Abstract

   It has been found that Streptococcus thermophilus (S. thermophilus)
   influenced the gut microbiota and host metabolism with strain
   specificity in C57BL/6J mice in the previous study, though it remains
   unclear whether lactose as a dietary factor associated with dairy
   consumption is involved as the mediator in the interaction. In the
   present study, integrated analysis of 16S rRNA gene sequencing and
   untargeted metabolomics by liquid chromatography–mass spectrometry of
   fecal samples in C57BL/6J mice was applied to evaluate the effect of
   lactose on the regulation of gut microbiota by two S. thermophilus
   strains (4M6 and DYNDL13-4). The results showed that the influence of
   lactose supplementation on gut microbiota induced by S. thermophilus
   ingestion was strain-specific. Although two S. thermophilus strains
   ingestion introduced similar perturbations in the fecal microbiota and
   gut microbial metabolism, the regulation of DYNDL13-4 on the gut
   microbiota and metabolism was more affected by lactose than 4M6. More
   specifically, lactose and 4M6 supplementation mainly enriched pathways
   of d-glutamine and d-glutamate metabolism, alanine, aspartate, and
   glutamate metabolism, and tryptophan and phenylalanine metabolism in
   the gut, whereas 4M6 only enriched tryptophan and phenylalanine
   metabolism. DYNDL13-4-L (DYNDL13-4 with lactose) had significant
   effects on sulfur, taurine, and hypotaurine metabolism in the gut and
   on phenylalanine, tyrosine, tryptophan biosynthesis, and linoleic acid
   metabolism in serum relative to the DYNDL13-4. Our study demonstrated
   the strain-specific effect of lactose and S. thermophilus
   supplementation on gut microbiota and host metabolism. However,
   considering the complexity of the gut microbiota, further research is
   necessary to provide insights to facilitate the design of personalized
   fermented milk products as a dietary therapeutic strategy for improving
   host health.

   Keywords: S. thermophilus, lactose, gut microbiota, metabolome

1. Introduction

   Streptococcus thermophilus (S. thermophilus) is a major dairy starter
   with rapid acidification capability used for the manufacture of yogurt
   and cheese. Previous studies have reported the beneficial impacts of S.
   thermophilus strains on the host health, such as improving lactose
   digestion, alleviating intestinal mucositis, and preventing colorectal
   tumorigenesis [[42]1,[43]2], demonstrating alteration of gut microbiota
   after the intervention of the S. thermophilus strains. In addition,
   probiotic combinations that included S. thermophilus strains have been
   found to have a regulatory effect on the host. The combination of S.
   thermophilus and Bifidobacterium lactis given to healthy infants aged 3
   to 24 months lowered the risk of their colitis [[44]3]. The probiotic
   bacterial mixture, including S. thermophilus, significantly reduced the
   diarrhea time by 45 h than the oral rehydration solution alone [[45]4].
   Clinical research indicated that probiotic mixes contained
   Bifidobacterium animalis subsp. lactis, Streptococcus thermophilus,
   Lactobacillus bulgaricus, and Lactococcus lactis subsp Lactis could
   regulate the responsiveness of an extensive brain network in healthy
   women [[46]5].

   Lactose is an exclusive constituent of mammalian milk, which has a
   concentration of approximately 4.6 g/100 mL in bovine milk [[47]6]. It
   has been observed that lactose does not undergo entire metabolism and
   absorption in the small intestine, and a proportion of dietary lactose
   might reach the colon [[48]7]. Dietary lactose supplementation could
   improve gut health and enhance host immunity. Recent studies have
   indicated that lactose impacts the composition and metabolome of the
   gut microbiota [[49]8,[50]9,[51]10]. Previous human studies have
   discovered a positive connection between lactose consumption and the
   presence of Bifidobacterium in adult fecal samples [[52]11]. Lactose
   treatment was observed to increase the abundance of Bifidobacterium and
   Lactobacillus, decrease the abundance of pathogens, and raise levels of
   acetate and lactate in the infant’s gut in vitro [[53]12,[54]13].
   Furthermore, lactose in human milk stimulates the transcription of
   cathelicidin antimicrobial peptides and then exhibits immunomodulatory
   properties in the neonate [[55]9].

