What is already known about this topic?

Whole-genome sequencing analysis of C. difficile RT027 strains has shown that two independent lines have emerged in North America between 2000 and 2003 [2]. A comparison with pre-epidemic RT027 strains has shown that the epidemic strains have acquired a mutation in the gyrA gene, that has led to an increased resistance to fluoroquinolone antibiotics. While the development of this resistance has certainly played a role in the spread of RT027 strains, it has also been observed in non-epidemic C. difficile ribotypes and identified in strains from the middle of the 1980s. Thus, other factors have probably contributed to the emergence of epidemic RT027 strains. The prevalence of a second C. difficile ribotype, RT078, has increased tenfold in hospitals and clinics between 1995 and 2007, and has been associated with increased severity [3]. However, the mechanisms involved in the increased virulence remain unknown. Since RT027 and RT078 lines are phylogenetically distant from each other, the changes that have simultaneously led to an increase in prevalence and severity of infection might be due to independent mechanisms.

What are the main results of this study?

RT027 strains have been shown to have a competitive advantage over other strains both in vitro and in mouse models of C. difficile infection. To investigate the mechanisms involved, the authors examined the use of different carbon sources by the various strains and highlighted an increased capacity of RT027 strains to metabolise the disaccharide, trehalose. By comparing the genomes of many C. difficile strains, the authors identified a putative responsible enzyme, phosphotrehalase enzyme (TreA), which metabolises trehalose-6-phosphate into glucose and glucose-6-phosphate. The authors then observed that this gene was activated in RT027 strains at a concentration of trehalose 500-fold lower than that for the other C. difficile strains. More detailed analyses have identified a polymorphism in the TreA transcriptional repressor (TreR) in all RT027 strains and in other closely related strains responsible for epidemics in Europe and Australia. To determine whether the capacity to metabolise trehalose has an impact on virulence, the authors administered it orally to mice transplanted with human microbiota and infected with either a RT027 strain (R20291) or the same strain deleted for the TreA gene (R20291Δ TreA), and therefore unable to metabolise trehalose. Mortality was much lower with the R20291Δ TreA strain (Figure 1).

In a second experiment, the authors infected mice transplanted with human microbiota with the RT027 strain (R20291) in the presence or absence of trehalose in the drinking water (given at a dose equivalent to that received during a standard human meal). Mortality was much higher in the presence of trehalose. The two combined experiments confirm the assumption that trehalose in food contributes to the severity of RT027 strains. The genetic analysis of RT078 strains demonstrated an insertion of 4 genes, encoding a second copy of phosphotrehalase (TreA2) and its repressor (TreR2), as well as 2 other related genes. A mutation- and overexpression-based approach confirmed that this insertion was responsible for the capacity of RT078 strains to grow in the presence of trehalose.

What are the practical consequences?

Trehalose is an extremely stable sugar that is resistant to both high temperatures and hydrolysis. Considered ideal for use in the food industry, it has been mainly used since 2000, when a new low-cost production process was discovered [3]. Its use has been authorized in food by the Food and Drug Administration (FDA) in 2000 and by European institutions in 2001. The wide adoption of trehalose coincides with the emergence of infection outbreaks with RT027 and RT078 strains. Overall, the results suggest a causal role for trehalose in food in the emergence of these hypervirulent epidemic strains of C. difficile.

The microbiota: a new player in our understanding of metabolic disorders

Like most African countries, Maghreb countries are experiencing a demographic, epidemiological, and nutritional transition. In 2018, as in developed countries, there were more deaths from non-communicable diseases (75%) than from infectious diseases. Excess weight, obesity, diabetes, and hypertension have become a public health concern, with prevalence exceeding 50, 20, 10 and 30%, respectively. Classic approaches to management have proved to be inadequate due to a number of determinant factors.

Although the first published studies on gut microbiota date back to the 1960s, it has only been in the last fifteen years or so that new work has highlighted the role of the microbiota in the maintenance of chronic inflammatory states, insulin resistance, or obesity, by acting through different mechanisms.[1] Phenomena such as metabolic endotoxemia and bacterial translocation, resulting from the passage of lipopolysaccharides (LPS) into the systemic circulation, appear to be implicated.

Whether the disorder is diabetes, obesity, or metabolic syndrome, the quantity, quality, and diversity of microbiota (especially the phyla Firmicutes, Bacteroidetes, and Actinobacteria) underlie the cascade of responses leading to increased intestinal permeability (“leaky gut”), mobilization of pro-inflammatory cells, and induction of specific cell transport proteins. Gut micro-organisms are even thought to play a positive role in the immune system through exposure to bacterial LPS, which may be tolerated in some cases. Metabolic disorders, such as those caused by a high-fat diet, for example, could be avoidable by inhibiting LPS receptors (CD14/Toll-like receptor 4; TLR-4).

Bacteria-host cell communication: impact on metabolism

When subjected to a high-fat, low-fibre diet, the bacteria that compose the gut microbiota undergo changes on their surface (LPS) that trigger immune responses and local inflammatory reactions. These events increase intestinal permeability, allowing inflammatory components to enter the bloodstream.[2] Recent work has highlighted this role of dietary fat in dysbiosis and endotoxemia, initially in the oral microbiota. The process is thought to be triggered first by glycoprotein CD36 (increased sensitivity to the taste of fat), and then augmented at the level of the microbiota of the gustatory papillae (rich in Streptococci), thus creating a local inflammatory process identical to that seen in the intestinal wall. In addition, the texture and type of fat – saturated or polyunsaturated – and the involvement of bile salts have also been suggested as factors that could help account for metabolic disorders and the risk of obesity.

