Luminal and mucosa-associated caecal microbiota of chickens after experimental Campylobacter jejuni infection in the absence of Campylobacter-specific phages of group II and III

Abstract

Campylobacteriosis is still the most commonly reported zoonosis in the European Union causing gastrointestinal disease in humans. One of the most common sources for these food-borne infections is broiler meat. Interactions between Campylobacter (C.) jejuni and the intestinal microbiota might influence Campylobacter colonization in chickens. The aim of the present study was to gain further knowledge about exclusive interactions of the host microbiota with C. jejuni in Campylobacter -specific phage-free chickens under standardized conditions and special biosafety precautions.

Therefore, 12 artificially infected ( C. jejuni inoculum with a challenge dose of 7.64 log₁₀ c.f.u.) and 12 control chickens of the breed Ross 308 were kept under special biosafety measures in an animal facility. At day 42 of life, microbiota studies were performed on samples of caecal digesta and mucus. No Campylobacter -specific phages were detected by real-time PCR analysis of caecal digesta of control or artificially infected chickens. Amplification of the 16S rRNA gene was performed within the hypervariable region V4 and subsequently sequenced with Illumina MiSeq platform. R (version 4.0.2) was used to compare the microbiota between C. jejuni -negative and C. jejuni -positive chickens. The factor chickens’ infection status contributed significantly to the differences in microbial composition of mucosal samples, explaining 10.6 % of the microbiota variation (P=0.007) and in digesta samples, explaining 9.69 % of the microbiota variation (P=0.015). The strongest difference between C. jejuni -non-infected and C. jejuni -infected birds was observed for the family Peptococcaceae whose presence in C. jejuni -infected birds could not be demonstrated. Further, several genera of the family Ruminococcaceae appeared to be depressed in its abundance due to Campylobacter infection. A negative correlation was found between Christensenellaceae R-7 group and Campylobacter in C. jejuni -colonised chickens, both genera potentially competing for substrate. This makes Christensenellaceae R-7 group highly interesting for further studies that aim to find control options for Campylobacter infections and assess the relevance of this finding for chicken health and Campylobacter colonization.

Data Summary

The following supplementary data are available. Table S1: Relative abundance (%) of bacterial phyla in luminal or mucosa-associated microbiota depending on C. jejuni status. Table S2: Comparison of bacterial relative abundance in samples of caecal content and mucus (at taxonomic level, family) between chickens of different C. jejuni status. Table S3: OTUs with significant (padj <0.05) different abundance between chickens of different C. jejuni status in samples of caecal content. Table S4: OTUs with significant (padj <0.05) different abundance between chickens of different C. jejuni status in samples of caecal mucus. Table S5: Real-time PCR results. Table S6: Sample overview.

The authors confirm all supporting data and protocols have been provided within the article or through supplementary data files.

Impact Statement.

Preliminary studies have shown interactions between intestinal microbiota and Campylobacter spp. influencing host susceptibility against this pathogen. As intestinal microbiota seems to impact Campylobacter (C.) jejuni colonization, load and clearance, the long-term goal is to identify bacterial species or specific microbiota compositions that might promote colonization resistance. Any reduction in Campylobacter load in chicken intestines is a further step towards increasing food safety because campylobacteriosis is the most commonly reported zoonosis in the European Union causing gastrointestinal disease. However, further knowledge is needed concerning interactions of the host microbiota and C. jejuni . When examining these relationships, it is important to ensure the absence of naturally occurring Campylobacter -specific bacteriophages, as they are present in about one third of C. jejuni positive conventional barn-reared broiler chickens and their presence is associated with a significant reduction in C. jejuni counts. For this reason, the present study was conducted under special biosafety measures. This study also explores possible interactions within the caecal luminal and mucosa-associated microbiota under standardized conditions and biosafety precautions, which is why the study could reveal important knowledge about exclusive interactions of the host microbiota and C. jejuni . The results of this study contribute to the goal of finding further control options for Campylobacter infections in chickens.

Introduction

In 2019, campylobacteriosis was the most commonly reported zoonosis in the European Union causing gastrointestinal disease in humans, while one of the most common sources for the food-borne infections was broiler meat [1]. For this reason, a process hygiene criterion was established for Campylobacter in 2017 [2], putting pressure on broiler meat producers to reduce the Campylobacter load. Chilled broiler carcases should comply with a limit of 1000 c.f.u. g⁻¹. However, currently no efficient control measures are available. Thus, control options for Campylobacter in broilers, focusing on primary production are still of major interest [3]. Since Campylobacter (C.) jejuni colonization in chickens is highly prevalent but asymptomatic, it appears to act as a commensal in this host (reviewed in Newell [4]). Various studies on animals and humans have shown that the composition of the intestinal microbiota plays a significant role in host susceptibility to Campylobacter species [5–8]. In humans, for example, a less diverse faecal microbiota was associated with a higher susceptibility to Campylobacter infections [8]. Even in chickens, studies revealed an influence of the microbiota on the outcome of a Campylobacter infection [5–7]. Compared to their hatchmates, higher C. jejuni counts and histopathological gut lesions were found in chickens that were raised under germ-free conditions or treated with an antibiotic cocktail [7]. C. jejuni invades the caecum of antibiotic-treated birds with limited gut microbiota and induces clinical signs and lesions, leading the authors to suggest that C. jejuni is a primary pathogen for chickens as well [6].