   S. thermophilus prefers lactose to other sugars as its primary carbon
   and energy source. A previous study has shown that lactose consumption
   contributed to the rapid and high level of S. thermophilus LMD9
   colonization in the fecal samples of gnotobiotic rats [[56]14]. Our
   previous research demonstrated that lactose could protect the bile salt
   stress response of S. thermophilus in a strain-independent manner in
   vitro [[57]15]. Furthermore, we have investigated the effect on the gut
   microbiota and host metabolism in C57BL/6J mice by multiple S.
   thermophilus strains based on inter-strain differences. Previously, it
   has been demonstrated that consumption of S. thermophilus DYNDL13-4 and
   DQHXNQ38M61 led to greater alterations in amino acid and lipid
   metabolism when compared to LMD9 and 4M6. This could be linked to the
   increased presence of Bifidobacterium, Coriobacteriaceae UCG-002,
   Rikenellaceae RC9 gut group, and Lactobacillus in the DYNDL13-4 and
   DQHXNQ38M61 group. Meanwhile, DYNDL13-4 and DQHXNQ38M61 are
   phylogenetically distinct from 4M6 and LMD9 based on genomic analysis.
   However, it remains unclear if lactose supplementation can influence
   the gut microbiota and host metabolism driven by S. thermophilus in
   vivo. In this research, the effect of lactose on the regulation of gut
   microbiota and metabolism in C57BL/6J mice by two S. thermophilus
   strains was evaluated.

2. Materials and Methods

2.1. Mice

   Malespecific pathogen-free (SPF) C57BL/6J mice (age 7–8 weeks) were
   obtained from the Beijing Vital River Laboratory Animal Technology Co.,
   Ltd. (Beijing, China). All the male C57BL/6J mice were housed in the
   SPF environment in the animal experimental center of the Department of
   Food Science and Technology, Jiangnan University, Wuxi, China. The mice
   were kept at 22 °C and were subjected to a standard 12 h/12 h
   light/dark cycle. Before experiments, they were fed adaptively for a
   week. All animals had free access to chow diet and water. The
   experimental procedures were conducted strictly by the guidelines of
   the Ethics Committee of Jiangnan University, China (JN.
   No20220915c1121202[333]).

2.2. Experimental Design

   The mice were randomly assigned to six treatment groups (n = 12 per
   group) as follows: the control was orally gavaged with 0.9% saline, the
   two S. thermophilus strains groups were orally gavaged with a dose of
   10^9 CFU/d for each mouse (referred to as 4M6 and DYNDL13-4), the
   lactose group was orally gavaged with a dose of 4 g/kg BW/d (referred
   to as lactose), and the combination groups were orally gavaged with 4
   g/kg BW/d lactose and S. thermophilus strain (referred to as 4M6-L and
   DYNDL13-4-L). For S. thermophilus preparation, the strains were
   cultured in M17 broth for 12 h, respectively, and then resuspended in
   sterile saline to a concentration of 5 × 10^9 CFU/mL. For lactose
   preparation, lactose was dissolved into the sterile saline to a
   concentration of 40% (w/w). All mice were subjected to the
   interventions for 28 days. At the end of the experiment, we placed each
   mouse individually in the sterilized and clean cage and then collected
   fresh feces using Eppendorf tubes. In addition, we euthanized the mice
   in each group on day 28 and collected the blood samples. The fecal and
   serum samples were frozen at −80 °C for DNA extraction and metabolome
   analysis.

2.3. 16S rRNA Gene Sequencing

   Fecal DNA was extracted using a FastDNA Spin Kit (MP Biomedicals,
   Solon, OH, USA). Using primers as previously described [[58]16], the
   V3-V4 region of the 16S ribosomal RNA gene was amplified. Alpha and
   beta diversity were determined via the MicrobiomeAnalyst platform
   [[59]17,[60]18]. The difference in the abundance of bacterial taxa
   among groups was determined through linear discriminant analysis (LDA).

