MKLN1 splicing defect in dogs with lethal acrodermatitis

Abstract

Lethal acrodermatitis (LAD) is a genodermatosis with monogenic autosomal recessive inheritance in Bull Terriers and Miniature Bull Terriers. The LAD phenotype is characterized by poor growth, immune deficiency, and skin lesions, especially at the paws. Utilizing a combination of genome wide association study and haplotype analysis, we mapped the LAD locus to a critical interval of ~1.11 Mb on chromosome 14. Whole genome sequencing of an LAD affected dog revealed a splice region variant in the MKLN1 gene that was not present in 191 control genomes (chr14:5,731,405T>G or MKLN1:c.400+3A>C). This variant showed perfect association in a larger combined Bull Terrier/Miniature Bull Terrier cohort of 46 cases and 294 controls. The variant was absent from 462 genetically diverse control dogs of 62 other dog breeds. RT-PCR analysis of skin RNA from an affected and a control dog demonstrated skipping of exon 4 in the MKLN1 transcripts of the LAD affected dog, which leads to a shift in the MKLN1 reading frame. MKLN1 encodes the widely expressed intracellular protein muskelin 1, for which diverse functions in cell adhesion, morphology, spreading, and intracellular transport processes are discussed. While the pathogenesis of LAD remains unclear, our data facilitate genetic testing of Bull Terriers and Miniature Bull Terriers to prevent the unintentional production of LAD affected dogs. This study may provide a starting point to further clarify the elusive physiological role of muskelin 1 in vivo.

Fig 1. LAD phenotype.

(A) Inflammatory skin lesions in the face of an affected Bull Terrier. (B) Similar lesions in the inguinal region. (C) LAD affected puppy in the middle of two non-affected littermates. A pronounced growth delay and a subtle coat color dilution are visible. (D, E) Fore paws of an LAD affected Bull Terrier puppy at necropsy. Symmetrical scaling and crusting of the skin including interdigital areas and foot pads is visible (F, G) Histopathological micrographs of the junction of interdigital haired skin and digital pad from an affected Bull Terrier puppy (F) and a control dog (G). Marked thickening of the epidermis, excessive layers of non-cornifying epithelium and a large pustule are evident in the affected dog. Hematoxylin-eosin, bar = 400 µm.

Fig 2. Mapping of the LAD locus.

(A) A GWAS was performed in a cohort of 22 LAD cases and 48 controls. The Manhattan plot shows a single significant signal at the beginning of chromosome 14. The red line indicates the Bonferroni significance threshold (P_Bonf = 6.5 x 10^-7). The quantile-quantile (QQ) plot in the inset shows the observed versus expected–log(p) values. The straight red line in the QQ plot indicates the distribution of p-values under the null hypothesis. The deviation of p-values at the right side indicates that these markers are stronger associated with the trait than it would be expected by chance. (B) Haplotype analysis in the 22 LAD cases. Each horizontal bar represents the chromosome 14 haplotypes of one dog. Twenty Bull Terriers and two Miniature Bull Terriers (MBT) had large homozygous intervals with allele sharing on chromosome 14 indicated in blue. The homozygous haplotype segment shared between all 22 dogs spanned ~1.11 Mb. The critical interval for the causative LAD variant corresponded to the interval between the first flanking heterozygous markers on either side or chr14:5,248,244–6,355,383 (CanFam 3.1 assembly). (C) Gene annotation for the critical interval. The NCBI annotation release 105 listed 4 protein coding genes (indicated in black or red) and 11 genes for non-coding RNAs (indicated in blue).

Fig 3. Sanger confirmation of the MKLN1:c.400+3A>C variant.

(A) Electropherograms from dogs with the three different genotypes. (B) Wildtype and mutant allele compared to the consensus sequence for the human U2 GT-AG type 5’-splice sites [13]. Subscript numbers in the consensus sequence indicate the percentage of the respective conserved nucleotide in 183,682 investigated human 5’-splice site motifs of the U2 GT-AG type. The additional difference to the optimal consensus in the U1 spliceosomal RNA recognition site in the mutant allele is highlighted in red. In human 5’-splice sites the most frequent base at position 3 is an A (60%). G is also common at this position (35%), while C and T are both rare (<3%) [13].

Fig 4. Experimental verification of the MKLN1 splice defect.

(A) The genomic organization of the MKLN1 gene. Exons 2–6 are enlarged and the position of the primers used for RT-PCR is indicated. (B) RT-PCR was performed using skin cDNA from a control and an LAD affected Bull Terrier. The picture shows a Fragment Analyzer gel image of the experiment. In the control animal, only the expected 366 bp product is visible. In the LAD affected dog, a 277 bp product representing a transcript lacking exon 4 is visible. The identity of the bands was verified by Sanger sequencing. Thus, the MKLN1:c.400+3A>C variant leads to complete skipping of exon 4 (MKLN1:r.312_400del89).