Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jun 5;27(11):1573-1584.e6.
doi: 10.1016/j.cub.2017.04.057. Epub 2017 May 25.

Canine Brachycephaly Is Associated with a Retrotransposon-Mediated Missplicing of SMOC2

Thomas W Marchant  1 , Edward J Johnson  1 , Lynn McTeir  1 , Craig I Johnson  1 , Adam Gow  1 , Tiziana Liuti  1 , Dana Kuehn  2 , Karen Svenson  3 , Mairead L Bermingham  4 , Michaela Drögemüller  5 , Marc Nussbaumer  6 , Megan G Davey  1 , David J Argyle  1 , Roger M Powell  7 , Sérgio Guilherme  8 , Johann Lang  9 , Gert Ter Haar  10 , Tosso Leeb  5 , Tobias Schwarz  1 , Richard J Mellanby  1 , Dylan N Clements  1 , Jeffrey J Schoenebeck  11
Affiliations

Canine Brachycephaly Is Associated with a Retrotransposon-Mediated Missplicing of SMOC2

Thomas W Marchant et al. Curr Biol. .

Abstract

In morphological terms, "form" is used to describe an object's shape and size. In dogs, facial form is stunningly diverse. Facial retrusion, the proximodistal shortening of the snout and widening of the hard palate is common to brachycephalic dogs and is a welfare concern, as the incidence of respiratory distress and ocular trauma observed in this class of dogs is highly correlated with their skull form. Progress to identify the molecular underpinnings of facial retrusion is limited to association of a missense mutation in BMP3 among small brachycephalic dogs. Here, we used morphometrics of skull isosurfaces derived from 374 pedigree and mixed-breed dogs to dissect the genetics of skull form. Through deconvolution of facial forms, we identified quantitative trait loci that are responsible for canine facial shapes and sizes. Our novel insights include recognition that the FGF4 retrogene insertion, previously associated with appendicular chondrodysplasia, also reduces neurocranium size. Focusing on facial shape, we resolved a quantitative trait locus on canine chromosome 1 to a 188-kb critical interval that encompasses SMOC2. An intronic, transposable element within SMOC2 promotes the utilization of cryptic splice sites, causing its incorporation into transcripts, and drastically reduces SMOC2 gene expression in brachycephalic dogs. SMOC2 disruption affects the facial skeleton in a dose-dependent manner. The size effects of the associated SMOC2 haplotype are profound, accounting for 36% of facial length variation in the dogs we tested. Our data bring new focus to SMOC2 by highlighting its clinical implications in both human and veterinary medicine.