2.4. Metabolite Analysis of Fecal and Serum Samples

   Fecal and serum extracts were prepared for metabolite profiling as
   previously described [[61]19]. In brief, 100 mg thawed fecal samples
   were placed in an Eppendorf tube with the sterilized beads added and
   extracted with 800 μL ice-cold methanol. The samples were homogenized
   and centrifuged at 12,000× g for 15 min. A 400 μL portion of the
   supernatant was transferred and subjected to drying using a vacuum
   evaporator. For serum, a volume of 100 μL was aliquoted from each
   sample into the Eppendorf tube. Subsequently, 400 μL of ice-cold
   methanol was added to the aliquoted serum, followed by vortexing and
   incubation at −20 °C for 1 h. The supernatant of extracts was obtained
   by centrifugation and then dried as described above. The dried extracts
   from feces and serum were resuspended in 200 and 100 μL methanol/water
   (4:1, v/v) for injection, respectively. To prepare the quality control
   (QC) sample, aliquots of the feces and serum samples were pooled,
   respectively.

   Extracts were separated by UHPLC and analyzed on a high-resolution
   Q-Exactive mass spectrometer. Electrospray ionization (ESI) was used to
   detect positive ions and negative ions. In the sample queue, a QC
   sample was set up in 10 experimental samples at intervals to correct
   the repeatability of the analysis. The raw data were uploaded to
   Compound Discovery 3.3 (Thermo Fisher Scientific, Waltham, MA, USA)
   with a metabolomics analysis pipeline, and tables with retention time,
   m/z, and peak area were generated for further analysis.

2.5. Statistical Analyses

   Statistical analyses were performed using the GraphPad Prism 8.0. The
   one-way ANOVA or two-way ANOVA was used for statistical analysis.
   Tukey’s test was applied, and a p-value less than 0.05 indicated
   statistical significance.

3. Results

3.1. Effect of Lactose Supplementation on Gut Microbiota Induced by S.
thermophilus Ingestion Is Strain-Specific

   The changes in intestinal microbiota were further explored from fecal
   samples collected from the mice administered with S. thermophilus
   strains with and without lactose supplementation (4M6, 4M6-L,
   DYNDL13-4, and DYNDL13-4-L). PCoA of Bray‒Curtis distance of
   genus-level taxonomic profiles of these samples showed that all
   interventions altered the structure of the indigenous gut microbiota
   after 28 days of treatments ([62]Figure 1a). Among these five groups,
   samples in the 4M6, DYNDL13-4, and 4M6-L groups showed a clear
   clustering, whereas the DYNDL13-4-L and lactose groups clustered
   together. This indicated that the ingestion of two S. thermophilus
   strains (4M6 and DYNDL13-4) introduced similar perturbations in the
   indigenous microbial community structure, and the regulation of
   DYNDL13-4 on gut microbiota was more affected by lactose than 4M6.
   Additionally, both 4M6 and DYNDL13-4 decreased alpha diversity relative
   to the control group, and those were restored under the supplement of
   lactose ([63]Figure 1b).

Figure 1.

   [64]Figure 1
   [65]Open in a new tab

   Effect of lactose supplementation on gut microbiota induced by S.
   thermophilus ingestion is strain-specific. (a) PCoA plot based on
   Bray–Curtis distance of taxonomic profiles of fecal samples on day 28;
   (b) Chao index of α-diversity of the fecal microbiota after 28 days of
   intervention. One-way ANOVA with Tukey’s multiple comparisons; * p <
   0.05, ** p < 0.01, **** p < 0.0001; (c) The heatmap of genus-level taxa
   with LEfSe analysis on day 28.

   The abundance of seven genera, including Lachnoclostridium, A2,
   Ruminiclostridium, Alistipes, [Eubacterium] xylanophilum group, ASF356,
   and Ruminococcaceae UCG-004 decreased compared to the control, with the
   abundance of only one common genera increased in all the treatments,
   i.e., Bifidobacterium ([66]Figure 1c). 4M6 and DYNDL13-4 shared common
   modulation effects on the gut microbiota of mice, such as increased
   abundance of Bifidobacterium, Akkermansia, Dubosiella, and
   Faecalibaculum, and decreased Prevotellaceae NK3B31 group and
   Ruminiclostridium 5, with variation in 16 genera between the
   treatments. With the supplementation of lactose and 4M6, the abundance
   of Bilophila, Ruminococcaceae NK4A214 group, Candidatus Saccharimonas,
   and [Eubacterium] coprostanoligenes group further increased in 4M6-L.
   On the other hand, the supplementation of lactose with DYNDL13-4
   resulted in a similar pattern of gut microbiota as the lactose group,
   with increased abundance of ten genera in DYNDL13-4-L (e.g.,
   Ruminococcaceae UCG-010, Coriobacteriaceae UCG-002, Prevotellaceae
   UCG-001, and [Eubacterium] coprostanoligenes group). Additionally, nine
   genera in 4M6-L and two genera in DYNDL13-4-L that had no significance
   in corresponding S. thermophilus strains and lactose group were
   affected notably compared to the control. This may be a complex
   regulation of gut microbiota by the combination of lactose and strains.

