CAG GM

WF co-abundance group gut microbiome principal coordinate analysis wild foods 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 The composition of the gut microbiome (GM) affects human health and varies among lifestyles. Adopting more “traditional” diets could lead to substantive and health-associated changes in the GM. However, research has focused on diets including domesticated foods. For most of our evolutionary history, humans consumed only wild foods. We explored the impact of a wild-foodonly (WF) diet on the GM composition. One participant collected daily fecal samples and recorded daily food consumption over an eight-week period, the middle four weeks of which he consumed only wild foods (nuts, fruits and leafy greens, wild deer, and fish). Samples were profiled through 16S rRNA amplicon sequencing and the species identified by oligotyping. The WF diet altered the GM composition, and the magnitude of the changes is larger than in other diet interventions. However, no new GM taxa, including “old friends” appeared; instead, the relative proportions of already-present taxa shifted. There is a clear successional shift from the pre-, during- and post-WF diet. The GM is very sensitive to the change from a “Western” diet to a WF diet, likely reflecting the different macro- and micronutrient properties of the consumed foods. 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

One of us (DP) who is an experienced forager of local wild foods collected daily stool samples during an 8-week period from 2018-09-14 until 2018-11-08. The first two weeks consisted of a normal diet, followed by four weeks of a WF diet, and a further two weeks of a return to a normal diet. The wild foods were prepared using “primitive” technologies - they were cooked on an open, wood fire and processed using grindstones and flint flakes instead of modern kitchen utensils, with the exception of a meat grinder. Other aspects of the author’s lifestyle remained unchanged; he performed his usual daily activities, continued living in his house, and interacted as usual with family members. In short, this study isolated the effects of a dietary alteration. The author is of Dutch ancestry, and at the time of the experiment was 46 yo, 1.82 m tall, and weighed approximately 76 kg. The author’s weight was measured daily, and he had daily contact with a medical doctor to monitor his health and well-being during the experimental stage. All consumed food and beverage items, except water, were recorded in a daily food log. The food items in the normal diet and in the wild-food-only diet are listed in Supplementary Table 1. Ethical evaluation of the project was conducted by the Ethics Committee of the Faculties of Humanities and Archaeology at Leiden University (Letter number 2022/23). 2.2 Fecal collection, DNA extraction and 16S rRNA amplicon sequencing 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 Kingdom), a disposable paper device to prevent sample contamination, and SMART eNAT® (Copan SpA, Brescia, Italy) for fecal sampling and preservation. These were sent on ice to the laboratory of the Unit of Microbiome Science and Biotechnology at the University of Bologna for further analysis. All specimens were stored at -20°C until processing. Total microbial DNA for each fecal sample was extracted through a method combining bead-beating and column purification, as described previously [15].

The V3–V4 hypervariable regions of the 16S rRNA gene were amplified and library preparation was performed following the 16S Metagenomic Sequencing Library Preparation protocol (Illumina) and the Nextera technology to index libraries. Indexed libraries were pooled at an equimolar concentration of 4 nM, denatured, and diluted to 5 pM prior to sequencing on an Illumina MiSeq

All sequences were processed using a pipeline that combined PANDASeq [16] and QIIME 2 [17]. After filtering the reads by length and quality, DADA2 was used to identify the amplicon sequence variants (ASVs) [ 18 ]. Taxonomic classification was performed using the VSEARCH algorithm [ 19 ] on the SILVA database (December 2017 release) [ 20 ]. Chloroplast, mitochondria, unknown, and eukaryote sequences were removed. Oligotyping [ 21 ] was then used for clustering the high-quality filtered fasta sequences from the QIIME 2 pipeline as previously illustrated by de Goffau and colleagues [ 22 ]. In particular, we used the ‘Minimum Entropy Decomposition’ (MED) option for sensitive partitioning of high-throughput marker gene sequences from the oligotyping software with the options -M 100 (to define the minimum abundance of an oligotype) and -V 2 (to define the maximum variation allowed in each node). The final node representative sequence of each oligotype was used for species profiling using the VSEARCH algorithm and the Genomes from Earth’s Microbiomes (GEM) catalog [ 23 ] as reference database. Alpha diversity was calculated using the number of observed ASVs, the Shannon index and the Faith phylogenetic diversity index. For beta diversity, the UniFrac dissimilarities were used to construct Principal Coordinates Analysis (PCoA) plots.