The role of certain bacterial phyla such as Firmicutes, Bacteroidetes, and Actinobacteria in metabolic endotoxemia, based on studies in both axenic mice and humans, is well established. For instance, in the mouse, a high-fat diet increases the concentration of circulating LPS, leading to metabolic changes linked to obesity, and analysis of gut microbiota reveals a significant decrease in Bifidobacterium spp. and Bacteroides-associated gut bacteria. In addition, a negative association has been observed between endotoxemia and the number of Bifidobacteria; the latter can reduce the level of LPS and improve intestinal barrier function,[3-5] as well as intestinal barrier integrity, which is crucial to prevent passage of bacterial components from the intestinal lumen into the bloodstream and host tissues.

What is the impact on management of metabolic disorders?

The major focus is currently the identification of specific bacteria, with a view to offering clinicians the tools to prevent or manage patients at risk of, or already suffering from, a metabolic disorder.[6- 7] Restoring equilibrium to the intestinal ecosystem or re-balancing the microbiota is a challenge in patients with intestinal dysbiosis, which is determined by epigenetics, the environment, diet, lifestyle, history of antibiotic treatments, and state of health.

Thus, Firmicutes and Bacteroidetes, which account for the majority of our gut microbiome, affect the risk of metabolic diseases, relative to their abundance. Moreover, recent studies on specific bacteria associated with energy and carbohydrate metabolism have been carried out in the laboratory. These studies seem to show, for example, that Akkermansia muciniphila, even pasteurized, improves intestinal barrier function and the thickness of the mucus layer, and may thus influence insulin resistance and obesity. The oxygen sensitivity of this species has so far limited its ability to be cultured and restricted its study in humans.

Other gut bacteria, such as Faecalibacterium prausnitzii, play a beneficial role and may offer therapeutic strategies based on the use of specific probiotics. In addition, other elements should be considered, such as the interactions between the host, microbiota, and brain, leading to the concepts of the taste cortex, pleasure circuit, and microbial agents that mediate obesity, which are of particular importance. The significant failure to manage obesity may be explained by approaches that lack optimal management of the dynamic states of the gut microbiota over time.

Consequently, today’s challenge is to better understand “obesogenic and diabetogenic microbiotic factors” in order to shape the way low-fat and low-carbohydrate nutritional diets, appropriate physical activity, probiotic and prebiotic supplementation, as well as faecal transplantation may be used in the future.

The role of these factors needs to be clarified with regards to their complementarity and based on an integrated preventative, but also therapeutic, approach.[8] Sufficiently large cohort studies should help confirm the role of the oral and gut microbiota in the development of transient and chronic inflammatory states underlying metabolic disorders – and probably other pathological states, such as cancers or some psychiatric conditions.

In medicine, the disciplines of neurology and microbiology have largely matured along distinct parallel trajectories, only interfacing in pathological situations when direct infections of the central nervous system occur. However, over the past decade, there has been a revolution in biomedicine with the realisation that the gut microbiota (the trillions of bacteria that reside within the gut) plays a key role in maintaining homeostasis and in programming the major body systems, including even the brain.

A growing body of research is focused on illuminating the bidirectional communication pathways between gut bacteria and the central nervous system, the microbiota– gut–brain axis, however, this is a field in its relative infancy [1]. Changes in the microbiome, its metabolites, and its interaction with the gut-brain axis are associated with a wide array of illnesses including brain disorders. Studying the microbiome requires close collaborative efforts of clinicians with basic scientists and bio-informaticians and works best when traditional discipline barriers between neurology, gastroenterology, and microbiology are broken down.

In preclinical research, a number of experimental models have proven essential to evaluate the microbiome within the context of brain and behaviour, including prebiotic and probiotic intervention, antibiotic administration, faecal transplantation, and the use of germ-free and gnotobiotic animals [1]. In clinical research, most data, especially in the area of neurology, have relied on cross-sectional studies of the microbiome in patients with disease versus healthy age-matched individuals.

The microbiota gut-brain axis mechanisms of communication

Much experimental effort is being placed at trying to dissect the pathways of communication between the gut and the brain. The gut bacteria influences central processes through a variety of mechanisms (Figure 1). Firstly, the microbiota’s ability to synthesize neurotransmitters (i.e. γ-amino butyric acid (GABA), noradrenaline, and dopamine) is an important avenue of communication. Secondly, microbes play a key role in activation of the immune system which can play a fundamental role in ageing, neurological disorders, and neurodegeneration. Finally, microbes produce metabolites including short-chain fatty acids (SCFAs) which are essential for the gut, the immune system, and potentially brain health. Moreover, the gut microbiota and the brain are linked through the vagus nerve, and through the modulation of key dietary amino acids, such as tryptophan.

Given the close association between the gut microbiota and the brain, it is not at all surprising that gut bacteria play key roles in neurological and psychiatric diseases. The strongest evidence for a role of the microbiome in brain function derives from germ-free mice. Studies from a number of research groups in Canada, Sweden, and Ireland have shown that, in germ-free animals, the brain fails to develop normally in the absence of the gut microbiome [2]. Moreover, fundamental brain processes, such as myelination, adult neurogenesis, and microglia activation, have also been shown to be critically dependent on microbiota composition.

Microbiota & Ageing

The relationship between the microbiome and the ageing brain is also receiving much attention, which is of particular interest in the field of neurology as many neurological and neurodegenerative disorders occur in old age. Once again, the concept of linking the microbiome to healthy ageing is not new, and this was championed over 100 years ago by Nobel prize-winning immunologist, Elie Metchnikoff, who observed that the villagers in a certain region of Bulgaria lived unusually long lives, a fact he attributed to the presence of lactic acid bacteria in their diet. We have recently revisited Metchnikoff’s original study [3], and shown that the behavioural deficits in aged animals coincide with changes in the microbiome. Moreover, the ELDERMET study has shown that the composition of the gut bacteria in the elderly correlated with their overall health – frailty and the immune system [4]. The greater the microbiome diversity, the better the health outcomes. These investigators went one step further to investigate what is driving the diversity of the microbiome, which they determined to be a diverse diet. As people eat processed bland food (often in nursing homes), the diversity of their microbiomes is reduced, whereas those with a diet rich in fruit and vegetables have better outcomes [4].

