Introduction

Elimination of potential pathogens by the innate and acquired immune systems involves an intrinsic compromise between host protection and tissue injury that requires mechanisms to control the nature, magnitude, duration and specificity of effector pathways. A key component of specificity is the capacity to distinguish between pathogens and the commensal microbiome, and nowhere is this challenge more pressing than in the intestine. The anti-inflammatory cytokine, interleukin 10 (IL10), has long been recognized as an essential feedback regulator of intestinal inflammation in mammals (reviewed in (1)). IL10-deficient mice generated by homologous recombination developed spontaneous enterocolitis (2). Subsequent studies revealed that components of the intestinal microbiome were essential to development of pathology and determined the extent and location of lesions (3). In humans, homozygous mutations in genes encoding IL10 or its receptor, IL10RA, have been associated with early-onset inflammatory bowel disease (1).

In mammals, IL10 is produced by both myeloid and T and B lymphoid cells following activation by immune stimuli (1, 4), with subtle differences between mice and humans (5). In the specific context of the intestine, conditional deletion of the IL10 and IL10RA genes in mice indicated that production of the ligand by macrophages was not required for pathology, consistent with T cells being the major source of intestinal IL10 (6, 7). By contrast, macrophage-specific deletion of the IL10RA gene encoding the IL10 receptor was sufficient to elicit spontaneous colitis in mice (8–10). The contribution of IL10 to feedback modulation of host defense is not restricted to the gut. There have been countless studies of the impact of IL10 deficiency in mouse parasite, bacterial, fungal and viral disease models (reviewed in (4)). The primary focus of these studies has been on the mitigation of immunopathology by IL10, with less attention given to pathogen elimination. By contrast to studies in mice, analysis of the quantitative importance of IL10 in other species has been largely correlative. Both ligand and receptor are conserved in vertebrates, and zebrafish has been discussed as a model for understanding IL10 biology (11). Here we focus on the chicken, both a model vertebrate and an economically important livestock species.

The original cloning and sequencing of chicken IL10 cDNA and genomic DNA (12) described the immunomodulatory activity of the recombinant protein on T cell production of interferon-gamma (IFNγ), its regulated expression in T cells and macrophages in vitro, and increased mRNA expression during parasite (Eimeria maxima) infection. Like the mammalian transcript, chicken IL10 mRNA contains multiple copies of the AUUUA instability motif in the 3’-untranslated region (UTR). We and others have provided evidence that quantitative variation in IL10 production may contribute to inter-individual or inter-breed variation in infectious disease susceptibility in chickens (13–17). To test this hypothesis directly, and more generally to determine the conservation of IL10 biological function across vertebrates, we have generated germ line modifications of the chicken IL10 locus that abolish or reduce IL10 expression. We describe the lack of dosage compensation in heterozygous mutation, and present evidence that the reduced expression of IL10 leads to increased pathogen clearance at the expense of increased inflammation.

Results

Generation of IL10-deficient chickens

The emergence of CRISPR/Cas9 technology and the ability to culture, cryopreserve, genetically modify and subsequently transfer chicken primordial germ cells (PGCs) into sterile surrogate host embryos has expedited the analysis of gene function in the chick (Figure 1A; (18–21)). We used this approach to edit the IL10 locus in chicken PGCs and create two independent lines of IL10-deficient chickens (Figure 1). We first created an IL10 knockout (IL10KO) line by introducing a premature in-frame stop codon in exon 1 of the IL10 gene in chicken PGCs, 75 bp downstream of the start codon (C26>*; Figure 1B and D, and SI Appendix, Figure S1). To generate a potential expression hypomorph, we created an IL10-Enhancer knockout (IL10EnKO) line by deleting a 533-bp genomic fragment encompassing a non-coding sequence located approximately 2.3-2.8 kb upstream of the IL10 transcription start site, in chicken PGCs (Figure 1B-C and SI Appendix, Figure S2). This sequence was targeted based upon alignment of sequences from multiple avian species that identified a core of 300 bp with conservation around 85%, containing perfectly conserved motifs for known transcriptional regulators, including AP1, FOXP3 and members of the NFAT/ETS family (SI Appendix, Figure S3). The size and species conservation of this element resembles the well-characterized intronic enhancer of the Csf1r locus (22).

Figure 1:

To create IL10KO and IL10EnKO chickens, male PGCs carrying the desired biallelic edit in IL10 exon 1 or in the putative IL10 enhancer region (as validated by PCR and DNA sequencing; Figure 1C-D and SI Appendix, Figure S4) were injected into iCaspase9 sterile surrogate host embryos, in which a chemically inducible caspase 9 protein is expressed from the PGC-specific DAZL locus, together with a GFP reporter (23). Embryos were treated with the iCaspase9 activation chemical AP20187, incubated to hatch, and the resulting chicks were raised to sexual maturity under specified pathogen-free (SPF) conditions at the National Avian Research Facility (NARF). Mating of the iCaspase9 surrogate host cockerels to Rhode Island Red (RIR) hens generated 100% IL10KO or IL10EnKO heterozygous (HET) offspring in the first generation (G1; Figure 1A and SI Appendix, Table S1). Upon sexual maturity, G1 HET birds were intercrossed to generate the second generation (G2) of IL10KO and IL10EnKO wildtype (WT), HET and homozygous (HOM) birds for analyses. Both lines produced WT, HET and HOM offspring at a 1:2:1 mendelian ratio (SI Appendix, Table S1).

