Alteration of Intestinal ...

[ Pobierz całość w formacie PDF ]
//-->Alteration of Intestinal Microbiota in Mice OrallyAdministered with Salmon Cartilage Proteoglycan, aProphylactic AgentKrisana Asano, Sayuri Yoshimura, Akio Nakane*Department of Microbiology and Immunology, Hirosaki University Graduate School of Medicine, Hirosaki, Aomori, JapanAbstractProteoglycan (PG) extracted from salmon nasal cartilage has potential to be a prophylactic agent. Daily oraladministration of the PG attenuates systemic inflammatory response in the experimental mouse models. In thisstudy, we applied the culture-independent approach to investigate an alteration of intestinal microbiota composition inPG-administered mice. The results indicated that the population level of bacilli increased in the small and largeintestine upon PG administration. On the other hand, the population level of clostridia decreased in the largeintestine. The proportion of bacteria that are able to ferment saccharides and produce short-chain fatty acidsincreased in the small intestine and decreased in the large intestine. Importantly, population level of probioticlactobacilli and bacteria exhibiting the immunomodulatory effect increased in the PG-administered mice. In addition,several disease-associated bacteria decreased upon PG administration. These results provided an understanding ofthe specific role of PG involved in host immune modulation and supported our hypothesis that daily oraladministration of PG improves the overall balance in composition of the intestinal microbial community.Citation:Asano K, Yoshimura S, Nakane A (2013) Alteration of Intestinal Microbiota in Mice Orally Administered with Salmon Cartilage Proteoglycan, aProphylactic Agent. PLoS ONE 8(9): e75008. doi:10.1371/journal.pone.0075008Editor:Stefan Bereswill, Charité-University Medicine Berlin, GermanyReceivedJanuary 28, 2013;AcceptedAugust 8, 2013;PublishedSeptember 9, 2013Copyright:© 2013 Asano et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Funding:This study was supported by the City Area Program for Promotion of Science and Technology in Regional Areas from Ministry of Education,Culture, Sports, Science and Technology (Grant number 2010110001). The funders had no role in study design, data collection and analysis, decision topublish, or preparation of the manuscript.Competing interests:The authors have declared that no competing interests exist.* E-mail: a27k03n0@cc.hirosaki-u.ac.jpIntroductionProteoglycan (PG) consists of core protein with one or morecovalently attached glycosaminoglycan chain(s). It is acomponent of extracellular matrix materials that exist inconnecting tissues such as skin, bone, cartilage and vascularwall by forming a complex with collagen, fibronectin, laminin,hyaluronic acid and other glycoproteins. In corporation withcollagen, fibronectin and laminin, PG has been shown to beinvolved in cellular proliferation and adhesion [1]. We havepreviously shown that PG extracted from salmon nasalcartilage has a potent effect on suppression of inflammatoryresponses induced by heat-killedEscherichia coliin mousemacrophages [2]. In addition, daily oral administration of PGattenuates the severity of experimental inflammatory colitis [3]and autoimmune encephalomyelitis [4]. Although our finding inboth mouse models suggested that PG attenuates systemicinflammation by suppressing T-helper 17 (Th17) linagedifferentiation and inducing Foxp3+regulatory T (Treg) cells,the mechanism of oral PG administration to regulate these Tcells is still elusive.Flora in the gastrointestinal tract (GIT) in mammals is highlycomplex and diverse. They have a profound effect onnutritional, physiological and immunological processes of thehost [5]. Keeping these communities in balance is most likelycrucial for health maintenance. Accumulating evidencessuggested that an unfavorable alteration of the commensalstructure of GIT microbiota implicates in chronic inflammatorybowel disease and also may be instrumental for thedevelopment of systemic immune diseases such as rheumatoidarthritis [6], autoimmune encephalomyelitis [7] and type-1diabetes [8]. Therefore we hypothesized that the prophylacticeffect of PG may result through the microbial communities inGIT. Between 40–60% of bacteria residing within the gut arereported to be un-culturable [9]. Thus in this