Chloride channel accessory 1 gene deficiency causes selective loss of mucus production in a new pig model

Abstract

Morbidity and mortality of respiratory diseases are linked to airway obstruction by mucus but there are still no specific, safe, and effective drugs to correct this phenotype. The need for better treatment requires a new understanding of the basis for mucus production. In that regard, studies of human airway epithelial cells in primary culture show that a mucin granule constituent known as chloride channel accessory 1 (CLCA1) is required for inducible expression of the inflammatory mucin MUC5AC in response to potent type 2 cytokines. However, it remained uncertain whether CLCLA1 is necessary for mucus production in vivo. Conventional approaches to functional biology using targeted gene knockout were difficult due to the functional redundancy of additional Clca genes in mice not found in humans. We reasoned that CLCA1 function might be better addressed in pigs that maintain the same four-member CLCA gene locus and the corresponding mucosal and submucosal populations of mucous cells found in humans. Here we develop to our knowledge the first CLCA1-gene-deficient (CLCA1^–/–) pig and show that these animals exhibit loss of MUC5AC⁺ mucous cells throughout the airway mucosa of the lung without affecting comparable cells in the tracheal mucosa or MUC5B⁺ mucous cells in submucosal glands. Similarly, CLCA1^–/– pigs exhibit loss of MUC5AC⁺ mucous cells in the intestinal mucosa without affecting MUC2⁺ mucous cells. These data establish CLCA1 function for controlling MUC5AC expression as a marker of mucus production and provide a new animal model to study mucus production at respiratory and intestinal sites.

INTRODUCTION

Chronic lower respiratory diseases primarily in the form of asthma and chronic obstructive pulmonary disease (COPD) are the fourth leading cause of death due to disease in the United States (1). The lung disease phenotype most closely linked to mortality is airway obstruction with mucus (2–4). Despite this connection, there is still an unmet need for a specific and effective method to correct this phenotype. This advance will require a better understanding of the control of mucus production in the airway and other mucosal sites. Similarly, there is still a need for the development of animal models with high fidelity to humans to allow for defining control of mucus production and testing possible therapeutic strategies.

A candidate pathway for controlling mucus production derives from the potent effects of type 2 cytokine IL-13 for stimulating mucus production in vivo in mouse models of postviral lung disease (5–8) and in cell culture with human airway epithelial cells (9). Some details of that pathway are defined, such as the transition from basal epithelial stem cells (basal-ESCs) to mucous cells based on IL-13R–STAT6 signaling (5, 9, 10). However, other key components are still uncertain. Our previous work indicated that a mucin granule protein known as chloride channel accessory 1 (CLCA1) is also upregulated in a postviral lung disease model (11, 12). Similarly, we found CLCA1 induction in human airway epithelial cells in response to IL-13R signaling to MAPK13 activation and in turn MUC5AC mucin expression (9). In addition, CLCA1 levels are increased in concert with IL-13, MAPK13, and MUC5AC biomarkers in clinical samples taken from the airway lumen and lung tissue of patients with asthma and COPD (9, 13). However, validation of CLCA function in vivo remains uncertain because of the need for an animal model that closely resembles human mucous cell architecture and signaling. In particular, genetic approaches using mouse models show that Clca1 gene expression is sufficient for MUC5AC⁺ mucus production but whether Clca1 is required is uncertain because of the functional redundancy of Clca genes in mice that are pseudogenes in humans (11, 12, 14, 15). Moreover, mice do not preserve baseline levels of airway mucous cells or submucosal gland structures as found in humans. Thus, a better experimental model and a modified approach are needed.

Here we reasoned that a study of pigs might provide a superior approach to defining the control of mucus production. In contrast to mice, there is the same lung anatomy (16) and airway mucosal and submucous gland morphology in the pig. Moreover, recent annotation of the pig genome shows that the CLCA genomic locus is similar to humans with the same nonfunctional CLCA3 pseudogene that limits the possibility of compensation in pigs in the absence of CLCA1 (12, 14, 15, 17). Similarly, Clca2 is expressed in bronchial epithelial cells in mice but the homologous CLCA2 is not found in humans and pigs (18). Furthermore, Clca4a is not found in mice lungs but the homologous CLCA4 is similarly expressed in human and pig lungs (19). Notably, CLCA1 is expressed similarly at baseline in pigs and humans (20). Accordingly, we evaluated the utility of a pig model beginning with studies at baseline. We generated to our knowledge the first CLCA1 gene knockout (CLCA1^–/–) pigs and defined the expression of key mucin genes in the airway and intestinal mucosal sites based on mRNA and protein expression levels. Comparison to heterozygous (CLCA1^+/–) and wild-type (CLCA^+/+) littermate controls established an absolute albeit selective functional role for CLCA1 in controlling MUC5AC⁺ mucous cells in the pulmonary airway and intestine that adds significantly to our understanding of the basis for mucus production. The study also establishes a new animal model for future studies of the critical issue of mucous cell biology in respiratory and gastrointestinal immunity and inflammatory disease.