Confirmation of the IL10 mutation in bone marrow-derived macrophages

Chicken bone marrow-derived macrophages (BMDMs) cultured in recombinant macrophage colony-stimulating factor respond to bacterial lipopolysaccharide (LPS) by secreting IL10 that can be detected by ELISA (24). The autocrine function of IL10 in this assay was demonstrated by the addition of neutralizing anti-IL10 antibody, which amplified the expression of nitric oxide synthase (NOS2) and production of nitric oxide (24). Figure 2A confirms the release of IL10 by LPS-stimulated BMDMs derived from WT embryos and the complete absence of detectable IL10 release by BMDMs derived from IL10KO HOM embryos. The absence of IL10 was associated with a significant increase in nitric oxide production that was not additive with the impact of added anti-IL10 antibody (Figure 2B). Since IL10 mediates feedback control it is conceivable that the loss of one functional allele could be compensated by increased expression from the other allele. However, we found that expression of IL10 was reduced by up to 50% in BMDMs derived from IL10KO HET embryos, with a corresponding partial impact on nitric oxide production (Figure 2A-B). The impact of the IL10EnKO HOM mutation was more variable between samples in both assays, but consistent with a partial expression hypomorph (Figure 2A-B).

Figure 2:

Loss of IL10 does not lead to spontaneous immunopathology

The IL10KO and IL10EnKO chicken lines were created and raised under SPF conditions in the NARF SPF avian facility. Close monitoring of the WT, HET and HOM birds over several months in that environment did not reveal any health issues or adverse phenotypes. The IL10KO and IL10EnKO edits did not significantly affect the post-hatch growth of HET and HOM birds compared to their respective WT controls (Figure 3A-B), although IL10EnKO HET and HOM hens were overall heavier than IL10EnKO WT controls (Figure 3B). Regular gross post-mortem examination of tissues and histopathological analyses of the gastrointestinal tract over several months (using a semiquantitative scoring system (62)) revealed no evidence of intestinal pathology in IL10KO HOM and IL10EnKO HOM birds compared to WT controls (Figure 3C). These results are in contrast with the early onset colitis and failure to thrive phenotypes observed in IL10-deficient mouse models in both conventional and SPF environments (2, 9, 25, 26).

Figure 3:

Spontaneous colitis in Il10^-/- mice is associated with and dependent upon dysbiosis in the intestinal microbiome (27–29). We compared the caecal microbiome of IL10KO WT and HOM birds at 4 weeks of age using 16S rDNA V3-V4 hypervariable amplicon sequencing as described (30, 31). The caecal microbiome of both genotypes was dominated by the phylum Firmicutes (SI Appendix, Figure S5A) and the most abundant genera were Lactobacillus and Faecalibacterium (SI Appendix, Figure S5B). Linear discriminant analysis (LDA) Effect Size (LEfSe) analysis showed enrichment of the phylum Firmicutes and the genera Defluviitaleaceae_UCG-011 and Paludicola in the IL10KO HOM group, while the phylum Bacteroidota and the genera Alistipes and UCG_008 were enriched in the WT group (SI Appendix, Figure S5C). Principal component analyses showed clear separation of the microbial communities based on beta diversity (SI Appendix, Figure S5D, R² = 0.25, P = 0.017). These findings confirm that mucosal IL10 regulates the composition of the chicken microbiome.

To evaluate a potential environmental impact on the phenotype of IL10-deficient birds, a cohort of IL10KO WT and HOM birds was hatched and raised in the NARF conventional avian facility. IL10KO HOM birds remained healthy in that environment as they aged. Post-growth hatch rates were comparable between IL10KO WT and HOM hens, with the exception of a marginal effect on growth that resolved with age (Figure 3D). Whilst there was no statistically significant evidence of overt intestinal pathology in IL10KO HOM birds compared to WT controls (Figure 3E), cellular infiltration scores were overall significantly higher in IL10KO WT and HOM birds raised in a conventional facility as opposed to a SPF environment (Figure 3F).

Two important differences exist between the IL10KO mouse and chicken models. Firstly, the birds are outbred. Most mouse studies are conducted on an inbred background and the impact of the IL10 mutation is strain-specific (26). Secondly, chickens in the NARF conventional avian facility are routinely immunised against major avian pathogens (SI Appendix, Table S2). The original analysis of IL10-deficient mice revealed little impact on T cell-dependent antibody responses (2). We measured antibody titres in 29-weeks old IL10KO WT and HOM birds raised in the NARF conventional facility for 7 of the pathogens that the birds were vaccinated against (SI Appendix, Figure S6): avian encephalomyelitis virus (AEV), chicken anaemia virus (CAV), duck adenovirus (the agent of egg drop syndrome, EDS), infectious bursal disease virus (IBDV), infectious bronchitis virus (IBV), infectious laryngotracheitis virus (ILTV), and Newcastle disease virus (NDV). The administration schedule, nature and delivery mode of these vaccines is summarised in SI Appendix, Table S2.

Statistical analyses revealed differences in vaccine response between IL10KO WT and HOM birds. Antibody titres to AEV and IBDV were significantly lower in IL10KO HOM birds compared to WT controls, whereas those to EDS and IBV were significantly higher in IL10KO HOM birds compared to WT controls (SI Appendix, Figure S6). AEV and IBDV vaccination were carried out with live vaccines delivered in the drinking water, whereas EDS and the last dose of IBV vaccination were done with inactivated vaccines given by intramuscular injection (SI Appendix, Table S2). Antibody titres to CAV, ILTV and NDV were not significantly different between IL10KO HOM birds and WT controls (SI Appendix, Figure S6).

