A Challenge Model for Shigella dysenteriae 1 in Cynomolgus Monkeys (Macaca fascicularis)

Shigella dysenteriae type 1 can cause devastating pandemics with high case fatality rates; a vaccine for Shigella is unavailable currently. Because of the risks associated with performing challenge studies with wild-type S. dysenteriae 1 in human clinical trials to advance vaccine development, an improved nonhuman primate model is needed urgently. In the present study, cynomolgus macaques (Macaca fascicularis) were challenged with various doses of S. dysenteriae 1 strain 1617 to establish a dose that would produce shigellosis. Further, different routes of delivery of S. dysenteriae 1 were compared to establish the most appropriate route for infection. Animals receiving 10¹¹ cfu S. dysenteriae 1 intragastrically consistently developed signs of shigellosis characterized by the onset of diarrhea and dysentery within 2 to 3 d. Administration of as many as 10⁹ cfu S. dysenteriae 1 intraduodenally did not elicit signs characteristic of infection in macaques despite fecal shedding of bacteria for as long as 10 d. S. dysenteriae 1 administered intraduodenally at 10⁹ cfu or intragastrically at 10¹¹ cfu elicited robust IgG and IgA antibody responses to LPS. We have developed a reliable challenge model of infection with wild-type S. dysenteriae 1 in cynomolgus macaques that reproducibly induces disease and elicits robust immune responses. We believe that this animal model may provide unique insights into the immunologic mechanisms of protection to S. dysenteriae 1 infection and in advancing development of a vaccine against shigellosis.Shigella has been classified by the National Institute of Allergy and Infectious Disease as a category B priority pathogen. Shigella has numerous features that characterize an effective biological weapon, including the potential to cause high morbidity and mortality, a low infectious dose (approximately 10 cfu) in humans, the ability to produce large outbreaks even in industrialized countries, ease of direct person-to-person transmission, the ability to contaminate food and water supplies, and the potential to be weaponized.¹⁵Currently a vaccine for shigellosis is unavailable, and antibiotic therapy remains the only means of treatment.^13,15 Continually emerging new and multiple-antibiotic-resistant strains pose a serious problem to treat this disease. The present studies were designed to establish an S. dysenteriae 1 challenge model for investigation of pathogenesis, immunogenicity, and protection from infection in cynomolgus macaques. Previous studies successfully established shigellosis models after challenge with wild-type strains of Shigella spp. in rhesus macaques,^{3,5–8,12,19,22} but a similar S. dysenteriae 1 challenge model in cynomolgus macaques has not yet been reported. This model has the potential to advance the design of novel Shigella vaccines. In the current study, we established and characterized an S. dysenteriae 1 challenge model in cynomolgus macaques. In addition, we determined the minimal challenge dose required to induce clinical signs of shigellosis, measured the duration of shedding of the organism from the macaques, and contrasted the effects (as evaluated by endoscopy) of intragastric versus intraduodenal inoculation on disease induction and immunogenicity.     

The efficacy and immunogenicity of a live transconjugant hybrid strain of <Emphasis Type="Italic">Shigella dysenteriae</Emphasis> type 1 in two animal models

In our earlier studies, we constructed a hybrid strain of Shigella dysenteriae type 1 by introducing a plasmid vector pPR 1347. After introduction of a lipopolysaccharide (LPS) biosynthesis gene, virulent Shigella dysenteriae type 1 strain became avirulent. In our present study, we have evaluated the immune response and protective efficacy of avirulent live transconjugant Shigella hybrid (LTSH) strain against wild type Shigella dysenteriae type 1, after four doses of oral (rabbit) and intranasal (mouse) immunizations. Serum IgG titers showed exponential increase during immunization and peaking on the 28th day and remained at that level till the 35th day in both the rabbit and the mouse models. When tested, serum IgG titers persisted for 63 days in mice and relatively high for 150 days in case of rabbits. Protection studies showed 100% protection against the challenge with wild type Shigella dysenteriae type 1 strain in rabbits and 80% protection in mice. Our results suggested that the LTSH strain could be a useful vaccine candidate strain in the future.     