A decline in microbial diversity is associated with a concomitant increase in microglial activation which is correlated with brain mass differences in the mouse. This contributes to an age-associated inflammatory response known as “inflammaging”, which in turn has been associated with neurodegenerative diseases such as Alzheimer’s (AD) and Parkinson’s disease (PD). Furthermore, the microbiome has been shown to regulate microglia activation; germ-free mouse brains were shown to express defective microglia, which was partially rescued upon restoration of the microbial community to control levels [5].

Parkinson's disease

There is a growing realisation that the aetiology of PD may actually originate in the gut [6]. Indeed, α-synuclein, the protein aggregate hallmark of PD pathology in the brain, has also been identified in the mucosal and submucosal nerve fibres and ganglia of Parkinsonian patients, with some preclinical evidence even suggesting that α-synuclein in the gut can be transported to the brain via the vagus nerve. Moreover, functional gut symptoms, such as constipation, often occur as prodromal symptoms years before any motor symptoms emerge. Since Scheperjans and colleagues first showed that there are specific alterations in microbiome composition in PD [7], many more studies have emerged [8]. However, to date, there is no consensus as to whether a specific microbial signature exists. When mice were colonized with the microbiota of PD patients via faecal microbiota transplantation, they developed motor deficits and neuroinflammation; two hallmark symptoms of PD [9]. Additionally, symptoms improved when the mice were treated with antibiotics. These studies implicated short-chain fatty acids as drivers of the neuroinflammatory processes in PD [9]. The vagus nerve is particularly well placed as the conduit for signals from the gut to the brain, involving either microbial or prionlike translocation of α-synuclein. Indeed, epidemiological studies based on Danish and Swedish patient registries have shown that truncal vagotomy is protective against PD. Although there has been much excitement in the field, caution is needed when examining the available data as it is largely derived from small cohorts and lacks a longitudinal perspective. Many more mechanistic studies are needed to understand how changes in the microbiota can moderate both the motor and non-motor symptoms of PD and its co-morbidities [10].

Alzheimer's disease

The concept that microbes may play a role in the pathophysiology of AD is not new and the notion that amyloid, the aggregation of which is one of the key hallmarks of AD, may act as an antimicrobial peptide in the brain is an intriguing concept [11]. However, it is ethically difficult from the perspective of Koch’s postulate to prove whether there is any infective cause of neuroinflammation and neurodegeneration. As in PD, the relationship between gut proteins and brain health is receiving increased attention with the realization that amyloid-like proteins can be produced by bacteria which has been shown to increase α-synuclein pathology in aged rats and worms [12]. Much more work is needed to validate such strategies in humans.

Recently, cross-sectional studies have identified that Escherichia/Shigella bacterial taxa, which are associated with mediating inflammation, were increased in faecal samples from AD patients relative to control subjects. Moreover, the microbiota changes correlated with pro-inflammatory cytokine levels in whole blood [13]. Such results suggest a causal link between dysregulation of the microbiota and systemic inflammation, which may initiate or exacerbate neurodegeneration in the brain in AD. However, these are still relatively small studies and much more research is needed in larger cohorts to assess the causal relationship between the gut microbiome and AD.

In parallel, a number of transgenic mouse models of AD have been shown to have an altered microbiome [14]. Seminal studies in germ-free mice showed that there is a marked absence of amyloid plaque buildup and neuroinflammation when microbes are not present [14]. Similarly, chronic treatment of APP/PS1 transgenic mice with an antibiotic cocktail reduced microglial and astrocyte accumulation surrounding amyloid plaques in the hippocampus and led to a decrease in insoluble A plaques [15]. Together, these studies unequivicolly place the microbiome as a regulator of key molecular components of AD.

Future perspectives

It is clear that the microbiome is critically important for the appropriate development and maintenance of brain function. Moreover, as outlined above, there is accumulating evidence from both animal and clinical studies implicating the microbiome in a variety of neurological and neurodegenerative diseases. Given the marked effects of the microbiota in regulating brain function, it is plausible that its composition affects the progression, susceptibility, and treatment of almost all neurological disorders. Nonetheless, there are marked gaps in our knowledge regarding the role of the microbiome in other neurodegenerative diseases, such as amyotrophic lateral sclerosis or Huntington’s disease, and caution is needed to not over-interpret such studies. The field needs to move away from corelative studies towards mechanistic causal approaches. Moreover, more interventional studies are needed using probiotic strains and prebiotics, and even faecal microbiota transplants may potentially be important in the field. It is possible that similar approaches could target different disorders, for example, the modulation of T lymphocyte signalling to the brain may be useful in dampening down the neuroinflammatory status in patients following stroke, as well as in patients with AD and during ageing.

With regards to clinical neurology, many patients are on multiple medications and there is a growing understanding of the relationship between the microbiome and drug action. Thus, all studies should aim to differentiate between the impact of drugs and that of disease on the microbiome. Moreover, temporal studies with presymptomatic individuals will be important to determine the potential role of the microbiome as a biomarker of disease.