Biostatistics analysis and graphical representation were performed in R using the base, vegan [ 24 ] and made4 [ 25 ] packages. Data separation in the PCoA was tested using a permutation test with pseudo-F ratios (function adonis in vegan). Kruskal-Wallis tests and Wilcoxon rank-sum test were 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 used to assess significant differences in alpha diversity and taxon relative abundance between groups. P-values were corrected for multiple testing using the Benjamini–Hochberg procedure. A false discovery rate (FDR) ≤ 0.05 was considered statistically significant.

Species-level bacterial co-abundance groups (CAGs) were identified as previously described [ 3,26,27 ]. Briefly, the associations among taxa were determined using the Kendall correlation test, visualized with an heatmap and a hierarchical Ward-linkage clustering based on Spearman correlation distance metrics. The network plots were created using Cytoscape [ 28 ]. Circle sizes were proportional to species- or genus-level abundance or overabundance, and connections between nodes represented positive (gray) or negative (red) significant correlations. Keystone species were identified taking into account the topology of the network and the relative abundance of each taxon. Specifically, keystone nodes were identified by looking at the combination of the highest values of closeness centrality, betweenness centrality and degree on Cytoscape as previously described [ 29,30 ] and selecting only the taxa with a mean relative abundance > 1%.

GM dynamics observed in this research were compared to the results from other studies on (i) travelers in a setting with a traditional diet and lifestyle [7], (ii) people that radically changed their diet to a completely plant-based or animal-based diet [9] and (iii) Western or traditional populations [ 3,14,31–41 ].

Data from [7] and [9] were directly downloaded from the Qiita website [42], selecting the tables “55266”, “63513” and “63516”. Each table contained the OTU abundance obtained using the QIIME pipeline with the closed-reference approach and the Greengenes database (version 13_8). Only the longitudinal samples from travelers and from people that radically changed their diet to an exclusively plant-based or animal-based diet were retained. Samples from our study were reanalyzed using the same parameters reported into the Qiita website and then merged in a new space and graphical representations were obtained using vegan.

For the meta-analysis using data from previous studies on subjects from different geographical locations following different subsistence strategies, we analyzed paired-end sequences using QIIME 2 [17]. The sequences were taxonomically assigned using the feature-classifier “classify-hybridvsearch-sklearn” option, implemented into the VSEARCH options [ 19 ] of the QIIME 2 pipeline, 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 followed by a “q2-feature-classifier” trained on the SILVA 138.1 database [ 20 ] previously processed with RESCRIPt [43] using the developer’s instructions. The resulting abundance tables, one for each study included, were merged and rarefied, resulting in a total of 966 repository-derived samples [ 3,14,31–41 ] that were then included in the analyses along with the 57 samples generated in this work. The dataset included 407 individuals from present-day tropical and subtropical huntergatherer groups and 24 from Inuit tribes – most of which are undergoing a rapid transition away from their traditional hunter-gatherer diet toward a more “Western” diet – 51 individuals from rural groups practicing small-scale or subsistence agriculture from Africa, South America and Papua New Guinea, 38 individuals from Native American tribes, 12 urban Nigerians and 434 urban dwellers from North America, Europe and Asia.

The PCoA analysis to produce the beta diversity plot was performed using the vegdist function in vegan, computing Bray-Curtis distances on relative abundances at the genus level, considering only genera showing more than 0.2% of relative abundance in at least 3 samples. Compositional data were fit onto the ordination implementing the envfit function in vegan, and only genera with FDRcorrected p-values < 0.001 were plotted.

KO (KEGG ortholog) gene abundances were predicted using the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) software [44] by applying the default parameters, including a Nearest Sequenced Taxon Index (NSTI) value of 2. Significant differences among periods are tested by Kruskall-Wallis test and represented by box plots. For beta diversity, the Bray-Curtis dissimilarity was used to build a PCoA graph and the separation verified with a permutational test with pseudo-F ratios (function adonis, in vegan).

