A specific protein substrate for a deubiquitinating enzyme: Liquid facets is the substrate of Fat facets

Abstract

Eukaryotic genomes encode large families of deubiquitinating enzymes (DUBs). Genetic data suggest that Fat facets (Faf), a Drosophila DUB essential for patterning the compound eye, might have a novel regulatory function; Faf might reverse the ubiquitination of a specific substrate, thereby preventing proteasomal degradation of that protein. Additional genetic data implicate Liquid facets (Lqf), a homolog of the vertebrate endocytic protein epsin, as a candidate for the key substrate of Faf. Here, biochemical experiments critical to testing this model were performed. The results show definitively that Lqf is the key substrate of Faf in the eye; Lqf concentration is Faf-dependent, Lqf is ubiquitinated in vivo and deubiquitinated by Faf, and Lqf and Faf interact physically.

Figure 1

Colocalization of Shi and Lqf proteins in eye discs. (A–J) Apical views of a third instar larval eye disc, double-labeled with anti-Shi and anti-Lqf. (A–C) Lqf and Shi colocalize in cells within and posterior to the morphogenetic furrow, indicated by the arrow in A. (D–F) An enlargement of the area near the furrow shows that Lqf and Shi are concentrated at the apical tips of cells within the furrow. (G–I) An enlarged view of the area posterior to the furrow shows that Lqf and Shi are concentrated in dots, which are the apical membranes of the photoreceptors (R-cells), where they meet (see below). In addition, Lqf and Shi colocalize in a lattice, which is made up of the membranes of the surrounding cells. (J) A further enlargement of G, showing Lqf membrane localization in four adjacent facets. (K,L) An apical view of four adjacent facets in eye discs double-labeled with anti-Lqf and anti-Elav (K), which labels R-cell nuclei, or anti-Lqf and anti-Cut (L), which labels cone cell nuclei. There is no overlap in the localization of Lqf protein and either Elav or Cut; Lqf is outside the nuclei (which fill the apical cytoplasm) and in the central region of the developing facet where the photoreceptor cell membranes meet.

Figure 2

Detection of Lqf and Shi proteins in faf⁻ clones in eye disc. Apical views of two different third instar larval eye discs are shown in A–D and E–H. (A,E) The faf⁻ clones are labeled by the absence of β-gal protein. (B,F) The clone shapes are apparent as areas with lower levels of Lqf protein. (C,G) The clone shapes in A and E were outlined in white and layered over the panels in B and F. (D,H) The levels of Shi protein are unaffected in the faf⁻ clones. We know that detection of the Lqf protein signal is unaffected by the β-gal protein signal, as clones are visible as areas of lower levels of Lqf signal in discs labeled only with anti-Lqf. A few of the clone areas in A and E are not obviously mirrored in B and F. This is because of the subtlety of the Lqf concentration difference (<twofold) often being detected; although there is only two- to threefold less Lqf in faf⁻/faf⁻ eye discs than in wild-type (faf⁺/faf⁺), the clones are faf⁻/faf⁻, but the surrounding cells are often faf⁻/faf⁺ (the clone twin spots are faf⁺/faf⁺). Slight variability in antibody penetration and so on within the disc can affect the staining and obscure concentration differences in parts of the disc. Nevertheless, it is clear that the clone shapes are generally present in the Lqf-stained discs (B,F), but not in the Shi-stained discs (D,H).

Figure 3

Comparison of protein levels in wild-type and faf⁻ eye discs. (A) A histogram showing the level of Lqf protein, normalized to tubulin, in wild-type and faf⁻ eye discs. The wild-type value was arbitrarily set to 1.0. Examples of Western blots used to generate this data are shown in Figure 4A below. (B) A histogram showing the level of Lqf protein, normalized to tubulin, in faf^BX4 eye discs, and in faf^BX4 eye discs with a copy of either a faf⁺ transgene or a faf⁻ transgene. The faf^BX4 value was arbitrarily set to 1.0. An example of a Western blot used to generate these data is shown in Figure 4B below. (C) A histogram showing the levels of four different endocytosis complex proteins and Armadillo (Arm), normalized to tubulin, in wild-type (wt; black bars) and faf^BX4 (gray bars) eye discs. The wild-type value was arbitrarily set to 1.0. Standard errors in A–C were calculated from differences in repeated experiments.

Figure 4

Deubiquitination and binding of Lqf by Faf. (A) Western blots of eye disc protein extracts, labeled with anti-Lqf and anti-tubulin, are shown. The lqf gene encodes two different proteins by alternate mRNA splicing, of predicted molecular weights ∼86 kD (Lqf1) and ∼70 kD (Lqf2; Cadavid et al. 2000); Lqf2 is the predominant form in eye discs. The small arrows indicate the rungs of the ladder of higher-molecular-weight forms of Lqf2. The size of the second rung of the ladder corresponds precisely to the size of Lqf1, which is the size predicted for Ub–Ub–Lqf2 (70 + 8 + 8 = 86). The two lanes shown for each genotype are different amounts of the same protein extract. These experiments were repeated 10 times, sometimes using faf^FO8, and identical results were always obtained. (B) A Western blot of eye disc protein extracts, labeled with anti-Lqf and anti-tubulin. Ubiquitinated forms of Lqf2 are stabilized in faf⁻ extracts, disappear when a faf⁺ transgene is introduced, but remain stabilized in the presence of a faf⁻ transgene. Results similar to these were obtained in 3/3 repetitions. (C) Western blots of an immunoprecipitation experiment, labeled with anti-myc. The extracts are from embryos transformed with the P{hs-myc-faf} transgene that were heat-shocked and thus express myc-Faf (left and right panels), or not heat-shocked (middle panel). The extracts were immunoprecipitated with anti-Lqf (left and middle panels), or with no antibody as a control (right panel). Preimmune serum also failed to immunoprecipitate Lqf or myc-Faf (data not shown). Coomassie-stained gels of heat-shocked and non-heat-shocked embryo extracts appeared identical. The predicted molecular weight of myc-Faf is ∼300 kD. (EX) 1/15 of a 150-μL crude extract from 150 μL of embryos, (S) 1/15 of the supernatant protein from 150 μL of crude extract, which was not immunoprecipitated by anti-Lqf, (B) total protein from the 150-μL extract bound to anti-Lqf beads.