The impact of IL10 deficiency on the response to enteric bacterial infection

Campylobacter jejuni (C. jejuni) and non-typhoidal Salmonella are zoonotic diarrhoeal pathogens of global importance (32). Breed-specific pathology associated with C. jejuni infection in chickens was associated with relatively low caecal IL10 mRNA (33). Similarly, IL10-deficient C57BL/6J mice exhibited severe inflammation in the caecum and colon upon C. jejuni infection, whereas congenic wild-type mice were persistently colonised in the absence of gut pathology (34). A similar outcome was reported in IL10-deficient mice infected with Salmonella Typhimurium (S. Typhimurium) (35).

To study the role of IL10 during Campylobacter infection in chickens, two cohorts of 20 IL10KO WT, HET or HOM chickens obtained from the NARF SPF avian facility were inoculated at 14 days-old with 10² colony forming units (CFU) C. jejuni strain M1 and analysed 1- and 2-weeks post-infection. The dose of strain M1 used is the lowest needed to achieve reliable caecal colonisation (36). In the first cohort, colonisation of the caeca by C. jejuni in IL10KO HOM birds was significantly lower at 1-week post-infection, with a median of 4.8 log₁₀ CFU g^-1 compared to >9 log₁₀ CFU g^-1 in WT birds Figure 4A). However, this was not sustained and by 2-weeks post-infection all groups had caecal colonisation at >9 log₁₀ CFU g^-1 (Figure 4A). In the second cohort, median counts were not statistically different between IL10KO HOM and WT birds at 1-week post-infection, but caecal colonisation decreased by ∼2 log₁₀ CFU g^-1 in IL10KO HOM birds at 2-weeks post-infection relative to WT birds (Figure 4A). The IL10KO HET birds did not differ significantly from WT in either cohort (Figure 4A). Histopathological analysis revealed that the caecal mucosa of IL10KO HOM birds presents with significantly increased numbers of lymphocytes and plasma cells, admixed with infiltrating heterophils, when compared to WT and IL10KO HET birds in both cohorts (SI Appendix, Figures S7 and S8). To confirm the apparent hyperinflammatory response in IL10KO HOM birds, we used multiplex quantitative reverse transcriptase-PCR (qRT-PCR) to measure expression of immunity-related genes in the caecum, as described (37). In both cohorts we detected large increases in expression of pro-inflammatory cytokines, chemokines and other effectors, notably Nos2, in the IL10KO HOM caecum relative to WT (Figure 4B).

Figure 4:

To test the role of IL10 in response to Salmonella infection, groups of 20 IL10KO WT, HET or HOM chickens were inoculated at 14 days-old with 10⁸ CFU S. Typhimurium strain ST4/74 nal^R (38). The impact of lack of IL10 on bacterial burden in caecum, liver and spleen was marginal at 1 week, but by 2 weeks the majority of IL10KO HOM birds had eliminated the pathogen, whereas WT birds retained bacteria in both caecum and spleen (Figure 4C). Enhanced lymphoplasmacytic and heterophil infiltration was observed in the ileum and caecum of IL10KO HOM compared to WT or IL10KO HET birds at both intervals after S. Typhimurium infection (SI Appendix, Figure S9). As observed following the C. jejuni challenge, analysis of inflammatory gene expression by multiplex qRT-PCR revealed a substantial increase in response in the IL10KO HOM birds relative to WT (Figure 4D). Volcano plots showing the direction and magnitude of gene expression differences detected between IL10KO HOM and WT birds following bacterial infection are shown in SI Appendix, Figures S10 and S11. In summary, the absence of IL10 leads to a sustained increase in inflammation in response to enteric bacterial infection that is also associated with more effective clearance.

The impact of IL10 deficiency on the outcome of enteric protozoan infection

Eimeria tenella (E. tenella) is one of ten protozoan parasite species that can cause coccidiosis in chickens (39). Eimeria infection of chickens induces a robust IFNγ-driven immune response with evidence of higher IL10 expression associated with increased susceptibility (12, 40). Following an initial titration, an infectious dose of 7000 sporulated E. tenella oocysts per bird was set as the standard challenge, as this induced measurable phenotypes in pathology, performance, and parasite replication in the absence of overt disease in WT birds. Oral challenge with this or a mock dose was administered to cohorts of 10 IL10KO HOM and IL10EnKO HOM chickens and corresponding WT and HET controls on day 21 post-hatch. The outcomes are shown in Figure 5. Average parasite replication was determined by quantitative PCR (qPCR) as the ratio of parasite genomes detected per host genome in total caecal tissue collected six days post-infection. Parasite burden was almost ablated in IL10KO HOM and significantly reduced in IL10EnKO HOM chickens compared to their respective WT controls, whereas HET mutations had no significant effect in either case (Figure 5A-B). Histology confirmed reduced parasite replication in IL10KO HOM chickens, with dense populations of late-stage gametocytes observable in WT caecal tissues and correspondingly very sparse numbers in IL10KO HOM chickens (Figure 5C-D). This selective reduction of parasite burden in IL10KO HOM chickens was associated with a significant increase in caecal lesion score and a substantial reduction in body weight gain over the 6-day challenge interval (Figure 5E and G). By contrast, both IL10KO HET and IL10EnKO HOM and HET (each anticipated to reduce IL10 expression) showed reduced caecal lesion scores and had no effect on body weight gain compared to their respective WT controls (Figure 5E-H).

