Alpha-COPI coatomer protein is required for rough endoplasmic reticulum whorl formation in mosquito midgut epithelial cells

Abstract

Background: One of the early events in midgut epithelial cells of Aedes aegypti mosquitoes is the dynamic reorganization of rough endoplasmic reticulum (RER) whorl structures coincident with the onset of blood meal digestion. Based on our previous studies showing that feeding on an amino acid meal induces TOR signaling in Ae. aegypti, we used proteomics and RNAi to functionally identify midgut epithelial cell proteins that contribute to RER whorl formation.

Methodology/principal findings: Adult female Ae. aegypti mosquitoes were maintained on sugar alone (unfed), or fed an amino acid meal, and then midgut epithelial cells were analyzed by electron microscopy and protein biochemistry. The size and number of RER whorls in midgut epithelial cells were found to decrease significantly after feeding, and several KDEL-containing proteins were shown to have altered expression levels. LC-MS/MS mass spectrometry was used to analyze midgut microsomal proteins isolated from unfed and amino acid fed mosquitoes, and of the 127 proteins identified, 8 were chosen as candidate whorl forming proteins. Three candidate proteins were COPI coatomer subunits (alpha, beta, beta'), all of which appeared to be present at higher levels in microsomal fractions from unfed mosquitoes. Using RNAi to knockdown alpha-COPI expression, electron microscopy revealed that both the size and number of RER whorls were dramatically reduced in unfed mosquitoes, and moreover, that extended regions of swollen RER were prevalent in fed mosquitoes. Lastly, while a deficiency in alpha-COPI had no effect on early trypsin protein synthesis or secretion 3 hr post blood meal (PBM), expression of late phase proteases at 24 hr PBM was completely blocked.

Conclusions: alpha-COPI was found to be required for the formation of RER whorls in midgut epithelial cells of unfed Aa. aegypti mosquitoes, as well as for the expression of late phase midgut proteases.

Figure 1. Amino acid feeding is sufficient to induce RER whorl unwinding.

Three-day-old Ae. aegypti females were kept unfed (sugar fed), or fed an amino acid meal for 30 min., and then sacrificed and dissected 30 min. or 120 min. later. The dissected midguts were fixed for electron microscopy preparation as described in Materials and Methods and representative electron micrographs of midgut sections are shown here. A) Unfed mosquito (25,000× magnification). B) Unfed mosquito (8,800× magnification). C) Amino acid fed mosquito dissected at 30 min. post-feeding (8,800× magnification). D) Amino acid fed mosquito dissected at 120 min. post-feeding (8,800× magnification).

Figure 2. Quantitation of whorl size and number in unfed and amino acid fed mosquitoes.

Whorls with a minimum of five stacked ER membranes were counted and the total area of each whorl was determined using NIH Image software. Data are shown for whorl size and number in 20 random EM fields covering 3335 µm² of mosquito midgut epithelial cell area. A) Size and number of RER whorls in 20 random fields from unfed and amino acid fed mosquitoes after 30 min. and 120 min. B) Total whorl in the same sample as shown in A. Statistical analysis using Pearson Chi-square test revealed that the total whorl area in amino acid fed mosquitoes at both 30 min and 120 min post-feeding was significantly less than the total whorl area in unfed mosquitoes (bars with different letters signify significance at p<0.001 comparing 30 min and 120 min to unfed mosquitoes).

Figure 3. Western blot analysis using a KDEL-specific antibody shows differential expression of four ER resident midgut proteins in response to amino acid feeding in Ae. aegypti mosquitoes.

Total midgut protein extracts were prepared from three day old female mosquitoes that were unfed, or amino acid fed and dissected at 30 min. or 120 min. post-feeding. Proteins were resolved by 12% SDS-PAGE gel and Western blotted as described in Materials and Methods. The four unique protein bands were labeled KDEL-1 through KDEL-4 to denote their relative molecular weight, with KDEL-1 being the largest protein. Western blotting with an antibody that recognizes the ubiquitously expressed glyceraldehyde-3P dehydrogenase protein (GAPDH) was performed to control for equal protein loading.