An Environmental Escherichia albertii Strain,DM104, Induces Protective Immunity to Shigella dysenteriae in Guinea Pig Eye Model

The environmental Escherichia albertii strain DM104, which cross-reacts serologically with Shigella dysenteriae was assessed for pathogenic properties, immunogenicity, and protective efficacy in different animal models to evaluate it as a vaccine candidate against S. dysenteriae, which causes the severe disease, shigellosis. The DM104 isolate was found to be non-invasive and did not produce any entero- or cyto-toxins. The strain also showed negative results in the mouse lethal activity assay. The non-pathogenic DM104 strain gave, however, a high protective efficacy as an ocularly administered vaccine in the guinea pig eye model against S. dysenteriae type 4 challenge. It also induced a high titer of serum IgG against S. dysenteriae type 4 whole cell lysate and lipopolysaccharide. Taken together, all these results indicate a good potential for the use of the DM104 as a live vaccine candidate against shigellosis.     

Serological Cross-Reaction Between O-Antigens of Shigella dysenteriae Type 4 and an Environmental Escherichia albertii Isolate

An environmental freshwater bacterial isolate, DM104, appearing as Shigella-like colonies on selective agar plates was found to show strong and specific serological cross-reactivity with Shigella dysenteriae type 4. Biochemical identification according to the analytical profile index, molecular serotyping by restriction of the amplified O-antigen gene cluster (rfb-RFLP), together with phylogenetic analysis of the 16S rRNA gene and multi-locus sequence analysis, identified the isolate as Escherichia albertii. rfb-RFLP of DM104, revealed a profile different from that of S. dysenteriae type 4. However, western blot analysis of extracted lipopolysaccharides demonstrated strong cross-reactivity with S. dysenteriae type 4 using specific monovalent antisera and a lipopolysaccharide gel banding profile similar to that of S. dysenteriae type 4. The observed O-antigen cross-reaction between an E. albertii isolate and S. dysenteriae extends our knowledge of the extent of O-antigen cross-reaction within the Escherichia/Shigella group of organisms, and offers the possibility of using DM104 and similar cross-reacting strains as shigellosis vaccine candidates.     

Passive immunity with multi‐serotype heat‐killed Shigellae in neonatal mice

The short‐ and long‐term passive protective efficacy of a mixture of heat‐killed cells of six serogroups/serotypes of Shigella strains (Shigella dysenteriae 1, S. flexneri 2a, S. flexneri 3a, S. flexneri 6, S. boydii 4, and S. sonnei) were studied in neonatal mice. Neonatal mice from immunized dams exhibited significant short‐ and long‐term passive protection against individual challenge by each of the six Shigella strains. High IgG and IgA titers against the lipopolysaccharide from each of the six Shigella strains were observed in sera from immunized dams.     

Diarrhea in young children from low-income countries leads to large-scale alterations in intestinal microbiota composition

Background

Diarrheal diseases continue to contribute significantly to morbidity and mortality in infants and young children in developing countries. There is an urgent need to better understand the contributions of novel, potentially uncultured, diarrheal pathogens to severe diarrheal disease, as well as distortions in normal gut microbiota composition that might facilitate severe disease.

Results

We use high throughput 16S rRNA gene sequencing to compare fecal microbiota composition in children under five years of age who have been diagnosed with moderate to severe diarrhea (MSD) with the microbiota from diarrhea-free controls. Our study includes 992 children from four low-income countries in West and East Africa, and Southeast Asia. Known pathogens, as well as bacteria currently not considered as important diarrhea-causing pathogens, are positively associated with MSD, and these include Escherichia/Shigella, and Granulicatella species, and Streptococcus mitis/pneumoniae groups. In both cases and controls, there tend to be distinct negative correlations between facultative anaerobic lineages and obligate anaerobic lineages. Overall genus-level microbiota composition exhibit a shift in controls from low to high levels of Prevotella and in MSD cases from high to low levels of Escherichia/Shigella in younger versus older children; however, there was significant variation among many genera by both site and age.

Conclusions

Our findings expand the current understanding of microbiota-associated diarrhea pathogenicity in young children from developing countries. Our findings are necessarily based on correlative analyses and must be further validated through epidemiological and molecular techniques.     

