In Vivo Expression of Chicken Gut Anaerobes Identifies Carbohydrate- or Amino Acid-Utilising, Motile or Type VI Secretion System-Expressing Bacteria.

Complex gut microbiota increases chickens' resistance to enteric pathogens. However, the principles of this phenomenon are not understood in detail. One of the possibilities for how to decipher the role of gut microbiota in chickens' resistance to enteric pathogens is to systematically characterise the gene expression of individual gut microbiota members colonising the chicken caecum. To reach this aim, newly hatched chicks were inoculated with bacterial species whose whole genomic sequence was known. Total protein purified from the chicken caecum was analysed by mass spectrometry, and the obtained spectra were searched against strain-specific protein databases generated from known genomic sequences. Campylobacter jejuni, Phascolarctobacterium sp. and Sutterella massiliensis did not utilise carbohydrates when colonising the chicken caecum. On the other hand, Bacteroides, Mediterranea, Marseilla, Megamonas, Megasphaera, Bifidobacterium, Blautia, Escherichia coli and Succinatimonas fermented carbohydrates. C. jejuni was the only motile bacterium, and Bacteroides mediterraneensis expressed the type VI secretion system. Classification of in vivo expression is key for understanding the role of individual species in complex microbial populations colonising the intestinal tract. Knowledge of the expression of motility, the type VI secretion system, and preference for carbohydrate or amino acid fermentation is important for the selection of bacteria for defined competitive exclusion products.

Colonization of chickens with competitive exclusion products results in extensive differences in metabolite composition in cecal digesta.

The concept of competitive exclusion is well established in poultry and different products are used to suppress the multiplication of enteric pathogens in the chicken intestinal tract. While the effect has been repeatedly confirmed, the specific principles of competitive exclusion are less clear. The aim of the study was to compare metabolites in the cecal digesta of differently colonized chickens. Metabolites in the cecal contents of chickens treated with a commercial competitive exclusion product or with an experimental product consisting of 23 gut anaerobes or in control untreated chickens were determined by mass spectrometry. Extensive differences in metabolite composition among the digesta of all 3 groups of chickens were recorded. Out of 1,706 detected compounds, 495 and 279 were differently abundant in the chicks treated with a commercial or experimental competitive exclusion product in comparison to the control group, respectively. Soyasaponins, betaine, carnitine, glutamate, tyramine, phenylacetaldehyde, or 3-methyladenine were more abundant in the digesta of control chicks while 4-oxododecanedioic acid, nucleotides, dipeptides, amino acids (except for glutamate), and vitamins were enriched in the digesta of chickens colonized by competitive exclusion products. Metabolites enriched in the digesta of control chicks can be classified as of plant feed origin released in the digesta by degradative activities of the chicken. Some of these molecules disappeared from the digesta of chicks colonized by complex microbiota due to them being metabolized. Instead, nucleotides, amino acids, and vitamins increased in the digesta of colonized chicks as a consequence of the additional digestive potential brought to the cecum by microbiota from competitive exclusion products. It is therefore possible to affect metabolite profiles in the chicken cecum by its colonization with selected bacterial species.

Host Species Adaptation of Obligate Gut Anaerobes Is Dependent on Their Environmental Survival.

The gut microbiota of warm-blooded vertebrates consists of bacterial species belonging to two main phyla; Firmicutes and Bacteroidetes. However, does it mean that the same bacterial species are found in humans and chickens? Here we show that the ability to survive in an aerobic environment is central for host species adaptation. Known bacterial species commonly found in humans, pigs, chickens and Antarctic gentoo penguins are those capable of extended survival under aerobic conditions, i.e., either spore-forming, aerotolerant or facultatively anaerobic bacteria. Such bacteria are ubiquitously distributed in the environment, which acts as the source of infection with similar probability in humans, pigs, chickens, penguins and likely any other warm-blooded omnivorous hosts. On the other hand, gut anaerobes with no specific adaptation for survival in an aerobic environment exhibit host adaptation. This is associated with their vertical transmission from mothers to offspring and long-term colonisation after administration of a single dose. This knowledge influences the design of next-generation probiotics. The origin of aerotolerant or spore-forming probiotic strains may not be that important. On the other hand, if Bacteroidetes and other host-adapted species are used as future probiotics, host preference should be considered.

