Identification of the santa maria gene. (A) Mapping of the santa maria mutation. The santa maria¹ gene was localized to 27F4-28A1, based on the failure of the following deficiencies to complement the santa maria phenotype: Df(2L)ED479 and Df(2L)Exel7031. CG12789 is indicated by the open box, and the other nine genes in this interval are indicated by the black boxes. The intron–exon structure of CG12789 is indicated below. The position of the mutation in CG12789 (GGT to GAT), which changes glycine 217 to an aspartic acid, is indicated. This mutation was identified by comparing the sequences in santa maria¹ and the original line used in the screen. (B) Alignment of the SANTA MARIA amino acid sequence with human SR-BI, human CD36, and NINAD. Identical residues, which are found in at least two proteins, are enclosed in black boxes. The running tallies of amino acids are indicated to the right. The two predicted transmembrane segments in SANTA MARIA are indicated by the black lines above the sequences. The position of the G217D change is indicated by the asterisk. (C–E) Rescue of the PDA phenotype in santa maria¹ by expression of a santa maria transgene under control of the heat shock protein 70 promoter (hs–santa maria). The ERG paradigm using orange (O) and blue light (B) is as described in Fig. 1 B. (C) PDA in wild type (w¹¹¹⁸). (D) Absence of a PDA in santa maria¹ flies. (E) A PDA was restored in santa maria¹;hs–santa maria/+ flies. (F) Rh1 expression increased in santa maria¹ flies upon expression of the hs–santa maria transgene (santa maria¹;hs–santa maria/+). We applied a 2-h heat shock to the flies immediately after eclosion. Head extracts were prepared from flies ∼2 d after eclosion and probed with anti-Rh1 and anti-tubulin antibodies.

santa maria but not ninaD appear to be functionally coupled with ninaB. (A–C) Western blots containing head extracts prepared from flies ∼2 d after eclosion were probed with anti-Rh1 and anti-tubulin antibodies. (A) Rh1 protein levels in ninaB^P315 flies after expression of UAS-ninaB under control of ninaB-GAL4, ninaD-GAL4, or santa maria–GAL4. (B) Rh1 protein levels in ninaD^P246 flies after expression of UAS-ninaD under the control of ninaB-GAL4, ninaD-GAL4, or santa maria–GAL4. (C) Rh1 protein levels in santa maria¹ flies after expression of UAS–santa maria under control of ninaB-GAL4, ninaD-GAL4, or santa maria–GAL4. (D–F) ERG paradigm to test for a PDA, as described in Fig. 1 B. All flies were in a w¹¹¹⁸ background. (D) santa maria¹,UAS–santa maria;santa maria–GAL4; (E) santa maria¹,UAS–santa maria;ninaB-GAL4; (F) santa maria¹, UAS–santa maria;ninaD-GAL4.

ninaD and santa maria cannot substitute for each other. (A) Expression of neither ninaD nor SR-BI restored Rh1 expression in santa maria¹ flies. The indicated UAS transgenes were expressed in santa maria¹;santa maria–GAL4 flies. The Western blot, processed as indicated in Fig. 5 C, was probed with anti-Rh1 and anti-tubulin antibodies. (B) The reduced Rh1 expression in ninaD^P246 flies was not rescued by expression of either santa maria or SR-BI. The indicated UAS transgenes were expressed in ninaD^P246;ninaD-GAL4 flies. The Western blot, processed as indicated in Fig. 5 C, was probed with anti-Rh1 and anti-tubulin antibodies.

Model of the pathway for production of vitamin A and the chromophore. Absorption of β-carotene into the gut cells is dependent on the NINAD scavenger receptor. The β-carotene is subsequently taken up into extra-retinal neurons and glia in the head via the SANTA MARIA scavenger receptor and cleaved by the NINAB BCO. Vitamin A is subsequently transported to the retinal pigment cells, where it is converted to the chromophore, through a process dependent on the PINTA retinoid binding protein (Wang and Montell, 2005) and the NINAG oxidoreductase (Ahmad et al., 2006). The chromophore is finally transported to the photoreceptor cells and binds to the opsin, resulting in the generation of rhodopsin. The proteins required for the indicated steps have not been identified (question marks).

Rhodopsin biosynthesis defects in santa maria¹. (A) Previously proposed model (Gu et al., 2004) suggesting that NINAB and NINAD are functionally coupled in extra-retinal neurons. (B–D) ERG paradigm that elicits PDA in wild-type but not in mutants that disrupt production of the major rhodopsin (Rh1). Flies (∼2 d after eclosion) were dark adapted for 1 min and subsequently exposed to five 5-s pulses of orange light (O) or blue light (B) interspersed by 7 s as indicated. A PDA is induced in wild type by blue light, as indicated by the arrows in B, and terminated by orange light. All the flies analyzed were in a w¹¹¹⁸ background. (B) Wild type (w¹¹¹⁸), (C) ninaE^P334, (D) santa maria¹. (E) Rh1 and Rh4 were reduced in santa maria¹ flies. The Western blot, which contained extracts prepared from the heads (∼2 d after eclosion) of wild-type, ninaE^P334, ninaD^P246, and santa maria¹ flies (all in a w¹¹¹⁸ background), were probed with anti-Rh1 and anti-tubulin antibodies. Molecular mass markers (kD) are indicated to the left. The same blot was reprobed with anti-Rh4 and anti-NORPA antibodies. The same samples were probed on a separate blot with anti–transient receptor potential (TRP) antibodies. (F) The ninaE mRNA levels were not affected in santa maria¹ flies. The Northern blot, which contained 2 μg of total RNA in each lane, was probed with ninaE DNA probe. Single-stranded RNA markers are indicated to the left.