MATERIALS AND METHODS

Generation of Gene-Knockout Pigs

We generated CLCA1 gene knockouts in domestic pigs using CRISPR-Cas9 technology as described previously (21) and modified them for the present experiments using guide RNAs that were designed to target regions in exon 1 of CLCA1. The DNA sequence was first analyzed using RepeatMasker software (v. open-4.0.9; https://www.repeatmasker.org) from A.F.A. Smit, R. Hubley, and P. Green (University of Missouri) to hide repetitive elements from being used as potential targets in the guide design. Two 20-bp guides were designed to target the start codon and downstream of the start codon. These guides were analyzed for optimal folding by using mFold (http://www.unafold.org). The two CLCA1 guides with the protospacer adjacent motif (PAM) indicated in bold type were GAGGGAGATGCACAGCCATG AGG and GTTTCATCTTCTGGCACGTT GGG. The distance between each guide was 120 bp.

A T7 promoter sequence was added upstream of the 20-bp guide and a gBlockgene fragment was synthesized to use as template DNA as described previously (22). Each gBlock gene fragment was diluted to a final concentration of 0.1 ng/µL and was PCR amplified with a forward primer ( ACTGGCACCTATGCGGGACGAC) and reverse primer ( AAAAGCACCGACTCGGTGCCAC) for use with a Q5 site-directed mutagenesis kit (New England Biolabs, Ipswich, MA). The PCR conditions were set at an initial denaturation of 98°C for 1 min followed by 35 cycles of 98°C (10 s), 68°C (30 s), and 72°C (30 s). Each gBlock amplimer was then purified using a Qiagen PCR purification kit (Valencia, CA). Purified gBlock amplimers were then used as a template for in vitro transcription with the MEGAshortscript T7 kit (Invitrogen, Thermo Fisher Scientific). The resulting RNA was purified using the MEGAclear Transcription Clean-Up Kit (Invitrogen, Thermo Fisher Scientific). The Cas9 RNA (TriLink Biotechnologies, San Diego, CA) the two CLCA1 guide RNAs were combined at final concentrations of 20 ng/µL and 10 ng/µL each, respectively.

Ovaries from prepubertal gilts were obtained from a local abattoir (Smithfield Foods, Milan, MO). Cumulus oocyte complexes were aspirated from follicles and subjected to in vitro maturation as described previously (23). Mature oocytes were selected and in vitro fertilized in a similar manner as described previously (24). The guide RNA and Cas9 RNA were coinjected into the cytoplasm of presumptive zygotes using a FemtoJet microinjector (Eppendorf, Hamburg, Germany). Injected zygotes were then transferred into MU2 medium for 5 days after fertilization. At that time, blastocysts and morula were selected for transfer into a recipient sow. Four embryo transfers were completed and resulted in three litters of piglets.

Gene deletions and insertions were determined using PCR-based amplification of CLCA1 sequence flanking a projected DNA-cutting site introduced by the CRISPR/Cas9 system. The forward primer ( TTGTCACCTGTGAGTCCTTTC) and reverse primer ( CCTCCATCCTGCAATAGTTGAT) were designed to amplify an 1,107-bp wild-type amplicon. The PCR assay conditions were set at an initial denaturation of 94°C (60 s) followed by 35 cycles of 94°C (30 s), 56°C (30 s), 68°C (2 min), and finally 68°C (2 min). For genotyping piglets after birth, the resulting PCR-based assay products were purified and then sequenced. The PCR product was also TOPO cloned into the pCR4 vector (Invitrogen) and transformed into chemically competent One Shot TOP10 E. coli (Invitrogen). Individual colonies were selected and sequenced to confirm genotypes. Because zygote-injected embryos can produce pigs that contain a mosaic genotype (21), we selected specific heterozygous null mating pairs to guarantee only two alleles would be present. A sow with a 125-bp deletion on one allele and intact CLCA1 in the second allele was crossed with a boar containing an 11-bp deletion and 195-bp insertion in a single allele to get two litters of F1 piglets. All experimental procedures involving animals were approved by the Animal Studies Committees of the University of Missouri in accordance with guidelines from the National Institutes of Health.

