Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inflammation, and Macrophage Dysfunction

Netusha Thevaranjan; Alicja Puchta; Christian Schulz; Avee Naidoo; J.C. Szamosi; Chris P. Verschoor; Dessi Loukov; Louis P. Schenck; Jennifer Jury; Kevin P. Foley; Jonathan D. Schertzer; Maggie J. Larché; Donald J. Davidson; Elena F. Verdú; Michael G. Surette; Dawn M.E. Bowdish

doi:10.1016/j.chom.2017.03.002

Thevaranjan, Netusha;Puchta, Alicja;Schulz, Christian;Naidoo, Avee;Szamosi, J.C.;Verschoor, Chris P.;Loukov, Dessi;Schenck, Louis P.;Jury, Jennifer;Foley, Kevin P.;Schertzer, Jonathan D.;Larché, Maggie J.;Davidson, Donald J.;Verdú, Elena F.;Surette, Michael G.;Bowdish, Dawn M.E.

2017-04-01 00:00:00

i An update to this article is included at the end Article Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inﬂammation, and Macrophage Dysfunction Graphical Abstract Authors Netusha Thevaranjan, Alicja Puchta, Christian Schulz, ..., Elena F. Verdu´ , Michael G. Surette, Dawn M.E. Bowdish Correspondence [email protected] In Brief Systemic inﬂammation increases with age, but the underlying causes are debated. Using young and old germ-free and conventional mice, Thevaranjan et al. demonstrate that age-related microbiota changes drive intestinal permeability, age-associated inﬂammation, and decreased macrophage function. Reducing TNF levels rescues microbiota changes and protects old mice from intestinal permeability. Highlights d Age-associated inﬂammation drives macrophage dysfunction and tissue damage d Mice under germ-free conditions are protected from age- associated inﬂammation d Co-housing germ-free mice with old, but not young, mice increases age-related inﬂammation d Age-related microbiota changes can be reversed by reducing TNF levels Thevaranjan et al., 2017, Cell Host & Microbe 21, 455–466 April 12, 2017 ª 2017 The Authors. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.chom.2017.03.002 Cell Host & Microbe Article Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inﬂammation, and Macrophage Dysfunction 1,2,3,8 1,2,3,8 1,2,3 1,2,3 4 Netusha Thevaranjan, Alicja Puchta, Christian Schulz, Avee Naidoo, J.C. Szamosi, 1,2,3 1,2,3 3,4,5 4,6 4,5 Chris P. Verschoor, Dessi Loukov, Louis P. Schenck, Jennifer Jury, Kevin P. Foley, 4,5 6 7 4,6 3,4,6 Jonathan D. Schertzer, Maggie J. Larche´ , Donald J. Davidson, Elena F. Verdu´ , Michael G. Surette, 1,2,3,9, and Dawn M.E. Bowdish * Department of Pathology and Molecular Medicine, McMaster University, Hamilton, ON L8N 3Z5, Canada McMaster Immunology Research Centre, McMaster University, Hamilton, ON L8N 3Z5, Canada Michael G. DeGroote Institute for Infectious Disease Research, McMaster University, Hamilton, ON L8N 3Z5, Canada Farncombe Family Digestive Health Research Institute, McMaster University, Hamilton, ON L8N 3Z5, Canada Department of Biochemistry & Biomedical Sciences, McMaster University, Hamilton, ON L8N 3Z5, Canada Department of Medicine, McMaster University, Hamilton, ON L8N 3Z5, Canada University of Edinburgh/MRC Centre for Inﬂammation Research, Queen’s Medical Research Institute, Edinburgh EH16 4TJ, UK Co-ﬁrst author Lead Contact *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chom.2017.03.002 SUMMARY phagocytes and that the resulting inﬂammatory response caused deterioration of surrounding tissues. Indeed, aging is Levels of inﬂammatory mediators in circulation are characterized by a state of chronic, low-grade, systemic inﬂam- mation (Franceschi et al., 2000). Higher than average levels of known to increase with age, but the underlying cause age-associated inﬂammation are a strong predictor of overall ill of this age-associated inﬂammation is debated. We health, development of chronic inﬂammatory conditions, and ﬁnd that, when maintained under germ-free condi- all-cause mortality in the elderly. Although age-associated tions, mice do not display an age-related increase inﬂammation inﬂuences the aging process, it is unclear why in circulating pro-inﬂammatory cytokine levels. levels of cytokines in the tissues and circulation increase with A higher proportion of germ-free mice live to age. It has been theorized that gradual, cumulative, sub-clinical 600 days than their conventional counterparts, and tissue damage occurs, which increases the burden of tissue macrophages derived from aged germ-free mice repair and results in increasing background levels of pro-inﬂam- maintain anti-microbial activity. Co-housing germ- matory cytokine production (Franceschi et al., 2000); however, free mice with old, but not young, conventionally the experimental evidence that would deﬁnitively prove this raised mice increases pro-inﬂammatory cytokines hypothesis is lacking. In contrast, studies in Drosophila demonstrate that age- in the blood. In tumor necrosis factor (TNF)-deﬁcient related changes in the microbiota increase intestinal perme- mice, which are protected from age-associated ability (Clark et al., 2015) and drive inﬂammation and mor- inﬂammation, age-related microbiota changes are tality (Rera et al., 2012). Although it has been demonstrated not observed. Furthermore, age-associated micro- that the microbial composition of the gut correlates with biota changes can be reversed by reducing TNF levels of circulating cytokines in the nursing home elderly using anti-TNF therapy. These data suggest that (Claesson et al., 2012) and in old mice (Conley et al., 2016), aging-associated microbiota promote inﬂammation it is not known whether this association is correlative or and that reversing these age-related microbiota whether the gut microbiota are a driver of age-associated changes represents a potential strategy for reducing inﬂammation. If the latter is true, it would indicate that these age-associated inﬂammation and the accompanying age-related changes in composition are a form of microbial morbidity. dysbiosis. Herein we report that intestinal permeability increases with age in mice due to age-related microbial dysbiosis. We INTRODUCTION demonstrate that microbial products enter the bloodstream of aged mice where they trigger systemic inﬂammation In 1907, Elie Metchnikoff proposed that tissue destruction and (i.e., elevated levels of serum interleukin 6 [IL6]). Chronic expo- senescence were consequences of chronic systemic inﬂamma- sure to inﬂammation alters macrophage function, rendering tion, which occurred as a result of increased permeability in the these cells poor killers of bacteria but potent producers of colon and the escape of bacteria and their products (Metchnik- inﬂammatory cytokines, which ultimately contributes to the off, 1907). He believed that these bacterial products activated inﬂammatory state of the aged host. Using old (18–22 months) Cell Host & Microbe 21, 455–466, April 12, 2017 ª 2017 The Authors. Published by Elsevier Inc. 455 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Figure 1. Inﬂammatory Responses Increase A B with Age (A) Killing of S. pneumoniae by resident peritoneal macrophages isolated from young and old mice (n = 5). (B) Killing of S. pneumoniae by bone marrow- derived macrophages from young and old C57BL/6 mice (n = 6). (C) Intact tetramethylrhodamine (TRITC)-labeled S. pneumoniae was observed in macrophages derived from old mice, but not young mice, up to 4 hr post-infection. (D–F) Levels of TNF (D) and IL6 (E) were higher in the CD plasma of old mice as was IL6 in slices of whole-lung homogenates from old mice (F). (G) H&E stain of formalin-ﬁxed histological sections from the lungs of young and old WT mice at 53 magniﬁcation. One representative image is at least ﬁve biological replicates. The degree of cellular inﬁltration within each image was measured by ex- pressing the total area of the cellular inﬁltrate within the lung as a percentage of the total lung area. (H) Lung slices were processed from the lungs of E F young and old mice and cultured in media. IL6 production was subsequently measured in the supernatant at 4 hr following stimulation with heat- killed S. pneumoniae or PBS control (n = 3, repre- sentative of two independent experiments). (I) IL6 production in the whole blood of young and old WT mice following stimulation with LPS or a vehicle control (PBS) (n = 5–9). (J) IL6 production from macrophages derived from young and old mice following 24-hr stimulation with a vehicle control (PBS), LPS, or S. pneumoniae as measured by ELISA (n = 6). Results represent mean ± SEM. Statistical signiﬁcance was deter- mined using the Mann-Whitney test or two-way ANOVA with Fisher’s post-test where appropriate (*p < 0.05, **p < 0.005, and ***p < 0.0005). RESULTS TNF Drives Age-Associated Defects in Macrophage Function We found that, after normalizing for differences in bacterial uptake between mice, resident peritoneal (Figure 1A) and bone marrow-derived (Figure 1B) macro- phages from old wild-type (WT) mice (18– 22 months) were impaired in their ability to kill Streptococcus pneumoniae as compared to those from young WT mice (10–14 weeks). Following internalization, bacterial lysis was observable in macro- phages from young mice but reduced or delayed in old mice (Figure 1C). Maturation markers on macrophages from young and old mice were expressed at equal levels, germ-free mice, which do not have age-associated inﬂam- indicating that differences observed with age were not due to mation, we demonstrate that colonization with the micro- altered differentiation or maturity (Figure S1, related to Figure 1). biota from old mice drives the inﬂammation that accompanies Levels of pro-inﬂammatory cytokines, such as IL6 and tumor aging. necrosis factor (TNF), in the circulation and tissues increase 456 Cell Host & Microbe 21, 455–466, April 12, 2017 Figure 2. Chronic Exposure to TNF Contrib- A B utes to Increased Inﬂammatory Responses and Tissue Damage that Occur with Age (A) Young and old murine bone marrow macro- phage-mediated killing of S. pneumoniae is decreased in the presence of 10 ng/mL exogenous TNF (n = 5). (B) Unlike old WT mice, old TNF KO mice do not have increased levels of plasma IL6 (n = 3–10 mice per group, one of two independent experiments shown). (C) IL6 production in the whole blood of young and C D old TNF KO mice following stimulation with LPS or a vehicle control (PBS) demonstrates that old TNF KO mice do not have