Elovl4 haploinsufficiency does not induce early onset retinal degeneration in mice

Abstract

ELOVL4 was first identified as a disease-causing gene in Stargardt macular dystrophy (STGD3, MIM 600110.) To date, three ELOVL4 mutations have been identified, all of which result in truncated proteins which induce autosomal dominant juvenile macular degenerations. Based on sequence homology, ELOVL4 is thought to be another member within a family of proteins functioning in the elongation of long chain fatty acids. However, the normal function of ELOVL4 is unclear. We generated Elovl4 knockout mice to determine if Elovl4 loss affects retinal development or function. Here we show that Elovl4 knockout mice, while perinatal lethal, exhibit normal retinal development prior to death at day of birth. Further, postnatal retinal development in Elovl4 heterozygous mice appears normal. Therefore haploinsufficiency for wildtype ELOVL4 in autosomal dominant macular degeneration likely does not contribute to juvenile macular degeneration in STGD3 patients. We found, however, that Elovl4+/- mice exhibit enhanced ERG scotopic and photopic a and b waves relative to wildtype Elovl4+/+ mice suggesting that reduced Elovl4 levels may impact retinal electrophysiological responses.

Figure 1

Targeted mutagenesis of the Elovl4 gene. A. The targeting vector used for disrupting the endogenous Elovl4 gene. A pGKneo cassette flanked by loxP (open triangles) was inserted into exon 2 of a 15 Kb Elovl4 fragment. The disrupted Exon 2 construct is flanked by 2.5 kb short arm (B) and 4 kb long arm (A) sequences for homologous recombination in embryonic stem cells. Transcription directions of neo and tk are indicated by one way arrows. B. The endogenous gene. The direction of Elovl4 transcription is from left to right. C. The Elovl4 locus after homologous recombination. Arrows indicate primers used for PCR genotyping.

Figure 2

Identification and characterization of Elovl4 mice. A. Genotypes determined by PCR amplification of DNA extracted from tail biopsies. The upper band is generated when the endogenous Elovl4 gene is present, and the lower band from the disrupted gene. The left lane shows Elovl4^+/+ genotype, the middle lane shows Elovl4^+/−, and the right lane represents Elovl4^−/− genotyping reactions. B. Size comparison of embryonic day 18 Elovl4^+/+, Elovl4^+/−, and Elovl4^−/− littermates. C. Mean weight differences are shown for Elovl4 knockout pups vs. wildtype and heterozygous pups. D. RT-PCR results show that Elovl4 RNA levels are reduced by 46% in Elovl4^+/− mice relative to wildtype mice. GAPDH amplification signals used for normalization between retinal samples. N=4

Figure 3

Western blotting using the YP2 (ELOVL4) antibody. A. YP2 recognizes the His-Elovl4 fusion protein expressed in bacterial lysates. B. Human and bovine retinal lysates showing ELOVL4 protein detection with the YP2 antibody and effective loss of signal with ELOVL4 protein preabsorption.

Figure 4

ELOVL4 immunolabeling in human and mouse retinal sections. ELOVL4 signal is green in all figures. A. Normal 27 yr old human retina showing that the YP2 antibody immunolabels photoreceptor inner segments (IS) (short arrow- cone IS, long arrow- rod IS), outer plexiform layer (OPL), and cells within the inner nuclear layer (INL) and ganglion cell layer (GCL). Cell nuclei are labeled with PI (red). B. A similar immunolabeling profile is evident in adult mouse retina. C. Preabsorption control in mouse retina. D-F. Embryonic day 18 retinas from Elovl4 ^+/+ (D), Elovl4 ^+/− (E), and Elovl4 ^−/− (F) mice immunostained with the YP2 antibody.

Figure 5

Retinal sections from Elovl4 ^+/+ (A) and Elovl4 ^+/− (B) 11 month old mice. Nuclei are stained with PI and outer segments are labeled with a CNGCa1 antibody (green). Retinal layers are normal in the Elovl4 ^+/− retina. Arrows indicate some disorganization in the outer segments.

Figure 6

Nine month old Elovl4^+/+ and Elovl4^+/− mice scotopic and photopic single flash ERG responses. The scotopic a and b waves were measured at intensities log −2, −1, 0, +1, and +2 cds/m². The filled symbols show trough a wave amplitude values (μV), and the open symbols represent the (adjusted) a wave values at 7msec after light flash. The dark line with squares represent the mean Elovl4^+/+ amplitude responses in μV (N=5) and the dashed line with diamonds represent the mean Elovl4^+/− (heterozygous) mouse responses (N=11). Note that in both sets of scotopic a wave plotted values the heterozygous mice show a tendency for enhanced responses at mesopic intensities. The scotopic b wave shows a similar enhanced response, although not significant) at log 0 cds/m². The Elovl4^+/− mice also show significantly enhanced photopic b wave amplitude at log 0 cds/m². Representative scotopic and photopic ERG traces (at intensity = log 0 cds/m²) are shown in the lower right corner.

Figure 7

Analysis of lipid differences in RBCs from Elovl4^+/+ and Elovl4^−/− mice. Elovl4^−/− mice show a significant increase in DHA, C24:2, and Adrenic acid levels, and decreased EPA levels compared to normal controls. Elovl4^−/− mice also show increased levels of C25:1 and C26:2. The inset table shows corresponding p-values and percent changes in Elovl4^−/− fatty acid levels compared to normal controls. (Control n=4, Elovl4^+/− n=5, Elovl4^−/− n=3)