Utilization of four galactans by Bacteroides thetaiotaomicron A4 based on transcriptome

The interaction between gut microbiota and polysaccharide is being paid more and more attention. Galactan is a kind of polysaccharide mainly composed of galactose, and it has been shown to play a significant role in regulating gut microbiota. Bacteroides thetaiotaomicron is considered as the best degrader of polysaccharides. The purpose of this paper was to investigate the utilization of four different galactans, including agarose, carrageenan, arabinogalactan, and glucofucogalactan, by our own isolated B. thetaiotaomicron A4. Samples of cultures grown on either four galactans groups or control groups were collected. The change of OD 600 , pH, and short-chain fatty acids (SCFAs) duringfermentationwere determined, andgrowthcurve and transcriptome of B. thetaiotaomicron A4 were studied. B. thetaiotaomicron A4 could utilize the four galac-tansandgrewwellonthem,withcarrageenanbeingmostutilized,followedbyarabino- galactan,glucofucogalactan,andagarose.SCFAs(mainlyaceticacidandpropionicacid) producedalongwiththedecreasedpHduringfermentation.Alargenumberofgenesof B. thetaiotaomicron A4 were upregulated and functioned in different pathways during the degradation of the four galactans. The carbohydrate metabolism-related pathways of B. thetaiotaomicron A4 were enriched after feeding the four galactans, although the specificpathwaysweredifferentamongfourgalactansgroups.Thedifferentstructural characteristicsoffourgalactansrequiredthat B.thetaiotaomicron A4couldexcretecor-respondingenzymestodegradethem.Theseresultshelptounderstandtheinteraction between galactans and gut microbe.

However, the molecular weight and sulfate content of galactans are vital for these bioactivities (Hatada, Ohta, & Horikoshi, 2006). Furthermore, they could also affect gut microbiota utilizing galactans (Benjdia, Martens, Gordon, & Berteau, 2011). Recently, the relationship between the structural characteristics of galactans and the ability of gut microbiota to utilize galactans have attracted extensive attention (Hu et al., 2006;Shang et al., 2017;Sun et al., 2019). In addition, it has been reported that agarose, carrageenans, and arabinogalactan could be utilized by Bacteroides (Hehemann, Kelly, Pudlo, Martens, & Boraston, 2012;Li, Shang, Li, Wang, & Yu, 2017;Martens et al., 2011). However, the differences between galactans with different structural characteristics and their utilization by Bacteroides need to be studied and discussed in detail. The purpose of this paper was to investigate the utilization of four different galactans, that is, agarose, carrageenan, arabinogalactan, and glucofucogalactan, by our own isolated B. thetaiotaomicron A4, as well as related utilization genes and function of those genes in pathways. In this study, four galactans were fermented by B. thetaiotaomicron A4. Microbial growth characteristics and short-chain fatty acids (SCFAs) production were determined during fermentation. Furthermore, the differentially expressed genes were screened based on transcriptome analysis, and the functional annotation Gene Ontology (GO) and pathway annotation Kyoto Encyclopedia of Genes and Genomes (KEGG) of these genes were performed to clarify the gene transcription level in metabolic pathway. Finally, CAZys involved in the degradation of the four galactans were investigated to study the effects of structural characteristics on the utilization of four galactans by B. thetaiotaomicron A4.

Chemicals and reagents
H. erinaceus fruiting body from Changbai Mountains was purchased from Jiangxi NanHua Medicine Co. Ltd, China; and was used to isolate and purify glucofucogalactan according to our method (Wang et al., 2018); arabinogalactan (larch source) was bought from Sigma-Aldrich, USA; carrageenan (mixtures of κ, ι, and λ) was obtained from Aladdin

Bacterial strain and medium
B. thetaiotaomicron A4 was obtained from feces of healthy people.
Growth medium was composed of BHI supplemented with 0.8 g/L of K-cysteine HCl; selective medium was composed of carbon-free BHI with 0.8 g/L of L-cysteine HCl and 2 g/L of specific carbon source. Solid medium was composed of growth medium supplemented with 15 g/L of agar. All the mediums were sterilized at 121 • C for 15 min. 0.005 g/L of hemin, 0.01 g/L of vitamin K1, 0.1 g/L of kanamycin, and 0.0075 g/L of vancomycin were filter-sterilized through a 0.22 μm filter, and then added to the autoclaved medium that cooling to 60 • C.