higher inﬂammatory responses to LPS compared to young mice (n = 5). (D) IL6 levels as detected by ELISA in whole-lung tissue homogenates were no higher in old TNF KO mice than in young TNF mice (n = 3). (E) H&E stain of formalin-ﬁxed histological sections of lungs of young and old TNF KO mice (203 magniﬁcation, one representative of at least four). The degree of cellular inﬁltration within each image was measured by expressing the total area of the cellular inﬁltrate within the lung as a percentage of the total lung area. (F) Bone marrow-derived macrophages from young and old TNF KO mice do not differ in their ability to kill S. pneumoniae (n = 5). Results represent pooled data and are shown as mean ± SEM. Statistical signiﬁcance was determined using the Mann- Whitney test or two-way ANOVA with Fisher’s post- test where appropriate (*p < 0.05, **p < 0.005, and ***p < 0.0005). of TNF into mice impairs anti-pneumo- coccal immunity and increases levels of S. pneumoniae in experimental models (Hinojosa et al., 2009). We therefore hy- pothesized that the chronically elevated levels of TNF that occur with age could have a direct effect on macrophage-medi- with age, both in humans and mice (Bouchlaka et al., 2013; Fran- ated killing of S. pneumoniae. The exogenous addition of TNF ceschi et al., 2007). In keeping with previous reports, levels of (10 ng/mL) to culture media reduced bacterial killing by macro- TNF and IL6 in the circulation (Figures 1D and 1E) and IL6 in phages from young or old mice (Figure 2A), indicating that the the lungs (Figure 1F) were higher in old mice. Peribronchiolar decreased ability of old macrophages to kill S. pneumoniae could be due to higher levels of TNF. cellular inﬁltration was observed in the lungs of old mice in the absence of stimulation or overt infection (Figure 1G). Consistent Since acute exposure to TNF impaired macrophage killing of with what others have shown, whole blood (Figure 1H) from old S. pneumoniae, we postulated that chronic age-associated mice produced higher baseline levels of IL6 than that from inﬂammation, characterized by high systemic levels of TNF, young mice, and it produced more IL6 when stimulated with might underpin the reduction in macrophage anti-bacterial activ- S. pneumoniae or lipopolysaccharide (LPS), demonstrating ity. In contrast to WT animals, aged TNF knockout (KO) mice did signiﬁcantly enhanced pro-inﬂammatory responses to live bac- not have greater levels of IL6 in the circulation in the steady state teria and bacterial products. This phenotype was also observed (Figure 2B), and when LPS was added to whole blood, old TNF using bone marrow-derived macrophages from old mice, KO mice did not produce higher amounts of IL6 than the stimu- which produced more IL6 following stimulation with LPS or lated blood of young mice (Figure 2C). Constitutive IL6 produc- S. pneumoniae compared to young mice (Figure 1I). tion was not affected by age in TNF KO lung slices (Figure 2D), It has been frequently observed that individuals with higher and pulmonary cellular inﬁltrates were not observed in old levels of age-associated inﬂammation are at increased risk of mice, demonstrating protection from inﬂammation in the lungs both developing and dying from S. pneumoniae infection (An- (Figure 2E) Finally, bone marrow-derived macrophages from tunes et al., 2002; Yende et al., 2005). Furthermore, infusion old TNF KO mice did not have impaired pneumococcal killing Cell Host & Microbe 21, 455–466, April 12, 2017 457 in contrast to old WT mice (Figure 2F). Thus, age-associated inﬂammation and, more speciﬁcally, chronic exposure to TNF contribute to changes in macrophage function, resulting in decreased S. pneumoniae killing capacity. Intestinal Permeability and Levels of Circulating C Bacterial Products Increase with Age Although our data demonstrated that the presence of TNF promoted systemic inﬂammation and impaired macrophage function, the cause of increased TNF production with age was unclear. Based on Metchnikoff’s hypothesis that bacterial components from the gut microbiota could cause systemic inﬂammation, we investigated whether increased intestinal permeability and translocation of bacterial products occurred in aged mice. Intestinal permeability was measured in WT mice (3, 12, 15, and 18 months old), by performing oral gavages with 3–5 kDa ﬂuorescein isothiocyanate (FITC)-labeled dextran and measuring translocation of ﬂuorescence into the plasma. Intesti- nal permeability increased with age (Figure 3A). We next as- sessed whether this was due to altered paracellular and/or pas- sive permeability in the ileum and colon of young and old WT mice. Although there were no gross differences in intestinal ar- chitecture (Figure 3B), paracellular permeability was higher in the colons of old mice (Figure 3C), as determined by mucosal- 51 51 to-serosal ﬂux using chromium-EDTA ( Cr-EDTA). Consistent with evidence of increased permeability in the colon, where bac- terial numbers are highest, levels of the bacterial cell wall component muramyl dipeptide (MDP) were also signiﬁcantly higher in the plasma of old WT mice compared to young mice (Figure 3D). Thus, increased leakiness of the gut is a conse- quence of aging. I J protein (NOD)-nuclear factor kB (NF-kB) promoter bioassay. Signiﬁcant changes shown are relative to young WT mice. (E) Mucosal-to-serosal ﬂux of Cr-EDTA as measured in Ussing chambers was used to measure the paracellular permeability of the colons from young and old GF mice (n = 5–8). There was no signiﬁcant increase in permeability in old GF mice. (F) Survival analysis showing all-cause mortality of WT and GF mice up to 600 days of life. Differences in the survival curves were analyzed by log rank (Mantel-Cox) test. (G) Plasma cytokines are not higher than young GF mice and are lower than WT SPF mice (n = 3–5). (H) Histological analysis of lung sections stained with H&E from young and old GF mice does not indicate any increased leukocyte inﬁltration with age (203 magniﬁcation; one representative image of at least ﬁve mice). (I) IL6 levels in the lung homogenates of old GF mice are not higher than in Figure 3. The Microbiota Increase Intestinal Permeability, Age- young WT or young GF mice. Associated Inﬂammation, Macrophage Function, and Longevity (J) IL6 production in the whole blood of young and old WT SPF and GF mice (A) Intestinal permeability of aging mice (3, 12, 15, and 18 months old) was following stimulation with LPS or a vehicle control (PBS). Old GF mice do not measured by FITC-dextran translocation to the circulation following oral have higher levels than young WT SPF or young GF mice (n = 5–9). Signiﬁcant gavage (n = 4–8), and it was found to increase signiﬁcantly with age (p < 0.007, changes shown are relative to young WT mice. one-way ANOVA). (K) Macrophages from young and old GF mice do not differ in their ability to kill (B) Colons of young and old mice do not have detectable changes in either S. pneumoniae (n = 3). Results represent the mean ± SEM of three biological epithelial architecture or inﬂammatory inﬁltrate when measured as described replicates. in the STAR Methods. (L) Bone marrow-derived macrophages from old GF mice do no not have (C) Mucosal-to-serosal ﬂux of Cr-EDTA as measured in Ussing chambers decreased killing or produce more IL6 following stimulation with LPS or a was used to measure the paracellular permeability of ileums and colons from vehicle control (PBS) in macrophages from young GF mice (n = 5). Statistical young and old WT mice (n = 6–12). signiﬁcance was determined using the Mann-Whitney test or two-way ANOVA (D) Circulating muramyl dipeptide (MDP) in the plasma of young and old WT with Fisher’s post-test where appropriate (*p < 0.05, **p < 0.005, and mice as measured by nucleotide-binding oligomerization domain containing ***p < 0.0005). 458 Cell Host & Microbe 21, 455–466, April 12, 2017 Germ-free Mice Are Protected from Age-Associated nized mice. The paracellular permeability was measured from all Inﬂammation and Dysregulated Macrophage Function the GF mice colonized with the microbiota sourced from old mice If the increase in circulating microbial products is a driving force (n = 23 total, n = 11 young mice and n = 12 old mice) compared to in age-associated inﬂammation and mortality, we reasoned that young microbiota (n = 13, n = 6 young mice and n = 7 old mice). germ-free (GF) mice, which have no detectable MDP in the circu- Microbiota sourced from old mice signiﬁcantly increased lation nor increased intestinal permeability with age (Figure 3E), paracellular permeability (Figure 4F). The age of the host did, would be protected. The proportion of GF mice that lived to however, inﬂuence the development of paracellular permeability. 