Cup blocks the precocious activation of the orb autoregulatory loop

Abstract

Translational regulation of localized mRNAs is essential for patterning and axes determination in many organisms. In the Drosophila ovary, the germline-specific Orb protein mediates the translational activation of a variety of mRNAs localized in the oocyte. One of the Orb target mRNAs is orb itself, and this autoregulatory activity ensures that Orb proteins specifically accumulate in the developing oocyte. Orb is an RNA-binding protein and is a member of the cytoplasmic polyadenylation element binding (CPEB) protein family. We report here that Cup forms a complex in vivo with Orb. We also show that cup negatively regulates orb and is required to block the precocious activation of the orb positive autoregulatory loop. In cup mutant ovaries, high levels of Orb accumulate in the nurse cells, leading to what appears to be a failure in oocyte specification as a number of oocyte markers inappropriately accumulate in nurse cells. In addition, while orb mRNA is mislocalized and destabilized, a longer poly(A) tail is maintained than in wild type ovaries. Analysis of Orb phosphoisoforms reveals that loss of cup leads to the accumulation of hyperphosphorylated Orb, suggesting that an important function of cup in orb-dependent mRNA localization pathways is to impede Orb activation.

Figure 1. Cup associates with Orb in vivo and affects the phosphorylation status of Orb.

Cup was initially identified as an Orb associated protein by analyzing Orb antibody immunoprecipitates from wild type ovaries using MudPIT mass spectrometry . Western blots were used to confirm this association. (A): Immunoprecipitation of ovarian extracts with Orb antibody co-immunoprecipitates Cup protein. Cup does not co-immunoprecipitate with a negative control, HA antibody. The amount of ovary extract loaded in panel A represents 5% of the ovary extract used in the Orb and HA antibody immunoprecipitates. Half of the sample recovered in each immunoprecipitate was loaded on to the “pellet” lanes. For HA, approximately 2.5% of the sample was loaded into the “supernatant” lane. For the Orb “supernatant” lane a portion of the sample was lost during loading. (B): The converse immunoprecipitation, using antibody against Cup, pulled down both Orb isoform, whereas β-galactosidase antibody does not. The Cup-Orb complexes are RNA-independent as they are resistant to treatment with RNAse A. The amount of ovary extract loaded in the first lanes in each panel consist of 10% of the amount of ovary extracts used in each immunoprecipitation experiment loaded in subsequent lanes. The sample loaded onto the HA “supernatant” lane represents in panel A represents 2.5% of the supernatant from the Cup antibody immunoprecipitation precipitation.

Figure 2. Cup negatively regulates orb.

orb is weakly haploinsufficient for the specification of dorsal-ventral polarity and a small percentage of the eggs produced by orb³⁴³/+ females are ventralized. This haploinsufficiency is enhanced when the orb³⁴³/+ females also carry a copy of the hsp83-LacZ orb 3′ UTR transgene and, depending upon temperature and growth conditions, between 20%–30% of the eggs laid by HD19G orb³⁴³/+ females are ventralized. To test for genetic interactions with cup, HD19G orb³⁴³/TM3Ser females (n = 10 in each cross) were mated with cup/CyO males. Five different cup alleles, cup¹, cup³, cup⁶, cup⁸ and cup¹³⁵⁵ were tested at 18°C and 25°C. When trans-heterozygous, all 5 alleles reduce the frequency of DV polarity defects in eggs produced by HD19G orb³⁴³/+ females. The severity of this defect ranges from completely fused to missing dorsal appendages. Suppression is stronger at 18°C (black bars) than at 25°C (gray bars). The p-values (by Chi-squared test) for each experiments are as such: 18°C (cup¹: 3.031E-141; cup³: 1.7326E-29; cup⁶: 7.6267E-36; cup⁸: 7.6611E-14; cup¹³⁵⁵: 1.72342E-28) and 25°C: (cup¹: 5.428E-118; cup³: 0.608012094; cup⁶: 4.5796E-09; cup⁸: 0.000309325; cup¹³⁵⁵: 4.52924E-07).

