Hostname: page-component-586b7cd67f-tf8b9 Total loading time: 0 Render date: 2024-12-03T19:20:36.089Z Has data issue: false hasContentIssue false

The effect of age and non-steroidal anti-inflammatory drugs on human intestinal microbiota composition

Published online by Cambridge University Press:  25 August 2009

Harri Mäkivuokko
Affiliation:
Danisco Finland Oy, Sokeritehtaantie 20, FIN-02460Kantvik, Finland Finnish Red Cross Blood Service, Research and Development, Kivihaantie 7, FIN-00310Helsinki, Finland
Kirsti Tiihonen
Affiliation:
Danisco Finland Oy, Sokeritehtaantie 20, FIN-02460Kantvik, Finland
Soile Tynkkynen
Affiliation:
Valio Ltd, Research and Development, Meijeritie 4, FIN-00370Helsinki, Finland
Lars Paulin
Affiliation:
Institute of Biotechnology, University of Helsinki, Viikinkaari 9, FIN-00014Helsinki, Finland
Nina Rautonen*
Affiliation:
Danisco Finland Oy, Sokeritehtaantie 20, FIN-02460Kantvik, Finland
*
*Corresponding author: Dr Nina Rautonen, fax +358 9 298 2203, email [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Ageing has been suggested to cause changes in the intestinal microbial community. In the present study, the microbiota of a previously well-defined group of elderly subjects aged between 70 and 85 years, both non-steroidal anti-inflammatory drugs (NSAID) users (n 9) and non-users (n 9), were further compared with young adults (n 14) with a mean age of 28 years, by two DNA-based techniques: percentage guanine+cytosine (%G+C) profiling and 16S rDNA sequencing. Remarkable changes in microbiota were described with both methods: compared with young adults a significant reduction in overall numbers of microbes in both elderly groups was measured. Moreover, the total number of microbes in elderly NSAID users was higher than in elderly without NSAID. In 16S rDNA sequencing, shifts in all major microbial phyla, such as lower numbers of Firmicutes and an increase in numbers of Bacteroidetes in the elderly were monitored. On the genus level an interesting link between reductions in the proportion of known butyrate producers belonging to Clostridium cluster XIVa, such as Roseburia and Ruminococcus, could be demonstrated in the elderly. Moreover, in the Actinobacteria group, lower numbers of Collinsella spp. were evident in the elderly subjects with NSAID compared both with young adults and the elderly without NSAID, suggesting that the use of NSAID along with age may also influence the composition of intestinal microbiota. Furthermore, relatively high numbers of Lactobacillus appeared only in the elderly subjects without NSAID. In general, the lowered numbers of microbial members in the major phyla, Firmicutes, together with changes in the epithelial layer functions can have a significant effect on the colon health of the elderly.

Type
Full Papers
Copyright
Copyright © The Authors 2009

The human gastrointestinal tract harbours a highly diverse population of commensal microbes, estimated as 1014 bacterial cells, participating in the breakdown of indigested food(Reference Hooper, Midtvedt and Gordon1, Reference Mueller, Saunier and Hanisch2). The composition of the microbiota develops during the first years of life and remains relatively stable throughout adulthood(Reference Ouwehand, Isolauri and Salminen3). Interestingly, after birth the maturation of immune responses reflects the microbial colonisation in the intestine(Reference Bauer, Williams and Smidt4, Reference Rautava and Walker5). It has been suggested that towards seniority specific changes in the microbial community(Reference Benno, Endo and Mizutani6) and immune defence occur again(Reference Gill, Darragh and Cross7). In the elderly, immune responses are weakened, predisposing seniors to a higher incidence of morbidity and mortality by infectious diseases(Reference Kumar and Burns8, Reference van Duin and Shaw9). Furthermore, the common use of non-steroidal anti-inflammatory drugs (NSAID) among the elderly is known to increase gastrointestinal ailments(Reference Bjarnason, Hayllar and MacPherson10Reference Laine, Connors and Reicin12). However, the elderly immune responses can be enhanced by supplementing the diet with probiotics, implying that not only in children are the immune responses modified by intestinal bacteria(Reference Gill, Rutherfurd and Cross13). With improved life expectancy the share of the population over the age of 60 years is estimated to reach 2 billion by 2050(Reference Cohen14). In order to improve the quality of life of the growing elderly population and decrease health-care costs, the role of intestinal microbes in health necessitates deeper understanding.

Age- and country-specific differences in intestinal microbiota have been reported by several studies. Most commonly, decreasing numbers of bifidobacteria and increasing numbers of enterobacteria have been suggested(Reference Mueller, Saunier and Hanisch2, Reference Benno, Endo and Mizutani6, Reference Hayashi, Sakamoto and Kitahara15, Reference Bartosch, Fite and Macfarlane16) in the elderly. Some reports indicate changes also in the Lactobacillus and Bacteroides groups(Reference Tiihonen, Tynkkynen and Ouwehand17). Recently, also a reduced microbial activity has been described in elderly subjects(Reference Tiihonen, Tynkkynen and Ouwehand17, Reference Woodmansey, McMurdo and Macfarlane18). It is noteworthy that changes in the microbial composition have not only been linked to age but also to a condition with multiple health risk factors, namely obesity(Reference Turnbaugh, Ley and Mahowald19), and several gastrointestinal diseases such as Crohn's disease, irritable bowel disease and irritable bowel syndrome (IBS)(Reference Malinen, Rinttilä and Kajander20, Reference Kassinen, Krogius-Kurikka and Makivuokko21). This underlies the importance of increasing knowledge of the intestinal microbial composition.