The main staples of the WF diet were chestnuts and acorns, which had to be shelled and were usually then ground to make porridge. These were supplementented by a few other nuts and seeds (hazelnuts and water lily seeds), a variety of fresh greens, dried berries and fruits, and a small amount of deer meat and ocean fish. During the WF period, DP gradually lost 4 kg, most during the first week of the WF diet. Two kg were quickly regained upon returning to a normal diet. Subjectively, DP became bored with the limited foods available to him, as there was little time to 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 prepare more than very basic meals or to gather foods from a wider area. This likely contributed to the overall caloric reduction and weight loss. During this period, DP kept a vlog of his experiences, which is available on Youtube [45]

In the pre-WF period, the GM was dominated by taxa belonging to the three major humanassociated phyla, i.e., Firmicutes, Bacteroidetes and Actinobacteria. In particular, the most representative families were Lachnospiraceae, Ruminococcaceae, Bacteroidaceae, Oscillospiraceae, Rikenellaceae and Bifidobacteriaceae (Fig. 1 A), which are commonly found in healthy people living a “Western” lifestyle [46]. Beta-diversity analysis revealed a clear pattern towards segregation of the microbial communities according to the sampling period, as shown by the unweighted and weighted UniFrac distances (permutation test with pseudo F-ratio, p-value ≤ 0.001) (Fig. 1 B-D). During the WF period, the GM configuration became significantly enriched in Lachnospiraceae, Streptococcaceae, Erysipelatoclostridiaceae, Butyricicoccaceae, and Eggerthellaceae and depleted in Bifidobacteriaceae, Rikenellaceae, Oscillospiraceae, of the modifications observed in the WF period returned to initial relative abundance values in the that remained at comparable levels to during the WF period (P < 0.05). The Akkermansiaceae family was even further enriched in the post-WF period compared to the two previous periods (P <0.05). In exploring differences in alpha diversity among periods, we observed a gradual increase of biodiversity from the pre-WF period, to WF and post-WF periods (P < 0.05, Kruskall Wallis test, Fig. 1 F), indicating that the intervention had an effect on the microbiome structure even after its conclusion.

These changes in the proportions of individual taxa were also mirrored by changes in clusters of coassociated bacteria, which is unsurprising given the high level of interdependence within the GM. To characterize these clusters of bacteria, we generated a heatmap based on the Kendall’s tau correlation coefficients between the different 57 genera and species with a minimum relative abundance of 0.1% in at least 20% of samples. We clustered correlated bacterial species into six coabundance groups (CAGs), indicated by different colored squares, whose relationships are represented by a Wiggum plot, where species/genus abundance is proportional to the circle diameter (Fig. 2 A and C). The dominant taxa for each CAG were Blautia (gray), Streptococcus (blue), Coprococcus comes (green), Erysipelatoclostridium (yellow), Faecalibacterium prausnitzii (cyan) 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 and Ruminococcus bicirculans (pink). The topological data analysis indicated that Faecalibacterium prausnitzii and Blautia are the two taxa with the highest combination of 1) closeness centrality (0.46 and 0.45, respectively), 2) betweenness centrality (0.04 and 0.03, respectively) and 3) degree (15 and 11 respectively), with a mean relative abundance > 1%. For these reasons they were reported as keystone taxa for the microbial community. The CAGs changed in relative abundance across the three dietary periods (Fig. 2 B). The overabundance plots of the CAG members in the 3 dietary periods showed the emergence of different patterns of correlated microorganisms, which were found to be associated with the dietary periods (i.e., pre-WF, WF and post-WF; Fig. 2 D). In particular, the GM of the pre-WF period was characterized by a F. prausnitzii-centered CAG, with several co-abundant glycan degraders, such as Bacteroides spp. (pectin, mannan, glucan, mucin) and Bifidobacterium (milk oligosaccharides) [47]. One auxiliary CAG was closely correlated to these bacteria and included R. bicirculans, Dialister invivus, Bacteroides stercoris, Romboutsia timonensis and CAG-83 taxon of the Oscillospiraceae family, which are eclectic bacteria with different substrate propensities [48–52].

