The Atg16l1 gene: characterization of wild type, knock-in, and knock-out phenotypes in rats

Abstract

ATG16L1 is a ubiquitous autophagy gene responsible, in part, for formation of the double-membrane bound autophagosome that delivers unwanted cellular debris and intracellular pathogens to the lysosome for degradation. A single, nonsynonymous adenine to guanine polymorphism resulting in a threonine to alanine amino acid substitution (T300A) directly preceded by a caspase cleavage site (DxxD) causes an increased susceptibility to Crohn's disease (CD) in humans. The mechanism behind this increased susceptibility is still being elucidated, however, the amino acid change caused by this point mutation results in increased ATG16L1 protein sensitivity to caspase 3-mediated cleavage. To generate novel rat strains carrying genetic alterations in the rat Atg16l1 gene, we first characterized the wild-type rat gene. We identified four alternative splice variants with tissue-specific expression. Using CRISPR-Cas9 genome editing technology, we developed a knock-in rat model for the human ATG16L1 T300A CD risk polymorphism, as well as a knock-out rat model to evaluate the role of Atg16l1 in autophagy as well as its potential effect on CD susceptibility. These are the first reported rat strains with alterations of the Atg16l1 gene. Consistent with studies of the effects of human ATG16L1 polymorphisms, models exhibit morphological abnormalities in both Paneth and goblet cells, but do not develop spontaneous intestinal permeability or inflammatory bowel disease. Analysis of the gut microbiota does not show inherent differences in bacterial composition between wild-type and genetically modified animals. These Atg16l1 strains are valuable new animal models for the study of both autophagy and CD susceptibility.

Keywords: CRISPR; Crohn's disease; genetic models; inflammatory bowel disease; phenotyping.

Figure 1.

Mutational analysis of the genetic alterations in Atg16l1 in the original em2 and T300A rat strains. The nucleotide sequence of each mutant Atg16l1 allele was compared to the wild type Atg16l1 sequence. A: the rat Atg16l1 gene has 19 exons. The T300A allele has a single basepair change (A to G) in exon 10 (arrow). B: The em2 allele has a 709 bp deletion spanning exons 5 through 14.

Figure 2.

Comparison of human, rat, and mouse Atg16l1. A: comparison of genomic organization of the Atg16l1 genes. Blue areas represent exons; white areas represent introns. Adapted schematics of Atg16l1 for human (Uniprot; E7EVC7), rat (Uniprot; D3ZFK6), and mouse (Uniprot; G9M4M6). All 19 exons are represented in the human, mouse, and rat full-length transcripts. B: comparison of homology at the DNA level and identity at the protein level for rat and mouse compared to human (top) and mouse compared to rat (bottom).

Figure 3.

Wild type (WT) splice variant DNA and protein analysis. A: representative schematic of the four WT splice variants of rat Atg16l1. Gray boxes designate exons. B: representative agarose gel images depicting splice variants detected in different tissues. C: summary of WT splice variants detected in various tissues. D: representative schematic of predicted protein isoform for each splice variant. Dotted lines denote the area of each protein isoform missing relative to the full-length protein (Atg16l1-A). shaded box, ATG5 binding motif; diagonal hashmark box, coiled coil domain; vertical hashline box, WD40 repeat domains (7 total). E: representative Western blot image showing protein expression when DNA constructs coding for each splice variant were transfected into HEK293 cells. A, Atg16l1-A; B, Atg16l1-B; C, Atg16l1-C; D, Atg16l1-D; E, nontransfected HEK293 cells (negative control). Ladder: Precision Plus Protein Kaleidoscope Prestained Protein Standards (1610375; BioRad); Loading control: β-actin.

Figure 4.

Fluorescein isothiocyanate (FITC) fluorescence for wild-type (WT) and heterozygous (HET) em2 (A) and T300A (B) rat strains. Average FITC fluorescence in relative fluorescence units (RFU) ± standard deviation. M, male; F, female; WT, wild type; HET, heterozygous; n = 6 for each group; samples were run in triplicate.

Figure 5.

Representative histologic images from intestinal characterization of the T300A rat strain. A: villus length measurement (M WT ×100; H&E). Villi were measured parallel to the center of the villus from the luminal tip to the crypt transition. B: crypt height measurement (F WT ×100; H&E). Crypts were measured from the crypt transition to the muscular layer. C: colonic mucosal thickness (M WT ×100; Alcian blue/PAS). D: Paneth cell (PC) count (M WT; ×400; Alcian blue/PAS). E and F: lysozyme IFA of ileal crypt PC. F WT, ×630 (E) and F HET, ×630 (F) HET rats exhibit inherent defects in PC granule packaging and number of granules present within the cytoplasm. Dotted lines (A–C) represent examples of how measurements were taken. Arrows (D) highlight individual PCs. Arrows (E) mark a few of many granules present. M, males; F, female; WT, wild type; HET, heterozygous.

Figure 6.

Quantitative intestinal histology in T300A strain. A: villus length as measured from the tip of the villus (luminal side) to the beginning of the crypt transition as measured by average villus length for WT and HET males and females and percent of epithelium comprised of the villus (villus length/villus length + crypt height). B: average colonic epithelium height as measured parallel from the luminal surface of the colonic mucosa to the base of the epithelium in contact with the muscular layer. C: abnormal Paneth cells (PC) on lysozyme IFA with results measured as abnormal (yes) or normal (no) and reported as the percentage of cells with abnormal PC granulation. D: PC counts as measured by both number of PC per crypt and a differential count of PC per 100 crypt cells. n = 6 for all groups; × in D represents a single outlier for the 6 WK HET group. Student’s t tests between 3 WK rats and between 6 WK rats used. WK, weeks.

Figure 7.

Gut microbiome analysis for 16-wk-old WT and HET T300A rats. A: stacked bar charts of genus-level OTUs for each rat per group. Each color represents a unique OTU in each group. B: principal component analysis plots. The amount of distance between 95% confidence ellipses show the amount of difference between groups. No significant differences were found. C: cluster heat map grouping samples based on the 25 most common OTUs present in the dataset from rats of WT and HET genotypes (n = 6/group, upper right) with samples arranged according to an unweighted pair group method with arithmetic mean (UPGMA) algorithm. HET, heterozygous; WT, wild type.