The availability of microbial DNA sequences has revealed that a great proportion of the gastrointestinal microbes belong to unknown and uncultivable genera and species(Reference Hayashi, Sakamoto and Benno22). Thus it is important to include DNA-based methods that are independent of prior knowledge of bacteria when evaluating the overall changes in the microbial communities. The percentage guanine+cytosine (%G+C) profiling technique has been previously used to reveal changes in the microbial community structure(Reference Malinen, Rinttilä and Kajander20, Reference Kassinen, Krogius-Kurikka and Makivuokko21, Reference Apajalahti, Kettunen and Kettunen23Reference Mäkivuokko, Saarinen and Ouwehand25). The technique exploits the diverging weight of microbial genomes due to different proportions of guanidine and cytosine bases in their genomes, and thus needs no prior information of specific sequences, allowing analysis of previously uncultivated microbes. The 16S rDNA sequencing technique, however, requires some prior information of the microbes present in the sample, since the broad-range primers used by the technique are common for known cultivated bacteria(Reference Lane, Pace and Olsen26). The combination of these two methods can provide useful information from samples with mixed populations containing hundreds of different species. In fact, the two methods have been used successfully previously in describing differences in the microbiota between IBS patients and healthy subjects(Reference Kassinen, Krogius-Kurikka and Makivuokko21).

In the present study we have characterised the intestinal microbiota of a subgroup of previously well-defined elderly subjects, with or without NSAID, and young adults(Reference Tiihonen, Tynkkynen and Ouwehand17) using %G+C profiling and 16S rDNA sequencing techniques in order to expose differences caused by age and NSAID use.

Materials and methods

Study subjects

A total of nine elderly NSAID users (mean age 80 (range 77–85) years) and nine elderly non-users (mean age 78 (range 70–83) years) living in a nursing home and fourteen young adults (mean age 28 (range 21–39) years) were recruited to the study in the city of Kuopio, Finland. The number of study subjects per group was originally planned to be fifteen but due to an unexpected high number of drop-outs, such as use of antibiotics, or diarrhoea, the number of especially elderly study subjects decreased. Significant differences in microbial composition and metabolites were found for the two elderly groups(Reference Tiihonen, Tynkkynen and Ouwehand17), prompting the separate analysis of the sequence data for the elderly NSAID users and non-users. The elderly were defined as NSAID users when they used the medication three or more times per week. The NSAID used included indomethacin, ibuprofen, ketoprofen, diclofenac, piroxicam, celecoxib, nimesulid and aspirin. The exclusion criteria were critical illnesses, inflammatory bowel disease, coeliac diseases and major malignancies in the gastrointestinal tract. Subjects using antibiotics during the study or within 2 months before the screening were excluded. The use of probiotic products was not allowed during the study. The inclusion and exclusion criteria were monitored during the run-in period before randomisation.

Throughout the sampling both the elderly and the young study subjects followed their habitual diet and most of them used the uniform meal service provided at least once per d. The calculated dietary intakes of the elderly living in the nursing home, based on the diets provided, were: energy, 6·17 MJ/d (6·5 MJ/d fibre included); protein, 66 g/d (18 % energy); fat, 52 g/d (32 % energy); cholesterol, 146 mg/d; carbohydrates, 181 g/d (50 % energy); dietary fibres, 20 g/d.

The present study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Research Ethics Committee of the Hospital Districts of Northern Savo, Finland. The purpose of the study was explained to the participants, and written consent to the study was given by all of the subjects. Information on the age, sex, bowel function, diseases, allergies, current medication and pre-trial use of probiotics of the study subjects was obtained at the beginning of the study. The pre-trial characteristics of the study groups are presented in Table 1.

Table 1 Sex, age, bowel function and medication of the thirty-two study subjects

NSAID, non-steroidal anti-inflammatory drugs.

* Several drugs can have gastrointestinal side effects. No side effects by the medications used by study subjects were observed.

Macrocol, sodium bicarbonate, sodium chloride, potassium chloride, sodium sulfate, lactulose, sodium picosulfate (common (>1/100) gastrointestinal side effects: diarrhoea; others – flatus).

Enalapril, bisoprolol, spironolactone, losartan, aspirin, warfarin, metoprolol, dipyridamole+aspirin, bisoprolol, clopidogrel, dipyridamole, dipyridamole, furosemide, isosorbide dinitrate, indapapamide, propranolol (common (>1/100) gastrointestinal side effects: diarrhoea; others – flatus, vomiting, gastrointestinal bleeding).

§ Magnesium hydroxide, finasteride, ranitidine, macrocol, sodium bicarbonate, sodium chloride, potassium chloride, sodium sulfate, lactulose, sodium picosulfate (common (>1/100) side effects: diarrhoea, constipation; others – flatus).

Olanzapine, mirtazapine, oxazepam, risperidone, melperone, thioridazine, chlorprothixene, temazepam, zopiclone, quetiapine, citalopram (very common (>1/10) gastrointestinal side effects: constipation; others – flatus, diarrhoea, vomiting).

Galantamine, oxcarbazepine, levodopa+benserazide, entacapone, pramipexole, donepezil, memantine, sodium valporate, memantine, rivastigmine (very common (>1/10) gastrointestinal side effects: vomiting, diarrhoea; others – constipation).