Construction of a trivalent candidateShigella vaccine strain with host-vector balanced-lethal system

A trivalent liveShigella vaccine candidate FSD01 against S.flexneri 2a, S.sonnei and S.dysenteriae I was constructed. This candidate strain was based on the S.flexneri 2a vaccine T32. By homologous recombination exchange, the chromosomalasd gene of T32 was site-specifically inactivated, resulting in the strain unable to grow normally in LB broth, while anotherasd gene of S.mutans was employed to construct an Asd⁺ complementary vector. This combination ofasd ^- host/Asd⁺ vector formed a balanced-lethal expression system in T32 strain. By use of this system, two important protective antigen genes coding for S.sonnei Form I antigen and Shiga toxin B subunit were cloned and expressed in T32, which led to the construction of trivalent candidate vaccine FSD01. Experimental results showed that this strain was genetically stable, but its recombinant plasmid was non-resistant. Moreover, it was able to effectively express trivalent antigens in one host and induce protective responses in mice against the challenges of the above threeShigella strains.     

Molecular cloning and biologically active production of IpaD N-terminal region

Shigella is known as pathogenic intestinal bacteria in high dispersion and pathogenic bacteria due to invasive plasmid antigen (Ipa). So far, a number of Ipa proteins have been studied to introduce a new candidate vaccine. Here, for the first time, we examined whether the N-terminal region of IpaD^72–162 could be a proper candidate for Shigella vaccine. Initially, the DNA sequence coding N-terminal region was isolated by PCR from Shigella dysenteriae type I and cloned into pET-28a expression vector. Then, the heterologous protein was expressed, optimized and purified by affinity Ni–NTA column. Western blot analysis using, His-tag and IpaD^72–162 polyclonal antibodies, confirmed the purity and specificity of the recombinant protein, respectively. Subsequently, the high immunogenicity of the antigen was shown by ELISA. The results of the sereny test in Guinea pigs showed that IpaD^72–162 provides a protective system against Shigella flexneri 5a and S. dysenteriae type I.     

Impact of Oral Typhoid Vaccination on the Human Gut Microbiota and Correlations with S. Typhi-Specific Immunological Responses

The resident microbial consortia of the human gastrointestinal tract play an integral role in modulating immune responses both locally and systemically. However, detailed information regarding the effector immune responses after vaccine administration in relation to the gastrointestinal microbiota is absent. In this study, the licensed oral live-attenuated typhoid vaccine Ty21a was administered in a clinical study to investigate whether oral immunization resulted in alterations of the microbiota and to identify whether a given microbiota composition, or subsets of the community, are associated with defined S. Typhi-specific immunological responses. The fecal microbiota composition and temporal dynamics were characterized using bacterial 16S rRNA pyrosequencing from individuals who were either immunized with the Ty21a typhoid vaccine (n = 13) or served as unvaccinated controls (n = 4). The analysis revealed considerable inter- and intra-individual variability, yet no discernible perturbations of the bacterial assemblage related to vaccine administration were observed. S. Typhi-specific cell mediated immune (CMI) responses were evaluated by measurement of intracellular cytokine production using multiparametric flow cytometry, and humoral responses were evaluated by measurement of serum anti-LPS IgA and IgG titers. Volunteers were categorized according to the kinetics and magnitude of their responses. While differences in microbial composition, diversity, or temporal stability were not observed among individuals able to mount a positive humoral response, individuals displaying multiphasic CMI responses harbored more diverse, complex communities. In line with this preliminary observation, over two hundred operational taxonomic units (OTUs) were found to differentiate multiphasic and late CMI responders, the vast majority of which classified within the order Clostridiales. These results provide an unprecedented view into the dramatic temporal heterogeneity of both the gut microbiota and host immune responses.     

Protective immunity by oral immunization with heat‐killed Shigella strains in a guinea pig colitis model

The protective efficacy of and immune response to heat‐killed cells of monovalent and hexavalent mixtures of six serogroups/serotypes of Shigella strains (Shigella dysenteriae 1, Shigella flexneri 2a, S. flexneri 3a, S. flexneri 6, Shigella boydii 4, and Shigella sonnei) were examined in a guinea pig colitis model. A monovalent or hexavalent mixture containing 1 × 10⁷ of each serogroup/serotype of heat‐killed Shigella cells was administered orally on Days 0, 7, 14 and 21. On Day 28, the immunized animals were challenged rectally with 1 × 10⁹ live virulent cells of each of the six Shigella serogroups/serotypes. In all immunized groups, significant levels of protection were observed after these challenges. The serum titers of IgG and IgA against the lipopolysaccharide of each of the six Shigella serogroups/serotypes increased exponential during the course of immunization. High IgA titers against the lipopolysaccharide of each of the six Shigella serogroups/serotypes were also observed in intestinal lavage fluid from all immunized animals. These data indicate that a hexavalent mixture of heat‐killed cells of the six Shigella serogroups/serotypes studied would be a possible broad‐spectrum candidate vaccine against shigellosis.     

