NULL MUTATIONS IN DROSOPHILA ... - Semantic Scholar

Copyright 0 1985 by the Genetics Society of America

MOLECULAR ANALYSIS O F X-RAY-INDUCED ALCOHOL DEHYDROGENASE ( A m ) NULL MUTATIONS IN DROSOPHILA MELANOGASTER MARK R. KELLEY,’ INKA P. MIMS,2 CHRIS M. FARNET,’ SHERRY A. DICHARRY4 AND WILLIAM R. LEE’

Genetics Program, Department of Zoology and Physiology, Louisiana Stale University, Baton Rouge, Louisiana 70803 Manuscript received April 16, 1984 Revised copy accepted October 15, 1984 ABSTRACT

We have attempted to analyze at the molecular level mutants previously determined as having intragenic lesions caused by X-ray mutagenesis. C. S. Aaron isolated 33 null mutations at the Adh locus and in collaboration with other investigators classified 23 as deletions. Of the eight mutants analyzed here, only two produced a detectable ADH protein using the two-dimensional electrophoresis technique. Restriction endonuclease and Southern blot analysis showed that three of the mutants were normal compared t o the wild-type restriction pattern, with one of the three producing a mutant ADH protein. Among the five mutants that had altered restriction patterns, only one mutant produced a detectable mutant ADH protein. All the mutants produced a hybridizable mRNA when probed with the genomic clones, suggesting that the mutant phenotype was not due to transcriptional inhibition. T w o probable explanations proposed for these observations are (1) mutations may be due to deletions of one o r a few bases resulting in frameshifts to nonsense codons and premature termination of ADH peptide synthesis or (2) mutations may be a result of transitions to nonsense codons, again producing shortened ADH proteins. Those mutants producing a mutant polypeptide may have resulted from mutations to missense rather than nonsense codons. T h e five mutants showing an abnormal endonuclease Southern blot along with the 23 mutants previously shown to be deletions (28/33 or 85%)are associated with multiple DNA chain breaks. Although all of the DNA chain breaks are not necessarily associated with the mutant phenotype of the Adh locus, multiple DNA chain breaks are the most consistent characteristic of ionizing radiation damage to DNA.

RAYS have long been known to produce both intra- and intergenic muX tations, including alkali-labile lesions, purine, pyrimidine base and sugar damage, cross-links between two DNA strands or DNA and protein and single and double-strand breaks. The radiolysis types of damage are the apyrimidinic Present address: Genetics Department, T h e Rockefeller University, New York, New York 10012. address: Genetics Program, State University of New York, Stony Brook, New York 11974. Present address: Department o f Biochemistry and Molecular Biology, Harvard University, Cambridge, Massachusetts 02 138. Present address: Biochemistry Department, University of Texas, Austin, Texas 78704. To whom reprint requests should be sent: Department o f Zoology and Physiology, Louisiana State University, Baton Rouge, Louisiana 70803. I

* Present

’ ’

Genetics 109: 365-377 February, 1985.

366

M. R.

KELLEY E T A L .

and apurinic sites along with formation of thymine glycols and urea from hydroxyl radical attacks on thymine. An important characteristic of ionizing radiation is that energy is not dissipated at random but instead is concentrated in tracks. As has been shown by genetic analysis, deletions are the predominant type of mutational lesion recovered (AARON1979; ASHBURNER, AARONand TSUBOTA 1982). However, of the intragenic mutations produced by X ray the question of whether these mutations are single-site o r small deletions has not been clearly shown. We have attempted to analyze, at the molecular level, the types of damage caused by X-ray mutagenesis on a nonrepairing germ cell: the mature Drosophila sperm. In this analysis w e have excluded those mutants previously determined as intergenic lesions. T h e Alcohol dehydrogenase ( A d h ) locus of Drosophila was chosen for this type of analysis predominantly because of its well-known molecular biology, ease of recovering mutants without alcohol dehydrogenase (ADH) activity and its apparent similarity with other loci in the types and frequency of mutations produced by X rays and chemicals (ASHBURNER, AARONand TSUBOTA 1982). In an X-ray mutagenesis study in 1979, AARONrecovered 33 mutations at the Adh locus. These included six at 500 rads, six at 1000 rads, 19 at 3000 rads and two spontaneous mutants. Twenty-two of the 3 1 (7 1%) induced and one of the two spontaneous mutants have been classified as deletions using inter se crosses for deletion mapping and standard genetic mapping techniques (AARON1979; ASHBURNER, AARONand TSUBOTA 1982). Two mutants were lost before the present experiment was initiated. T h e remaining seven induced mutants and one spontaneous mutant ( A d l ~ ~ ' ,would ~ ~ ' ~normally ) be classified as point mutations; however, we have shown that mutants AdhnLA' and Adh"LA24H show additional bands, and two mutants, Adh"L"74and AdhnrAAxo show missing bands by restriction endonuclease and Southern blot analysis using a 4.8-kb SAC1 genomic probe. More differences were observed when the iarger 1 i .8kb sAF2 genomic probe was used, perhaps indicating alterations farther from the structural gene. Only mutants AdhnLAX0and Adh"L"'4g made a detectable protein by two-dimensional gel electrophoresis (O'FARRELL1975) analysis, whereas all produced a hybridizable messenger RNA (mRNA). Therefore, only three of the original 31 (10%) induced mutants at the Adh locus showed no alterations from the wild type except for lack of a functional ADH protein. Of these three, only mutant Adh"L"249produced a nonfunctional ADH protein with the normal molecular weight and total charge, whereas AdhnLA7"and AdvL.4405 did not produce a recognizable ADH protein as determined by twodimensional gel electrophoresis. Since these three intragenic mutants might be due to very small deletions, the final analysis concerning these Adh null mutants must await our in vitro translation and detailed mRNA and DNA sequence studies. Certainly most of the X-ray-induced mutants were induced through a mechanism of multiple chain breaks in DNA. MATERIALS AND METHODS