3.2. Effect of Lactose Supplementation on Fecal Metabolic Profile Induced by
S. thermophilus Ingestion Is Similar to Gut Microbiota

   To characterize the effect of S. thermophilus strains with and without
   lactose supplementation on metabolic profile in gut microbiota, PLS-DA
   was performed on the fecal metabolome data ([67]Figure 2a and
   [68]Figure S1a). The results revealed that all the treatments altered
   the fecal metabolic profile compared to the control. The differential
   metabolites between each treatment and the control group are shown in
   [69]Figure S1c. In line with fecal microbiota, S. thermophilus strain
   groups (4M6 and DYNDL13-4), as well as 4M6-L, clustered closely,
   whereas DYNDL13-4-L and lactose group clustered more closely
   ([70]Figure 2a and [71]Figure S1c). There were 14 commonly changed
   compounds among the five treatment groups ([72]Figure S2a). Lactose
   intervention resulted in the most abundant alterations (64 specific
   metabolites) in feces, with the least metabolite changes in the two S.
   thermophilus strain groups (10 specific metabolites by 4M6 and 19 by
   DYNDL13-4). Lactose supplementation with S. thermophilus strains
   increased the number of altered metabolites in the feces of mice.

Figure 2.

   [73]Figure 2
   [74]Open in a new tab

   Effect of lactose supplementation on fecal metabolic profile induced by
   S. thermophilus ingestion. (a) PLS-DA score plots based on the fecal
   metabolomic profiles in negative ionization modes; (b) the enrichment
   pathways established using 84 metabolites in fecal samples on day 28
   between each treatment and the control group; (c–f) Box and Whisker
   plots of fecal metabolites on tryptophan (c), histidine (d), arginine
   (e), and glutamate metabolism (f). * p < 0.05, ** p < 0.01, *** p <
   0.001, **** p < 0.0001.3.3. Effect of Lactose Supplementation on Serum
   Metabolome Induced by S. thermophilus Ingestion Is Strain-Specific.

   Among the differentially abundant pathways, phenylalanine, tyrosine and
   tryptophan biosynthesis and metabolism, arginine biosynthesis and
   metabolism, and histidine metabolism were commonly enriched pathways in
   fecal samples of the five intervention groups ([75]Figure 2b). Lactose
   supplementation enriched more pathways during the ingestion of S.
   thermophilus strains, with differentiated responses on 4M6 and
   DYNDL13-4, respectively. 4M6-L significantly enriched pathways of
   d-glutamine and d-glutamate metabolism, alanine, aspartate, and
   glutamate metabolism, and DYNDL13-4-L had significant effects on
   sulfur, taurine, and hypotaurine metabolism when compared to their
   corresponding strain groups ([76]Figure 2b).

   Among the metabolites of tryptophan metabolism in feces,
   indole-3-ethanol, indole-3-acetic acid, 3-methylindole, and serotonin
   were reduced in five treatment groups ([77]Figure 2c). 2-Aminomuconate,
   involved in kynurenine catabolic metabolism, was lowered only in 4M6-L
   and DYNDL13-4-L, with a reduction of L-tryptophan in groups containing
   lactose consumption (lactose, 4M6-L, and DYNDL13-4-L). Moreover, indole
   derivatives (including indole, 3-indole propionic acid,
   indole-3-pyruvic acid, methyl indole-3-acetate, and
   4-indolecarbaldehyde) and kynurenic acid decreased in DYNDL13-4-L,
   which had a similar trend with those in lactose group. In addition,
   other indole derivatives, such as indole-2-carboxylic acid and
   5-methoxyindole, decreased exclusively in the lactose group. These
   results revealed that tryptophan metabolism was down-regulated to
   varying degrees by five treatment groups, and lactose supplementation
   enhanced these changes, especially in indole derivatives, with
   DYNDL13-4 more pronounced than 4M6. What is more, metabolites involved
   in arginine, histidine, and glutamate metabolism in DYNDL13-4 were also
   more influenced by lactose supplementation than those in 4M6 in feces
   ([78]Figure 2d–f).

