The level of DLDB/CHIP controls the activity of the LIM homeodomain protein apterous: evidence for a functional tetramer complex in vivo

Abstract

The LIM homeodomain (LIM-HD) protein Apterous (Ap) and its cofactor DLDB/CHIP control dorso- ventral (D/V) patterning and growth of Drosophila wing. To investigate the molecular mechanisms of Ap/CHIP function we altered their relative levels of expression and generated mutants in the LIM1, LIM2 and HD domains of Ap, as well as in the LIM-interacting and self-association domains of CHIP. Using in vitro and in vivo assays we found that: (i) the levels of CHIP relative to Ap control D/V patterning; (ii) the LIM1 and LIM2 domains differ in their contributions to Ap function; (iii) Ap HD mutations cause weak dominant negative effects; (iv) overexpression of ChipDeltaSAD mutants mimics Ap lack-of-function, and this dominant negative phenotype is caused by titration of Ap because it can be rescued by adding extra Ap; and (v) overexpression of ChipDeltaLID mutants also causes an Ap lack-of-function phenotype, but it cannot be rescued by extra Ap. These results support the model that the Ap-CHIP active complex in vivo is a tetramer.

Fig. 1. Mutations made in the Ap LIM domains and homeodomain. (A) The LIM domain as a double zinc-finger structure (Jurata and Gill, 1998). The amino acids interacting with zinc are indicated by circles. Residues shown in reverse type were mutated (H into R, and C into S) to disrupt the zinc-finger structure. (B) The homeodomain interacting with DNA (Kissinger et al., 1990). Two different regions of the homeodomain were mutated and identified as HA and HB. HA corresponds to mutations made on the N-terminal arm and HB represents mutations made on the third helix. The amino acid substitutions and their corresponding positions are indicated. (C) Structural organization of the ap mutant constructs.

Fig. 2. Transcriptional activity of Ap mutant constructs. Drosophila S2 cells were cotransfected with pGSU-AdhCAT and the indicated ap effector constructs. CAT activities relative to that given by the pMKapwt vector alone are shown at the top. Values are the mean (± SEM) of three independent experiments. Details of the assay are given in Materials and methods. Lhx2, which is the vertebrate ortholog of ap, recognizes the GSU regulatory element and was used as a positive control. Mock, cells without transfected DNA. A schematic representation of each effector construct is also shown.

Fig. 3. Functional analysis of wild-type and mutant ap constructs in ectopic assays. Panels show anti-β-gal immunostaining for the wg-lacZ marker following ectopic expression in wing imaginal discs, as well as the resulting phenotypes. (A and B) Wing discs showing UAS-lacZ expression driven by 32B-GAL4 and dpp-GAL4, respectively. (C) Wing disc showing the wg-lacZ wild-type expression. The arrowhead points to the expression domain along the D/V compartment boundary. (D) Wild-type wing. (E and G) Wing discs expressing UAS:apwt and UAS:apL1 from 32B-GAL4, respectively. Note that wild-type ap completely, whereas apL1 partially eliminates the wg-lacZ expression along the D/V boundary. (I and K) Wing discs expressing UAS:apwt and UAS:apL1 from dpp-GAL4, respectively. Note that wild-type ap induces a large stripe of wg-lacZ expression across the ventral compartment, whereas apL1 only induces a small stripe (arrows). (F and H) Wing phenotypes caused by ectopic expression of ap and apL1 using 32B-GAL4, respectively. (J and L) Wing phenotypes caused by ectopic expression of ap and apL1 using dpp-GAL4, respectively.

Fig. 4. Functional analysis of Ap mutant proteins. (A) Summary of phenotypic rescues. The ability of each protein to rescue the indicated ap mutant phenotypes is displayed at the right. See the text and Materials and methods for details. Stars indicate the mutation site of the proteins. The number of lines tested per construct is also indicated. ++, complete rescue; +, partial rescue; –, no rescue. (B–E) The rescue results of the wing phenotype. (B) Wing of a wild-type fly. (C) Notum showing the wing phenotype of an apGAL4/ap^UGO35 mutant fly. Arrows point to wing rudiments. (D) Rescued wing of an apGAL4/ap^UGO35 fly carrying the UAS:apwt construct. (E) Partially rescued wing phenotype of an apGAL4/ap^UGO35 fly carrying the UAS:apL1 mutant construct.

Fig. 5. Ap homeodomain mutations behave as weak dominant negatives. (A) Wild-type wing. (B) Wing of an apGAL4; UAS:apHA/B fly. Note the blistering and the pointed shape of the wing. (C) Wing of an apGAL4/ap^ts78j fly. (D) Wing of an apGAL4/ap^ts78j; UAS:apHA/B fly. Note that the size of the wing (arrows) is greatly reduced. (E) ApHA/B impairs Ap transcriptional activity at high concentration. S2 cells were cotransfected with a fixed amount of an Ap expression vector (pMKapwt) and increasing amounts of a vector expressing an Ap homeodomain mutant protein (pMKapHA/B). Note the relatively high concentration of ApHA/B required to reduce Ap transcriptional activity to basal levels. CAT activities relative to that produced by transfecting with the pMKapwt vector alone are shown at the top. Values are the mean (± SEM) of three independent experiments. Various amounts of a carrier plasmid (pMK26) were added to equalize the amount of transfected DNA.

Fig. 6. CHIP/DLDB controls Ap transcriptional activity in a dosage-dependent manner. (A) An excess of CHIP diminishes Ap transcriptional activity. S2 cells were cotransfected with a fixed amount of an ap expression vector (MKapwt) and increasing amounts of a vector expressing Chip (MKChip). Note that an increase in the concentration of CHIP reduces Ap transcriptional activity. (B) The addition of Ap overcomes the inhibitory effect produced by excess CHIP. A fixed amount of pMKChip was cotransfected with increasing amounts of pMKapwt. Note that an augmentation in the concentration of Ap gradually restores CAT activity. CAT activities relative to that produced by transfecting with the pMKapwt vector alone are shown at the top. Values are the mean (± SEM) of three independent experiments. Various amounts of a carrier plasmid (pMK26) were added to equalize the amount of transfected DNA.

Fig. 7. Different requirements for the LIM1 and LIM2 domains in Ap–CHIP interactions. (A) The LIM2 domain is most important for Ap transcriptional activity. S2 cells were cotransfected with MKapwt and MKChip at a 1:4 molar ratio, plus a similar excess of MKapL1, MKapL2 and MKapL1/2. An excess of MKapwt was also added as positive control. CAT activities relative to that produced by transfecting with the MKapwt vector alone are shown at the top. Values are the mean (± SEM) of three independent experiments. Various amounts of a carrier plasmid were added to equalize the amount of transfected DNA. (B–E) The LIM2 domain is absolutely required for Ap–CHIP interactions in vivo. (B) Wild-type wing. (C) Notum showing the wing rudiments (arrows) of an apGal4/+; UAS:Chip/+ fly. Note the severe reduction in the size of the wings caused by overexpression of Chip. (D) Wing of an apGal4/+; UAS:Chip/UAS:apwt fly. (E) Wing of an apGal4/UAS:apL1; UAS:Chip/+ fly. Note that the phenotype caused by Chip overexpression is completely and partially rescued by the UAS:apwt (D) and UAS:apL1 (E) constructs, respectively.

Fig. 8. Functional analysis of CHIP mutants. (A) The CHIP deletion constructs. For comparison, the location of the major blocks of homology between the vertebrate NLI/LDB and its Drosophila ortholog CHIP are shown in shaded boxes. The vertebrate SAD, NLS (nuclear localization signal) and LID motifs are indicated by brackets. The CHIP deletions are represented by continuous lines and the positions of the corresponding amino acids are indicated. Black boxes represent human Myc epitope tags introduced within the UAS vector for monitoring expression. (B–G) In vivo analysis of the LID and SAD domains of CHIP. (B) Wild-type wing. (C) Wing phenotype caused by overexpression of ChipΔSAD from the dppGAL4 driver; the arrow points to the ectopic margin. (D) Wing phenotype caused by dppGAL4 overexpression of ChipΔLID; the arrow points to the nicking of the wing. (E) Wing phenotype resulting from overexpression of ChipΔSAD with the apGAL4 driver. Note the severe reduction in the size of the wing. (F) Rescued wing phenotype of a fly overexpressing ChipΔSAD and ap from the apGAL4 driver. (G) Wing phenotype caused by overexpression of ChipΔLID from the apGAL4 driver.

Fig. 9. Ap and CHIP mutants in the context of the tetramer model (A). Two CHIP molecules interact with each other through their SAD domains and bridge two Ap molecules through interactions between the CHIP LID and Ap LIM domains. The CHIP homodimer may be required for bringing together distant Ap-binding sites and/or for facilitating contacts with target promoters to activate downstream genes. (B) Ap LIM1 is only required for full Ap activity. Thus, the LIM1 domain may stabilize the tetramer complex. In Ap LIM1 mutants, the complex would form, but it would be less efficient at regulating target genes. (C) Ap LIM2 is absolutely required for interactions with CHIP LID. Thus, Ap LIM2 mutations are likely to disrupt the complex. (D) Mutations in Ap HD behave as dominant negatives. As indicated in the figure, they are likely to result in the formation of non-functional complexes. Our experiments show that a higher concentration of wild-type Ap overcomes this dominant negative effect. (E) Deletion of the CHIP SAD motif should lead to the titration of Ap. In this context, we showed that the addition of Ap rescues the CHIPΔSAD overexpression phenotype, probably by transforming abnormal trimeric complexes into functional tetramers. (F) Deletion of the CHIP LID domain should also cause formation of non-functional trimers due to recruitment of CHIP molecules that cannot interact with Ap LIM domains. In this context, we showed that the addition of Ap does not rescue the CHIPΔLID overexpression phenotype.