Chemicals and enzymes: Restriction en7ymes were purchased from either Bethesda Research Laboratories or New England Biolabs. Ampholines were obtained from LKB. Acrylamide, Apro-

MOLECULAR ANALYSIS OF ADH MUTANTS

367

tinin, agarose, formaldehyde and forniamide were from Sigma. Silver nitrate was obtained from Aldrich and Cooniassie blue R from Bio-Rad. RNA markers (16s and 23s) were from Miles Laboratories and Cronex intensifying screens were from DuPont. Stocks: X-ray mutants AdhnLA2,AdhnLA7',Adh"LA74,AdhnLAso,AdVUZ4', AdhnLAZ4', AdhnLA3I9 and Adh"LA379 were obtained from C. S. AARON(1979) and carried the markers b cn bw as did mutant ~dh"LA405from M. ASHBURNER. AdPLA3"is a spontaneous mutant and arose on chromosome AdhD pr cn (these mutants will be referred to by number only.) These strains were maintained over the second chroniosonie balancer In(2LRX1, Cy dp'"'AdhnBpr cn2 (LINDSLEY and GRELL1967). Mutants 2, 73, 74, 80, 249, 319 and 405 were placed over chromosome b, Dx2L)64j, a very large second chromosome deletion including the Adh region (WOODRUFF and ASHBURNER 1979a,b). Mutant 248 is homozygous viable and mutant 379, a known deletion for noc osp Adh, was carried over the formaldehyde deletion mutant A d P 3 p r cn, which uncovers 122 noc osp Adh I ? (ASHBURNER, AARONand TSUBOTA 1982). Mutant 405 is carried over the deletion mutant 379. T h e wild-type stock b AdhF cn bw, maintained frozen at -20" since 1979, was used as the control for all of the mutants since this was the stock in which the mutants were induced, except for mutant 319 (AARON 1979). Mutant 319 was also compared to b AdhF cn bw strain since the restriction pattern of AdhD pr cn had the same restriction-banding pattern for all enzymes used (M. R. KELLEY,unpublished data). Probes: T h e 4.8-kb genomic sACl (GOLDBERG1980) and Adh.cDNA (BENYAJATIet al. 1980) clones in pBR322 were obtained from W. SOFER.T h e 11.8-kb genomic clone sAF2 in plasmid pSV2 was obtained from T . MANIATIS.Probes and plasmid DNAs were prepared according to MANIATIS, FRITSCHand SAMBROOK (1982). Two-dimensional gel electrophoresis and silver staining: Five adult flies at least 5 days old were decapitated and homogenized in a 50-rl sample buffer with Aprotinin added to inhibit protease activity. T h e solution was centrifuged for 5 min at 12,000 X g. T h e supernatant was removed, and 2 5 pl were applied to the focusing gel. Each sample for the two-dimensional protein analysis consisted of 100-150 pg of crude soluble protein. Isoelectric focusing was performed according to the method of O'FARRELL(1975), except the second dimension gel was an 8% SDS-acrylamide gel. Silver staining was accomplished according to a procedure developed by M. ADAMS,Louisiana State University (personal communication). Isolation of DNA and RNAfrom adult Drosophila: DNA was isolated by grinding 1 g of flies in liquid nitrogen and homogenizing in 15 ml of 50 mM Tris-HCI, pH 8.5, 100 mM NaCI, 10 mM EDTA, 0.5% B-niercaptoethanol and 1% SDS. This solution was chloroform/isoamyl alcohol/ phenol (24: 1:25) extracted three times. DNA was ethanol precipitated from the aqueous phase, resuspended and centrifuged on a K1-ethidium bromide isopycnic gradient (KELLEYand LEE 1983) in the Sorvall TV-865 vertical rotor for 8-16 hr at 55,000 rpm. T h e nuclear DNA band was collected, butanol extracted to remove the ethidium bromide and dialyzed against 10 mM TrisHCI, pH 7.5, 1 mM EDTA (TE). RNA was isolated by the procedure of DEELEYet al. (1977) with the following modifications. After precipitation from the guanidine-HCI, the pellet was resuspended in T E , and 1 g/ml of CsCl was immediately added. This was placed over 0.3 volumes of a 1.8-g/ml CsCl cushion (FRYBERG et al. 1980) and centrifuged in a TV-865 vertical rotor for 5-8 hr at 55,000 rpm. T h e RNA was removed and ethanol precipitated. Proteinase K (100 pg/nil) was added to the resuspension for 1 hr at 65", followed by a phenol extraction and dialysis. T h e RNA was stored frozen at -20". Restriction endonuclease analysis: DNA, 10 pg, from each of the mutants and from the wild-type stock was cut with the various restriction enzymes for 12 hr and electrophoresed on a 1.5% agai-ose gel in 0.04 M Tris-acetate, 2 niM EDTA buffer (pH 8.0) with 0.5 pg/nil of ethidium bromide. Hind111 digests of A-DNA were used as molecular weight markers. DNA was bidirectioilally transferred to nitrocellulose according to the procedure of SMITHand SUMMERS (1 980). RNA dot blot analysis: Total adult RNA was spotted onto nitrocellulose according to the method described by THOMAS (1980). Hybridizations: Hybridizations of the SOUTHERN (1975) transfers and the RNA spot blots were done according to the method of MANIAITS, FRITSCHand SAMBROOK (1982) under stringent temperature (65") and washing conditions. Autoradiography was done at -70" with Dupont Cronex intensifying screens. SAC1, Adh.cDNA and sAF2 plasmids were nick-translated according to the procedure of RIGBYet al. (1977).

