www.nature.com/npjbiofilms ARTICLE OPEN Vaccination with an HIV T-cell immunogen induces alterations in the mouse gut microbiota Alessandra Borgognone 1,10, Aleix Elizalde-Torrent 1,10, Maria Casadellà 1, Luis Romero1,2, Tuixent Escribà 1, Mariona Parera1, Francesc Català-Moll1, Marc Noguera-Julian1,3,4, Christian Brander 1,3,4,5,6, Alex Olvera 1,3,4,11 ✉ and Roger Paredes 1,2,3,4,7,8,9,11 ✉ The gut microbiota is emerging as a crucial factor modulating vaccine responses; however, few studies have investigated if vaccines, in turn, can alter the microbiota and to what extent such changes may improve vaccine efficacy. To understand the effect of T-cell vaccination on the gut microbiome, we administered an HIV-1 T-cell immunogen (HTI arm) or PBS (control, Mock arm) to C57Bl/6 mice following a heterologous prime-boost scheme. The longitudinal dynamics of the mice gut microbiota was characterized by 16 S ribosomal RNA sequencing in fecal samples collected from cages, as well as from three gut sections (cecum, small and large intestine). Serum and spleen cells were obtained at the last time point of the study to assess immune correlates using IFNγ ELISPOT and cytokine Luminex® assays. Compared with Mock, HTI-vaccinated mice were enriched in Clostridiales genera (Eubacterium xylanophilum group, Roseburia and Ruminococcus) known as primary contributors of anti-inflammatory metabolites, such as short-chain fatty acids. Such shift was observed after the first HTI dose and remained throughout the study follow-up (18 weeks). However, the enriched Clostridiales genera were different between feces and gut sections. The abundance of bacteria 1234567890():,; enriched in vaccinated animals positively correlated with HTI-specific T-cell responses and a set of pro-inflammatory cytokines, such as IL-6. This longitudinal analysis indicates that, in mice, T-cell vaccination may promote an increase in gut bacteria known to produce anti-inflammatory molecules, which in turn correlate with proinflammatory cytokines, suggesting an adaptation of the gut microbial milieu to T-cell-induced systemic inflammation. npj Biofilms and Microbiomes (2022)8:104 ; https://doi.org/10.1038/s41522-022-00368-y INTRODUCTION between the pre-existing gut microbiota and T cell responses to It has been well established that the resident gastrointestinal vaccination13. microbiota plays a pivotal role in the development and regulation The effect of vaccine administration on the intestinal microbiota of both innate and adaptive immune responses1. Of the different has gathered even less attention. Previous works investigating the determinants of the adaptive immune response, T cell response effects of HIV-114,15 and oral typhoid16 vaccines on the human gut polarization might be especially affected by the gut bacteria microbiota reported no discernible perturbations of the microbial composition2. In fact, the absence of specific bacteria has been community structure induced by vaccination. In contrast, an associated with reduced frequencies of Treg cells in the gut3 or increase in the Firmicutes/Bacteroidetes ratio following HIV-1 DNA/ dysregulated Th17 cell responses in germ-free and antibiotic- protein immunization was observed in the rectal microbiome of treated mice4. Abnormal ratios of these T cell subpopulations non-human primates17. Also, a recent study comparing two SARS- (Th17 and Treg) have been frequently associated with metabolic CoV-2 vaccination strategies reported changes in the human gut or immunologic diseases5. microbiota 1 month after immunization, characterized by lower Given the tight interaction and co-evolution of gut microbiota alpha diversity and increased abundance of a certain bacterial and immune system6, increasing evidence suggests that certain species, such as Bacteroides caccae18. Collectively this provides at bacterial taxa might also modulate vaccine-induced immune least circumstantial evidence that vaccination can influence the responses7, by exerting local or systemic effects8. In recent years, resident microbiota. Therefore, understanding how vaccines several clinical and animal studies have shown that the composi- promote changes in the host intestinal immune milieu and alter tion and functions of the gut microbiota are crucial determinants the microbial community structure, may provide important insight of the immunogenicity and efficacy of oral and injectable into this interplay and, in turn, the identification of specific vaccines9,10. So far, most work in this area has been focused on bacteria that positively modulate vaccination outcome. the role of the microbiota in modulating vaccine-induced In the present study, HIV-1 T-cell immunogen (HTI) expressing antibody responses7. Of note, the importance of T cell-mediated vectors19,20 were used to assess vaccine-induced changes in the immunity has been increasingly recognized in the protection gut microbiota and to identify potential correlates with the induced by several immunization strategies, including those immunological response to vaccination in a murine model. The against pathogens such as HIV-111 or SARS-CoV-212. Yet, a limited HTI immunogenicity has been tested in both mice21,22 and number of studies have directly investigated the relationship humans23 using different vaccine vectors (such as DNA, 1 IrsiCaixa AIDS Research Institute, Badalona, Spain. 2Universitat Autònoma de Barcelona (UAB), Barcelona, Catalonia, Spain. 3Universitat de Vic-Universitat Central de Catalunya (UVic-UCC), Vic, Spain. 4Ciber of Infectious Diseases CIBERINFEC, ISCIII, Madrid, Spain. 5Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain. 6AELIX Therapeutics, Barcelona, Spain. 7Center for Global Health and Diseases, Department of Pathology, Case Western Reserve University, Cleveland, OH, USA. 8Fundació Lluita contra les Infeccions, Hospital Universitari Germans Trias i Pujol, Badalona, Catalonia, Spain. 9Department of Infectious Diseases, Hospital Universitari Germans Trias i Pujol, Badalona, Catalonia, Spain. 10These authors contributed equally: Alessandra Borgognone, Aleix Elizalde-Torrent. 11These authors jointly supervised this work: Alex Olvera, Roger Paredes. ✉email: aolvera@irsicaixa.es; rparedes@irsicaixa.es Published in partnership with Nanyang Technological University A. Borgognone et al. 2 Fig. 1 Study design and sample collection. Female and male mice were immunized with an HIV T-cell vaccine (three DNA.HTI primes followed by a ChAd.HTI and MVA.HTI booster; n = 20) or a mock vaccine (PBS; n = 20). Antibiotic conditioning followed by fecal microbiota transfer was performed prior to vaccination to homogenize the intestinal content across all animals. Collection of samples, including stool, luminal content, serum and spleen was completed at the indicated time points. Abbreviations: FMT fecal microbiota transfer, Atb Antibiotic, PBS phosphate-buffered saline. The Figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license. Chimpanzee Adenovirus, ChAdOx1; Modified Vaccine virus experiment (Fig. 2a). Samples from both fecal and luminal content Ankara, MVA and Bacillus Calmette–Guérin, BCG). Based on this were dominated by Muribaculaceae (49.8% and 57.9%), Lachnos- previous evidence, mouse strain C57Bl/6 and a three DNA prime piraceae (14.2% and 11.4%), Bacteroidaceae (9.4% and 4.8%) and followed by a ChAdOx.1—MVA boost were selected to achieve Bifidobacteriaceae (7.1% and 6.2%) (Fig. 2a and Supplementary Fig. maximal immunogenicity. 