We recognize that the annotation of the pig genome build is not as robust as for mouse, human, or cattle. However, for off-site cleavage concerns, the completeness of the genome build is more important than the annotation. In that regard, the content of the pig genome is nearly and satisfactorily complete even though not all polymorphisms are known. Similarly, the degree of identity between the present genetics and the swine reference genome is not defined. Given these conditions, our methods still show that all known potential secondary targets in the pig genome based on homology (sequence identity) exceed three mismatches with our guide RNAs (gRNAs). In addition, any off-target effects would appear as random effects in animals produced by breeding because these effects would segregate independently from the present modification. The exception would be in the case of a closely linked unintended target. Thus, as part of the gRNA selection criteria, we screened for unintended targets and excluded any gRNA that had a potential target with 0, 1, or 2 mismatches and the presence of a PAM (protospacer adjacent motif). If we consider our selected gRNAs and do not require a PAM to identify potential unintended targets, the selected gRNAs have no known targets that have 0 or 1 mismatch. There are a few known regions (1 for one gRNA, 4 for the other gRNA) that have two mismatches. However, these regions of homology harbor the differences in the seed region (generally intolerant of mismatches) and do not have a PAM. In addition, these five regions appear to be intergenic. Given our stringent selection criteria for gRNAs and the absence of a segregating phenotype, we therefore conclude that unintended target mutations are not involved with the present CLCA1 genetic model.

Necropsy and Histology

A full necropsy was performed on six F1 pigs with careful examination of any organs containing mucosal epithelium. All six pigs were from the same litter and were selected to include a male and female from each CLCA1 genotype: CLCA1^–/– (n = 2), CLCA1^+/– (n = 3), and CLCA1^+/+ (n = 2). Tissue samples were collected from trachea, lung, duodenum, jejunum, and ileum, and colon of necropsy pigs and a second litter of pigs with each CLCA1 genotype and sex represented. The total number of pigs for tissue analysis included CLCA1^–/– (n = 3, 2 males and 1 female), CLCA1^+/– (n = 3, 1 male and 2 females), and CLCA1^+/– (n = 3, 1 male and 2 females) for trachea, duodenum, jejunum, ileum, and colon samples and CLCA1^–/– (n = 5, 2 males and 3 females), CLCA1^+/– (n = 3, 1 male and 2 females), and CLCA1^+/– (n = 5, 2 males and 3 females) for lung samples. All tissue samples were fixed with 10% formalin, embedded in paraffin, cut into 5-μm sections, and adhered to charged slides for staining as described previously (10). Conventional tissue staining was performed using Periodic acid-Schiff (PAS) and hematoxylin. Immunostaining was performed after sections were prepared using Fisherbrand CitroSolv (Thermo Fisher Scientific) for deparaffinization followed by rehydration and heat-induced antigen retrieval in citric acid-based solution (Vector Laboratories, Inc). Immunostaining was performed with the following primary antibodies: biotin-labeled mouse anti-MUC5AC (Invitrogen MA5-12175), rabbit anti-MUC5B (Abcam ab87276), rabbit anti-MUC2 (Abcam ab134119), and rabbit anti-CLCA1 antibody described previously (9). Primary antibodies were selected based on specificity at a working concentration of 2 µg/mL as described previously (9, 25–27). Primary antibodies were detected with secondary antibodies labeled with Alexa Fluor 488 (Thermo Fisher Scientific) or Alexa Fluor 555 (Thermo Fisher Scientific) at a concentration of 1 µg/mL followed by DAPI counterstaining. Stained sections were imaged by light microscopy using a Leica DM6B equipped with a Leica DFC7000T camera and immunofluorescent microscopy using an Olympus BX51. In addition, whole slides were scanned using a NanoZoomer S60 slide scanner (Hamamatsu), and staining was quantified by ImageJ software.

