Conditional deletion of Des1 in the mouse retina does not impair the visual cycle in cones

Abstract

Cone photoreceptors are essential for vision under moderate to high illuminance and allow color discrimination. Their fast dark adaptation rate and resistance to saturation are believed to depend in part on an intraretinal visual cycle that supplies 11- cis-retinaldehyde to cone opsins. Candidate enzymes of this pathway have been reported, but their physiologic contribution to cone photoresponses remains unknown. Here, we evaluate the role of a candidate retinol isomerase of this pathway, sphingolipid δ4 desaturase 1 (Des1). Single-cell RNA sequencing analysis revealed Des1 expression not only in Müller glia but also throughout the retina and in the retinal pigment epithelium. We assessed cone functional dependence on Müller cell-expressed Des1 through a conditional knockout approach. Floxed Des1 mice, on a guanine nucleotide-binding protein subunit α transducin 1 knockout ( Gnat1^-/-) background to allow isolated recording of cone-driven photoresponses, were bred with platelet-derived growth factor receptor α (Pdgfrα)-Cre mice to delete Des1 in Müller cells. Conditional knockout of Des1 expression, as shown by tissue-selective Des1 gene recombination and reduced Des1 catalytic activity, caused no gross changes in the retinal structure and had no effect on cone sensitivity or dark adaptation but did slightly accelerate the rate of cone phototransduction termination. These results indicate that Des1 expression in Müller cells is not required for cone visual pigment regeneration in the mouse.-Kiser, P. D., Kolesnikov, A.V., Kiser, J. Z., Dong, Z., Chaurasia, B., Wang, L., Summers, S. A., Hoang, T., Blackshaw, S., Peachey, N. S., Kefalov, V. J., Palczewski, K. Conditional deletion of Des1 in the mouse retina does not impair the visual cycle in cones.

Keywords: Müller glia; photoreceptor; retinoid cycle; sphingolipidδ(4) desaturase.

Figure 1

Pathways contributing to cone visual pigment regeneration in the mammalian retina. Cone pigment regeneration is mediated by both the classic visual cycle as well as an intraretinal visual cycle involving Müller glia. Des1, an enzyme originally recognized for its involvement in de novo ceramide production, is reported to catalyze the isomerization of retinol and has been proposed to be the physiologic isomerase of the cone-specific, intraretinal visual cycle. Here, we tested this proposal through the use of Müller cell Des1 conditional knockout mice. Chemical intermediates of the classic visual cycle are omitted for simplicity.

Figure 2

Single-cell RNAseq analysis of Des1 expression in the adult mouse retina. A) Analysis of ∼17,000 cells revealed 17 different cell clusters from the adult whole mouse retina. B) Des1 expression in the whole mouse retina at the single-cell level (red points). C) Comparative levels of Des1, Rlbp1, Lrat1, and Rpe65 transcripts in the Müller glia and RPE cell clusters. AC, amacrine cell; BC, bipolar cell; HC, horizontal cell; RGC, retinal ganglion cell; UMI, unique molecular identifier.

Figure 3

Generation and genotyping of floxed Des1 mice. A) Exon 2 of the Des1 gene was targeted for deletion by insertion of parallel flanking LoxP sequences. The targeting vector also contained a neomycin-resistance cassette, flanked by FLP recombinase targeting sequences, to allow for selection of successfully targeted embryonic stem cells. The neomycin cassette was subsequently removed by crossing floxed Des1 mice with FRT germline deleter mice to yield mice lacking the neomycin cassette that were used for further breeding and analysis. The green and red triangles represent LoxP and FRT sequences, respectively. PCR primer binding sites and sizes of expected products are shown on the figure. Primers sequences are shown in Supplemental Table S1. HSV-TK, herpes simplex virus thymidylate kinase gene. B) PCR genotyping results for wild-type (Des1^+/+), heterozygous (Des1^F/+), and homozygous (Des1^F/F) mice.

