Identification of the molecular lesions in four EMS-induced alleles of the daughterless gene of Drosophila melanogaster.

Baker, S.T., and C. Cronmiller.  Department of Biology, Gilmer Hall, University of Virginia, P.O. Box 400328, Charlottesville, VA 22904-4328.

            The Drosophila daughterless (da) gene encodes a basic helix-loop-helix (bHLH) transcription factor, and although this protein regulates many aspects of development, the molecular mechanisms of its multifunctionality have not been fully elucidated.  Using polymerase chain reaction (PCR) and DNA sequencing, we analyzed four ethyl methane sulfate (EMS)-induced da alleles (da^s22, da⁵, da², and da^f75) to determine their molecular lesions and possibly identify functionally significant regions of the da protein

            We chose these alleles because they have been extensively characterized genetically, or their genetic behavior suggested that their mutations appear to affect some of the daughterless functions more severely than others.  da² and da⁵ have been used in numerous developmental studies (e.g., Cronmiller and Cline, 1987;  Caudy et al., 1988;  Cummings and Cronmiller, 1994;  Brown et al., 1996).  Not only do these alleles behave like genetic nulls, but both also appear to be protein nulls, as determined by immunohistochemical staining (Cronmiller and Cummings, 1993).  da^s22 is a strong hypomorph, whose effects on oogenesis are somewhat more severe than those on sex determination or viability (Cummings and Cronmiller, 1994).  Of all of the alleles analyzed, da^f75 is the least well characterized genetically, but it was included in this analysis because it had been found previously to exhibit reduced viability with da^lyh (K. Curran and C. Cronmiller, unpublished observations), which otherwise appears to be an ovary-specific allele with no adverse effects on viability (Smith and Cronmiller, 2001).

            Because each of the four da alleles chosen for sequence analysis is recessive lethal, we took advantage of their viability in combination with the da^lyh insertion allele to recover sequencing template DNA by PCR.  DNA was prepared from each transheterozygous combination (e.g., da²/da^lyh), and primers that flanked the springer insertion site of da^lyh (in the intron, upstream of the protein-coding exon: Smith and Cronmiller, 2001) were used to amplify da gene DNA.  DNAs amplified from the two alleles would be distinguishable following electrophoretic separation of the PCR products, since the 8 kb transposon inserted in da^lyh would substantially increase the size of the DNA fragment amplified from that allele.  In each case, however, PCR product was recovered only from the allele to be sequenced, indicating that the larger (>11 kb) product was not amplifiable by the PCR conditions used.  As a precaution against contamination, each experimental band was excised, purified and re-amplified;  either the original primers or nested primers were used for re-amplification.

            Automated DNA sequencing was carried out at the University of Virginia Biomolecular Research Facility (Charlottesville, VA).  Sequencing primers (Operon Technologies, Sigma) were spaced 500-700 nucleotides apart and spanned the entire da gene, from the 3´ end of mRpS7 (located upstream of the 5´ end of da) to the 3´ end of the da protein coding region.  DNA sequence analysis was accomplished with the MacVector, AssemblyLIGN (Sequence Analysis Software), and BLAST (NCBI) programs.  The sequence of this entire region was determined for da^s22, while only the protein coding regions for da⁵, da² and da^f75 were sequenced.  Single point mutations were identified for all four da alleles analyzed (Figure 1).

                                                                          [TAT]

A     da⁵         ATG GGC CAA TCG GCG TAA CAG AAT AGC GGC CCG  574

              106     M      G      Q      S       A      * 

                                                                          [CAG]

B     da²          GCG GCA CTG CGG CAA TAG ATG TAC ATG CCG GCG  1102

              282      A      A       L      R       Q      *

                                                                       [CAG]

C     da^f75       CTG GCC GGT GTC AAT TAG TCG CTG GCC TCG ATC  1416

              387       L      A      G      V      N      *

                                                                       [CGG]

D     da^s22       GAA CGT ATC CGC ATT TGG GAT ATT AAC GAG GCG  1945

              563       E     R      I       R       I      W      D      I      N       E      A

Figure 1.  Mutational lesions in EMS-induced alleles of da.  DNA sequences of (A) da⁵, (B) da², (C) da^f75 and (D) da^s22 are shown with each single base change indicated in bold.  The affected codons are underlined with their corresponding wild-type amino acids bracketed above.  All sequence alterations were confirmed by numerous independent sequence analyses.  Position numbers on the right refer to the nucleotide sequence of the 3.0-kb da cDNA, MN6 (Cronmiller et al., 1988); numbers on the left refer to the amino acids of Da.