RNA Analysis

Lung samples were collected from CLCA1^–/– (n = 5, 2 males and 3 females), CLCA1^+/– (n = 3, 1 male and 2 females), and CLCA1^+/– (n = 5, 2 males and 3 females) pigs, and flash-frozen in liquid nitrogen for storage at −80°C. Stored samples were homogenized by bead-beating, and RNA was purified using Trizol (Invitrogen) and used to generate cDNA with the high-capacity cDNA Archive Kit (Life Technologies). Target mRNA was quantified using the ABI 7500 real-time PCR system (Applied Biosystems) and pig-specific real-time qPCR assays with fluorogenic probe-primer combinations and Fast Universal PCR Master Mix (Applied Biosystems). Forward and reverse primers and probes for the qPCR-based assays were: CLCA1, CTGACGTGGACGGCTCCTGGGGAT, GAACTTGTCTCTGAGATCAAGAATATTTGTGCT, and TTACGACCACGGAAGAGCTGACAGGT; MUC5AC, CGTAGAGCACAGGTGCAAGT, GCAGGGTCACGTTTCTCAG, and probe No. 17 (Roche Universal Probe Library); MUC5B, CAGCCAACAACTGCACAGA, CAGCCAACAACTGCACAGA, and probe No. 3 (Roche Universal Probe Library); and GAPDH, CCCCGCGATCTAATGTTCT, CTTCACCATCGTGTCTCAGG, and probe No. 6 (Roche Universal Probe Library). All samples were run with internal standard plasmids and thereby quantified for copy number per copy of GAPDH.

Statistical Analysis and Reproducibility

All data presented in bar-graph format were expressed as means ± SE. For this data, statistical differences between means for sample conditions were assessed using one-way analysis of variance (ANOVA) with Tukey’s correction for multiple comparisons. For all data, the significance threshold was set at P < 0.05. The number of pigs for each condition is defined in the figure legends.

RESULTS

CLCA1 Genomic Structure and Gene Expression in Mice, Humans, and Pigs

As a first step in choosing a higher-fidelity animal model, we reexamined genomic concordance, lung morphology, and gene expression in pigs compared with humans and mice. As first noted in the Introduction, CLCA genomic organization was distinct in mice and similar between humans and pigs (Fig. 1A). In addition, healthy mice (C56Bl/6J strain) showed only submucosal mucous glands without mucosal mucous cells in tracheal sections and lacked both of these mucus-producing sources in lung sections based on PAS⁺ staining (Fig. 1B). In contrast, constitutive mucus production at both mucosal and submucosal locations was detectable in humans even without lung disease in a pattern that was also found in healthy domestic pigs (Fig. 1B). Similarly, Clca1 or Muc5ac expression is not found by immunostaining in tracheal or lung sections from healthy mice (Fig. 1C, top row), recognizing that this method is satisfactory to detect Clca1 and Muc5ac induction during inflammatory conditions such as respiratory viral infection (10, 12, 28). Here again distinct from the findings in mice, CLCA1 was abundantly expressed in mucous cells in the mucosal epithelium and submucosal glands of both healthy humans and pigs at baseline (Fig. 1C, top row). Moreover, MUC5AC expression was confined to the mucosal epithelium (colocalized to CLCA1) in the lung in a pattern that was also conserved across humans and pigs (Fig. 1C, top row). In contrast to MUC5AC, lung immunostaining for MUC5B was confined to submucosal glands found in mouse trachea as noted previously (29) and in human and pig lungs (Fig. 1C, bottom row) even under baseline, nondisease conditions. Together, the similarities in CLCA1 genomic structure, lung morphology, and CLCA1 and mucin gene expression in human and pig lung tissue suggested the suitability of pigs as a model to study the regulation of mucus production, and in particular any possible role of CLCA1 in controlling airway mucus levels under baseline conditions.

Figure 1.

Comparison of CLCA gene loci and gene expression patterns in mice, humans, and pigs under baseline conditions. A: scheme for CLCA gene loci in mouse, human, and pig (dashed lines indicate homologous gene groups; white boxes indicate pseudogenes). B: staining with PAS and hematoxylin of tracheal sections from mice and lung sections from mice, humans, and pigs under baseline (nonlung disease) conditions. C: corresponding immunostaining for Ccla1/CLCA1, Muc5ac/MUC5AC, and Muc5b/MUC5B with DAPI counterstaining of tracheal sections from mice and lung sections from mice, humans, and pigs for conditions in B. Results are representative of three to five animals or human subjects per group. CLCA, chloride channel accessory 1; DAPI, 4′,6-diamidino-2-phenylindole; PAS, Periodic acid-Schiff.