16 S rRNA gene sequencing was used to 6a), except for the small intestine (Supplementary Fig. 6b). In this characterize longitudinal fecal microbiome and intestinal content gut section, reduction in Lachnospiraceae NK4A136 group (0.2%) profiles in mice receiving a prime-boost HTI immunization and Bacteroides (0.05%) was concomitant to an overall increase in regimen or PBS (control group). Correlations between gut Lactobacillus (7.7%) and Ligilactobacillus (3.7%) (Supplementary microbial signatures, T cell vaccine response and cytokines after Fig. 6c, d). Despite overall compositional similarities during the 1234567890():,; full immunization were also assessed. vaccine administration (0 week–15 week, Fig. 2a and Supplemen- tary Fig. 6e), samples clustered separately by experimental group RESULTS (p = 0.022), but not by timepoint within each group (p = 0.594) Study design and sequencing analysis (Fig. 2b), as indicated by both univariate and multivariate analyses To investigate the effects of HTI vaccination on the gut microbiota, (Supplementary Table 5). Cage and Sex covariates also showed a two experimental groups receiving the HTI immunogen (Immu- significant impact on the microbiota composition (Supplementary nized, n = 20) or PBS (Mock, n = 20) were compared (Fig. 1). Fecal Table 5). Despite the lack of individual-level contribution in our samples were collected longitudinally from cages and from three assessments, a trend toward higher Bray–Curtis dissimilarity in the gut sections for gut microbiome profiling; serum and splenocytes Immunized compared to the Mock group was observed (Supple- were obtained at the end of the study to assess vaccine mentary Fig. 6f), potentially indicating a strong impact on the gut immunogenicity and cytokine profiles, respectively (Supplemen- microbiota of vaccinated mice. Also, no differences in alpha tary Fig. 1). Two weeks before vaccination, the mice gut diversity (Shannon index) were observed between groups during microbiota was homogenized by antibiotic conditioning followed vaccination (Supplementary Fig. 6g). Clustering by experimental by mouse-to-mouse fecal microbiota transfer (FMT) to reduce group (p = 0.021) and intestinal section (p = 0.001) was also baseline variability, as detailed in Supplementary Text and observed at the last time point of the study, being samples from Supplementary Fig. 2. Baseline mouse characteristics at animal small intestine separated from those from caecum and large arrival (-3w; Supplementary Tables 1 and 2) and pre-vaccination intestine (Fig. 2c). Moreover, samples from the small intestine (0w; Table 1 and Supplementary Table 3) are reported. showed lower alpha diversity in both Mock and Immunized Mock and Immunized longitudinal samples had no differences groups when compared to caecum and large intestine, although in the number of high-quality filtered reads, with a median of no differences between the two experimental groups were 52,092 and 46,584 reads, respectively (Supplementary Fig. 3a and detected (Fig. 2d). Table S4). Negative extraction (n = 12) and no-template PCR (n = 11) controls had significantly lower number of reads (5966 and 1444 mean read number, respectively) compared to the experimental groups (Supplementary Fig. 3a). Despite the Gut microbiota alterations associated to HTI vaccination sequencing depth was higher in batch 1 (Supplementary Fig. The longitudinal discriminant analysis showed that the abundance 3b), no substantial impact on data structure was observed after of specific bacteria changed after vaccination. While only the A2 assessing for differences between the two batches (Supplemen- genus was differentially abundant between experimental groups tary Fig. 4) (batch was included as a covariate in testing for pooled at the baseline (Supplementary Fig. 5a), the Immunized group fecal samples collected from cages). showed an enrichment in a set of Clostridiales genera after vaccination (Fig. 3). The most remarkable observation was an Longitudinal assessment of microbiota composition and enrichment in Ruminococcus in fecal samples from cages (week diversity after HTI vaccination 3–15) and Roseburia in gut sections (caecum and large intestine), Prior vaccination, samples displayed similar bacterial composition while Eubacterium xylanophilum group remained consistently (0 week, Supplementary Fig. 2) and only the genus A2 was higher in fecal samples from cages and gut sections after week increased in the Mock group (Supplementary Fig. 5a). After 3. Also, Ligilactobacillus was increased specifically in samples from segregating by sex, female mice showed higher abundance of the small intestine (Fig. 3). Longitudinal assessment of such Lactobacillus and Muribaculum before vaccination (Supplementary genera abundance indicated an increasing trend starting at Fig. 5b). vaccination in the Immunized group (Supplementary Fig. 7). At family-level, the overall microbial community composition Conversely, no specific genera were consistently higher in the between Mock and Immunized groups was similar over the Mock group over time. npj Biofilms and Microbiomes (2022) 104 Published in partnership with Nanyang Technological University A. Borgognone et al. 3 Bacteroides (19.2%), Bifidobacterium (10.4%), Lachnospiraceae NK4A136 group (10.1%), Alistipes Correlations between gut microbial signatures and Bacteroides (11.7%), Lachnospiraceae NK4A136 group (11.7%), Bifidobacterium (6.6%), Alistipes Bacteroides (11.4%), Lachnospiraceae NK4A136 group (7.9%), Bifidobacterium (6.2%), Alistipes Lachnospiraceae NK4A136 group (14.1%), Bifidobacterium (7.4%), Bacteroides (6.7%), Alistipes immunological responses To examine the HTI immunogenicity, IFNγ assay assessing Bacteroides (16.3%), Lachnospiraceae NK4A136 group (12.2%), Bifidobacterium (5.1%), Bacteroides (14.1%), Bifidobacterium (12.9%), Lachnospiraceae NK4A136 group (8.5%), Lachnospiraceae NK4A136 group (10.8%), Bifidobacterium (5.7%), Bacteroides (5.3%), magnitude and breadth of the specific T cell response against HTI peptides was performed in splenocytes collected at the last time point of the study. No differences in the background IFNγ- producing splenocytes were found