study, the DNA-based culture-independent method was applied to analyze thecomposition of the GIT microbiota in PG-administered mice.Diversity and population levels of intestinal microbiota werecompared between PG-administered and control mice.Bacterial phylotypes in the small and large intestine whosepopulation levels were associated with PG-administration wereinvestigated. The phylotypes that may have potential toPLOS ONE | www.plosone.org1September 2013 | Volume 8 | Issue 9 | e75008Proteoglycan Administration Alters Gut Microbiotaregulate immune response upon daily PG administration weresuggested.Sequence analysisThe 16S rRNA gene sequences reported herein have beendeposited in the DDBJ Sequence Read Archive (accessionnumber of submission; DRA000939 and DRA001079,accession number of study; DRP000976 and DRP001126,accession numbers of samples; DRS005602 to DRS005609and DRS011995 to DRS012006, accession numbers ofexperiments; DRX005456 to DRX005463 and DRX012124 toDRX012135; accession numbers of runs; DRR006232 toDRR006239 and DRR013297 to DRR013308). The sequencesfrom each labeled-tag were clustered using a sequencesimilarity at 96.00% with the software of cdhit-v4.5.4. Theclusters less than 350 nucleotides in length were discarded.The representative sequence of each cluster was searched forhomology using BLASTN against Genbank database extractedas a product name of 16S rRNA or 16S ribosomal RNA withthe length of more than 299 bases. The taxon for eachrepresentative sequence was assigned from the closest related16S rRNA sequence with a minimum alignment length of 350nucleotides and minimum percent identity of 85.0. Percentrelative count of each taxon was calculated from count numberof each specific taxon and total count number in each sample.Materials and MethodsMice and ethics statementC57BL/6 mice were purchased from Clea Japan, Inc., Tokyo,Japan and maintained under specific pathogen-free conditionsat the Institute for Animal Experimentation, Hirosaki UniversityGraduate School of Medicine. All experiments were carried outin strict accordance with the Guidelines for AnimalExperimentation of Hirosaki University. The protocol wasapproved by the committee on the ethics of the Institute forAnimal Experimentation, Hirosaki University Graduate Schoolof Medicine (Permit number: M08028). For group A, mice werebred at the Institute for Animal Experimentation, HirosakiUniversity Graduate School of Medicine for 3 generations. Forgroup B-group E, mice were purchased at different times andthey were used after 1 week of purchase.PG administration and sample collectionPG extracted from salmon nasal cartilage was purchasedfrom Kakuhiro Co., Ltd. (Aomori, Japan). It was dissolved inphosphate-buffered saline (PBS) at the concentration of 10mg/ml. Six week-old female mice were administered with 200µl of PG or PBS as control per os daily. To prevent fecalconsumption between groups, mice administered with PBS andPG were housed in separate cages. On day 30, mice werestarved for 16 h and then sacrificed by cervical dislocation.Luminal contents of small intestine (from 1 cm below stomachto 1 cm over caecum) and large intestine (from caecum toanus) were collected by gently scraping. Intestinal contentsfrom 3 mice in each group were mechanically homogenizedwith sterile spatula and kept at -80°C until use.Statistical analysisAssociations between bacterial phylotypes and PGadministration were examined by Fisher exact test.Pvaluesless than 0.05 were used to indicate the statistical difference ofbacterial counts between PG-administered and control mice.ResultsNumbers of 16S rRNA gene sequence and taxonomicassignmentFrom DNA sequencing data generated by the FLX TitaniumSequencer, a total of 1,171,101 counts were obtained from 20samples. As shown in Figure S1, the sequences longer than349 nucleotides in length were predominant in group A and B.Thus the sequences having less than 350 nucleotides werediscarded. The bacterial phylotype or taxon for therepresentative sequence was then assigned from the closestrelated 16S rRNA sequence with threshold of 350 nucleotidesin alignment length and 85.0% identity. Total count numberswith or without threshold from each sample are listed in TableS1. A total of 1,107,491 sequences with acceptable thresholdwere obtained from 20 samples with an average of 55,375counts (range: 