Impact of Campylobacter spp. on the Integrity of the Porcine Gut

Simple Summary Campylobacteriosis is recognised as a leading food-borne zoonotic enteric disease of humans, mainly caused by Campylobacter jejuni, and to a minor extent by C. coli. In general, pigs are symptomless carriers primarily of C. coli, but may also harbour C. jejuni. In a swine infection model, weaned piglets were intragastrically inoculated with C. coli (ST-5777; n = 8), with C. jejuni (ST-122; n = 10), or with both strains (n = 8) and 11 piglets served as control. The health status was monitored and the influence on the intestinal barrier was investigated using the Ussing chamber technique and histological examinations. After inoculation, no clinical manifestations were noted. No gross lesions were observed during dissection four weeks post inoculation, and no pathohistological changes were detected in the intestinal mucosal sections. On the other hand, in the caecum of C. jejuni mono-inoculated pigs, we recognized an impact on transepithelial transport processes. We observed an increased Cl− secretion by using the Ussing chamber technique. Abstract Campylobacter (C.) is the most common food-borne zoonosis in humans, which mainly manifests with watery to bloody diarrhoea. While C. jejuni is responsible for most cases of infection, C. coli is less frequently encountered. The object of the study was to prove the clinical impact of mono- and co-colonisation of C. coli and C. jejuni on weaned piglets in an infection model and to investigate the impact on transepithelial transport processes in the jejunum and caecum. At an age of eight weeks, eight pigs were infected with C. coli (ST-5777), 10 pigs with C. jejuni (ST-122), eight pigs with both strains, and 11 piglets served as control. During the four-week observation period, no clinical signs were observed. During dissection, both strains could be isolated from the jejunum and the caecum, but no alteration of the tissue could be determined histopathologically. Mono-infection with C. jejuni showed an impact on transepithelial ion transport processes of the caecum. An increase in the short circuit current (Isc) was observed in the Ussing chamber resulting from carbachol- and forskolin-mediated Cl− secretion. Therefore, we speculate that caecal colonisation of C. jejuni might affect the transport mechanisms of the intestinal mucosa without detectable inflammatory reaction.


Introduction
Campylobacter (C.) has replaced Salmonella as the most common human diarroheal pathogen in all developing and non-developing countries [1,2]. In 2019, 61,526 cases

Campylobacter Strains and Inoculum Preparation
Two different Campylobacter strains were used for this experiment: C. coli: ST5777/CT828 and C. jejuni: ST122/CT206. Both strains were isolated from poultry and had been used in a previous study [28]. The genotypes, characterised by multilocus sequence typing (MLST), had previously been isolated several times in association with human disease (http://pubmlst.org/campylobacter/, accessed on 18 May 2021). The C. coli strain was isolated from a sporadic case of gastroenteritis in humans in Luxembourg in 2012. The C. jejuni strain was isolated from faeces of a human case associated with Guillain-Barré syndrome. It has also been found in numerous cases in the United States, the Netherlands, and Germany in association with gastroenteritis in humans. Most recently, the strain was isolated in one case of systemic disease in the United Kingdom.
Different laboratory-induced antibiotic resistance patterns of the strains enabled differentiation when grown on agar supplemented with the corresponding antibiotic substance. The C. coli ST-5777 strain is resistant against nalidixic acid, and C. jejuni ST-122 against streptomycin.