The ideal donor 

Despite the fact that up until now nobody really knew how to precisely define the normal gut microbiota (“eubiosis”), we do know that high microbial diversity and gene richness are of key importance in the host-microbiota equilibrium. Therefore, the ideal fecal donor should be screened for bacterial richness. A surrogate marker for this property was proposed by Marie Joossen (Leuven, Belgium) who pointed out that the presence of Blastocystis hominis correlates with a higher microbial richness.[1] This finding – if confirmed by others – may change our current practice to avoid carriers of this commensal as fecal donors. Enriching the donor’s microbiota by prebiotics or using multiple donors may also ensure (theoretically) a higher baseline diversity of the donated material. This was also observed by Karakan et al. (Ankara, Turkey) who performed an open trial in ulcerative colitis with an overall complete response rate of 32% which was particularly influenced by a high bacterial diversity in the donated fecal material.

The new brown gold 

Strict adherence to current guidelines for selection of donors for fecal material necessitates the rejection of most donors. Terveer et al. (Leiden, the Netherlands) report that only 3,5% of possible donors are suitable at the end of the line [2]. The main reasons for not being accepted as a donor are: age above 50, high BMI and carrier status of non-pathogenic germs (Blastocystis hominis, Dientamoeba fragilis) and MDROs (multidrug resistant organisms).[2]

Clostridium Difficile 

The main indication for fecal microbiota transplantation remains refractory infection with Clostridium difficile (C. difficile). Ianiro et al. (Rome, Italy) showed, in a retrospective series of 282 C. difficile patients comparing antibiotic treatment and fecal transplant, that this latter treatment resulted in significant shorter hospital stay, significantly lower mortality and specifically less sepsis-related mortality.

Antonio Gasbarrini (Rome, Italy) therefore suggests that the time has come to promote fecal microbiota transplantation as the first line therapy in C. difficile infection.

Expanding the scope of FMT 

Antonio Gasbarrini gave a nice overview of promising indications for fecal micro- biota transplantation. It has been shown that fecal transplants restore the human microbiota better than probiotics after antibiotic-induced dysbiosis in humans. This was also the case following combined chemotherapy and antibiotic induced dysbiosis in the setting of hematological stem cell therapy and in patients with liver cirrhosis receiving antibiotics. Mice models even show evidence for restoration of immune function and intestinal integrity after chemotherapy induced intestinal damage. Therefore, Antonio Gasbarrini makes the case for pre-emptive stool conservation for subsequent autologous fecal microbiota transplantation, e.g. following antibiotic treatment or bone marrow transplantation. In vivo evidence from clinical trials must be awaited before this futuristic but not unrealistic strategy can be implemented. Anyhow, it seems quite logical that collecting one’s own stool for autologous transplantation later in life, is the way to go.

Ulcerative colitis 

In 3 out of 4 published randomized controlled trials fecal microbiota transplantation was superior to placebo in refractory ulcerative colitis (UC) patients [3]. Mean remission rate in these studies, however, was only 25-30 % and Rainer et al. (Graz, Austria) presented a study with similar complete remission rates, showing no added value of administering fresh stool in these patients. Nevertheless, up until now there was no standardized stool transplantation protocol in UC. Remission rates of 30% seem low but to put things in perspective: this is also the remission rate that is achieved with the expensive and widely used biologicals [...]. The importance of colonic microbiota in UC was demonstrated by Herrera-de Guise et al. (Barcelona, Spain) who showed that patients in long-term stable remission (more than 5 years) present with an abundance of Akkermansia muciniphila and Faecalibacterium prausnitzii, similar to healthy controls. These authors even suggest that we should think of a paradigm shift in treating UC: our therapeutic endpoint should perhaps no longer be immune suppression, but we should aim to reach eubiotic microbiota characteristics.

Irritable bowel syndrome 

Irritable bowel syndrome (IBS) is certainly the condition for which the expectations of a cure by fecal microbiota transplantation are very high in both patients and health care practitioners. Still, conflicting results from randomized controlled trials [4, 5] do not currently support a widespread use of this treatment in IBS. Intestinal dysbiosis is present in IBS but a clear causality between these microbial changes and symptoms are lacking. Halkjaer et al. (Copenhagen, Denmark) performed a randomized controlled trial in 52 adult IBS patients; an increase in biodiversity (comparable to the donors) was observed in the patients being actively transplanted.[5] Unfortunately, the placebo group had a significantly better clinical outcome at 3 and 6 months than the patients receiving fecal capsules.[5] In a small cohort of 16 IBS patients, Holster et al. (Orebro, Sweden) were also unable to demonstrate efficacy for fecal microbiota transplant vs. placebo. They also studied rectal sensitivity by means of a barostat and showed no difference between the active and control groups, concluding that changing the microbiota does not contribute to the visceral hypersensitivity in IBS.

Bacterial pills 

Ianiro et al. performed an open label trial in C. difficile infection with a synthetic microbiota suspension (10 patients only) demonstrating efficacy. Khanna et al. (Rochester, USA) have also demonstrated efficacy in prevention of C. difficile infection’s relapse with a non-frozen, lyophilized, orally administrated microbiota restoring drug (RBX7455) in an open-label phase 1 trial. These are promising results that need to be confirmed in large scale randomized trials.

Explain fecal transplantation to your patients with this dedicated content: 

What is already known about this topic? 

From birth, certain factors influence the development of obesity. Breastfeeding has a protective effect, partly because breast milk is low in protein. The gut microbiota (GM) must also be taken into consideration because it is involved in food absorption and energy metabolism. The GM is formed during the first 2 to 3 years of life, and the method of feeding (breast milk vs. infant formula milk) is one of the main factors that modulates GM composition. The GM of obese adults is less diverse and has a higher ratio of Firmicutes/Bacteroidetes.

What are the main results of this study? 

This study is based on data from 1,087 infants included in the CHILD birth cohort (Canadian Healthy Infant Longitudinal Development). Infant faecal samples were collected for microbial analysis at 3 to 4 months (n=996), 12 months (n=821),  and at both time points (n=730). Mothers completed questionnaires on the method of feeding at 3 and 6 months, which made it possible to define different groups according to breastfeeding practices (Table 1). Among these infants, 74.2% were delivered vaginally, and 39.8% of mothers were overweight or obese. Rates of exclusive breastfeeding were 53.8% at 3 months and 17.6% at 6 months.