368

M.

R.

KELLEY ET AL. RESULTS

Two-dimensionalprotein analysis: The modified two-dimensional O’Farrell and silver-staining techniques were very sensitive in the analysis of ADH protein from the normal and wild-type mutants. Four mutants analyzed by us which have been previously cited in the literature as CRM- (antibody cross-reacting material negative) (SCHWARTZ and SOFER1976) were found not to produce a recognizable protein in the two-dimensional analysis used here (data not shown). Three found to be CRM+, or having ADH protein recognizable by the antibody, were detected as distinct spots on our gels in the ADH region. This is seen with ,yonB (Figures l b and 2b), an EMS-induced nonsense-transition mutant lacking the terminal 20 amino acids (KUBLI et al. 1982) in which molecular weight of the ADH monomer changed from 26,000 to 24,000 daltons. This loss of 20 amino acids (9%) is near the limit of detection of the ADH protein on the two-dimensional system (Figure lb). There was also a slight change in the total charge toward the anode, as evidenced by its movement in the horizontal direction (compare this with AdhF in Figure la). Single amino acid changes can easily be seen (Figure 1). AdhF is distinguishable from Adhs due to a single amino acid change of a threonine (neutral) in AdhF to a lysine (positive) in AdhS (Figure IC).Figure Id again shows AdhF and AdhS but also shows another isozyme of the ADH protein, AdhD, which has a glutamic acid (negative) instead of the glycine (neutral) in AdhF. Initial two-dimensional protein analysis of the null mutants revealed that only mutant 249 produced a recognizable protein (Figure 2e). However, when we refined our assay using only the soluble protein from five freshly killed flies instead of a lengthier extract preparation, mutant 80 was also found to make a detectable protein (Figure 2d). Mutant 80 showed a shift toward the anode which probably represented a change in charge due to substitutions for at least one amino acid from the AdhF-stock in which the mutants were induced (Figure 2a). These results with mutant 80 were checked by preparing the mutant protein by both methods coelectrophoresed on a two-dimensional gel. The results were consistent with our hypothesis that mutant 80 was highly labile and was quickly degraded in the absence of protease inhibitor. Early examples of ADH instability were seen by THATCHER (1977). Some mutants that did not have any recognizable ADH protein are also seen in Figure 2. Here, mutants 2 and 73 (Figures 2c and 2f) are shown without any recognizable ADH protein in comparison to AdhF (Figure 2a). This technique, especially when examining mutants at a particular locus, is much more informative and sensitive than the CRM test and can be performed quickly from just five flies. Furthermore, it is especially applicable when the protein being analyzed is a predominant one, such as ADH, which makes up 2% of Drosophila soluble proteins (BENYAJATI et al. 1980). mRNA analysis: Spot blot analysis (data not shown) showed that when probed with the sACl clone all of the mutants make hybridizable Adh-mRNA except mutant 379, which is a known Adh deletion and did not hybridize to the probe. The observation that the nondeletion mutants did make hybridizable AdhmRNA leads us to conclude that the mutations were not due to transcriptional inhibition.

