Focal Duodenal Necrosis Replication in Layer Chickens Using a Bacterial Mixture in Challenge Experiments Open Access

Focal duodenal necrosis (FDN) is an underresearched and underdiagnosed disease in layer chickens that causes egg production drops and decreased egg weight (1). Egg production may fall short of the anticipated peak by 2%–3%, or producers may experience a decline ranging from 1% to 10% (2). Additionally, egg weights tend to fall behind the expected range by approximately 0.6–1.3 g per egg in young flocks (3). There are no characteristic clinical signs in affected chickens, although some nonspecific signs such as lower body weight and pale comb have been reported (4). Although FDN has been found in all major genetic lines of laying birds under various management systems, including cage, cage-free, and organic, organic farms tend to have a higher prevalence of FDN compared to nonorganic farms (5). Internationally, FDN has been reported to have a comparable flock-level prevalence in Europe, Canada, and the United States (6,7,8).

The lesion associated with FDN is characteristic and easy to identify at necropsy. It consists of single or multiple 3–15-mm-diameter, irregularly shaped patches of grayish ulceration (described as cigarette-burn-like) in the mucosa of the duodenal loop and can sometimes extend to the first section of the jejunum (9). Histopathological changes include necrosis and loss of enterocytes from the villous tips, along with infiltration of the lumen by fibrinoheterophilic inflammatory cells mixed with clusters of long, rod-shaped bacteria, sloughed degenerated cells, and red blood cells (1). Additionally, large amounts of both Gram-positive and Gram-negative rod-shaped bacteria may also be found in the exudate or attached to the villi (1). It has often been described as a “volcano-like lesion” due to the loss of enterocytes from villous tips accompanied by fibrinoheterophilic exudate (M. Franca, pers. comm).

The etiology of FDN is undetermined but has been linked to Clostridia species in previous studies, especially Clostridium perfringens (CP) and Clostridium colinum (CC) (1, 4,10,11,). A study in 2016 reported isolation of CP from FDN lesions of layers (1). Additionally, the CP isolates harbored genes associated with alpha and beta2 toxins. As for CC, a nonculture molecular profiling study using PCR of the duodenal microbiota observed that FDN-affected groups exhibited a higher prevalence of CC (10). In 2017, a study attempted to reproduce FDN in challenge experiments using necrotic enteritis B–like toxin (netB)–positive and netB-negative CP isolates and duodenal homogenates obtained from FDN lesions. A questionnaire survey of FDN-affected farms identified potential predisposing factors (12). Most flocks received over 12 feed formulations from prelay to 60 wk; distiller’s dried grain with solubles (DDGS) was commonly included. Limestone was used universally for calcium supplementation (12). Therefore, the birds in the current study were fed a custom feed that included high protein, high DDGS, and additional ingredients that were considered to be FDN predisposing factors (12). The chickens challenged with CP and/or duodenal homogenates developed mild enteritis, but the study failed to reproduce the characteristic FDN lesions in layers. These results suggested that CP might not be the only factor linked with FDN, and there could be multiple causative agents, including other bacterial species, that contribute to the intestinal condition. Therefore, our research focused on elucidating the missing pieces, from a bacteriologic perspective, that may contribute to FDN.

Since then, our research group has sought to identify other potential bacterial agents that contribute to the etiology of FDN. Through analysis of the 16S ribosomal (rRNA) patterns of microbiota associated with FDN lesions using laser capture microdissection, we found differences in the duodenal microbiota between layers with FDN and healthy birds suggesting that FDN-affected birds show an intestinal dysbiosis pattern (13). Fresh FDN lesions were also collected for bacteriologic analysis, specifically focusing on Escherichia coli (EC) and other Gram-negative organisms present. PCR for EC virulence–associated genes found that 53.8% of the EC isolates were identified as avian pathogenic EC–like strains. Further PCR analysis for 19 EC virulence genes associated with inflammatory bowel disease (IBD) found that several isolates contained multiple IBD-related virulence genes as well. The EC strain included in this study possesses a total of 14 genes. The list of genes and PCR results can be found in Supplemental Table S1. Additionally, multiple Enterococcus spp. have been isolated (13). Among them, we selected an Enterococcus faecalis strain with the highest number of potential virulence determinants as identified by PCR assay (unpublished work; 14). Recently, from another collection of FDN samples, we used total bacterial sequencing directly from the FDN lesion itself to determine the organisms present. The duodenal tissue with the FDN lesion was trimmed to contain only the lesion, and total DNA extraction was performed. The 16S rRNA sequencing using PacBio technology and subsequent analysis found that CP, EC, and Tyzzerella/CC comprised 15% to 80% of the total bacterial composition in the sequenced FDN lesions (unpublished work). Gallibacterium anatis and Enterococcus spp. were detected with Enterococcus spp. also being isolated from fresh FDN lesion samples (13).