Figure 5:

To better characterise the differences in pathology, performance and parasite replication detected between IL10KO WT and HOM chickens, we undertook a time course study, analysing parasite burden by qPCR at 0-, 2-, 4-, 6-, 8- and 10-days post-infection. In parallel, excreted oocysts per gram (OPG) were counted in litter samples collected from IL10KO HOM and WT groups during the period of peak oocyst excretion at 7-, 8- and 9-days post-infection. Figure 6A confirms the greatly reduced E. tenella replication in IL10KO HOM, that was reflected in corresponding reduced excretion of oocysts (Figure 6B). Lesions associated with E. tenella infection were accelerated in IL10KO HOM chickens compared to WT (Figure 6C). The adverse impact of IL10KO HOM on body weight gain was sustained to day 8 and showed signs of recovery by day 10 (Figure 6D), likely due to complete pathogen elimination (Figure 6A). Consistent with the mild inflammatory cell infiltration in IL10KO HOM birds noted in Figure 5D, inflammatory gene expression was elevated in the caecum prior to challenge (Day 0). The greatest difference in IL10KO HOM birds relative to WT was detected on day 6, during the period of peak pathology (Figure 6C and E). Volcano plots showing the direction and magnitude of gene expression differences detected between IL10KO HOM and WT birds following E. tenella infection are shown in SI Appendix, Figure S12, highlighting genes associated with inflammation (Il6) and lysosomes (Hps5). Transcription of genes associated with movement of molecules across membranes (Pkd2l1, Abcg2) and tubulin production (Tubat) was lower in IL10KO HOM birds 6 days post-infection.

Figure 6:

Discussion

An intrinsic compromise exists between effective host defence leading to pathogen elimination and the induction of immune-mediated pathology. The central importance of IL10 in mediating this balance has been studied extensively in the mouse, notably in the context of parasite infections (1, 4, 5, 11, 41). The regulatory function of IL10 is especially important in preventing inappropriate responses to the intestinal microbiome with the potential to drive spontaneous enterocolitis (2, 8, 10, 26). Here we have generated two IL10 mutations in the chicken genome, a definitive knockout and a putative enhancer mutation. Notwithstanding the development of robust technologies based upon CRISPR/Cas9, the development of models of immune deficiency in chickens remains time consuming and logistically challenging, and relatively few informative mutations have been generated (20, 42). The enhancer mutation was intended to generate a quantitative impact on IL10 expression to model selection for reduced expression in production lines of birds. Based upon the mouse Il10^-/- phenotype, we also anticipated a possible severe post-hatch failure to thrive due to spontaneous colitis in IL10KO HOM chicks. An unexpected outcome was the discovery that the IL10KO in chicken is not dosage compensated, therefore the HET provides a second model of the impact of reduced expression. In overview, neither the IL10KO HET nor the IL10EnKO HOM had a large effect on pathogen clearance or inflammation in either of the challenge models, indicating that IL10 biology is not highly sensitive to changes in expression of the magnitude that has been detected in individual lines of commercial broilers and layers (14, 43, 44).

Detailed analysis of the IL10KO HOM birds in SPF or conventional avian facilities revealed no evidence of spontaneous intestinal inflammation or compromised body weight gain, and only a small and inconsistent effect on specific antibody responses to routine live or inactivated vaccines. Since spontaneous colitis in mice depends upon resident enteric bacteria (3), it is possible that the difference compared to the rodent models lies in the nature of the chicken intestinal microbiome (45). In any case, the results indicate that deletion of IL10 from the genome or selection in favour of low IL10 expression alleles (14, 17, 43, 44) need not necessarily affect production traits.

In both the bacterial challenge models, and in the Eimeria challenge, the response of IL10KO HOM bird highlights for the first time in a non-rodent species the compromise between pathogen elimination and pathology. Accelerated pathogen clearance is associated with increased inflammatory gene expression and local inflammation, and in the case of Eimeria, reduced body weight gain. The models we have chosen are not lethal and there is effective pathogen clearance even in WT birds. Findings with Campylobacter in IL10KO HOM birds are consistent with elevated intestinal inflammation in a broiler breed with relatively low IL10 expression, albeit a significant reduction in caecal colonisation relative to breeds with higher IL10 expression was not observed (33). In mammals, Salmonella has been reported to gain a competitive advantage from intestinal inflammation, both owing to availability of tetrathionate that Salmonella can uniquely use as a respiratory electron acceptor (46) and remodelling of the indigenous microbiota (47). It is unclear if Salmonella derives the same benefits during colonisation of the avian intestine and the counts of viable bacteria recovered from the caeca of WT and IL10KO HOM birds can only be interpreted as the balance of bacterial growth and killing.

Our results with Eimeria are relevant to commercial studies aimed at targeting IL10 to mitigate disease. Coccidiosis has been estimated to exert costs in excess of £10 billion every year to the chicken production industry (48), with E. tenella among the most common and pathogenic species (49, 50). Feeding neutralizing anti-IL10 antibody to chickens during Eimeria infection was reported to mitigate reduced body weight gain (BWG) but had an inconsistent impact on parasite replication and pathology (51, 52). Similarly, feeding antibody to IL10 receptor 2 (but not receptor 1) reduced lost BWG, but had no effect on oocyst shedding (53). Feeding chickens genetically modified corn (Zea mays) expressing IL10 improved BWG and reduced oocyst shedding and caecal lesions (54). These results are consistent with the phenotypes associated with a reduction, rather than complete absence, of IL10 observed here.

Given the excessive inflammatory activation we observed in non-lethal infections, and the literature in mice, IL10KO HOM birds or hypomorphs might be hyper-sensitive to more virulent bacterial infections and viral pathogens, such as avian influenza, that can elicit a cytokine storm. On the other hand, they may be primed for a more rapid and effective response. These possibilities can now be tested using the models we have developed.