Faecal samples

Spot samples of faeces were obtained three times from each subject over 3 consecutive weeks. Subjects were personally instructed to take the faecal samples according to the written protocol. All the equipment and documentation papers were provided to the subjects before sampling. All faecal samples were initially stored at − 20°C, and, transferred to the laboratory within 12 h of defecation and stored at − 70°C until analysed. The three samples obtained from each study subject were pooled to result in a representative individual sample.

Percentage guanine+cytosine profiling

Total genomic DNA of faecal microbes was extracted as previously described(Reference Tiihonen, Tynkkynen and Ouwehand17). In short, 1 g of faecal material was washed, microbes were collected and freeze-thawed to loosen their cell membranes, after which the microbes were lysed, the cell debris was removed and genomic DNA was precipitated.

The composition of microbiota in the faecal samples was monitored by %G+C profiling(Reference Apajalahti, Sarkilahti and Maki24). The method enables fractionation of total microbial DNA into bacterial subpopulations by their genomic G+C content. For fractioning, three individual DNA samples were pooled for each elderly donor pool (five pools in total; three pools of elderly without NSAID and two with NSAID) and four or five individual DNA samples in each of the young adult donor pools (three pools in total). One pool of three subjects for the elderly with NSAID group had to be removed from the %G+C-profile analysis due to breakdown of genomic DNA into smaller fragments during the extraction phase, making %G+C-profiling of that particular pool inaccurate compared with pools with intact DNA. However, this fragmented pool was included in the sequencing procedure, in which only intact 16S rDNA was needed. Each of the pools contained 300 μg genomic DNA.

Microbial enumeration

The total bacteria cell counts in faecal samples were determined by flow cytometry (FACSCalibur; Becton Dickinson, Franklin Lakes, NJ, USA) as described previously(Reference Apajalahti, Sarkilahti and Maki24). The bacterial fractions were recovered by suspending faecal samples in a buffer, followed by centrifugation and washing. The cell samples were diluted, fixed and stained with the fluorescent nucleic acid-binding dye SYTO 24 (Molecular Probes, Leiden, The Netherlands). The results are expressed as the quantity of bacteria/g faeces.

Sequencing procedure

For sequencing, six fractions containing %G+C areas of 20–30, 30–40, 40–50, 50–60, 60–70 and 70–80 %G+C were collected from the pools. All corresponding %G+C fractions in one study group (i.e. young adults, elderly and elderly with NSAID) were pooled, resulting in six %G+C fractions for each study group, and totally eighteen fractions. The DNA of all the eighteen fractions were used as templates to amplify 900 bp 16S rDNA regions for sequencing. The PCR amplification was performed using a Dynazyme II polymerase kit (Finnzymes, Espoo, Finland) according to the manufacturer's instructions and broad-range primers pA 5′-AGAGTTTGATCCTGGCTCAG-3′ and pE′ 5′-CCGTCAATTCCTTTGAGTTT-3′(Reference Lane, Pace and Olsen26, Reference Edwards, Rogall and Blocker27) corresponding to the 8–28 and 928–908 positions in the Escherichia coli 16S rDNA sequence. Amplification products were purified with Sephacryl S-400 colon (Amersham Biosciences, Little Chalfont, Bucks, UK) and cloned in pDrive cloning vector (Qiagen, Hilden, Germany) in E. coli. Colonies were picked and sequenced from both directions with T7 and SP6 primers using BigDye Chemistry and analysed on an ABI 3730 Capillary DNA Sequencer (Applied Biosystems, Foster City, CA, USA). The obtained sequences were assembled and edited using Gap4 from the Staden package(Reference Staden, Beal and Bonfield28).

Spermidine analysis

The spermidine concentrations in faeces were analysed using the method described by Saarinen(Reference Saarinen29).

Phylogenetic and statistical analysis of sequence data

%G+C fractions were analysed with ANOVA applying linear mixed models(Reference Pinheiro and Bates30) inside each fraction, and then computing post hoc t tests whenever a significant difference between the groups was found. Spermidine data were analysed with ANOVA applying a linear model and post hoc t tests. The sequences were phylogenetically aligned with the Ribosomal Database Project (RDP) II My RDP on-line-tool (https://rdp.cme.msu.edu/login/myrdp) and statistically compared using the RDP II Library Compare Tool Analysis on-line-tool (http://rdp.cme.msu.edu/seqmatch/) with a 95 % confidence threshold for genus determination, which uses a RDP-naive Bayesian classifier to provide rapid classification of library sequences into the bacterial taxonomy. The equality of the sequence proportions between the groups in Fig. 1 was tested with a standard multi-population proportions test whenever all the frequencies were larger than five, and if statistically significant non-equality was found, the proportions were analysed further with a pair-wise proportions test. If the proportions were smaller than five, a special proportions test for small counts was applied as explained in Wang et al. (Reference Wang, Garrity and Tiedje31), and first introduced for digital gene expression profiles in Audic & Claverie(Reference Audic and Claverie32). The comparison analysis enabled the calculation of the most abundant strain of each genus using a similarity score or a sequence match score (S_ab) comparison reporting the number of (unique) seven-base oligomers shared between an input sequence and a given RDP sequence divided by the lowest number of unique oligos in either of the two sequences. In addition to RDP software (Center for Microbial Ecology at Michigan State University, East Lansing, MI, USA), the statistical programming language R was used to perform analyses (http://www.r-project.org/).

