Aubergine is a component of a nanos mRNA localization complex

Abstract

Localization of nanos (nos) mRNA to the posterior pole of the Drosophila oocyte is essential for abdominal segmentation and germline development during embryogenesis. Posterior localization is mediated by a complex cis-acting localization signal in the nos 3' untranslated region that comprises multiple partially redundant elements. Genetic analysis suggests that this signal is recognized by RNA-binding proteins and associated factors that package nos mRNA into a localization competent ribonucleoprotein complex. However, functional redundancy among localization elements has made the identification of individual localization factors difficult. Indeed, only a single direct-acting nos localization factor, Rumpelstiltskin (Rump), has been identified thus far. Through a sensitized genetic screen, we have now identified the Argonaute family member Aubergine (Aub) as a nos localization factor. Aub interacts with nos mRNA in vivo and co-purifies with Rump in an RNA-dependent manner. Our results support a role for Aub, independent of its function in RNA silencing, as a component of a nos mRNA localization complex.

Figure 1

Identification of aub as a nos mRNA localization factor. (A) Schematics of the 3′UTR of wild-type nos mRNA and the sensitized nos+1+3 transgene. The four previously defined localization signal elements are indicated, the black bar represents the remainder of the nos 3′UTR. (B) Bar chart showing the average number of abdominal segments in embryos from wild-type females (8 segments, n = 106), nos+1+3 females (5.8 segments, n = 332), and nos+1+3 females that are heterozygous for Df(2L)BSC32 (2.4 segments, n = 70), aub^QC42 (n = 64), aub^HN2 (n=145), or squ^HE47 (6.2 segments, n = 216). Df(2L)BSC32, aub^QC42, and aub^HN2 significantly reduce segment number in nos+1+3 derived embryos (*P < 0.0001). Error bars indicate SEM.

Figure 2

aub affects localization of a sensitized nos mRNA. (A, B) Localization of wild-type nos or nos+1+3 mRNA was evaluated by in situ hybridization to embryos from wild-type females (WT), females heterozygous for aub^QC42 (aub⁻/+) and nos+1+3 females heterozygous for aub^QC42 (aub⁻/+; nos+1+3). A typical example of the nos RNA localization pattern in each genotype is shown (A). Localization was classified as wild-type (+++), moderate (++), weak (+), or undetectable (−). Localization of wild-type nos is unaffected by heterozygosity for aub whereas localization of the nos+1+3 transgene is further compromised by reducing aub (wild-type, n = 200; aub⁻/+, n = 97; nos+1+3, n = 305; aub⁻/+; nos+1+3, n = 341). (C) Northern blot of total mRNA isolated from embryos of the indicated genotypes, probed simultaneously for nos and for rp49 as a loading control. The blot was stripped and reprobed for osk. The relative nos and osk mRNA levels for each sample pair, after normalization to rp49, are indicated below. (D) In situ hybridization to bcd mRNA embryos from wild-type or aub^QC/+ females. Heterozygosity for aub does not affect bcd mRNA localization.

Figure 3

osk regulation is unaffected in aub heterozygotes. (A) In situ hybridization to osk mRNA in embryos from wild-type or aub^QC/+ females. (B) Immunoblot of total protein from wild-type and aub^QC/+ embryos. After transfer, the membrane was cut at the 37 kDa marker. The top portion was blotted with anti-Osk, which recognizes both the short and long Osk isoforms, and the bottom with anti-Snf as a loading control. Quantitation of the blot showed that Osk expression is not affected by heterozygosity for aub. (C) Confocal projections of anti-Vas immunofluorescence (red) in wild-type or aub^QC/+ embryos. Nuclei are labeled with DAPI (blue). (D) Scatterplot of the number of pole cells per embryo in wild-type (n = 54) or aub^QC/+ (n = 31) embryos, with the mean indicated by a horizontal bar. The average number of pole cells is not significantly different between wild-type (30.2 ± 6.7) and aub^QC/+ (n = 29.2 ± 5.0) as determined by the Student’s t-test.

Figure 4

The effect of aub on nos is independent of the DNA damage response pathway. (AC) Immunoblots of total protein extracted from 0–2 hour embryos of the indicated genotypes, probed with either anti-Nos or anti-Osk antibodies. Each membrane was also probed with anti-Snf or anti-Khc to control for loading. (A) Extracts from wild-type, mnk, aub, and mnk, aub double mutant embryos. Nos levels are reduced in aub and mnk, aub mutant embryos, however the reduction in Osk levels observed in aub mutants is partially suppressed by eliminating mnk. (B) Extracts from wild-type or nos^BN embryos, which lack nos mRNA, confirming specificity of the anti-Nos antibody as previously shown (Gavis et al., 2008). (C) Extracts from wild-type or osk^A87 heterozygous embryos, which contain half the wild-type level of osk mRNA. Quantitation of the blot shows that Nos protein levels are reduced to a lesser extent than Osk (70% versus 50% of wild-type levels, respectively).

Figure 5

Aub associates with nos mRNA and Rump in vivo. (A) RT-PCR to detect nos or osk mRNA co-immunoprecipitated from ovaries expressing either GFP alone (G) or GFP-Aub (A) using anti-GFP antibody. his3.3b serves as a negative control. Total RNA from the extracts used for immunoprecipitation was used as a positive control for the RT-PCR reaction. Reactions were performed in the presence (+RT) or absence (−RT) of reverse transcriptase. (B) Immunoblot of anti-GFP immunoprecipitates from ovaries expressing either MCP-GFP or GFP-Aub in the absence (−) or presence (+) of RNase. Duplicate immunoblots were prepared for detection with either anti-Rump or anti-GFP antibodies. Lanes with extract used for immunoprecipitation (E) contain 1/20 volume equivalent of the immunoprecipitate sample (IP). (C) Bar graph with results from a genetic interaction assay for aub and rump. The percentage of embryos developing with 8 or fewer than 8 (≤8) abdominal segments in rump¹ embryos (10%, n = 185), rump¹ embryos heterozygous for aub^QC (31%, n = 326), rump¹ embryos heterozygous for nos^BN (27%, n = 391), or rump¹ embryos heterozygous for both aub^QC and nos^BN (56%, n = 268) was determined from embryonic cuticle preparations. ***P<0.0001 as determined by the Chi squared test.