Generation and Gross Characterization of CLCA1 Gene-Knockout Pigs

As a next step in studying CLCA1 function, we generated a CLCA1 gene-knockout (CLCA1^–/–) pig by using adapted CRISPR/Cas9 technology (21) and established phenotype when compared with heterozygous (CLCA^+/–) and wild-type (CLCA^+/+) littermate control animals. The CLCA1^–/– pigs were found to develop normally and were uniformly healthy throughout the 2-yr observation period (including the absence of any clinical signs of aspiration or respiratory infection at any age). We also found normal reproduction at least for CLCA1^+/– heterozygotes that were used here for breeding to generate CLCA1^–/– pigs. These data were consistent with the lack of any reported phenotype for genetic deficiency of CLCA1 in humans and normal development and reproduction in Clca1^–/– mice (12). Furthermore, we detected no gross abnormalities of skin or internal organs (including lung and intestine) during complete necropsy of CLCA1^–/–or CLCA1^+/– pigs at 2–4 mo of age and therefore chose this time point for further histological studies as described in the next two sections of the Results.

Effect of CLCA1 Gene-Knockout in Pig Airway Tissues

We defined CLCA1 expression and function in the pulmonary system based on gene expression at mRNA and protein tissue levels. Analysis of lung mRNA by real-time qPCR assay showed the expected decrease in CLCA1 mRNA in CLCA^+/– pigs and the absence of detectable levels in CLCA^–/– pigs compared with wild-type CLCA^+/+ littermate control animals. In concert with CLCA1 mRNA levels, we observed the same pattern of mRNA levels for MUC5AC (Fig. 2A). In contrast, MUC5B mRNA was expressed at markedly lower levels (∼1,000-fold less abundance) in each genotype (Fig. 2A). Together, the results indicated high concordance between CLCA1 and MUC5AC but not MUC5B expression, thereby providing evidence that CLCA1 gene function is required albeit selectively for MUC5AC gene expression at baseline in pigs. These changes in MUC5AC expression were not reflected in staining for PAS⁺ mucin staining in the trachea (Fig. 2, B and C) but were accompanied by significant decreases in PAS⁺ staining in lung tissue (Fig. 2, D and E). Of note, the decrease in CLCA1-dependent PAS⁺ staining was found in the mucosal epithelium but not submucosal gland sites, further indicating that CLCA1 control was selective for specific type and site of mucin gene expression.

Figure 2.

Selective effect of CLCA1-deficiency on mucin gene expression and mucus levels in pig airways. A: levels of indicated mRNAs in lung tissue from CLCA1^–/– and control littermate heterozygous (CLCA1^+/–) and wild-type CLCA1^+/+ pigs. Values represent means ± SE (n = 3–5 per group with samples from both sexes). B: PAS and hematoxylin staining of tracheal sections from CLCA1^+/+, CLCA1^+/–, and CLCA1^–/– pigs. C: quantitation of PAS⁺ mucin staining in B. Values represent means ± SE for three technical replicates from three pigs per group with samples from both sexes. D: PAS and hematoxylin staining of lung sections for conditions in B. E Quantitation of PAS⁺ mucin staining in D. Values represent means ± SE for two technical replicates from three to five pigs per genotype group with samples from both sexes. *P < 0.05 by ANOVA, ns, not significant. CLCA1, chloride channel accessory 1; PAS, Periodic acid-Schiff.

To pursue the issue of selective control, we next assessed protein expression using specific immunostaining of the relevant mucin targets in trachea and lung. Similar to our analysis of gene expression and PAS⁺ staining, we observed levels of CLCA1⁺ immunostaining that correlated with genotype but no significant differences in MUC5AC⁺ or MUC5B⁺ immunostaining across genotypes in the trachea (Fig. 3, A and B). In contrast to this pattern, but again in close agreement with PAS⁺ staining, we also found that CLCA1⁺ expression correlated with corresponding MUC5AC⁺ expression without changing MUC5B⁺ expression based on immunostaining in lung tissue (Fig. 3, C and D). As noted for PAS⁺ staining, CLCA1-dependent expression of MUC5AC was localized to the mucosal epithelium (Fig. 3C) as a basis for the overall decrease in MUC5AC⁺ immunostaining (Fig. 3D). Thus, the decreased PAS⁺ staining observed in the lung of CLCA^+/– and CLCA^–/– pigs was attributed to a decrease in MUC5AC and not MUC5B expression. These findings further support a requirement for CLCA1 function selective for MUC5AC (vs. MUC5B) expression in the lung (vs. the trachea) and thereby for mucosal (but not submucosal) epithelial sites. In addition, CLCA1 regulation of mucin production at baseline in the absence of inflammatory lung disease provides in situ evidence of mucus control under baseline physiological conditions. This fine-tuning is also supported by the significant decrease in MUC5AC levels even with heterozygous conditions for the CLCA1 genotype.