santa maria is required for carotenoid metabolism outside of the retina. (A) santa maria functions outside of the compound eye for production of Rh1. The level of the Rh1 protein was normal in santa maria¹ mosaic eyes: santa maria¹ FRT/P[GMR-hid]G1 P[neoFRT]40 l(2)CL-L; P[GAL4-ey.H]SS5 P[UAS-FLP1.D]JD2. The samples were prepared from flies ∼2 d after eclosion, and the blot was probed with ant-Rh1 and anti-tubulin antibodies. (B) Expression of santa maria in different sets of retinal cells using the GAL4/UAS system (Brand and Perrimon, 1993) did not restore Rh1 levels. Expression of UAS–santa maria was driven ubiquitously in the eye using the GMR-GAL4 (Freeman, 1996) or specifically in pigment or photoreceptor cells using the CG7077-GAL4 (not depicted) or the ninaE-GAL4, respectively. To examine the level of Rh1, we prepared head extracts from flies ∼2 d after eclosion, and a Western blot was probed with anti-Rh1 and anti-tubulin antibodies. (C) Normal Rh1 levels were restored in santa maria¹ flies by feeding all-trans-retinal but not by feeding with β-carotene. The ninaB^P315, ninaD^P246, and santa maria¹ flies were fed either 0.2 mM all-trans-retinal or 0.2 mM β-carotene. To perform the Western blot, we prepared extracts from flies ∼4 d after eclosion and probed the filter with anti-Rh1 and anti-tubulin antibodies. (D–I) ERGs using a series of orange (O) and blue (B) light stimuli as de scribed in Fig. 1 B (see event markers at the bottom of G–I). All the flies were in a w¹¹¹⁸ background. Flies were fed normal food unless indicated otherwise. (D) wild-type; (E) santa maria¹; (F) santa maria¹,UAS–santa maria;santa maria–GAL4/+; (G) santa maria¹,UAS–santa maria;GMR-GAL4/+; (H) all-trans-retinal fed santa maria¹; (I) β-carotene fed santa maria¹.

ninaB and santa maria were expressed and functioned in neuronal and glia cells. (A) Examination of ninaB and santa maria expression in heads and bodies using a UAS-GFP reporter. Extracts, which were prepared from the heads and bodies of flies containing UAS-GFP and either the ninaB-GAL4 or the santa maria–GAL4, were fractionated by SDS-PAGE, and a Western blot was probed with anti-GFP antibodies. The same membrane was stained with Ponceau S to compare total proteins in each lane (not depicted). (B) Testing for colocalization of the santa maria– and ninaB-GFP reporters with neuronal and glial cell markers (anti-ELAV and anti-REPO, respectively) in adult heads. Because both REPO and ELAV are nuclear proteins, we used a GFP with a nuclear localization signal (GFP.nls). To conduct the experiments, sections of adult heads were prepared from ninaB-GAL4/UAS-GFP.nls and santa maria–GAL4/UAS-GFP.nls flies and stained with rabbit anti-GFP (green) and either rat anti-ELAV (red) or mouse anti-REPO (red). br, brain; co, cornea; la, lamina; me, medulla; re, retina. The arrowheads indicate one example in each set, in which there is overlap between the green (anti-GFP) and red (anti-ELAV or anti-REPO) signals. The labeling of the cornea is nonspecific because of autofluorescence. Similar patterns of anti-GFP staining were observed with different santa maria–GAL4 and ninaB-GAL4 lines, in combination with UAS-GFP.nls (not depicted). (C) Western blot showing that expression of UAS-ninaB or UAS–santa maria in neurons, using the elav-GAL4, restores Rh1 expression in the ninaB^P315 and santa maria¹, respectively. Head extracts were prepared from flies ∼2 d after eclosion and probed with anti-Rh1 and anti-tubulin antibodies. (D) Western blot demonstrating that expression of UAS-ninaB or UAS–santa maria in glia, under control of the repo-GAL4, increases Rh1 expression in ninaB^P315 and santa maria¹. The blot was performed as indicated in C.

ninaD was expressed and functioned in the gut. (A) GFP fluorescence in the midgut of ninaD-GAL4/UAS-GFP flies but not in flies containing only the UAS-GFP transgene. (B) Expression of ninaD was enriched in the gut. Extracts were prepared from the heads, bodies, guts, and bodies without guts of ninaD-GAL4/UAS-GFP flies and from the heads and bodies of UAS-GFP flies. The total proteins in each lane was detected by staining with Ponceau S (not depicted). The Ponceau S staining in the gut lane was lighter than in the other lanes because the same number of bodies and guts (rather than the same mass) was loaded into each lane. (C) Expression of UAS-ninaD using a midgut GAL4 line (drm-GAL4) rescued the ninaD^P246 phenotype. The Western blot, which contained head extracts prepared from flies ∼2 d after eclosion, was probed with anti-Rh1 and anti-tubulin antibodies.