Animals and Experimental Set-Up
In total, 37 (21 male, 16 female) crossbred piglets ((Danish Landrace × German Large White) × German Landrace) were used for the experiment. They were derived by Caesarean section to exclude a Campylobacter field colonisation and raised in isolation units. The pigs were housed and treated in accordance with the German Animal Welfare Act which complies with the German Research Council's criteria and the EC Directive 2010/63/EU for animal experiments. The absence of Campylobacter spp. was ensured by regular rectal swabs. Strict hygiene standards were established to avoid transmission of Campylobacter between units after infection.
CulinaMilk ® (H. Bröring GmbH & Co. KG, Dinklage, Germany) was used as milk replacer. After weaning in the fifth week of life, the pigs were fed twice daily with a conventional rearing diet (Deuka GmbH & Co. KG, Düsseldorf, Germany) and had access to drinking water ad libitum. The piglets were injected with 2 mL iron dextrane (Ursoferran ® 100 mg/mL, Serumwerk Bernburg AG, Bernburg, Germany) subcutaneously (s.c.) on their third day of life.
The piglets were randomly divided into four different groups (Table 1) and inoculated with the respective Campylobacter strain in week eight. Hereafter, the control group is referred to as Group 0, the group infected with C. coli is called Group 1, and the group infected with C. jejuni is called Group 2. In addition, the group infected with both strains is called Group 3. Table 1. Infection scheme of the animal groups.
During the experiment, the pigs underwent daily monitoring including evaluation of the general health condition and rectal measurement of body temperature in the morning. If the rectal temperature rose above 40 • C, the temperature was rechecked after 12 h. If it was still elevated thereafter, the piglets were treated orally with Metacam ® (Meloxicam, 15 mg/mL; 0.4 mg/kg body weight, Boehringer Ingelheim Vetmedica GmbH, Ingelheim, Germany).
At necropsy four weeks after inoculation, the animals had an average body weight of 30.9 ± 5.6 kg.

Sampling and Cultivation of Campylobacter Strains from Faeces and Ingesta
During the experimental period, rectal swabs were collected both before and after infection to demonstrate the absence of Campylobacter spp. and to confirm intestinal colonisation after infection, respectively.
All animals were stunned by captive bolt shot and killed by exsanguination in the 13th week of life (four weeks post-inoculation with Campylobacter spp.). Immediately thereafter, Animals 2021, 11, 2742 4 of 14 the abdomen was opened by a midline incision and the intestinal tract was removed. During dissection, the intestinal tract was examined macroscopically for pathological abnormalities.
The first three metres of jejunum were discarded, and the following 30 cm were removed. Additionally, the middle part of the caecum was excised for Ussing chamber experiments.
Previous experiments showed that Campylobacter spp. colonise mainly in these two intestinal sections [28]. Therefore, the experiments were carried out on the jejunum and caecum using the Ussing chamber.
The intestinal segments were rinsed with cold physiological saline solution (4 • C, 0.9% NaCl (w/v)) to remove ingesta and afterwards stored in cooled serosal buffer solution ( Table 2) for Ussing chamber experiments. After longitudinal incision of the jejunum along the mesenteric border, the mucosa was stripped of the underlying muscle layers. The tissue from the caecum was also opened and the mucosa separated. The specimens were then clamped in the Ussing chambers. For the histological examinations, samples were taken from the duodenum, jejunum, ileum, caecum and colon. The sampling sites of the jejunum and caecum were adjacent to the samples taken for the Ussing chamber. Intestinal samples were rinsed with physiological saline solution and placed in 4% (w/v) phosphate buffered formalin for at least 48 h.
For the microbiological examination, faecal and intestinal samples (intestinal mucosa + intestinal contents) were placed in a tube filled with 10 mL Preston broth for enrichment and incubated for 48 h at 37.4 • C in a micro-aerophilic atmosphere (5% O 2 , 10% CO 2 and 85% N 2 ). After this time, the enrichment was spread onto modified charcoal cefoperazonedeoxycholate agar plates (mCCDA, Thermo Scientific Inc., Waltham, MA, USA) and incubated again for 48 h as described above (the procedure can be seen in Figure S1). The CCDA agar was enriched with the corresponding antibiotic (C. coli = nalidixic acid; C. jejuni = streptomycin; detailed ingredients can be found in Tables S1-S3). Subsequently, the presence of Campylobacter-specific colonies was examined, and a native smear was performed if the results were questionable. In addition, random samples from Groups 1 and 2 were spread on the CCDA agar enriched with the other antibiotic in order to exclude cross-contamination of the mono-infection groups.