Figure 3. Orb protein levels are altered in cup mutant ovaries.

A Western blot of ovary extracts from wild type and cup mutants as indicated was probed sequentially with Orb, dFMR1 and α-tubulin antibodies. Orb protein levels are reduced in cup⁸ ovaries compared to wild type. The ratio of cup⁸/WT is 0.56 using dFMR1 as a loading control and 0.34 using α-tubulin as a loading control. In contrast, Orb levels are elevated in cup³ and cup¹³⁵⁵ ovaries. For cup³, the level of Orb protein is elevated 1.8-fold using dFMR1 as a loading control and 1.9 using α-tubulin as the loading control. For cup¹³⁵⁵, the level of Orb protein is elevated 1.6-fold using dFMR1 as a loading control, and 2.2-fold using α-tubulin as the loading control. Even larger increases were observed in another experiment: cup³: 3.0-fold using dFMR1 as the loading control and 2.25-fold using α-tubulin; cup¹³⁵⁵: 2.9-fold using dFMR1 as the loading control and 2.8-fold using α-tubulin as the loading control.

Figure 4. Orb is misexpressed in cup ovaries.

Green: Nucleic acids. Red: Orb. Panels A–B′: In wild type, Orb is first expressed during the formation of the 16-cell cyst (arrowhead in A) and though it is present in most cells in the cyst, it is greatly enriched in the presumptive oocyte. The localization pattern is refined during early pre-vitellogenic stages (arrow in A); Orb becomes concentrated at the posterior of the oocyte (>and * in A, arrowhead in B′), while only low levels of protein are detected in the nurse cells closest to the oocyte (arrow in B′). Panels B and B′: successive confocal sections through the same two egg chambers. Note that there are high levels of Orb protein in all of the germ cells in these two chambers. Panels C, D and E: Chambers from cup⁸ ovaries. In each case several focal planes are shown and are designated as ‘. In cup⁸ 16-cell cysts Orb is present at high levels in most of the cells (C, C′). Unlike wild type the Orb localization pattern does not refine during the early pre-vitellogenic stages and high levels of Orb persist in most of the cells in the chamber. The distribution of Orb in two “early” cup⁸ chambers is shown in panels D-D″″. Note that many cells in these two chambers have high levels of Orb protein. Two other cup⁸ chambers are shown in panels E-E″. In the chamber on the left Orb is clearly concentrated in a subset of the cells. In the chamber on the right the level of Orb protein is low. A marked reduction in the level of Orb protein is typically observed in older cup⁸ chambers. Panels F–J show the pattern of Orb accumulation in cup³ chambers. Like cup⁸, chambers that have high levels of Orb in most germ cells are often observed in the cup³ mutant (arrowhead in F, F′ and H). Orb also accumulates in rings around the nurse cell nuclei (* in F, F′) or in clumps in the nurse cell cytoplasm (* in I and J). In some cup³ chambers Orb concentrates in several cells near the posterior (arrow heads in I, arrows in J) while in other chambers the oocyte has an abnormal elongated shape (arrow in I). Panels K and L show cup¹³⁵⁵ chambers while panel M shows a wild stage 7–8 chamber. Orb accumulation during the early stages of oogenesis usually appears normal in cup¹³⁵⁵ (panel K); however in older chambers, high than normal levels Orb protein are seen in the nurse cells (arrows in L) and there are often clumps of Orb protein (arrowheads in L). In older wild type chambers (panel M) Orb is localized to the oocyte cortex, while there is little Orb protein in the nurse cells.

Figure 5. orb mRNA is mislocalized in cup mutants.