between Mock and Immunized mice, indicating that the vaccination regimen did not induce noticeable unspecific T cell activation (Supplementary Fig. 8a). After specific re-stimulation with peptide pools covering the HTI sequence, Mock group did not show any IFNγ response above the thresholds. In contrast, Immunized mice showed strong responses, in both magnitude (median = 1560 IFN-γ SFC/million splenocytes, p = 0.00031, Supplementary Fig. 8b) and breadth (median = 8.0 out of 17 positive pools, p = 0.00028, Supplementary Fig. 8c) compared to Mock. Lactobacillus (3.4%), Faecalibaculum (2.1%) After assessing serum levels of 22 cytokines in Mock and Immunized groups (Supplementary Fig. 9a), IL-6 was the only Lactobacillus (3.1%), Mucispirillum (2.7%) cytokine showing an increasing trend in the Immunized group Lactobacillus (3.4%), Turicibacter (1.7%) after segregating by sex (p = 0.052) (Supplementary Fig. 9b), although significance did not pass multiple testing correction (BH- adjusted p = 0.69). Specifically, females from the Immunized (3.2%), Lactobacillus (1.6%), group showed lower IL-6 levels than males in both Immunized (3.2%), Lactobacillus (2.1%) (1.8%), Lactobacillus (3.9%) (2.5%), Anaerostipes (1.7%) Sequencing Batch Microbiota composition2 and Mock groups (Supplementary Fig. 9b). We next examined potential links between HTI-specific responses, serum cytokine levels and discriminant bacterial genera. Eubacterium xylanophi- lum group and Roseburia from the three intestinal sections were positively correlated with vaccine-induced HTI-specific T cell response (Fig. 4), measured as magnitude and breadth. In addition, most of the tested cytokines showed positive correla- tions with Eubacterium xylanophilum group, Roseburia and Ruminococcus, although there were some exceptions (i.e., IL-33). An opposite trend was observed for Ligilactobacillus, which overall showed negative associations with the serum cytokine levels (Fig. percentage of top 5 most abundant taxa; complete list of taxa is provided in Supplementary Table 3. 4). Of note, IL-6 was positively associated with Eubacterium xylanophilum group, Ruminococcus and especially Roseburia in the large intestine. Moreover, Roseburia showed positive correlations B1 B1 B2 B2 B1 B1 B2 with other cytokines, including GM-CSF, IL-2, IFNγ, IL-17A and IL-25 (small intestine), IL-27 (caecum and large intestine) and IL-22 Age (weeks) Cage N mice Sex Weight (g ± SD) (large intestine). Significant correlations were confirmed by individual scatterplots (Supplementary Fig. 10). Moreover, Eubac- 19.3 ± 0.9 25.4 ± 1.4 25.3 ± 1.3 19.5 ± 0.9 17.2 ± 0.6 terium xylanophilum group, Roseburia and Ruminococcus from the 19.4 ± 1 25.6 ± 2 three intestinal compartments clustered close to the HTI-specific Baseline characteristics at baseline (0 week, pre-vaccination). responses (magnitude and breadth) in a hierarchical clustering, 11 also displaying positive associations with a group of cytokines (Supplementary Fig. 11). M M M F F F F DISCUSSION 5 5 5 5 5 5 5 To date, limited studies have investigated how vaccine adminis- C19 C20 C31 C32 C21 C22 C29 tration may alter resident gut microbial communities. Herein, we used a prime-boost regimen combining the HTI insert in different vaccine vectors, to induce strong HTI-specific responses and study their impact on the microbiota. These results showed, in mice, an increase in Clostridiales bacteria following HIV T-cell immunization, which in turn associate with proinflammatory cytokines. Except for C57BL/6JOlaHsd 9 C57BL/6JOlaHsd 9 C57BL/6JOlaHsd 9 C57BL/6JOlaHsd 9 Immunized C57BL/6JOlaHsd 9 Immunized C57BL/6JOlaHsd 9 Immunized C57BL/6JOlaHsd 9 Eubacterium xylanophilum group, Clostridiales genera found in feces were different from those in the three gut sections in the Mouse Line Immunized group. Such differences might be related to the presence of distinct species in the intestinal epithelium compared mean grams and sd. to the lumen24 and the presence of a 50-60% spore forming bacteria, like Ruminococcaceae and Lachnospiraceae, most likely to be present in feces25. Remarkably, associations between enriched bacteria and serum cytokines varied across the three intestinal Table 1. sections, being representative of distinct oxygen requirements Group Mock Mock Mock Mock and a different crosstalk between mucosal cells and co-existing microbial communities26. The most prominent difference was 1 2 Published in partnership with Nanyang Technological University npj Biofilms and Microbiomes (2022) 104 A. Borgognone et al. 4 Fig. 2 Temporal dynamics in microbiome composition after vaccination. a Stacked bar chart displaying the relative abundance of bacterial families (top 15 most abundant) at different timepoints in HTI-vaccinated (Immunized) or PBS-vaccinated (Mock) mice. CageID (pooled feces) and individual samples for each cage (intestinal content) are reported in the x-axis. PCoA biplot based on the first and second components (Bray–Curtis distances at genus level) of (b) fecal samples from cages (c) and from gut sections (small intestine, caecum and large intestine). Biplots report genera with the top 5 effects on the community composition, with vector position indicating the direction of the effect. Arbitrary ellipses were drawn to facilitate the interpretation of the figure. PERMANOVA r-squared and p-values are shown. d Alpha diversity based on Shannon entropy index (ASV level) of individual fecal samples obtained at the last point of the study from small intestine, caecum and large intestine. Boxplots are colored by the experimental group. Median values and interquartile ranges are shown in boxplots. p-values from Wilcoxon signed rank tests are reported. observed in the small intestine, which was depleted in obligate Our data indicated that bacteria increased after HTI vaccination anaerobes bacteria, such as Ruminococcacae and Lachnospiraceae, (Eubacterium xylanophilum group, Roseburia and Ruminococcus), compared to the caecum and large intestine, as previously located in the three distinct gut sections, positively correlated with reported27. HTI-specific T cell response and a group of serum cytokines. Supporting our results, a recent study in rhesus macaques Specifically, the genus Roseburia, known as a butyrate-producing showed that vaccination with three HIV-1 Env expressing DNA bacteria30, in the caecum and large intestine, positively correlated plasmid vaccines followed by a gp140 protein booster induced with the magnitude and breadth of HTI specific T-cell response in changes in the rectal microbiota, resulting in increased Firmicutes/ spleen cells. At serum level, Roseburia correlated with IL-27, which Bacteroidetes ratio as well as associations of rectal anti-HIV-1 IgG is produced by antigen presenting cells and regulates B and vaccine responses with increased