Genes were expressed in Drosophila through the standard GAL4/UAS system (16). The following flies and DNA have been described previously: armadillo-GAL4 (17), uas-myc-EDEM1 (14), uas-myc-EDEM2 (14) and uas-NHK (14). The DNA encoding ATZ (18) was subcloned into a pUAST plasmid.

Drosophila Schneider 2 (S2) cells were grown in Schneider’s medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (Invitrogen, Grand Island, NY, USA). Treatment with double-stranded RNA (dsRNA) was performed as described in a previous study (19). Briefly, the cells were plated into 6-well plates at a density of 1×10⁶ cells/well before treatment with dsRNA (day 1). After 24 h, 20 μg of EGFP, EDEM1 or EDEM2 dsRNA were added to each well following another boost with 20 μg dsRNA on day 4. The cells were then transiently transfected with either NHK or ATZ using Effectene™ (Qiagen, Valencia, CA, USA) on day 5. The cells were split on day 8 and lysed to examine the level of NHK or ATZ on day 9. The EDEM1 dsRNA consisted of a 516-nt region (Amplicon ID: DRSC18573), as described by the Drosophila RNAi Screening Center (http://www.flyrnai.org). The following primers were used to amplify this sequence from an embryo cDNA library: ‘R’ primer, CAATGTTGTCACCCACGAAA; ‘S’ primer, TCGAAGTT GCTTACTAACAGA. This amplicon has no predicted off-targets. The EDEM2 dsRNA consisted of a 520-nucleotide region (amplicon ID: DRSC02877). The following primers were used to amplify this sequence from an embryo cDNA library: ‘R’ primer, 5′-CATGCGCGGGTTAAT CTC-3′; ‘S’ primer, 5′-GATAGAGCATCTCGTGTGTC-3′. To induce ER stress, Drosophila S2 cells were treated with dithiothreitol (DTT; Cat. no. 43815; Sigma, St. Louis, MO, USA) or thapsigargin (Tg; Cat. no. T9033; Sigma) for the indicated periods of time.

All fluorescent images were captured under a Zeiss LSM 510 confocal microscope, using x100 objective lenses. The following antibodies were used: guinea pig anti-Hsc3 antibody (1:500) as previously described (17), mouse anti-myc (1:1,000 for immunohistochemistry; 9E10; Cat. no. 11667149001; Roche Diagnostics GmbH, Mannheim, Germany), rhodamine red anti-mouse secondary antibody (Cat. no. 715-295-150; 1:500) and FITC anti-guinea pig secondary antibody (Cat. no. 706-095-148; 1:500) (both from Jackson ImmunoResearch, West Grove, PA, USA).

For western blot analyses, Drosophila S2 cells were extracted with 1% SDS lysis buffer (10 mM Tris pH 7.5, 1 mM EDTA, 150 mM NaCl and 1% SDS; Sigma). Following centrifugation at 16,100 x g for 10 min, the supernatants were fractionated by SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Billerica, MA, USA). The membranes were then probed with the indicated antibodies: polyclonal rabbit anti-A1AT (1:5,000 for western blot analysis; Cat. no. A0012; Dako, Glostrup, Denmark), mouse anti-profilin (1:2,000 for western blot analysis; Developmental Studies Hybridoma Bank, chi 1J, University of Iowa, Iowa City, IA, USA), rat anti-HA antibody (Cat. no. 11867423001; 1:1,000; Roche Diagnostics GmbH) and rabbit anti-GFP antibody (1:5,000; Cat. no. A6455; Molecular Probes, Eugene, OR, USA). The fractionation of the Drosophila S2 cells was performed as previously described (20). For immunoprecipitation, the Drosophila S2 cells were extracted with 1% Triton X-100 lysis buffer (50 mM Tris pH 8.0, 150 mM NaCl, digitonin and 1% Triton X-100; Sigma) for 20 min on ice, and centrifuged at 16,100 × g. The supernatant was quantified by Bradford assay, and used for immunoprecipitation. Immunoprecipitation was performed with anti-A1AT antibody and protein-G-agarose beads (Roche Diagnostics GmbH). The beads were washed 3 times in low-ionic-strength buffer (50 mM Tris-Cl pH 8.0, 100 mM NaCl and 1% Triton X-100; Sigma). Rat anti-HA antibody was used to detect the ubiquitination of A1AT. For quantification of western bands, we used ImageJ software (http://rsbweb.nih.gov/ij). The intensity of the band of interest was normalized with an anti-profilin band.

Total RNA was isolated using TRIzol reagent, and reverse transcription was performed from 200 ng of total RNA using the SuperScript First-Strand Synthesis kit (both from Invitrogen). The following primer sequences were used for amplification and quantification: dEDEM1-F, ACGCCTACGATGGTTACCTG; dEDEM1-R, ACACGTTGATGTCCCTGTCA; dEDEM2-F, CTTAGCACCGAAACCACCAT; dEDEM2-R, ACTCCTCGGTACCGTCCTTT.

To determine the subcellular distribution of EDEMs in Drosophila, we generated transgenic EDEM lines with epitope tags and expressed them in Drosophila embryo amnioserosa cells. Immunolabeling with anti-myc antibody revealed that the Drosophila EDEMs co-localized with Hsc3, the Drosophila orthologue of mammalian BIP. This result is consistent with the hypothesis that Drosophila EDEMs reside in the ER (Fig. 2A and B).