Figure 4. Isolation of microsomal proteins for LC-MS/MS analysis from unfed and amino acid fed Ae. aegypti mosquitoes.

Three-day-old female mosquitoes were unfed or amino acid fed and dissected after 30 min. Microsomal protein samples were prepared as described in Materials and Methods and resolved on a 12% SDS-PAGE gel. Following protein staining, four equal slices of polyacrylamide gel were prepared from each lane (A, B, C, or D) and proteins were subjected to in-gel trypsin digestion prior to LC-MS/MS analysis. Note that minimal differences exist in the quality and quantity of abundant proteins in the two protein samples, suggesting that the 30 min. time point is prior to the onset of feeding-induced protein synthesis.

Figure 5. QRT-PCR analysis of alpha-COPI transcript levels in midgut tissue of individual Ae. aegypti mosquitoes that were either uninjected, or injected with dsRNA targeted against the control Fluc gene or the alpha-COPI gene.

Transcript analysis revealed that 100% of the alpha-COPI dsRNA injected mosquitoes had <10% the level of alpha-COPI transcripts present in uninjected or Fluc dsRNA injected mosquitoes.

Figure 6. Knock down of alpha-COPI expression is associated with disorganization of RER whorls in unfed mosquitoes.

Representative electron micrographs are presented here. A) Unfed Fluc dsRNA injected mosquito (25,000× magnification). B) Unfed Fluc dsRNA injected mosquito (8,800× magnification). C) Unfed alpha-COPI dsRNA injected mosquito (25,000× magnification). D) Unfed alpha-COPI dsRNA injected mosquito (8,800× magnification).

Figure 7. Knock down of alpha-COPI expression is associated with disorganization of endosomal structures and appearance of extended regions of swollen RER in amino acid fed mosquitoes.

Representative electron micrographs are presented here. A) Fluc dsRNA injected mosquito dissected 120 min. after amino acid feeding (25,000× magnification). B) Fluc dsRNA injected mosquito dissected 120 min. after amino acid feeding (8,800× magnification). C) alpha-COPI dsRNA injected mosquito dissected 120 min. after amino acid feeding (25,000× magnification). D) alpha-COPI dsRNA injected mosquito dissected 120 min. after amino acid feeding (8,800× magnification).

Figure 8. Quantitation of whorl size and number in dsRNA injected unfed and amino acid fed mosquitoes.

Whorls with a minimum of five stacked ER membranes were counted and the total area of each whorl was determined as described in figure 2 legend. A) Size and number of RER whorls in 20 random fields from unfed and amino acid fed mosquitoes after 120 min. that were injected with Fluc or alpha-COPI dsRNA. B) Total whorl in the same sample as shown in A. Statistical analysis using Pearson Chi-square test revealed that the total whorl area in alpha-COPI dsRNA injected unfed and amino acid fed mosquitoes was significantly less than the total whorl area in Fluc dsRNA injected unfed mosquitoes (bars with different letters signify significance at p<0.001). In addition, the total whorl area in Fluc dsRNA injected fed mosquitoes was found to be significantly less than in Fluc dsRNA injected unfed mosquitoes using the Pearson Chi-square test (bars with different letters signify significance at p<0.001).

Figure 9. A deficiency in alpha-COPI inhibits feeding-induced expression of late phase midgut proteases.

A) Representative Western blot of AeET protein expression in the midguts of mosquitoes injected with 400 ng of Fluc or alpha-COPI dsRNA and maintained on sugar for 3 days and then dissected (unfed), or blood fed after 3 days of sugar feeding and dissected at 3 hr PBM (fed). The GAPDH antibody was used as a protein loading control. Each lane contains the same midgut equivalents obtained from pooled mosquitoes. B) Representative Western blots of protein extracts prepared from pooled mosquito midguts dissected at 24 hr PBM and analyzed with antibodies against late phase serine proteases (AaSPVI, AaSPVII, AaLT) or GAPDH. Mosquitoes were injected with the indicated amount of dsRNA 3 days prior to blood feeding.