Keywords: GWAS; QTL; SMOC2; brachycephaly; craniofacial; dog; morphology; retrotransposon; selection; whole-genome sequencing.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Capturing Gross Interbreed and Subtle Intrabreed Variation in Skull Shape (A) Three-dimensional isosurfaces of canine skulls are reconstructed from computed tomography (CT) scans of referral patients. (B–E) Lateral images of a smooth collie (B; dolichocephalic), Bernese mountain dog (C; mesocephalic), border terrier (D; mesocephalic), and pug (E; brachycephalic) with corresponding isosurfaces were included in our analysis. Head images and isosurfaces are not to scale. (F) Lateral and dorsoventral views of the canine skull with wireframe diagrams superimposed, representing the changes in viscerocranium shape for negative and positive viscerocranium PC1 scores (“ve PC”). Red circles indicate surface landmarks of the rostrum. Connecting blue lines are added to provide visual context to shape. Circles connected by black dotted lines indicate landmarks of the hard palate. (G) Individual breed members cluster together when viscerocranium shape (viscerocranium PC1) is plotted against body size (neurocranium centroid). BMD-, Bernese mountain dog; BORD, border collie; BORT, border terrier; BOX-, boxer; COLL, smooth collie; PUG-, pug; YORK, Yorkshire terrier. See also Figures S1 and S2 and Table S1.
Figure 2
Figure 2
Morphology of Skull Substructures Are Associated with Multiple QTL Manhattan plots for neurocranium centroid size (A) and viscerocranium PC1 GWASs (B). Red dashed line (3.6 × 10−7) indicates threshold for multiple testing with significant SNPs colored red. The associated SNPs and candidate genes at each locus are summarized in Table 1. Insets: skull schematics indicate the region of landmarks used for datasets. Expected (x axis) and observed (y axis) −log10(p) values are plotted for all SNPs (black circles) and pruned SNPs (gray circles). Histograms depict the frequency (y axis) of viscerocranium PC1 and neurocranium centroid, respectively. See also Figures S1 and S3 and Table S2.
Figure 3
Figure 3
Regional Association and Critical Interval Determination of the CFA1 Viscerocranium QTL (A) SNP associations with viscerocranium PC1 are shown for ∼1 Mb on either side of significant SNPs on CFA1. SNPs are colored depending on the degree of LD (r2) with the index SNP (BICF2P250912; 1.91 × 10−20). (B) Ten-SNP sliding window haplotype association. (C) Genotypes between 55,881,672 and 56,020,217 (including ∼500 kb of flanking sequence) were phased and ranked by their viscerocranium PC1 value. Only haplotypes from brachycephalic dogs (viscerocranium PC1 ≤ −0.2; see Figure S3) were considered. Haplotypes are paired by subject and ranked by viscerocranium PC1 value. Alleles colored light gray match the consensus haplotype; dark gray alleles are variant. A 187.7-kb critical interval is defined by at least three meiotic recombinations (indicated above by black bar). The 12 SNPs that constitute the associated haplotype (red bar) are distributed within or up to ∼44 kb downstream of SMOC2. Black arrows indicate 3 of 37 dogs that have a homozygous variant haplotype. These dogs are registered as two Dogues de Bordeaux and a Chihuahua. The red arrow indicates a Japanese Chin that is homozygosed for a recombinant haplotype within the critical interval. See also Figures S2 and S6 and Table S3.
Figure 4
Figure 4
Characterization of the Intronic LINE-1 Retrotransposon within SMOC2 (A) Schematic of a full-length canine LINE-1 element consisting of 5′ UTR/3′ UTR, open reading frames 1 and 2 (ORF1/ORF2), and a polyadenylated tail (AAAAn) flanked by target site duplications (TSD). The structural variant within SMOC2 is 1,506 bp in length (in addition to a poly(A) tail) and has a 99.1% match to the consensus sequence of canine LINE-1. (B and C) Distribution of the SMOC2 LINE-1 fragment for (B) viscerocranium PC1 and neurocranium centroid size (C) across all individuals. (D) Ventral-dorsal view of the canine hard palate and its constituent bones. (E) Length and width of the canine palate and constituent bones normalized by the neurocranium centroid for homozygous ancestral (white), heterozygotes (gray), and homozygous-derived (black) individuals for the SMOC2 LINE-1 insertion. (F) Relative expression levels of SMOC2 both up- and downstream of the LINE insertion (<0.05 ∗; <0.01 ∗∗; <0.001 ∗∗∗). Error bars represent SEM. (G) RNA sequencing (RNA-seq) data reveal four genes with significant changes in mRNA levels (red) for homozygous SMOC2 LINE-1 carriers compared to non-carriers (three each). Neighboring genes to SMOC2 are colored green. (H) Schematic of genomic DNA (gDNA) spanning exon 8 and 9 of SMOC2, including the LINE-1 fragment. mRNA transcripts include the canonical splicing of SMOC2 (i) followed by the three most abundant SMOC2 isoforms when the LINE-1 element is present (ii–iv). All isoforms have premature stop codons prior to exon 9. C/T indicates the SNP in exon 8. Schematic is not to scale. (I) Exons 1–13 of SMOC2 contribute to a follistatin-like module (FS), thyroglobulin-like modules (TY), a unique SMOC module, an extracellular calcium-binding module (EC), and a signal peptide (SP). See also Figures S5 and S7 and Tables S5 and S6.
Figure 5
Figure 5
Size Effects of the Viscerocranium Shape and Neurocranium Centroid Size QTL (A and C) Boxplots depicting the distribution of normalized size-corrected viscerocranium PC1 (A) and normalized neurocranium centroid size (C) for 11 loci linked with body size and skull shape. Distributions are subdivided by genotype —homozygous ancestral (AA), heterozygotes (AD), and homozygous derived (DD). ∗∗∗ denotes p < 0.001 in Mann-Whitney-Wilcoxon and Kolmogorov-Smirnov tests. (B and D) A stepwise linear regression model for viscerocranium PC1 (B) and neurocranium centroid (D) determined the best explanatory model for ancestral (left) and derived (right) genotypes for each positional candidate.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Stockard C.R. The Wistar Institute of Anatomy and Biology; Philadelphia: 1941. The Genetic and Endocrinic Basis for Differences in Form and Behaviour.
    1. Wayne R.R. Phylogeny and origin of the domestic dog. In: Ruvinsky A., Sampson J., editors. The Genetics of the Dog. CABI; 2001. pp. 1–14.
    1. Harvey R.G., ter Haar G. Brachycephalic obstructive airway syndrome. In Ear, Nose and Throat Diseases of the Dog and Cat. In: Harvey R.G., ter Haar G., editors. CRC Press; Devon: 2016. pp. 290–293.
    1. Poncet C.M., Dupre G.P., Freiche V.G., Estrada M.M., Poubanne Y.A., Bouvy B.M. Prevalence of gastrointestinal tract lesions in 73 brachycephalic dogs with upper respiratory syndrome. J. Small Anim. Pract. 2005;46:273–279. - PubMed
    1. Sanchez R.F., Innocent G., Mould J., Billson F.M. Canine keratoconjunctivitis sicca: disease trends in a review of 229 cases. J. Small Anim. Pract. 2007;48:211–217. - PubMed

MeSH terms