MOLECULAR ANALYSTS OF ADH MUTANTS

369

d

C

FIGURE 1 .--I'w[)-diiiieilsion;il gel electrophoresis of (a) Adhb', (b) (;yO"', (c) AdhF and Adh' and (d) ,4dhF,Adhsand Adh". Isoelectric focusing is in the horizontal direction (anode to the right) and S1)S-pol).;icryl;imide gel electrophoresis in the vertical (direction of migration is from top to bottoin) direction. Gels were silver stained a s described in text. Arrowheads indicate either the A D H protein or where it is expected on the gel. The abundant and unidentified protein above A D H scrves ;IS a reference marker.

e

f

d FIGURE'L.-Silver-st;iined t\~o-tlimensionaIgels showing nilitants with shift in protein charge or those that produce no detectable AI)H protein. AdhF and CyOnBare shown a s references (a and b, respectively). Panel (c) 2. (d) 80. (e) 249 and (9 75.

370

M. R . KELLEY E T A L .

11.8 kb

s AF2

SAC 1 1 kb

Adb

f f hf I 893 399

ffpo I Hiin f I

1

Exons

87

2

3

4

96 405 264

FIGURE3.-Comparison of the 4.8-kb sACl and 11.8-kb sAF2 genomic probes. Also shown are the Hinfl and HpaI restriction sites. T h e 1536-bp fragment produced by the Hinfl includes the promoter, first 654-bp intron and the 5' half of the gene. T h e 399-bp fragment contains the 3' half of' the A d h structural gene. Cutting with HpaI and Hznfl splits the 1536-bp fragment into 643 and 893 bp. T h e 893-bp fragment includes the 654-bp intron and the 5' half of the gene, whereas the 643-bp fragment includes the 5' end of the adult mRNA and the promoter site. Also shown are the four exons of the adult niRNA, the leader sequence before exon 2 and the 3' untranslated region after exon 4. T h e coding and splice regions are depicted at the bottom of the figure.

Restriction endonuclease analysis: The sensitivity of restriction endonuclease mapping has been estimated to be no better than detecting a change of 50 base pairs (bp) or greater (ZACHAR and BINGHAM 1982). This does not include those mutations involving the restriction site where smaller changes, including single base changes, can be detected. The wild-type b AdhF cn bw stock which was used as the control had been maintained frozen at -20" since the initial experinient more than 5 yr ago, whereas the mutant stocks have been constantly expanded and crossed to various deletions. This change in maintenance between the control and mutant stocks was originally thought to account for some of the unusual aberrations we report here, but recent experiments on the expanded wild-type stock show no differences from those using the frozen flies (data not shown). The strain that spontaneous mutant 319 was recovered from, Adh" p r cn, also showed no differences from b AdhF cn bw. Two genomic probes were used in the analysis of the control and mutant restriction digests. The 4.8-kb sACl probe contains approximately 2 kb of DNA on either side of the Adh structural gene and is bordered by two EcoRI restriction sites. The 11.8-kb sAF2 genomic probe contains an additional 3.5 kb of DNA outside each EcoRI site and was bordered by Sac1 restriction sites (Figure 3). Other restriction sites important to this analysis are shown in Figure 3. Digestion with HhaI (Figure 4) and probing with the sACl clone gives normal restriction patterns for mutants 73, 74, 80, 249, 319 and 405, whereas mutant 2 has an additional 300-bp band. Mutant 248 has an additional 450-

371


1

2

3

4

5

r

7

FI(.ilRE 4.-/IpaI digest of wild-type and niittant DNAs probed with 52P-sAC:I. Lane ( I ) wild tylw, (2) 2. (3) 73, (4) 74, (5) 248, (6) 249. (7) 319, (8) 80. (9) 40.5. Arrowheads indicate additional b;~ndstiybritliping t o the sAC1 prohe. Molecul;tr weights of fragments determined froni Hind111 digests of X-DNA ( n o t sliown 011 gels).