PCR for Detection of Shigella spp. in Mayonnaise

The use of PCR to amplify a specific virA gene fragment serves as a highly specific and sensitive method to detect virulent bacteria of the genus Shigella and enteroinvasive Escherichia coli. Amplification of a 215-bp DNA band was obtained by using isolated genomic DNA of Shigella, individual cells of Shigella dysenteriae, and mayonnaise contaminated with S. dysenteriae. Moreover, a multiplex PCR with specific (virA) and bacterium-restricted (16S ribosomal DNA) primers generated an amplification product of approximately 755 bp for all bacteria tested and an additional 215-bp product for Shigella and enteroinvasive E. coli.     

Fecal Bacterial Microbiota of Healthy Free-Ranging,Healthy Corralled,and Chronic Diarrheic Corralled Rhesus Macaques (Macaca mulatta)

A clinical challenge to nearly every primate facility in North America is chronic idiopathic diarrhea (CID), the pathogenesis of which has yet to be fully elucidated. However, wild macaques appear resistant to CID, a trend that we observed in the free-ranging population of the Caribbean Primate Research Center. The gastrointestinal microbiota has been shown to have a significant role in the pathogenesis of disease and in maintaining normal health and development of the gut. In humans, chronic diarrhea is associated with alteration of the gut microbiota, which has lower bacterial diversity than does the microbiota of healthy humans. The current study was designed to describe and compare the fecal bacterial microbiota of healthy corralled, CID corralled, and healthy, free-ranging macaques. Fresh fecal samples were collected from healthy corralled (HC; n = 30) and CID (n = 27) rhesus macaques and from healthy macaques from our free-ranging colony (HF; n = 43). We excluded macaques that had received antibiotics during the preceding 60 d (90 d for healthy animals). Bacterial DNA was extracted, and the V4 region of the 16S rRNA gene was sequenced and compared with known databases. The relative abundance of Proteobacteria was higher in CID animals than HC animals, but otherwise few differences were found between these 2 groups. HF macaques were differentially enriched with Christensenellaceae and Helicobacter, which are highly associated with a ‘healthy’ gut in humans, as compared to corralled animals, whereas CID animals were enriched with Proteobacteria, which are associated with dysbiosis in other species. These results indicate that environment has a greater influence than health status on the gut microbiota. Furthermore, the current data provided targets for future studies on potential clinical interventions, such as probiotics and fecal transplants.

Chronic idiopathic diarrhea (CID; also called idiopathic chronic diarrhea and chronic enterocolitis) is a clinical challenge that plagues nearly every large primate facility in North America. For example, the Oregon National Primate Center reports that CID comprises nearly 30% of their clinical caseload.²⁰ At the Caribbean Primate Research Center, a review of the medical records database at the Sabana Seca Field Station (SSFS), where animals are housed in large, outdoor corrals, indicates that treatment for diarrhea comprises nearly 50% of the clinical caseload.Information on CID in wild macaques is sparse, and an exact cause for CID in research macaques has not been identified, despite extensive study. Fecal bacterial culture has yielded mixed results, with no specific pathogen consistently isolated from animals with CID. An increased prevalence of Campylobacter, Shigella, and Yersinia species in animals with chronic diarrhea compared with healthy animals has been reported.⁵⁹ However, the overall prevalence in diarrheic animals was around 25% for Campylobacter and well below 25% for Shigella and Yersinia.⁵⁹ Similarly, one study reported that approximately 30% of chronic diarrheic animals had at least one historic bout of diarrhea that was culture positive and 40% culture positive for Campylobacter at the time of necropsy.³⁸ Others have reported that fecal cultures are regularly negative for these and other common gastrointestinal pathogen,^28,38 which is consistent with our experience.The collective, interacting genomes of the symbiotic microorganisms in the gastrointestinal tract are referred to as the gastrointestinal microbiome.³⁴ The microbiome has a significant role in the pathogenesis of disease and contributes to normal health and development of the gut.^19,67 In humans, chronic diarrhea due to Clostridium difficile infection is associated with alteration of the gut microbiota (also known as dysbiosis), which has lower bacterial diversity than does the microbiota of healthy humans. This finding led to the successful use of fecal bacterial transplantation to restore the flora to normal.^17,39 Similarly, our group identified significant differences in the bacterial microbiota and enrichment of Proteobacteria (a phylum associated with dysbiosis) in diarrheic calves and horses as compared with healthy ones.^3,23 We also reported that diarrheic calves had lower relative abundance of genes responsible for metabolism of various nutrients, indicating that nutrient availability can be altered in diarrheic states.²¹ A better understanding of the organisms present in the gut of healthy and diarrheic macaques may offer new insights into the pathogenesis of this condition, and lead to new approaches to prevent and treat CID in NHP.The current study was designed to describe and compare the fecal bacterial microbiota of healthy free-ranging, semiwild rhesus macaques (HF group), healthy macaques living in large, outdoor corrals (HC group), and corralled macaques with CID. The composition of the fecal bacterial microbiota from these 3 groups was compared to determine whether differences in bacterial composition are present among the groups. Identification of such changes may provide feasible starting points for studying the role of the intestinal microbiota in the pathophysiology of CID and possible treatment and preventive measures.     