As intestinal microbiota seem to have an influence on C. jejuni colonization, load and clearance, the long-term goal is to identify bacterial species or specific microbiota compositions that could induce the colonization resistance. However, these efforts still require further knowledge concerning interactions of C. jejuni and the host microbiota.

Characteristics of C. jejuni suggest an adaptation to the environment of the mucus, which might also influence C. jejuni pathogenicity [9]. In addition, distinct community structures are observed in the luminal and mucosal samples, which necessitate studying the variations between the bacterial communities of the lumen and mucosa, thus improving the understanding of host–microbe interactions [10].

Along with bacteria, viruses are present in the chicken’s intestines that can influence Campylobacter [11]. It could be shown that naturally occurring Campylobacter -specific lytic bacteriophages are present in about one third of C. jejuni positive caecal samples of conventional barn-reared broiler chickens, and C. jejuni counts in the presence of bacteriophages were associated with a significant reduction [11].

The aim of the present study was to investigate the luminal and mucosa-associated caecal microbiota of experimentally C. jejuni -infected chickens under the absence of Campylobacter -specific phages. Results from this study may help to better understand the interactions of the host microbiota and C. jejuni in chickens, with the aim of finding further control options for primary production in the long term.

Methods

A total of 24 newly hatched chickens (day 0) of both sexes (Ross 308) were obtained from a commercial hatchery (BWE-Brüterei Weser Ems, PHW Gruppe/Lohmann and Co. AG, Visbek, Germany). They were kept in two separate rooms in a biosafety level (BSL) 2 animal facility of the Research Center for Emerging Infections and Zoonoses (RIZ) at the University of Veterinary Medicine Hannover, Foundation, Hanover, Germany. The RIZ animal facility including BSL-2 stables harbours a special technology to reduce the pressure in all labs/stables. Directional air flow is established from clean areas into contaminated areas. If multiple containment zones exist within an area, sequentially more negative pressure differentials are established so that the more contaminated spaces are maintained at a negative pressure with respect to less contaminated areas. Access to the area was controlled and restricted to designated employees. Effective control of vectors was accomplished via a threefold barrier system. All items brought into the rooms were subjected to germ reduction measures. Each room had its own equipment. In both rooms, an identical box was installed, which was divided by a continuous wall and the 12 birds were kept first exclusively on one side of the box. Each half of the box was littered with wood shavings (1 kg/m²). Stocking density amounted to a maximum of 30 kg per square metre. The birds were maintained under a 16-hour light, 8-hour dark lighting schedule during the entire experiment.

The animals were fed in three phases with conventional complete diets. The starter diet was offered in the first week of life, the first grower diet in the second week of life and the second grower diet from day 14 onwards. The diets were analysed by standard procedures in accordance with the official methods of the VDLUFA [12] as described in Visscher, Klingenberg [13]. The second grower diet was mainly composed of wheat, extracted soybean meal, corn, extracted rapeseed meal, peas, soybean oil, and palm oil and contained 19.7 % crude protein and 12.5 MJ AME per kg diet.

At day 17 of life, each bird in the first room (12 birds) was administered orally with one millilitre of a C. jejuni inoculum (challenge dose of 7.64 log₁₀ c.f.u./1 ml of challenge inoculum). The C. jejuni inoculum was a mixture of the C. jejuni strain C356 (DSM 24306, Leibniz Institute DSMZ – German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany) and a strain that was isolated from sheddings of chickens participating in a previous microbiota study by Hankel et al. [5]. Prior to the described experimental challenge, Campylobacter exclusion diagnosis had been performed via qualitative bacteriological examination. In order to ensure identical environmental and at the same time to guarantee controlled infection conditions for each animal (non-infected and artificially infected birds), two animals from each box changed rooms every 3 days (see Fig. 1). The continuous wall prevented direct contact between the non-infected and artificially infected birds when kept in one box during the whole experiment. This effort guaranteed that each animal was kept once in each room and box, therefore exposed to identical environmental conditions.

Fig. 1.