Conversely, the WF-type GM was found to be centered around the Blautia CAG, which included a plethora of well-known fiber-degrading and SCFA-producing bacteria, such as taxa within the Coprococcus eutactus group, Agathobaculum butyriciproducens group and Lachnospira rogosae group, as well as Blautia, Anaerostipes hadrus, Fusicatenibacter saccharivorans and Lachnobacterium bovis [ 53,54 ]. Strongly associated to this cluster, the increase in members of the C. comes CAG and the concomitant presence of all the other CAGs enriches the WF group with a wider metabolic potential than in the previous period.

As expected, the post-WF period was characterized by an intermediate configuration between the pre-WF and the WF periods, suggesting a reapproach, although not complete, to the initial profile. Indeed, we observed a strong increase of members of the F. prausnitzii CAG, such as Bacteroides spp., to values comparable with the pre-WF period, together with the resilience of some members of the Blautia and C. comes CAGs. Notably, this period was also characterized by a higher abundance of the mucin degrader Akkermansia muciniphila, than the previous two periods.

Unlike the previous CAGs, the Streptococcus and the Erysipelatoclostridium CAGs did not show appreciable variations among the different dietary regimes, but only some relevant associations of specific members to each period. For instance, higher proportions of the proteolytic and animal fatdegrading taxa, such as Alistipes shahii, Alistipes putredinis and Clostridium disporicum were representative of the pre-WF period, whereas higher abundances of the SCFA producers 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327

Eubacterium hallii, Eubacterium ramulus, Streptococcus and Erysipelatoclostridium were characteristic of the WF period [9].

Collectively, the WF consumption caused a deep modification of the GM structure, but the GM nevertheless maintained a relevant level of inertia to partially return to the initial configuration when DP resumed a normal diet. However, there were some traits that differentiated the GM between the pre- and post-WF periods: (i) some members of the Blautia and C. comes CAGs remained at higher abundances respect to the pre-WF period; (ii) the post-WF microbiome was characterized by new traits respect to the initial period, e.g., the higher abundance of A. muciniphila.

Putative functional changes corresponding to the observed taxonomic variations were obtained by inferred metagenomics. Given the changes in individual taxa and CAGs, some of the functional aspects of the GM were altered during the WF period as well. The GM of the WF period showed a higher propensity for starch and atrazine degradation, and phenylalanine, tyrosine and tryptophan biosynthesis, compared to the other periods, as indicated by functional assignment of GM genes using the PICRUSt2 tool (P<0.05, Kruskall-Wallis test) (Fig. 3). These changes seemed to mirror the changes to the diet that included a very heavy reliance on starch-rich nuts (acorns and chestnuts) and reduced consumption of animal products (limited meat and fish, and no poultry, dairy or eggs), which may have increased the need for amino acid biosynthesis. Phenylalanine and tyrosine are both common in milk, eggs, and some meat products, while tryptophan is common in eggs and meats. These food items were limited during the WF period. While atrazine has been banned in the EU since 2004, this herbicide is highly persistent in groundwater [55]. The location where DP acquired the wild leafy greens included field borders and previous agricultural land. Herbicidedegrading microorganisms could be possibly acquired through the ingestion of food sources endowed with these specific microbiome components, allowing their adaptation to environments under xenobiotic threat of anthropic origin [56].

The patterns observed above do not represent a stochastic change to the GM, but instead reflect a distinct successional pattern from pre-WF, to WF, to post-WF diets. The weighted proportion of species maintained, gained and lost in the temporal succession of paired samples revealed a constant ratio of species shared or newly acquired between two consecutive timepoints, while the proportion of species lost decreased slightly but significantly (P<0.05, Wilcoxon test) in the WF and post-WF periods with respect to the pre-WF period (Fig. 4). This result indicates distinct 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 successional dynamics of the GM after the dietary modifications, with the GM keeping more diversity during and after the WF intervention. In particular, we found an increase in the number of persistent species during and after the WF period (i.e., species present in at least 70% of samples for each group). The pre-WF period was characterized by the constant presence of F. prausnitzii, Collinsella, Bacteroides vulgatus, Blautia, Gemmiger variabile group, Christensenella group, Oscillispiraceae F23-B02, Oscillospiraceae CAG-110, whereas the WF period was characterized by group. Finally, the post-WF was characterized by a higher number of persistent species, with the Notably, the list of the resilient taxa of this latter period included both some taxa characteristic of the pre-WF group, consistent with the resumption of a normal diet, but also completely new taxa that emerged after the WF diet (see the paragraph above).