1200

300

400

FIGLIRE~.-Airtara[liogr;i1,11 of' I f h n l genoniic digests Idlowii1g Southern transfer froni 1.5% gel. Filters were ~ ~ r o l xwith d "P-sAF2. I . a w ( I ) 379, (2) wild type. (3) 319. (4) 248, ( 5 ) 80. (6) 74, (7) 73, (8) 2. (9) 249. Arrodie;ids indicate missing and adtlitioiral bands. Molecul;ir w*iglits of ft;tgnienis tlelimnined froni Hind111 digests o f I;tmbd;i D N A (not showii on gels).

;ig;iiwse

bp band when digested with HhaI (Figure 5). Using the larger sAF2 probe (Figure 5) and digesting with the same enzyme shows that mutant 74 is missing a 1200-bp band fragment and has a new 900-bp fragment. This could be the result of a 300-bp deletion internal to two Hhal sites but must be outside the

372

M. R . KELLEY E T A L .

region covered by the sACl probe since these alterations are not seen when using the smaller SAC1 probe. The spontaneous mutant 319 has a pattern similar to that of mutant 74. It is missing the same 1200-bp fragment and has the extra 900-bp fragment but also additional 620- and 800-bp fragments. Again, none of these changes in 319 are seen with the SAC1 probe. The extra fragment in mutant 2 seen with the SAC1 probe is not seen on this gel because of its size. We suspect the change in the restriction pattern of spontaneous mutant 319 may be due to the loss of an insertion element that may be present near the Adh gene. This loss can leave behind an alteration in the restriction endonuclease pattern. Since the change is detected with only the sAF2 probe, we feel an enhancer region may be involved or the X rays stimulated transposon movement from this region. However, we have not proved either of these two theories to our satisfaction and they are currently under investigation. Only mutant 2 has any alterations in banding patterns, using the restriction enzyme HinfI and the sACl probe. This mutant has an extra large fragment of approximately 1600 bp. This extra fragment is also seen when the larger sAF2 probe is used (data not shown). Spontaneous mutant 319 is missing a band of close to 1100 bp and has an additional one of 520 bp when probed with sAF2 after HinfI digestion. Double digestion with HpaI and Hinfl and probing with sAF2 again reveals the extra 1600-bp fragment for mutant 2 (data not shown). This extra fragment is also seen when the shorter sACl probe is used (Figure 6). This extra Adh DNA must be inserted somewhere within the region maintained either by the CyOnBbalancer chromosome or the deletion 64J. The DNA must have originated from the 4.8-kb region since it hybridizes with the sACl probe. However, mutant 319 which again shows the missing and additional fragments when probed with sAF2 after HpaI/Hinfl digestion (not shown) appears normal when probed with sAC1. The alterations for this mutant are outside the 4.8-kb (sAC1) region but within the 11.8-kb (sAF2) region. The fragments of 643, 893 and 399 bp span the region from the promoter past the 3’ end of the Adh gene. These appear normal in all of the mutants analyzed with HpaI/Hinfl digestion, at least within the limits of detection. T h e possibility of any gross mutations (deletions) spanning the adult promoter region which lies in a 387-bp SalI-HpaI fragment (BENYAJATIet al. 1982, 1983) can be ruled out since any deletion spanning the HpaI-cutting site would be detected by the loss of this site. This is not seen in any of the mutants analyzed. Furthermore, mutants 74 and 80, after HpaI/Hinfl digestion and sACl hybridization, are missing the same 1400-bp fragment with no other noticeable alterations (Figure 6). Mutant 248 has all of the expected bands but also additional bands of 520 and 310 bp. Another digestion with EcoRI, Sal1 and BamHI shows that only mutant 2 has any detectable alterations. With the sACl probe mutant 2 has an additional fragment, and when the sAF2 probe is used it has two additional bands, one of which is the same as is seen with the SAC1 probe. Other digests with either additional enzymes or other combinations of enzymes support the aforementioned results but give no further information on alterations that may be the cause of the mutations.