In this study, we hypothesized that CP, EC, CC, Enterococcus faecalis (EF), and Gallibacterium anatis (GA) are potential etiologic candidates for FDN. This study aimed to assess different bacterial mixtures with these bacterial species to determine which bacterial combination could cause the highest lesion scores and the development of lesions resembling FDN lesions seen in the field. Moreover, customized feed was used as a predisposing factor to facilitate the effect of the challenge experiment.

A CP strain positive for alpha and netB toxins and EC strain FDN-4 were obtained from field cases of FDN (15). The CC strain ATCC 27770 was obtained from the American Type Culture Collection. EF and GA were isolated from FDN cases between 2021 and 2023 (13). All isolates were confirmed for identification using either matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) or 16S rRNA sequencing.

The challenge strains were retrieved from −80 C frozen stocks and cultivated in specific culture media suitable for each bacterium. Thioglycolate broth (Oxoid; Remel, Lenexa, KS) was used as the growth medium for EC, CP, EF, and GA, and tryptone soy agar (TSA) plates with 5% sheep blood (Remel) were used for CC. Subsequently, all broth and agar plates were incubated at 37 C for at least 24 hr, either aerobically (for EC) or anaerobically (for the remaining isolates) using Mitsubishi AnaeroPak-Anero^TM gas packs and jars (Mitsubishi, Tokyo, Japan). The resulting inoculum was then measured using a spectrophotometer (with optical density of 600 nm) to assess the concentration of cells present as colony-forming units (CFUs) per milliliter. Each bacterial culture (20 ml) was centrifuged to precipitate the cells (8000 × g for 10 min at 4 C; Sorvall, ThermoFisher, Waltham, MA), followed by resuspension in 20 ml of phosphate-buffered saline (PBS; MP Biomedicals, Irvine, CA). Using the design described in the experimental design section below, the various cell suspensions were mixed in equal volumes within each experimental group. All cell suspensions were kept on ice until administration and prepared no more than 1 hr in advance. Prior to oral administration, the mixed suspensions were vigorously vortexed to ensure thorough mixing. Bacterial mixtures used in the study were made fresh daily.

Thirty laying hens (Hy-line W-36) were obtained from a local commercial flock at 35 wk of age and randomly assigned into one of five groups (six layers in each group, housed in cages with three chickens per cage). The treatment groups were as follows: T1: CP + EC, T2: CP + CC, T3: CP + EC + CC, and T4: CP + EC + CC + EF + GA and a control group inoculated with sterile PBS. The birds were provided with water and feed ad libitum and maintained under a standard lighting schedule (8 am to 10 pm) typically employed for commercial layers. All groups in this experiment received a custom mixed corn-soy diet containing 10% DDGS, 5% animal protein, and 20% fine and 80% coarse limestone particles as described previously (5).

After a 4-day acclimation period, the challenge phase started. The treatment groups were subjected to daily oral challenges using 1 ml of bacterial mixtures formulated at a concentration of 10⁸–10⁹ CFU/ml, and the control group received 1 ml of sterile PBS. The inoculum was delivered by oral gavage needle into the crop. The daily oral challenge continued for 14 days, after which the challenge was paused for a week to assess recovery. Throughout the experiment, the birds were monitored daily for both egg production and manifestation of clinical signs such as lethargy or a pale comb. At days 7, 14, and 21 in the challenge period, three birds from each group were euthanatized using American Veterinary Medical Association–approved methods and examined for any potential lesions present in the duodenum. Sample collection involved fixing a segment of the duodenum with lesions in formalin for histopathology, taking swabs from the mucosal surface, and streaking the samples onto MacConkey agar plates (MAC; Becton and Dickinson, Franklin Lakes, NJ) and prereduced TSA with 5% sheep blood plates (Remel). Aerobic and anaerobic cultures were incubated at 37 C overnight to detect the presence of the challenge pathogens. All birds used in this study was approved by the University of Georgia’s Institutional Animal Care and Use Committee under permit no. A2023 04-014-Y1-A0.