   A total of 525 metabolites were detected in serum, of which 229 were
   identified in the negative mode and 296 in the positive mode. In
   contrast to fecal data, samples in the lactose group could not be
   discriminated from that in the control in serum metabolome according to
   PLS-DA plots, S. thermophilus strain samples with or without lactose
   supplement, especially DYNDL13-4 and DYNDL13-4-L, were separated from
   the control ([79]Figure 3a and [80]Figure S1b). In the case of lactose
   supplementation, serum metabolic pattern was more influenced in
   DYNDL13-4 than 4M6 ([81]Figure 3a and [82]Figure S1b,d). Two
   metabolites (including N-isovalerylglycine and indole-3-acetic acid)
   were shared by DYNDL13-4, 4M6, DYNDL13-4-L, 4M6-L, and the lactose
   group. Among these five groups, DYNDL13-4 (8 specific metabolites) and
   DYNDL13-4-L (38 specific metabolites altered more metabolites than 4M6
   (4 specific metabolites) and 4M6-L (7 specific metabolites), with no
   specific alteration observed as a result of lactose consumption. Like
   fecal data, the lactose supplement promoted more changes in serum
   metabolome during S. thermophilus strains consumption.

Figure 3.

   [83]Figure 3
   [84]Open in a new tab

   Effect of lactose supplementation on serum metabolome induced by S.
   thermophilus ingestion is strain-specific. (a) PLS-DA score plots based
   on the serum metabolomic profiles in positive ionization modes; (b) Box
   and Whisker plots of serum metabolites on tryptophan metabolism; (c)
   the enrichment pathways established using 84 metabolites in serum
   samples on day 28 between each treatment and the control group. * p <
   0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Pathway enrichment
   analysis revealed that tryptophan, sulfur, pyruvate metabolism, steroid
   hormone, unsaturated fatty acids, and arginine biosynthesis were
   enriched in the serum of the strain treatment groups (4M6, DYNDL13-4,
   4M6-L, and DYNDL13-4-L) compared to the control group. More pathways
   were enriched in S. thermophilus strain groups with lactose supplement
   than the corresponding strain groups. The enrichment of taurine and
   porphyrin metabolism was observed in 4M6-L and DYNDL13-4-L, with no
   change in 4M6 and DYNDL13-4. Additionally, 4M6-L significantly enriched
   histidine and arginine metabolism relative to the 4M6, and DYNDL13-4-L
   remarkably enriched phenylalanine, tyrosine, and tryptophan
   biosynthesis and linoleic acid metabolism relative to the DYNDL13-4
   ([85]Figure 3c).

   In view of metabolites on tryptophan metabolism in serum among four
   groups (4M6, DYNDL13-4, 4M6-L, and DYNDL13-4-L), serotonin was
   elevated, whereas indole-3-acetic acid and 5-hydroxyindole-3-acetic
   acid were generally reduced compared to the control. Indole-3-lactic
   acid decreased in 4M6-L and DYNDL13-4-L but showed no significance in
   4M6 and DYNDL13-4 compared to the control ([86]Figure 3b). In addition,
   DYNDL13-4 and DYNDL13-4-L, especially DYNDL13-4-L, resulted in the
   decrease of other indole derivatives (including 4-indolecarbaldehyde
   and indole-3-acetaldoxime), kynurenine catabolic metabolites (including
   xanthurenic acid and 4-(2-aminophenyl)-2,4-dioxobutanoate), and
   tryptophan derivatives (including 5-hydroxy-dl-tryptophan and
   N-acetyl-dl-tryptophan), most of which were decreased in groups of
   fecal samples containing lactose consumption, especially DYNDL13-4-L
   and lactose group. These results indicated that S. thermophilus strains
   differentially altered the catabolic flux of tryptophan in serum. The
   change was stimulated with continued lactose supplementation, as
   evidenced by the decreased kynurenine branch and indole derivatives and
   the increased serotonin. Likewise, serum metabolites of fatty acids
   were differentially increased in S. thermophilus strains consumption,
   and the lactose supplementation further enriched more fatty acids, such
   as lignoceric, docosanoic, and palmitoleic acid. These changes were the
   most prominent in DYNDL13-4-L, and the fatty acids profiles in 4M6-L
   tended to be those in DYNDL13-4. This suggested that lactose intake may
   reduce differences in lipid metabolism induced by strain variability.