A tale of two bacterial enteropathogens and one multivalent vaccine

Shigella and enterotoxigenic Escherichia coli (ETEC) are among the top four enteric pathogens that cause diarrheal illness in young children in developing countries and are major etiologic agents of travellers' diarrhoea. A single vaccine that could target both of these pathogens would have significant public health impact. In this review, we highlight the many pivotal contributions of Phillippe Sansonetti to the identification of molecular mechanisms of pathogenesis of Shigella that paved the way for the development of rationally designed, novel vaccines candidates. The CVD developed a series of live attenuated Shigella vaccine strains based on the most prevalent serotypes associated with disease. Shigella vaccine strains were engineered to express critical ETEC antigens to form a broadly protective Shigella‐ETEC multivalent vaccine.     

An antibody targeting type III secretion system induces broad protection against Salmonella and Shigella infections

Salmonella and Shigella bacteria are food- and waterborne pathogens that are responsible for enteric infections in humans and are still the major cause of morbidity and mortality in the emerging countries. The existence of multiple Salmonella and Shigella serotypes as well as the emergence of strains resistant to antibiotics requires the development of broadly protective therapies. Recently, the needle tip proteins of the type III secretion system of these bacteria were successfully utilized (SipD for Salmonella and IpaD for Shigella) as vaccine immunogens to provide good prophylactic cross-protection in murine models of infections. From these experiments, we have isolated a cross-protective monoclonal antibody directed against a conserved region of both proteins. Its conformational epitope determined by Deep Mutational Scanning is conserved among needle tip proteins of all pathogenic Shigella species and Salmonella serovars, and are well recognized by this antibody. Our study provides the first in vivo experimental evidence of the importance of this common region in the mechanism of virulence of Salmonella and Shigella and opens the way to the development of cross-protective therapeutic agents.     

Diet and environment shape fecal bacterial microbiota composition and enteric pathogen load of grizzly bears

Background

Diet and environment impact the composition of mammalian intestinal microbiota; dietary or health disturbances trigger alterations in intestinal microbiota composition and render the host susceptible to enteric pathogens. To date no long term monitoring data exist on the fecal microbiota and pathogen load of carnivores either in natural environments or in captivity. This study investigates fecal microbiota composition and the presence of pathogenic Escherichia coli and toxigenic clostridia in wild and captive grizzly bears (Ursus arctos) and relates these to food resources consumed by bears.

Methodology/Principal Findings

Feces were obtained from animals of two wild populations and from two captive animals during an active bear season. Wild animals consumed a diverse diet composed of plant material, animal prey and insects. Captive animals were fed a regular granulated diet with a supplement of fruits and vegetables. Bacterial populations were analyzed using quantitative PCR. Fecal microbiota composition fluctuated in wild and in captive animals. The abundance of Clostridium clusters I and XI, and of C. perfringens correlated to regular diet protein intake. Enteroaggregative E. coli were consistently present in all populations. The C. sordellii phospholipase C was identified in three samples of wild animals and for the first time in Ursids.

Conclusion

This is the first longitudinal study monitoring the fecal microbiota of wild carnivores and comparing it to that of captive individuals of the same species. Location and diet affected fecal bacterial populations as well as the presence of enteric pathogens.     

