Knockdown of Dehydrodolichyl Diphosphate Synthase in the Drosophila Retina Leads to a Unique Pattern of Retinal Degeneration

Abstract

Dehydrodolichyl diphosphate synthase (DHDDS) is a ubiquitously expressed enzyme that catalyzes cis-prenyl chain elongation to produce the poly-prenyl backbone of dolichol. It appears in all tissues including the nervous system and it is a highly conserved enzyme that can be found in all animal species. Individuals who have biallelic missense mutations in the DHDDS gene are presented with non-syndromic retinitis pigmentosa with unknown underlying mechanism. We have used the Drosophila model to compromise DHDDS ortholog gene (CG10778) in order to look for cellular and molecular mechanisms that, when defective, might be responsible for this retinal disease. The Gal4/UAS system was used to suppress the expression of CG10778 via RNAi-mediated-knockdown in various tissues. The resulting phenotypes were assessed using q-RT-PCR, transmission-electron-microscopy (TEM), electroretinogram, antibody staining and Western blot analysis. Targeted knockdown of CG10778-mRNA in the early embryo using the actin promoter or in the developing wings using the nub promoter resulted in lethality, or wings loss, respectively. Targeted expression of CG10778-RNAi using the glass multiple reporter (GMR)-Gal4 driver (GMR-DHDDS-RNAi) in the larva eye disc and pupal retina resulted in a complex phenotype: (a) TEM retinal sections revealed a unique pattern of retinal-degeneration, where photoreceptors R2 and R5 exhibited a nearly normal structure of their signaling-compartment (rhabdomere), but only at the region of the nucleus, while all other photoreceptors showed retinal degeneration at all regions. (b) Western blot analysis revealed a drastic reduction in rhodopsin levels in GMR-DHDDS-RNAi-flies and TEM sections showed an abnormal accumulation of endoplasmic reticulum (ER). To conclude, compromising DHDDS in the developing retina, while allowing formation of the retina, resulted in a unique pattern of retinal degeneration, characterized by a dramatic reduction in rhodopsin protein level and an abnormal accumulation of ER membranes in the photoreceptors cells, thus indicating that DHDDS is essential for normal retinal formation.

Keywords: DHDDS; Drosophila; N-glycosylation; RNAi; photoreceptor degeneration.

FIGURE 1

Amino acid sequence alignment of the human DHDDS and its Drosophila ortholog proteins. The amino acid sequences of the human DHDDS and the Drosophila CG10778 proteins were aligned by BLASTp. Positions with identical amino acids were labeled with (*), while positions that preserved the physico-chemical properties of the original residue were marked with (:). Marked with arrows and in color are the DHDDS sites known to be involved in non-syndromic Retinitis Pigmentosa: K42 (red), R98 (green), T206 (blue). Only R98 and T206 are identical in the Drosophila orthologue CG10778. NCBI accession numbers are as follows: NP_572425.1 (CG10778) and NP_995583.1 (DHDDS).

FIGURE 2

Knockdown of DHDDS in the developing wing primordium resulted in a complete wing loss. Nub-Gal4 flies were mated with either w¹¹¹⁸ or UAS-DHDDS-RNAi flies (VDRC #104188). (A,B) Representative photographs taken under a stereo-microscope of the wings of F1 progenies, females (left) and males (right). (A) Nub/+ control flies. (B) Nub-DHDDS-RNAi flies showing wing absence, indicating that DHDDS is essential for wing formation. (C) A different UAS-DHDDS-RNAi line with a weaker phenotype. Nub-Gal4 flies were also mated with UAS-DHDDS-RNAi flies of a different line (VDRC #3166). The F1 progenies showed a weaker wing phenotype having twisted wings. (D–E) A graphical representation of the significant decrease in DHDDS mRNA levels following its RNAi-mediated knockdown in the wing imaginal discs. q-RT-PCR analysis of DHDDS mRNA expression levels in wing discs of control (Nub/+, black) and experimental (Nub-DHDDS-RNAi, red) third instar larvae. (F) Average expression of DHDDS mRNA in wing discs of experimental group (Nub-DHDDS-RNAi, red), relative to the control group (Nub/+, black). q-RT-PCR analysis was done using the Comparative threshold Cycle quantification that was used to calculate differential mRNA. Values were normalized to the control group. Statistical significance (*P ≤ 0.05) was determined using the Mann–Whitney U test in a one-tailed test. Data is presented as mean ± standard error of the mean (SEM), n = 3. The Mann–Whitney U test is a non-parametric test that allows two groups, conditions or treatments to be compared without assuming that values are normally distributed.