At 3 months, exclusive breastfeeding protected against the risk of being overweight at 12 months (defined by measured weight: expected weight ratio > 85th percentile) compared to exclusively formula-fed infants: 19.2% vs. 33.3%, respectively, with no significant effect of adjustment (Table 1). At 6 months, formula milk in addition to breastfeeding increased the risk of being overweight at 12 months, but solid foods did not have the same effect. Prolonged breastfeeding was found to confer a protective benefit.

As expected, GM richness and diversity at age 3 to 4 months differed according to the method of feeding, and the composition was significantly different between exclusively breastfed and non-breastfed infants (Figure 1). An increase in breastfeeding was accompanied by an increase in relative abundance of Bifidobacteriaceae and Enterobacteriaceae and a decrease in Lachnospiraceae, Veillonellaceae, and Ruminococcaceae.

By 12 months, the GM had become more homogeneous but there were still differences relative to the method of feeding at 6 months, i.e. increased richness for infants at least partially fed with formula milk, and a relative abundance of Actinobacteria and Proteobacteria that was higher in exclusively breastfed infants and lower in non-breastfed infants.

Greater GM richness at 3 to 4 months correlated with an increase in excess weight or a risk of becoming overweight at 12 months, particularly with regards to Lachnospiraceae, with a median relative abundance of 5.9% (overweight), 4.7% (risk of becoming overweight), and 1.9% (normal weight) (p=0.01) (Figure 2).

What are the practical consequences?

This study firstly demonstrates the benefit of breastfeeding on gaining excess weight at 1 year, and secondly that this benefit is related to modulation of the GM.

Additionally, it is important to promote exclusive breastfeeding from birth, by limiting supplementation with infant formula milk in the maternity unit. This benefit is enhanced by prolonged breastfeeding. Whereas the use of formula milk has a negative impact in infants, this is not the case for solid foods.

What is already known about this topic? 

The human gut microbiota forms a complex and balanced ecosystem. Perturbations of this ecosystem can play a role in the development of infections, obesity, diabetes as well as neurological and inflammatory disorders. It is estimated that antibiotics have added 2 to 10 years to our life expectancy, but early exposure to these drugs has also been associated with noxious metabolic, inflammatory and neurological effects, both in animal models and in humans. When microbial communities are exposed to antibiotics, not only do they react by shifting their composition, but also by evolving, optimizing and disseminating antibiotic resistant genes (ABR genes) which collectively form the resistome.[2] The human gut microbiota is a reservoir of ABR genes which are exchanged between the resident strains, thereby propagating resistance.[3] The development and spread of antibiotic-resistant bacteria constitute a serious threat to public health. Only a few studies have investigated the effects of specific antibiotics on intestinal ecosystems and their associated resistomes. In previous work it was shown that antibiotic administration induces a decrease in microbiota diversity and an in- crease in the carriage of ABR genes.[4, 5] However, the effects of a combination of antibiotics on the microbiota and the role of ABR genes in microbial persistence have not yet been studied. In this study, 12 healthy men aged 18 to 40 years received a cocktail of three last-resort antibiotics (vancomycin, gentamicin and meropenem). The authors analysed the impact of this treatment on the gut microbiota by shotgun sequencing of faecal samples taken before and at four time points over a 6-month period following antibiotic administration.

What are the main results of this study? 

At D4, immediately after the intervention, microbiota diversity and richness were notably reduced compared to baseline. However, despite the use of very broad spectrum antibiotics, many species were still detectable at D4 (Figure 1a). By D8, microbiota diversity (measured by the Shan- non index) had considerably increased, suggesting that surviving microorganisms had begun to regrow (Figure 1b). At 6 months, microbiota diversity was almost completely restored to baseline levels but this was not the case for richness, suggesting that some strains had been permanently (or at least extendedly) eradicated.

Some of the early observable changes included blooms of normally subdominant commensals like Escherichia coli, Veillo- nella spp., Klebsiella spp., Enterococcus faecalis and Fusobacterium nucleatum and a sharp depletion of butyrate-producers like Faecalibacterium prausnitzii, Roseburia hominis, Anaerostipes hadrus, Coprococcus spp. and Eubacterium spp. These shifts in microbiota composition were no longer significant at D42.

The authors then investigated the role of ABR genes in microbiota recovery. In particular, they found that β-lactamase-harbouring metagenomic species had significantly higher odds of survival (OR = 1.64 [1.24-2.17]) at D8. In addition, metagenomic species not detected at baseline had better odds of subsequent de novo colonization if they harboured ABR genes against one of the three antibiotic classes used.

What are the practical consequences? 

These findings indicate that the gut micro- biota of healthy young adults is resilient to a 4-day intervention with broad spectrum antibiotics with recovery of the majority of bacterial communities after about 6 months. The recovery of individual species is modulated by carriage of ABR genes. Further studies are needed to assess the impact of repetitive perturbations and/or over longer periods and to determine whether these findings also hold true in children whose immune system and microbiota are immature. It is possible that repetitive use of antibiotics over long periods selects for bacteria carrying ABR genes at the expense of other commensals, with prolonged or permanent effects on the microbiota. In such cases, corrective intervention with exogenous supply of microorganisms could be considered. The effects of antibiotics on the intestinal microbiota are therefore important and their use must be rationalized.