Materials and Methods

Animals

All chicken lines used in this study were bred and maintained at the National Avian Research Facility (NARF), University of Edinburgh, under UK Home Office Establishment Licence and in compliance with the Animals (Scientific) Procedures Act (ASPA) 1986. All regulated procedures were approved by local ethical review committees and carried out by UK Home Office Personal Licence holders under UK Home Office Project Licences 70/8528 and PP9565661 (creation and maintenance of genetically altered chicken lines), PCD70CB48 (Campylobacter and Salmonella challenges performed at the Moredun Research Institute) and PDAAD5C9D (Eimeria challenges performed at the Royal Veterinary College). Animals were humanely culled in accordance with Schedule 1, ASPA 1986.

The Rhode Island Red (RIR) line was maintained as a closed outbred population in the NARF specified pathogen-free (SPF) avian facility. The CSF1R-eGFP line (55) was created and maintained on a Hy-Line background in the NARF conventional avian facility by HOM x HOM crosses; CSF1R-eGFP fertile eggs were used for PGC derivation. The iCaspase9 line (23) was created and maintained on a RIR background in the NARF SPF avian facility by crossing iCaspase9 HET cockerels to RIR hens; iCaspase9 fertile eggs were used as hosts for PGC injection. The IL10KO and IL10EnKO lines were created in the NARF SPF avian facility and maintained on a mixed Hy-Line / RIR background by crossing iCaspase9 surrogate host cockerels carrying gene-edited PGCs (generation G0) to RIR hens to produce the first generation of IL10KO HET and IL10EnKO HET chickens (G1); the second generation (G2) of IL10KO and IL10EnKO chickens was obtained by crossing G1 HET cockerels to G1 HET hens from each line, resulting in WT, HET and HOM offspring.

Chicken PGC derivation and culture

PGC lines were derived from blood collected from CSF1R-eGFP HOM embryos at Hamburger-Hamilton (HH) stage 16 and expanded in FAOT medium as previously described (56). A W-chromosome-specific PCR (57) was used to determine the sex of the PGC lines. PGC lines were tested for Mycoplasma contamination by PCR using primers specific for Mycoplasma synoviae (MS; Myco5_14713A and Myco3_14709) and Mycoplasma gallisepticum (MG; Myco5_14712 and Myco3_14709); see SI Appendix, Table S3 for primer sequences.

CRISPR plasmid, guide RNAs and ssODN donor

Guide RNAs (gRNAs) were designed using CHOPCHOP (58) and CRISPOR (59) and cloned into the High Fidelity CRISPR/Cas9 plasmid HF-PX459 V2.0 (60) as described previously (61). A gRNA (ggIL10_exon1_g3) and a 143-bp single-stranded oligonucleotide DNA (ssODN; IL10_exon1_HDRoligo2; IDT Ultramer^TM DNA Oligonucleotide) containing a STOP codon (TAG) and AvrII restriction site (CCTAGG), were used to edit IL10 exon 1. Two gRNAs (IL10_enhancer_guide1 and IL10_enhancer_guide2) were used in combination to delete the IL10 putative enhancer region. See SI Appendix, Table S4 for gRNA and ssODN sequences.

PGC transfection and screening

To edit IL10 exon 1, 200,000 CSF1R-eGFP male PGCs were transiently transfected with 1.5 μg gRNA plasmid ggIL10_exon1_g3 and 10 pmol ssODN IL10_exon1_HDRoligo2. To delete the IL10 putative enhancer region, 200,000 CSF1R-eGFP male PGCs were transiently transfected with 1.0 μg gRNA plasmid IL10_enhancer_guide1 and 1.0 μg gRNA plasmid IL10_enhancer_guide2. Transient transfections were performed with Lipofectamine 2000 (Invitrogen), followed by treatment with 0.4 μg/mL puromycin to enrich for transfected PGCs, as previously described (60). Bulk PGC populations were screened by PCR to confirm the presence of the desired IL10KO or IL10EnKO edits. Single-cell clonal populations were subsequently expanded as previously described (60) and screened by PCR and sequencing. Primers IL10_Exon1_F4 and IL10_Exon1_R4 were used to amplify a 1826-bp fragment encompassing IL10 exon 1, followed by digestion with AvrII to identify PGCs carrying the IL10KO edited allele (SI Appendix, Figure S4A); primers IL10_Exon1_F1 and IL10_Exon1_R1 were subsequently used to sequence the IL10 exon 1 fragment. Primers IL10-Enhancer_F2 and IL10_Enhancer_R2 were used to amplify and sequence a fragment encompassing the IL10 putative enhancer region (SI Appendix, Figure S4C). See SI Appendix, Table S3 for primer sequences and expected fragment sizes. PGCs were transfected and cultured under SPF conditions and regularly tested for mycoplasma contamination, as described above.

Generation of surrogate host chickens and establishment of the IL10KO and IL10EnKO lines under SPF conditions

Male PGCs carrying the desired IL10KO or IL10EnKO biallelic edits were injected into iCaspase9 embryos at HH stage 16, together with the iCaspase 9 activator AP20187 (Takara Bio) as previously described (23), to create surrogate host cockerels carrying IL10KO HOM or IL10EnKO HOM germ cells (G0 founders); all manipulations were performed under SPF conditions in the NARF SPF avian facility. G0 founders were hatched and raised to sexual maturity in negative pressure isolators and released to floor pens at 22 weeks. The SPF status of the G0 founders was externally assessed at three timepoints (10-, 20- and 30-weeks post-hatch) by screening for 17 different pathogens (Sci-Tech Ireland; SI Appendix, Table S5).

