Drosophila KDEL receptor function in the embryonic salivary gland and epidermis

Abstract

Core components of the secretory pathway have largely been identified and studied in single cell systems such as the budding yeast S. cerevisiae or in mammalian tissue culture. These studies provide details on the molecular functions of the secretory machinery; they fail, however, to provide insight into the role of these proteins in the context of specialized organs of higher eukaryotes. Here, we identify and characterize the first loss-of-function mutations in a KDEL receptor gene from higher eukaryotes. Transcripts from the Drosophila KDEL receptor gene KdelR - formerly known as dmErd2 - are provided maternally and, at later stages, are at elevated levels in several embryonic cell types, including the salivary gland secretory cells, the fat body and the epidermis. We show that, unlike Saccharomyces cerevisiae Erd2 mutants, which are viable, KdelR mutations are early larval lethal, with homozygous mutant animals dying as first instar larvae. KdelR mutants have larval cuticle defects similar to those observed with loss-of-function mutations in other core secretory pathway genes and with mutations in CrebA, which encodes a bZip transcription factor that coordinately upregulates secretory pathway component genes in specialized secretory cell types. Using the salivary gland, we demonstrate a requirement for KdelR in maintaining the ER pool of a subset of soluble resident ER proteins. These studies underscore the utility of the Drosophila salivary gland as a unique system for studying the molecular machinery of the secretory pathway in vivo in a complex eukaryote.

Figure 1. Drosophila KdelR is conserved.

(A) Drosophila KdelR is highly conserved with respect to its vertebrate and C. elegans counterparts (A) and is homologous to S. cerevisiae Erd2. (Mm-mouse, Hs-human , Gg-chicken, Xl-Xenopus, Dr-zebrafish, Dm-Drosophila, Ce- C. elegans and Sc- S. cerevisiae). Vertebrates encode two-three Kdel Receptors, whereas only a single gene is found in flies, worms and yeast. Black bars over the sequences indicate membrane spanning regions [3]. Red dots denote residues involved in ligand (KDEL) binding [4]. (Green= completely conserved residues, yellow= identical residues, purple=similar residues). (B) A Phylip unrooted tree analysis of the KDEL Receptors from the major model organisms reveals that the Drosophila Kdel Receptor is slightly more related to the vertebrate proteins than are the worm and yeast receptors.

Figure 2. KdelR expression profile.

KdelR is detected early in embryogenesis at the cellular blastoderm stage (A,E) and to high levels in the salivary gland beginning at embryonic stage 10 and continuing throughout embryogenesis (b-d, f-h, black arrows). KdelR is also expressed to elevated levels in the epidermis (cells on the embryo surface), fat body (clusters of staining in each segment in stage 11 embryos, indicated by black stars), proventriculus (white star) and a subset of gut endoderm cells (arrowheads). b-gal and RNA staining in the neural tube (nt) is also observed at late stages, although the RNA expression in the nt is not in the plane of focus of the embryos shown. Left column shows mRNA expression detected with a probe made from the KdelR cDNA clone CK00230 and the right column shows β-gal staining of l(2)k00311embryos.

Figure 3. Identification of KdelR alleles.

KdelR maps to region 31E in the Drosophila genome and l(2)k00311 is inserted in the 5’ UTR of the KdelR transcript. l(2)k00311 fails to complement deficiencies Df(2L)J3, Df(2L)J106 and Df(2L)J27, but complements deficiencies Df(2L)J17 and Df(2L)J16. Both EMS alleles of KdelR (31Em1 and 31Em ²) encode ORFs with premature stop codons.

Figure 4. KdelR mutant cuticles are smaller than wild-type cuticles and are grossly underdeveloped.

Dark field (ventral) images of wild-type, l(2)k00311 and 31Em ² larvae (A-C). Note that l(2)k00311 and 31Em ² mutants are approximately 60% the length of wild type larvae and the ventral denticles are not as prominent as in their wild-type siblings (A-C). Mouthparts (MP) of l(2)k00311 (G) and 31Em ² (J) are underdeveloped and less pigmented than corresponding wild-type mouth parts (D). The filtzkörper (FK) of l(2)k00311 (I) and 31Em ² (L) are underdeveloped and do not protrude from the larval body as in wild-type (F).

Figure 5. KdelR and the retention of PH4αSG1 and other soluble resident proteins in the ER.

(A) Staining with antibodies to the resident ER protein PH4αSG1 revealed cellular expression in WT SGs beginning at embryonic stage 11 and continuing through embryogenesis (data not shown; left panels). By embryonic stage 13, PH4αSG1 staining was observed at low levels in the salivary lumens of l(2)k00311 mutants (middle top panel) and at high levels in the salivary lumens of l(2)31Em ¹ and l(2)31Em ² mutants (top right panel and data not shown). By stage 17, almost all detectable PH4αSG1 was lumenal in both l(2)k00311 and l(2)31Em ¹ salivary glands. The same staining patterns were observed in all embryos examined of each genotype. (B) Stage 15 embryonic salivary glands were co-stained with antibodies to PH4αSG1, Boca, and Windbeutel. PH4αSG1 protein was observed entirely in the ER in wild-type salivary glands, in both the ER and lumen in l(2)k00311 salivary glands, and predominantly in the lumen in 31Em ² mutant salivary glands (left panels). Boca protein was barely detected in the salivary glands of 31Em ² mutants, compared to the WT and l(2)k00311 mutants, although some Boca protein can be detected in the lumens of 31Em ² mutants when the image is overexposed (second column, last row, inset). Wbl localization was largely unaffected in l(2)k00311 and 31Em ² mutant salivary glands; only minimal lumenal staining of Wbl protein was observed, even when the cellular staining was at high levels (third column, second row). The changes in PH4αSG1 and Boca protein localization are more apparent in the merged images (last column). Again, the same patterns of accumulation were observed in all of the stage 15 embryos examined for each genotype.

Figure 6. Boca and Wbl require C-terminal KDEL/KEEL sequences for ER retention in S2 cells

(A) Immunoblots of cell pellets (P) and supernatants (S) from Drosophila S2 cells transfected with empty vector (left lanes), full length Boca (center lanes) or KDEL deleted Boca (right lanes) incubated with aBoca (top gel) or abTub antibodies (bottom gel). Note that Boca protein is easily detected in the cell pellets from all three samples but that Boca protein is also detected in the supernatant in only cells transfected with the KDEL-deleted Boca construct. Note also that bTub is detected in only the cell pellets from each transfected cell type, as expected for a cytosolic protein. (B) Immunoblots of cell pellets (P) and supernatants (S) from Drosophila S2 cells transfected with empty vector (left lanes), full length Wbl (center lanes) or KEEL-deleted Wbl (right lanes) incubated with aWbl (top gel) or abTub antibodies (bottom gel). Note that Wbl protein is detected in the cell pellets from S2 cells transfected with the full length Wbl construct, whereas Wbl protein is detected in the supernatant from S2 cells transfected with the KEEL-deleted Wbl construct. Note also that bTub is again detected in only the cell pellets from each transfected cell type.