Sixty-six laying hens (Hy-line W-36) at 26 wk of age were randomly assigned into one of five groups as follows: treatment groups T1: CP (n = 18), T2: EC + CC (n = 18), and T3: CP + EC + CC (n = 18) and control groups Control 1: sterile PBS + customized feed (n = 6), and Control 2: sterile PBS + commercial feed (n = 6). The birds were provided with water and feed ad libitum and a standard lighting schedule (8 am to 10 pm) typically employed for commercial layers. Birds were reared in cages, with three chickens per cage. All treatment groups and Control 1 received a corn-soy diet containing 10% DDGS, 5% animal protein, and 20% fine and 80% coarse limestone particles. Control 2 received a commercial layer feed (Kalmbach, Sandusky, OH). After a 6-day acclimation period, the challenge phase started. The treatment groups were subjected to daily oral challenges using 1 ml of bacterial mixtures formulated at a concentration of 10⁸–10⁹ CFU/ml, and control groups received 1 ml of PBS. This daily oral challenge continued for 21 days. Throughout experiment, the birds underwent daily monitoring for both egg production and the manifestation of clinical signs such as lethargy or a pale comb. At days 4, 8, 11, 15, 18, and 22 (two intervals per week) after the first oral challenge, three birds from each treatment group and one from each control group were randomly selected and removed for analysis. The preparation of inoculum, euthanatization, necropsy, and sample collection procedures all followed the same procedure as described for Experiment 1.

Duodenal samples were evaluated for the presence of gross lesions including mucosal hyperemia, red foci/patches, and mucosal erosion. Each of these lesions was scored 0 = not present or 1 = present.

For histopathologic lesion scores, fixed duodenal samples were examined and scored by a trained pathologist with a second pathologist faculty member using a blinding method where no indication of challenge was specified on the slides. Intestinal samples were evaluated for the presence of lymphoplasmacytic inflammation, heterophilic inflammation, hemorrhage, necrosis of enterocytes, cystic crypts and/or crypt necrosis, and inflammatory infiltrate in the lumen. Each of these lesions was scored as 0 = no lesion, 1 = minimal, 2 = mild, 3 = moderate, and 4 = marked; the sum of lesion scores was used to determine the total microscopic lesion score per chicken.

Collected duodenal samples were struck to MAC or blood agar plates and incubated aerobically while a second set of blood agar media was incubated anaerobically at 37 C for 18–24 hr. Suspect colonies on MAC and blood agar were confirmed as EC using 16S rRNA PCR. Suspect colonies from MAC were selected and struck onto TSA (Becton and Dickinson) and incubated at 37 C for 18–24 hr. DNA was extracted using the boil prep method. Briefly, a single bacterial colony was picked and added to 1 ml of brain heart infusion (Becton and Dickinson) broth with further incubation at 37 C for 18–24 hr. Following incubation, the bacterial suspension was pelleted by centrifugation at 8000 × g for 10 min, and the supernatant was removed. The cell pellet was resuspended in 200 μl of sterile water and vortexed. The cell suspension was then placed on a heating block and boiled at 100 C for 10 min and then allowed to cool before centrifugation again to pellet the cell debris. The supernatant containing DNA (180 μl) was transferred to a new tube and used as a template for further analysis.