Celiac disease is a common inflammatory and autoimmune reaction that occurs in genetically predisposed individuals after consuming gluten (Figure 1). The characteristic lesion is the destruction of the finger-like projections of small intestinal lining (enteropathy). Clinical manifestations of the disease are varied and include both intestinal or extra-intestinal symptoms. Celiac disease is unique in that it is the only autoimmune disease where the triggering antigen (gluten) is known. The mechanisms that explain HLA genetic risk and the steps triggered by the dietary trigger that ultimately leads to the development of pro-inflammatory gluten specific T-cells and autoantibodies, are well understood (Figure 2). One unsolved question relates to the rapid recent increase in prevalence as well as the fact that celiac disease only develops in a fraction of genetically susceptible individuals, suggesting there must be additional genetic or environmental factors involved in activating the inflammatory cascade. In particular, there has been a growing interest in the role of the microbial factors in celiac disease development [1]. In this review we focus on bacterial alterations and how they could play a role in disease mechanisms as well as how they constitute potential therapeutic targets for celiac disease.

Correlation of dysbiosis and celiac disease: lessons from clinical studies 

One of the first studies to suggest a microbial contribution to celiac disease described the presence of rod-shaped bacteria in duodenal biopsies of children born during a celiac disease “epidemic” in Sweden. These bacteria were not observed in children without celiac disease, or in children born following the epidemic and it was thought that their presence may have contributed to the increase in the disease incidence observed in Sweden.[2] However, the mechanisms underlying this association remain unknown.

A number of studies have since been published that analyzed the fecal and small intestinal microbiota composition in celiac disease patients compared to healthy controls. Some relatively consistent findings across studies include increases in proportions of Bacteroides and members of Proteobacteria, and decreases in Lactobacillus and Bifidobacteria in celiac patients compared to controls.[1] Increased abundance of Proteobacteria was also found in patients who suffered from persistent symptoms, despite adhering to a gluten-free diet.[3] More recently, children at a high genetic risk for developing celiac disease were shown to have a different microbiota composition compared to children who were at a low genetic risk.[4-6] Finally, at-risk children who went on to develop celiac disease were suggested to have higher basal microbial diversity that did not further diversify over time, suggestive of a “premature maturation” of the gut microbiota.[7] While the results suggest that an altered early trajectory of the microbiota could predispose to celiac disease, larger trials with increased sample sizes are needed to confirm the findings. Despite the findings of an altered composition of the microbiota in celiac disease or in at-risk children, no celiac “microbial signature” has been established. Differences in the locations of study population, status of control subjects, fecal vs. small intestinal samples, and methodology may contribute to inconsistencies between studies. Inconsistent findings have also been reported regarding associations between events that can alter the development of the microbiota and celiac disease develop- ment. While early studies suggested anti- biotic use and delivery by C-section could increase celiac disease risk, more recent larger clinical studies did not confirm these associations.[8]

The long-term follow-up of at-risk infants may provide insight into the factors that may contribute to disease onset. Despite no evidence for causation, these clinical associations have stimulated the study of basic causative mechanisms in reductionist systems and animal models.

Mechanisms by which microbiota alterations can influence celiac disease: lessons from basic research 

How microbes could contribute to the pathogenesis or development of celiac disease can be better understood by studying the function of the microbial community in celiac disease patients vs. healthy subjects. Isolation of bacteria from the human small intestine allows for translation into reductionist models. For instance, strains of Enterobacteriaceae isolated from celiac patients were more virulent than those isolated from healthy controls.[9] Moreover, E. coli strains isolated from celiac children had more in vitro pro-inflammatory capacity compared to strains of Bifidobacterium that were isolated from control children.[10] Microbiota “humanized” mouse models of germ-free mice add a layer of complexity and allow in vivo comparison of phenotypes induced by a microbiota of interest. Moreover, these mice can express features of the human immune system (such as MHC-class II expression) that are critical for celiac disease development. Transgenic mice that express the human celiac risk gene, HLA-DQ8, were protected from gluten-induced pathology when they were minimally colonized with a microbiota that was free from pathogens or opportunistic bacteria. However, if an adherent strain of E. coli, isolated from the celiac gut, was added to the protective bacteria, mice developed gluten-induced pathology. Similarly, treatment of mice harbouring a diverse murine microbiota with the antibiotic vancomycin led to an increase in Proteobacteria, including E. coli, and a worsening of gluten-induced pathology.[11]

Recent translational work in mice expanded on the observation that bacteria are capable of degrading gluten (Figure 3). The study used gnotobiotic mice that were colonized with either opportunistic pathogens, such as Pseudomonas aeruginosa, or with commensals, such as Lactobacillus. The authors showed that different bacteria can degrade gluten in vivo, but the protein fragments they produce are distinct. The study further demonstrated that enzymes from P. aeruginosa, which was isolated from a celiac disease patient, could degrade gluten. This digestion process produced gluten fragments that stimulated an inflammatory immune response in cells isolated from celiac disease patients and were better able to cross the small intestinal barrier, where interaction with immune cells would occur. Several peptides generated by P. aeruginosa digestion that were subsequently digested with lactobacilli, isolated from a healthy subject and a core member of the healthy microbiome, no longer induced inflammatory immune responses in vitro. This study provided a key mechanism where both opportunistic pathogens and commensals may modify the repertoire and immune properties of gluten peptides in the gut, thereby impacting disease susceptibility.[12]

Can microorganisms be used to treat or prevent celiac disease? 

A diagnosis of celiac disease means strict, life-long avoidance of gluten-containing foods because exposure to small amounts of gluten can trigger a variety symptoms and enteropathy in affected people. Gluten is used ubiquitously in processed foods, which makes strict adherence difficult, and has prompted the search for alternative or adjuvant therapies. Given the key role for microorganisms in regulating immunity and the association between celiac disease and altered composition and function of the microbiota, the therapeutic potential of several probiotics has been tested. A strain of Bifidobacterium longum, previously shown to have anti-inflammatory effects in vitro,[10-13] was tested in children on a gluten-free diet in a double-blind placebo-controlled trial.