MOLECULAR ANALYSIS O F ADH MUTANTS

373

FIGURE6.-Autoradiograph of Hpal- and Hinfl-digested DNAs run on a 1.5% agarose gel, transferred to nitrocellulose and probed with the '*P-sACl genomic clone. Arrowheads show the additional bands of mutants 2 and 248 and the missing bands of mutants 80 and 74. Lane (1) 80, (2) wild type, (3) 319. (4) 249. (5) 248, (6) 74, (7) 73, (8) 2. Molecular weights of fragments determined from Hindlll digests of lambda DNA (not shown on gels). DISCUSSION

T h e X-ray-induced mutants presented here were all initially isolated by AARON(1979) over an intragenic Adh null mutation therefore allowing for the recovery of deletions, both large and small. Most of the recovered mutants (7 1%) were characterized by classical genetic mapping techniques as deletions. T h e remaining set of mutants that showed no cytological aberrations (ASHBURNER, AARONand TSUBOTA 1982) or reduced recombination frequencies were analyzed on the molecular level for changes in the DNA, RNA or protein produced by the Adh locus. Of the seven induced mutants studied here, only two (80 and 249) produced a recognizable ADH protein by the highly sensitive O'FARRELL( 1975) electrophoresis and silver-staining technique modified for detection of Drosophila ADH protein. One of these two, mutant 80, produced a highly labile protein, whereas the other mutant, 249, was very stable. These et al. 1977) to find two are currently undergoing further analysis (CLEVELAND which amino acids are altered from the normal ADHF protein. Failure to detect an ADH protein can be interpreted as (1) a change in the ADH protein that is greater than substitution of a single amino acid with opposite charge, (2) a change of more than 10% in molecular weight, otherwise the ADH protein is indistinguishable from other proteins or (3) at least an 80% reduction in the amount of protein produced. Analysis on X-ray-induced mutants at the white locus in Drosophila resulted in six mutants recovered (ZACHAR and BINGHAM1982). Three were large

374

M.

R.

KELLEY ET AL. \

deletions (1.2, 16 and 30 kb), whereas the others were essentially identical to white+ by restriction analysis except one which was missing a Hind111 site (ZACHAR and BINGHAM1982). Using seven allozyme loci and 7-irradiation, (1980) showed that, of the seven null mutants RACINE,LANGLEY and VOELKER recovered, three of five were CRM+ and two of six formed active heterodimers with an active normal allele. However, since a low chronic dose was used over a period of 15 generations, the accumulation of spontaneous mutants cannot be ruled out. They also proposed that the CRM- nulls could be either physical deletions of a locus, missense or nonsense mutations or frameshifts with the resulting inactive protein degraded. Since we have analyzed the mutants presented here by restriction mapping and Southern blotting we can rule out the first idea of physical deletions of the Adh locus, at least for any large deletion. Searching the known DNA sequence using the NIH-Prophet Genbank System, we find three possible early nonsense codons if a frameshift of either + I or -2 nucleotides is incurred. Two are within the first ten amino acids and all three occur in exon 2, the amino terminal end of the ADH enzyme (using the nomenclature of KREITMAN 1983). Exon 3 has seven potential nonsense codons with a shift of + 1 or -2 nucleotides and two with a shift of -1 or +2 nucleotides. Therefore, any mutation, whether a single base or a small deletion causing a frameshift early in the Adh structural gene results in an extremely small polypeptide that would not be detected by the methods employed here or by the CRM test. Furthermore, these polypeptides are most likely very labile. Induced mutants 2, 74, 80 and 248 are different from the wild type using the SAC1 probe (Figure 6), and mutants 2, 74 and spontaneous mutant 319 show differences using the longer sAF2 genomic clone. Mutant 2 consistently shows additional Adh material in Southern blot experiments, perhaps representing a rearrangement or duplication of Adh-DNA during replication subsequent to X-ray mutagenesis. This possibility is also currently under analysis. The five mutants showing differences in the restriction pattern have normal bands associated with the Adh locus, suggesting that the structural changes lie outside the Adh locus. However, the mutant 248 has recently been partially sequenced by W. CHIA,who finds it to be a partial duplication of exons 2 and 3. This partial duplication would give a near normal banding pattern for the three bands associated wiih the Adh locus plus an additional small band as seen in our Southern blot experiments (Figure 6). We cannot exclude similar explanations for the other mutants until they are sequenced, although we question whether all of the alterations in the restriction pattern are associated with the cause of the Adh null activity. Adh null activity can be the result of changes some distance from the structural gene. For example, GOLDBERG, POSAKONY and MANIATIS(1983) using P element transformation showed that the 11.8-kb fragment between the Sac1 sites (Figure 3) contains cis-acting regions that were necessary for the correct regulation of Adh gene expression. Whether the alterations seen when probing with the sAF2 clone are indicative of alterations affecting these regulatory regions is currently under investigation. This latter hypothesis is not supported by the results of the spot blot experiments which