Chickens kept in two rooms under special biosafety measures in the animal facility. Every 3 days, two animals from each box changed rooms. The continuous wall prevented direct contact between the artificially C. jejuni -infected and non-infected birds kept in one box. The figure was created with biorender.com.

Three weeks after the experimental challenge with C. jejuni , all chickens were weighed and dissected (day 42). Anaesthesia was performed by head stroke. After bleeding, caecal contents were removed under sterile conditions and placed in reaction vessels. The caecal contents were immediately frozen and stored at −80 °C for further microbiota analyses and examination of the presence or absence of Campylobacter -specific bacteriophages. In addition, mucus samples from the caecum were collected in accordance with the protocol described below.

Bacteriological analyses

Qualitative bacteriological examination of C. jejuni performed in caecal contents of all birds was based on the DIN EN ISO 10272–1 : 2006, taken from the official collection of analysis methods in accordance with § 64 LFBG. Quantitative bacteriological examination was performed as described in Hankel et al. [5].

Campylobacter -specific bacteriophages

Detection of Campylobacter -specific group II and III bacteriophages in caecal content was carried out as described by Jäckel et al. [14] with some modifications. In brief, primers and probes were received from biomers.net (biomers.net GmbH, Ulm, Germany), using primer and probe sequences described by Jäckel et al. [14]. For real-time PCR, a 1 g sample of each caecal content was thoroughly mixed with 9 ml of 0.9 % sodium chloride and centrifuged at 13.000  g for 10 min at 4 °C to remove debris. Subsequently, the supernatant was heated at 95 °C for 20 min. For CPGIII- and CPGII/GIII-detection, multiplex real-time PCR was used: 1 µl of each sample suspension was mixed with 5 µl QuantiNova Multiplex PCR Master Mix (Quiagen GmbH, Hilden, Germany), each 0.8 µl of CPGIII and CPGII/GIII forward and reverse primers, 0.25 µl CPGIII and CPGII/GIII probe and 10.30 µl water. Amplification was done as described for the QuantiNova Multiplex PCR Kit. For CPGII-detection, one microlitre of each sample suspension was mixed with 5 µl QuanitNova Multiplex PCR Master Mix, each 0.8 µl of CPGII forward and reverse primers, 0.25 µl CPGII probe and 12.15 µl water. The amplification conditions were as follows: initial denaturation at 95 °C for 2 min followed by 40 cycles of 95 °C for 5 s, and 55 °C for 30 s. Reactions were performed using the LightCycler 480 instrument II (Roche Diagnostics GmbH, Mannheim, Germany). Analysis of ct values and fluorescence emission intensity curves was performed using LightCycler 480 software Release 1.5OSP3 IDEAS 2.0. A ct value cut-off of ≤37 as validated by Jäckel et al. [14] was used. Samples with ct values above 37 or no rising fluorescence emission intensity curves were considered to contain no Campylobacter -specific group II or III phages (Table S5).

Preparation of mucosal samples

Cell-wall-associated bacteria were obtained based on the described procedure by Gong et al. [15]. Both caeca were opened longitudinally and immediately washed in saline three times, then washed twice in saline containing 0.1 % Tween 80 under vigorous hand shaking for 30 s per wash. In contrast to the method described, only the last washing solution was centrifuged (27000  g , 20 min) at 4 °C to pellet the bacterial cells that were released from the caecal wall. This fraction of bacterial cells was referred to as mucosa-associated bacteria in the present investigation and immediately frozen and stored at −80 °C for further microbiota analyses.

16S rRNA gene sequencing

DNA extraction

A total of 48 samples of caecal contents and mucus taken from 24 birds were immediately frozen and stored until simultaneous analysis at a temperature of −80 °C. A phenol-chloroform based protocol was used to isolate total DNA from samples. Therefore, the obtained samples were centrifuged (3500 r.p.m.) and suspended with 500 µl of extraction buffer (200 mM Tris, 20 mM EDTA, 200 mM NaCl, pH 8.0), 200 µl of 20 % SDS (sodium dodecyl sulphate), 500 µl of phenol:chloroform:isoamyl alcohol (24 : 24 : 1) and 100 µl of zirconia/silica beads (0.1 mm diameter). The samples were homogenized with a bead beater (BioSpec) for 2 min, DNA was precipitated with absolute isopropanol and washed with 70 % vol. ethanol. DNA extracts were suspended in TE Buffer with 100 µg ml⁻¹ RNAse I and additionally column purified. Finally, total DNA was quantified and diluted to 25 ng µl⁻¹.