Clostridium IV and reduced T cells31. Interestingly, despite its different effector function, IL-27 Prevotella17. Enrichment in Ruminococcus in the intestinal micro- shares the gp130 receptor subunit with IL-632 and both cytokines biota of piglets was also reported after immunization with and IL-22 were associated with the content of Roseburia in the Lawsonia intracellularis vaccine28. In parallel, increased Bacteroides large intestine. Although the levels of Roseburia in small intestine abundance along with alterations in the microbial community did not correlate with HTI-specific T cell responses, strong structure were found following Mtb vaccine (with ESAT6 correlations were observed with inflammatory cytokines (IL-6, adjuvanted with TLR8 agonists) administration in mice29. Similarly, GM-CSF), two Th1 polarization cytokines (IFNγ and IL-2) and the an enrichment in Bacteroides caccae and reduction in clostridial Th17 polarization cytokines (IL-17A and IL-22). However, none of species, such as Coprococcus comes, Dorea longicatena and these cytokines were differentially produced between the Mock Ruminococcus obeum were reported in Covid-19 vaccinees18. Such and the Immunized group and only IL-6 showed a marginal apparently conflictive trends indicate that a more thorough significance, which did not pass multiple comparison correction. understanding is needed to decipher specific microbial shifts We hypothesize that the abundance of Roseburia may increase induced by distinct immunization strategies. due to T cell vaccine stimulation and relate to Th1 responses, as npj Biofilms and Microbiomes (2022) 104 Published in partnership with Nanyang Technological University A. Borgognone et al. 5 Fig. 3 Discriminative genera between HTI- and PBS-immunized mice following vaccination. LEfSe analysis comparing the relative abundance of bacterial genera between Immunized and Mock groups during vaccine administration (pooled fecal samples at 3 week, 9 week, 6 week and 15 week) and at end of study at week 15 for individual fecal samples from small intestine, caecum and large intestine. Horizontal bars represent the effect size for each discriminant genus (p < 0.05 and LDA scores > 2.0). The bar length represents the log10 transformed LDA score, indicated by vertical lines. indicated by positive correlation with IFNγ and IL-2 in plasma, but in turn, alter the associated microbiota structure38. Furthermore, most remarkably with the HTI-specific T cell responses in spleen immune cells, such as dendritic cells or regulatory T cells, activated cells. The increase in Roseburia may be also related to the at the site of infection, can mediate inflammatory signaling to the presence of inflammation mediators such as IL-633, likely lymph nodes, including mesenteric lymph nodes or the Peyer produced after vaccination and Th17 (IL-17A and IL-22) cytokines Patches39. in serum, which could be affecting mucosal immune responses in Notably, bacteria correlated with pro-inflammatory cytokines the gut. Additional butyrate producing Clostridiales genera34 (such as IL-6, IL-27 and IFNγ) in our study (Roseburia, Ruminococcus enriched in vaccinated animals showed positive correlations with and Eubacterium) are known as primary contributors of anti- the HTI-specific T cell responses (Eubacterium xylanophilum group inflammatory metabolites, mainly short chain fatty acids (SCFAs), in small intestine and caecum) and serum levels of IL-6 in the gut40. This finding may appear inconsistent with previous (Ruminococcus and Eubacterium xylanophilum group in large evidence, such as in inflammatory bowel diseases, in which SCFAs intestine). Also, although Ligilactibacillus did not associate with T inhibit the production of pro-inflammatory cytokines41; although, cell responses, correlations with cytokines representatives of Th17 several mechanisms by which commensal bacteria can adapt to and Treg population (IL-10 and IL17A), known as gut inflammation inflamed gut environment have been discussed elsewhere42. For biomarkers35, might indicate specific changes in the microbiota example, specific components of the gut microbiota have been composition. While the exact mechanism by which HTI- implicated in the production of pleiotropic cytokines by innate immunization would induce changes in the intestinal milieu and immune cells, such as IL-6, and subsequent expansion of Th17 promote an increase of specific bacteria remains unclear, cytokine cells43, which are critical for vaccine-induced memory immune signaling after innate immunity and T cell activation by vaccina- responses44. Also, flagellin and peptidoglycans produced by gut tion is a plausible hypothesis. In fact, intramuscular vaccine bacteria, such as Ruminococcus spp.45, can be sensed by pattern administration may induce local ‘depot’ effect. This consists in a recognition receptors (PRRs) expressed by T cells and B cells and prolonged antigen release at the site of injection, which induce act as natural adjuvants to vaccines7. In fact, higher abundances of persistent stimulation of the immune system and recruitment of bacteria with flagella and fimbriae, including butyrate-producing macrophages that, in turn trigger activation of inflammatory such as Roseburia and Eubacterium, can boost vaccine immuno- mediators36. Of note, administration of viral vaccine vectors rich in genicity serving as adjuvants through immunomodulatory TLR Pathogen Associated Molecular Patterns (PAMP), as used in this agonists46. study, can also trigger the activation of Pathogen Recognition Notably, members of these bacterial groups, such as Roseburia Receptors (PRRs), stimulating the production of pro-inflammatory intestinalis and Eubacterium hallii have been listed as probiotic cytokines, such as IL-1 and TNF-α by dendritic cells and candidates by the International Scientific Association for Probiotics macrophages37. Inflammation mediators can then migrate and Prebiotics47. Indeed, some butyrate-producing bacteria, through the bloodstream and lymph vessels to reach intestinal including those identified in this study, have been shown to tissues and promote local changes in the immune response and, modulate the colonic luminal metabolome, regulate T-cell Published in partnership with Nanyang Technological University npj Biofilms and Microbiomes (2022) 104 A. Borgognone et al. 6 Fig. 4 Correlation analysis between discriminant bacteria, serum cytokines and immunological responses at the last time point of the study. Spearman rank correlations between bacterial genera (only bacteria longitudinally enriched in HTI-immunized mice were evaluated), HTI immunogenicity (magnitude and breadth of the response) and serum cytokine levels. Associations are assessed for each gut section (small intestine, caecum and large intestine). Cytokines for each gut section are