Figure 3.

Selective effect of CLCA1-deficiency on MUC5AC expression level in pig airways. A: immunofluorescent staining for CLCA1, MUC5AC, and MUC5B with DAPI counterstaining in tracheal sections from CLCA1^+/+, CLCA1^+/–. and CLCA1^–/– pigs. B: quantitation of immunostaining in A. C: immunofluorescent staining of lung sections for conditions in A. D: quantitation of immunostaining in C. For B, D, values represent means ± SE for three technical replicates from three pigs per genotype group with samples from both sexes. *P < 0.05 by ANOVA, ns, not significant. CLCA1, chloride channel accessory 1; DAPI, 4′,6-diamidino-2-phenylindole.

Effect of CLCA1 Gene-Knockout in Pig Intestinal Tissues

To ascertain any role of CLCA1 in the control of mucus production at other tissue sites, we also applied our analysis of PAS staining and mucin immunostaining to samples from the intestinal epithelium. For PAS⁺ staining, we observed no change in the duodenum but significant decreases in the jejunum, ileum, and colon in CLCA^–/– pigs and to a slightly lesser extent in CLCA^+/– pigs compared with wild-type controls (Fig. 4, A and B). This pattern of expression suggested increasing CLCA1 control over at least some component of mucin gene expression and mucus production in the intestinal epithelium.

Figure 4.

Selective effect of CLCA1-deficiency on mucus levels in pig intestines. A: PAS and hematoxylin staining of intestine (duodenum, jejunum, ileum, and distal colon) from CLCA1^+/+, CLCA1^+/–, and CLCA1^–/–pigs. B: quantitation of PAS⁺ mucin staining in A. Values represent means ± SE for two technical replicates from three pigs per genotype group with samples from both sexes. *P < 0.05 by ANOVA, ns, not significant. CLCA1, chloride channel accessory 1; DAPI, 4′,6-diamidino-2-phenylindole; PAS, Periodic acid-Schiff.

Analysis of immunostaining showed that CLCA1 expression was found at all intestinal mucosal sites in proportion to CLCA1 genotype, i.e., level in CLCA1^–/– < CLCA1^+/– < CLCA1^+/+ (Fig. 5, A and B). The same pattern of expression was found for MUC5AC⁺ immunostaining, whereas the predominant intestinal mucin MUC2 (30) was unchanged by CLCA1 genotype (Fig. 5, A and B). This pattern was preserved across intestinal sites with relatively low immunostaining (e.g., jejunum) and high immunostaining (e.g., colon) for CLCA1 (Fig. 5, A and B). Immunostaining for MUC5B was barely detectable at 10- to 100-fold lower levels than MUC5AC (data not shown). Together, these observations indicated that CLCA1 controls MUC5AC expression in the intestine similar to the lung under baseline conditions. Moreover, the lack of CLCA1 regulation of MUC2 expression further highlights the selective function of CLCA1 for mucin-type, in this case affecting MUC5AC under baseline conditions in addition to the conventional connection to type 2 immunity and inflammation.

Figure 5.

Selective effect of CLCA1-deficiency on MUC5AC expression level in pig intestines. A: immunofluorescent staining for CLCA1, MUC5AC, and MUC2 with DAPI counterstaining of intestine (duodenum, jejunum, ileum, and distal colon) from CLCA1^+/+, CLCA1^+/–. and CLCA1^–/– pigs. B: quantitation of immunostaining in A. Values represent means ± SE for two technical replicates from three pigs per genotype group. *P < 0.05 by ANOVA, ns, not significant. CLCA1, chloride channel accessory 1; DAPI, 4′,6-diamidino-2-phenylindole.