Characterization of bacterial growth on four galactans
B. thetaiotaomicron A4 was spread on solid medium. After 48 h anaerobic cultivation at 37 • C, single and well-formed colony was picked into growth medium. After 48 h anaerobic cultivation at 37 • C, inoculum (10%, v/v) was added to selective medium (agarose, carrageenan, arabinogalactan, or glucofucogalactan as the sole carbon source. D-Glucose and no carbon source were used as positive control [PC] and negative control [NC]). During cultivation at 37 • C under anaerobic conditions, fermentation broths were taken out for further analysis at 0, 6, 12, 24, and 36 h. The analysis was performed in triplicates and all operations were carried out in anaerobic workstation.
OD 600 and pH were recorded to characterize the growth of B. thetaiotaomicron A4. The sugar contents of fermentation broths were analyzed using the phenol-sulfuric acid colorimetric method at the wavelength of 490 nm (Dubois, Gilles, Hamilton, Rebers, & Smith, 1956). D-Galactose and D-glucose were used as the standard of four galactans and the control groups, respectively.

Determination of SCFAs
The determination of SCFAs was slightly modified according to the method of Fenster et al. (2003). In brief, samples were centrifuged at 12,000 rpm for 10 min at 4 • C, and then 0.5 ml of each supernatant was mixed with 1.2 ml of ethanol using ice bath. Next, 0.1 ml of H 2 SO 4 (18 mol/L) and 1 mL of n-hexane were added. The esterification lasted for 1 h at 60 • C with shaking (250 rpm). Afterward, the reaction liquid was standing for 15 min, and the supernatant was filtered with 0.22 μm filter membrane.
A SCFA standard mix (Supelco, Bellefonte, PA) was used as the external standards.

Growth curve analysis
Absorbance at 600 nm was measured every 2 h and in triplicate during the anaerobic cultivation using selective medium. The growth curve was generated by GraphPad Prism software.
Sequencing service were provided by Personal Biotechnology Co., Ltd. Main sequencing process included RNA extraction, rRNA remove, mRNA fragmentation, cDNA reverse transcription, PCR amplification, and Illumina Novaseq sequencing. The gene expression level was normalized with FPKM method (Trapnell et al., 2012). HTSeq was used to analyze the expression levels of genes (Anders, Pyl, & Huber, 2015).
The reproducibility of genes expression between biological replicates was examined by using Pearson correlation. DESeq was used to screen for differentially expressed genes (DEGs), where the fold change of the transcription in the galactan groups relative to that in the blank control should be over 2 and the p value be less than 0.05 (Anders & Huber, 2010

Statistical analysis
The results were presented as mean ± standard deviation and were analyzed using one-way analysis of variance with SPSS 20.0. Least significant difference method and Dunnett's T3 method were used for the post-hoc test, in the case of homogeneity and heterogeneity of variance, respectively (Zhao, Liu, Liang, Feng, & Xu, 2019). p < 0.05 was considered statistically significant.

The utilization of four galactans by B. thetaiotaomicron A4
The physicochemical properties of four galactans were shown in Table 1. The molecular weight of carrageenan was 4331 kDa, and its monosaccharide composition was mainly galactose, which contained a small amount of glucose. Agarose was polymerized by galactose, and its molecular weight was 191 kDa. The molecular weight of arabinogalactan was 186 kDa, and its monosaccharide composition was mainly galactose, which contained a small amount of arabinose. The molecular weight of glucofucogalactan was 15 kDa, and its monosaccharide com- position was mainly galactose, followed by glucose and fucose. To test the effects of four galactans on the growth of B. thetaiotaomicron A4, the strain was cultured with each of the four galactans as the sole carbon source, with glucose and no carbon source as PC and NC. All the four galactans significantly promoted the growth of B. thetaiotaomicron A4, with the best effect seen in glucofucogalactan, followed by arabinogalactan, carrageenan, and agarose ( Figure 1A).
pH was also an indicator that microbes could utilize polysaccharides (Kim, Shin, Lee, Moon, & Lee, 2018). The change of pH in the fermentation process was directly reflected by the decrease of pH of the fermentation liquid at each fermentation time point (6, 12, 24, and 36 h) compared with that at 0 h ( Figure 1B). The pH of the glucose group, the carrageenan group, and the glucofucogalactan group decreased significantly in the first 12 h, whereas the pH of the agarose group and the arabinogalactan group decreased significantly in the first 24 h. The pH of the arabinogalactan group decreased to more extent when compared with the other three galactan groups.
The sugar content of all groups decreased significantly after 12 h ( Figure 1C). The decrease rate of sugar content of four galactans reach the peak during 6-12 h, which was also seen for the growth rate. The increase of OD 600 , along with decrease of pH and sugar content, provided solid proof for the utilization of four galactans by B. thetaiotaomicron A4.