The amplification of EC 16S rRNA was performed using a Mastercycler X50s thermocycler (Eppendorf, Germany) in a final volume of 25 μl containing 2.5 μl of 10x PCR buffer, 1.25 μl of deoxyribonucleotide triphosphate (dNTP) mixture (10 mM), 0.4 μl of Taq DNA Polymerase (DreamTaq; Life Sciences, ThermoFisher), 18.65 μl of nuclease-free water, 0.1 μl of forward primer (UAL1939-187, 100 μM) ATGGAATTTCGCCGATTTTGC, and 0.1 μl of reverse primer (UAL2105-187, 100 μM) ATTGTTTGCCTCCCTGCTGC (16). PCR amplification was carried out using a Mastercycler X50s thermocycler and a protocol consisting of an initial denaturation of 94 C (5 min), followed by 30 cycles of 94 C denaturation (30 sec), 60 C annealing (30 sec), and 72 C extension (1 min), and then a final extension step at 72 C (10 min). Amplification products were separated on a 1.5% agarose gel using electrophoresis and stained with 0.25% ethidium bromide solution (Sigma-Aldrich, St. Louis, MO) for 20 min, and the fragment of 187 bp was visualized under ultraviolet light (UVP BioDock-It² Imager; Analytik Jena, Jena, Germany). A positive control strain of EC (lab strain) was included in the PCR analysis, and water in place of DNA was used as the negative control. Isolates that had a positive band size were considered EC, while isolates that failed to amplify as EC were sent for MALDI-TOF analysis at The University of Georgia Athens Veterinary Diagnostic Laboratory.

Cohen's kappa and inter-rater reliability tests were applied in comparing the microscopic lesion scores from two different pathologists. Cohen suggested the kappa result should be interpreted as follows: Values ≤0 indicate no agreement; 0.01–0.20 indicate no to slight agreement; 0.21–0.40 indicate fair agreement; 0.41–0.60 indicate moderate agreement; 0.61–0.80 indicate substantial agreement; and 0.81–1.00 indicate almost perfect agreement between scorers. Both macroscopic and microscopic lesion scores in Experiments 1 and 2 were subjected to nonparametric Kruskal-Wallis tests using Stata 18 (Release 18; StataCorp LLC, College Station, TX), with significance reported at P < 0.05. The data were analyzed by comparing the different group and week interactions. The final statistical graphs were made using GraphPad Prism version 10.0.0 for macOS (GraphPad Software, Boston, MA, www.graphpad.com).

No clinical signs or mortality cases were observed in the treatment or control groups of the study. All three gross lesions of the scoring system (mucosal hyperemia, red foci/patches, and mucosal erosion) were observed in the challenged groups (Fig. 1). On the other hand, mucosal hyperemia was the only lesion observed in the control group. The result of gross lesion scores is shown in Table 1. The findings indicated that T3 (CP + EC + CC) resulted in the highest lesion score (2 pts) on average. The lowest average lesion score among the treatment groups was observed for T4 (CP + EC + CC + EF + GA) with 1 pt recorded throughout the three necropsies. Lesion scores increased in T2 and T4 over the course of the study, while the lesion score for T3 plateaued from second to third necropsy period. Statistically, the average lesion score for T3 was significantly higher than that of the control group (Fig. 2A). Additionally, when the group data were compared by week, week 3 had significantly higher lesion scores than week 1, demonstrating an increase in overall lesion score even when the oral challenge was stopped at the third week (Fig. 2B).

Microscopically, enteric changes were mild to moderate (Fig. 1). Two separate pathologists evaluated the slides, and tests of agreement were also conducted (Table 2). The most significant and consistent findings were a mild to moderate number of lymphocytes, plasma cells, and heterophils infiltrating the lamina propria and occasionally the mucosa of the duodenum. Multifocally, there were dilated crypts filled with a moderate amount of eosinophilic material, cellular debris, and scant heterophils. Enterocyte necrosis and inflammatory cell infiltration of the duodenal lumen were absent to minimal, and some tissues had scant bacterial rods within the lumen with no association with duodenal villi or inflammation.

For the total lesion score by group (Fig. 2C), there was a significant difference between T2 (CP + CC) and the control group. No significant differences were observed between the different weeks. When examined for all five treatment categories and histopathologic lesions, T2 was significantly different to the control group in lymphoplasmacytic inflammation and necrosis of enterocytes categories (Fig. 2E). For all other categories scored, differences in scores were numerically greater for T2 compared to T1, T3, T4, and the control group, but these differences were not significant (P > 0.05).