Administration of the probiotic led to some immune changes, as well as lower levels of potentially harmful bacteria (B. fragilis). However, no changes in symptoms were observed between children that received the probiotic compared to those that received placebo.[14] Because the probiotic was administered together with the gluten-free diet, it is difficult to discern between changes induced by the dietary restriction or the probiotic. Two other studies tested the effects of a strain of Bifidobacterium infantis. The first randomized double-blind placebo-controlled trial demonstrated that patients receiving the probiotic showed significant improvement of symptoms following the 3-week trial, but no differences were found in intestinal permeability.[15] A follow-up trial tested whether the same probiotic could modulate innate immune responses, which could be responsible for the symptomatic improvement previously observed. Administration of a strain of B. infantis led to a decrease in the number of small intestinal Paneth cells, that paralleled a decrease in antimicrobial peptides. These effects of the probiotic were independent of the gluten free diet.[16] Due to the nature of these few studies including only small groups, there is no evidence to date to recommend any probiotic in celiac disease. Moreover, probiotics consumed by patients with celiac disease need to be rigorously certified gluten-free, and this is not the case for every over the counter preparation. Prior to patient consumption, we need a better understanding of mechanisms of action, and those chosen for further testing should be selected due to their involvement in celiac disease pathways. For example, bacteria that aid in the detoxification of gluten could be selected and used to complement the gluten-free diet. However, to date, no single bacterium tested has shown optimal gluten digestion in vitro. Studies have focused on bacterial strains that produce enzymes capable of degrading gluten, but fungal species such as Aspergillus niger, also produce gluten-degrading enzymes, and rational combinations of fungal and bacterial organisms may offer an attractive avenue of therapeutic research in celiac disease.

Recent review examined the role of esophageal microbiome in BE and esophageal cancer.[1] Esophagus is exposed to swallowed oral microorganisms and also microbes of the refluxed gastric contents. Esophageal microbiota is not similar to oral nor gastric microbiota. The first bacteria detected in the esophagus were Strectococcus viridans and group D Streptococcus. Later six phyla were observed by broad-range 16S ribosomal DNA gene clone sequencing including Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, Fusobacteria, and Streptococcus. BE and high-grade dysplasia were associated with the highest number of bacteria. In patients with esophagitis and BE the number of Streptococcus was diminished while gram-negative anaerobes and microaerophiles were increased.

BE and esophageal adenocarcinoma are associated with increased number of Escherichia coli. Another gram-negative species detected in esophageal cancer patients is Fusobacterium nucleatum. Also oral dysbiosis may be related to the increased risk of esophageal cancer, while gastric Helicobacter pylori seems to protect against esophageal cancer. Gastric dysbiosis such as increase of Clostridiales and Erysipelotrichaceae is associated with esophageal squamous carcinoma. Also fungi, for example Candica albicans and C. glabrata are often detected in esophageal samples from patients with esophaghageal adenocarcinoma. An epidemiologic study showed dose-dependent association between penicillin use and increased risk of esophageal cancer. Also proton pump inhibitors modify gastric and esophageal microbiome.

The present data of esophageal microbiota were obtained from small, selected, and symptomatic patient populations in cross-sectional studies. Thus no conclusions of causality between esophageal microbiota and esophageal diseases can be made. Only a small portion of patients with BE develop adenocarcinoma and further studies are needed to define the role of esophageal dysbiosis in the pathogenesis of cancer. One topic for further research is the impact of proton pump inhibitors on esophageal microbiota and on the risk for esophageal diseases.[1]

Eosinophilic esophagitis (EoE) is an allergic chronic inflammatory disease that is the most common cause of dysphagia in children and young adults in developed countries. EoE has common inflammatory features with other allergic diseases and allergen exposure probably has a central role in EoE pathogenesis. Capucilli and Hill have reviewed EoE epidemiology, pathogenesis and treatment.[2] The esophageal microbiota may be involved in the pathogenesis of EoE. The esophagus is colonized by hundreds of bacterial species and members of Firmicutes and Bacteroidetes phyla are the commonest [2]. In patients with active EoE, the genera Streptococcus and Atopobium are decreased while Neisseria and Corynebacterium are increased. Another study showed that the total amount of esophageal bacteria and Haemophilus genus specifically were increased in EoE. Proton pump inhibitors that are used in the treatment of EoE, cause an enrichment of Proteobacteria phylum. Esophageal bacterial load is increased in EoE patients irrespective of treatment or the severity of esophageal mucosal eosinophilia. Like in other allergic and autoimmune diseases, antibiotic treatment and caesarean delivery are associated with increased risk of EoE.[2]

Professor Friedman has reviewed developmental programming.[1] Maternal obesity, diabetes, and western-style diet have an impact on infant stem cells, immune system and gut microbiota. The gut of the newborn is first colonized by aerobes and facultative anaerobes, which are replaced by strict anaerobes. This modifies innate immune signalling, T helper cell immune responses, and endotoxin tolerance. Maternal obesity may disrupt normal microbial colonization and increase the risk of immunologic and metabolic diseases later in the life. Antibiotic treatment during pregnancy increases the risk of childhood obesity. Children born for obese mothers have lesser abundance of two families of fecal Proteobacteria. Furthermore, maternal high-fat diet causes the loss of key bacteria and decrease in bacterial diversity of the infant fecal microbiota.

Also paternal diet may have impact on the health status of the following generations. Zhang and co-workers examined the impact of unhealthy diet in animal model.[2] They fed male rats in two successive generations (F0 and F1) with with high fat, sucrose and salt diet. The control group was fed with normal diet. The high fat-sucrosesalt diet was associated with increased aspartate aminotransferase levels in the next generation (F2). Unhealthy diet was also associated with higher body weight. In female F2 rats, the Shannon index of gut microbiota indicated significantly higher diversity. Variation in the abundance of bacterial genus was associated with liver function abnormalities. Unhealthy diet in F0 and F1 generations was associated with increased serum cholesterol and lipoprotein levels in male F2 rats.