G0 founder cockerels were bred to RIR hens to produce the first generation of IL10KO HET and IL10EnKO HET chickens (G1); the second generation (G2) of IL10KO and IL10EnKO chickens was obtained by crossing G1 HET cockerels to G1 HET hens, resulting in IL10KO and IL10EnKO WT, HET and HOM offspring. All G1 and G2 chicks were genotyped by PCR using genomic DNA extracted from chorioallantoic membrane or 4-μL blood samples as previously described (23). Primers IL10_Exon1_F1 and IL10_Exon1_R1 were used to amplify a 246-bp fragment encompassing IL10 exon 1, followed by digestion with AvrII to identify the IL10KO WT and edited alleles (SI Appendix, Figure S4B). Primers IL10-Enhancer_F2 and IL10_Enhancer_R2 were used to amplify the IL10EnKO WT and edited alleles (SI Appendix, Figure S4D). Primers LTR_U3_F, 235_F and 235_R were used to amplify the CSF1R-eGFP WT and transgenic alleles. See SI Appendix, Table S3 for primer sequences and expected fragment sizes. The SPF status of G1 and G2 chickens was verified at point of lay (20-24 weeks) and at end of life (around 40 weeks).

Animal monitoring, post-mortem examination and histopathological analyses under SPF and conventional conditions

All IL10KO and IL10EnKO chicks hatched in the NARF SPF avian facility were closely monitored for signs of reduced growth, health issues, adverse phenotypes or unexpected behaviours. All chicks were weighed daily in the first week post-hatch and twice a week until four weeks; a small cohort of G2 IL10KO and IL10EnKO chicks was subsequently weighed weekly, up to a maximum of 30 weeks. Post-mortem examination of one WT and one HOM bird from both IL10KO and IL10EnKO lines was performed monthly from 16 to 40 weeks, with a particular focus on the gastrointestinal tract. Tissue samples from the proximal colon, ileum and duodenum were collected for histopathological analyses at all timepoints and scored in a blinded manner using a semiquantitative scoring system (62) to grade the extent of destruction of normal mucosal architecture (0–3), the presence and degree of cellular infiltration (0–3), the extent of muscle thickening (0–3), the presence or absence of crypt abscesses (0–1), and the presence or absence of goblet cell depletion (0–1), for a maximum possible score of 11.

In addition to animals raised in the NARF SPF avian facility, a single cohort of IL10KO chicks was hatched in the NARF conventional avian facility and closely monitored as described above. Post-mortem examination of one WT and one HOM IL10KO birds was performed monthly from 10 to 50 weeks, with a particular focus on the gastrointestinal tract. Tissue samples from the proximal colon, duodenum and caecum were collected for histopathological analyses and scored as described above.

Generation of bone marrow-derived macrophage from chicken embryos

Bone marrow cells were isolated from the tibias and femurs of Day 18 embryos (by pooling three embryos of the same genotype per sample) by flushing the bone marrow through a 40 μm cell strainer with RPMI-1640 medium (Sigma R5886) supplemented with 10% fetal bovine serum ultra-low IgG (Gibco 011-90035M), 2 mM L-glutamine (Gibco 25030081) and 0.1X penicillin/streptomycin (Gibco 15140122) using a syringe and blunt needle. Bone marrow cells were washed, resuspended at 1×10⁶ cells/mL in supplemented RPMI-1640 medium, seeded at 1 mL per well in 6-well plates or 150 μL per well in 96-well plates, and cultured for 7 days at 41°C, 5% CO₂ in the presence of 200 ng/mL colony stimulating factor-1 (CSF-1) to induce formation of bone marrow-derived macrophages (BMDM). Cell culture supernatants were subsequently used for capture ELISA and nitric oxide assay, as detailed below.

Detection of IL10 by capture ELISA

On day 7 of culture, BMDMs cultured in 6-well plates were stimulated with 0.5 μg/mL lipopolysaccharide (LPS from E. coli O55:B5; Sigma L2880) for 2 hours at 41°C, 5% CO₂ to induce IL10 expression; supernatants were then harvested and IL10 protein levels were measured by capture ELISA, as described previously (24). Briefly, assay plates were coated with 3 μg/mL capture antibody (ROS-AV164; (24)) incubated overnight at 4°C, washed and blocked. Plates were then incubated with recombinant IL10 standards or 50 μl of 10-fold diluted cell culture supernatants, followed by incubation with 1 μg/mL biotinylated detection antibody (ROS-AV163; (24)) and then incubation with Pierce^TM High Sensitivity Streptavidin-HRP (1:5000; Thermo Fisher Scientific, cat #34028), before adding 1-Step^TM TMB ELISA Substrate Solution (Thermo Fisher Scientific, cat #21130) and then sulfuric acid stop solution. Absorbance was measured at 450 nm (550 nm as a reference) in a SpectraMax 250 microplate spectrophotometer system (Molecular Devices, Sunnyvale, CA, USA). Data were fitted with GraphPad Prism 7.0 using a second-order polynomial (quadratic) model; statistical significance was calculated using two-tailed unpaired t tests.

Nitric oxide assay

On day 7 of culture, BMDMs cultured in 96-well plates were stimulated with 0.5 μg/mL LPS in the presence or absence of 2.5 μg/mL neutralizing anti-IL10 antibody (ROS-AV163; (24)) for 24 hours at 41°C, 5% CO₂; supernatants were then harvested and nitrite levels were measured by Griess assay, according to the manufacturer’s instructions (Promega G2930); absorbance was measured at 550 nm. Statistical significance was calculated using two-tailed unpaired t tests.