FIGURE 3

Retinal TEM sections of GMR-DHDDS-RNAi knockdown flies revealed a unique pattern of photoreceptors degeneration. Thin TEM sections at a ∼40 μm depth (from the corneal surface) of dark-raised newly eclosed w¹¹¹⁸ (A,C) and experimental GMR-DHDDS-RNAi (B,D) retina at high magnification (×2000, A and ×1500, B) and low magnification (×500, C,D). The identity of the various photoreceptors was determined by the position of the easily identified R7. (A) A representative TEM section of a single ommatidium of a newly eclosed dark-raised w¹¹¹⁸ fly in which the different photoreceptors are identified by their numbers. (B) A representative TEM section of a single ommatidium of a newly eclosed dark-raised GMR-DHDDS-RNAi fly in which the different photoreceptors are identified by their numbers. Note the normal appearance of the rhabdomeres of R2 and R5 cells and the degenerated appearance of the other cells showing very short microvilli (arrows). (C) Seven neighboring ommatidia are presented at a lower magnification (×500). Similar TEM with a normal retinal structure was previously described for the white-eyed GMR/+ control strain of older age (Weiss et al., 2012). Note the normal appearance of the rhabdomeres of R2 and R5 cells and the degenerated appearance of the other cells showing very short microvilli (arrows). (D) Several neighboring ommatidia of the GMR-DHDDS-RNAi fly are presented at lower magnification (×500).

FIGURE 4

The mean cross-section area of rhabdomeric slices measured from R1-7 photoreceptors as a function of retinal depth. Analysis of serial TEM retinal cross sections of newly eclosed flies. Four sections were analyzed for the control group (w¹¹¹⁸), at the depths of 20, 40, 55, and 65 μm from corneal surface. Six sections were analyzed for the experimental group (GMR-DHDDS-RNAi), at the depths of 10, 30, 40, 55, 70, and 85 μm from corneal surface. For each depth, 4–19 ommatidia were analyzed. The values in the graphs represent the mean area (in μm²) of every R1–7 rhabdomere, as a function of depth, ±STDEV (Standart deviation). Measurements were done using the ImageJ and Origin softwares. A comparison was performed between the mean area of R2 or R5 at the retinal depth of 40 μm and the mean area of R1, serving as a representative rhabdomere from the group of R1, R3, R4, and R6, at the same retinal depth. While the mean area of R2 and R5 at 40 μm was 2.1 ± 0.4 and 2.4 ± 0.5 μm², respectively, the mean area of R1 at the same depth was 0.6 ± 0.1 μm² (average ± SEM, n = 6 for each group) and the differences between the groups were significant p ≤ 0.01 (**) for the difference between R2 and R1, and p ≤ 0.05 for the difference between R5 and R1 (a two tailed, unpaired student’s t-test). At all other depths, R2 and R5 of GMR-DHDDS-RNAi retinal sections exhibited similar areas in comparison to R1, R3, R4, and R6. Note the smaller cross section area of all rhabdomeres of GMR-DHDDS-RNAi flies (red) relative to the w¹¹¹⁸ control flies (black) at all measured retinal depth except for R2 and R5 rhabdomeres at ∼40 μm depth, where cross-section areas were similar to that of w¹¹¹⁸ flies. The cell nucleus resides at ∼ 40 μm depth from the corneal surface.

FIGURE 5

Immunostaining of the GMR-DHDDS-RNAi imaginal eye discs. Eye discs of GMR-DHDDS-RNAi and control (GMR/GFP) late third instar larvae were collected and stained with DAPI (nuclei, magenta), ELAV (developing photoreceptors, red), and Sal (R3/R4 differentiation and planar polarity, blue). (A,B) ×20 magnification of control GMR/GFP (A) and GMR-DHDDS-RNAi (B) imaginal eye discs. (C,D) ×40 magnification of sections of control (C) and GMR-DHDDS-RNAi (D) imaginal eye discs. The sections were selected to include Sal expression at R3/R4. Note that ELAV and Sal are normally expressed in the GMR-DHDDS-RNAi and control tissues.

FIGURE 6

TEM sections of GMR-DHDDS-RNAi flies revealed an enhanced degeneration with increasing age and an abnormally large accumulation of ER membranes in the affected retinae. (A,B) Representative thin TEM cross retinal sections at a ∼55 μm depth (from the corneal surface) showing an ultrastructure of a 12-days-old GMR-DHDDS-RNAi fly at a high (×2500, A) and low (×800, B) magnification. Missing rhabdomeres are marked with asterisks in (A). The ommatidium shown in (A) is marked with a red circle in (B). (C,D) Representative thin TEM cross retinal sections at a ∼55 μm depth (from the corneal surface) of dark-raised newly eclosed control (w¹¹¹⁸) flies (C) and GMR-DHDDS-RNAi flies (D) at a high magnification. The red arrows show the normal appearance of ER membranes in (C) and the abnormal accumulation of ER membranes in (D). (E) An enlarged image of the indicated area in (C) showing a normal manifestation of ER membranes in the control (w¹¹¹⁸) ommatidium. (F) An enlarged image of the indicated area in (D) showing an abnormal accumulation of ER membranes in the affected retinae of GMR-DHDDS-RNAi fly.