Probiotic Lactobacilli Do Not Protect Chickens against Salmonella Enteritidis Infection by Competitive Exclusion in the Intestinal Tract but in Feed, Outside the Chicken Host.

Lactobacilli are commonly used as probiotics in poultry to improve production parameters and to increase chicken resistance to enteric infections. However, lactobacilli do not efficiently colonise the chicken intestinal tract, and also, their anti-infection effect in vivo is sometimes questionable. In this study, we therefore evaluated the potential of a mixture of four Lactobacillus species (L. salivarius, L. reuteri, L. ingluviei and L. alvi) for the protection of chickens against Salmonella Enteritidis infection. Whenever the chickens were inoculated by lactobacilli and S. Enteritidis separately, there was no protective effect of lactobacilli. This means that when lactobacilli and S. Enteritidis are exposed to each other as late as in the crop of chickens, lactobacilli did not influence chicken resistance to S. Enteritidis at all. The only positive effect was recorded when the mixture of lactobacilli and S. Enteritidis was used for the inoculation of feed and the feed was anaerobically fermented for 1 to 5 days. In this case, chickens fed such a diet remained S. Enteritidis negative. In vitro experiments showed that the protective effect was caused by acidification of feed down to pH 4.6 due to lactobacilli fermentation and was associated with S. Enteritidis inactivation. The probiotic effect of lactobacilli was thus expressed in the feed, outside the chicken host.

Eggshell and Feed Microbiota Do Not Represent Major Sources of Gut Anaerobes for Chickens in Commercial Production.

In this study, we addressed the origin of chicken gut microbiota in commercial production by a comparison of eggshell and feed microbiota with caecal microbiota of 7-day-old chickens, using microbiota analysis by 16S rRNA sequencing. In addition, we tested at which timepoint during prenatal or neonatal development it is possible to successfully administer probiotics. We found that eggshell microbiota was a combination of environmental and adult hen gut microbiota but was completely different from caecal microbiota of 7-day-old chicks. Similarly, we observed that the composition of feed microbiota was different from caecal microbiota. Neither eggshell nor feed acted as an important source of gut microbiota for the chickens in commercial production. Following the experimental administration of potential probiotics, we found that chickens can be colonised only when already hatched and active. Spraying of eggs with gut anaerobes during egg incubation or hatching itself did not result in effective chicken colonisation. Such conclusions should be considered when selecting and administering probiotics to chickens in hatcheries. Eggshells, feed or drinking water do not act as major sources of gut microbiota. Newly hatched chickens must be colonised from additional sources, such as air dust with spores of Clostridiales. The natural colonisation starts only when chickens are already hatched, as spraying of eggs or even chickens at the very beginning of the hatching process did not result in efficient colonisation.

The distribution of antibiotic resistance genes in chicken gut microbiota commensals.

Antibiotic resistance in bacterial pathogens or several indicator bacteria is commonly studied but the extent of antibiotic resistance in bacterial commensals colonising the intestinal tract is essentially unknown. In this study, we aimed to investigate the presence of horizontally acquired antibiotic resistance genes among chicken gut microbiota members in 259 isolates with known whole genomic sequences. Altogether 124 isolates contained at least one gene coding for antibiotic resistance. Genes coding for the resistance to tetracyclines (detected in 101 isolates), macrolide-lincosamide-streptogramin B antibiotics (28 isolates) and aminoglycosides (25 isolates) were the most common. The most frequent tetracycline resistance genes were tet(W), tet(32), tet(O) and tet(Q). Lachnospiraceae and Ruminococcaceae frequently encoded tet(W). Lachnospiraceae commonly coded also for tet(32) and tet(O). The tet(44) gene was associated with Erysipelotrichaceae and tet(Q) was detected in the genomes of Bacteroidaceae and Porphyromonadaceae. Without any bias we have shown that antibiotic resistance is quite common in gut commensals. However, a comparison of codon usage showed that the above-mentioned families represent the most common current reservoirs but probably not the original host of the detected resistances.

