GTPBP2 in-frame deletion in canine model with non-syndromic progressive retinal atrophy

Abstract

Progressive retinal atrophy (PRA), caused by aberrant functioning of rod/cone photoreceptors, leads to blindness affecting mammals, including dogs. We identified a litter of three Labrador retrievers affected by non-syndromic PRA; the parents and three other siblings were unaffected. Homozygosity mapping and whole-genome sequencing detected a homozygous 3-bp deletion in the coding region of GTPBP2, located in CFA12 (NC_049233.1:12,264,348_12,264,350del, c.1606_1608del, p.Ala536del). The variant was absent from the online European Variation Archive (EVA) database, the Dog Biomedical Variants Database Consortium, and the Dog10k database. We tested 91 non-affected dogs from the same kennel and found 75 wild-type (WT) and 16 carriers, all clinically normal, and 569 Labradors from the general population (USA), all WT. GTPBP2 is associated with Jaberi-Elahi syndrome (JES) in Homo sapiens, and splice variants in Mus musculus are associated with neurodegeneration; in both cases photoreceptor degeneration may be included in its manifestation. Heterologous cellular systems were transfected with cDNA encoding WT or A536del mutant GTPBP2 protein and immunoblot analysis of total cell lysate with anti-GTPBP2 antibodies showed that the expression level of the GTPBP2 mutant protein A536del is slightly but not significantly reduced compared to WT. Immunofluorescent methods and confocal analysis of cells transfected with WT or A536del GTPBP2 protein revealed that the WT form is diffuse throughout the cytosol, while the mutant form resulted in the formation of cytoplasmic aggregates in ~70-80% of cells. The deleted amino acid falls within a conserved interval outside the GTP domain of GTPBP2, suggesting a potentially novel role of the sequence on cellular localization of the protein.

Keywords: Animal model; Non-syndromic; Phenotype variability; Progressive retinal atrophy; Protein domain; Protein function.

Fig. 1

Phenotype. Fundus pictures of unaffected (a) and affected (b) dogs. a - Picture of the left retina of a 15-months old unaffected and unrelated Labrador from the same kennel. b– Fundus the right eye of a PRA-affected case at 25 months of age demonstrating late-stage disease. Note the marked thinning of the retinal vessels and pale atrophic optic nerve head. Both normal and affected dogs have a normal variation of the tapetum color which appears brown to dull gray.

Fig. 2

Mapping. (a) Family tree of the affected dogs. Males in squares, females in circles, affected in blue. Obligate carriers (and unaffected sibling later identified as carriers) as half-filled shapes. Autosomal recessive inheritance mechanism was suspected. The whole family was genotyped on canine SNP chip, whole-genome sequenced dogs are marked with red asterisk. (b) Homozygosity mapping of the three cases, compared against the three available unaffected siblings and two parents was carried out. The homozygous regions shared by all the cases and exclusive to them compared to the controls are marked in red (37 Mb region in CFA12 marked with an asterisk).

Fig. 3

DNA sequencing. (a) Detail of the of the 3-bp in-frame GTPBP2 deletion variant exclusive for the cases. Screenshots of the interval visualized with IGV. On the top, a sequenced case is shown, homozygous of the haplotype and variant, and the deleted 3-bp variant. Further down, a carriers visible. Bottom panel of (a) normal retinal RNA-seq shows the variant falling within the coding region of the GTPBP2 gene (position based on the c-terminal part of isoform 1: c.1606_1608del, p.Ala536del). (b) Sanger results for the variant showing the electropherograms for wild-type, carrier and affected. In the wild-type (top), the red rectangle marks the triplet deleted in the affected, in which the deletion position is marked with an arrow (bottom).

Fig. 4

Variant position and GTPBP2 sequence. (a) Alignment of the c-terminal Isoform 1 (NCBI: XP_038539566.1, entry under UU_Cfam_GSD_1.0) of canine GTPBP2 with the same interval in other five different mammalian species. Observe the high degree of conservation of the protein. The deleted Alanine is highlighted. (b) 3D structure of human GTPBP2 with the known human syndromic diseases associated variants marked in red for frameshifts and premature stop codons, and in yellow for amino acid changes. alongside the canine variant transposed, in green. In purple, D1 and D2 show the boundaries of the GTP domain, with variants numbered 15 and above outside of it. Bottom right corner, whole structure without the cutoff of the N-terminal loop. 1-Ser3X; 2-Leu93P; 3-Arg121X; 4-Lys125R; 5-Arg131X; 6-Arg144X; 7-Arg219X; 8-Asp319N; 9-Gln332fs; 10-Glu352fs; 11-Gln407X; 12-Val413fs; 13-Arg423X; 14-Arg470X; 15-Glu509fs; 16-Arg520X; 17-Arg521fs; 18-Ala536del.

Fig. 5

Expression level analysis of GTPBP2 protein in HeLa cells transfected with WT and mutant A536del GTPBP2 constructs. Cells were transfected with WT GTPBP2 or with A536del GTPBP2 cDNAs. Cells transfected with A536del GTPBP2 were then treated with the proteasome inhibitor MG132 (+). Untransfected cells were used as control. a) Total protein lysates from untransfected and transfected cells were obtained by solubilization. An equal quantity of protein was separated by SDS-PAGE and blotted onto nitrocellulose paper. The blots were incubated with polyclonal antibodies to GTPBP2. A cropped representative Western blot is shown. The lower panel is a Ponceau Red staining of the same cell lysates, used as loading control. b) Quantification of GTPBP2 protein bands was performed by densitometric analysis on western blots. Data (mean values from at least three independent experiments + S.D.) are reported as the percentage of values from WT GTPBP2 protein. Statistical analysis was performed by One-way ANOVA test, followed by multiple comparisons Dunnett’s test. * p < 0.05.

Fig. 6

Cellular localization of WT and A536del mutant GTPBP2 proteins. HeLa cells were transfected with WT (a,b) or with A536del mutated GTPBP2 cDNAs (c,d,e,f). Untransfected cells were used as control (g,h). Cells were immunolabelled with polyclonal antibodies to GTPBP2 (red fluorescence) and subsequently with monoclonal antibodies to Lamp2 (green fluorescence), were then incubated with the appropriate secondary antibodies. Nuclear morphology was demonstrated by staining with Hoechst. Images were recorded at the same setting conditions and magnification (scale bar 10 um).