23,042-96,496) per sample (Table S1 andFigure 1A). The data indicated that 94.6% of total sequencescould be assigned to domain Bacteria. The variety ofphylotypes assigned from each sample is shown in Figure 1B.The relative alteration of the bacterial variety upon PGadministration was not found among all 5 groups. The resultssuggested that there was not an association between PGadministration and biodiversity of GIT microbiota. Thedistribution of intestinal bacteria at phylum-level is shown inFigure 2 and Table S2-Table S3. Excluding unclassifiedphylum, majority of GIT microbial population belonged to 5phyla,Actinobacteria,Bacteroidetes,Firmicutes,ProteobacteriaandVerrucomicrobia.However, the member inDNA isolation, 16S rRNA gene amplification and DNAsequencingFrozen intestinal contents were thawed on ice and DNA wasisolated using QIAamp DNA Stool Mini kit (QIAGEN, Hilden,Germany) according to the manufacturer’s instruction. Primer27F (5’-AGAGTTTGATCMTGGCTCAG-3’, where M = C/A) and519R (5’-GWATTACCGCGGCKGCTG-3’, where W = A/T andK = G/T) were used to amplify the variable region, V1 to V3, of16S rRNA gene [10]. Primer 27F and 519R were tagged with amultiplex identifier (MID) sequence (Table S1) at 5’ end alongwithadaptersequence(5’-CGTATCGCCTCCCTCGCGCCATCAG-3’and5’-CTATGCGCCTTGCCAGCCCGCTCAG-3’, respectively) toallow all samples to be included in a single FLX Titaniumsequencing plate. PCR reaction was performed using FastStartHigh Fidelity PCR system (Roche Diagnostics, Indianapolis, IN)and PCR Nucleotide Mix (Roche Diagnostics). The conditionswere 95°C for 2 min, 35 cycles of 95°C for 30 sec, 55°C for 30sec and 72°C for 30 sec, followed by 72°C for 4 min. DNAproducts were quantified and pooled for sequencing with theGenome Sequencer FLX Titanium system (Roche Diagnostics)PLOS ONE | www.plosone.org2September 2013 | Volume 8 | Issue 9 | e75008Proteoglycan Administration Alters Gut MicrobiotaphylumVerrucomicrobiawas found in only group A and groupB. To address the effect of salmon cartilage PG on intestinalmicrobiota, bacterial phylotypes whose population levelshowed relative alteration at least in 4 of 5 groups wereinvestigated. As shown in Figure 2 and Table S2-Table S3, thepopulation of phylumActinobacteriaand unclassified phylum inthe small intestine and the population of phylumFirmicutesinthe large intestine showed relative increasing upon PGadministration. On the other hand, the population of phylumFirmicutesin the small intestine and the population ofunclassified phylum in the large intestine showed negativeassociation with PG. We further analyzed the alterations in thepopulation levels of each taxon. Distribution of all intestinalphylotypes in each phylum is given in Materials S1. Obviously,the population of classBacilliincreased in both small and largeintestine. On the other hand, the population of classClostridiadecreased in the large intestine (Table S4-Table S5).autistic children suffering from gastric intestinal ailments [20].They probably associate with the disease.Population level of only 4 phylotypes including adisease-associated bacterium decreased in the smallintestine upon PG administration.At least in 4 groups of mice, there were 6 phylotypes whosepopulation level relatively decreased in the small intestine uponPG administration (Table S7) and only 4 phylotypes showing astatistical difference in population between PG-administeredand control mice (P values less than 0.05) at least in 2 groups(Figure 4). Among these populations,Clostridiumsp. ID4 isonly one phylotype belonging to classClostridiaphylumFirmicutes.The biological properties and species of this strainhave not yet been characterized. Two phylotypes in thispopulation belong to genusLactobacillus.They are probioticLactobacillus reuteri[21] and folate-producingLactobacillussp.ID9203 (referred to 16S rRNA gene GenBank: AY862434).Gordonibacter pamelaeaein phylumActinobacteriahas beenshown to associate with diseases. It has been isolated frompatients with Crohn’s disease and rectosigmoid carcinoma[22,23].In the small intestine, the population level ofsaccharolytic bacteria,Lactobacillus intestinalisandbacteria exhibiting immunomodulatory effect increasedupon oral administration of PG.There were 30 phylotypes whose population level increasedin the small intestine upon PG administration, relating at leastin 