600 days was higher than the speciﬁc pathogen free (SPF) Young mice colonized with the microbiota from old mice had mice (Figure 3F). These GF mice were protected from age-asso- higher paracellular permeability, demonstrating that the compo- ciated inﬂammation, lacking the high circulating IL6 levels found sition of the microbiota can increase permeability. Although old in old control animals (Figure 3G). They also did not have peri- mice colonized with the old microbiota had the greatest degree bronchiolar cellular inﬁltrates (Figure 3H) or increased levels of of permeability, old mice colonized with the young microbiota IL6 in the lungs (Figure 3I) compared to young GF mice. Further- also had an increase in permeability, implying that there more, baseline and LPS-induced IL6 in whole blood did not in- are age-related changes in the gut that predispose to barrier crease with age in GF mice, in contrast to the signiﬁcantly higher dysfunction (Figure 4G). levels in old SPF/WT mice (Figure 3J). Finally, bone marrow- In addition to the paracellular permeability, the age of the host derived macrophages from old GF mice did not have impaired contributed to levels of circulating TNF (Figure 4H). Circulating S. pneumoniae killing capacity (Figure 3K) or produce more IL6 TNF was measured from all the young GF mice (n = 13 total, than young GF mice either basally or after stimulation with LPS n = 5 colonized with young microbiota and n = 8 colonized ex vivo (Figure 3L). These data demonstrate that chronic age- with old microbiota) and old GF mice (n = 11 total, n = 5 colonized associated inﬂammation requires the presence of microbiota. with the young microbiota and n = 6 colonized with the old micro- biota). Old recipients had higher levels of circulating TNF than The Composition of the Microbial Community Inﬂuences young recipients (Figure 4H). Young GF mice that were colonized Intestinal Permeability and Age-Associated with old SPF microbiota had higher levels of plasma TNF than Inﬂammation those recolonized with young SPF microbiota (Figure 4H), indi- We envisioned two possibilities that could explain how the mi- cating that the speciﬁc composition of the aging microbial com- crobiota drive age-associated inﬂammation. In the ﬁrst, the pres- munity contributes to age-associated inﬂammation but there are ence of any microbiota, even minimal microbiota, could result in age-related changes that predispose old mice to increased increased intestinal permeability. The second hypothesis was inﬂammation when they are exposed to any microbiota (Fig- that microbial dysbiosis occurs with age to drive increased intes- ure 4I). These data indicate that the composition of the aged mi- tinal permeability. crobiota altered intestinal permeability but that the composition To test these hypotheses, mice with a minimal microbiome of the microbiota interacts with other age-related changes in the were used. Mice were colonized with the altered Schaedler ﬂora host to enhance systemic inﬂammation. (ASF) (Dewhirst et al., 1999) on the C57BL/6 background and bred for two generations, during which time their microbiota A Reciprocal Relationship between Age-Associated diversiﬁed naturally as previously described (Slack et al., Inﬂammation and Microbial Dysbiosis 2009). The result is a low-diversity microbial community. Similar Our data demonstrate that the gut microbiota and/or age-related to old SPF WT mice, old ASF-derived mice had greater intestinal microbial dysbiosis can lead to increased gut permeability with permeability (Figure 4A), higher levels of plasma IL6 (Figure 4B), age and result in age-associated inﬂammation. However, since and higher IL6 production in whole blood following PBS or LPS expression of TNF has been shown to increase intestinal perme- stimulation (Figure 4C) than did young mice. Despite having min- ability in vitro (Soderholm et al., 2004) and anti-TNF treatment imal microbiota these mice also experienced age-related micro- can alter intestinal permeability in vivo (Noth et al., 2012), we bial dysbiosis (Figure 4D). also considered the possibility that age-associated increases Although our data demonstrated that colonization with micro- in TNF could exacerbate intestinal permeability and subsequent biota of initially limited diversity was sufﬁcient to elicit age-asso- release of bacterial products. We hypothesized that, if age-asso- ciated changes in permeability and inﬂammation, we next inves- ciated increases in TNF promoted increased intestinal perme- tigated whether the microbial composition changes with age to ability, old TNF KO mice would be protected and would not determine if microbial dysbiosis might contribute to these phe- have higher levels of circulating bacterial components than notypes. Similar to what others have reported, we found that young TNF KO mice. Consistent with this, intestinal barrier func- there were changes in both community structure (Figure 4E) tion in old TNF KO mice was equivalent to young TNF KO mice, and speciﬁc operational taxonomic units (OTUs) in the SPF young SPF WT mice, and young or old GF mice (Figure 5A), and mice between young and old mice (Table 1). To determine circulating levels of MDP in these mice did not increase with age whether this dysbiosis could increase age-associated inﬂamma- (Figure 5B). tion, young and old GF mice were colonized, via co-housing, with Although TNF is proposed to alter permeability of epithelial microbiota from either young or old SPF mice. The microbial dys- barriers, the mechanism remains unclear. We hypothesized biosis that was evident in the fecal microbiota of the donor mice that TNF may be a driver of microbial dysbiosis and that this was maintained in the colonized recipient mice over the time might be an indirect way in which barrier function is decreased. course of this study (Figure 4E). After a minimum of 6 weeks, If this hypothesis is correct then age-related microbial dysbiosis changes in paracellular permeability were assessed in the colo- would be less pronounced in old TNF KO mice. To determine Cell Host & Microbe 21, 455–466, April 12, 2017 459 Figure 4. Age-Associated Inﬂammation Is A BC Dependent on the Composition of the Intestinal Microbial Community (A and B) Mice with minimal ASF-derived microbiota were aged, and their intestinal permeability was measured via FITC-dextran oral gavage assay (A; n = 5). Old (18 months) ASF mice had higher intestinal permeability than young (10–14 weeks) ASF mice in addition to higher levels of plasma IL6 (B). (C) IL6 production after LPS stimulation in whole blood was higher in old ASF mice (n = 3). (D) The taxa summary of microbiota of young and old ASF mice indicates that age-related microbial dys- biosis occurs. (E) Taxa summaries illustrate that the composition of the young and old microbiota is retained upon transfer to young or old GF mice (n = 5–16 mice/group over four independent colonization experiments). (F) Paracellular permeability was measured in GF mice colonized with young and old microbiota (n = 6–8 mice per group). There was a statistically signiﬁcant in- crease in paracellular permeability in mice colonized with old microbiota (n = 23 total, n = 11 young mice and n = 12 old mice) compared to young microbiota (n = 13, n = 6 young mice and n = 7 old mice). This demon- strates that the age of the microbiota alters barrier function. E (G) Colonization of young GF mice with old microbiota increased paracellular permeability compared to those colonized with young microbiota; however, old mice colonized with either young or old microbiota demonstrated increased permeability, indicating that age-related changes in the host increased suscepti- bility to the microbiota. (H) Circulating TNF was measured from all the young GF mice (n = 13 total, n = 5 colonized with young mi- crobiota and n = 8 colonized with old microbiota) and old GF mice (n = 11 total, n = 5 colonized with the young microbiota and n = 6 colonized with the old microbiota). Old recipient mice had higher levels of circulating TNF than young recipient mice. (I) The microbiota contributed to the increased TNF in F G the circulation of young mice, since young GF mice colonized with old microbiota had higher circulating levels of TNF than those colonized with young micro- biota. In contrast, colonization with either the young or old microbiota increased circulating TNF in old GF mice, indicating that the age of the host interacts with the age of the microbiota to induce systemic inﬂam- mation. Bars represent the mean ± SEM. Statistical signiﬁcance was determined using the Mann-Whitney test or two-way ANOVA with Fisher’s post-test or un- paired t test where appropriate (*p < 0.05, **p < 0.005, and ***p < 0.0005). H I 460 Cell Host & Microbe 21, 455–466, April 12, 2017 Table 1. OTUs that Were Signiﬁcantly Changed in Old SPF and TNF KO Mice Increased Family Genus Number of OTUs (WT) Old > Young (WT) Old > Young (TNF KO) Number of OTUs (TNF KO) Ruminococcaceae Ruminococcus 6 **** NS 7 Lachnospiraceae Clostridium 34 **** ** 34 Ruminococcaceae Clostridium 14 **** NS 16 Prevotellaceae Prevotella 39 ** ** 44 Erysipelotrichaceae Allobaculum 54 **** * 60 Lachnospiraceae many identiﬁed 354 * * 343 Biﬁdobacteriaceae Biﬁdobacterium 17 *** ** 17 Ruminococcaceae Oscillospira 54 **** NS 56 Lactobacillaceae Lactobacillus 53 *** **** 57 Bacteroidaceae Bacteroides 52 ** *** 71 Coriobacteriaceae Adlercreutzia 22 *** NS 23 Peptococcaceae not identiﬁed 14 * NS 15 Catabacteriaceae not identiﬁed 51 **** ** 65 Coriobacteriaceae not identiﬁed 14 *** NS 14 Decreased Family Genus Number of OTUs (WT) Old < Young (WT) Old < Young (TNF KO) Number of OTUs (TNF KO) Rikenellaceae Alistipes 48 ** NS 52 Verrucomicrobiaceae Akkermansia 41 **** * 49 Lachnospiraceae Blautia 11 **** * 10 *p < 0.05, **p < 0.005, ***p < 0.0005, and ****p < 0.0001; NS, not signiﬁcant. Number of OTUs is the number of OTUs found in the samples that are attributed to the family or genus. OTUs that were signiﬁcantly decreased in old TNF KO mice. OTUs that were signiﬁcantly increased in old TNF KO mice. whether the fecal microbiota of TNF KO mice were less divergent intestinal permeability by directly altering intestinal barrier with age than the WT microbiota, the distances between young- function, but they do indicate that elevated levels of TNF old pairs were calculated using the Bray-Curtis distance matrix. contribute to age-associated microbial dysbiosis. Collectively Average distances between the young and old WT mice were these data are most consistent with dysbiosis causing increased greater than the young and old TNF KO mice, indicating that intestinal permeability and translocation of bacterial products, the beta diversity was greater between young and old WT mice which increases systemic inﬂammation that ultimately impairs (p = 0.04, Mann-Whitney test). The divergence in the microbiota macrophage function, a model for which is presented in that occurs in aging WT mice and the decreased divergence that Figure 5E. occurs in aging TNF KO mice is visualized in Figure 5C and listed in Table 1. DISCUSSION To further evaluate the potential of TNF to induce changes to the intestinal microbiota, young and old WT mice were treated Age-associated inﬂammation is a strong risk factor for overall with the anti-TNF drug Humira for 2 weeks, which reduced mortality in older adults. In fact, individuals having higher than TNF levels in old mice to below the limit of detection. Anti-TNF, age-average levels of inﬂammatory markers are more likely to but not an IgG control, altered the composition of the intestinal be hospitalized (de Gonzalo-Calvo et al., 2012a), have higher microbiota of old mice (Figure 5D), as there was a signiﬁcant all-cause mortality rates (Giovannini et al., 2011), be frail (Leng (p = 0.045) interaction on