Sphingosine kinases and their metabolites modulate endolysosomal trafficking in photoreceptors

Abstract

Internalized membrane proteins are either transported to late endosomes and lysosomes for degradation or recycled to the plasma membrane. Although proteins involved in trafficking and sorting have been well studied, far less is known about the lipid molecules that regulate the intracellular trafficking of membrane proteins. We studied the function of sphingosine kinases and their metabolites in endosomal trafficking using Drosophila melanogaster photoreceptors as a model system. Gain- and loss-of-function analyses show that sphingosine kinases affect trafficking of the G protein-coupled receptor Rhodopsin and the light-sensitive transient receptor potential (TRP) channel by modulating the levels of dihydrosphingosine 1 phosphate (DHS1P) and sphingosine 1 phosphate (S1P). An increase in DHS1P levels relative to S1P leads to the enhanced lysosomal degradation of Rhodopsin and TRP and retinal degeneration in wild-type photoreceptors. Our results suggest that sphingosine kinases and their metabolites modulate photoreceptor homeostasis by influencing endolysosomal trafficking of Rhodopsin and TRP.

Figure 1.

Targeted expression of Sk2 leads to photoreceptor degeneration and decrease in Rh1 and TRP levels. (A) 7-d-old photoreceptors overexpressing Sk1 show a mild effect on rhabdomeres, whereas photoreceptors overexpressing Sk2 show severe degeneration. The percentage of intact rhabdomeres for w¹¹¹⁸ is 100%, for Sk1 overexpression is 80%, and for Sk2 overexpression is 4%. R1–R7 label the rhabdomeres of R1–R7 photoreceptor cells; N represents the nucleus, M represents mitochondria, and black arrows show adherens junctions. (B) Sk2 overexpression results in decreased Rh1 and TRP levels in 7-d-old flies. The TRP blot is reprobed with an antibody to IPP as a loading control. Sk2 overexpression is also seen in blots probed with an Sk2 antibody. (C) A 65–70% reduction in Rh1 level and a 60% reduction in TRP level are observed in Sk2 overexpressors compared with w¹¹¹⁸. (D) Graph shows the extent of Sk2 overexpression compared with w¹¹¹⁸. (E) Photoreceptors expressing Sk2 at 30°C degenerate, whereas those at 18°C expressing Gal80^ts do not degenerate. (F) TRP level is reduced at 30°C compared with 18°C. Sk2 overexpressor (Sk2 OE) represents GMR-Gal4–driven Sk2 at 25°C. The middle blot shows Sk2 overexpression, and the bottom blot shows loading controls. (G and H) Quantitation of TRP bands (G) and Sk2 (H) at permissive and nonpermissive temperature, and Sk2 overexpression at 25°C. n = 3; bars denote standard deviation.

Figure 2.

Overexpression of Sk2 in a car¹ mutant restores Rh1 and TRP levels, and photoreceptor degeneration is suppressed. (A) Rh1 and TRP levels are recovered when Sk2 is overexpressed in a car mutant. (B) The percentage of expression of Rh1 and TRP in an Sk2 overexpressor (Sk2 OE) and an Sk2 overexpressor in a car mutant background compared with car mutant alone. n = 3; bars denote standard deviation. (C) Degeneration is suppressed in 7-d-old photoreceptors of the Sk2 overexpressor in a car mutant. 4% of rhabdomeres are intact in Sk2 overexpressors, whereas 88% are intact in Sk2 overexpressors in car. (D) S2 cells were stained with antibodies to Sk1 and cadherin, a plasma membrane marker. Overlay shows the colocalization of Sk1 with cadherin. S2 cells constitutively expressing GFP-tagged Lamp1 were stained with antibodies to Sk2. Overlay shows the colocalization of Sk2 with Lamp1-GFP.

Figure 3.

Increase in DHS1P leads to retinal degeneration in w¹¹¹⁸ photoreceptors accompanied by a decrease in Rh1 and TRP levels, and Sk2 overexpression increases the DHS1P level. (A) Transmission electron micrograph of photoreceptors of w¹¹¹⁸ fed sphingosine, S1P, DHS, and DHS1P. 100% of rhabdomeres are intact in w¹¹¹⁸ fed sphingosine and S1P, 90% of rhabdomeres are intact in DHS-fed flies, and 27% are intact in DHS1P-fed flies. (B) Rh1 and TRP are reduced in flies fed DHS1P compared with solvent and flies fed with other lipids. (C) Rh1 and TRP levels are reduced by 40% in w¹¹¹⁸ flies fed DHS1P compared with solvent. Bars denote standard deviation; n = 3. (D) UFLC MS/MS measurements show that Sk2 overexpressors have a higher level of DHS1P compared with control or Sk1 overexpressors. Bars denote standard error; n = 3.

Figure 4.

Increasing DHS1P induces Sk1-expressing photoreceptors to degenerate, and photoreceptor degeneration in Sk2 overexpressors is suppressed in a CalX mutant. (A) Photoreceptors of Sk1 overexpressors degenerate when DHS1P is increased. In DHS1P-fed Sk1 flies, 32% of the rhabdomeres are intact compared with flies raised in solvent. (B) Photoreceptor degeneration in 7-d-old Sk2 overexpressors is suppressed in a calx^B mutant, suggesting altered calcium homeostasis is a likely cause of degeneration. 81% of the rhabdomeres are intact in Sk2 overexpressors in calx^B; however, many rhabdomeres are elongated in shape.

Figure 5.

Altered DHS1P to S1P ratio in Sk1 mutants. (A) The targeting vector used to generate Sk1 mutants by ends-out homologous recombination. (B) sk1 mutants do not show Sk1 protein, indicating that they are null mutants. (C) Quantification of Sk1 in overexpressors and sk1 mutant alleles. (D) Rh1 and TRP are low in 1-d-old sk1¹ compared with control cinnabar.brown (cn.bw) flies. (E) Quantification of Rh1 and TRP shows that Rh1 is decreased to 30% and TRP to 40% in sk1 mutants compared with control cn.bw flies. (F) Rh1 and TRP in sk2 compared with control flies. (G) Quantification of Rh1 and TRP bands in sk2. (C, E, and G) n = 3; bars denote standard deviation. (H) Mass spectrometric measurements show a 20% reduction in S1P level in sk1. n = 3; bars denote standard error. OE, overexpressor.