4 groups of mice (Table S6). After statistical analysis, therewere 24 phylotypes showing a statistical difference inpopulation between PG-administered and control mice (Pvalues less than 0.05) at least in 2 groups (Table S6 andFigure 3). Among these populations, majority of phylotypeshave not yet been characterized and classified, although theyare found as normal flora in humans or animals. They wereBacteroidessp. SLC1-38,Clostridiumsp. Clone-17,Clostridiumsp. Clone-25,Clostridiumsp. Clone-44,Clostridiumsp. Clone-9,Clostridiumsp. Culture Jar-13,Clostridiumsp.Culture-41, Lachnospiraceae bacterium 607, LachnospiraceaebacteriumDJF_VP30,Ruminococcussp.CO28,Ruminococcussp. CO41, Clostridiales bacterium oral taxon085, Gram-negative bacterium cL10-2b-4, human intestinalfirmicute CJ6 and TM7 phylum sp. oral taxon 351. Obviously,several phylotypes whose population level increased upon PGadministration have been shown to provide a health benefits onthe hosts.Alistipes putredinis[11] andClostridiumsaccharolyticum[12] are able to utilize carbohydrates orpolysaccharides as carbon source. These bacteria are able toproduce acetate, propionate and butyrate as fermentedproducts.Adlercreutzia equolifacienshas ability to produceequol [13], a nonsteroidal estrogen which has beneficial effectson the incidence of prostate cancer [14] and reducesexperimental cutaneous inflammation in mice [15].Lactobacillus intestinalisin classBacilli,phylumFirmicutesis apotential probiotic strain [16].Anaerovorax odorimutansis aputrescine-fermenting bacterium which changes putrescine toacetate, butyrate, and ammonia as metabolic products [17].Lachnospiraceae bacterium A4 is negatively associated withdevelopment of colitis [18]. Excluding the phylotypesmentioned above,Enterorhabdus caecimurishas been isolatedfrom mice with spontaneous colitis [19].Clostridium bolteaehas been shown to be overabundant in the intestinal tract ofPrevalence proportion of probiotic lactobacilli andbacteria with immunomodulatory effect was enhancedin the large intestine upon PG administration.At least in 4 groups of mice, there were 16 phylotypes whosepopulation level relatively increased in the large intestine uponPG administration (Table S8). Only 11 phylotypes had astatistical difference in population between PG-administeredand control mice (P values less than 0.05) at least in 2 groups(Table S8 and Figure 5). The role ofPrevotellasp. canine oraltaxon 298,Desulfotomaculumsp. CYP1 andRuminococcussp. M10 in host immune response is unknown.Odoribactersplanchnicusis a butyric acid-producing bacterium which isreclassifiedfromBacteroidessplanchnicus[24,25].Parabacteroides distasonisis a bacterium which has beenshown to attenuate experimental murine colitis throughmodulation of immunity [26].Lactobacillus intestinalis[16],Lactobacillus johnsonii[27] andLactobacillus reuteri[21] are acharacterized by various probiotic properties.Clostridiumfusiformisis a strain of gut barrier flora against pathogen,Clostridium difficilein mouse model (referred to 16S rRNAgene GenBank: AF028349).Roseburia intestinalisissaccharolytic, butyrate producing bacterium [28]. Itcorresponds to a reduction of plasma interleukin-6 [29].Besides these bacteria, only one phylotype probablyassociates with disease that isClostridium bolteae.Asmentioned above, overabundance of this bacterium is found inthe intestinal tract of autistic children suffering from gastricintestinal ailments [20].Population level of saccharolytic clostridia anddisease-associated bacteria decreased in the largeintestine of PG-administered mice.At least in 4 groups of mice, there were 42 phylotypes whosepopulation level decreased in the large intestine upon PGadministration (Table S9). Only 30 phylotypes had a statisticalPLOS ONE | www.plosone.org3September 2013 | Volume 8 | Issue 9 | e75008Proteoglycan Administration Alters Gut MicrobiotaFigure 1. Total counts (A) and number of assigned phylotypes (B) obtained from each intestinal sample.