Set-Up for Ussing Chamber Technique
Stripped jejunal and caecal mucosal tissues were mounted in the Ussing chambers. Three chambers were used per intestinal segment and per animal (exposed serosal area of 1.00 cm 2 ). By mounting the tissues between the two halves of the Ussing chamber, a barrier was formed between the serosal and mucosal compartment, each of which were filled with 10 mL buffer. For the serosal side, the same buffer was used for the jejunum and caecum. On the mucosal side, different buffers were used for the two intestinal segments. The buffer composition is shown in Table 2.
All buffer solutions had an osmolarity of 297 mosmol kg −1 , were heated to 37 • C, and were continuously aerated with carbogen. The pH value ranged between 7.45 and 7.47.
All chemicals for the buffer solutions were obtained from Merck KGaA, Darmstadt, Germany, except mannitol, which was purchased from Sigma-Aldrich Inc., St. Louis, MO, USA, and HEPES, which was purchased from Carl Roth GmbH + Co.KG, Karlsruhe, Germany.
After filling the chambers with buffer solutions, they were connected to computercontrolled voltage clamps (K. Mussler Scientific Instruments, Aachen, Germany) through which the transepithelial potential difference (PD t ), tissue conductance (G t ), and short circuit current (I sc ) were measured. The experiment was performed under short-circuited conditions.
Firstly, the tissues were equilibrated for 30 min after attaching the voltage clamps and filling in the buffers. After the 30-min equilibration phase, the voltage was clamped to zero. The further experiment was performed in accordance with two different protocols, respectively for the tissue from the jejunum and caecum.
According to the first protocol, the mucosal glucose concentration of the jejunal tissues was adjusted to 10 mmol/L after equilibration. In order to avoid transepithelial osmotic gradients, this was compensated by adding mannitol to the serosal side. After an incubation period of 30 min, 5 × 10 −6 M forskolin (1 µL, serosal) was added.
The second protocol was applied to the caecal tissues: 10 5 M carbachol (10 µL, serosal) was added, incubated for another 30 min and then 5 × 10 −6 M forskolin (1 µL, serosal) was added. Forskolin and carbachol were obtained from Sigma-Aldrich Inc., St. Louis, MO, USA. After adding forskolin, another 30 min incubation period was performed before the experiment was terminated.

Histological Investigation
Tissue samples from duodenum, jejunum, ileum, caecum, and colon were stored in formalin solution (100 mL formaldehyde 37%, 900 mL bidest, 4 g NaH 2 PO 4 H 2 O, 6.5 g Na 2 HPO 4 ) for at least 48 h. Post-fixation, tissue drainage and embedding of the fixed tissue with paraffin was performed in the Shandon Pathcentre ® Tissue Processor (Thermo Scientific Inc., Waltham, MA, USA). The samples were then poured into paraffin wax blocks at a paraffin pouring station (TES 4004 Tissue Embedding Station, PFM, Cologne, Germany). Tissue sections with a thickness of 2 µm were prepared from the paraffin wax blocks on a slide microtome (2646, Reichert & Jung GmbH, Heidelberg, Germany). The sections were placed in a water bath at 40 • C, applied to Menzel Superfrost ® Ultra Plus slides (Thermo Scientific Inc.) and dried for at least 24 h at room temperature.
The sections were dewaxed, dehydrated and stained automatically with haematoxylinand eosin (H&E) (Varistan 24-2, Shandon GmbH, Frankfurt, Germany) in accordance with a protocol (Table 3). Subsequently, the sections were covered with ROTI ® Mount and coverglass, dried for at least 24 h and examined microscopically.
For the histological specimens, a scoring key of 1-5 was used for semiquantitative analysis of the inflammatory infiltrate for microscopic examination. Specimens were graded according to the degree of neutrophilic and increased lymphoplasmacytic cell infiltration, and possible differences between groups were evaluated (1 = slight infiltration, 2 = mild infiltration, 3 = moderate infiltration, 4 = marked infiltration, 5 = severe infiltration).

Dry Matter Determination of Ingesta (Caecum)
To determine the dry matter (DM) of ingesta from the caecum, approximately 1 g of the sample (E) was weighed into a metal crucible (T1) and then dried at 103 • C for a period of at least four hours in a drying oven.
The samples were allowed to cool in a desiccator before the porcelain crucibles were weighed again (T2). Ten samples were treated per group.
The following formula was used to determine the dry matter content.