4. Discussion

   Most studies just focused on evaluating the protective effect of
   prebiotic supplementation on the individual strains in the gut
   microbiota [[87]20,[88]21,[89]22]. In this research, we considered the
   strain-specific responses to lactose supplements, and the regulation of
   DYNDL13-4 on the gut microbiota was more affected by lactose than 4M6.
   Various prebiotics may have a protective effect on the probiotics by
   enhancing their antioxidant capacity, adhesive capacity, resistance to
   digestive fluids, anti-bacterial effects, and cross-feeding behavior in
   the gastrointestinal tract [[90]23,[91]24,[92]25,[93]26]. Thus, it is a
   reasonable hypothesis that these factors may contribute to the
   differential regulation of strains by lactose. Bifidobacterium was the
   only one increased genera shared by five treatment groups in our
   results. This is in accordance with the previous studies that S.
   thermophilus ATCC 19,258 enhanced Bifidobacterium in the gut of murine
   models of intestinal tumorigenesis [[94]1], and lactose could regulate
   the Bifidobacterium in healthy adults independently of donors [[95]27].

   Alteration in the gut microbiota composition can change metabolic
   profiles, affecting host health [[96]28]. We observed that tryptophan
   metabolism in feces and serum was regulated by four treatment groups
   (DYNDL13-4, 4M6, DYNDL13-4-L, 4M6-L), and the combination of lactose
   and strains enhanced more changes, especially in indole derivatives.
   Evidence from animal studies has suggested that tryptophan metabolism
   may affect host metabolic health via a host–microbiota interaction
   [[97]29]. As the tryptophan catabolites by gut microbiota, indole
   derivatives have been shown to have beneficial effects on host
   diseases, such as inflammatory bowel disease (IBD) and neurological
   diseases [[98]30]. It was found that IBD patients have reduced fecal
   concentrations of the indole-3-acetate [[99]31]. 3-Indolepropionicacid
   was reduced in the serum of patients with colitis when compared to
   healthy subjects [[100]32], and oral treatment of indole and
   3-indolepropionicacid was found to alleviate colonic inflammation in
   mice [[101]32,[102]33]. Plasma levels of 3-indole propionic acid were
   significantly diminished significantly in subjects with Huntington’s
   disease compared to healthy controls [[103]34]. Additionally, a
   previous study demonstrated a positive correlation between elevated
   milk consumption and increased levels of 3-indole propionic acid in the
   serum of genetically lactase non-persistent individuals. Our results
   emphasized the regulation of indole derivatives by lactose and
   DYNDL13-4-L and suggested that dietary milk consumption may affect
   intestinal and neurological function by modulating tryptophan and its
   derivatives in the host.

   The combination of lactose and S. thermophilus strain significantly
   increased the metabolites of arginine and histidine metabolism than the
   individual strain in feces in a strain-specific pattern. Arginine has
   been found to improve intestinal mucosal barrier function and is an
   essential amino acid for the growth and development of intestinal stem
   cells [[104]35]. In contrast, higher arginine levels may be associated
   with an increased risk of ischemic heart disease [[105]36].
   Multi-cohort analysis showed that abnormal histidine metabolism was
   related to the occurrence of colorectal cancer and diabetes
   [[106]37,[107]38]. Thus, the influence of S. thermophilus and lactose,
   key elements associated with fermented milk, on hosts needs to be
   verified further, including with altered health models, which would
   provide insights to facilitate designing personalized fermented milk
   products as a promising dietary therapeutic strategy for improving host
   health.

   In addition, we observed the most abundant metabolites in feces but no
   significant alteration in serum in the lactose treatment. On the one
   hand, physiological changes, including chemical and nutritional
   gradients, are changed in various regions of the small intestine and
   colon [[108]39]. Fecal samples are only a snapshot of the gut
   microbiota and are not comprehensive to represent the circulating
   chemicals. On the other hand, a serum metabolism analysis showed that
   the amount of lactose peaked in the blood of the healthy volunteers
   having consumed milk in the first 6 h and dropped off afterward
   [[109]40]. Our test for serum metabolism was at least 12 h, and
   differential metabolites may not be observed in the lactose group. We
   noted that lactose supplementation remarkably enriched amino acid
   metabolism in serum compared to the individual strain treatments, such
   as histidine, arginine, phenylalanine, tyrosine, and tryptophan
   metabolism. Previous studies have shown that the consumption of
   fermented dairy products could regulate amino acid metabolism in the
   serum of humans. Tyrosine, valine, and proline increased in the serum
   of healthy humans after cheese intake [[110]40]. A higher abundance of
   plasma proline and phenylalanine were observed after the yogurt intake
   compared to the milk intake [[111]41]. These studies mostly attributed
   to the free amino acids released during the fermentation of dairy
   products. Our research further revealed several amino acid changes in
   serum may be caused by a combination of lactose and S. thermophilus
   strains.