The production of SCFAs
To understand the metabolism of four galactans from B. thetaiotaomicron A4, we analyzed the production of SCFAs during fermentation (Table 2). AA and PA were the main SCFAs, which was consistent with the literature (Zhang, Chen, & Ding, 2019). In addition, there were small amounts of IBA and IVA, which were the product of protein fermentation. IBA and IVA are branched-chain fatty acids that are associated with insulin resistance (Turroni, Brigidi, Cavalli, & Candela, 2018).
Regarding the yield of total SCFAs, four galactans groups had a significantly higher value than NC group at 12 and 24 h, and arabinogalactan group was the highest among all groups at 36 h. The level of AA in the, PC-, arabinogalactan-, and glucofucogalactan groups increased continuously with the extension of fermentation time, whereas that in the agarose-and carrageenan groups increased continuously in the first 24 h and later decreased. In the aspect of yield of PA, all groups increased continuously during fermentation. The yield of PA in the agarose-, arabinogalactan-, and glucofucogalactan groups were higher than that in PC group at 36 h, with the highest yield found in the agarose group.

Growth curve
The profile of the lag phases of the four galactans fermentation were similar and the stationary phases appeared after 12 h ( Figure 2). Carbohydrate utilization related genes transcription are the essential indicators for investigating the polysaccharides degradation capabilities of microbes (Tan, Zhao, Zhang, Zhai, & Chen, 2018). Studies have shown that more genes involved in the utilization of polysaccharides appear in the mid-logarithmic phase than in the late-logarithmic phase (Despres et al., 2016), so the fermentation broth in the mid-logarithmic phase was selected for transcriptome analysis.

GO enrichment analysis
The upregulated genes of four groups were annotated in the GO database for functional classification. The top 20 GO terms with the most significant enrichment were displayed in Figure 5. After treating with agarose, a large number of genes were enriched in B. thetaiotaomicron A4, for example, those involved in catalytic activity, hydrolase activity, hydrolase activity acting on glycosyl bonds, hydrolase activity of hydrolyzing O-glycosyl compounds, carbohydrate metabolic process, carbohydrate derivative metabolic process, and other functions ( Figure 5A).
For those treated with carrageenan, genes related to conjugation/unidirectional conjugation, polyphosphate metabolic process, polyphosphate kinase activity, phosphotransferase activity, polysaccharide biosynthetic process, genetic transfer, and other functions were enriched ( Figure 5B).
In the arabinogalactan group, cells were significantly enriched in genes targeting transporter activity, carbohydrate metabolic pro-cess, carbohydrate catabolic process, small molecule catabolic process, monosaccharide catabolic/metabolic process, pentose metabolic process, and other functions ( Figure 5C).
After treating with glucofucogalactan, genes related to carbohydrate metabolic process, cell outer membrane/outer membrane, carbohydrate binding, glucose catabolic/metabolic process, hexose catabolic/metabolic process, monosaccharide catabolic/metabolic process, and other functions were significantly enriched ( Figure 5D).

KEGG pathway enrichment analysis
The DEGs involved in biological functions were further analyzed by KEGG pathways, and the top 20 pathways with the most significant enrichment were predicted ( Figure 6). After treating with agarose, a large number of genes in B. thetaiotaomicron A4 were involved in amino acid metabolism (cyanoamino acid metabolism, tyrosine metabolism, beta-alanine metabolism, and phenylalanine

F I G U R E 1
The utilization of four galactans by Bacteroides thetaiotaomicron A4. OD 600 (A), pH decrease (B), and change of sugar content (C). * in pH decrease (B) indicated a significant difference compared with NC group at the same time point and * in change of sugar content (C) indicated a significant difference that each group of 6, 12, 24, 36 h compared with 0 h. *: p < 0.05; **: p < 0.01; ***: p < 0.001 F I G U R E 2 Growth curve of Bacteroides thetaiotaomicron A4 in four galactans metabolism), energy metabolism (carbon fixation pathways in prokaryotes and methane metabolism), carbohydrate metabolism (ascorbate and aldarate metabolism, butanoate metabolism, TCA cycle, glyoxylate, and dicarboxylate metabolism), and lipid metabolism (fatty acid biosynthesis and sphingolipid metabolism) ( Figure 6A).
Only six pathways were enriched after treating with carrageenan ( Figure 6B) Figure 6D).
In this study, the gene encoding GH16 of B. thetaiotaomicron A4 was  (Marcobal et al., 2010). In present study, the same phenomenon was occurred and sugar content of BHI medium with no added carbon (the blank control medium) was about 1 mg/ml ( Figure 1C).