Data Analysis and Statistics
Results from the Ussing chamber experiment were expressed as the mean value ± standard deviation (MW ± SD) of the single values of the three chambers used per bowel segment. For the calculation of ∆I sc , the maximum changes after adding the respective substance were used. The statistical analysis for the results of the Ussing chamber, histological samples and dry matter determination was performed with a one-way ANOVA and the post-hoc Dunnetts t-test using SAS v. 9.1 (SAS Inst. Inc., Cary, NC, USA).
The differences were assumed to be statistically significant if p ≤ 0.05.

Clinical and Microbiological Findings during the Test Period and Macroscopic Findings in the Section
No abnormalities in the general condition of the animals were observed during the experimental period. Diarrhoea or increased internal body temperature did not occur (Table S4). Faecal samples of the control group (Group 0) remained Campylobacter negative throughout the trial. In the infected groups, the respective strains could be detected in the faeces as early as two days after infection (p. inf.) in some cases, but reliably seven days after inoculation. In Group 2, the detection rate decreased slightly again after day 7 (Table 4). Cross infection with the other Campylobacter strain was not detected in Groups 1 and 2.
On dissection, there were no macroscopic changes to the intestinal tract, liver, or spleen.

Distribution of Campylobacter spp. in Different Intestinal Segments
Cultural investigation of the gut demonstrated different detection frequencies of the Campylobacter strains in the intestinal sections (Table 5).  C. coli as well as C. jejuni could be reisolated from all samples from the caecum and colon. This also applied for mono-and co-infection (Groups 1-3).
Only C. jejuni was found in the duodenum. In Group 2, 40% (n = 4) of the samples were positive, while in Group 3, 62.5% (n = 5) were positive. C. coli was not detected in the duodenum in either mono-or co-infection.
C. jejuni could be isolated from the jejunum samples for Group 3 in all pigs. However, only 12.5% of the samples showed C. coli (n = 1). Moreover, in the mono-infected pigs, the detection of C. jejuni in the jejunum was higher (Group 2; 70%; n = 7) than that of C. coli (Group 1; 25%; n = 2).
C. jejuni was detected in the ileum of all animals in Groups 2 and 3. In contrast, the identification of C. coli showed a difference between mono-and co-infection. While C. coli was detected in the ileum of all pigs in Group 1, only 12.5% of the animals in Group 3 were positive (n = 1).
For Group 0, the Campylobacter negative status was confirmed by the negative results of the microbiological investigation of all intestinal sections.

Basal, Glucose-Induced and Forskolin-Induced Short-Circuit Currents (I sc ) in Jejunal Epithelia
After equilibration, the basal values for I sc were at a similar level in all four groups (Figure 1a). An increase in I sc could be measured after adding glucose and forskolin in all groups (Figure 1a) without a statistically significant difference between them (p > 0.05). The course of short-circuit currents and tissue conductance during the experimental period in the jejunum in an animal of the control group (Group 0) is exemplarily shown in Figure 2. The time course was similar for animals in the infected groups (all values can be found in Table S5). The course of short-circuit currents and tissue conductance during the experimental period in the jejunum in an animal of the control group (Group 0) is exemplarily shown in Figure 2. The time course was similar for animals in the infected groups (all values can be found in Table S5).

Basal, Carbachol-Induced and Forskolin-Induced Short-Circuit Currents (Isc) in Caecal Epithelia
Between Group 1, Group 3, and the control group (Group 0), no statistical differences in the Isc of the caecal epithelia could be demonstrated. The basal caecal Isc in Group 2, mono-infected with C. jejuni, was statistically significantly higher (p < 0.05) than in Group The course of short-circuit currents and tissue conductance during the experimental period in the jejunum in an animal of the control group (Group 0) is exemplarily shown in Figure 2. The time course was similar for animals in the infected groups (all values can be found in Table S5).