the Bray-Curtis distances of the old et al., 2007), be less independent (de Gonzalo-Calvo et al., Humira and IgG microbiota, but not the young Humira- and 2012b), and are more likely to have a variety of late-life diseases IgG-treated microbiota. This demonstrates that the microbiota (Bruunsgaard, 2006; Bruunsgaard et al., 1999) Age-associated can be manipulated by altering the inﬂammatory status of the inﬂammation has also been shown to increase susceptibility to host. Changes in speciﬁc OTUs during anti-TNF treatment are pneumococcal infection (Yende et al., 2005, 2013), and it is asso- described in Table 2. Despite altering the microbiota, anti-TNF ciated with increased disease severity and decreased survival treatment had no measurable effect on intestinal permeability, from pneumococcal infection in older adults (Antunes et al., as measured by translocation of FITC-dextran (data not shown), 2002; Reade et al., 2009). Despite the clinical importance of indicating that these changes to the composition of the intestinal age-associated inﬂammation, the etiological factors that lead microbiota were not a consequence of altered TNF levels having to its development have not been identiﬁed. This study demon- any direct effect on intestinal permeability. These experiments strates that age-associated inﬂammation and microbial dysbio- do not rule out a role for age-associated increases in TNF driving sis drive intestinal permeability and translocation of bacterial Cell Host & Microbe 21, 455–466, April 12, 2017 461 Figure 5. Microbial Dysbiosis Occurs with 600 600 A B ** *** * 80 80 Age and Inﬂammation ** *** * (A) Intestinal permeability, as measured by plasma 60 60 400 400 FITC-dextran following oral gavage, was increased in old WT/SPF mice, but not old TNF KO or old GF 40 40 mice. 200 200 (B) Circulating MDP is increased in old SPF/WT 20 20 mice. Old TNF KO and GF mice are not signiﬁ- cantly different from young WT/SPF mice (n = 5– 0 0 0 0 10). GF mice do not have any detectable MDP in Yo You un ng g Old Old Yo You un ng g Old Old Y Yo oung ung Old Old Y Yo oung ung Old Old Y Yo oung ung Old Old the circulation. WT WT WT WT TNF TNF TNF TNF WT WT WT WT TNF TNF TNF TNF GF GF GF GF (C) Principal coordinate analysis based on Bray- Curtis demonstrates that the microbial commu- Wildtype TNF KO nities of old WT mice diverge from young mice, but 0.2 0.2 this is not the case in old TNF KO mice. Mice were sampled from multiple cages. Chi-square of the 0.1 0.1 likelihood ratio test in DESeq2 shows old and young Ag Age e ol old d microbiota are signiﬁcantly different (p < 0.001). yo you un ng g (D) Anti-TNF (Adalimumab) or a human IgG isotype -0.1 control was administered at a dose of 50 ng/g of -0.1 -0.2 body weight every other day for 2 weeks. Principal -0.2 0 0.2 -0.2 -0.1 0 0.1 0.2 co-ordinate analysis was used to visualize differ- Axis 2 [25.3%]] Axis 2 [11.2%]] ences in the microbial communities after 2 weeks 0.2 Yo You un ng g I Ig gG G of anti-TNF treatment. Anti-TNF treatment altered 0.2 Yo You un ng g a an nt ti i- -T TN NF F the composition of the fecal microbiota of old, but Ol Old d I Ig gG G 0.1 0.1 not young, mice. Ol Old d a an nt ti i- -T TN NF F (E) Basal translocation of microbial products oc- -0.4 0.2 0.4 -0.2 0 curs throughout life; however, with age, these Axis 1[44.5%] -0.1 induce an inﬂammatory response, which contrib- -0.1 utes to microbial dysbiosis. Microbial dysbiosis -0.2 increases intestinal permeability, which increases -0.6 -0.2 0 0.2 -0.4 -0.25 0 0.25 Axis 1 [39.2%]] Axis 1 [67.0%] bacterial translocation. This feed-forward process increases with age. translocation and subsequent inﬂamma- tion are a driver of intestinal permeability, rather than a readout of intestinal damage (Kristoff et al., 2014). These data are consistent with our model in which both GF and TNF KO mice (which do not have increased levels of circulating bacterial products) are protected from age-associ- ated inﬂammation. Unlike the GF mice, components, further fueling inﬂammation and impairing cellular which are protected by virtue of not being exposed to bacteria, antibacterial function. the TNF mice may be protected because they do not undergo In humans, transient endotoxemia occurs naturally after inges- microbial dysbiosis with age, which we demonstrate confers in- testinal permeability and systemic inﬂammation in the context of tion of high-fat meals, vigorous exercise, and in many diseases (Kelly et al., 2012). Microbial translocation has been shown to the aged host. This may be an evolutionarily conserved compo- occur in HIV patients due to a loss of immune control at the nent of the aging process, since intestinal permeability has been gut mucosa, and this translocation leads to a state of immune demonstrated to precede systemic inﬂammation and to be a activation and systemic inﬂammation that is reminiscent of marker of premature death in Drosophila (Rera et al., 2012). what is observed in normal aging (Brenchley et al., 2006). This in- Metchnikoff made careful observations that, in acute inﬂam- crease in chronic inﬂammation correlates with early mortality, mation, macrophage-mediated phagocytosis seemed to be which is often due to premature development of diseases of impaired. In examining autopsy samples of the elderly, he age such as cardiovascular disease (Sandler et al., 2011). In sim- noticed that brain tissue macrophages seemed to be associated ian models of HIV, translocation of bacterial products is a precur- with areas of damage, and he hypothesized that their presence sor of immune activation and macrophage dysfunction (Estes might do more harm than good. He also observed that the integ- et al., 2010). In these models, reducing levels of circulating rity of the gut changed with age and concluded that, ‘‘it is indu- LPS by chelation with the drug sevelamer prevents immune bitable, therefore, that the intestinal microbes or their poisons dysfunction, systemic inﬂammation, and, most relevant to our may reach the system generally and bring harm to it’’ (Metchnik- study, reduces intestinal permeability, implying that bacterial off, 1907). He believed that this macrophage ‘‘intoxiﬁcation’’ had 462 Cell Host & Microbe 21, 455–466, April 12, 2017 Axis 3 [13.5%] Ax Axi is s 3 3 [ [13.5%] Gut-to-Plasm Gut-to-Plasma a F FITC-Dextran ITC-Dextran Axis 3 [9.7%] Axis 3 [9.7%]] Muramyl Dipeptide (ng/ml) Muramyl Dipeptide (ng/ml) Axis 3 [6.29%] Table 2. OTUs Altered by Anti-TNF Treatment in Old Mice Increased Family Genus Old SPF > Young SPF Decreased by Anti-TNF Treatment Erysipelotrichaceae not identiﬁed **** NS Ruminococcaceae Ruminococcus **** NS Lachnospiraceae Clostridium **** NS Ruminococcaceae Clostridium **** NS Prevotellaceae Prevotella ** NS Erysipelotrichaceae Allobaculum **** NS Lachnospiraceae many identiﬁed * NS Biﬁdobacteriaceae Biﬁdobacterium *** NS Ruminococcaceae Oscillospira **** NS Lactobacillaceae Lactobacillus *** NS Bacteroidaceae Bacteroides ** * Coriobacteriaceae Adlercreutzia *** * Peptococcaceae not identiﬁed * ** Catabacteriaceae not identiﬁed **** * (increased) Coriobacteriaceae not identiﬁed *** * Ruminococcaceae Subdoligranulum NS * Decreased Family Genus Old < Young Increased by Anti-TNF Treatment Rikenellaceae Alistipes ** NS Verrucomicrobiaceae Akkermansia **** NS Lachnospiraceae Blautia **** NS Lachnospiraceae Roseburia NS * Eubacteriaceae Anaerofustis NS * Families that are higher in old SPF mice are listed. Those that are decreased by anti-TNF treatment in old mice are labeled with an asterisk. *p < 0.05, **p < 0.005, ***p < 0.0005, and ****p < 0.0001; NS, not signiﬁcant. systemic effects and led to deterioration of even distal tissues. The authors suggest that these changes may occur following Our observations are consistent with his, as we observed an in- overgrowth of gut microbes and/or threshold production of crease in circulating bacterial products as our WT mice aged and bacterial products, resulting in their systemic translocation, evidence of systemic and distal inﬂammation. Although we only increased inﬂammation, and ensuing pulmonary endothelial measured the presence of bacterial products in the serum, it is damage. The bacterial taxa that were mainly implicated in this entirely possible that they also enter the lymphatics. It has pathogenicity were members of Clostridia, which others have recently been demonstrated that acute infections can perma- also demonstrated have distinct abundance patterns in the ag- nently remodel the lymphatics, causing them to become more ing gut microbial community (Claesson et al., 2011, 2012). permeable (Fonseca et al., 2015). Whether age-related microbial Although it has been suggested that changes in the microbiota dysbiosis increases lymphatic permeability is unknown. Regard- might drive the ills of aging, determining cause and effect has less of how bacterial products enter the periphery, the systemic been challenging. Numerous studies have demonstrated that inﬂammation they cause has profound effects on myelopoiesis, there are characteristic changes in gut microbial communities since macrophages derived from bone marrow precursors in in elderly humans (Claesson et al., 2011; Ma €kivuokko et al., the absence of the aging microenvironment become hyper-in- 2010; Mariat et al., 2009; Zwielehner et al., 2009) and that these ﬂammatory and have poor killing capacity. Although Metchnikoff changes correlate with health status in the elderly population imagined that loss of macrophage function was a result of age- (Bartosch et al., 2004; Claesson et al., 2012). Furthermore, ther- associated inﬂammation, he did not predict that they may also apeutic manipulation of the gut microbiota appears to improve contribute to the global inﬂammatory state. In fact, it appears immune function in the elderly. For example, oral supplementa- as though both aged monocytes (Puchta et al., 2016) and mac- tion with Biﬁdobacterium increased lymphocyte proportions in rophages (Mirsoian et al., 2014) contribute to chronic inﬂamma- the circulation, improved the anti-tumoricidal activity of natural tion, as their depletion reduces levels of inﬂammatory cytokines. killer cells, and restored phagocytosis in peripheral blood mono- Consistent with our ﬁndings that the gut microbiota can also nuclear cells