5. Conclusions

   Our results showed that the effect of lactose supplementation on the
   regulation of S. thermophilus on the gut microbiota and host metabolism
   is strain-dependent. Although consumption of the two S. thermophilus
   strains caused similar changes in the fecal microbiota and metabolism,
   lactose had a greater impact on DYNDL13-4 regulation than it did on
   4M6. More specifically, the combination of lactose and 4M6 mainly
   enriched pathways of d-glutamine and d-glutamate metabolism, alanine,
   aspartate and glutamate metabolism, and tryptophan and phenylalanine
   metabolism in feces, whereas 4M6 only enriched tryptophan and
   phenylalanine metabolism. DYNDL13-4-L had significant effects on
   sulfur, taurine, and hypotaurine metabolism in feces and on
   phenylalanine, tyrosine, tryptophan biosynthesis, and linoleic acid
   metabolism in serum relative to the DYNDL13-4. Our study provided
   insight into the complexity of how fermented milk product consumption
   may influence host health through modifications of the metabolites and
   gut microbiota.

Supplementary Materials

   The following supporting information can be downloaded at:
   [112]https://www.mdpi.com/article/10.3390/nu15224767/s1. Figure S1:
   Effect of lactose supplementation on metabolic profile induced by S.
   thermophilus related to [113]Figure 2 and [114]Figure 3. (a) PLS-DA
   score plots based on the fecal metabolomic profiles in positive
   ionization modes. (b) PLS-DA score plots based on the serum metabolomic
   profiles in negative ionization modes. (c,d) Heatmap of the
   differential metabolites of each treatment relative to the control in
   feces (c) and serum (d). Figure S2: (a,b) Venn diagram of all
   identified metabolites of each treatment relative to the control in
   feces (a) and serum (b).
   [115]Click here for additional data file.^ (418.4KB, zip)

Author Contributions

   Conceptualization, P.Y., X.L., J.Z., H.Z. and W.C.; methodology, P.Y.,
   Y.P., M.G., X.L., J.Z. and H.Z.; software, P.Y., Y.P., Z.P. and M.G.;
   validation, P.Y., Y.P. and Z.P.; formal analysis, P.Y., Y.P. and Z.P.;
   investigation, P.Y., Y.P., M.G. and X.L.; resources, B.Y., X.L., J.Z.,
   H.Z. and W.C.; data curation, P.Y., Y.P. and M.G.; writing—original
   draft preparation, P.Y.; writing—review and editing, X.L. and Y.-K.L.;
   visualization, P.Y. and Y.P.; supervision, X.L., J.Z., H.Z. and W.C.;
   project administration, X.L.; funding acquisition, X.L., J.Z., H.Z. and
   W.C. All authors have read and agreed to the published version of the
   manuscript.

Institutional Review Board Statement

   This research was approved by the Ethics Committee of Jiangnan
   University, China (JN. No20220915c1121202[333]).

Informed Consent Statement

   Not applicable.

Data Availability Statement

   Data supporting reported results are available upon reasonable request
   and in accordance with ethical principles.

Conflicts of Interest

   The authors declare no conflict of interest.

Funding Statement

   The present research work was supported by the National Natural Science
   Foundation of China (No. 31871829, 31820103010, and 32021005) and the
   Collaborative Innovation Centre of Food Safety and Quality Control in
   Jiangsu Province (Jiangsu, China).

Footnotes

   Disclaimer/Publisher’s Note: The statements, opinions and data
   contained in all publications are solely those of the individual
   author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI
   and/or the editor(s) disclaim responsibility for any injury to people
   or property resulting from any ideas, methods, instructions or products
   referred to in the content.

References