            Nonsense mutations were found in da^f75 and the null alleles, da² and da⁵.  The most severe truncation of the da protein product would occur in da⁵, which had a T to A transversion at position 559 (at the beginning of exon 2).  This converted a Tyr codon (TAT) to a stop (TAA), truncating the open reading frame to encode only the N-terminal 110 amino acids (~16%) of the normal protein.  The nonsense mutation of da2 resulted from a C to T transition at position 1085, which changed a Gln codon (CAG) to a stop (TAG) and reduced the protein coding potential of the mutant allele to 286 amino acids (~41% of the normal length).  Finally, da^f75 was found to result from a C to T transition at position 1399, replacing another Gln codon with a stop codon.  This stop codon would truncate the protein to 391 amino acids, approximately 55% of normal Da.  On western blots of extracts from heterozygous flies, we were able to detect only one of the predicted mutant proteins, that from da²; however, the amount of truncated protein was drastically reduced, relative to that of the wild-type protein.  Thus, if translated, the truncated proteins appear to be unstable.  Nevertheless, since the essential bHLH domain has been deleted in all three nonsense mutations, any small/trace amounts of mutant protein that might be stable would be nonfunctional.

            A missense mutation was found in the strong hypomorph, da^s22.  In this mutant a C to T transition at position 1928 changed an Arg codon (CGG) in the HLH domain into a Trp codon (TGG).  The substituted amino acid is part of a stripe of basic residues that are located on the outer surface of helix 1 of the bHLH domain.  The amphipathic helices of bHLH transcription factors mediate dimerization, a requirement for DNA binding (Murre et al., 1989;  Voronova and Baltimore, 1990), and stripes of acidic and basic residues of helix 1 may be important for heterodimer formation (Ellenberger et al., 1994;  Shirakata et al., 1993).  Although the Arg568>Trp substitution of da^s22 probably does not change the protein’s structure (because this residue sits on the surface of helix 1), the change in the helix 1 surface charge that resulted from replacing a positively charged amino acid with a nonpolar residue might disrupt normal protein interactions and thereby hamper dimerization with normal binding partners.  Alternatively, since Arg568 is located only two residues away from the highly conserved Arg566 in the basic domain, which mediates DNA binding, perhaps the charge alteration that results from the da^s22 missense mutation reduces the efficiency of target site recognition and/or binding.

            Acknowledgments:  We thank T. Grigliatti for providing da^f75 and L.Yue for providing the fs(2)lyh mutant, which was identified subsequently as da^lyh by Smith and Cronmiller (2001).  This work was supported by a grant from the National Science Foundation.

            References:  Brown, N.L., S.W. Paddock, C.A. Sattler, C. Cronmiller, B.J. Thomas, and S.B. Carroll 1996, Dev. Biol. 179: 65-78;  Caudy, M., E.H. Grell, C. Dambly-Chaudière, A. Ghysen, L.Y. Jan, and Y.N. Jan 1988, Genes Dev. 2: 843-852;  Cronmiller, C., and T.W. Cline 1987, Cell 48: 479-487;  Cronmiller, C., P. Schedl, and T.W. Cline 1988, Genes Dev. 2: 1666-1676;  Cummings, C.A, and C. Cronmiller 1994, Development 120: 381-394;  Cronmiller, C., and C.A. Cummings 1993, Mech. Dev. 42: 159-169;  Ellenberger, T., D. Fass, M. Arnaud, and S.C. Harrison 1994, Genes Dev. 8: 970-980;  Murre, C., P.S. McCaw, and D. Baltimore 1989, Cell 56: 777-783;  Shirakata, M., F.K. Friedman, Q. Wei, and B.M. Paterson 1993, Genes Dev. 7: 2456-2470;  Smith, J.E., and C. Cronmiller 2001, Development 128: 4705-4714;  Voronova, A., and D. Baltimore 1990, Proc. Natl. Acad. Sci. U S A 87: 4722-4726.