Sequencing and data processing

16S rRNA gene analysis was performed as described in Hankel et al. [5]. A purification step (Kit: BS 365, Bio Basic, Markham, Ontario, Canada) was performed before the hypervariable region V4 of the 16S rRNA gene was amplified in accordance with previously described protocols using the primer pair F515 (5′-GTGCCAGCMGCCGCGGTAA-3′) and R806 (3′-TAATCTWTGGGVHCATCAGG-5′) [16]. Amplicons were sequenced on the Illumina MiSeq platform (PE250) and the Usearch8.1 software package (http://www.drive5.com/usearch/) was used to assemble, quality control and cluster the obtained reads. The reads were merged and chimeric sequences were identified and removed. Quality filtering was set up with fastq_filter (-fastq_maxee 1) according to a minimum read length of 200 bp before reads were clustered into 97 % ID operational taxonomic units (OTUs). OTU clusters and representative sequences were determined with the UPARSE algorithm [17]. Taxonomy assignment was carried out with the help of Silva database v128 [18] and the Naïve Bayesian Classifier from the Ribosomal Database Project (RDP) [19]. The bootstrap confidence cut-off was set at 70 %.

Samples with fewer than 999 total reads were removed. Therefore, 47 out of 48 samples were included in statistical analyses of microbiota. The dataset contained 1 529 785 reads (mean number of reads: 32548; range: 17 644–40 847) mapped to 191 OTUs. The rarefaction curves were plotted using with the R-package ‘vegan’ (version 2.5.6, Figure S1).

Statistical analyses

Statistical analyses of microbiota were performed using R (version 4.0.2, www.r-project.org) with the R-package ‘phyloseq’ (version 1.32.0 [20]). Ordination was performed using Bray–Curtis dissimilarity-based principal coordinate analysis (PCoA) also provided in the R-package ‘phyloseq’. Factors contributing to the differences in microbial composition of the samples were identified with permutational multivariate analysis of variance (PERMANOVA) on Bray–Curtis distances via the adonis function of the ‘vegan’ package (version 2.5.6 [21]), whereby to evaluate the contribution of the factor C. jejuni infection, OTU_90 (genus Campylobacter ) was pruned from the dataset. Sample diversity was measured with the species richness estimators Observed Species, Chao 1 and Shannon index. Data were checked for normality by analysing the model residuals with the Shapiro–Wilk normality test before pairwise comparisons were conducted, all implemented in the package ‘rstatix’ (version 0.6.0 [22]). To identify bacterial phyla with significantly different abundance, multiple testing also included in the R-package ‘phyloseq’ was used on normalized counts. P-values from this test were adjusted by the Benjamini and Hochberg (BH) method to control for the false discovery rate (FDR) of 5 %. To find OTUs with significantly different abundance between birds, abundance counts were compared using the R-package ‘DESeq2’ (version 1.22.2), which uses tests based on the negative binomial distribution [23]. OTUs were filtered using false discovery rate (FDR) cut-off 0.05. Results were visualized with the help of ggplot2 (version 3.3.5 [24]). Finally, relative abundance of bacterial genera within luminal and mucosa-associated microbiota in C. jejuni -positive chickens that had a relative abundance >1 % in at least one sample were tested for association with the R-package ‘corrplot’ (version 0.84 [25]) using Spearman’s rank correlation.

Results

Performance data

The experiments ran without complications. The final body weight at day 42 amounted on average to 3 015 g±313 ( C. jejuni -negative birds) and 3 140 g±361 ( C. jejuni -positive birds) exceeding performance objectives of Aviagen at day 42 [26].

Bacteriological examination and detection of Campylobacter -specific bacteriophages

Qualitative bacteriological examination revealed all caecal samples of experimentally C. jejuni -infected birds as being C. jejuni positive (5.31±0.69 log₁₀ c.f.u. g⁻¹), while test results of the non-infected birds were C. jejuni negative.

The presence of Campylobacter -specific phages of group II and III could be excluded.

Luminal and mucosa-associated caecal microbiota

Independent of C. jejuni status, Firmicutes dominated the microbiota of caecal contents at phylum level, while Firmicutes and Proteobacteria dominated the microbiota of caecal mucus. Relative abundance of bacterial phyla within all samples is shown in Fig. 2. Luminal and mucosa-associated microbiota showed significant differences regarding bacterial abundance (Table S2).

Fig. 2.

Relative abundance of (a) bacterial phyla and (b) families in luminal and mucosa-associated caecal microbiota of chickens with and without experimental C. jejuni infection.

Both factors, the sampling point (P=0.001) and the chickens’ infection status (P=0.001) contributed significantly to the differences in microbial composition of the samples. The factor sampling point explained 15.6 % of the sample’s variability, whereas the factor infection status 7.39 % thereof. At both sampling points, luminal and mucosa-associated caecal microbiota, the tested variable C. jejuni status altered the microbiota to almost the same extent. The tested variable C. jejuni status explained 9.69 % of the microbiota variation in caecal contents (P=0.015), while the C. jejuni status explained 10.6 % of the microbiota variation in samples of caecal mucus (P=0.007). Fig. 3 shows PCoA based on Bray–Curtis dissimilarity of samples separated by the two sampling points, caecal content and mucus, and infection status.