Basal, Carbachol-Induced and Forskolin-Induced Short-Circuit Currents (Isc) in Caecal Epithelia
Between Group 1, Group 3, and the control group (Group 0), no statistical differences in the Isc of the caecal epithelia could be demonstrated. The basal caecal Isc in Group 2, mono-infected with C. jejuni, was statistically significantly higher (p < 0.05) than in Group

Basal, Carbachol-Induced and Forskolin-Induced Short-Circuit Currents (I sc ) in Caecal Epithelia
Between Group 1, Group 3, and the control group (Group 0), no statistical differences in the I sc of the caecal epithelia could be demonstrated. The basal caecal I sc in Group 2, mono-infected with C. jejuni, was statistically significantly higher (p < 0.05) than in Group 0. After adding glucose and forskolin, the I sc in Group 2 continued to differ statistically significantly from Group 0 (Figures 1b and 3; detailed data in Table S6). 0. After adding glucose and forskolin, the Isc in Group 2 continued to differ statistically significantly from Group 0 (Figures 1b and 3; detailed data in Table S6).

Conductance (Gt) of Jejunal and Caecal Epithelia
The basal Gt values in jejunal and caecal samples did not differ statistically significantly between groups. The changes after adding glucose to the jejunal tissue and carbachol to the caecal tissue were also not statistically significantly different (p > 0.05). Accordingly, adding forskolin 30 min later did not result in statistically significant changes in the Gt values (p > 0.05). A time course of the changes in the Gt-value in caecal tissue is shown in Figure 3 (detailed data in Tables S7 and S8).

Results of the Histological Examination
Almost all tissue samples of the intestinal sections (duodenum, jejunum, ileum, caecum, and colon) of all pigs were given the scores 1 or 2 in the applied scoring system, corresponding to a slight to mild neutrophilic and increased lymphoplasmacytic cell infiltration. Only two samples from the duodenum (both from Group 2) were found to have a score of 3. No statistically significant differences between the infected groups (Groups 1, 2, and 3) and the control group (Group 0) were detected (p > 0.05).

Conductance (G t ) of Jejunal and Caecal Epithelia
The basal G t values in jejunal and caecal samples did not differ statistically significantly between groups. The changes after adding glucose to the jejunal tissue and carbachol to the caecal tissue were also not statistically significantly different (p > 0.05). Accordingly, adding forskolin 30 min later did not result in statistically significant changes in the G t values (p > 0.05). A time course of the changes in the G t -value in caecal tissue is shown in Figure 3 (detailed data in Tables S7 and S8).

Results of the Histological Examination
Almost all tissue samples of the intestinal sections (duodenum, jejunum, ileum, caecum, and colon) of all pigs were given the scores 1 or 2 in the applied scoring system, corresponding to a slight to mild neutrophilic and increased lymphoplasmacytic cell infiltration. Only two samples from the duodenum (both from Group 2) were found to have a score of 3. No statistically significant differences between the infected groups (Groups 1, 2, and 3) and the control group (Group 0) were detected (p > 0.05).

Dry Matter of the Caecal Content
Due to the differences between Group 2 and Group 0 in the tests using the Ussing chamber, a dry substance determination from the caecal content was carried out. Samples from each group were averaged and compared (Group 0: 143.52 ± 17.68; Group 1: 135.86 ± 37.28; Group 2: 108.87 ± 17.23; Group 3: 139.29 ± 28.33). The dry matter content of the caecum from Group 2 was statistically significantly lower than that from the control group (p < 0.05).