ordered based on unsupervised hierarchical clustering. Color and size of the square represent the magnitude of the correlation. Significance is indicated by white asterisks (*p < 0.05; **p < 0.01; ***p < 0.001 after Benjamini–Hochberg adjustment for multiple comparisons). responses and enhance regulatory T-cell functions48. Collectively, to T-cell vaccination and, ultimately provide new potential targets these data prompted the hypothesis that bacteria producing anti- for optimizing vaccine efficacy. inflammatory molecules may adapt to increased local inflamma- tion following vaccination, colonize the intestinal milieu and thrive in it, thereby modulating and quenching the inflammatory METHODS processes. However, further work is needed to understand how Mice and housing vaccine-induced inflammation promotes an increase of anti- Six-week-old C57BL/6JOlaHsd mice (n = 40, 20 females and 20 inflammatory bacteria in the gut. In this context, Th17/Treg cell males), grown in the same breeding facility, were purchased from balance, T cell polarization and activation profiles, ILC responses Envigo and housed under specific pathogen-free (SPF) conditions and Kyn/Trp ratio49 in the gut tissues are promising candidates for at the Centre for Comparative Medicine and Bioimage (CMCiB) future validations. Also, additional omics approaches, including animal facility at IGTP in Badalona, Spain. Mice were randomly metabolite and targeted SCFA measurements, would help to assigned to cages according to a combination of sex (female/ identify potential anti-inflammatory pathways induced after HTI male) and treatment group (Mock/Immunized), with a maximum vaccination. We fully acknowledge the limitation of the small of 5 animals per cage (Supplementary Fig. 1) and allowed to sample size and lack of individual-level contribution in this study. acclimate for one week before the initial treatment. Also, the lack of longitudinal cytokine measurement did not allow us to identify temporal variations of immune correlates with specific bacterial signatures at each stage of the vaccine regimen. Pre-intervention gut microbial homogenization Despite a group of bacteria showed strong correlations with HTI- To reduce potential baseline differences in the microbiota specific T cell responses, another limitation was that the composition, the mice gut microbiota was homogenized by experimental design did not allow to differentiate to which extent antibiotic conditioning followed by mouse-to-mouse FMT (1 week vaccine-induced microbiota changes were due to an adjuvant after arrival and 2 weeks prior to vaccination). The antibiotic effect of the vectors or to insert-specific responses. Future treatment consisted of a combination of ampicillin, amikacin, experiments including a control arm using empty vectors will metronidazole and vancomycin (10 mg/ml each antibiotic), help to discriminate the direct contribution of the vaccine insert administered via oral gavage (200 µl/animal) for 5 days prior to on gut microbiota changes. In addition, another intrinsic limitation FMT. Mixed stools from five male and five female donor mice was the inability of the 16 S rRNA gene sequencing to distinguish (cages MF1 and MM1, Supplementary Fig. 1) were used for FMT in live organisms from transient microorganism colonization50. all animals in the study. Briefly, feces from donors were collected, Finally, host-specificity of the core gut microbiota may limit pooled, aliquoted (350 mg vials) and preserved at −80 °C until use. potential extrapolation of these results from a mouse model to On the day of the fecal transfer, thawed feces were resuspended humans. Thus, clinical trials including microbiota characterization in sterile physiological saline serum (2.3 ml per vial), homogenized, are needed to validate such findings in the human gut microbiota. centrifuged (500 × g, 1 min) and supernatant administered by oral In summary, our data suggest that gut bacteria can adapt to gavage (100 µl/animal). Bacterial load in FMT was measured using vaccine-induced inflammation. Also, they establish a framework the LIVE/DEAD® BacLight™ Bacterial Viability and Counting Kit for future studies to fully capture the dynamics of microbial shifts (Invitrogen, Carlsbad, MA, USA). npj Biofilms and Microbiomes (2022) 104 Published in partnership with Nanyang Technological University A. Borgognone et al. 7 Immunization strategy CA, USA). PCR were performed in 25 μL reaction volume, contain- Animals were vaccinated with the HIV T-cell immunogen (HTI), a ing 2.5 μL of DNA template, 12.5 μL of KAPA HiFi HotStart Ready synthetic protein designed as the fusion of different HIV-1 protein Mix (KAPA HiFi HotStart DNA Polymerase, buffer, MgCl2, and fragments associated with T-cell control of HIV infection20. Briefly, dNTPs, KAPA Biosystems Inc., Wilmington, MA, USA) and 5 μL of the HTI sequence encodes for a 529 amino acid immunogen, each primer at 1 μM. Thermocycling conditions were as follows: designed as the concatenation of 16 clade B consensus HIV Gag, initial denaturation step at 95 °C for 3 min, followed by 30 cycles of Pol, Vif and Nef protein fragments, which were preferentially denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, extension targeted by people showing natural control of HIV infection, at 72 °C for 30 s and a final extension step at 72 °C for 10 min. linked by alanine triplets. The amino acid sequence was back- Once the expected amplicon size was confirmed on a 1.0% translated into nucleotides, codon usage optimized for expression agarose gel electrophoresis (~460 base pairs), PCR products were in human and the resulting open reading frame (ORF) synthetized stored at −30 °C until sequencing library preparation. Negative to be inserted in different vectors22. Specifically, it was inserted extraction and PCR controls (blank controls using DNA-free water into a plasmid DNA under the control of a CMV promotor as the template) were loaded to assess for potential contamina- (DNA.HTI, D) and into two viral vectors: Chimpanzee Adenovirus tion and processed in the same conditions as any other sample. Ox1 (ChAdOx1.HTI, C) and Modified Vaccinia Ankara (MVA.HTI, Amplified DNA templates were cleaned-up with AMPure XP M)22. HTI vaccine immunogenicity has been previously tested in reagent (Beckman Coulter Life Sciences, Indianapolis, USA) for C57BL/6 mouse model and non-human primates22 as well as in non-DNA molecules and Illumina sequencing adapters and dual humans23. In this study, twenty mice (10 females and 10 males) indices were attached using Nextera XT index Kit (Illumina Inc.), were immunized intramuscularly (caudal thigh muscle) with a followed by a corresponding PCR amplification program as heterologous prime-boost regime consisting of three different described in MiSeq 16 S rRNA Amplicon Sequencing protocol (Illumina Inc., San Diego, CA, USA). After the second round of vaccine vectors expressing HTI: (i) three DNA.HTI primes (100 μg/ clean-up, amplicon libraries were quantified using a Quant-iTTM animal), (ii) one boost with ChAd.HTI (1 × 109 VP/animal), and (iii) a PicoGreen® dsDNA Assay Kit (Invitrogen, Carlsbad, MA, USA) and second boost with MVA.HTI (1 × 106 pfu/animal). DNA.HTI diluted in equimolar concentrations (4 nM) for further pooling. vaccinations were administered at weeks 0, 3, and 6, followed Pooled libraries were sequenced on an Illumina MiSeqTM platform by ChAd.HTI and MVA.HTI, separated by a 6-week interval (weeks (Illumina Inc., San Diego, CA, USA) at the genomics core facility in 9 and 15, respectively) (Fig. 1 and Supplementary Fig. 1). Mice in Germans Trias i Pujol research campus (Badalona, Spain) using a the mock group (n = 20, 10 females and 10 males) were paired-end 300 base-length protocol. vaccinated with PBS as a control group, following the same immunization schedule. Microbiome data analysis Sample collection and processing The quality of raw sequencing data was estimated using FastQC51 and the analysis of 16 S rRNA sequences performed using the Microbiota data were generated in two batches (samples DADA2 pipeline (v1.10.1)52. The pipeline was executed according extracted and sequenced in 2019 and 2020), as illustrated in to default parameters using maxEE = 4.10 in the filtering step. Supplementary Figs. 1 and 2. The first sequencing batch included After chimeric reads removal, unique amplicon sequence variants pooled fecal samples longitudinally collected from cages, whereas (ASV) were assigned a taxonomy and aligned to the SILVA rRNA the second batch included longitudinal pools of feces from cages gene database53. Downstream analyses were conducted in the R and individual fecal samples obtained from different sections of environment (v3.5.2)54 using multiple packages, including phylo- the intestinal tract at the last time point of the study. Longitudinal seq (v1.26.1)55, vegan (v2.5-5)56 ade4 (v1.7-13)57 and ggplot2 fecal samples were collected (one pool per cage, n = 5) at animal (v3.2.0)58. ASV counts were normalized and transformed to relative arrival (week-3), following antibiotic conditioning prior to FMT abundance using the phyloseq function transform_sample_counts (week-2) and prior each vaccination (weeks 0, 3, 6, 9 and 15). (100 × (x/sum(x))). Alpha diversity (Shannon entropy index) was Donor pool fecal sample (week-1) was also collected and evaluated using the estimate_richness on rarefied ASV counts. Beta sequenced in batch 1. Three weeks after the last immunization diversity was calculated based on Bray–Curtis distances (normal- (week 18), mice were euthanized and luminal contents from three ized data) and distance matrix used to perform an ordination intestinal sections (small, caecum and large intestine), whole principal coordinate analysis (PCoA). blood and spleen were harvested. Serum was separated from the whole blood by centrifugation (10,000 × g, 5 min) in Serum Gel S/ Mouse IFNγ ELISPOT assay 1.1 tubes (Sarstedt) and frozen at −80 °C until use. Splenocytes were isolated by mechanical disruption and pressed through a cell ELISPOT assays were performed using the mouse IFNγ ELISPOT kit strainer (Falcon) using a 5 ml-syringe rubber plunger. Following (ALP) (Mabtech AB, Stockholm, SE), following the manufacturer’s red blood cell lysis, splenocytes were washed, resuspended in R10 instructions with minor modifications. Briefly, splenocytes were (RPMI 1640 supplemented with 10% fetal bovine serum, 2 mM L- thawed, washed, counted (NucleoCounter® NC-3000, Chemome- glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin) and tec), cell density adjusted with R10 at 4 × 105 cells/well and plated cryopreserved in liquid N2 until use. in 96-well polyvinylidene plates (Millipore Corp., Bedford, MA), previously coated with IFNγ capture antibody (clone AN-18). Cells were stimulated with 17 HTI- specific peptide pools (14 μg/ml final DNA extraction and 16 S rRNA sequencing concentration for each peptide) for 16 h at 37 °C in 5% CO2. A set Genomic DNA from fecal samples and intestinal content was of 147 overlapping peptides (OLP) of 15 amino acids in length, extracted using the QIAamp DNA Stool Kit, according to the overlapping by 11 amino acids and covering the entire HTI manufacturer’s instructions, and stored at −80 °C until sequen- sequence, were designed using the PeptGen algorithm (Los cing. Amplification of the V3–V4 hypervariable region of the 16 S Alamos HIV database) and synthesized (Synepeptide). To assess rRNA gene was performed using universal primers flanked by responses to HTI, OLP were assembled into 17 separated peptide standard forward and reverse adapters (16S_F 5′-TCG TCG TCG pools consisting of 7 pools for Gag (8-11 peptides/pool), 7 for Pol GCA GCG TCA GAT GTG TAT AAG AGA CAG CCT ACG GGN GGC (11 or 5 peptides/ pool), 2 for Vif (8 and 6 peptides/pool) and 1 for WGC AG-3′; 16S_R 5′-GTC TCG TGG GCT CGG AGA TGT GTA TAA Nef (2 peptides/pool). Concanavalin A (Sigma-Aldrich Corp., Saint GAG ACA GGA CTA CHV GGG TAT CTA ATC C-3′) described in the Louis, MO), at 5 µg/ml, was used as a positive control and R10 in Illumina MiSeq rRNA Amplicon Sequencing protocol (San Diego, triplicate as a negative control. After stimulation, spot-forming Published in partnership with Nanyang Technological University npj Biofilms and Microbiomes (2022) 104 A. Borgognone et al. 8 cells (SFC) were revealed by adding a biotinylated IFNγ detection 3. Atarashi, K. et al. Induction of colonic regulatory T cells by indigenous Clostridium antibody (clone R4-6A2), streptavidin conjugated with alkaline species. Science (1979) 331, 337–341 (2011). phosphatase (AP) and AP Conjugate Substrate Kit (Bio-Rad 4. Ivanov, I. I. et al. Induction of intestinal Th17 cells by segmented filamentous Laboratories, Inc., Irvine, CA). SFC per well were counted using bacteria. Cell 139, 485–498 (2009). 5. Cheng, H., Guan, X., Chen, D. & Ma, W. The th17/treg cell balance: a gut an automated ELISPOT reader system (CTL Analyzers LLC, Cleve- microbiota-modulated story. Microorganisms 7, 583 (2019). land, OH) together with the ImmunoSpot software and the 6. Gill, N. & Finlay, B. B. The gut microbiota: challenging immunology. Nat. Rev. magnitude of responses expressed as SFC per 106 splenocytes. Immunol. 11, 636–637 (2011). Threshold for positive responses was defined as responses that 7. Lynn, D. J., Benson, S. C., Lynn, M. A. & Pulendran, B. Modulation of immune were at least 50 SFC/106 spleen cells, the mean SFC/106 spleen responses to vaccination by the microbiota: implications and potential cells in negative control wells plus 3 standard deviations of the mechanisms. Nat. Rev. Immunol. 