Fig. 1 Numbers of partial 16S rRNA gene sequences obtained from the three different study groups organised into bacterial phyla by the Ribosomal Database Project (RDP) II My RDP on-line-tool (https://rdp.cme.msu.edu/login/myrdp). ↔ , Statistical significance determined by standard proportions test (P < 0·001) between two study groups. NSAID, non-steroidal anti-inflammatory drugs.

Results

The total microbial numbers and the corresponding %G+C-profiles were compared to reveal the proportional differences in the total microbiota profiles of the three study groups (Fig. 2). The young adult group had statistically higher numbers of microbes in all cumulated 10 %G+C-fractions between the 30–69 %G+C area when compared with either of the elderly groups (P < 0·001 and P < 0·0003, respectively, in each 5 % fraction) and in the 20–29 %G+C area when compared with the elderly group without NSAID (P < 0·006). The two elderly groups differed from each other in the 20–59 %G+C area (P < 0·0002, respectively, in each 5 % fraction), the microbial numbers being higher in the elderly group with NSAID.

Fig. 2 Average number of microbes in the percentage guanine+cytosine (%G+C) fractions in the three study groups: elderly subjects without non-steroidal anti-inflammatory drugs (NSAID) (); elderly subjects with NSAID (); young adults (□). Due to problems in DNA extraction (see Materials and methods), the final number of subjects included in the DNA pool of the elderly with NSAID was six, and nine from the elderly without NSAID. Statistically significant differences were determined by ANOVA and post hoc t tests. * Mean value was significantly different from those of the two elderly groups (P < 0·01). † Mean value was significantly different from that of the elderly group without NSAID (P < 0·01).

A total of 2112 high-quality sequences were obtained by sequencing of the six fractioned DNA samples of each study group. The total sums of sequences obtained were 803 sequences in the young adults, 702 sequences in the elderly without NSAID and 610 sequences in the elderly with NSAID. In Fig. 1 these aligned sequences are divided into six phyla and fifty-one presently known microbial genera; the bacterial genera differing significantly (P < 0·001) between the study groups are marked with arrows. The majority of the microbes in each study group were found to be associated to the Firmicutes phylum followed by the Bacteroidetes and Actinobacteria phyla. Comparison of young adults with either of the elderly groups revealed more differences in the number of genus-specific sequences than the comparison of the two elderly groups. The young adults exhibited higher faecal numbers of Rumicococcus, Roseburia, Coprobacillus and Dialister spp., all belonging to Firmicutes, than the elderly without NSAID, suggesting an age-dependent shift in genera. Lactobacillus, Streptococcus and Bacteroides were present in higher numbers in the elderly. Compared with the elderly NSAID users, the samples obtained from the young adults exhibited more Dialister, Coprobacillus and Collinsella. The significant differences between the elderly groups were the higher amounts of Lactobacillus and Collinsella in the elderly without NSAID and Roseburia in the NSAID users.

In order to indicate the degree of match of study sequences to each named phylotype in the RDP-II database (http://rdp.cme.msu.edu/), S_ab-scores were calculated on single species in each genus with statistical significance between study groups present in Fig. 1; species with S_ab-scores higher than 0·97 were considered significant. The S_ab-scores for the Ruminococcus and Coprobacillus genera had no significant matches, while the Roseburia, Lactobacillus, Streptococcus, Dialister, Bacteroides and Collinsella genera each had one or two matching strains with the RDP-II database; the most abundant bacterial species were, respectively, Roseburia intestinalis, Lactobacillus ruminis, L. sobrius, Streptococcus lutiensis, Dialister invisus, Bacteroides coprocola, B. dorei and Collinsella aerofaciens.

The concentrations of spermidine in the faeces of the elderly (median 76·7 and 67·6 nmol/g faeces, respectively, in the elderly without NSAID and with NSAID) were monitored to be significantly lower when compared with the concentrations in samples obtained from the younger adults (median 160·4 nmol/g faeces; P = 0·004 and P = 0·007, respectively).

Discussion

In the present study we characterised the intestinal microbiota of a subgroup of previously well-defined elderly subjects, with or without NSAID drugs, and young adults(Reference Tiihonen, Tynkkynen and Ouwehand17) using %G+C profiling and 16S rDNA sequencing techniques in order to expose differences caused by age and NSAID use. Our previous study(Reference Tiihonen, Tynkkynen and Ouwehand17) focused mainly on the microbial metabolism and intestinal immunological status but analysed also differences in faecal microbiota. The study showed that specific immune responses as well as generally the microbial metabolism were decreased, and furthermore that, for example, the proportion of facultative anaerobic microbes was increased in the elderly.

The %G+C-profiling technique has recently been used as a tool for describing the total microbial community in human(Reference Kassinen, Krogius-Kurikka and Makivuokko21, Reference Mäkivuokko, Saarinen and Ouwehand25), mouse(Reference Apajalahti, Kettunen and Kettunen23) and piglet(Reference Fava, Makivuokko and Siljander-Rasi33) intestines. The %G+C-profiling enables microbiota analysis that includes previously unknown or presently unculturable microbial genera since it exploits only the different proportion of guanidine and cytosine bases in their genomes. Moreover, %G+C fractioning of microbiota DNA before 16S rRNA gene sequencing enhances the possibility of less abundant microbial species to be revealed in specific fractions.