Different Bacteroides Species Colonise Human and Chicken Intestinal Tract.

Bacteroidaceae are common gut microbiota members in all warm-blooded animals. However, if Bacteroidaceae are to be used as probiotics, the species selected for different hosts should reflect the natural distribution. In this study, we therefore evaluated host adaptation of bacterial species belonging to the family Bacteroidaceae. B. dorei, B. uniformis, B. xylanisolvens, B. ovatus, B. clarus, B. thetaiotaomicron and B. vulgatus represented human-adapted species while B. gallinaceum, B. caecigallinarum, B. mediterraneensis, B. caecicola, M. massiliensis, B. plebeius and B. coprocola were commonly detected in chicken but not human gut microbiota. There were 29 genes which were present in all human-adapted Bacteroides but absent from the genomes of all chicken isolates, and these included genes required for the pentose cycle and glutamate or histidine metabolism. These genes were expressed during an in vitro competitive assay, in which human-adapted Bacteroides species overgrew the chicken-adapted isolates. Not a single gene specific for the chicken-adapted species was found. Instead, chicken-adapted species exhibited signs of frequent horizontal gene transfer, of KUP, linA and sugE genes in particular. The differences in host adaptation should be considered when the new generation of probiotics for humans or chickens is designed.

Gut Anaerobes Capable of Chicken Caecum Colonisation.

Chicks in commercial production are highly sensitive to enteric infections and their resistance can be increased by administration of complex adult microbiota. However, it is not known which adult microbiota members are capable of colonising the caecum of newly hatched chicks. In this study, we therefore orally inoculated chicks with pure cultures of 76 different bacterial isolates originating from chicken caecum on day 1 of life and determined their ability to colonise seven days later. The caecum of newly hatched chickens could be colonised by bacteria belonging to phyla Bacteroidetes, Proteobacteria, Synergistetes, or Verrucomicrobia, and isolates from class Negativicutes (phylum Firmicutes). On the other hand, we did not record colonisation with isolates from phyla Actinobacteria and Firmicutes (except for Negativicutes), including isolates from families Lachnospiraceae, Ruminococcaceae, Erysipelotrichaceae, and Lactobacillaceae. Representatives of genera commonly used in probiotics such as Lactobacillus, Enterococcus, or Bacillus therefore did not colonise the chicken intestinal tract after a single dose administration. Following challenge with Salmonella enterica serovar Enteritidis, the best protecting isolates increased the chicken's resistance to S. Enteritidis only tenfold, which, however, means that none of the tested individual bacterial isolates on their own efficiently protected chicks against S. Enteritidis.

Systematic Culturomics Shows that Half of Chicken Caecal Microbiota Members can be Grown in Vitro Except for Two Lineages of Clostridiales and a Single Lineage of Bacteroidetes.

Epidemiological data show that the composition of gut microbiota influences host health, disease status, and even behaviour. However, to confirm these epidemiological observations in controlled experiments, pure cultures of gut anaerobes must be obtained. Since the culture of gut anaerobes is not a simple task due to the large number of bacterial species colonising the intestinal tract, in this study we inoculated 174 different culture media with caecal content from adult hens, and compared the microbiota composition in the original caecal samples and in bacterial masses growing in vitro by 16S rRNA sequencing. In total, 42% of gut microbiota members could be grown in vitro and since there were some species which were not cultured but for which the culture conditions are known, it is likely that more than half of chicken gut microbiota can be grown in vitro. However, there were two lineages of Clostridiales and a single lineage of Bacteroidetes which were common in chicken caecal microbiota but resistant to culture. Of the most selective culture conditions, nutrient broths supplemented with mono- or di-saccharides, including those present in fruits, positively selected for Lactobacillaceae. The addition of bile salts selected for Veillonellaceae and YCFA (yeast casitone fatty acid agar) enriched for Desulfovibrionaceae. In addition, Erysipelotrichaceae were positively selected by colistin, trimethoprim, streptomycin and nalidixic acid. Culture conditions tested in this study can be used for the selective enrichment of desired bacterial species but also point towards the specific functions of individual gut microbiota members.

Contact with adult hen affects development of caecal microbiota in newly hatched chicks.