FIGURE 7

Prolonged Depolarizing Afterpotential (PDA) revealed a highly reduced Rh1 rhodopsin level in GMR-DHDDS-RNAi flies in vivo. (A) A representative ERG response to intense blue and orange lights of a 21-days-old white-eyed control (GMR/+) fly. A similar ERG response was observed in a newly eclosed fly of the same strain, indicating the preservation of the structure and function of the control flies (not shown). The orange bars represent orange light pulses (O; Schott OG 590 edge filter), while the blue bars represent blue light pulses (B; Schott, BG 28 broad-band filter). The first blue light pulse induced a PDA in R1–6 photoreceptors by converting Rh1 rhodopsin to its dark stable state, metarhodopsin. An on transient is indicated (arrow). The following blue light-pulse elicited neither PDA nor on-and-off transients, but only an ERG response of the R7, 8 cells. An additional orange light-pulse converted metarhodopsin back to Rh1 rhodopsin and suppressed the PDA at the cessation of the light. The following orange light did not convert a net amount of photopigment and did not elicit a PDA. (B) An ERG response of a newly eclosed white-eyed GMR-DHDDS-RNAi fly showed a very short PDA, on-and-off transients (see arrows) and a smaller ERG response amplitude under an identical experimental paradigm, indicating a highly reduced Rh1 rhodopsin level. (C) A histogram comparing the PDA amplitude of 21-days-old white-eyed GMR/+ (black) and newly eclosed white-eyed GMR-DHDDS-RNAi flies (red). The baseline was set to 0mV, and the voltage was measured 55 seconds after the first blue light was turned off. Data is presented as mean ± standard error of the mean (SEM), n = 11. The difference between the two group was significant (p ≤ 0.05, a two-tailed, unpaired student’s t-test).

FIGURE 8

A drastic reduction of Rh1 rhodopsin levels in GMR-DHDDS-RNAi flies. Western blot analysis that compares the expression levels of Rh1 rhodopsin and TRP channel of newly eclosed white-eyed GMR-DHDDS-RNAi fly relative to positive control of newly eclosed white-eyed w¹¹¹⁸ control. (A) A histogram showing a drastic reduction of Rh1 rhodopsin level in GMR-DHDDS-RNAi flies relative to positive control. The first and second columns show the Rh1 level of w¹¹¹⁸ and GMR/+, while the third column shows the Rh1 level of GMR-DHDDS-RNAi flies. The 4th column shows a negative control of the ninaE^I17 Rh1 null mutant. For each group, the histogram shows the average value of normalized αRh1 denseties from all experiments, relative to the control (w¹¹¹⁸) group. Statistical analysis was done using the Student’s t-test, in a two-tailed paired test. Data is presented as mean ± standard error of the mean (SEM), n = 6 or 7 for each group. A significant (*P ≤ 0.05) difference in αRh1 densities was found between the GMR/+ and w¹¹¹⁸ groups, and between the GMR-DHDDS-RNAi and the GMR/+ groups. (B) A histogram showing a reduction of TRP channel protein level in GMR-DHDDS-RNAi flies relative to the positive control. The first and second columns show TRP level of w¹¹¹⁸ and GMR/+ of the same flies used in section (A), while the third column shows the TRP level of GMR-DHDDS-RNAi flies. The 4th column shows a negative control of the trp^P343 a TRP null mutant. For each group, the histogram shows the average value of normalized α-TRP denseties from all experiments, relative to the control (w¹¹¹⁸) group. Statistical analysis was done using the Student’s t-test, in a two-tailed paired test. Data is presented as mean ± standard error of the mean (SEM), n = 6 or 7 for each group. A significant (*P ≤ 0.05) difference in α-TRP densities was found between the GMR/WT and WT groups, and between the GMR-DHDDS-RNAi and GMR/WT groups (***P ≤ 0.001). (C) A representative Western blot showing an example of a blot used in the histograms. Heads or eyes were lysed and run on gel. Following the protein transfer from the gel to a PVDF membrane, we dissected the membrane into three sections. Each section was probed with a different antibody: αRh1 (monoclonal, 1:1,000 dilution), αTRP (polyclonal, 1:1000 dilution), and α-dMoesin (polyclonal, 1:10,000 dilution). The density of the α-Rh1 and α-TRP bands were corrected by the dMoesin signal serving as a protein loading control, and calculated as a percentage of control fly (w¹¹¹⁸) signals. (D) A histogram showing the difference between Rh1 rhodopsin and TRP channel protein levels in the GMR-DHDDS-RNAi flies, and their levels in the GMR/+ flies. For each target protein (Rh1 rhodopsin or TRP channel) the mean difference between expression levels in the GMR/+ flies and expression levels in the GMR-DHDDS-RNAi flies was calculated. In each gel (n = 6), the expression level of each target protein (Rh1 rhodopsin or TRP channel) in the GMR/+ flies was set as 100%, from which the expression level of the same protein in the GMR-DHDDS-RNAi was subtracted. The mean difference between Rh1 rhodopsin levels in the GMR/+ flies and Rh1 rhodopsin levels in the GMR-DHDDS-RNAi flies was 97.2 ± 1.2% (SEM). The mean difference between the TRP protein levels in the GMR/+ flies and the mean difference of the GMR/+ was 72.1 ± 3.3% (SEM). A significant (***P ≤ 0.001, a two tailed, unpaired student’s t-test) difference was found between the mean difference of Rh1 rhodopsin levels and the mean difference of the TRP channel protein levels.