The obtained %G+C-profiles support the sequencing data well by depicting that young adults have the highest numbers of microbes especially in the Firmicutes-rich %G+C area (low %G+C bacteria). In the young adults 87 % of sequences were classified to Firmicutes. Similarly, in the elderly groups, 78–82 % of the sequences grouped to this phylum; however, the significantly higher number of bacteria in the NSAID users shown in %G+C profiling did not result in higher numbers of sequences. Also in the high %G+C areas (>50 %G+C), a difference between the %G+C and sequencing data can be noted. In the elderly NSAID users the 50–59 %G+C area was significantly higher compared with that of the non-users, but our sequencing data do not reveal an increase in bacterial species. As %G+C profiling is a totally culture- and probe-independent analysis method portraying only the concentration differences of the microbial genomic DNA, this discrepancy is most probably due to higher numbers of bacteria belonging to the same genera or problems in amplifying high-G+C sequences, which affected especially the Bifidobacterium genus. One explanation for the lack of bifidobacteria in the present study is due to the primer used in PCR. The pE′ primer 5′-CCGTCAATTCCTTTGAGTTT-3′ has a mismatch C/T (highlighted in bold) in the middle. This mismatch can abolish the amplification of bifidobacteria 16S rRNA genes. Others have also reported the lack of bifidobacteria in 16S rRNA libraries, possibly due to mismatches in primers or imperfect DNA denaturation during PCR(Reference Hayashi, Sakamoto and Kitahara15, Reference Suau, Bonnet and Sutren34).

The %G+C profiling showed that the young adults had statistically higher numbers of microbes in the %G+C area of Firmicutes, Bacteroidetes and Actinobacteria, the major phyla reported in the human intestine, compared with either elderly group. The most prominent difference dependent on age, however, was a clear reduction in numbers of Ruminococcus and Roseburia spp., the known butyrate producers in Clostridium cluster XIVa, in the elderly. This is consistent with our earlier studies showing a decrease in the Clostridium coccoidesEubacterium rectale group in the elderly(Reference Tiihonen, Tynkkynen and Ouwehand17). Similarly, the elderly with NSAID had higher numbers of bacteria in most %G+C fractions compared with the non-users. Therefore, both ageing and the use of specific medication can deeply alter the abundance and composition of the microbial community. The sequencing revealed higher amounts of lactobacilli in the elderly groups compared with the young adults, which is in accordance with our earlier findings(Reference Tiihonen, Tynkkynen and Ouwehand17).

The major anaerobic microbes in the gastrointestinal tract are Bacteroides species. These organisms can function as symbiotic partners in the host physiology or they can cause serious diseases(Reference Wexler35). Recent studies have indicated the link between levels of the Bacteroides and Firmicutes and obesity(Reference Ley, Backhed and Turnbaugh36Reference Turnbaugh, Backhed and Fulton38). Ley et al. (Reference Ley, Backhed and Turnbaugh36) showed that obesity alters gut microbiota by increasing the relative abundance of Firmicutes over Bacteroidetes. Evidently in mice with diet-induced obesity the Western diet nutrient milieu provides a competitive advantage to the Firmicutes(Reference Turnbaugh, Backhed and Fulton38).

Interestingly, also in the present study, major shifts in the microbiota occurred in the levels of Bacteroidetes and Firmicutes, suggesting perhaps differences in the diet of the elderly and young adults and also in the physiology of the intestinal tract. This is consistent with our earlier finding of a higher proportion of anaerobic bacteria in faecal samples from elderly subjects(Reference Tiihonen, Tynkkynen and Ouwehand17, Reference Woodmansey, McMurdo and Macfarlane18). Furthermore, some of the major differences between the microbiota of healthy elderly and young adults reported previously in German(Reference Mueller, Saunier and Hanisch2) and Japanese populations(Reference Hayashi, Sakamoto and Kitahara15) correlate with those presented in the present study, that is, a higher proportion of Bacteroides in the elderly and a higher proportion of Ruminococcus in the young adults. However, Italian elderly were reported to have lower levels of Bacteroides when compared with young adults(Reference Mueller, Saunier and Hanisch2). These results indicate strong country specificity in diversity of the elderly microbiota.

The metabolic activity of the intestinal microbes has been reported by us to be decreased in elderly subjects(Reference Tiihonen, Tynkkynen and Ouwehand17, Reference Woodmansey, McMurdo and Macfarlane18). Changes in the microbial community structure described in the present study may offer some explanations or hypotheses to the detected differences. R. intestinalis and Eubacterium rectale are known butyrate-producing species belonging to Firmicutes(Reference Barcenilla, Pryde and Martin39, Reference Duncan, Hold and Barcenilla40). Roseburia sp. can convert acetate to butyrate, which is the major energy source for colonocytes and has been reported to reduce risk for developing colon cancer and colitis(Reference Louis, Scott and Duncan41). During a low-carbohydrate diet, the levels of Roseburia and Eubacterium rectale decrease, causing a drop in butyrate concentrations(Reference Duncan, Belenguer and Holtrop42), which emphasises the importance of the diet. In the present study, reduction of Roseburia was described in the elderly, and it may partially explain the described reduced concentrations of butyrate in elderly subjects compared with young adults as reported by Tiihonen et al. (Reference Tiihonen, Tynkkynen and Ouwehand17). C. aerofaciens is abundantly found in the human intestine(Reference Kageyama, Sakamoto and Benno43). Kassinen et al. (Reference Kassinen, Krogius-Kurikka and Makivuokko21) reported that the amount of Collinsella in IBS patients was significantly reduced. In the present study the number of C. aerofaciens was reduced in the elderly subjects with NSAID, while elderly subjects without NSAID and young adults had high numbers of C. aerofaciens. The use of NSAID in the study with IBS patients was not recorded but due to common gastrointestinal discomfort and pain in subjects with IBS, NSAID may have been used by the subjects.