DISCUSSION

This study provides to our knowledge the first evidence that CLCA1 gene expression is required for MUC5AC mucin expression and mucus production in vivo at mucosal epithelial sites in the lung and intestine. This conclusion is made possible by several advances: 1) development of a CLCA1 gene-knockout model in pigs to better recapitulate the tissue morphology and CLCA1 genomic structure found in humans versus previously and more generally used mouse genetic models (12, 31); 2) quantitative morphology to define significant decreases in PAS⁺ mucin staining and MUC5AC⁺ mucous cell immunostaining in the mucosal epithelium of the pulmonary airways and intestines in CLCA1^–/– pigs; 3) extension of this precise histological approach to demonstrate preservation of MUC5B expression in lung and MUC2 expression in the intestine as a sign of selective CLCA1 control of MUC5AC⁺ mucin regulation and consequent mucus production; and 4) demonstration of CLCA1 function under baseline conditions thereby defining the mucus regulation paradigm in homeostatic conditions versus previous studies of inflammatory cytokine-driven conditions (9, 12). Together, the findings significantly extend understanding of CLCA1 gene function and consequent control of mucus production in a model that more closely resembles humans. The data thereby provide a new scheme (Fig. 6) and experimental model for further studies of human mucus production and possible attenuation with therapeutics targeted to the CLCA1-regulated pathway. Here we develop each of these points and summarize existing and future strategies.

Figure 6.

Scheme for CLCA1-selective control of mucous cells. A: scheme for CLCA1 requirement for development of MUC5AC⁺ mucous cells in relation to CLCA1 genotype. B: corresponding scheme for lack of CLCA1 requirement for the generation of MUC5B⁺ and MUC2⁺ mucous cells for A. CLCA1, chloride channel accessory 1.

First, we underline the strategy that we took to achieve close correspondence between pig and human CLCA1 genomics and gene expression as noted previously (11, 14, 20). Here we also find that CLCA1 is colocalized to MUC5AC⁺ mucous granules in vivo similar to the case in cultured human airway epithelial cells in vitro based on confocal microscopy (9). This pattern of coexpression is evident in the mucosal epithelium in the respiratory system at tracheal and lung sites. However, CLCA1 influence on MUC5AC expression level is restricted to the lung site with no significant effect at the tracheal site at least under baseline homeostatic conditions. The present findings rule against an absolute requirement of CLCA1 for mucin granule formation based on intact tracheal mucous cell formation in submucosal glands in the absence of CLCA1 expression. However, further studies of mucin regulation and mucus production will be required to better define the basis for CLCA1-dependent control. Of note, we also obtained evidence for CLCA1-independent expression of MUC5B levels that is found primarily in submucosal glands in the respiratory airway tissue. Thus, additional studies of signal transduction are also needed to determine the differential regulation of MUC5B under baseline conditions. In each case, it will be necessary to develop new methods and reagents for studies of porcine cell and molecular biology.

Second, we also established that CLCA1 gene deficiency did not interfere with the normal development of the organism based on observations of pigs housed under clean conditions up to 2 yr of age. A complete necropsy established that all organs and tissues were normal in appearance, including epithelium at all mucosal surfaces at 2–4 mo of age that was used for more extensive histology. We also found no abnormalities in reproductive capacity based on observations of heterozygotes that exhibit an intermediate decrease in CLCA1-dependent expression of MUC5AC⁺ mucous cells. However. we did not yet study reproductive function in homozygotes. Thus, additional studies of homozygotes under baseline and stressed conditions will be needed to determine whether CLCA1 function can influence reproduction. Similarly, a comprehensive study of CLCA1 expression and function during normal and stressed development will be useful to define any influence of CLCA1 during infancy, childhood, and adulthood.

Third, we highlight the present evidence for CLCA1 function in pigs versus previous data for lack of such function in mice. As first noted in the Introduction, we propose that Clca gene redundancy in mice is a suitable explanation for the species difference. However, we also note that mice do not manifest significant numbers of mucous cells including MUC5AC⁺ mucous cells in the lungs under baseline conditions. Thus, we cannot properly assess a baseline phenotype in mice. As a consequence, studies of airway mucous cells in mice are generally performed under conditions where there is stimulated production of pro-mucus cytokines, particularly potent type 2 cytokines IL-4 and IL-13 (6, 12, 32). Thus, further studies of pigs will be needed to define the role of CLCA1 under cytokine-stimulation conditions. Nonetheless, studies in vitro already indicate that CLCA1 expression is highly inducible with IL-13 stimulation and is required for consequent MUC5AC and mucus formation and secretion in primary culture hTECs (9). Our results extend this work by demonstrating that CLCA1 not only regulates IL-13-induced MUC5AC but also the baseline expression of this mucin under homeostatic conditions. Relevant to this point, CLCA1 regulation of MUC5AC and mucus levels found in the intrapulmonary (bronchial) airways were not present in the trachea of pigs at baseline. Further study of mucus regulation will need to also address any differences in tracheal versus bronchial epithelial cells for possible regional differences in the epithelial stem cell to mucous cell axis. Similarly, additional studies will need to define CLCA1-independent factors for the residual level of MUC5AC expression as found at tracheal sites.