and neutrophils (Gill et al., 2001a, 2001b). Interest- inﬂuence systemic (i.e., lung) inﬂammation and tissue damage, ingly, these beneﬁts were most strongly evident in individuals 70 it has been shown that increased circulating bacterial toxins years of age and older, as well as those individuals who demon- result in reduced tight junction gene expression and lethal pul- strated the greatest degree of cellular immunosenescence. monary damage following fecal transplantation (Ji et al., 2014). Furthermore, dysbiosis in HIV patients, which shows many Cell Host & Microbe 21, 455–466, April 12, 2017 463 d KEY RESOURCES TABLE parallels to that which occurs in the elderly (including decreased d CONTACT FOR REAGENT AND RESOURCE SHARING Biﬁdobacteria frequency and increased clusters of Clostridium), d EXPERIMENTAL MODEL AND SUBJECT DETAILS decreases following prebiotic administration. This led to a B Ethics statement decrease in the overall degree of microbial translocation and B Mouse Experiments ultimately improved immune cell function (Gori et al., 2011). d METHOD DETAILS The microbial communities of the elderly gut appear to be B TNF ablation strongly inﬂuenced by diet (Claesson et al., 2012), and dietary in- B Histological analysis terventions designed to restore a robust microbiota may improve B Measurement of cytokine production anti-bacterial immunity by reducing age-associated inﬂamma- B Macrophage culture tion and macrophage immunosenescence (Clements and Card- B Bacterial killing assays ing, 2016; Vaiserman et al., 2017). B In vitro and in vivo permeability Although manipulation of the microbiota may improve health B MDP Detection Bioassay in the elderly, until now it has not been clear whether microbial B Germ-free Mouse Recolonization dysbiosis is a driver of immune dysfunction. For example, it has B Bacterial proﬁling by deep sequencing analysis of 16S been demonstrated that gut microbial composition correlates rRNA with Illumina with levels of circulating cytokines and markers of health in d QUANTIFICATION AND STATISTICAL ANALYSIS the elderly (Claesson et al., 2012) and that intestinal perme- d DATA AND SOFTWARE AVAILABILITY ability and systemic inﬂammation increase in old mice (Scott et al., 2017), but not whether the microbiota drive these changes. Our data demonstrate that microbial dysbiosis occurs SUPPLEMENTAL INFORMATION with age, even in minimal microbiota, and these changes are Supplemental Information includes one ﬁgure and can be found with this sufﬁcient to promote age-associated inﬂammation, although article online at http://dx.doi.org/10.1016/j.chom.2017.03.002. we have not determined whether this is due to enrichment of speciﬁc species, changes in microbe-microbe interactions, AUTHOR CONTRIBUTIONS alterations in the functional capacity of the aging microbiota (e.g., changes in short-chain fatty acid production), or loss of Conceptualization, D.M.E.B., M.G.S., E.F.V., and D.J.D.; Methodology, J.D.S., compartmentalization of the microbiota as is found in E.F.V., and M.G.S.; Investigation, N.T., A.P., A.N., C.S., J.C.S., C.P.V., D.L., Drosophila (Li et al., 2016). Interestingly there may be a causal L.P.S., J.J., and K.P.F.; Resources, M.J.L.; Review & Editing, D.M.E.B., M.G.S., and D.J.D.; Supervision, D.M.E.B., M.G.S., J.D.S., and E.F.V. relationship between age-associated inﬂammation and microbi- al dysbiosis, since we found that TNF KO mice had a less diver- ACKNOWLEDGMENTS gent microbiota with age and treatment with anti-TNF altered the microbial communities of aged mice. Although there were A.P. was supported by an Ontario Graduate Scholarship. C.P.V. was sup- signiﬁcant changes in the composition of the microbiota with ported by a fellowship from the Canadian Thoracic Society. A.N. was sup- anti-TNF treatment, we have not yet identiﬁed which members ported by a scholarship from the Canadian Institutes of Health Research of the microbial community alter barrier function with age. (CIHR). N.T. was supported by an Early Researcher Award from the Ontario Ministry of Research and Innovation. This work was funded by grants from Further experiments will need to be performed to determine if the CIHR to D.M.E.B. (FRN 123404 and 224026). E.F.V. is supported by the it is the loss of beneﬁcial members of the microbial community, CIHR (MOP 142773). E.F.V., M.G.S., and D.M.E.B. are supported by the overgrowth of harmful members, or a shift in metabolism that CIHR and hold Canada Research Chairs. Work in the Bowdish laboratory is contributes to this phenomenon. supported by the McMaster Immunology Research Centre (MIRC) and the Metchnikoff had great faith that the appropriate experiments M.G. DeGroote Institute for Infectious Disease Research (IIDR). D.J.D. is an could be performed to demonstrate that manipulation of the MRC Senior Research Fellow (G1002046). The authors would like to thank intestinal microbiota would extend life. Until that time he sug- Kate Manners and Laura Rossi for isolation and preparation of DNA for micro- biome analysis and Fiona Whelan for advice on microbiome analysis. gested, ‘‘. those who wish to preserve their intelligence as long as possible and to make their cycle of life as complete Received: October 4, 2016 and as normal as possible under present conditions, must Revised: February 1, 2017 depend on general sobriety and on habits conforming to the Accepted: March 2, 2017 rules of rational hygiene.’’ The experiments he envisioned remain Published: April 12, 2017; corrected online: April 9, 2018 to be performed, and, until they are, the only reliable ways to reduce age-associated inﬂammation, delay the onset of inﬂam- REFERENCES matory diseases, and prolong life are a sensible diet (Fontana Antunes, G., Evans, S.A., Lordan, J.L., and Frew, A.J. (2002). Systemic cyto- and Hu, 2014; Fontana and Klein, 2007) and exercise (Lee kine levels in community-acquired pneumonia and their association with dis- et al., 1995). For those of us less inclined to live a lifestyle of gen- ease severity. Eur. Respir. J. 20, 990–995. eral sobriety, targeting age-associated inﬂammation may pro- Bartosch, S., Fite, A., Macfarlane, G.T., and McMurdo, M.E. (2004). vide an attractive alternative. Characterization of bacterial communities in feces from healthy elderly volun- teers and hospitalized elderly patients by using real-time PCR and effects of STAR+METHODS antibiotic treatment on the fecal microbiota. Appl. Environ. Microbiol. 70, 3575–3581. Detailed methods are provided in the online version of this paper Bartram, A.K., Lynch, M.D., Stearns, J.C., Moreno-Hagelsieb, G., and Neufeld, and include the following: J.D. (2011). Generation of multimillion-sequence 16S rRNA gene libraries from 464 Cell Host & Microbe 21, 455–466, April 12, 2017 complex microbial communities by assembling paired-end illumina reads. Franceschi, C., Bonafe` , M., Valensin, S., Olivieri, F., De Luca, M., Ottaviani, E., Appl. Environ. Microbiol. 77, 3846–3852. and De Benedictis, G. (2000). Inﬂamm-aging. An evolutionary perspective on immunosenescence. Ann. N Y Acad. Sci. 908, 244–254. Bouchlaka, M.N., Sckisel, G.D., Chen, M., Mirsoian, A., Zamora, A.E., Maverakis, E., Wilkins, D.E., Alderson, K.L., Hsiao, H.H., Weiss, J.M., et al. Franceschi, C., Capri, M., Monti, D., Giunta, S., Olivieri, F., Sevini, F., (2013). Aging predisposes to acute inﬂammatory induced pathology after tu- Panourgia, M.P., Invidia, L., Celani, L., Scurti, M., et al. (2007). Inﬂammaging mor immunotherapy. J. Exp. Med. 210, 2223–2237. and anti-inﬂammaging: a systemic perspective on aging and longevity emerged from studies in humans. Mech. Ageing Dev. 128, 92–105. Brenchley, J.M., Price, D.A., Schacker, T.W., Asher, T.E., Silvestri, G., Rao, S., Kazzaz, Z., Bornstein, E., Lambotte, O., Altmann, D., et al. (2006). Microbial Gill, H.S., Rutherfurd, K.J., and Cross, M.L. (2001a). Dietary probiotic supple- translocation is a cause of systemic immune activation in chronic HIV infection. mentation enhances natural killer cell activity in the elderly: an investigation of Nat. Med. 12, 1365–1371. age-related immunological changes. J. Clin. Immunol. 21, 264–271. Bruunsgaard, H. (2006). The clinical impact of systemic low-level inﬂammation Gill, H.S., Rutherfurd, K.J., Cross, M.L., and Gopal, P.K. (2001b). in elderly populations. With special reference to cardiovascular disease, de- Enhancement of immunity in the elderly by dietary supplementation with the mentia and mortality. Dan. Med. Bull. 53, 285–309. probiotic Biﬁdobacterium lactis HN019. Am. J. Clin. Nutr. 74, 833–839. Bruunsgaard, H., Andersen-Ranberg, K., Jeune, B., Pedersen, A.N., Skinhøj, Giovannini, S., Onder, G., Liperoti, R., Russo, A., Carter, C., Capoluongo, E., P., and Pedersen, B.K. (1999). A high plasma concentration of TNF-alpha is Pahor, M., Bernabei, R., and Landi, F. (2011). Interleukin-6, C-reactive protein, associated with dementia in centenarians. J. Gerontol. A Biol. Sci. Med. Sci. and tumor necrosis factor-alpha as predictors of mortality in frail, community- 54, M357–M364. living elderly individuals. J. Am. Geriatr. Soc. 59, 1679–1685. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Gori, A., Rizzardini, G., Van’t Land, B., Amor, K.B., van Schaik, J., Torti, C., Costello, E.K., Fierer, N., Pen˜ a, A.G., Goodrich, J.K., Gordon, J.I., et al. Quirino, T., Tincati, C., Bandera, A., Knol, J., et al. (2011). Speciﬁc prebiotics (2010). QIIME allows analysis of high-throughput community sequencing modulate gut microbiota and immune activation in HAART-naive HIV-infected data. Nat. Methods 7, 335–336. adults: results of the ‘‘COPA’’ pilot randomized trial. Mucosal Immunol. 4, Claesson, M.J., Cusack, S., O’Sullivan, O., Greene-Diniz, R., de Weerd, H., 554–563. Flannery, E., Marchesi, J.R., Falush, D., Dinan, T., Fitzgerald, G., et al. Hinojosa, E., Boyd, A.R., and Orihuela, C.J. (2009). Age-associated inﬂamma- (2011). Composition, variability, and temporal stability of the intestinal micro- tion and toll-like receptor dysfunction prime the lungs for pneumococcal pneu- biota of the elderly. Proc. Natl. Acad. Sci. USA 108 (Suppl 1 ), 4586–4591. monia. J. Infect. Dis. 200, 546–554. Claesson, M.J., Jeffery, I.B., Conde, S., Power, S.E., O’Connor, E.M., Cusack, Ji, Y., Sun, S., Goodrich, J.K., Kim, H., Poole, A.C., Duhamel, G.E., Ley, R.E., S., Harris, H.M., Coakley, M., Lakshminarayanan, B., O’Sullivan, O., et al. and Qi, L. (2014). Diet-induced alterations in gut microﬂora contribute to lethal (2012). Gut microbiota composition correlates with diet and health in the pulmonary damage in TLR2/TLR4-deﬁcient mice. Cell Rep. 8, 137–149. elderly. Nature 488, 178–184. Kelly, C.J., Colgan, S.P., and Frank, D.N. (2012). Of microbes and meals: the Clark, R.I., Salazar, A., Yamada, R., Fitz-Gibbon, S., Morselli, M., Alcaraz, J., health consequences of dietary endotoxemia. Nutr. Clin. Pract. 