Discussion
The aim of this study was to evaluate whether colonisation of weaning pigs with different Campylobacter strains has an impact on intestinal health and mucosal barrier function.
Prerequisite for the Campylobacter infection model was the rearing of the piglets without infection with Campylobacter field strains, which was confirmed by the absence of Campylobacter spp. before inoculation and in the control group. In experiments with gnotobiotically reared piglets [9][10][11], a method for maintaining absence from Campylobacterstrains for days up to a few weeks was established in a highly sterile environment. In the present study, it was shown that Campylobacter-free rearing over a period of 12 weeks after Cesarean section is possible under strict hygiene standards without sterile conditions.
In previous studies, it was observed that a low infectious dose of 800 cfu was sufficient to cause disease in 50% of the subjects in humans. At a dose of 10 8 , the disease rate was 100% [29,30].
An infectious dose of at least 10 8 cfu per animal was obtained. In previous studies, this infectious dose has been shown to be sufficient to achieve intestinal colonisation in pigs as safely as possible [28,31].
Although the piglets were assumed to have no maternal immunity against Campylobacter, comparable to colostrum-deprived or gnotobiotic piglets, none of them developed clinical signs throughout the experiment, but were successfully colonised. Furthermore, during post-mortem examination, internal organs did not show any gross lesions.
These observations are consistent with those of previous studies [28,31,32]. For the most part, Campylobacter spp. are considered commensals of the swine intestine [4]. Thus, as previously mentioned, it could be demonstrated, that Campylobacter spp. have no clinical effects on pigs. Therefore, the importance of Campylobacter infections in diarrhoel diseases of pigs can be neglected. However, when colonising gnotobiotic piglets with C. jejuni on the day of birth, significant clinical abnormalities were observed [9], comparable to other Campylobacter infection models with very young pigs [10,11]. In contrast to the histopathological results of the presented study, other trials with gnotobiotic piglets showed distinct histological alterations of the intestinal epithelium. These changes were described as an increase in mucous secretory cells, signs of villi shortening and blunting. In addition, some villous cells showed damaged nuclei and some crypt abscesses were present in the caecum [9,10]. These changes were not found in the present study. Thus, it seems that the intestine of gnotobiologically reared animals is more affected by Campylobacter spp. colonisation than is the case in animals that have not been kept under sterile conditions. The Ussing chamber technique allows the measurement of the electrophysiological tissue parameters (conductance (G t )) and short circuit current (I sc ). In order to eliminate electrochemical gradients across the tissues, the experiments were performed under shortcircuited conditions, so that a change in short circuit current (I sc ) implies a change in net transepithelial ion transfer. Additionally, tissue conductance (G t ) provides an indication of epithelial integrity [33].
Changes in short circuit currents depend on the transepithelial ion transport. Adding glucose to the mucosal side of jejunal samples stimulates the electrogenic SGLT1 transport system, resulting in an increase in I sc [22,23].
For caecal tissues, the addition of glucose was replaced by carbachol, since glucose is almost exclusively absorbed in the small intestine [23,34]. Forskolin and carbachol stimulate Cl − secretion: forskolin activates cAMP-controlled secretion [24] whereas carba-chol acts on Ca-controlled secretion [25][26][27]. This ion release into the intestinal lumen is always passively followed by H 2 O. Thus, when the secretion of Cl − is increased, a greater amount of water is also released into the intestinal lumen. This manifests itself clinically as diarrhea [35].
It can be concluded that jejunal glucose transport by the secondary active co-transporter 1 (SGLT1 transport) [36,37] was not affected by Campylobacter colonisation. Similarly, cAMPmediated chloride secretion was not altered. The I sc increase from the jejunal epithelia was not statistically significantly different among all groups after adding forskolin (p > 0.05).
Microbiological investigation of ingesta showed that predominantly C. jejuni seems to colonise the jejunum, although C. coli was also detected, but to a lesser extent, which has also been demonstrated in a previous study [28]. Both agents did not affect the integrity of the studied intestinal barrier and the transport mechanisms.
Investigations on the caecal tissue showed that only Group 2 (mono-infected with C. jejuni) was found to differ statistically significantly from the control group in forskolinand carbachol-induced increases in I sc (p < 0.05) due to an increased permeability for ions (Cl − ). Forskolin induces cAMP-regulated chloride secretion, whereas carbachol acts on calcium-regulated chloride secretion. In both mechanisms, water passively follows chloride into the intestinal lumen [27], initially causing a more fluid ingesta. Dry matter determination of caecal content confirmed these findings, although clinically, diarrohea could not be detected. One potential explanation for this fact might be that a substantial amount of water is reabsorbed along the intact colon [34], which could compensate the water loss in the caecum. The fact that the colon is an intestinal segment with a high water resorption rate has already been extensively described [34]. The mechanism of the caecal epithelium alteration remains unclear.
In connection with this, it can be speculated, that an infection with C. jejuni could have a subclinical effect on pigs and that the loss of water in the caecum could also have an influence on nutrient intake. This can lead to losses in performance and weight development, which could well be decisive in conventional pig farming. To be more precise, longer-term studies should be carried out to study the weight development of the pigs in more detail. In earlier experiments by Leonhard-Marek et al. [38], pigs were infected with parasites. It was found that chloride secretion was increased during the penetration phase of the parasites into the intestinal mucosa. It is possible that C. jejuni has a similar effect on the intestinal barrier, resulting in an increase in chloride secretion.
Despite the differences observed in the Ussing chamber experiments, corresponding histological changes could not be found, leading to the assumption that colonisation of C. jejuni affects the transport mechanisms of the intestinal mucosa without any inflammatory reaction. Van Deun et al. (2008) demonstrated in an in-vitro experiment that C. jejuni strains could invade chicken primary caecal epithelial crypt cells without showing any signs of necrosis or intestinal inflammation [39]. Transmigration might be possible by the cleavage of occludin, a tight junction protein impairing intestinal barrier function, which is assumed by Harrer et al. [40]. Further studies are needed to investigate the mechanism of Campylobacter spp. concerning disturbance of the intestinal barrier in swine. In addition, further studies should clarify whether the subclinical effect that the C. jejuni strain used in the present study has on pigs can be transferred to other C. jejuni strains or only applies to the one used here.
No significant increase in I sc was found either in mono-infected pigs with C. coli or in co-infected animals with C. coli and C. jejuni. A previous study by Leblanc et al. [32] proved that C. coli has a higher colonisation potential than C. jejuni in pigs. This could result in C. coli overriding C. jejuni colonisation in co-infection to the extent that a smaller amount of C. jejuni colonising the caecum did not have any impact on the intestinal transepithelial chloride transport. Enumeration of Campylobacter spp. in the ingesta may be able to substantiate this assumption.