22, 33–46 (2021). negative control wells, or three times the mean of negative 8. de Jong, S. E., Olin, A. & Pulendran, B. The impact of the microbiome on immunity control wells, whichever was higher. Results were given as total to vaccination in humans. Cell Host Microbe 28, 169–179 (2020). magnitude (accumulated response to all the pools; SFC/106 9. Harris, V. et al. Rotavirus vaccine response correlates with the infant gut micro- biota composition in Pakistan. Gut Microbes 9, 93–101 (2018). PBMCs) and breadth (number of positive pools). 10. Hagan, T. et al. Antibiotics-driven gut microbiome perturbation Alters immunity to vaccines in humans. Cell 178, 1313–1328.e13 (2019). Luminex assay 11. Siegrist, C.-A. Vaccine immunology. Vaccines https://doi.org/10.1016/b978-1- Concentrations of CD40 ligand (CD154), GM-CSF, IFN γ, IL-1ß, IL-2, 4160-3611-1.50006-4 (2008). 12. Cox, R. J. & Brokstad, K. A. Not just antibodies: B cells and T cells mediate IL-4, IL-5, IL-6, IL-10, IL-12 (p70), IL-13, IL-15, IL-17A, IL-17E/IL-25, IL- immunity to COVID-19. Nat. Rev. Immunol. 20, 581–582 (2020). 21, IL-22, IL-23, IL-27, IL-31, IL-33, TNFα, TNFß at the end of the 13. Lynn, M. A. et al. Early-life antibiotic-driven dysbiosis leads to dysregulated vac- study (week 21 wpi) were simultaneously measured in mice sera, in cine immune responses in mice. Cell Host Microbe 23, 653–660.e5 (2018). a single batch. Briefly, serum samples were thawed, vortexed and 14. Pastor-Ibáñez, R. et al. Impact of transcriptome and gut microbiome on the centrifuged at 10000 xg for 10 min before collecting 25 µl of serum response of hiv-1 infected individuals to a dendritic cell-based hiv therapeutic for further testing. The serum concentration of the 22 cytokines was vaccine. Vaccines (Basel) 9, 694 (2021). measured simultaneously using sample duplicates in a customized 15. Borgognone, A. et al. Gut microbiome signatures linked to HIV-1 reservoir size mouse Th17 magnetic bead panel kit (Milliplex) and a Luminex® and viremia control. Microbiome 10, 1–15 (2022). 200 (Luminex Corp) reader, following manufacturer’s instructions. 16. Eloe-Fadrosh, E. A. et al. Impact of oral typhoid vaccination on the human gut microbiota and correlations with S. Typhi-specific immunological responses. PLoS One 8, e62026 (2013). Statistical analysis 17. Elizaldi, S. R. et al. Rectal microbiome composition correlates with humoral Differences in alpha diversity and taxa abundance between experi- immunity to HIV-1 in vaccinated Rhesus Macaques. mSphere 4, e00824–e00819 mental groups were assessed using the Wilcoxon signed-rank and (2019). 18. Ng, S. C. et al. Gut microbiota composition is associated with SARS-CoV-2 vaccine Kruskal-Wallis test (two sided). For beta diversity, statistical signifi- immunogenicity and adverse events. Gut 71, 1106–1116 (2022). cance was assessed using the pairwise permutational multivariate 19. Mothe, B. et al. Definition of the viral targets of protective HIV-1-specific T cell analysis of variance (PERMANOVA, adonis) test. The linear discriminant responses. J. Transl. Med. 9, 208 (2011). analysis effect size (LEfSe)59 was performed to identify discriminant 20. Mothe, B. et al. Ctl responses of high functional avidity and broad variant cross- bacterial signatures between Mock and Immunized groups (α = 0.05 reactivity are associated with hiv control. PLoS One 7, e29717 (2012). and LDA score > 2.0). Associations between cytokines, immune 21. Saubi, N. et al. Priming with recombinant BCG expressing HTI enhances the response and bacteria were computed based on Spearman correla- magnitude and breadth of the T-cell immune responses elicited by MVA.HTI in tion coefficients in combination with Benjamini–Hochberg multiple BALB/c mice. Vaccines (Basel) 8, 678 (2020). testing correction using the rcorr function within the R/hmisc 22. Mothe, B. et al. A human immune data-informed vaccine concept elicits strong and broad T-cell specificities associated with HIV-1 control in mice and maca- package. Statistical analyses were performed in the R environment ques. J. Transl. Med. 13, 60 (2015). and two-tailed 5% level of significance (p-value ≤ 0.05) was con- 23. Bailón, L. et al. Safety, immunogenicity and effect on viral rebound of HTI vac- sidered significant, unless otherwise stated. cines in early treated HIV-1 infection: a randomized, placebo-controlled phase 1 trial. Nat. Med. https://doi.org/10.1038/s41591-022-02060-2 (2022). Reporting summary 24. Wu, M. et al. The differences between luminal microbiota and mucosal micro- biota in mice. J. Microbiol Biotechnol. 30, 287–295 (2020). Further information on research design is available in the Nature 25. Blander, J. M., Longman, R. S., Iliev, I. D., Sonnenberg, G. F. & Artis, D. Regulation of Research Reporting Summary linked to this article. inflammation by microbiota interactions with the host. Nat. Immunol. 18, 851–860 (2017). 26. Schuijt, T. J., van der Poll, T., de Vos, W. M. & Wiersinga, W. J. The intestinal DATA AVAILABILITY microbiota and host immune interactions in the critically ill. Trends Microbiol. 21, The raw 16 S rRNA sequencing reads used in this study have been deposited at the 221–229 (2013). Sequence Read Archive (SRA) repository (Bioproject No. PRJEB52963). 27. Pereira, F. C. & Berry, D. Microbial nutrient niches in the gut. Environ. Microbiol 19, 1366–1378 (2017). 28. Guevarra, R. B. et al. Oral vaccination against lawsonia intracellularis changes the CODE AVAILABILITY intestinal microbiome in weaned piglets. Animals 11, 2082 (2021). R analysis scripts and data associated with this paper are available at https://doi.org/ 29. Guo, J. et al. Different immunization methods lead to altered gut flora and varied 10.5281/zenodo.7017193. responses to Mycobacterium tuberculosis infection in mice. J. Infect. Dev. Ctries 14, 1170–1177 (2020). 30. Tamanai-Shacoori, Z. et al. Roseburia spp.: a marker of health? Future Microbiol. Received: 25 August 2022; Accepted: 12 December 2022; 12, 157–170 (2017). 31. Lucas, S., Ghilardi, N., Li, J., Dé, F. & de Sauvage, R. J. IL-27 regulates IL-12 responsiveness of naïve CD4 T cells through Stat1-dependent and-independent mechanisms. Proc. Natl Acad. Sci USA 100, 15047–15052 (2003). 32. Dainichi, T., Richards, C. D., Karin, N. & West, N. R. Coordination of immune-stroma REFERENCES crosstalk by IL-6 family cytokines. Front. Immunol. 1, 1093 www.frontiersin.org (2019). 1. Zheng, D., Liwinski, T. & Elinav, E. Interaction between microbiota and immunity 33. Shohan, M. et al. Interleukin-22 and intestinal homeostasis: protective or in health and disease. Cell Res. 30, 492–506 (2020). destructive? IUBMB Life 72, 1585–1602 (2020). 