27, 215–225. Rana, A., Rera, M., Pellegrini, M., Ja, W.W., and Walker, D.W. (2015). Kristoff, J., Haret-Richter, G., Ma, D., Ribeiro, R.M., Xu, C., Cornell, E., Stock, Distinct Shifts in Microbiota Composition during Drosophila Aging Impair Intestinal Function and Drive Mortality. Cell Rep. 12, 1656–1667. J.L., He, T., Mobley, A.D., Ross, S., et al. (2014). Early microbial translocation blockade reduces SIV-mediated inﬂammation and viral replication. J. Clin. Clements, S.J., and Carding, S.R. (2016). Diet, the intestinal microbiota and Invest. 124, 2802–2806. immune health in ageing. Crit. Rev. Food Sci. Nutr. Lee, I.M., Hsieh, C.C., and Paffenbarger, R.S., Jr. (1995). Exercise intensity Conley, M.N., Wong, C.P., Duyck, K.M., Hord, N., Ho, E., and Sharpton, T.J. and longevity in men. The Harvard Alumni Health Study. JAMA 273, (2016). Aging and serum MCP-1 are associated with gut microbiome compo- 1179–1184. sition in a murine model. PeerJ 4, e1854. Leng, S.X., Xue, Q.L., Tian, J., Walston, J.D., and Fried, L.P. (2007). de Gonzalo-Calvo, D., de Luxa´ n-Delgado, B., Martı´nez-Camblor, P., Inﬂammation and frailty in older women. J. Am. Geriatr. Soc. 55, 864–871. Rodrı´guez-Gonza´ lez, S., Garcı´a-Macia, M., Sua´ rez, F.M., Solano, J.J., Rodrı´guez-Colunga, M.J., and Coto-Montes, A. (2012a). Chronic inﬂammation Li, H., Qi, Y., and Jasper, H. (2016). Preventing Age-Related Decline of Gut as predictor of 1-year hospitalization and mortality in elderly population. Eur. J. Compartmentalization Limits Microbiota Dysbiosis and Extends Lifespan. Clin. Invest. 42, 1037–1046. Cell Host Microbe 19, 240–253. de Gonzalo-Calvo, D., de Luxa´ n-Delgado, B., Rodrı´guez-Gonza´ lez, S., Garcı´a- Makivuokko, € H., Tiihonen, K., Tynkkynen, S., Paulin, L., and Rautonen, N. Macia, M., Sua´ rez, F.M., Solano, J.J., Rodrı´guez-Colunga, M.J., and Coto- (2010). The effect of age and non-steroidal anti-inﬂammatory drugs on human Montes, A. (2012b). Interleukin 6, soluble tumor necrosis factor receptor I intestinal microbiota composition. Br. J. Nutr. 103, 227–234. and red blood cell distribution width as biological markers of functional depen- Mariat, D., Firmesse, O., Levenez, F., Guimaraes, V., Sokol, H., Dore´ , J., dence in an elderly population: a translational approach. Cytokine 58, Corthier, G., and Furet, J.P. (2009). The Firmicutes/Bacteroidetes ratio of the 193–198. human microbiota changes with age. BMC Microbiol. 9, 123. Dewhirst, F.E., Chien, C.C., Paster, B.J., Ericson, R.L., Orcutt, R.P., Schauer, Martin, M. (2011). Cutadapt removes adapter sequences from high-throughput D.B., and Fox, J.G. (1999). Phylogeny of the deﬁned murine microbiota: altered sequencing reads. EMBnet.journal 17, 10–12. Schaedler ﬂora. Appl. Environ. Microbiol. 65, 3287–3292. Masella, A.P., Bartram, A.K., Truszkowski, J.M., Brown, D.G., and Neufeld, Estes, J.D., Harris, L.D., Klatt, N.R., Tabb, B., Pittaluga, S., Paiardini, M., J.D. (2012). PANDAseq: paired-end assembler for illumina sequences. BMC Barclay, G.R., Smedley, J., Pung, R., Oliveira, K.M., et al. (2010). Damaged Bioinformatics 13,31. intestinal epithelial integrity linked to microbial translocation in pathogenic simian immunodeﬁciency virus infections. PLoS Pathog. 6, e1001052. McMurdie, P.J., and Holmes, S. (2013). phyloseq: an R package for reproduc- ible interactive analysis and graphics of microbiome census data. PLoS ONE Fonseca, D.M., Hand, T.W., Han, S.J., Gerner, M.Y., Glatman Zaretsky, A., 8, e61217. Byrd, A.L., Harrison, O.J., Ortiz, A.M., Quinones, M., Trinchieri, G., et al. (2015). Microbiota-Dependent Sequelae of Acute Infection Compromise Metchnikoff, E. (1907). The Prolongation of Life: Optimistic Studies (G. P. Tissue-Speciﬁc Immunity. Cell 163, 354–366. Putnam’s Sons). Fontana, L., and Hu, F.B. (2014). Optimal body weight for health and longevity: Mirsoian, A., Bouchlaka, M.N., Sckisel, G.D., Chen, M., Pai, C.C., Maverakis, bridging basic, clinical, and population research. Aging Cell 13, 391–400. E., Spencer, R.G., Fishbein, K.W., Siddiqui, S., Monjazeb, A.M., et al. (2014). Fontana, L., and Klein, S. (2007). Aging, adiposity, and calorie restriction. Adiposity induces lethal cytokine storm after systemic administration of stim- JAMA 297, 986–994. ulatory immunotherapy regimens in aged mice. J. Exp. Med. 211, 2373–2383. Cell Host & Microbe 21, 455–466, April 12, 2017 465 Noth, R., Stuber, € E., Hasler, € R., Nikolaus, S., Kuhbacher, € T., Hampe, J., Bewig, Ubeda, C., Lipuma, L., Gobourne, A., Viale, A., Leiner, I., Equinda, M., Khanin, B., Schreiber, S., and Arlt, A. (2012). Anti-TNF-a antibodies improve intestinal R., and Pamer, E.G. (2012). Familial transmission rather than defective innate barrier function in Crohn’s disease. J. Crohn’s Colitis 6, 464–469. immunity shapes the distinct intestinal microbiota of TLR-deﬁcient mice. J. Exp. Med. 209, 1445–1456. Novakowski, K.E., Loukov, D., Chawla, V., and Bowdish, D.M. (2017). Bacterial Binding, Phagocytosis, and Killing: Measurements Using Colony Forming Vaiserman, A.M., Koliada, A.K., and Marotta, F. (2017). Gut microbiota: A Units. Methods Mol. Biol. 1519, 297–309. player in aging and a target for anti-aging intervention. Ageing Res. Rev. Puchta, A., Naidoo, A., Verschoor, C.P., Loukov, D., Thevaranjan, N., Mandur, 35, 36–45. T.S., Nguyen, P.S., Jordana, M., Loeb, M., Xing, Z., et al. (2016). TNF Drives Verdu, E.F., Huang, X., Natividad, J., Lu, J., Blennerhassett, P.A., David, C.S., Monocyte Dysfunction with Age and Results in Impaired Anti-pneumococcal McKay, D.M., and Murray, J.A. (2008). Gliadin-dependent neuromuscular and Immunity. PLoS Pathog. 12, e1005368. epithelial secretory responses in gluten-sensitive HLA-DQ8 transgenic mice. Reade, M.C., Yende, S., D’Angelo, G., Kong, L., Kellum, J.A., Barnato, A.E., Am. J. Physiol. Gastrointest. Liver Physiol. 294, G217–G225. Milbrandt, E.B., Dooley, C., Mayr, F.B., Weissfeld, L., and Angus, D.C.; Verschoor, C.P., Naidoo, A., Wallace, J.G., Johnstone, J., Loeb, M., Bramson, Genetic and Inﬂammatory Markers of Sepsis Investigators (2009). J.L., and Bowdish, D.M. (2015). Circulating muramyl dipeptide is negatively Differences in immune response may explain lower survival among older associated with interleukin-10 in the frail elderly. Inﬂammation 38, 272–277. men with pneumonia. Crit. Care Med. 37, 1655–1662. Rera, M., Clark, R.I., and Walker, D.W. (2012). Intestinal barrier dysfunction Whelan, F.J., Verschoor, C.P., Stearns, J.C., Rossi, L., Luinstra, K., Loeb, M., links metabolic and inﬂammatory markers of aging to death in Drosophila. Smieja, M., Johnstone, J., Surette, M.G., and Bowdish, D.M. (2014). The loss Proc. Natl. Acad. Sci. USA 109, 21528–21533. of topography in the microbial communities of the upper respiratory tract in the elderly. Ann. Am. Thorac. Soc. 11, 513–521. Sandler, N.G., Wand, H., Roque, A., Law, M., Nason, M.C., Nixon, D.E., Pedersen, C., Ruxrungtham, K., Lewin, S.R., Emery, S., et al.; INSIGHT Ye, Y. (2011). Identiﬁcation and quantiﬁcation of abundant species from pyro- SMART Study Group (2011). Plasma levels of soluble CD14 independently sequences of 16S rRNA by consensus alignment. Proceedings (IEEE Int. Conf. predict mortality in HIV infection. J. Infect. Dis. 203, 780–790. Bioinformatics Biomed.) 2010, 153–157. Scott, K.A., Ida, M., Peterson, V.L., Prenderville, J.A., Moloney, G.M., Izumo, Yende, S., Tuomanen, E.I., Wunderink, R., Kanaya, A., Newman, A.B., Harris, T., Murphy, K., Murphy, A., Ross, R.P., Stanton, C., et al. (2017). Revisiting T., de Rekeneire, N., and Kritchevsky, S.B. (2005). Preinfection systemic Metchnikoff: Age-related alterations in microbiota-gut-brain axis in the mouse. inﬂammatory markers and risk of hospitalization due to pneumonia. Am. J. Brain Behav. Immun., S0889-1591(17)30034-X. Respir. Crit. Care Med. 172, 1440–1446. Slack, E., Hapfelmeier, S., Stecher, B., Velykoredko, Y., Stoel, M., Lawson, Yende, S., Alvarez, K., Loehr, L., Folsom, A.R., Newman, A.B., Weissfeld, L.A., M.A., Geuking, M.B., Beutler, B., Tedder, T.F., Hardt, W.D., et al. (2009). Wunderink, R.G., Kritchevsky, S.B., Mukamal, K.J., London, S.J., et al.; Innate and adaptive immunity cooperate ﬂexibly to maintain host-microbiota Atherosclerosis Risk in Communities Study; Cardiovascular Health Study; mutualism. Science 325, 617–620. Health, Aging, and Body Composition Study (2013). Epidemiology and long- So¨ derholm, J.D., Streutker, C., Yang, P.C., Paterson, C., Singh, P.K., McKay, term clinical and biologic risk factors for pneumonia in community-dwelling D.M., Sherman, P.M., Croitoru, K., and Perdue, M.H. (2004). Increased epithe- older Americans: analysis of three cohorts. Chest 144, 1008–1017. lial uptake of protein antigens in the ileum of Crohn’s disease mediated by tumour necrosis factor alpha. Gut 53, 1817–1824. Zwielehner, J., Liszt, K., Handschur, M., Lassl, C., Lapin, A., and Haslberger, Stearns, J.C., Davidson, C.J., McKeon, S., Whelan, F.J., Fontes, M.E., A.G. (2009). Combined PCR-DGGE ﬁngerprinting and quantitative-PCR Schryvers, A.B., Bowdish, D.M., Kellner, J.D., and Surette, M.G. (2015). indicates shifts in fecal population sizes and diversity of Bacteroides, biﬁdo- Culture and molecular-based proﬁles show shifts in bacterial communities of bacteria and Clostridium cluster IV in institutionalized elderly. Exp. Gerontol. the upper respiratory tract that occur with age. ISME J. 9, 1246–1259. 44, 440–446. 