Fig. 3.

Bray–Curtis dissimilarity-based principal coordinate analysis (PCoA). Each point represents a different bird with C. jejuni infection (filled points) or a C. jejuni -non-infected bird (open points); coloured lines connect samples of one sampling point: caecal content (green) and caecal mucus (blue).

Pairwise comparisons of measured species richness estimators Observed Species, Chao 1 and Shannon index revealed no statistically significant differences between birds with and without experimental C. jejuni infection, neither in luminal nor in mucosa-associated microbiota (Fig. 4).

Fig. 4.

Alpha diversity of luminal and mucosa-associated caecal microbiota of C. jejuni -positive and C. jejuni -negative chickens. Box-plots showing alpha diversity in samples using the species richness estimators Observed Species, Chao1 and Shannon index.

Multiple testing on normalized counts on each phylum of luminal microbiota yielded no significant differences between animals of different C. jejuni status. In contrast to the luminal-associated microbiota, mucosal microbiota showed a shift to an 7.7% points higher relative abundance of the phylum Firmicutes in C. jejuni -infected birds, as well as an 7.85% points reduction in Proteobacteria relative abundance (relative abundance of Firmicutes in C. jejuni -non-infected birds: 87.3 % and C. jejuni -infected birds: 95.0 %; relative abundance of Proteobacteria in C. jejuni -non-infected birds: 12.1 % and C. jejuni -infected birds: 4.25 %; Table S1). Nevertheless, differences in these phyla between non-infected and infected birds were not significant.

At species level, 24 out of 191 OTUs showed a significantly different abundance between caecal contents of C. jejuni -infected and non-infected birds (Table S3), while in samples of caecal mucus, 30 OTUs were significantly different between C. jejuni -infected and non-infected birds (Table S4). Log₂-fold changes for these OTUs grouped by genus are shown in Fig. 5.

Fig. 5.

Differential analysis of OTUs using the DESeq2 package (significance threshold for padj <0.05) comparing C. jejuni -negative and C. jejuni -positive chickens in (a) caecal content, (b) caecal mucus. Each point represents a single OTU grouped by genus and by colour according to which taxonomic family the OTU originates.

With two exceptions, especially sequences of OTUs belonging to genera within the family Ruminococcaceae were enriched in caecal contents of non-infected compared to C. jejuni -infected chickens (Fig. 5a). These bacterial sequences could be assigned to five different genera, Ruminococcaceae UCG-004, Intestinimonas , Ruminococcaceae UCG-014, Anaerotruncus and Subdoligranulum, and seem to be depressed as a result of the artificial Campylobacter infection. Even if bacterial members of the genus Ruminococcaceae UCG-014 were enriched in non-infected birds, one bacterial species of this genus was clearly enriched in caecal contents of C. jejuni -infected chickens. Sequences of one OTU assigned to the genus Enterococcus , as the only member of this genus in the present study, was enriched C. jejuni -infected compared to non-infected chickens. This observation was limited to samples of caecal content. In addition, sequences of OTUs belonging to two genera of the family Clostridiales vadinBB60 group were enriched in caecal contents of C. jejuni -infected birds. A positive correlation was found between one further unknown genus belonging to this family Clostridiales vadinBB60 group and the genus Campylobacter in caecal contents of C. jejuni -infected chickens (Fig. 6a).

Fig. 6.

Spearman’s correlation between 33 genera (relative abundance >1 % in at least one sample) of (a) luminal and (b) mucosa-associated microbiota in C. jejuni -positive chickens. Positive correlations are displayed in blue and negative correlations in red, while colour intensity and the circle size are proportional to the correlation coefficients. *: P-value <0.05, **: P-value <0.01, ***: P-value <0.001.

In both, luminal as well as mucosa-associated microbiota, sequences of the bacterial member of an uncultured genus belonging to the family Peptococcaceae differed clearly between non-infected and C. jejuni -infected chickens (Fig. 5). This genus of the family Peptococcaceae is its only bacterial member in the present study, whose existence in C. jejuni -infected birds was not detected. Not as clear as in caecal contents, but just as noticeable, sequences of OTUs belonging to genera within the family Ruminococcaceae were enriched in caecal mucus of non-infected compared to C. jejuni -infected chickens (Fig. 5b). Within mucosa-associated microbiota of C. jejuni -infected chickens, a strong negative association was found between the genus Ruminococcaceae UCG-014 (−0.80, P=0.003), one futher unknown genus of the same family (−0.77, P=0.005) and the genus Campylobacter (Fig. 6b). Additional genera were enriched within mucosa-associated microbiota of non-infected chickens and therefore seem to be depressed due to Campylobacter infection only at this site of the caecum; under them Ruminiclostridium 5, Blautia , Akkermansia , Alistipes , Staphylococcus , Lactobacillus , unclassified genus of the order Clostridiales , and [Eubacterium] hallii group.