Chickens in commercial production are hatched in a clean hatchery environment in the absence of any contact with adult hens. However, Gallus gallus evolved to be hatched in a nest in contact with an adult hen which may act as a donor of gut microbiota. In this study, we therefore addressed the issue of microbiota development in newly hatched chickens with or without contact with an adult hen. We found that a mere 24-hour-long contact between a hen and newly hatched chickens was long enough for transfer of hen gut microbiota to chickens. Hens were efficient donors of Bacteroidetes and Actinobacteria. However, except for genus Faecalibacterium and bacterial species belonging to class Negativicutes, hens did not act as an important source of Gram-positive Firmicutes. Though common to the chicken intestinal tract, Lactobacilli and isolates from families Erysipelotrichaceae, Lachnospiraceae and Ruminococcaceae therefore originated from environmental sources instead of from the hens. These observation may have considerable consequences for the evidence-based design of the new generation of probiotics for poultry.

Antibiotic Resistance, Core-Genome and Protein Expression in IncHI1 Plasmids in Salmonella Typhimurium.

Conjugative plasmids from the IncHI1 incompatibility group play an important role in transferring antibiotic resistance in Salmonella Typhimurium. However, knowledge of their genome structure or gene expression is limited. In this study, we determined the complete nucleotide sequences of four IncHI1 plasmids transferring resistance to antibiotics by two different next generation sequencing protocols and protein expression by mass spectrometry. Sequence data including additional 11 IncHI1 plasmids from GenBank were used for the definition of the IncHI1 plasmid core-genome and pan-genome. The core-genome consisted of approximately 123 kbp and 122 genes while the total pan-genome represented approximately 600 kbp. When the core-genome sequences were used for multiple alignments, the 15 tested IncHI1 plasmids were separated into two main lineages. GC content in core-genome genes was around 46% and 50% in accessory genome genes. A multidrug resistance region present in all 4 sequenced plasmids extended over 20 kbp and, except for tet(B), the genes responsible for antibiotic resistance were those with the highest GC content. IncHI1 plasmids therefore represent replicons that evolved in low GC content bacteria. From their original host, they spread to Salmonella and during this spread these plasmids acquired multiple accessory genes including those coding for antibiotic resistance. Antibiotic-resistance genes belonged to genes with the highest level of expression and were constitutively expressed even in the absence of antibiotics. This is the likely mechanism that facilitates host cell survival when antibiotics suddenly emerge in the environment.

Complete sequence of multidrug resistance p9134 plasmid and its variants including natural recombinant with the virulence plasmid of Salmonella serovar Typhimurium.

In this study we determined the complete nucleotide sequence of multidrug-resistance plasmid p9134, and its variants p9134dT and p9134dAT which spontaneously lost either tetracycline or both tetracycline and ampicillin resistance, respectively. The plasmids were 133,802 bp, 109,512 bp and 127,291 bp in size, respectively, and their basic backbone was similar to that of IncI plasmids. Genes coding for ampicillin (blaTEM), chloramphenicol (catA1), streptomycin (strA, strB), tetracycline (tetA(A)) and gentamicin (aac(3)-IV) resistance were confirmed in wild-type p9134. Moreover, a gene for hygromycine resistance (hph) and a putative gene for apramycin resistance were newly determined. In p9134dAT, a continuous sequence coding for ampicillin and tetracycline resistances was lost. Genetic rearrangements in p9134dT were more complex and 2 recombination events must have occurred. During the first one, the tetracycline resistance locus was replaced with rck, srgB, srgA, orf7 and pefI originating from Salmonella virulence plasmid pSLT. During the second one, ydjA, pifA and repC genes from p9134 were replaced with repA2, PSLT025 and PSLT026 genes from pSLT. Our findings indicate that recombination event between unrelated plasmids might be quite common and may lead to the generation and selection of plasmids both transferring antibiotic resistance and increasing virulence of their host.

allB, allantoin utilisation and Salmonella enterica serovar Enteritidis and Typhimurium colonisation of poultry and mice.