375

show the mutants do make Adh-mRNA that will hybridize with the genomic probes. If the regulatory regions had been affected, we would have expected the action to be on the transcriptional rather than the translational level. The alterations observed outside the structural gene might not be directly related to the Adh null mutation but may have been associated with the X-ray ionization track that did multiple damage to this region of the chromosome. A single track could result in losses or additions of a few bases in the structural gene resulting in a frameshift yielding chain termination and larger changes outside the structural gene yielding an alteration in the restriction patterns. T h e two that produce a nonactive ADH protein are probably the result of base substitutions or point mutations. The idea that base alterations leading to amino acid substitutions produce Adh null mutants is supported by KREITMAN'S (1983) work in which, with 43 polymorphisms from five natural populations studied by DNA sequence analysis, only one (AdhF to Adh") resulted in an amino acid change. This implies that amino acid substitutions are normally deleterious for the Adh locus. Recently N-ethyl-N-nitrosourea (ENU) Adh mutants induced in both larval and adult Drosophila have been recovered. Of the three mutants recovered from treatment of adults, two are intragenic with one producing an ADH protein, and the third appears to be a deletion uncovering the two adjacent loci osp and noc. The fact that a direct-acting alkylating agent can produce such a large deletion is intriguing and may have been missed in past experiments with EMS, since none of the 13 mutants recovered were cytologically detectable deletions (ASHBURNER, AARONand TSUBOTA 1982). The frequency of deletions produced by chemical mutagens such as EMS or ENU may have been underestimated due to the method of mutant recovery since some experiments used deletions instead of Adh null chromosomes for recovery of the Adh mutant. Only two of the eight mutants (25%) analyzed here or two of the original 3 1 (6%)induced mutants produced a detectable protein using two-dimensional gel technique. Restriction endonuclease and Southern blot studies using the shorter SAC1 genomic probe showed that four of the eight mutants have altered restriction patterns from the wild types; however, none of the mutants appeared altered in the region of the Adh structural gene. When the larger sAF2 genomic probe was used, one additional mutant also showed an altered pattern, again outside of the structural gene region. Therefore, five of the eight (63%) of the mutants previously determined not to be large deletions had an altered DNA restriction pattern, including the spontaneous mutant. Of the 31 X-ray-induced mutations at the Adh locus, only three (10%) were observed to be normal using restriction endonuclease and Southern blot analysis. Two mutants were lost before they could be tested. The three normal appearing mutants (73, 249 and 405) had only one (249) that made a detectable ADH protein. The five mutants that had altered restriction patterns (2, 74, 80, 248 and 319) outside the Adh structural gene region also only had one (80) that produced a detectable ADH protein. There are two possible explanations for mutants with transcription (mRNA)

376

M. R. KELLEY ET AL.

and normal restriction patterns but no observable ADH proteins: (1) the mutations may be due to deletions of one or a few bases resulting in frameshifts to nonsense codons that allow normal restriction patterns and mRNA production but produce premature terminations of ADH polypeptide synthesis as described or (2) these mutations may be a result of transitions within the Adh structural gene to nonsense codons. A search of the wild-type Adh structural sequence in Genbank showed that there are 20 potential C to T and 14 G to A transition sites that would result in early termination. If mutation to nonsense codons occur, then normal restriction patterns, mRNA synthesis and splicing would be observed, but shortened ADH polypeptides would result and might not be detected by two-dimensional gel electrophoresis. The two mutants that produced a detectable ADH protein might be the result of a missense rather than a nonsense mutation. We would like to thank M. ADAMSfor help on the O’FARRELL protein analysis and silver staining procedure. This investigation was supported in part by contract DE-AS05-76EV03728 from the Department of Energy, Public Health Service research grant PO1 ES03347 from the National Institute of Environmental Health Sciences and in part by a Sigma XI Grant-in-Aid Research to M.R.K. LITERATURE CITED