Strengths of an association between bacterial genera and the direction of the relationship with the Spearman’s rank correlation shown in Fig. 6 revealed a strong negative correlation between the genus Campylobacter and Christensenellaceae R-7 group in both luminal and mucosa- associated caecal microbiota of C. jejuni -positive chickens. The strength of association was greater in luminal as in mucosa-associated microbiota (caecal content: −0.84, P <0.001; caecal mucus: −0.67, P=0.023). Interestingly, this genus was enriched in luminal and mucosa-associated microbiota of C. jejuni -infected compared to non-infected chickens (Fig. 5). In addition, there was a positive association of the genera Campylobacter and Faecalibacterium in luminal as well as in mucosa-associated caecal microbiota of C. jejuni -positive chickens (caecal content: 0.66, P=0.019; caecal mucus: 0.70, P=0.016).

Discussion

The present infection trial took place under absolutely controlled conditions constituting the prerequisites for an exclusive investigation of possible interactions between caecal microbiota and Campylobacter .

Influence of C. jejuni on composition of microbial communities in caeca of chickens

In the present study, it could be shown that the C. jejuni infection had significant influence on microbial composition of the samples. Similar results were seen in investigations by Awad et al. [27] where the gut microbial communities changed as a result of infection as well. Only 25–36 % of the observed OTUs in the jejunum and caecum were shared between the control and infected birds. Nevertheless, the number of shared OTUs between the control and infected birds at day 21 (1 week after infection) was smaller compared to day 28 (2 weeks after infection). In addition, Thibodeau et al. [28] observed that caecal beta-diversity was only moderately modified (visual analysis of the NMDS graph) albeit significantly after C. jejuni colonisation (UniFrac significance P=0.001). This shows that the impact of Campylobacter on microbiota composition seems to vary according to the age of the chickens and the time passed after infection, being less prominent for older chickens and the time after infection.

Alpha diversity did not differ between birds with and without experimental C. jejuni infection in the present study, neither in luminal nor in mucosa-associated microbiota. Still, C. jejuni -infected birds showed slightly higher richness and evenness. These results are consistent with previous investigations. Thibodeau et al. [28] observed no impact of C. jejuni colonization on alpha diversity in caecal microbiota of Ross 308 chickens (means of Chao1 index of C. jejuni -negative birds: 644 and C. jejuni -positive birds: 663; means of Shannon index of C. jejuni -negative birds: 3.8 and C. jejuni -positive birds: 3.7). In contrast, Awad et al. [27] noticed a significantly higher species richness in caecal contents of infected Ross 308 chickens 14 days after infection [Sobs, Chao1, and ACE (P=0.047)]. The authors concluded from the observed higher diversity, an indication that the Campylobacter infection increased the microbiota complexity. One further study investigated the microbiota of chickens at an industrial farm environment with the aim to investigate Campylobacter appearance within a natural habitat setting [29]. McKenna et al. [29] also recognized that the presence of Campylobacter was linked with an increased microbial diversity (P <0.05). At this point, however, the question arises whether with the introduction of Campylobacter onto the farm, other bacteria were introduced as well, that may have brought about this observation on microbiota diversity. In the present study, the risk of bacterial introduction from outside the infection unit was reduced to a minimum, which is why the results of the present study can be seen under the sole factor of a Campylobacter infection.

Another difference between the studies could reveal an additional factor that should be taken into account in studies on Campylobacter infections. In the present study and the study conducted by Thibodeau et al. [28], 25 and 21 days passed between experimental Campylobacter infection and intestinal microbiota analyses, respectively, in contrast to Awad et al. [27], who investigated the microbiota already 14 days after infection. Additionally, Awad et al. [27] investigated the intestinal microbiota at a younger chicken age (28 days) compared to the present study (42 days) and Thibodeau et al. [28], who collected the samples at an age of 35 days. Taking these observations and the results of the present study into account, it seems that age at the time of examination and the time passed after Campylobacter infection are two important factors that should be considered when investigating intestinal microbiota and its impact on Campylobacter and vice versa. Similar to the comments on microbiota composition, the impact of Campylobacter on microbiota diversity might vary according to the age of the chickens and the time passed after infection, being less prominent for older chickens and the time after infection.