Figure 4

Genome- and protein-level analysis of efficiency and tissue specificity of Des1 conditional knockout. A) Multiplex PCR strategy for detecting Cre recombinase activity directed toward the floxed Des1 allele. The green and red triangles represent LoxP and FRT sequences, respectively. Primer binding positions are shown as arrows above and below the gene schematics. Primer sequences are given in Supplemental Table S1. B) Assay used to assess the level of Des1 activity in targeted tissues. Des1 activity against a tritiated dihydroceramide substrate releases tritiated water, which can be detected by scintillation counting. C) Multiplex PCR showing the susceptibility of Des1 exon 2 deletion in wild-type or floxed Des1 mice expressing Cre recombinase under the albumin promoter. Primers used for the reactions are listed below the gel image. D) Des1 activity assays conducted by using liver microsome samples from 1- to 2-mo-old mice as the enzyme source. Aqueous radioactivity produced during the reaction was normalized to the total protein concentration in the assay mixture. The “Des1 Alb-Cre” bar shows pooled data from Cre-negative Des1^+/+, Des1^F/+, and Des1^F/F mice, which exhibited comparable activity levels. Des1 Alb-Cre⁻, n = 3; Des1^F/+ Alb-Cre⁺, n = 3; Des1^F/F Alb-Cre⁺, n = 4. E) Multiplex PCR showing the susceptibility of Des1 exon 2 deletion in wild-type or floxed Des1 mice expressing Cre recombinase under the Pdgfrα promoter. Primers used for the reactions are listed below the gel image. F) Des1 activity assays conducted by using whole retinal lysates, free of RPE contamination, from 1- to 2-mo-old mice of the indicated genotypes as the enzyme source. Aqueous radioactivity produced during the reaction was normalized to the total protein concentration in the assay mixture. Des1^+/+ Pdgfrα-Cre⁺, n = 2; Des1^F/+ Pdgfrα-Cre⁺, n = 10; Des1^F/F Pdgfrα-Cre⁺, n = 4. Data (D, F) are presented as means ± sem. ^# P < 0.01 vs. Cre negative (D) or wild-type (E) Des1 control groups.

Figure 5

Analysis of Des1 mouse retinal structure by OCT. A) Representative OCT images from Des1^F/+ Gnat1^−/− Pdgfrα-Cre⁺ (top) and Des1^F/F Gnat1^−/− Pdgfrα-Cre⁺ (bottom) mice. The left and right portions of the images represent the inferior and superior regions of the retina, respectively. The bracket demarcates the ONL. B) Quantification of ONL thickness. The ONL measurements were made 500 μm from the optic nerve head in the superior, inferior, nasal, and temporal quadrants of the retina and then averaged. The data are presented as overall averages ± sem. Des1^F/+ Gnat1^−/− Pdgfrα-Cre⁺, n = 6; Des1^F/F Gnat1^−/− Pdgfrα-Cre⁺, n = 8.

Figure 6

Impact of retinal Des1 conditional deletion on single-flash ERG b-wave responses of murine cone photoreceptors. Full-field single-flash ERGs were recorded from 2- to 3-mo-old mice. The group labeled “control” was aggregated from littermates expressing at least 1 functional copy of Des1, all of which exhibited comparable ERG responses. After recovery from initial dark- and light-adapted ERG recordings mice were administered a single 160 μg dose of MB-001 by intraperitoneal injection and 2 h later administered mydriatic eye drops and then exposed to 10,000 lux illumination for 10 min followed by a 4 h dark adaptation period. Dark-adapted ERGs were then recorded from these mice. Data are shown as average ± sem. Control mouse group, n = 6; Des1^F/F Gnat1^−/− Pdgfrα-Cre⁺ group, n = 4.

Figure 7

Physiologic characterization of cones in Des1^F/+ Gnat1^−/− Pdgfrα-Cre⁺ control and Des1^F/F Gnat1^−/− Pdgfrα-Cre⁺ mutant mice and their dark adaptation. A) Kinetics of activation and inactivation of cone phototransduction in control and Des1-deficient mice measured by transretinal (ex vivo) ERG. Population-averaged (mean ± sem) dim flash responses to test stimuli of 7.0 × 10³ photons/μm² (n = 13 for controls, and n = 10 for mutants). B) Comparison of saturated cone responses from control and Des1-deficient mice obtained by ex vivo ERG. Population-averaged (mean ± sem) responses to test stimuli of 2.0 × 10⁶ photons/μm² (n = 13 for controls, and n = 10 for mutants). C) Recovery of cone ERG a-wave flash sensitivity (S_f) in isolated retinas of control (n = 12) and Des1-deficient (n = 10) animals after bleaching >90% of cone pigment at time 0 with 505 nm LED light. Data were fitted with single-exponential functions that yielded the recovery time constants of 1.2 ± 0.2 and 1.1 ± 0.2 min, respectively. S_f^DA denotes the sensitivity of dark-adapted cones. D) Recovery of photopic ERG b-wave S_f in vivo in control (n = 10) and Des1-deficient (n = 10) mice after bleaching >90% of cone pigment at time 0 with 520 nm LED light. S_f^DA designates the sensitivity of dark-adapted cones.