Natural variation in the presence or the absence of STM0517-0529 genes allowing allantoin utilisation has been described in field isolates of the multidrug resistant Salmonella enterica serovar Typhimurium belonging to the phage type DT104. Interestingly, S. enterica subspecies enterica serovar Typhimurium DT104 is quite frequent in pigs and cattle, but rarely present in egg-laying hens. Taking into account the different mode of allantoin metabolism in birds and mammals, we were interested in whether the absence of STM0517-0529 genes may disable this clone in poultry colonisation. We have therefore constructed the allB (also designated as STM0523) mutants in S. enterica subspecies enterica serovar Typhimurium and S. enterica subspecies enterica serovar Enteritidis, and with these, we infected mice, newly hatched chickens and adult egg-laying hens to show that the defect in allantoin utilisation does not influence S. enterica virulence for mice or adult hens, but slightly decreases virulence of S. enterica for chickens. The decrease in virulence of the allB mutant was relatively minor as it could be observed only after a mixed infection model, consistent with a lower prevalence, but not a total absence of such clones in poultry flocks.

Retron reverse transcriptase (rrtT) can be lost in multidrug resistant Salmonella enterica serovar Typhimurium DT 104 strains and influences virulence for mice.

In Salmonella enterica serovar Typhimurium, retron reverse transcriptase (rrtT), which is part of St 85 retron, is quite ubiquitous and is located in the thdF -yidY intergenic region. In this study, we showed that rrtT is relatively unstable in multidrug resistant, Salmonella genomic island 1 (SGI 1) positive strains. Out of 365 field strains, 55 were free of retron. In 54 of the rrtT negative strains, the excision must have occurred by the same mechanism in which the rrtT together with five other genes was excised. Altogether 8164 bp was missing in the chromosome of the rrtT negative strains. Since the deletion happened exactly between the right inverted repeat of IS 6100 and inside the yieE gene, we propose that intramolecular transposition of IS 6100 followed by homologous recombination was responsible for the excision. Excision of retron together with the right end of SGI 1 may also result in its stabilisation in the Salmonella typhimurium genome. Experimental deletion of rrtT resulted in an accelerated course of infection in orally infected mice. Since the retron excision occurred exclusively in multidrug resistant S. typhimurium, it cannot be excluded that such strains may increase their virulence in the future.

Salmonella enterica subsp. enterica serovar Enteritidis harbours ColE1, ColE2, and rolling-circle-like replicating plasmids.

Using DNA hybridization, at least three distinct groups of low molecular mass plasmids were identified in Salmonella enterica subsp. enterica serovar Enteritidis. After sequencing representative plasmids from each group, we concluded that they belonged to ColE1, ColE2, and rolling-circle-like replicating plasmids. Plasmid pK (4245 bp) is a representative of widely distributed ColE1 plasmids. Plasmid pP (4301 bp) is homologous to ColE2 plasmids and was present predominantly in single-stranded DNA form. The smallest plasmids pJ (2096 bp) and pB (1983 bp) were classified as rolling-circle-like replicating plasmids. Both encoded only a single protein essential for their own replication, and they must have existed in an unusual molecular structure, as (i) they were capable of hybridization without denaturation, (ii) their DNA could be linearized with S1 nuclease, and (iii) even after such treatment, the ability to hybridize without denaturation did not disappear.

Retron reverse transcriptase rrtT is ubiquitous in strains of Salmonella enterica serovar Typhimurium.

Bacterial retron reverse transcriptases are unusual enzymes which utilise the same RNA molecule as a template and also as a primer for initiation of the reverse transcription. Except for their relatively frequent presence in Myxococcus spp., they are considered as quite rare proteins. However, in this study we proved that retron reverse transcriptase is frequently found in certain serovars of Salmonella enterica. Using polymerase chain reaction (PCR), in strains of serovar Typhimurium, the rrtT (retron reverse transcriptase Typhimurium) gene was detected in 158 out of 175 tested field strains. On the other hand, in none of the 18 tested serovar Enteritidis strains the rrtT was detected in their genome. Detailed computer analysis allowed us to predict the sequence of msDNA and to propose that the final msDNA is free of any RNA. Furthermore, we predict that there are at least three different classes of retron reverse transcriptases.