Diet and the gut microbiome

Diet is a key element for the symbiotic interactions between gut microbes and the host, and it is considered as one of the main drivers in shaping the gut microbiota across lifetime, as reviewed by Jack A. Gilbert (UC, San Diego), Susan Devkota (Cedars-Sinai, Los Angeles), and Lipping Zhao (Rutgers, New Jersey). Foods deliver numerous substrates for microbial metabolism and the microbiome is a chemical factory that synthesizes metabolites important for human health. Macro- and micronutrients in food influence the structure and functions of the gut microbial ecosystem in such way that diet appears to be the most important determinant of similarity in gut microbial composition across humans.[1]

Self-reported dietary data from the American Gut project [2] suggest that the number of unique plant species that a subject consumes is associated with microbial diversity, rather than self-reported categories such as “vegan” or “omnivore”. Higher microbial diversity and higher abundance of short chain fatty acid (SCFA) producer species was found in individuals eating more than 30 types of plants per week as compared to those eating less than 10 types of plants per week. The faecal metabolome also differed between both groups. In addition, individuals who consume more than 30 types of plants compared to those who consume 10 or fewer plants had significantly lower abundance of antibiotic resistance genes.

Dysbiosis of the gut microbiome is a definable state with mechanistic implications. It is not just a change in microbial diversity but a rupture of the mutualistic balance between microbiota and host, where inadequate diet plays a detrimental role. During homeostasis, colonocyte metabolism is directed towards oxidative phosphorylation, resulting in high epithelial oxygen consumption. The consequent epithelial hypoxia helps maintain a microbial community dominated by obligate anaerobes, which provide benefit by converting fibre into fermentation products (SCFA) absorbed by the host. Conditions that alter metabolism of the epithelium, such as a fibre poor diet, increase epithelial oxygenation, thereby driving an expansion of facultative anaerobes, a hallmark of dysbiosis in the colon.[3] The shift in the colonic microbiota composition from obligate to facultative anaerobes, associated with many chronic human illnesses, might have a common underpinning in colonocyte dysfunction. As highlighted by Susan Devkota, if choosing a strict or extreme dietary regime, consuming mixed fibre types can support the microbiome and prevent nutrient deficiencies.

The "Foundation Guild"

Lipping Zhao pointed out that our ancestors had much higher intake of dietary fibres than current consumption rates. Reduced intake of fibres and diminished prevalence of SCFA-producing bacteria may underlie many chronic diseases such as type 2 diabetes. In a randomised controlled intervention trial with Chinese type 2 diabetes patients,[4] high intake of diverse dietary fibres (WTP diet) selectively promoted abundance of a group of acetic and butyric acid producer strains in the gut. The WTP diet is based on wholegrains, traditional Chinese medicinal foods and prebiotics. The WTP diet improved glucose homeostasis by reducing glycated haemoglobin, fasting blood glucose and meal tolerance test.

Abundance of the SCFA producers in faeces correlated with the metabolic outcomes and the blood levels of glucagon- like peptide-1 and peptide YY, which induce insulin secretion. Moreover, reduction of faecal pH by SCFA production correlated with inhibition of detrimental bacteria that promote inflammation and suppress glucagon-like peptide-1 production. In addition to providing SCFAs to directly benefit the host, this group of SCFA producers played important ecological functions in the gut microbiota. Lipping Zhao suggested that they work as the “foundation guild” for structuring the healthy gut microbiota. To help patients regain a healthy gut microbiota, “this foundation guild must be re-seeded and re-established”, he said.

FODMAPS and IBS

As reviewed by Magnus Simren (University of Gothenburg), the low fermentable oligosaccharides, disaccharides, monosaccharides and polyols (FODMAPs) diet is now being recommended by up to 85% of doctors to treat functional abdominal symptoms. Clinical trials suggest that some patients have a short-term favourable response to a low FODMAP diet, but whether this dietary advice is clearly better than the first line dietary therapy for IBS is uncertain. Of concern, short-term use of the low FODMAP diet has been associated with potentially unfavourable changes in gut microbiota composition, including reduction of fermentative species (Bifidobacterium, Faecalibacterium and Clostridium cluster XIVa) and increased dysbiotic index scores.[5]

A randomized controlled trial compared effects of the low FODMAP diet or the prebiotic GOS on gut microbiota composition.[6] Changes in faecal microbiota differed between both groups after a 4-week treatment period, particularly in relation to bifidobacteria (increase in the prebiotic group and decrease in the low FODMAP group) and Bilophila wadsworthia (the opposite pattern). Despite distinct effects on microbiota, reductions of symptoms were very similar in both groups. Of interest, the decrease in symptoms persisted during the 2-week follow-up after cessation of prebiotic intake, but reappeared immediately after discontinuation the low FODMAP diet. Modulation of gut microbiota as a treatment strategy for IBS seems promising, but long-term safety aspects need to be taken into account. Diets that reduce symptoms but deteriorate gut health (and general health in the long term) should not be the first choice.

Symbiotic trial to prevent newborn sepsis 

Sepsis in early infancy results in one million annual deaths worldwide, most of them in developing countries. Pinaki Panigrahi presented an intervention study to prevent sepsis among infants in rural India.[7] An oral symbiotic preparation (Lactobacillus plantarum plus fructooligosaccharide) significantly reduced sepsis and death in newborns (risk ratio 0.60, 95% confidence interval 0.48–0.74). This finding suggests that a large proportion of neonatal sepsis in developing countries could be effectively prevented using probiotic- prebiotic treatment.