For microbiologic analysis, 84 bacterial isolates were cultured from the lesions and were identified by 16S PCR or MALDI-TOF. Bacterial identification found 22/84 (26.2%) isolates typed as EC, 13/84 (15.5%) were identified as Ligilactobacillus salivarius, and the remaining isolates included Enterococcus, Staphylococcus, and miscellaneous bacterial species. Fifteen isolates (17.9%) could not generate a confident identity by MALDI-TOF (Table 3).

No clinical signs or mortality cases were observed in the challenged or control groups. The result of gross lesion scores is shown in Table 1. The findings indicated that T3 (CP + EC + CC) had the highest lesion score (1.67 pts) on average. The lowest average lesion score among the challenged groups was observed for T1 (CP only) with 1.22 pts. Control 1 had a higher score (0.5 pts) than Control 2 (0.17 pts). When lesion scores were examined throughout the length of the study, every group had different patterns. T1 lesion scores increased through each week; T2 appeared to drop in week 2 but increased in week 3; T3 lesion scores remained steady at 1.67 pts throughout the study. On statistical analysis, lesion scores for T2 and T3 were significantly higher than Control 2 but not Control 1 (Fig. 3A). No significant differences were observed between lesion scores in the same group when compared over the weeks sampled (Fig. 3B).

Microscopically, two separate pathologists evaluated the slides, and tests of agreement were assessed (Table 2). The findings were similar to Experiment 1 and included lymphocytes, plasma cells, and heterophils infiltrating the lamina propria and occasionally the mucosa of the duodenum. Dilated crypts were also found. Scant bacterial rods within the lumen were also observed without association with duodenal villi or inflammation. Statistical analysis showed lesion scores of T1 (CP only) were significantly lower than those of T2 (EC + CC), T3 (CP + EC + CC), and Control 1 (PBS + customized feed) (Fig. 3C). Moreover, T3 and Control 1 had significantly higher lesion scores than Control 2. There were no significant differences for intervals sampled within the same groups. When the total lesion score by group was examined, significant differences between groups were observed for necrosis of enterocytes and cystic crypts/crypt necrosis (Fig. 3E).

For microbiologic analysis, 63 bacterial isolates were cultured from the lesions collected and were identified by 16S PCR or MALDI-TOF. Bacterial identification found 36/63 (57.1%) isolates recovered were identified as EC, and 7/63 (11.1%) were identified as L. salivarius. Five isolates were identified as Klebsiella pneumoniae (7.9%), and four isolates (6.3%) could not generate a confident identity by MALDI-TOF (Table 3).

The first experiment in these studies aimed to identify the combination of bacterial species that resulted in higher lesion scores and would be the most promising candidates for use in the second experiment. We used bacterial mixtures instead of a single bacterial strain because previous studies suggested that the etiology of FDN might involve multiple bacterial pathogens (13,15). Moreover, CP has been suggested as the most common bacteria to be isolated from FDN lesions, so all four treatment groups included CP in the mix with different combinations of other bacteria to allow us to compare the lesion scores as a result of the challenges. Experiment 1 attempted to compare CP with EC (T1) or CC (T2) alone, CP with both EC and CC (T3), and a mixture of CP, EC, CC, EF, and GA (T4). There was a difference observed in gross and microscopic lesions in terms of the treatment group that resulted in significantly higher scores than the control group. In the gross lesion analysis, T3 (CP + EC + CC) scored significantly higher than the control group, while in microscopic lesion analysis, T2 (CP + CC) scored significantly higher (Fig. 2A,E). This difference could possibly have resulted from variation in scoring by different pathologists, and the scoring criteria. However, one consistent observation was that T2 and T3 both contained CP and CC. A recent study identified coinfection with CP and CC in commercial laying hens experiencing severe enteritis (17). Postmortem examination revealed round intestinal ulcerations with necrohemorrhagic centers, ranging in diameter from 0.5 to 3 mm. Histopathology showed erosions at the mucous surface, with the aggregation of bacteria and lymphoplasmacytic infiltration. The authors proposed that there was a synergistic effect between these bacterial species in causing severe necrotic enteritis (17). In contrast, the lesions observed in our challenge model exhibited mild to moderate lymphoplasmacytic inflammation without mucosal erosion, indicating a potentially less severe intestinal inflammatory response in our model compared to the one described in the literature, but the result is somewhat similar to another FDN challenge experiment conducted by Villegas et al. (15).