Ruminococci are abundant in the human gastrointestinal tract, comprising 2·5 % of total bacteria(Reference Zoetendal, Ben-Amor and Harmsen44). These Clostridium coccoides group XIVa-belonging bacteria are cellulolytic, found on solid particles in faeces(Reference Walker, Duncan and Harmsen45) and have been linked to acetate and butyrate production(Reference Duncan, Holtrop and Lobley46). Furthermore, R. intestinalis has been suggested to be a polyamine producer(Reference Matsumoto and Benno47). Putrescine, spermidine and spermine are important for cell proliferation and differentiation and mucosal barrier function (for reviews, see Thomas & Thomas(Reference Thomas and Thomas48) and Larque et al. (Reference Larque, Sabater-Molina and Zamora49)). In the present study we found that the concentration of spermidine in the faecal samples was decreased in the elderly individuals when compared with the young adults. The reduced production of this polyamine, which is essential for the maintenance and function of rapidly renewing gastrointestinal epithelia(Reference Wang, Feith and Welsh50, Reference Seiler and Raul51), can most probably alter the structure of the epithelial layer, thus affecting the adherence of the commensal intestinal microbiota, and therefore reducing the overall microbial numbers in the gut. Lowered microbial numbers, together with an altered epithelial layer surface, can have a significant effect on the digestibility and breakdown of food components entering the colon, enhancing development of constipation and reducing nutrient intake.

In conclusion, further research using a larger set of faecal donors of different ages could confirm the relationship between changes in shares of specific species and the consequences in their activity, especially regarding metabolism of important compounds for colon health, such as butyrate or spermidine. We propose that in future research, measurements of, for example, specifically two prominent intestinal strains, R. intestinalis and C. aerofaciens, should be included as potentially useful biomarkers for monitoring changes in the intestinal conditions.

Acknowledgements

We especially appreciate the help of the organisers of the sampling in the Pohjois-Savo Vocational Institute in Siilinjärvi and the Harjula Hospital in the District of Northern Savo, and Essi Sarkkinen and Henna Karvonen from Foodfiles Ltd for organising the sampling procedure. Tanja Horko is acknowledged for technical assistance and Janne Nikkilä from Genolyze Ltd for statistical analyses. The present study was supported by the National Technology Agency of Finland (TEKES), grant no. 129/31/03.

H. M. was responsible for the design of the study, performing %G+C-analyses, sequence comparisons, interpretation of the results and writing the manuscript. K. T. was responsible for the design of the study, managing the clinical trial, interpretation of the results and writing the manuscript. S. T. was responsible for the design of the study, interpretation of the results and writing the manuscript. L. P. was responsible for the sequencing procedure, interpretation of results and writing the manuscript. N. R. was responsible for the design of the study, interpretation of the results and writing the manuscript.

There are no conflicts of interests.