Response to vaccination

IL10KO WT and HOM chicks hatched and raised in the NARF conventional avian facility were vaccinated following the routine vaccination schedule in place in the facility (SI Appendix, Table S2). One-mL blood samples were collected from the brachial vein of 29-weeks old hens and sent fresh to Sci-Tech Ireland for antibody titre measurements by ELISA. Statistical significance was calculated using Mann-Whitney U tests.

Microbiome analysis

Caecal contents were harvested from 4-weeks old IL10KO WT and HOM chicks raised in the NARF SPF avian facility. DNA isolation, amplification of the V3-V4 region of the 16S rDNA gene and pair-end sequencing were performed as described previously (31). Raw reads were quality assessed using fastQC, and the DADA2 plugin in QIIME2 was used to generate an ASVs table against the SILVA138 database. Alpha diversity and weighted uniFrac distances were calculated using the Phyloseq package, while Bray-Curtis metrics were generated using the vegan package. PERMANOVA (Bray-Curtis) analysis was conducted to test for significant effects of the group on overall microbiome community composition. LEfSe analysis was performed to identify differentially abundant phyla and genera between both groups using a linear discriminant analysis (LDA) score equal to 4 as a threshold value. P values were declared significant at P < 0.05. The sequencing data have been deposited in the European Nucleotide Archive (ENA) at EMBL-EBI under accession number PRJEB71960.

Bacterial strains

Campylobacter jejuni strain M1 was used as the minimum number of viable bacteria required for reliable intestinal colonisation of chickens has been defined (36). Strain M1 was grown on charcoal cefoperazone deoxycholate agar (CCDA) at 41 °C for 48 hours in a microaerophilic atmosphere (90 % N₂, 5 % CO₂ and 5 % O₂). For challenge, 10 mL of Mueller Hinton broth was inoculated with single colonies and incubated for 16 hours with shaking at 41 °C in the same atmosphere. Prior to inoculation, bacteria were confirmed to have spiral morphology and to be motile by phase contrast microscopy. Cultures were diluted in phosphate-buffered saline (PBS) based on a standard curve of colony-forming units (CFU/mL) relative to optical density at 600 nm and inocula were retrospectively confirmed by plating serial ten-fold dilutions onto CCDA. A spontaneous nalidixic acid-resistant mutant of Salmonella enterica serovar Typhimurium strain 4/74 was used, as its colonisation kinetics are well defined in chickens (38). This was cultured on Luria Bertani (LB) agar containing 20 µg/mL nalidixic acid at 37 °C overnight and a single colony was transferred to LB broth with nalidixic acid and incubated at 37 °C with shaking for 20 hours. 100 µL of this culture was used for inoculation and the number of viable bacteria given was determined retrospectively by plating serial ten-fold dilutions onto LB agar.

Experimental infection of chickens

Animal experiments using Campylobacter or Salmonella were conducted at the Moredun Research Institute. Birds were provided with access to water and irradiated feed based on vegetable protein ad libitum. Groups of 20 to 22 birds of each genotype (IL10KO WT, HET, HOM) were wing-tagged for identification and separately housed in biosafety level 2 facilities from day 7 of age in colony cages. At 14 days of age, birds were inoculated by oral gavage with 100 µL of bacterial suspension. In two separate trials with C. jejuni, birds were inoculated with 1×10² CFU of strain M1. In a single trial with S. Typhimurium, birds were inoculated with 1×10⁸ CFU of strain 4/74 nal^R. At 7 days post-inoculation, half of the birds in each group were euthanised by cervical dislocation. The remainder of the birds were euthanised at 14 days post-inoculation. On post-mortem examination, samples from the ileum and the distal end of the caecum were stored in RNAlater® or fixed in 10% neutral-buffered formalin for histopathology. The contents of the caeca were pooled at the bird level, homogenised in PBS, and ten-fold serial dilutions cultured on CCDA at 41 °C in a microaerophilic atmosphere to enumerate CFU/g of C. jejuni or MacConkey agar containing 20 µg/mL nalidixic acid at 37 °C aerobically to enumerate CFU/g of Salmonella.

Histopathology

Formalin-fixed tissue samples were embedded in paraffin, sectioned, mounted on glass slides, and stained by haematoxylin-eosin (HE) staining. Histopathological analyses were performed by a certified veterinary pathologist blind to the genotype of individuals and the inoculum. Samples of ileum and caecum were assessed for the presence of lymphoplasmacytic and heterophil infiltration on a scale of 0 to 5 (where infiltration is 0 normal, 1 slight, 2 mild, 3 moderate, 4 marked or 5 severe).

RNA extraction and multiplex PCR

Caecal tissue samples in RNAlater® were homogenised with a hand-held tissue homogeniser and RNA was extracted using RNeasy kit following the instructions from the manufacturer (QIAGEN). The concentration and integrity of the RNA were respectively analysed using a NanoDrop microvolume spectrophotometer and TapeStation automated electrophoresis system. Reverse transcription, preamplification and high-throughput quantitative PCR (qPCR) for avian immune-related transcripts and reference genes were performed as previously described using a 96.96 Dynamic Array IFC for Gene Expression ((37); Standard BioTools, San Francisco, CA, United States).

