in different groups (n=7-8). B. Ace index (n=7-8). C. Bacterial composition at the family level in the gut were displayed (n=7-8). Each dot represents an individual mouse (A and B) and one-way ANOVA was performed, followed by Tukey test (A and B). ***p < 0.001 indicate significance. ns, no significance. Figure S10. Top 50 metabolism pathways enriched in different groups using Tax4Fun. Figure S11. Sodium tungstate treatment reduces intestinal Enterobacteriaceae abundance. A. The gut microbial compositions at the family level from different treatment groups (n=4). B. A Wilcoxon rank-sum test was performed to identify the differential bacterial taxa in the AN and AN+T groups (FDR < 0.05) (n=4). Figure S12. The ruminal and intestinal microbial compositions in the donor and recipient mice, and the effect of RMT on systemic inflammation in mice. A-D. Alpha diversity indices, including observed species (A), Shannon (B), Chao1 (C) and Simpson index (D), from Health and SARA groups (n=6). E. The ruminal microbial compositions at the family level from the indicated groups (n=6). F. Venn diagram showed the observed OTUs in different RMT groups. G-I. Alpha diversity indices, including Shannon (G), Chao1 (H) and Simpson index (I), from different treatment groups (n=7). J. The gut microbial compositions at the family level from different RMT groups (n=7). K-L. Serum TNF-α (K) and IL-1β (L) levels by ELISA (n=7). Data are expressed as boxplot (A-D and G-I) or the mean ± SD (K-L). A Student’s t test (A-D) or one-way ANOVA was performed, followed by Tukey test (G-I and K-L). *p < 0.05, **p < 0.01, ***p < 0.001 indicate significant difference. Figure S13. A high-starch diet promotes SA production and mastitis in SRMT mice. A. Neu5Ac levels from different treatment groups (n=5). B. Representative images of H&E-stained mammary sections. C. Histological scores based on H&E-stained mammary sections (n=5). D-F. Mammary TNF-α (D), IL-1β (E), MPO activity (F) were assessed (n=5). Data are expressed as the mean ± SD (C-F) and one-way ANOVA was performed, followed by Tukey test (C-F). **p < 0.01, ***p < 0.001 indicate significant difference. Table S1. Identified number of differential metabolites in Health and SARA samples. Table S2. Metabolites significantly upregulated in SARA cows and ranked according to the P-value. Table S3. Metabolites significantly downregulated in SARA cows and ranked according to the P-value. Table S4. The oligonucleotides used in this study.