References

1Hooper, LV, Midtvedt, T & Gordon, JI (2002) How host–microbial interactions shape the nutrient environment of the mammalian intestine. Annu Rev Nutr 22, 283307.CrossRefGoogle ScholarPubMed
2Mueller, S, Saunier, K, Hanisch, C, et al. (2006) Differences in fecal microbiota in different European study populations in relation to age, gender, and country: a cross-sectional study. Appl Environ Microbiol 72, 10271033.CrossRefGoogle ScholarPubMed
3Ouwehand, A, Isolauri, E & Salminen, S (2002) The role of the intestinal microflora for the development of the immune system in early childhood. Eur J Nutr 41, Suppl. 1, I32I37.CrossRefGoogle ScholarPubMed
4Bauer, E, Williams, BA, Smidt, H, et al. (2006) Influence of dietary components on development of the microbiota in single-stomached species. Nutr Res Rev 19, 6378.CrossRefGoogle ScholarPubMed
5Rautava, S & Walker, WA (2007) Commensal bacteria and epithelial cross talk in the developing intestine. Curr Gastroenterol Rep 9, 385392.CrossRefGoogle ScholarPubMed
6Benno, Y, Endo, K, Mizutani, T, et al. (1989) Comparison of fecal microflora of elderly persons in rural and urban areas of Japan. Appl Environ Microbiol 55, 11001105.CrossRefGoogle Scholar
7Gill, HS, Darragh, AJ & Cross, ML (2001) Optimizing immunity and gut function in the elderly. J Nutr Health Aging 5, 8091.Google ScholarPubMed
8Kumar, R & Burns, EA (2008) Age-related decline in immunity: implications for vaccine responsiveness. Expert Rev Vaccines 7, 467479.CrossRefGoogle ScholarPubMed
9van Duin, D & Shaw, AC (2007) Toll-like receptors in older adults. J Am Geriatr Soc 55, 14381444.CrossRefGoogle ScholarPubMed
10Bjarnason, I, Hayllar, J, MacPherson, AJ, et al. (1993) Side effects of nonsteroidal anti-inflammatory drugs on the small and large intestine in humans. Gastroenterology 104, 18321847.CrossRefGoogle ScholarPubMed
11Gabriel, SE, Jaakkimainen, RL & Bombardier, C (1993) The cost-effectiveness of misoprostol for nonsteroidal antiinflammatory drug-associated adverse gastrointestinal events. Arthritis Rheum 36, 447459.CrossRefGoogle ScholarPubMed
12Laine, L, Connors, LG, Reicin, A, et al. (2003) Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use. Gastroenterology 124, 288292.CrossRefGoogle ScholarPubMed
13Gill, HS, Rutherfurd, KJ, Cross, ML, et al. (2001) Enhancement of immunity in the elderly by dietary supplementation with the probiotic Bifidobacterium lactis HN019. Am J Clin Nutr 74, 833839.CrossRefGoogle ScholarPubMed
14Cohen, JE (2003) Human population: the next half century. Science 302, 11721175.CrossRefGoogle ScholarPubMed
15Hayashi, H, Sakamoto, M, Kitahara, M, et al. (2003) Molecular analysis of fecal microbiota in elderly individuals using 16S rDNA library and T-RFLP. Microbiol Immunol 47, 557570.CrossRefGoogle ScholarPubMed
16Bartosch, S, Fite, A, Macfarlane, GT, et al. (2004) Characterization of bacterial communities in feces from healthy elderly volunteers and hospitalized elderly patients by using real-time PCR and effects of antibiotic treatment on the fecal microbiota. Appl Environ Microbiol 70, 35753581.CrossRefGoogle ScholarPubMed
17Tiihonen, K, Tynkkynen, S, Ouwehand, A, et al. (2008) The effect of ageing with and without non-steroidal anti-inflammatory drugs on gastrointestinal microbiology and immunology. Br J Nutr 100, 130137.CrossRefGoogle ScholarPubMed
18Woodmansey, EJ, McMurdo, ME, Macfarlane, GT, et al. (2004) Comparison of compositions and metabolic activities of fecal microbiotas in young adults and in antibiotic-treated and non-antibiotic-treated elderly subjects. Appl Environ Microbiol 70, 61136122.CrossRefGoogle ScholarPubMed
19Turnbaugh, PJ, Ley, RE, Mahowald, MA, et al. (2006) An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 444, 10271031.CrossRefGoogle ScholarPubMed
20Malinen, E, Rinttilä, T, Kajander, K, et al. (2005) Analysis of the fecal microbiota of irritable bowel syndrome patients and healthy controls with real-time PCR. Am J Gastroenterol 100, 373382.CrossRefGoogle ScholarPubMed
21Kassinen, A, Krogius-Kurikka, L, Makivuokko, H, et al. (2007) The fecal microbiota of irritable bowel syndrome patients differs significantly from that of healthy subjects. Gastroenterology 133, 2433.CrossRefGoogle ScholarPubMed
22Hayashi, H, Sakamoto, M & Benno, Y (2002) Phylogenetic analysis of the human gut microbiota using 16S rDNA clone libraries and strictly anaerobic culture-based methods. Microbiol Immunol 46, 535548.CrossRefGoogle ScholarPubMed
23Apajalahti, JH, Kettunen, H, Kettunen, A, et al. (2002) Culture-independent microbial community analysis reveals that inulin in the diet primarily affects previously unknown bacteria in the mouse cecum. Appl Environ Microbiol 68, 49864995.CrossRefGoogle ScholarPubMed
24Apajalahti, JH, Sarkilahti, LK, Maki, BR, et al. (1998) Effective recovery of bacterial DNA and percent-guanine-plus-cytosine-based analysis of community structure in the gastrointestinal tract of broiler chickens. Appl Environ Microbiol 64, 40844088.CrossRefGoogle ScholarPubMed
25Mäkivuokko, HA, Saarinen, MT, Ouwehand, AC, et al. (2006) Effects of lactose on colon microbial community structure and function in a four-stage semi-continuous culture system. Biosci Biotechnol Biochem 70, 20562063.CrossRefGoogle Scholar
26Lane, DJ, Pace, B, Olsen, GJ, et al. (1985) Rapid determination of 16S ribosomal RNA sequences for phylogenetic analyses. Proc Natl Acad Sci U S A 82, 69556959.CrossRefGoogle ScholarPubMed
27Edwards, U, Rogall, T, Blocker, H, et al. (1989) Isolation and direct complete nucleotide determination of entire genes. Characterization of a gene coding for 16S ribosomal RNA. Nucleic Acids Res 17, 78437853.CrossRefGoogle ScholarPubMed