Mucosal microbiota of C. jejuni -infected birds in the present study showed a higher relative abundance of the phylum Firmicutes and a lower relative abundance of Proteobacteria . Even if the differences in these phyla between non-infected and infected birds were not significant, similar observations were previously made in other studies for luminal-associated microbiota of the caecum. Awad et al. [27] observed similar shifts in the two major phyla in caecal contents towards an enrichment of Firmicutes (relative abundance in control birds: 70.86 % and infected birds: 97.65 %; P=0.031) with a concomitant reduction in Proteobacteria (relative abundance in control birds: 4.42 % and infected birds: 0.71 %; P=0.029) due to Campylobacter colonization at day 28 of life. Besides these major shifts, Awad et al. [27] also observed an alteration in low abundant phyla (e.g. Actinobacteria and Tenericutes ) by Campylobacter infection. The authors assume that these shifts could also disequilibrate the microbiome composition. In the current study, significant differences due to C. jejuni infection were only seen in caecal mucus for the low abundant phyla Bacteroidetes and Verrucomicrobia .

Enrichment of bacteria within caecal microbiota of chickens due to Campylobacter infection

The genus Enterococcus seemed to be enriched due to C. jejuni infection. In investigations by Pandit et al. [30], a repeated positive correlation was found between Campylobacter and Enterococcus , well in line with the present study. In Kaakoush et al. [31] Enterococcus was associated with the presence of C. jejuni in the chicken gastrointestinal tract as well. However, in contrast to the present study, also a decreased relative abundance of Enterococcus with C. jejuni presence in caecal contents was observed [27, 31, 32]. Within the order Lactobacillales and the family Enterococcaceae , the genus Enterococcus is known for its fermentative metabolism with the predominant end product of glucose: l-lactate [33]. A variety of bacterial species have been used as probiotics in poultry including Enterococcus [34]. On the other hand, infections with pathogen Enterococcus cecorum strains form an important emerging disease in modern broiler chicken lines associated with arthritis and osteomyelitis, leading to high mortality rates [35–37]. There are indications that the presence of certain bacterial species of the microbiota may encourage the presence of Enterococci by providing nutrients. Vancomycin-resistant Enterococci are unable to ferment mucin-derived complex polysaccharides; their growth was supported after a pre-digestion of mucins with enzyme mixtures obtained from human faeces, which resulted in a release of monosaccharides [38]. Some intestinal mucolytic bacteria use their specific enzymatic activities to degrade mucins and release monosaccharides attached to the mucin glycoproteins that can be used by other resident bacteria [39]. It has been shown that C. jejuni upregulates putative mucin-degrading enzymes in the presence of mucins [40]. For this reason, it can be suspected that in consequence of the presence of C. jejuni, mucins were degraded and monosaccharides released, which may have been used by Enterococcus . In any case, what can be assumed from the results of the present study is that environmental changes due to the presence of C. jejuni promoted the growth of Enterococcus.

In addition, the Clostridiales vadin BB60 group seem to benefit from a Campylobacter infection as its abundance was enriched in caecal contents of C. jejuni -infected compared to non-infected birds of the present study. This genus is known to be a later colonizer of the chickens caecum, however this genus is poorly classified and little is known about its metabolism or role within microbiota [41].

A positive association between Faecalibacterium and Campylobacter, as seen in the present study, was observed before in other studies with C. jejuni -infected chickens [28, 31]. The abundance of Faecalibacterium increased within the caecum of chickens that were colonized by C. jejuni [28]. Faecalibacterium reported in chickens, its 16S rRNA sequence is only 96–97% similar to 16S rRNA of Faecalibacterium prausnitzii from humans, however, both human and chicken Faecalibacterium isolates are efficient butyrate producers [42]. This ability of Faecalibacterium to produce butyrate appears to be in contradiction with its positive association with C. jejuni [28] as it was reported to be detrimental to C. jejuni [43, 44]. The mutualistic relationship between these two bacterial genera ( Faecalibacterium and Campylobacter ), seems to occur by means of an intermediate microbial conglomerate ( Limnobacter , Parabacteroides , Pseudomonadaceae , Sutterella , Sphingobium and Oxalobacteraceae ) that appears to modulate their mutual interactions [45]. The authors discuss that their identified network topology suggests the possibility of a commensalistic relationship between Faecalibacterium and this intermediate microbial conglomerate; conversely, an amensalistic relationship could be in place between the intermediate microbial conglomerate and Campylobacter [45]. Nevertheless, the reverse relation between both genera was found as well [46].