AARONC. S., 1979 X-ray induced mutations affecting the level of the enzyme alcohol dehydrogenase in Drosophila melanogaster: frequency and genetic analysis of the null enzyme mutants. Mutat. Res. 63: 127-137. ASHBURNER, M., C. S. AARONand S. TSUBOTA, 1982 T h e genetics of a small autosomal region of Drosophila melanogaster, including the structural gene for alcohol dehydrogenase. V. Characterization of X-ray-induced Adh null mutations. Genetics 102: 42 1-435. BENYAJATI, C., A. R. PLACE,N. WANG,E. PENZand W. SOFER,1982 Deletions at intervening sequence splice sites in the alcohol dehydrogenase gene of Drosophila. Nucleic Acids Res. 10: 7261-7272. 1983 T h e messenger RNA for BENYAJATI, C., N. SPOEREL,H. HAYMERLE and M. ASHBURNER, alcohol dehydrogenase in Drosophila melanoguster differs in its 5’ end in different developmental stages. Cell 33: 125-133. BENYAJATI, C., N. WANG,A. REDDY,E. WEINBERC and W. SOFER,1980 Alcohol dehydrogenase in Drosophila: isolation and characterization of messenger RNA and cDNA clone. Nucleic Acids Res. 8: 5649-5667. and U. K. LAEMMLI, 1977 Peptide mapping CLEVELAND, D. W., S. G. FISCHER,M. W. KIRSCHNER by limited proteolysis in sodium dodecyl sulfate and analysis by gel electrophoresis. J. Biol. Cheni. 252: 1102-1 106. DEELEY,R. G., J. I. GORDON,A. T. H. BURNS,K. P. MULLINIX,M. BINA-STEINand R. F. GOLDBERGER, 1977 Primary activation of the vitellogenin gene in the rooster. J. Biol. Chem. 252: 8310-8319. FYRBERG, E. A., K. L. KINDLE,N. DAVIDSON and A. SODJA,1980 T h e actin genes of Drosophila: a dispersed multigene family. Cell 19: 365-378. GOLDBERG, D. A., 1980 Isolation and partial characterization of the Drosophila alcohol dehydrogenase gene. Proc. Natl. Acad. Sci. USA 77: 5794-5798. GOLDBERG, D. A., J. W. POSAKONY and T. MANIATIS,1983 Correct developmental expression of a cloned alcohol dehydrogenase gene transduced into the Drosophila germ line. Cell 34: 5973.


377

KELLEY,M. R. and W. R. LEE, 1983 Mutagenesis in oocytes of Drosophila melanogaster. I. Scheduled synthesis of nuclear and mitochondrial DNA apd urxheduled DNA synthesis. Genetics 104: 279-299. KREITMAN, M., 1983 Nucleotide polymorphism at the alcohol dehydrogenase locus of Drosophila melanogaster. Nature 304 4 12-4 17. KUBLI,E., T . SCHMIDT,P. F. MARTINand W. SOFER,1982 I n vitro suppression of a nonsense mutant of Drosophila melanogaster. Nucleic Acids Res. 1 0 7145-7152. LINDSLEY, D. L. and E. H. GRELL,1967 Genetic variations of Drosophila melanogaster. Carnegie Inst. Wash. Publ. 627. MANIATIS,T . , E. F. FRITSCHand J. SAMBROOK, 1982 Molecular Cloning. Cold Springs Harbor Laboratories, Cold Spring Harbor, New York ~ ' F A R R E LP. L , H., 1975 High resolution two-dimensional electrophoresis of proteins. J. Biol. Cheni. 250: 4007-4021. RACINE,R. R., C. H. LANGLEY and R. A. VOELKER,1980 Enzyme mutants induced by low doserate gamma-irradiation in Drosophila: frequency and characterization. Environ. Mutagen. 2: 167-178. RIGBY,P. W. J., M. DIECKMANN, C. RHODESand P. BERG, 1977 Labeling deoxyribonucleic acid to high specific activity in vitro by ick translation with DNA polymerase I. J. Mol. Biol. 113: 237-25 1 . SCHWARTZ,M. and W. SOFER, 1976 Alcohol dehydrogenase-negative mutants in Drosophila: defects at the structural locus? Genetics 83: 125-136. SMITH,G. and M. D. SUMMERS, 1980 T h e bidirectional transfer of DNA and RNA to nitrocellulose or diazobenzyloxymethyl-paper. Anal. Biochem. 109: 123-1 29. SOUTHERN,E., 1975 Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503-517. THATCHER, D. R., 1977 Enzyme instability and proteolysis during the purification of an alcohol dehydrogenase from Drosophila melanogaster. Biochem. J. 163: 3 17-323. THOMAS, P. S., 1980 Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc. Natl. Acad. Sci. USA 77: 5201-5205.

WOODRUFF,R. C. and M. ASHBURNER,1979a T h e genetics of a small autosomal region of Drosophila melanogaster containing the structural gene for alcohol dehydrogenase. I. Characterization of deficiencies and mapping of Adh and visible mutations. Genetics 92: 1 17-1 32. WOODRUFF,R. C. and M. ASHBURNER,1979b T h e genetics of a small autosomal region of Drosophila melanogaster containing the structural gene for alcohol dehydrogenase. 11. Lethal mutations in the region. Genetics 92: 133-149. ZACHAR,2. and P. M. BINGHAM,1982 Regulation of white locus expression: the structure of mutant alleles at the white locus of Drosophila melanogaster. Cell 3 0 529-541. Communicating editor: B. W. GLICKMAN