Parasites

Animal experiments using Eimeria were conducted at the Royal Veterinary College, using chicks hatched at the NARF SPF avian facility. All chickens were provided with access to water and ammonia fumigated feed based on vegetable protein ad libitum. The Eimeria tenella Houghton (H) reference strain was used (63). Parasites were passaged by oral inoculation of specified pathogen-free Lohmann Valo chickens, purified and prepared for use as described elsewhere (50). At 21 days post-hatch, chicks were weighed and infected by oral inoculation with 7000 sporulated E. tenella oocysts or mock-infected using sterile water. Group sizes (minimum eight to ten per group) were determined following dose titration in IL10KO and IL10EnKO HET chickens using power calculations with alpha set at 0.05 and power at 80%. At the desired timepoint, birds were culled and whole blood was collected for serum. Caeca were collected, scored for severity of infection (64) and preserved using RNAlater (Invitrogen). Total genomic DNA (gDNA) was extracted from thawed complete caeca stored in RNAlater from all studies. Caeca were homogenised in Buffer ATL using a TissueRuptor homogenizer (Qiagen, Hilden, Germany) and then digested overnight at 56 °C in Buffer ATL and proteinase K, prior to extraction using the Qiagen DNeasy Blood and Tissue DNA Kit according to manufacturer’s instructions.

Quantitative PCR for assessment of E. tenella genome copy number was performed as previously described (65). Briefly, gDNA extracted from caecal tissue was used as template for qPCR targeting E. tenella (RAPD-SCAR marker Tn-E03-116, primers F: 5’-TCGTCTTTGGCTGGCTATTC-3’, R: 5’-CAGAGAGTCGCCGTCACAGT-3’), normalised against the number of chicken tata-binding protein (TBP) genomic copies detected (F: 5’-TAGCCCGATGATGCCGTAT-3’, R: 5’-GTTCCCTGTGTCGCTTGC-3’). Quantitative PCR was performed in 20 μL reactions in triplicate containing 10 μL 2 × SsoFast EvaGreen Supermix (Bio-Rad, Hercules, CA, USA), 1 μL of primers (F: 3 μM, R: 3 μM), 8 μL of molecular biological grade water (Invitrogen) and 1 μL of gDNA, or water as negative control. Hard-shelled 96-well reaction plates (Bio-Rad) were sealed with adhesive film (Bio-Rad) and loaded into a Bio-Rad CFX qPCR cycler. Reactions were heated to 95°C for 2 min, prior to 40 cycles consisting of 95°C for 15 s then 60°C for 30 s with a fluorescence reading taken after each cycle. Melting curve analysis was performed for 15 s at 95°C, before cooling to 65°C for 60 s, then heating to 95°C in 0.5°C increments for 0.5 s. Absolute quantification was performed against a standard curve generated using serially diluted plasmid DNA containing the amplicon of interest (EtenSCAR or ChickenTBP), to generate a standard curve ranging from 10⁶ to 10¹ genome copies per mL. Parasite genome copy number was normalised by division with host (chicken) genome copy number.

Statistical analyses

Minitab Statistical Software was used to analyse IL10 protein levels and nitric oxide production in BMDMs using two-tailed unpaired t tests, growth curves were analysed using two-tailed unpaired t tests or one-way ANOVA with Bonferroni multiple comparison tests, and histopathological scores were analysed using Mann-Whitney U tests or Kruskal-Wallis tests. Analysis of bacterial counts, histopathology, and heatmaps was performed with GraphPad Prism software version 10.2.0. Data from multiplex RT-qPCR was analysed with GenEx6 software. The distribution of data was analysed by D’Agostino-Pearson and Shapiro-Wilk normality tests. Eimeria replication and chicken body weight gain during parasite infection were analysed using one-way ANOVA with Tukey’s multiple comparison tests, and differences in lesion scores were tested by Kruskal-Wallis with Dunn’s multiple comparison tests. Bacterial counts and histopathology scores were analysed by Mann-Whitney U tests or Kruskal-Wallis tests. Fold change in transcript abundance was analysed by t tests using normalised data.

Data Availability Statement

All study data are included in the article and/or SI Appendix. IL10KO and IL10EnKO edited PGCs (HET and HOM) are cryopreserved and available to researchers under MTA.

Acknowledgements

We thank staff at the National Avian Research Facility of the Roslin Institute, at the Moredun Research Institute, and at the Biological Services Unit of the Royal Veterinary College, for care and maintenance of the chickens used in this study.

Additional information

Funding

This research was enabled by responsive mode funding from the Biotechnology & Biological Sciences Research Council (BBSRC; grant references BB/P021638/1 (to DPB & FT) and BB/P022049/1 (to AP, DH, KB, KW, MJM & MPS). The work was also supported by Cobb Vantress-LLC (MJM, DM and KW; grant reference 13393976), a Roslin Institute Early Career Researcher Grant (DM; grant reference 13302777) and by BBSRC Institute Strategic Programme grants (grant references BBS/E/D/20002174 and BBS/E/RL/230002B).

Author Contributions

Designed research: AP, DPB, DAH, DM, FT, KB, KW, MJM, MPS, RCT

Performed research: AAA, DPB, DM, GSA, JJO, KB, LG, LT, MV, RCT, ZW

Contributed new reagents/analytic tools: MJM

Analysed data: AAA, DB, DPB, DAH, DM, JDP, LG, LV, MJM, MPS, RCT, ZW

Wrote the paper: DPB, DAH, DM, FT, MJM, MPS, RCT

Secured funding: AP, DPB, DAH, DM, FT, KB, KW, MJM, MPS

Additional files

Supporting Information Appendix