(A) Total countswere obtained after removing the sequences under threshold of 350 nucleotides in length and 85.0% identity. (B) Bacterialphylotypes from each sample were assigned from the closest related 16S rRNA sequence. The number of counts and the numberof bacterial phylotypes in each phylum were indicated. Sequences that could not be classified into known phyla were assigned as‘Unclassified’. A, B, C, D and E indicate the results obtained from group A, group B, group C, group D and group E mice,respectively.doi: 10.1371/journal.pone.0075008.g001difference in population between PG-administered and controlmice (P values less than 0.05) at least in 2 groups (Table S9and Figure 6). Among these populations, major phylotypes (27phylotypes) are members in classClostridiaphylumFirmicutes.PLOS ONE | www.plosone.org4September 2013 | Volume 8 | Issue 9 | e75008Proteoglycan Administration Alters Gut MicrobiotaFigure 2. Relative abundance at phylum-level of GIT microbiota in PBS- and PG-administered mice.Relative count of eachphylum was calculated from counts of specific phylum and total counts of each sample. Sequences that could not be classified intoany known bacterial phyla were assigned as ‘Unclassified’. A, B, C, D and E indicate the results obtained from group A, group B,group C, group D and group E mice, respectively.doi: 10.1371/journal.pone.0075008.g002The role in GIT immunity of the predominant phylotypesincludingBarnesiellasp. NSB1,Clostridiumsp. Clone-strains,Clostridiumsp.Culture-strains,Eubacteriumcoprostanoligenes,Lachnospiraceae bacterium DJF_VP30,Oscillibacterspecies, Peptostreptococcaceae bacterium canineoral taxon 125 and 221, and Gram-negative bacteriumcL10-2b-4 is unclear. Obviously, the proportion of 5 phylotypesalso increased in the small intestine upon PG administration.They areClostridiumsp. Clone-17,Clostridiumsp. Clone-25,Clostridiumsp. Clone-44, Lachnospiraceae bacteriumDJF_VP30 and Gram-negative bacterium cL10-2b-4 (Figure 3).Among these 30 phylotypes, one phylotype belongs tophylumProteobacteriathatisParasutterellaexcrementihominis. Parasutterella excrementihominisis amember in classBetaproteobacteria.It is a strictly anaerobic,non-spore-forming, Gram-negative coccobacillus found inhuman feces. Biochemically, this strain is largely unreactive aswell as asaccharolytic [30]. On the other hand, 5 strains whosepopulation level decreased upon PG administration are able todigest saccharides and/or produce butyrate. They areClostridium xylanolyticum[31],Eubacterium plexicaudatum[32],Ruminococcus flavefaciens[33], Clostridiales bacterium21-4c [34] and Clostridiales bacterium 37-2a [34].Interestingly, 4 phylotypes whose population level decreasedin the large intestine have been shown to be associated withdiseases.Clostridium hathewayicauses bacteremia andsepticemia [35,36]. Multiple studies have identified thatRuminococcus gnavusis enriched-bowel inflammatory disease[37].Clostridiumsp. A9 and Lachnospiraceae bacterium 6-1are flagellated bacteria which implicate in Crohn’s disease(referred to 16S rRNA gene GenBank:DQ789124 and DQ789123, respectively).DQ789119.1,DiscussionIntestinal microbiota constitutively and profoundly affects thedevelopment and balance of the host immune system. Theyhave been implicated in prevention of damage induced byopportunistic microbes, in repair of damage to mucosal barrierand in influencing systemic autoimmune diseases. Thecomposition of a host’s intestinal microbiota directs the type ofmucosal and systemic immune response by affecting theproportion and number of functionally distinct T cells subsets.In particular, the microbiota composition affects thedifferentiation of intestinal Th17 cells and Treg cells, both ofwhich play crucial roles in maintaining mucosal barrier offunctions and in controlling immunological homeostasis [38].We have previously reported that salmon cartilage PG has apotential to be a prophylactic agent because it exhibitedimmunomodulatory activity by suppressing systemic hostinflammation in colitis and autoimmune encephalomyelitis [3,4].In both models, Th17 cells are important for progression of thediseases [39,40]. From the chemokine and cytokine productionas well as T cell analysis, we found that salmon cartilage PGsuppresses Th17 linage differentiation and enhances of Tregexpansion [3,4]. To find the link between oral administration ofsalmon PG and regulation of these T cells, the composition ofGIT microbiota in PG-administered mice was investigated inthis study.The bar-coded pyrosequencing technique is culture-independent molecular tool that is widely used to analyze thePLOS ONE | www.plosone.org5September 2013 | Volume 8 | Issue 9 | e75008 [ Pobierz całość w formacie PDF ]