Metabolites of gut microbiota play an important role in regulating
host health by linking polysaccharides to the host. SCFAs are important metabolites of gut microbiota, mainly including AA, PA, and BA . To further understand the relationship between the growth of B. thetaiotaomicron A4 and its metabolism of the four galactans, we quantitatively analyzed SCFAs from the supernatants of the culture medium. The four galactans selectively promoted the production of AA and PA. Regarding AA, the yield of AA in nonlinear galactan (arabinogalactan and glucofucogalactan) groups increased continuously along with the extension of fermentation time, whereas that in linear galactan (agarose and carrageenan) groups, it increased continuously in the first 24 h and later decreased. Regarding PA, its yield in all groups increased continuously during fermentation. AA is a source of energy for the brain, heart, and surrounding tissues, as it could be absorbed by the colon epithelium and transferred via the F I G U R E 5 GO enrichment analysis of DEGs. The vertical axis indicates GO term, and the horizontal axis indicates rich factor. The dot size indicates the number of differentially expressed genes in the term, and the color of the dots corresponds to FDR values circulatory system to the target organs as energy substrate (Di et al., 2018). PA regulates cholesterol synthesis, reduces the production of proinflammatory factors, reduces DNA damage in colon cells, and interacts with fatty acid receptors in the liver to regulate glucose production (Fang, Hu, Nie, & Nie, 2019 Rogowski et al., 2015). After that, they were transported from the outer membrane to the periplasm via SusC-like TonB-dependent transporters (Foley, Cockburn, & Koropatkin, 2016;Glenwright et al., 2017). These two enzymes have been reported to break down agarose in previous studies, unlike the present study, they were found in B.

F I G U R E 7
Venn of the upregulated genes of degrading agarose (A), carrageenan (K), arabinogalactan (Ag), and glucofucogalactan (Rp) by Bacteroides thetaiotaomicron A4 uniformis NP1 and B. plebeius DSM 17135 (Hehemann et al., 2010;Pluvinage et al., 2018). In addition to β-galactosidase, we also found that some α-galactosidases might have agarose degradation activity, such as GH36. Studies have shown that GH36 had catalytic activity for p-nitrophenyl-α-D-galactopyranoside, galacto-oligosaccharides, and galactomannans (Xie, Wang, He, & Pan, 2020), but there were few reports on the degradation of agarose. In this study, only GH16 was predicted to degrade carrageenan. β-Porphyranase (GH16) was significantly upregulated during the degradation of linear galactans (agarose and carrageenan). The ability of β-porphyranase (GH16) to degrade this type of galactans might be related to the lack of side chain substitution. GH43 was predicted to degrade larch arabinogalactan.
Among four galactans, types of enzymes associated with glucofucogalactan were the most, which might be related to complex monosaccharide composition of glucofucogalactan. These enzymes might be potential industrial GHs. Moreover, β-galactosidase from Streptococcus thermophilus have been reported to have tumor suppressive effects . Therefore, future research is needed on effects of the predicted enzymes related to galactans degradation on human health, whereas structural changes and metabolites after galactans degradation also require attention.

CONCLUSION
Four kinds of galactans, including agarose, carrageenan, arabinogalactan, and glucofucogalactan, could be used by our own isolated B. thetaiotaomicron A4. The most utilized by B. thetaiotaomicron A4 was carrageenan, followed by arabinogalactan, glucofucogalactan, and agarose. SCFAs (mainly AA and PA) produced along with the decreased pH during fermentation. A large number of genes of B. thetaiotaomicron A4 were upregulated and functioned in different pathways during the degradation of the four galactans. The carbohydrate metabolismrelated pathways were enriched after utilizing four galactans by B.
thetaiotaomicron A4, although the specific pathways were different among four galactans groups. The different structural characteristics of four galactans required that B. thetaiotaomicron A4 could excrete corresponding enzymes to degrade them. Results of this study could help to understand the utilization of galactans by gut microbe. In future study, the predicted enzymes found in this study will be further studied to understand the detailed degradation mechanism of galactans by B.

ACKNOWLEDGMENTS
This study was supported by the National Science Fund for Distinguished Young Scholars of China (31825020) and Technological Innovation Guidance Plan of Jiangxi Province (20203AEI91007).

CONFLICTS OF INTEREST
The authors declare no conflict of interest.