Despite these considerable changes to the GM communities during the dietary shifts, no old-friend taxa, e.g., Treponema, Prevotella and Succinivibrio [3,13,14], increased during or after the WF diet. The GM changes almost exclusively involved the taxa already present into the microbial community, without the addition of new taxa. The most relevant aspect of the GM rearrangement during the transition from a normal to a WF diet was the switch of keystone species (i.e., the most important taxa in defining the microbiome structure as highlighted by network analysis – see Methods for how we identified keystone taxa), from F. prausnitzii to Blautia. This supports the emerging interest in the Blautia genus, which has recently been proposed as a next-generation probiotic candidate, also due to its role in ameliorating inflammatory and metabolic diseases [ 53,57 ]. These changes were also associated with an overall rearrangement of butyrate-producing bacteria, from a configuration dominated by F. prausnitzii to a configuration where the contribution of A. hadrus and E. hallii were more relevant. 3.4 GM shift caused by wild-food diet is larger than in other dietary perturbations Previous studies have explored the changes in GM structure in individuals commonly consuming a Western, industrial diet after adopting a new, more traditional diet while traveling [7], or shifting entirely to a plant-only or animal-only diet [9]. Interestingly, when compared to these, the shift in beta diversity between the initial “Western” diet and a WF-only diet was significantly larger (Fig. 5 A-C, P<0.001 permutation test with pseudo F-ratio and Kruskall-Wallis test). Furthermore, we 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 compared the GM configurations, specifically the genus-level relative abundance profiles, of our entire study to published data from hunter-gatherer, rural agriculturalist and urban-industrial communities [ 3,14,31–41 ]. The PCoA of Bray–Curtis distances showed a clear separation between traditional and urban-industrial GMs, consistent across the different studies (Fig. 5 D). In addition, the samples from our study nested within the other urban-industrial populations, in an intermediate position between the majority of the urban-industrial samples and the GM from Native Americans and rural agriculturists. Together, our analysis highlighted how the changes in diet during the WF period instigated a rearrangement of the individual GM that did not alter in depth the microbiome structure, but acted more on the present species changing their relative abundance.

[8]

Afolayan AO, Biagi E, Rampelli S, Candela M, Brigidi P, Turroni S, et al. The Gut Microbiota of an Individual Varies With Intercontinental Four-Month Stay Between Italy and Nigeria: A

[9]

David LA, Maurice CF, Carmody RN, Gootenberg DB, Button JE, Wolfe BE, et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature 2014;505:559–63. [10] Garcia-Mantrana I, Selma-Royo M, Alcantara C, Collado MC. Shifts on Gut Microbiota Associated to Mediterranean Diet Adherence and Specific Dietary Intakes on General Adult

[11] Crittenden AN, Schnorr SL. Current views on hunter-gatherer nutrition and the evolution of the human diet. American Journal of Physical Anthropology 2017;162:e23148. [12] Rampelli S, Turroni S, Mallol C, Hernandez C, Galván B, Sistiaga A, et al. Components of a Neanderthal gut microbiome recovered from fecal sediments from El Salt. Commun Biol 2021;4:1–10. https://doi.org/10.1038/s42003-021-01689-y. [13] Jha AR, Davenport ER, Gautam Y, Bhandari D, Tandukar S, Ng KM, et al. Gut microbiome transition across a lifestyle gradient in Himalaya. PLOS Biology 2018;16:e2005396. [14] Ayeni FA, Biagi E, Rampelli S, Fiori J, Soverini M, Audu HJ, et al. Infant and Adult Gut Microbiome and Metabolome in Rural Bassa and Urban Settlers from Nigeria. Cell Reports [15] Viciani E, Barone M, Bongiovanni T, Quercia S, Gesu RD, Pasta G, et al. Fecal Microbiota Monitoring in Elite Soccer Players Along the 2019–2020 Competitive Season. Int J Sports Med 2022. https://doi.org/10.1055/a-1858-1810. [16] Masella AP, Bartram AK, Truszkowski JM, Brown DG, Neufeld JD. PANDAseq: paired-end assembler for illumina sequences. BMC Bioinformatics 2012;13:31. [17] Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al.

Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572

Human gut microbiome viewed across age and geography. Nature 2012;486:222–7. [42] Gonzalez A, Navas-Molina JA, Kosciolek T, McDonald D, Vázquez-Baeza Y, Ackermann G, et al. Qiita: rapid, web-enabled microbiome meta-analysis. Nat Methods 2018;15:796–8. [43] Robeson II MS, O’Rourke DR, Kaehler BD, Ziemski M, Dillon MR, Foster JT, et al.

RESCRIPt: Reproducible sequence taxonomy reference database management. PLOS Computational Biology 2021;17:e1009581. https://doi.org/10.1371/journal.pcbi.1009581. [44] Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, et al. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol 2020;38:685–8. [45] Archaeologist Diederik Pomstra’s Wild Food (Palaeodiet) Experiment (part 1): introduction.

2018. [46] Lloyd-Price J, Abu-Ali G, Huttenhower C. The healthy human microbiome. Genome

Medicine 2016;8:51. https://doi.org/10.1186/s13073-016-0307-y. [47] Flint HJ, Scott KP, Duncan SH, Louis P, Forano E. Microbial degradation of complex carbohydrates in the gut. Gut Microbes 2012;3:289–306. https://doi.org/10.4161/gmic.19897. [48] Taylor H, Serrano-Contreras JI, McDonald JAK, Epstein J, Fell J, Seoane RC, et al. Multiomic features associated with mucosal healing and inflammation in paediatric Crohn’s disease.

[49] Gaundal L, Myhrstad MCW, Rud I, Gjøvaag T, Byfuglien MG, Retterstøl K, et al. Gut microbiota is associated with dietary intake and metabolic markers in healthy individuals.

Food Nutr Res 2022;66. https://doi.org/10.29219/fnr.v66.8580. [50] Salonen A, Lahti L, Salojärvi J, Holtrop G, Korpela K, Duncan SH, et al. Impact of diet and individual variation on intestinal microbiota composition and fermentation products in obese men. ISME J 2014;8:2218–30. https://doi.org/10.1038/ismej.2014.63. [51] Ricaboni D, Mailhe M, Khelaifia S, Raoult D, Million M. Romboutsia timonensis, a new species isolated from human gut. New Microbes and New Infections 2016;12:6–7. [52] Atzeni A, Bastiaanssen TFS, Cryan JF, Tinahones FJ, Vioque J, Corella D, et al. Taxonomic and Functional Fecal Microbiota Signatures Associated With Insulin Resistance in NonDiabetic Subjects With Overweight/Obesity Within the Frame of the PREDIMED-Plus Study. Frontiers in Endocrinology 2022;13. potential probiotic properties? Gut Microbes 2021;13:1875796. [54] Vacca M, Celano G, Calabrese FM, Portincasa P, Gobbetti M, De Angelis M. The Controversial Role of Human Gut Lachnospiraceae. Microorganisms 2020;8:573. [55] Jablonowski ND, Schäffer A, Burauel P. Still present after all these years: persistence plus potential toxicity raise questions about the use of atrazine. Environ Sci Pollut Res 2011;18:328–31. https://doi.org/10.1007/s11356-010-0431-y. [56] Henry LP, Bruijning M, Forsberg SKG, Ayroles JF. The microbiome extends host evolutionary potential. Nat Commun 2021;12:5141. https://doi.org/10.1038/s41467-02125315-x.

Safety of Blautia producta DSM 2950. Microorganisms 2021;9:908. [57] Liu X, Guo W, Cui S, Tang X, Zhao J, Zhang H, et al. A Comprehensive Assessment of the