Predicted seminal astacin-like protease is required for processing of reproductive proteins in Drosophila melanogaster

Abstract

During mating, males provide females with seminal fluids that include proteins affecting female physiology and, in some cases, reproductive behavior. In several species these male-derived modulators of reproduction are processed upon transfer to the female, suggesting molecular interaction between the sexes. Males could increase their reproductive success by contributing to regulation of this processing; consistent with this hypothesis, seminal fluids are rich in proteolysis regulators. However, whether these molecules carry out processing of male-derived reproductive modulators is unknown. We tested for this role using RNAi to knock down individually 11 Drosophila seminal fluid proteases and protease inhibitors. We found that CG11864, a predicted astacin-type metalloprotease in seminal fluid, is necessary to process two other seminal proteins: the ovulation hormone ovulin (Acp26Aa) and the sperm storage protein Acp36DE. This processing occurs only after all three proteins have entered the female. Moreover, CG11864 itself is processed inside males while en route to the female and before its action in processing ovulin and Acp36DE. Thus, processing of seminal proteins is stepwise in Drosophila, beginning in the male after the proteins leave their site of synthesis and continuing within another organism, the mated female, and the male-donated protease CG11864 is an agent of this latter processing.

Fig. 1.

Processing of ovulin, Acp36DE, and CG11864 in the male ejaculatory duct and bulb (lane ED & EB) or in the female reproductive tract (lane ♀-RT). At 8 min ASM only full-length ovulin is seen. Acp36DE shows the full-length 122-kDa band and a cross-reactive band at 64 kDa (36) (data not shown) at 8 min ASM in the ED & EB samples, and the 122- and 64-kDa bands and a 68-kDa processed product in the mated female. Previously, the cross-reactive 64-kDa band was suggested to be derived from the female (36), but our result clearly shows that it is a male-derived non-Acp (thus, from either sperm or other seminal proteins cross-reacting with Acp36DE antibody), because it is seen in ejaculatory duct/bulb samples. For CG11864 both the full-length protein (32 kDa) and the processed 30-kDa band (35) are detected in male and female (ED & EB and ♀-RT) by 8 min ASM. Protein equivalent to eight male ejaculatory ducts/bulbs or four mated female reproductive tracts from 8 min ASM or 10 min ASM were loaded in each lane.

Fig. 2.

Alignment of the N-terminal sequence predicted for the secreted form of CG11864 with the propeptide cleavage region of astacin and other astacin-like metalloproteases. Astacin family members are cleaved at ≈30 aa from N terminus (42) of their secreted form at a basic region (boxed). Based on the sequence alignment and its basic residue cluster, CG11864's propeptide cleavage site is predicted to be after amino acid 45 in the CG11864 sequence [amino acid 27 in the figure; because CG11864's signal peptide is predicted to be amino acids 1–18 (as per by SignalP 3.0), amino acid 19 is indicated as “1” in this figure]. This size is consistent with the 3-kDa difference seen between full-length and processed CG11864 (35). Alignments were done by using the Clustal W algorithm of the Megalign program (Lasergene; DNASTAR) and used the sequences of astacin-family members from the GenBank database (accession nos. CAA64981, CAA46637, 2112204A, AAC46482, P42664, and CAA70854).

Fig. 3.

Processing of ovulin in mates of RNAi and control males. (A) Processing of ovulin at 45 min ASM in females mated to knockdown males (lane RNAi) of different Acp protease lines in comparison with their controls (lane Control). Female reproductive tracts were dissected at 45 min ASM, and protein equivalents of one pair of accessory glands (lane ♂ in this and subsequent figures) or two female reproductive tracts were loaded in each lane and processed for Western blotting with anti-ovulin. (B) Analysis of female reproductive tract samples from later time points after mating confirm that ovulin (and similarly Acp36DE; data not shown here) processing is blocked in females mated to CG11864 knockdown males. Mated female reproductive tracts were dissected at the indicated times ASM from mates of knockdown males (lane CG11864 RNAi) or control males (lane CG11864 control) and probed with anti-ovulin. Samples from females mated to males knocked down for a different Acp that is not a proteolysis regulator (CG1656) were run as a control to rule out the effects of the tubulin-GAL4 background.

Fig. 4.

Processing of ovulin in females mated to knockdown males (lane RNAi) of different Acp protease inhibitor lines in comparison with their controls (lane Control) at 15 min ASM (A) and 45 min ASM (B) (left side). To test whether knockdown of any of the protease inhibitors results in the premature processing of Acps in males, protein samples from male accessory glands (AG) were probed from knockdown males (RNAi) and control males. Analogous to the results shown here, the processing of Acp36DE and CG11864 was also not affected in any of the protease inhibitor knockdown males and their mates (data not shown). Protein equivalents of one pair of male accessory glands or two female reproductive tracts were loaded in each lane.

Fig. 5.

Example Western blots (A) and RT-PCR (B) showing the levels of Acp knockdown in experimental males (RNAi) compared with control males. (A) In Western blots, the levels were quantified by running serial dilutions to the level of 2.5% (lane 1), 5% (lane 2), 10% (lane 3), 15% (lane 4), 20% (lane 5), and 25% (lane 6) of Acps from control male (100%; lane 7) in parallel with Acps equivalent to one experimental male (RNAi). In all but one case, knockdowns were specific to the targeted Acp. The one exception was for the sequence related protease inhibitors (CG8137 and CG9334). Knockdown of CG8137 knocks down CG9334 and vice versa; an example is shown (CG8137 in CG9334 RNAi). (B) For PCR amplification, cDNA prepared from RNA extracts of 20 control males or experimental males was used. RP49 primers (48) were used as positive control for the quality/quantity of cDNA. The figure shows exemplar of data for one protease and one protease inhibitor. We observed similar results for all other proteases and protease inhibitors tested.