Negative association between bacteria and the genus Campylobacter or in the context of the Campylobacter infection

The existence of members belonging to the family Peptococcaceae could not be demonstrated in C. jejuni -infected birds in the present study. Peptococcaceae colonize the caecum of chickens in the third week of life [41, 47] and can be found in caecal digesta as well as mucosa samples [48], being more abundant in the lumen [10]. Peptococcaceae found in human and animal intestines are obligate anaerobes with a fermentative metabolism, producing caproic acid as a terminal metabolite of glucose [49]. To our knowledge, a connection between the presence or absence of the family Peptococcaceae and an infection with C. jejuni has not yet been described.

Various genera of the family Ruminococcaceae seemed to be depressed under C. jejuni infection in the present study. Additionally, a strong negative relation was found between two genera of this family and the genus Campylobacter . Other studies also suggest that the co-occurrence in high levels of certain genera of this family and C. jejuni does not seem to be compatible. Only 16 weeks after oral infection with C. jejuni , the Ruminococcaceae family exhibited an increased relative abundance that plateaued and thus was stably present within the caecal microbiota [50]. This occurred at the same time when most laying hens became negative and remained negative for C. jejuni [50]. The role of bacterial members of the family Ruminococcaceae within the intestinal microbiota of chickens was reviewed by Rychlik [42]. Ruminococcaceae represent major butyrate producers of which the vegetative cells are highly sensitive to oxygen, and are therefore among the first bacteria to disappear from gut microbiota during inflammatory diseases due to the production of reactive oxygen species by macrophages and granulocytes [42]. This means that in most cases, the decrease of Ruminococcaceae is not the cause of the inflammation but its consequence [42].

Furthermore, a strong negative correlation between the genus Campylobacter and Christensenellaceae R-7 group of the family Christensenellaceae was seen in the present study. The name Christensenellaceae is derived from the isolate named Christensenella minuta , a saccharolytic bacterium with acetic acid and a small amount of butyric acid as end products of glucose fermentation [51]. Christensenellaceae form a family of bacteria within the phylum Firmicutes , class Clostridia , order Clostridiales [51]. In 2015, Thibodeau et al. [28] observed a relationship between an unclassified genus of the family Christensenellaceae and C. jejuni colonization of chickens. This unknown genus as well as the family Christensenellaceae decreased in relative abundance in birds colonized by C. jejuni , but its involvement in chicken health was still unknown [28]. In the meantime, Christensenellaceae have emerged as an important player in human health [52], and more knowledge has been gained about this family within the microbiota of chickens as well. Christensenellaceae are later colonizers of the caecal microbiota, being most apparent from day 14 post-hatch [41], while the chicken’s breed has been shown to have an impact on its abundance [41, 53, 54]. Even if Richards et al. [41] observed a higher abundance of Christensenellaceae in the chicken’s mucus, the present study revealed a higher relative abundance of Christensenellaceae in luminal compared to mucosa-associated microbiota, this family being the fourth most common one in caecal contents of chickens (after Ruminococcaceae , Lachnospiraceae and Erysipelotrichaceae ). Recently, an association between Christensenellaceae R-7 group and another enteric pathogen was found in chickens [55]. The family Christensenellaceae and its genus Christensenellaceae R-7 group was more abundant in low compared to high Salmonella enterica Enteritidis-carrying chickens and found to be higher in a White Leghorn inbred line, which are more resistant to Salmonella [55]. In contrast to Thibodeau et al. [28], we observed an enrichment of bacterial sequences of the genus Christensenellaceae R-7 group in C. jejuni -infected compared to non-infected birds. Even if the differences were significant, the log₂-fold changes were small (with one exception <2). In addition, the relative abundance of the genus Christensenellaceae R-7 group was significantly negatively related to the relative abundance of Campylobacter in C. jejuni -infected chickens in luminal and mucosa-associated microbiota. It can be hypothesized that members of both families compete for substrate. Amino acids are used by C. jejuni as carbon and energy sources (reviewed in [56]) or directly for protein synthesis [57], while Christensenellaceae have recently been positively associated with gut metabolites typical of amino acid degradation [58].

Conclusion

In the present study, exclusive interactions between caecal luminal and mucosa-associated microbiota and C. jejuni could be assessed in an infection model with high explanatory power that was achieved by the applied biosecurity measures and assurance of identical conditions between both groups that were only separated by a continuous wall. At the same time, chickens were obtained from a commercial hatchery, feeding was adapted to usual field conditions, common bedding material was used and different C. jejuni prevalences were generated via the boxes in order to simulate practical conditions. Based on the obtained data, it can be concluded that C. jejuni infection contributed significantly to the differences in microbial composition, while the contribution to mucosa-associated bacteria was greater compared to its influence on luminal microbiota. The negative correlation found between the genus Christensenellaceae R-7 group and Campylobacter in C. jejuni -colonized chickens makes this family highly interesting for further studies that aim to find control options for Campylobacter infections in chickens.