In wild type, orb mRNA can first be detected in the germarium (panel A). It appears concentrated in a single cell that is thought to correspond to the presumptive oocyte. During the previtellogenic stages (panel A), orb mRNA is localized at the posterior end of the oocyte, while there are only low levels in the nurse cell cytoplasm. After the onset of vitellogenesis (panel B), orb mRNA accumulates along the anterior margin of the oocyte. In many cup⁸ chambers orb mRNA is not properly localized and reduced in level (arrowheads in C). Sometimes a single cell can be seen which has high levels of orb mRNA (D & E); however, this cell is not always positioned correctly (arrow in D). Mislocalized orb mRNA is also seen in cup³. In the examples shown in F and G the level of orb mRNA in the nurse cells is high, while the level in the presumptive oocyte is relatively low (compare levels of orb mRNA in nurse cells and oocyte in the wild type (arrow in panel A) and mutant (arrow in F and G) chambers). The pattern of orb mRNA accumulation in cup¹³⁵⁵ chambers (panels H and I) more closely resembles wild type; however, in older chambers the mRNA is not properly localized to the anterior margin and higher than normal levels are seen in the nurse cells (panel I).

Figure 6. osk mRNA is reduced in cup mutants while Osk protein levels remain the same as in wild type.

(A) osk mRNA levels in pre-vitellogenic wild type ovaries and in ovaries from the three cup alleles, cup⁸, cup³ and cup¹³⁵⁵, were measured by quantitative real time RT-PCR using actin mRNA as the control. osk mRNA levels in the mutant ovaries were found to be decreased compared to wild type levels. The largest reduction is seen cup⁸, where the level of osk mRNA is only about 20% that in previtellogenic wild type ovaries. In the less severe alleles, osk mRNA levels are 50% (cup³) and 70% (cup¹³⁵⁵) of wild type previtellogenic levels. The amount of osk mRNA normalized to actin mRNA in pre-vitellogenic chambers is essentially the same (98%) as it is in ovaries from aged wild type females. (B) The effects of cup mutations on the expression of Osk protein. The relative amount of Osk protein in wild type and cup mutant ovaries was estimated by comparing the levels of Osk and α-tubulin in pre -vitellogenic wild type ovaries and in ovaries from the three cup mutants, cup⁸, cup³ and cup¹³⁵⁵. In all three cases the relative amount of Osk protein was marginally greater than the amount in wild type ovaries. (C) Panel C shows the ratio of Osk protein to osk mRNA levels. Assuming that the rate of Osk turnover in wild type is equivalent to that in the cup mutants, more Osk protein must produced on average from each osk mRNA in the cup mutants than in wild type.

Figure 7. The poly(A) tails of orb in cup⁸ and cup³ mutants are elongated compared to wild type.

Poly(A) tails of orb mRNA from wild type, cup⁸ and cup³ (as indicated) were analyzed using the ligation-mediated poly(A)-tail assay . Anchor primers for reverse transcription and amplification from the poly(A) tail were as described in . To increase specificity for the orb 3′ UTR we used a two step PCR amplification with a nested primer pair derived from the orb 3′UTR (F2:GATTGTCCGCTAAGCGTTTATCAGGA and F4:CCTTGTGAACATTAACGCGATG). The orb 3′UTR products from the 2^nd PCR amplification were analyzed on an agarose gel, and after blotting to nitrocellulose were detected by hybridizing the filter with a radioactive orb 3′UTR probe.

Figure 8. The ratio of the upper and lower Orb isoforms is altered in cup ovaries.

In wild type, the lower isoform is typically most abundant, while in cup mutant ovaries the level of the upper isoform is increased. The ratio of upper to lower phosphoisoforms in the WT is 0.75. In other experiments (n = 6) the range for WT was 0.74–0.84 with an average value of 0.77. In the experiment shown here, the ratio for the cup mutants is increased to 1.3, 1.9 and 1.1 for cup³, cup⁸ and cup¹³⁵⁵, respectively. Two different exposures of the cup⁸ lane are shown. The average of three experiments for cup³ was 1.2, cup⁸ was 1.7 and cup¹³⁵⁵ was 1.1. The average ratio is shown in the figure.