CC is recognized as the primary causative agent of acute to chronic ulcerative enteritis (UE) in various avian species (18). While most commonly affecting bobwhite quail (Colinus virginianus), a case of UE was diagnosed in broiler chickens with CC isolated from the lesions in Japan (19). Though established as the cause of avian UE, particularly in quails, the virulence mechanisms of CC remain poorly understood (20). While suggestions of a toxin role in UE pathogenesis exist (21,22), its identification and characterization are currently lacking. Elucidation of specific virulence factors and their role in the disease is crucial for understanding the pathogenesis of CC and its effect on the host. The EC strain used in this challenge model was isolated from a bird with FDN lesions. This strain exhibits high genetic similarity to O25b:H4 ST131 uropathogenic EC strains and other EC strains associated with IBD in humans. Importantly, the strain possesses virulence genes like cea (colicin E1), cnl (colicin N lysis protein), and ompW. These genes may contribute to the strain's survival within the host, ability to compete with other bacteria, and, potentially, increased resistance to the avian immune response (23,24,25). Further investigation is warranted to elucidate the specific roles of these virulence factors in FDN pathogenesis. Our experiment demonstrates a positive correlation between the presence of CC and EC in the challenge mixture and increased lesion scores. These scores were significantly higher compared to the control group.

Based on the findings of Experiment 1, the combinations CP + CC and CP + CC + EC were identified as the most likely candidates for inducing higher lesion scores. Experiment 2 was designed to elucidate the impact of coinfection of CP with or without CC and EC on lesion severity. In this experiment, T1 consisted of CP only, while T2 contained only the combined inoculum of EC and CC. Finally, T3 (CP + EC + CC) represented the combination of strains that yielded the highest score in Experiment 1. Additionally, control groups were modified in Experiment 2. Because Experiment 1 used a control group that contained the customized feed designed to address potential predisposing factors of FDN, it may not fully represent the true control for the challenge experiment. Therefore, Experiment 2 employed two control groups with Control 1 maintained on the custom feed described in Experiment 1 and Control 2 receiving a standard commercial layer feed. As anticipated, T3 exhibited the highest overall histopathologic lesion score among the treatment groups and displayed a statistically significant difference compared to Control 2 (Fig. 3C). However, an unexpected observation was the significantly elevated lesion score in Control 1 compared to Control 2 and even T1. One potential explanation for this increased score could be a statistical artifact caused by the difference in sample size between the groups. Control groups (n = 6) had a substantially lower number of birds compared to treatment groups (n = 18). Smaller sample sizes can lead to increased variability and potentially underestimate the true population mean lesion score (26). As this was an initial investigation focused on replicating FDN lesions, we opted for a small bird sample size in the experiments. Nevertheless, the results provide the insight that CP alone could not cause as high a lesion score compared to the combination of CP + EC + CC.

The potential predisposing factors influencing the development of FDN were previously investigated by Villegas et al. (12). A questionnaire survey gathered information from farms affected by FDN and identified several potential predisposing factors that may contribute to the disease. From a dietary perspective, this study included several key ingredients referenced from that study in our customized feed formula. For example, the formula comprised 10% DDGS, 5% animal protein, 20% fine limestone, and 80% coarse limestone. DDGS is a rich source of crude protein, amino acids, phosphorus, and other nutrients. It can be included at levels of 5%–20% in diets and can supply up to one third of the protein requirement without negatively impacting egg production and egg weight (27,28,29). However, a study in broilers suggested that higher DDGS inclusion rates increased the incidence of necrotic enteritis (30). Moreover, the survey of FDN-affected layer farms indicated that DDGS was used in the majority of these operations (81%). Animal protein is also a common ingredient in the poultry diet and has been used for a long time (31). Increased dietary protein intake is associated with elevated protein concentration within the gastrointestinal tract. However, this enriched protein environment may selectively promote the growth of specific bacterial populations, such as CP (32). Studies in broilers have demonstrated a positive correlation between high inclusion levels of animal-based protein sources (e.g., fishmeal, meat and bone meal) and the incidence of CP and subsequent necrotic enteritis (33,34). Regarding the inclusion of limestone, which is used as a calcium supplement during egg production in layers (12), higher dietary calcium from limestone has been reported to negatively affect mortality associated with necrotic enteritis and broiler performance (34).