466 Cell Host & Microbe 21, 455–466, April 12, 2017 STAR+METHODS KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies Anti-mouse F4/80-APC eBioscience Cat#17-4801-82; RRID: AB_469452 Anti-mouse Ly6G-PE BD Biosciences Cat#551461; RRID: AB_394208 Anti-mouse CD45-eﬂuor450 eBioscience Cat#48-0451; RRID: AB_1518806 Anti-mouse CD11b-PeCy7 BioLegend Cat#301321; RRID: AB_830643 Anti-mouse TLR4-FITC eBioscience Cat#53-9041-82; RRID: AB_469944 Anti-mouse TLR2-PeCy7 eBioscience Cat#25-9024-80; RRID: AB_469687 Anti-mouse CD14-PerCpCy5.5 eBioscience Cat#45-0141; RRID: AB_925733 Anti-MARCO-PE AbSerotec Cat#0310 Anti-beta actin Cell Signaling Technologies Cat#4970; RRID: AB_2223172 Bacterial and Virus Strains Streptococcus pneumoniae strain P1547 Prof. Jeffrey Weiser N/A Biological Samples Mouse bone marrow derived or peritoneal macrophages C57BL/6J or Collected in house, age either 10-14 wk B6.129S-Tnftm1Gkl/J (young) or 18-22 mo (old) Chemicals, Peptides, and Recombinant Proteins Adalimumab/Humira Abbott Laboratories N/A Lidocaine powder Sigma Cat# L7757 Human IgG BioLegend Cat# 403102 Tryptic Soy Agar BD Cat# 211822 4kDa FITC-Dextran Sigma Cat#46944 Heparin Sodium Injection 1000 USP Units/mL Sandoz DIN 02303086 Chromium-51 Radionuclide, 1mCi, EDTA Complex in Perkin-Elmer Cat#NEZ14700 0.005M EDTA Eschericia coli 055:B5, ultrapure Invivogen Cat# tlrl-pb5lps Taq polymerase and buffer solution Life Technologies Cat# J00273 Critical Commercial Assays Mouse IL6 ELISA Ready-SET-Go eBioscience Cat# 88-7064 QiaQuick Gel Extraction QIAGEN Cat#28704 Milliplex Catalog ID.MCYTOMAG-70K-02.Mouse Milliplex Cat#MCYTOMAG-70K-02 Cytokine MAGNETIC Kit Deposited Data Microbiome data submitted Bioproject ID: PRJNA379319 Bioproject ID: PRJNA379319 Experimental Models: Cell Lines HEK293T-NOD2/pNifty2-SEAP Created in house N/A Experimental Models: Organisms/Strains Mouse: C57BL/6J 10-14 wk (young) or 18-22 mo Jackson labs 000664 (old), raised under either speciﬁc pathogen free or germ-free conditions Mouse: B6.129S-Tnftm1Gkl/J 10-14 wk (young) or Jackson labs 005540 18-22 mo (old) Recombinant DNA Software and Algorithms Custom, in-house Perl scripts to process the Whelan et al., 2014 Whelan et al. Ann Am Thorac Soc 11, 513-521. sequences after Illumina sequencing PANDAseq Masella et al., 2012 BMC Bioinformatics 13, 1-7. (Continued on next page) Cell Host & Microbe 21, 455–466.e1–e4, April 12, 2017 e1 Continued REAGENT or RESOURCE SOURCE IDENTIFIER Cutadapt Martin, 2011 https://github.com/marcelm/cutadapt Ribosomal Database Project (RDP) classiﬁer Michigan State University https://rdp.cme.msu.edu/classiﬁer/classiﬁer.jsp Greengenes reference database http://greengenes.lbl.gov/cgi-bin/JD_Tutorial/ nph-Alignment.cgi AbundantOTU+ Indiana University http://omics.informatics.indiana.edu/AbundantOTU/ Quantitative Insights into Microbial Ecology (QIIME) Caporaso et al., 2010 Nat Methods 7, 335-336. Other Illumina 16 s rRNA sequencing This paper http://www.science.mcmaster.ca/mobixlab/ CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for resources and reagents should be directed to and will be fulﬁlled by the lead author, Dawn Bowdish ([email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS Ethics statement All experiments were performed in accordance with Institutional Animal Utilization protocols approved by McMaster University’s Animal Research Ethics Board as per the recommendations of the Canadian Council for Animal Care. Mouse Experiments –/– WT young (10-16 wk) and old (18-22 mo) C57BL/6 and TNF mice (originally from Jackson Laboratories), were bred in house. To protect from age-related obesity, aging SPF mice (and corresponding young controls) are fed with a low protein diet Teklad Irradiated Global 14% protein Maintenance Diet and provided with an exercise wheel. The average weight of a young SPF mouse (8-14 wks) in this study is 20 g+/–1g and the old SPF mice (18-22 mo) are on average, 27 g+/–2.5g. Mice were housed in pathogen-free conditions and pathogen-free status of mice within the aging colony was conﬁrmed in mice through constitutive monitoring of sentinel mice and speciﬁc testing of fecal samples for common mouse pathogens. Mice were maintained in the same animal room, with the exception –/– of germ-free and ASF mice (all C57BL/6), which were bred in the Gnotobiotic Facility of McMaster. Because the WT and TNF mice were originally from different breeding colonies from Jackson labs, and the ASF mice had fundamentally different microbiota, –/– changes in the microbiota were made within strains (i.e., differences between young and old WT or young and old TNF or young and old ASF mice) but not between strains. For experiments described in Figures 1, 2, 3, and 5 female mice were used. For exper- iments described in Figure 4 groups of sex matched male and female mice were used. For studies of the microbiota, in order to mini- mize cage effects or familial transfer of the microbiota (as described in (Ubeda et al., 2012)), mice were selected from multiple cages and multiple breeding pairs. No evidence of cage effects was found in any of the studies of the microbiota. Due to the limited avail- ability of aged ASF mice, experiments were performed in groups of 2-4 mice from multiple breeders over 2 years, again minimizing cage effects. METHOD DETAILS TNF ablation Adalimumab (HUMIRA, Abbott Laboratories), a humanized anti-TNF antibody, or the human IgG isotype control diluted in sterile sa- line were administered to mice. A dose of 40 mg per gram of body weight was given intraperitoneally in a volume of 200 ml every other day, for a period of 3 weeks to young and old WT mice. Histological analysis Histopathological analysis was carried out on samples from the lungs of old WT, TNF KO and germ-free mice, and their young con- trols. Upon collection, lungs were formalin-inﬂated and these, alongside formalin ﬁxed spleens, were parafﬁn-embedded. Tissue blocks were cut into 5-mm sections that were stained with hematoxylin-eosin (HE). The slides were blinded/coded and the colon epithelial architecture and inﬂammation were histologically scored. Histological scoring was performed using the following system: tissue architectural changes: 0, normal; 1, blebbing; 2, loss of epithelium; 3, complete loss of crypt architecture; and inﬂammation: 0, normal; 1, increased number of inﬂammatory cells in lamina propria; 2, increased number of inﬂammatory cells in submucosa; 3, dense inﬂammatory cell mass, but not transmural in nature; 4, transmural inﬂammation. The average score for all mice was 0 for both inﬂammation and epithelial cell architectural changes. Cellular inﬁltration in the lungs was quantitated in H&E stained sections. The degree of inﬂammation within each lung was measured by expressing the total area of the cellular inﬁltrate within the lung as a e2 Cell Host & Microbe 21, 455–466.e1–e4, April 12, 2017 percentage of the total lung area using ImageJ. Images were acquired with a Leica DM LB2 microscope at a magniﬁcation of 20X and captured using a Leica DFC 280 camera. Measurement of cytokine production For circulating levels of cytokines, blood samples from naive animals were collected by retro-orbital bleeding into heparin, and spun at 1500 x g for 5 min. 100 mL of plasma was then collected, and IL6 levels assayed using ELISA as per the manufacturer’s direction (eBioscience). To measure the TNF and IL6 cytokine concentrations within the plasma samples of the colonized germ-free mice, Milli- plex immunoassay Kits were used and completed as recommended by the manufacturer’s instructions (Millipore, Etobicoke, ON). For whole blood stimulation studies, 100 mL of whole blood samples collected in heparin from young and old WT, TNF KO and germ-free mice were stimulated with 100 ng/ml of LPS (Eschericia coli 055:B5, ultrapure Invivogen), or left unstimulated. Samples were incubated for 24 hr at 5% CO and 37 C, then centrifuged at 1500 x g for 5 min. 50 mL of plasma samples were assayed for the presence of IL6 using ELISA. To measure constitutive levels IL6 in the lung, right lobe samples of lung were mechanically homog- enized in 500 mL of PBS and assayed by ELISA. To measure inducible cytokine production in lung tissue, lungs were perfused with low melt agarose and sliced into 10 micron sections. 3 slices were cultured in 1 mL of media for 24 hr; supernatants were then removed and assayed for IL6 production using 100 mL of sample ELISA. To measure cytokine production by bone marrow macro- phages, 3.5 3 10 mature bone-marrow-derived macrophages were seeded in a 24-well tissue culture-grade plate (Fisher) in 1.5 mL of media and allowed 24 hr to recover. Cells were then stimulated with either 100 ng/ml of LPS (Eschericia coli 055:B5, ultrapure In- vivogen), whole heat-killed P1547 at an MOI of 50 or 50 mL of media control. Supernatants were collected at 24 hr post-stimulation. Levels of TNF or IL6 were measured by ELISA. Macrophage culture Bone marrow derived-macrophages were isolated according to previously published methods(60) and differentiated in the presence of L929 conditioned media for 8 days as per standard protocols. After 8 days the cells were incubated with 4 mg/ml lidocaine (Sigma) for 15 min at 4 C and gently lifted using a cell lifter. Cells were then centrifuged, counted and re-suspended in medium at a concen- tration appropriate for measurement of cytokine production, bacterial uptake, ﬂow cytometry or bacterial killing assays. Macrophage maturation was assessed by ﬂow cytometry using APC-conjugated anti-F4/80, PE-conjugated anti-Ly6G or -CCR2, FITC-conju- gated Ly6C, eFluor 450-conjugated CD45 and PE-Cy7-conjugated CD11b, or corresponding isotype controls. Pattern recognition receptor (PRR) expression was measured using anti-TLR4-FITC, anti-TLR2-PE-Cy7 and anti-CD14-PerCpCy5.5 (eBioscience), as well as anti-MARCO-PE (RND systems). To visualize S. pneumoniae uptake by macrophages, TRITC labeled bacteria were incubated with bone marrow derived macrophages for 2h at an MOI of 200. Cells were ﬁxed and stained using an anti-beta actin antibody (Cell Signaling). Images were acquired at 40X magniﬁcation using an inverted Zeiss LSM510 laser confocal microscope. Bacterial killing assays To measure macrophage killing of S. pneumoniae, 5 3 10 bone marrow derived macrophages were pre-incubated with an multi- plicity of infection (MOI) of 10 bacteria per macrophage for 60 min at 37 C with gentle inversion as outlined above to allow for inter- nalization of bacteria (Novakowski et al., 2017). Viable CFUs were determined by culturing of supernatants on TS agar plates. In vitro