Marked seasonal variation in the wild mouse gut microbiota

Recent studies have provided an unprecedented view of the microbial communities colonizing captive mice; yet the host and environmental factors that shape the rodent gut microbiota in their natural habitat remain largely unexplored. Here, we present results from a 2-year 16 S ribosomal RNA gene sequencing-based survey of wild wood mice (Apodemus sylvaticus) in two nearby woodlands. Similar to other mammals, wild mice were colonized by 10 bacterial phyla and dominated by the Firmicutes, Bacteroidetes and Proteobacteria. Within the Firmicutes, the Lactobacillus genus was most abundant. Putative bacterial pathogens were widespread and often abundant members of the wild mouse gut microbiota. Among a suite of extrinsic (environmental) and intrinsic (host-related) factors examined, seasonal changes dominated in driving qualitative and quantitative differences in the gut microbiota. In both years examined, we observed a strong seasonal shift in gut microbial community structure, potentially due to the transition from an insect- to a seed-based diet. This involved decreased levels of Lactobacillus, and increased levels of Alistipes (Bacteroidetes phylum) and Helicobacter. We also detected more subtle but statistically significant associations between the gut microbiota and biogeography, sex, reproductive status and co-colonization with enteric nematodes. These results suggest that environmental factors have a major role in shaping temporal variations in microbial community structure within natural populations.     

Shigella Strains Are Not Clones of Escherichia coli but Sister Species in the Genus Escherichia

Shigella species and Escherichia coli are closely related organisms. Early phenotyping experiments and several recent molecular studies put Shigella within the species E. coli. However, the whole-genome-based, alignment-free and parameter-free CVTree approach shows convincingly that four established Shigella species, Shigella boydii, Shigella sonnei, Shigella felxneri and Shigella dysenteriae, are distinct from E. coli strains, and form sister species to E. coli within the genus Escherichia. In view of the overall success and high resolution power of the CVTree approach, this result should be taken seriously. We hope that the present report may promote further in-depth study of the Shigella-E. coli relationship.     

Dysbiosis Signature of Fecal Microbiota in Colorectal Cancer Patients

The human gut microbiota is a complex system that is essential to the health of the host. Increasing evidence suggests that the gut microbiota may play an important role in the pathogenesis of colorectal cancer (CRC). In this study, we used pyrosequencing of the 16S rRNA gene V3 region to characterize the fecal microbiota of 19 patients with CRC and 20 healthy control subjects. The results revealed striking differences in fecal microbial population patterns between these two groups. Partial least-squares discriminant analysis showed that 17 phylotypes closely related to Bacteroides were enriched in the gut microbiota of CRC patients, whereas nine operational taxonomic units, represented by the butyrate-producing genera Faecalibacterium and Roseburia, were significantly less abundant. A positive correlation was observed between the abundance of Bacteroides species and CRC disease status (R?=?0.462, P?=?0.046?<?0.5). In addition, 16 genera were significantly more abundant in CRC samples than in controls, including potentially pathogenic Fusobacterium and Campylobacter species at genus level. The dysbiosis of fecal microbiota, characterized by the enrichment of potential pathogens and the decrease in butyrate-producing members, may therefore represent a specific microbial signature of CRC. A greater understanding of the dynamics of the fecal microbiota may assist in the development of novel fecal microbiome-related diagnostic tools for CRC.     

Ecological Modeling from Time-Series Inference: Insight into Dynamics and Stability of Intestinal Microbiota

The intestinal microbiota is a microbial ecosystem of crucial importance to human health. Understanding how the microbiota confers resistance against enteric pathogens and how antibiotics disrupt that resistance is key to the prevention and cure of intestinal infections. We present a novel method to infer microbial community ecology directly from time-resolved metagenomics. This method extends generalized Lotka–Volterra dynamics to account for external perturbations. Data from recent experiments on antibiotic-mediated Clostridium difficile infection is analyzed to quantify microbial interactions, commensal-pathogen interactions, and the effect of the antibiotic on the community. Stability analysis reveals that the microbiota is intrinsically stable, explaining how antibiotic perturbations and C. difficile inoculation can produce catastrophic shifts that persist even after removal of the perturbations. Importantly, the analysis suggests a subnetwork of bacterial groups implicated in protection against C. difficile. Due to its generality, our method can be applied to any high-resolution ecological time-series data to infer community structure and response to external stimuli.