Overall, although this study did not successfully reproduce FDN-specific histopathologic lesions, particularly the characteristic “volcano-like lesions” described as villous tip necrosis, enterocyte loss, luminal infiltration with fibrinoheterophilic inflammatory cells, bacterial aggregates, degenerated cells, and erythrocytes. However, the study results suggest that some of the bacterial species examined here play a key role in the development of lesions and damage to the duodenum of challenged birds. Further investigation is needed to refine the challenge model to accurately represent the disease pathology. There are some modifications that could be done in the future. In several necrotic enteritis challenge experiment models, the inoculum of CP was mixed in feed at a ratio of 1.25–1.5 fluid thioglycollate medium:feed (v/w) (33,35). Advantages of in-feed challenge models include eliminating the need for individual bird handling, which can be stressful for the birds, and they ensure a continuous intake of bacterial inoculum throughout the challenge period, unlike the once-a-day high volume of oral challenges described in this study. However, disadvantages include the unpleasant odors produced by the bacterial inoculum, such as hydrogen sulfide produced by CP, which can make the feed less palatable and potentially reduce its consumption by birds (33). On the other hand, infecting individual birds using a gavage needle directly into the crop with bacterial cultures has also been applied in several necrotic enteritis challenge models (36,37,38) and does result in an accurate, controlled dose in each bird compared to the effect if the challenge is based on feed intake alone. Notably, a successful challenge using CP depends on initiating intestinal damage through preformed toxins rather than toxins produced within the intestine. Therefore, it is essential to use whole broth cultures that contain these preformed toxins rather than relying solely on vegetative CP cells (39). In our experiment, we centrifuged the broth cultures and took only the pellet for resuspension in PBS. This approach may have contributed to a loss of toxin in the broth and might be another reason why the lesions observed in our experiment were limited, and it may also explain why we were unable to culture the Clostridium strain postinoculation. The challenge strain may also have been outcompeted in the avian gut. This is not an unusual observation since the gut contains such a wide and varied microflora, and we were culturing inocula from lesions that were not as definitive as desired. However, a previous FDN challenge experiment used the complete CP culture and broth and could only generate mild duodenal lesions similar to those observed in the current study (15). Taken together, these findings suggest that all these additional factors may need to be considered for the successful reproduction of FDN in layers.

This is the first report to use a bacterial mixture in an FDN challenge experiment. The results showed that the combination of CP, CC, and EC caused the highest overall lesion scores, even though our challenge model failed to reproduce the specific FDN lesions in the duodenum. For future successful reproduction of FDN, additional factors may need to be considered, such as other unknown predisposing factors, toxins produced by challenge strains, longer challenge periods, different challenge routes, and modified inoculum preparation. These elements could potentially play a role in developing a successful model.

Supplemental data associated with this article can be found at https://doi.org/10.1637/aviandiseases-D-24-00069.s1.

This work was funded by a grant from the Egg Industry Center at Iowa State University under grant number SG2706767. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the author(s) and do not necessarily reflect views of the Egg Industry Center or Iowa State University.

CC =

Clostridium colinum;

CFU =

colony-forming units;

CP =

Clostridium perfringens;

DDGS =

distiller’s dried grain with solubles;

EC

Escherichia coli;

EF =

Enterococcus faecalis;

FDN =

focal duodenal necrosis;

GA =

Gallibacterium anatis;

IBD =

inflammatory bowel disease;

MAC =

MacConkey agar plates;

MALDI-TOF =

matrix-assisted laser desorption ionization time-of-flight;

netB =

necrotic enteritis B–like toxin gene;

PBS =

phosphate-buffered saline;

rRNA =

ribosomal RNA;

TSA =

tryptone soy agar;

UE =

ulcerative enteritis