28Staden, R, Beal, KF & Bonfield, JK (2000) The Staden package, 1998. Methods Mol Biol 132, 115130.Google ScholarPubMed
29Saarinen, MT (2002) Determination of biogenic amines as dansyl derivatives in intestinal digesta and feces by reversed phase HPLC. Chromatographia 55, 297300.CrossRefGoogle Scholar
30Pinheiro, JC & Bates, DM (2000) Mixed-Effects Models in S and S-PLUS. New York: Springer-Verlag.CrossRefGoogle Scholar
31Wang, Q, Garrity, GM, Tiedje, JM, et al. (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 73, 52615267.CrossRefGoogle ScholarPubMed
32Audic, S & Claverie, JM (1997) The significance of digital gene expression profiles. Genome Res 7, 986995.CrossRefGoogle ScholarPubMed
33Fava, F, Makivuokko, H, Siljander-Rasi, H, et al. (2007) Effect of polydextrose on intestinal microbes and immune functions in pigs. Br J Nutr 98, 123133.CrossRefGoogle ScholarPubMed
34Suau, A, Bonnet, R, Sutren, M, et al. (1999) Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut. Appl Environ Microbiol 65, 47994807.CrossRefGoogle ScholarPubMed
35Wexler, HM (2007) Bacteroides: the good, the bad, and the nitty-gritty. Clin Microbiol Rev 20, 593621.CrossRefGoogle ScholarPubMed
36Ley, RE, Backhed, F, Turnbaugh, P, et al. (2005) Obesity alters gut microbial ecology. Proc Natl Acad Sci U S A 102, 1107011075.CrossRefGoogle ScholarPubMed
37Ley, RE, Hamady, M, Lozupone, C, et al. (2008) Evolution of mammals and their gut microbes. Science 320, 16471651.CrossRefGoogle ScholarPubMed
38Turnbaugh, PJ, Backhed, F, Fulton, L, et al. (2008) Diet-induced obesity is linked to marked but reversible alterations in the mouse distal gut microbiome. Cell Host Microbe 3, 213223.CrossRefGoogle ScholarPubMed
39Barcenilla, A, Pryde, SE, Martin, JC, et al. (2000) Phylogenetic relationships of butyrate-producing bacteria from the human gut. Appl Environ Microbiol 66, 16541661.CrossRefGoogle ScholarPubMed
40Duncan, SH, Hold, GL, Barcenilla, A, et al. (2002) Roseburia intestinalis sp. Nov., a novel saccharolytic, butyrate-producing bacterium from human faeces. Int J Syst Evol Microbiol 52, 16151620.Google Scholar
41Louis, P, Scott, KP, Duncan, SH, et al. (2007) Understanding the effects of diet on bacterial metabolism in the large intestine. J Appl Microbiol 102, 11971208.CrossRefGoogle ScholarPubMed
42Duncan, SH, Belenguer, A, Holtrop, G, et al. (2007) Reduced dietary intake of carbohydrates by obese subjects results in decreased concentrations of butyrate and butyrate-producing bacteria in feces. Appl Environ Microbiol 73, 10731078.CrossRefGoogle ScholarPubMed
43Kageyama, A, Sakamoto, M & Benno, Y (2000) Rapid identification and quantification of Collinsella aerofaciens using PCR. FEMS Microbiol Lett 183, 4347.CrossRefGoogle ScholarPubMed
44Zoetendal, EG, Ben-Amor, K, Harmsen, HJ, et al. (2002) Quantification of uncultured Ruminococcus obeum-like bacteria in human fecal samples by fluorescent in situ hybridization and flow cytometry using 16S rRNA-targeted probes. Appl Environ Microbiol 68, 42254232.CrossRefGoogle ScholarPubMed
45Walker, AW, Duncan, SH, Harmsen, HJ, et al. (2008) The species composition of the human intestinal microbiota differs between particle-associated and liquid phase communities. Environ Microbiol 10, 32753283.CrossRefGoogle ScholarPubMed
46Duncan, SH, Holtrop, G, Lobley, GE, et al. (2004) Contribution of acetate to butyrate formation by human faecal bacteria. Br J Nutr 91, 915923.CrossRefGoogle ScholarPubMed
47Matsumoto, M & Benno, Y (2007) The relationship between microbiota and polyamine concentration in the human intestine: a pilot study. Microbiol Immunol 51, 2535.CrossRefGoogle ScholarPubMed
48Thomas, T & Thomas, TJ (2001) Polyamines in cell growth and cell death: molecular mechanisms and therapeutic applications. Cell Mol Life Sci 58, 244258.CrossRefGoogle ScholarPubMed
49Larque, E, Sabater-Molina, M & Zamora, S (2007) Biological significance of dietary polyamines. Nutrition 23, 8795.CrossRefGoogle ScholarPubMed
50Wang, X, Feith, DJ, Welsh, P, et al. (2007) Studies of the mechanism by which increased spermidine/spermine N1-acetyltransferase activity increases susceptibility to skin carcinogenesis. Carcinogenesis 28, 24042411.CrossRefGoogle ScholarPubMed
51Seiler, N & Raul, F (2007) Polyamines and the intestinal tract. Crit Rev Clin Lab Sci 44, 365411.CrossRefGoogle ScholarPubMed
Figure 0

Table 1 Sex, age, bowel function and medication of the thirty-two study subjects

Figure 1

Fig. 1 Numbers of partial 16S rRNA gene sequences obtained from the three different study groups organised into bacterial phyla by the Ribosomal Database Project (RDP) II My RDP on-line-tool (https://rdp.cme.msu.edu/login/myrdp). ↔ , Statistical significance determined by standard proportions test (P < 0·001) between two study groups. NSAID, non-steroidal anti-inflammatory drugs.

Figure 2

Fig. 2 Average number of microbes in the percentage guanine+cytosine (%G+C) fractions in the three study groups: elderly subjects without non-steroidal anti-inflammatory drugs (NSAID) (); elderly subjects with NSAID (); young adults (□). Due to problems in DNA extraction (see Materials and methods), the final number of subjects included in the DNA pool of the elderly with NSAID was six, and nine from the elderly without NSAID. Statistically significant differences were determined by ANOVA and post hoct tests. * Mean value was significantly different from those of the two elderly groups (P < 0·01). † Mean value was significantly different from that of the elderly group without NSAID (P < 0·01).