Conclusions
The results of the presented study demonstrate that the inoculated C. coli and C. jejuni strains of poultry origin are capable of colonising the intestine of weaned pigs and that both species integrated well after co-infection. The pigs did not show any clinical signs of disease after inoculation, although investigations using the Ussing chamber demonstrated an impact of C. jejuni on the caecal epithelium and the chloride transport mechanism. However, this effect appears to be compensated by the intact colon of the piglets so that no clinical disease occurs. Additionally, this study showed that the raising of Campylobacter free pigs is possible under high hygienic standards and that this pig colonisation trial seems to be suitable for studying Campylobacter infections without the need for gnotobiotic or specific-pathogen-free (SPF) animals.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/ani11092742/s1, Figure S1: Microbiological procedure for the detection of Campylobacter spp. in the fecal samples of the test animals on modified charcoal cefoperazonedeoxycholate agar plates (CCDA), Table S1: Ingredients of the nalidixic acid stock solution (1%), Table S2: Ingredients of the streptomycin stock solution (1%), Table S3: Quantity ratio for the preparation of the specific antibiotics-CCDA agar with the respective stock solution (nalidixic acid or streptomycin sulphate), Table S4: Body temperature of the infection groups over the experimental period as mean +/-standard deviation (Group 0: n = 11; Group 1: n = 8; Group 2: n = 10; Group 3: n = 8), Table S5: ∆I sc values for jejunal tissue, both basal values for the groups and the changes after addition of glucose and forskolin, the mean values and standard deviations for the respective groups (0-3) are shown (Group 0: n = 11; Group 1: n = 8; Group 2: n = 10; Group 3: n = 8), Table S6: ∆I sc values for caecal tissue, both basal values for the groups and the changes after addition of carbachol and forskolin, the mean values and standard deviations for the respective groups (0-3) are shown (Group 0: n = 11; Group 1: n = 8; Group 2: n = 10; Group 3: n = 8), Table S7: ∆G t values for jejunal epithelia, both basal values for the groups and the changes after addition of glucose and forskolin, the mean values and standard deviations for the respective groups (0-3) are shown (Group 0: n = 11; Group 1: n = 8; Group 2: n = 10; Group 3: n = 8), Table S8

Institutional Review Board Statement:
The study was conducted in accordance with the guidelines of the Declaration of Helsinki. Ethical authorization was granted following notification to the national veterinary authorities (33.9-42502-04-16/2132).

Data Availability Statement:
The data presented in this study are available in Supplementary Material. Additional data are available on request from the corresponding author.