Fourth, we point out that our study focused primarily on the respiratory system, but we also performed a comparative analysis of CLCA1 control of mucus production in the developmentally related site in the intestines. Previous study of Clca1^–/– mice concluded that Clca1 had no functional role in regulating mucin expression in the intestines either at baseline or in a mouse model of colitis (31). In contrast, we observed MUC5AC and PAS⁺ mucus levels were concordant with CLCA1 expression in CLCA1^+/– and CLCA1^–/– pigs. We also found no significant change in the predominant intestinal mucin MUC2 across CLCA1 genotypes. This finding is similar to the selectivity of CLCA1 regulation of MUC5AC but not MUC5B in the pulmonary airways as well as the difference between mice and pig CLCA1 function. Thus, functional redundancy of Clca1 genes and proteins in mice can explain the lack of phenotype in both lung and intestine in that species (31) versus the present CLCA1^–/– phenotype found in pigs at least under baseline conditions. In that regard, intestinal MUC2 expression is variable whereas MUC5AC expression is generally increased in patients with inflammatory bowel disease (33–37). Whether MUC5AC functions in a protective or pathological capacity in IBD remains uncertain based on studies of mouse models (38). The present results suggest that well-characterized pig models of inflammatory bowel disease and related conditions (39, 40) would inform the role of CLCA1 regulation of mucus production with perhaps higher fidelity to humans.

In summary, this study provides a new animal model suitable for studying the role of CLCA1 in regulating MUC5AC⁺ mucus production comparable with humans. Some future areas for study include: 1) characterization of CLCA1 function at other mucosal sites, including the GU system, and any implications for reproductive capacity in homozygous knockout animals; 2) the impact of CLCA1 signaling function during immune and inflammatory conditions given the implications for inflammatory airway and bowel disease; and 3) the development of therapeutics that target CLCA1 function as a means of regulating baseline and inducible mucus production. There is also a possible role for CLCA1 in regulating chloride ion transport during mucus production and perhaps compensating for this function in ion transport disorders such as cystic fibrosis (41–44). Other studies suggest a role for CLCA1 as a proinflammatory signaling protein (45). Indeed, the present study highlights the role of CLCA1 in controlling mucus production under baseline homeostatic conditions in addition to inflammatory conditions described previously (9, 12, 28). Defining these controls will be critical to designing therapeutics to downregulate excess mucus production to physiological levels. The CLCA1-deficient pigs developed here should provide a key tool for addressing these questions in the control of mucus production at baseline and disease at pulmonary, intestinal, and other mucosal sites using this new resource for studies in vivo or for cells and organoids derived from this resource for studies in vitro.

DATA AVAILABILITY

The data that support the findings of this study will be made available upon reasonable request from the corresponding author.

GRANTS

This work was supported by the National Institute of Allergy and Infectious Diseases Grant R01 AI130591, National Heart, Lung, and Blood Institute Grant R35 HL145242, Department of Defense (PR190726), the Cystic Fibrosis Foundation, the Hardy Trust, and the Schaeffer Fund. The National Swine Resource and Research Center (to R.S.P. and K.D.W.) was supported by the National Institute of Allergy and Infectious Disease, the National Institute of Heart, Lung and Blood, and the Office of the Director (U42OD011140).

DISCLOSURES

M.J.H. is Founder and President of NuPeak Therapeutics and is a member of the Data Safety Monitoring Board for AstraZeneca. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

S.P.K., J.Y., Y.Z., K.M.W., R.S.P., and M.J.H. conceived and designed research; S.P.K., J.Y., S.L.Y., S.P., D.M., Y.Z., K.M.W., B.R., M.S.S., and K.D.W. performed experiments; S.P.K., S.L.Y., S.P., K.M.W., R.S.P., and M.J.H. analyzed data; S.P.K., B.J.G, R.S.P., and M.J.H. interpreted results of experiments; S.P.K. and M.J.H. prepared figures; S.P.K. and M.J.H. drafted manuscript; R.S.P. and M.J.H. edited and revised manuscript; S.P.K., J.Y., B.J.G., S.L.Y., B.R., M.S.S., K.D.W., R.S.P., and M.J.H. approved final version of manuscript.