and in vivo permeability Sections of colon and ileum were excised, opened along the mesenteric border, and mounted in Ussing chambers (World Precision Instruments, Sarasota, Florida). Tissues (ileum and colon) were allowed to equilibrate for 15-25 min before baseline values for poten- tial difference (PD) and short circuit current (Isc) were recorded. Tissue conductance (G) was calculated by Ohm’s law using the PD 51 51 and Isc values. Mucosal to serosal ﬂux of the small inert probe (360 Da) -chromium-ethylenediaminetetraacetic acid ( Cr-EDTA) was used to assess paracellular permeability. After equilibration, time zero samples were taken from the serosal buffer and 6mCi/ml CR-EDTA was added to the mucosal compartment. A ‘‘hot sample’’ was taken from the mucosal buffer then samples were then taken every 30 min from the serosal buffer for 2 hr and counted in a liquid scintillation counter (Beckman). Counts from each 30 min were averaged and compared to the ‘‘hot sample’’(100%). Data expressed as mucosal-to-serosal ﬂux (%ﬂux/cm /hr). Each sample was completed in duplicates. Recordings were performed as described previously (Slack et al., 2009; Verdu et al., 2008). For non-terminal studies, tracer FITC-labeled dextran (4kDa; Sigma-Aldrich) was used to assess in vivo intestinal permeability. Mice were deprived of food 4 hr prior to and both food and water 4 hr following an oral gavage using 200 ml of 80 mg/ml FITC-dextran. Blood was retro-orbitally collected after 4 hr, and ﬂuorescence intensity was measured on ﬂuorescence plates using an excitation wavelength of 493nm and an emission wavelength of 518 nm. MDP Detection Bioassay HEK293T cells stably were transfected with mNod2 (a kind gift from Dr. Jonathan Schertzer) and pNifty2-SEAP plasmids (Invivogen) to create a reporter system. Binding of the intracellular mNod2 receptor with its ligand, MDP, results in downstream activation and translocation of NFkB. Activation of this transcription factor leads to SEAP expression via the ELAM proximal promoter, which is de- tected via absorbance spectroscopy. Plates were seeded with cells 24 hr prior to addition of heat-inactivated mouse plasma, diluted Cell Host & Microbe 21, 455–466.e1–e4, April 12, 2017 e3 1 in 200 in HEK Blue Detection Media (Invivogen) to a ﬁnal volume of 200 ml, in a 96-well plate format. Readings were performed at 630nm, 24 hr subsequent to stimulation as described in (Verschoor et al., 2015). Germ-free Mouse Recolonization For recolonization studies, one young and old germ-free mice were transferred to individually ventilated racks and co-housed with either a young or old mouse. Due to the availability of aged germ free mice, 8 independent colonization experiments of 2-6 young or old germ-free mice were performed over 3.5 years. Consequently the SPF mice that were used were from different breeding pairs, ensuring that cage effects or changes speciﬁc to a particular breeding pair were minimized. The mice were left undisturbed for two week following the start of the colonization and then maintained for a minimum of 6 weeks at which point fecal pellets were collected for microbiome analysis (as described below), plasma cytokines were assayed and intestinal permeability was measured as described above. Bacterial proﬁling by deep sequencing analysis of 16S rRNA with Illumina Fecal pellets were collected and the V3 region of the 16S rRNA gene was ampliﬁed by PCR as in Bartram et al. (2011); Stearns et al. (2015); and Whelan et al. (2014). Brieﬂy, each 50 mL PCR reaction mixture contained 1.5 mM of MgCl (50mM), 200 mM dNTPs, 4 mM of BSA, 25 pmol of each primer, 1U of Taq polymerase (Life Technologies), and 200 ng of DNA. The reaction was then run for 30 cycles (94 C for 2 min, 94 C for 30 s, 50 C for 30 C, 72 C for 30 s), with a ﬁnal polymerization step at 72 C for 10 min (Eppendorf). The prod- ucts were separated by electrophoresis in 2% agarose gel and visualized under a UV transilluminator and the products correspond- ing to the ampliﬁed V3 region (300 base pairs) were excised and puriﬁed using standard gel extraction kits (QIAGEN). Illumina sequencing and initial quality control were carried out by the MOBIX-McMaster Genome Center (McMaster University). Custom, in-house Perl scripts were developed to process the sequences after Illumina sequencing (Whelan et al., 2014). Brieﬂy, Cutadapt was used to trim the forward and reverse paired-end reads at the opposing primers for input into PANDAseq for assembly (Martin, 2011; Masella et al., 2012). Mismatches and ambiguous base attributions in the assembly from speciﬁc set of paired end sequences were discarded. Operational taxonomic units (OTUs) were picked using AbundantOTU+ and taxonomy-assigned using the Ribo- somal Database Project (RDP) classiﬁer against the Greengenes reference database (Ye, 2011). OTU number are generated in order from most abundant (OTU 1) when clustering using AbundantOTU +. Single sequence OTUs (singletons) were removed prior to all analyses using Quantitative Insights into Microbial Ecology (QIIME) (Caporaso et al., 2010). QUANTIFICATION AND STATISTICAL ANALYSIS Unless otherwise mentioned in the ﬁgure legend, statistical signiﬁcance was determined by two-way analysis of variance with Fischer’s post-test and unpaired t tests (two tailed). Statistical signiﬁcance was deﬁned as a p value of 0.05. All data were analyzed with Prism (Version 6; GraphPad). Differences in the survival curves were analyzed by Log-rank (Mantel-Cox) test Microbiota changes were analyzed with Quantitative Insights into Microbial Ecology (QIIME) software using principal component analysis as measured by Bray-Curtis. The Chi-square of the likelihood ratio test in phyloseq DESeq2 was used to determine differences between groups as in (McMurdie and Holmes, 2013). In order to avoid the challenges of multiple testing correction, two datasets for the young and old microbiota samples were generated from samples gathered approximately 6 months apart from at least 5 different cages of mice. A list of OTUs representing families or genuses which changed in abundance in old SPF mice was created and statistically signiﬁcant differences in the second dataset were determined using Welch’s unequal variances t test. Data from the second dataset are pre- sented in Tables 1 and 2. No evidence of cage effects was found. Bray-Curtis distances were calculated and interactions between age and treatment were tested using the permanova test ‘adonis’ from the ‘vegan’ package in R. DATA AND SOFTWARE AVAILABILITY All data are available upon request to the lead contact author. No proprietary software was used in the data analysis. The accession number for the data reported in this paper is Bioproject ID: PRJNA379319. e4 Cell Host & Microbe 21, 455–466.e1–e4, April 12, 2017 Update Cell Host & Microbe Volume 23, Issue 4, 11 April 2018, Page 570 DOI: https://doi.org/10.1016/j.chom.2018.03.006 Cell Host & Microbe Corrections Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inﬂammation, and Macrophage Dysfunction Netusha Thevaranjan, Alicja Puchta, Christian Schulz, Avee Naidoo, J.C. Szamosi, Chris P. Verschoor, Dessi Loukov, Louis P. Schenck, Jennifer Jury, Kevin P. Foley, Jonathan D. Schertzer, Maggie J. Larche´ , Donald J. Davidson, Elena F. Verdu´ , Michael G. Surette, and Dawn M.E. Bowdish* *Correspondence: [email protected] https://doi.org/10.1016/j.chom.2018.03.006 (Cell Host & Microbe 21, 455–466; April 12, 2017) The authors would like to clarify that several experiments in the paper as detailed here were performed simultaneously. Speciﬁcally, the histology slides represented in Figures 1G, 2E, and 3H were analyzed by a blinded reviewer and quantiﬁed using the same pathology scale. The experiments yielding data on intestinal permeability (Figures 3D and 5B) and ELISA data (Figures 1E, 1F, 2B, 3G, and 3I) were performed simultaneously to minimize inter-experimental error. The relevant comparisons (e.g., age, genotype, SPF/germ-free) were presented in separate ﬁgures/panels to facilitate the narrative of the manuscript, but the statistical analysis was performed and presented based on analysis of the entire dataset. Additionally, the sentence ‘‘Mice were deprived of food 4 hr prior to and both food and water 4 hr following an oral gavage using 200 ml of 0.8 mg/ml FITC-dextran’’ should state ‘‘80 mg/ml FITC-dextran.’’ The authors apologize for this error, and it has been corrected online. ª 2018 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Dysbiosis-Associated Change in Host Metabolism Generates Lactate to Support Salmonella Growth Caroline C. Gillis, Elizabeth R. Hughes, Luisella Spiga, Maria G. Winter, Wenhan Zhu, Tatiane Furtado de Carvalho, Rachael B. Chanin, Cassie L. Behrendt, Lora V. Hooper, Renato L. Santos, and Sebastian E. Winter* *Correspondence: [email protected] https://doi.org/10.1016/j.chom.2018.03.013 (Cell Host & Microbe 23, 54–64; January 10, 2018) In the original publication, Tables S1–S3 in the Supplemental Information ﬁle were inadvertently omitted. A corrected Supplemental Information ﬁle has been made available at the journal website. The authors apologize for any inconvenience this may have caused. ª 2018 Elsevier Inc. 570 Cell Host & Microbe 23, 570, April 11, 2018

http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Cell Host & Microbe Unpaywall

http://www.deepdyve.com/lp/unpaywall/age-associated-microbial-dysbiosis-promotes-intestinal-permeability-XJ0SS47vc8

Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inflammation, and Macrophage Dysfunction

Loading next page...

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher: Unpaywall
ISSN: 1931-3128
DOI: 10.1016/j.chom.2017.03.002
Publisher site: See Article on Publisher Site

Abstract

Journal

Cell Host & Microbe – Unpaywall

Published: Apr 1, 2017

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inflammation, and Macrophage Dysfunction

Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inflammation, and Macrophage Dysfunction

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inflammation, and Macrophage Dysfunction

Age-Associated Microbial Dysbiosis Promotes Intestinal Permeability, Systemic Inflammation, and Macrophage Dysfunction

References

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies