IV. Discussion of Results

Analysis of data reveals several significant changes in phenotype and cases of synthetic lethality. Although there are no bristles on the halteres of wildtype D. melanogaster, the control crosses which carried mutated B52 and Ubx alleles produced an extremely mild Ubx phenotype (Appendix B). Increased numbers of bristles on the halteres of the progeny from the designated crosses imply a change in the Ubx splicing pattern. The increased appearance of bristles on the fly haltere also indicates a loss of function of UBX protein. B52^ED Ubx¹⁹⁵ has a mild Ubx phenotype: approximately two bristles per haltere. The effects of splicing factors, mutants, and Minutes were tested for a modification of this haltere phenotype.
In order to determine the significance of the enhancement due to Minutes and B52 interaction, the average number of bristles per haltere for the progeny of each of the fly crosses was compared to that of the control cross to the wildtype (ORE-R). The program Statistica was used to quantify whether the average number of bristles per haltere was significantly different in the Minute and control crosses. The nonparametric Mann-Whitney U Test was administered to analyze the small sample sizes. If the P value for a specific cross was less than 0.001, the significance of the bristle enhancement on the halteres of the fly was extremely high. If the P value for the specific cross was less than 0.01, the enhancement was also significant. If the P value was approximately 0.05, the enhancement was considered to be borderline significance (see Tables 1-8 in Appendix B).
Several Minutes and mutations in genes of known functions caused a significant increase in the number of bristles on the flies' halteres. The Ubx phenotype was enhanced for both mutations of the Ubx gene by the following: su(s)^R39, U170K¹, M(2)41A², M(3)76A¹, and M(3)X¹ (Table 2, Results) Since the products of these genes enhance the Ubx phenotype for both Ubx alleles, they are considered prime candidates as splicing factors.
The U170K¹ encodes the snRNP U1-related protein. The mutation demonstrated enhancement of Ubx phenotype with both alleles, while U170K⁶² only affected the halteres of Ubx^9.22. Since both of these mutations in the genes for a known general splicing factor have caused a change in the Ubx splicing pattern, it is probable that U170K interacts with B52. Another gene implicated in universal splicing pathways, fl(2)d², resulted in a significant increase in bristles on the haltere in conjunction with the Ubx¹⁹⁵ allele (Table 3, Appendix B).
Also of interest is the result that M(2)32A, which encodes ribosomal protein S13, enhanced the Ubx phenotype in the B52^EDUbx^9.22 heterozygotes (Table 5, Appendix B). The mutation l(2)06694, a P element insertion in a tRNA synthetase gene, is another mutation in a translation factor that modifies the Ubx phenotype. These results suggest that a mutation in a translation factor combined with the mutation in the splicing factor B52 may together contribute to the enhancement of the Ubx phenotype. Many proteins have been found to have multiple functions in the cell; ribosomal protein S3, encoded by M(3)95A, has been associated with the nuclear matrix, implying a role in DNA repair in addition to its involvement in translation. Perhaps some of these translation factors also have roles in splicing.
The effects on Ubx splicing by the Minutes were indicated not only by a significant increase in
the number of bristles on the haltere, but also by synthetic lethality. Much information can be drawn from the summary graphs compiled from the synthetic lethal tables (Tables 1 and 2, Appendix A). One Minute, M(3)65F¹ was synthetic lethal when crossed with B52^EDUbx¹⁹⁵ and with B52^EDUbx^9.22. This result suggests that M(3)65F¹, may be a splicing factor. To test this hypothesis, reverse transcription-polymerase chain reaction (RT-PCR) analysis would be performed in order to determine which Ubx isoforms are produced in this strain.
Crosses utilizing M(3)86D also revealed an interesting result. This Minute was lethal when crossed with B52^EDUbx¹⁹⁵ but enhanced the Ubx phenotype when crossed with B52^EDUbx^9.22, qualifying M(3)86D as a possible splicing factor gene. This Minute is known to be located near the gene for RBP1, another SR protein that may be involved in Ubx splicing. It is possible that M(3)86D codes for RBPI.
Five Minutes were synthetic lethal with B52^EDUbx¹⁹⁵, and nine Minutes enhanced the Ubx phenotype. However, only two Minutes were synthetic lethal with B52^EDUbx^9.22, and fourteen crosses enhanced the Ubx phenotype. In general, the B52^EDUbx¹⁹⁵ flies seemed to be more sensitive to synthetic lethality than were B52^EDUbx^9.22. This may suggest that there is a threshold of isoform IV that can be tolerated. The only functional protein that can be made from Ubx¹⁹⁵ is isoform IV, but no functional Ubx can be made from Ubx^9.22. This implies that the B52^EDUbx^9.22 will not acquire enough isoform IV from its one wildtype allele to result in synthetic lethality. If this threshold exists, the Ubx¹⁹⁵ allele will contribute to the isoform IV content but the Ubx^9.22 alelle cannot, and therefore there will be more lethals with the B52^EDUbx¹⁹⁵ heterozygotes due to isoform IV overdose. The B52^EDUbx^9.22 would be more likely to live, but UBX activity will be compromised due to the increase of the isoform IV caused by the presence of the interacting Minute. Another explanation for the higher incidence of lethality with B52^EDUbx¹⁹⁵ is that there is another unidentified lethal modifier on the chromosome. To verify our synthetic lethality results, the original stock would have to be outcrossed to remove modifiers on the B52 Ubx mutant chromosome and the experiment repeated with this new "clean" stock.
B52^EDUbx¹⁹⁵ crossed with su(w[a])¹ showed intriguing effects on the phenotypes of resulting flies. Flies of this combined genotype had halteres with a significantly low bristle count making them similar to wild type halteres (Table 3, Appendix B). Halteres tending towards the formation of a wing, with a higher count of bristles, were expected in this cross. The data reveals that the suppressor-of-white-apricot gene suppresses B52^EDUbx¹⁹⁵, resulting in an increase in Ubx function. B52 protein works in conjunction with the suppressor white apricot protein to modify Ubx splicing.
Although the results from the sythetic lethal screen and bristle counting implicate several Minutes in splicing mechanisms, there still remains the possibility for error. Since D. melanogaster has a relatively long period of development, the statistics and averages could be distorted due to the low number of progeny obtained. A larger number of progeny would have shown correlations more accurately; however, due to time restraints, it was impossible to conduct many trials. In addition, human error may have skewed data results. Lab teams may not have identified markers and Ubx phenotypes correctly. For example, the difference between wildtype eyes and mutated henna eyes was extremely difficult to see. Another eye phenotype, moire, was also challenging to identify even under high magnification. Despite these drawbacks, the data did contain trends from which conclusions could be drawn.