2. Wojciech, L., Tan, K. S. W. & Gascoigne, N. R. J. Taming the sentinels: microbiome- 34. Fu, X., Liu, Z., Zhu, C., Mou, H. & Kong, Q. Nondigestible carbohydrates, butyrate, derived metabolites and polarization of t cells. Int. J. Mol. Sci. 21, 1–17 (2020). and butyrate-producing bacteria. Crit. Rev. food Sci. Nutr. 59, S130–S152 (2019). npj Biofilms and Microbiomes (2022) 104 Published in partnership with Nanyang Technological University A. Borgognone et al. 9 35. Cheng, H., Guan, X., Chen, D. & Ma, W. The Th17/Treg cell balance: a gut microbiota- ACKNOWLEDGEMENTS modulated story. Microorganisms https://doi.org/10.3390/microorganisms7120583 The project received funding from the European Union’s Horizon 2020 Research and (2019). Innovation programme under Grant Agreement No. 847943 (MISTRAL) and was 36. Awate, S., Babiuk, L. A., Mutwiri, G., Fournel, S. & Massari, P. Mechanisms of action sponsored in part by Grifols. of adjuvants. Front. Immunol. https://doi.org/10.3389/fimmu.2013.00114 (2013). 37. Pulendran, B., S. Arunachalam, P. & O’Hagan, D. T. Emerging concepts in the science of vaccine adjuvants. Nat. Rev. Drug Discov. 20, 454–475 (2021). AUTHOR CONTRIBUTIONS 38. Ansar, W. & Ghosh, S. Infl Ammation and Infl Ammatory Diseases, Markers, and M.N.-J., C.B., A.O. and R.P. designed the study and supervised the research. A.B. Mediators: Role of CRP in Some Infl ammatory Diseases Abstract. https://doi.org/ analyzed the data and wrote the manuscript. A.E.-T. contributed to data 10.1007/978-81-322-2680-2_4 (2016). interpretation and redaction of the manuscript. A.E.-T., M.C. and A.O. set up the 39. Mazzoni, A. & Segal, D. M. Controlling the toll road to dendritic cell polarization. J. animal experiments and collected samples. A.E.-T., M.C. and M.P. performed fecal Leukoc. Biol. 75, 721–730 (2004). DNA extraction, library preparation and sequencing. L.R., T.E. and A.O. performed 40. David, L. A. et al. Diet rapidly and reproducibly alters the human gut microbiome. cytokine profiling and immune response determinations. F.C.-M., M.N.-J. and A.B. Nat. 2013 505:7484 505, 559–563 (2013). contributed to data processing, management, and storage. A.B. and A.E.-T. are co-first 41. di Sabatino, A. et al. Oral butyrate for mildly to moderately active Crohn’s disease. authors of the manuscript. A.O. and R.P. are co-corresponding authors of the Aliment Pharm. Ther. 22, 789–794 (2005). manuscript. All authors edited drafts critically, revised the manuscript, and approved 42. Guo, Y., Kitamoto, S. & Kamada, N. Microbial adaptation to the healthy and inflamed the final version as submitted. gut environments. https://doi.org/10.1080/19490976.2020.1857505 (2020). 43. Omenetti, S. & Pizarro, T. T. The Treg/Th17 axis: A dynamic balance regulated by the gut microbiome. Front. Immunol. 6, 639 (2015). 44. Lin, Y., Slight, S. R. & Khader, S. A. Th17 cytokines and vaccine-induced immunity. COMPETING INTERESTS Semin. Immunopathol. 32, 79–90 (2010). The authors declare no competing interests. 45. Henke, M. T. et al. Ruminococcus gnavus, a member of the human gut micro- biome associated with Crohn’s disease, produces an inflammatory poly- saccharide. Proc. Natl Acad. Sci. USA 116, 12672–12677 (2019). ETHICAL APPROVAL 46. Ashkar, A. A., Mossman, K. L., Coombes, B. K., Gyles, C. L. & Mackenzie, R. FimH All experiments involving animals were approved by the Animal Experimentation adhesin of type 1 fimbriae is a potent inducer of innate antimicrobial responses Ethics Committee of the Germans Trias i Pujol Hospital (IGTP) and the Generalitat de which requires TLR4 and type 1 interferon signalling. PLoS Pathog. 4, 14430 (2008). Catalunya (Project number: 10559). 47. Hill, C. et al. Expert consensus document: the international scientific association for probiotics and prebiotics consensus statement on the scope and appropriate use of the term probiotic. Nat. Rev. Gastroenterol. Hepatol. 11, 506–514 (2014). ADDITIONAL INFORMATION 48. Luu, M. et al. Regulation of the effector function of CD8+ T cells by gut Supplementary information The online version contains supplementary material microbiota-derived metabolite butyrate. Sci. Rep. https://doi.org/10.1038/s41598- available at https://doi.org/10.1038/s41522-022-00368-y. 018-32860-x (2018). 49. Jenabian, M.-A. et al. Immunosuppressive tryptophan catabolism and gut Correspondence and requests for materials should be addressed to Alex Olvera or mucosal dysfunction following early HIV infection. J. Infect Dis. https://doi.org/ Roger Paredes. 10.1093/infdis/jiv037 (2015). 50. Tarabichi, Y. et al. The administration of intranasal live attenuated influenza Reprints and permission information is available at http://www.nature.com/ vaccine induces changes in the nasal microbiota and nasal epithelium gene reprints expression profiles. Microbiome 3, 74 (2015). 51. Andrews, S., Krueger, F., Seconds-Pichon, A., Biggins, F. & Wingett, S. FastQC. A Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims quality control tool for high throughput sequence data. Babraham Bioinfor- in published maps and institutional affiliations. matics. Babraham Inst. 1, 1 (2015). 52. Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods 13, 581–583 (2016). 53. Quast, C. et al. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. https://doi.org/10.1093/nar/ Open Access This article is licensed under a Creative Commons gks1219 (2013). Attribution 4.0 International License, which permits use, sharing, 54. R Foundation for Statistical Computing. R: A Language and Environment for Sta- adaptation, distribution and reproduction in any medium or format, as long as you give tistical Computing. http://www.R-project.org/ (2018). appropriate credit to the original author(s) and the source, provide a link to the Creative 55. McMurdie, P. J. & Holmes, S. Phyloseq: An R package for reproducible interactive Commons license, and indicate if changes were made. The images or other third party analysis and graphics of microbiome census data. PLoS One 8, e61217 (2013). material in this article are included in the article’s Creative Commons license, unless 56. Oksanen, J. et al. vegan: Community ecology package. R package version 2.5-2. indicated otherwise in a credit line to the material. If material is not included in the Cran R. 1, 2 (2019). article’s Creative Commons license and your intended use is not permitted by statutory 57. Dray, S. & Dufour, A.-B. The ade4 Package: implementing the duality diagram for regulation or exceeds the permitted use, you will need to obtain permission directly ecologists. J. Stat. Softw. https://doi.org/10.18637/jss.v022.i04 (2015). from the copyright holder. To view a copy of this license, visit http:// 58. Wickham, H. Package ‘ggplot2’: Elegant Graphics for Data Analysis (Springer-Ver- creativecommons.org/licenses/by/4.0/. lag, 2016). 59. Segata, N. et al. Metagenomic biomarker discovery and explanation. Genome Biol. 12, R60 (2011). © The Author(s) 2022 Published in partnership with Nanyang Technological University npj Biofilms and Microbiomes (2022) 104