Mammalian EDEMs are stress-regulated proteins that are induced by ER stress (21,22). To determine whether the Drosophila EDEM homologs are similarly regulated, we treated the Drosophila S2 cells with dithiothreitol (DTT), which imposes ER stress by reducing disulfide bonds. Under these conditions, Drosophila EDEM2 expression was transcriptionally induced (Fig. 2C). Similar results were obtained with independent ER stress-causing chemicals; tunicamycin (10 μg/ml), which inhibits the glycosylation of proteins in the ER and thapsigargin (Tg; 1 μM), which perturbs ER-calcium homeostasis (data not shown). On the other hand, we were not able to detect the induction of Drosophila EDEM1 under these conditions.

The degree of ER stress in cells can be indirectly assessed by the extent of the transcriptional response that induces ER chaperones and ERAD genes. To examine the protective role of Drosophila EDEMs under ER stress conditions, we treated the Drosophila S2 cells with dsRNAs that target either EGFP (as a control), EDEM1 or EDEM2, or both EDEM1 and EDEM2 and subsequently exposed them to Tg. The level of Hsc3 increased after 4 h, and this increase was even more pronounced when both EDEM1 and EDEM2 were knocked down (Fig. 2D). These results suggest that Drosophila EDEMs play protective roles against ER stress.

Excessive protein misfolding in the ER triggers the activation of signaling pathways referred to as the unfolded protein response (UPR). One such pathway involves the mRNA splicing of X-box binding protein 1 (XBP1), which causes a shift in the reading frame of the downstream sequences and the generation of a distinct protein isoform (17,23). We have previously exploited this property to develop a UPR sensor, XBP1-EGFP, in which EGFP is expressed in frame when UPR is activated (17). We thus employed this UPR sensor to assess whether the mutant variants of A1AT cause ER stress in Drosophila. As we have reported previously (14), this UPR marker was activated by NHK expression (Fig. 3A, lane 4) to a similar extent as that induced by DTT treatment (Fig. 3A, lane 6). Intriguingly, ATZ expression in the Drosophila S2 cells did not trigger XBP1 mRNA splicing (Fig. 3A, lane 5). Previous studies on mammalian cells have also reported that, for some reason, ATZ does not activate UPR; instead, ATZ expression activates NF-κB signaling via ER overload response (EOR) and ERK signaling (24–26). To determine whether this difference is derived from the solubility of misfolded A1AT, we simply fractionated the cell extracts as supernatants and pellets. We found that the ratio of ATZ protein in the soluble versus the insoluble fraction was roughly 1:1 (Fig. 3B). On the other hand, the majority of NHK protein was found in the soluble fraction (Fig. 3C). These observations support the hypothesis that the two disease alleles of A1AT have distinct biochemical properties.

We have previously demonstrated that Drosophila EDEM1 and EDEM2 are homologs of mammalian EDEM2 and EDEM3, respectively. Moreover, we demonstrated that the overexpression of Drosophila EDEMs helps to lower the levels of the A1AT NHK variant (14). In this study, to investigate whether Drosophila EDEMs also show distinct specificity toward two misfolded A1AT variants, we examined the effects of EDEM1 and EDEM2 on the degradation of the misfolded ATZ and NHK variants. The level of ATZ increased by approximately 2-fold after the knockdown of EDEM1 and EDEM2 by dsRNA in Drosophila S2 cells (Fig. 4A and B). Similarly, the knockdown of Drosophila EDEM1 and EDEM2 increased the levels of another misfolded A1AT variant, NHK (Fig. 4D and E). Of note, a slightly higher molecular weight band of ATZ was detected after the knockdown of either Drosophila EDEM1 or EDEM2. As the deglycosylation of ERAD substrates occur after being dislocated back into the cytoplasm (27), the slower migrating A1AT band suggests defective ERAD and the accumulation of glycosylated ATZ species that accumulate in the ER.

We also overexpressed EDEMs and found that mutant A1AT degradation was accelerated by the overexpression of EDEMs. Intriguingly, the two EDEMs from Drosophila had additive effects on the degradation of both the NHK (Fig. 5A) and ATZ variants of A1AT (Fig. 5A). Subsequently, we wished to determine whether Drosophila EDEM1 or EDEM2 affects the solubility of ATZ or NHK. We did not observe any significant change in the solubility of ATZ or NHK by knocking down EDEM1 or EDEM2 (Fig. 5B and C). These results indicate that Drosophila EDEMs regulate the degradation of misfolded A1AT variants without affecting the solubility of misfolded A1AT.

Previous studies have suggested that ATZ can be degraded by either the ubiquitin-proteasomal pathway, or through autophagy (13,28,29). In order to further confirm that the Drosophila EDEMs act by stimulating the ubiquitin proteasomal degradation of misfolded A1AT variants, we co-expressed Drosophila EDEMs and ATZ with HA-tagged ubiquitin. The co-expression of EDEM1 or EDEM2 with ATZ increased the level of ubiquitinated ATZ (Fig. 5D). The levels of ubiquitinated NHK also increased, albeit to a different extent than that observed for ATZ (Fig. 5E). These results again suggest that Drosophila EDEM1 and EDEM2 target misfolded A1AT variants for proteasomal degradation.