Evidence for Regulatory Variants of the Dopa Hypersensitive Loci in Drosophila Decarboxylase and Alpha-methyldopa

We have analyzed two variants of Drosophila melanogaster (RS and RE) which lead to the dual phenotype of elevated DDC activity and increased resistance to dietary alpha-methyldopa relative to Oregon-R controls. Both phenotypes show tight genetic linkage to the dopa decarboxylase, Ddc, and l(2)amd genes (i.e., < 0.05 cM distant). We find that low (Oregon-R), medium (RS) and high (RE and Canton-S) levels of DDC activity seen at both pupariation and eclosion in these strains are completely accounted for by differences in accumulation of DDC protein as measured by immunoprecipitation. Genetic reconstruction experiments in which Ddc+ and amd+ gene doses are varied show that increasing DDC activity does not lead to a measurable increase in resistance to dietary alphamethyldopa. This suggests that the increased resistance to dietary alpha-methyldopa is not the result of increased DDC activity but, rather, results from increased 1(2)amd+ activity. Both cytogenetic and molecular analyses indicate that these overproduction variants are not the result of small duplications of the Ddc and amd genes, nor are they associated with small (2100 bp) insertions or deletions. Measurements of DDC activity in wild-type strains of Drosophila reveal a unimodal distribution of activity levels with the Canton-S and RE strains at the high end of the scale, the Oregon-R control at the low end and RS near the modal value. We conclude that accumulated changes in a genetic element (or elements) in close proximity to the Ddc+ and amd+ genes lead to the coordinated changes in the expression of the Ddc and amd genes in these strains. OCALIZATION and characterization of mutations which lead to altered L gene regulation can identify elements involved in control of gene expression and reveal aspects of the mechanisms of control. In metazoans, lesions are frequently recovered which lead to moderate alterations in the level of gene expression. Some of these affect the temporal pattern of expression (e.g., aldehyde oxidase, DICKINSON 1975; dopa decarboxylase, ESTELLE and HODCETTS 1984a, b) or the tissue specificity of expression (e.g., amylase, DOANE et al. 1983; xanthine dehydrogenase, CLARK et al. 1984). However, most of the Abbrevidtions: DDC = Dopa decarboxylase en7yme; cM = centimorgan; CRM = Cross-reacting material. ' To whom correspondence should be addressed. Genetics 112: 249-265 February, 1986 250 J. L. MARSH AND T. R. F. WRIGHT regulatory mutants described in Drosophila, as well as in other metazoans, have been recovered as naturally occurring variants from various populations or strains rather than from direct mutagenesis (e.g., in Drosophila, RABINOW and DICKINSON 1981; DOANE 1980; ESTELLE and HODGETTS 1984a; in mice, PAIGEN 1979; and in maize, CHANDLEE and SCANDALIOS 1984). LAURIE-AHLBERG et al. (1982) have suggested that naturally occurring polymorphisms of both cisand trans-acting activity modifiers may provide an important source of variation for adaptive evolutionary change. However, few putative cis-acting activity regulators have actually been subjected to the detailed analysis necessary to distinguish between altered gene expression and structural alterations. We have examined low, medium and high activity variants of the Ddc and, possibly, amd genes to determine whether these changes might be due to altered gene regulation. Dopa decarboxylase (DDC) (EC 4.1.1.26) in Drosophila is under stringent tissue and temporal control (LUNAN and MITCHELL 1969; MARSH and WRIGHT 1980) and is vital to the organism (WRIGHT, BEWLEY and SHERALD 1976). The 1(2)amd(amd) gene is located immediately adjacent (0.002 cM) to the dopa decarboxylase gene (Ddc) and is functionally related to Ddc by virtue of its interaction with alpha-methyldopa (alpha MD) and by its affect on the cuticle. Mutants of the amd locus do not affect Ddc activity in any tissue, nor do they affect soluble phenol oxidase or dopamine acetyltransferase activity (MARSH and WRIGHT 1979). However, embryos homozygous for amd die at the embryonic/larval boundary exhibiting exceptionally friable cuticles and necrotic anal organs, suggesting a role in cuticle formation (WRIGHT 1977). The amd+ product is required for resistance to dietary administration of structural analogs of dopa (e.g., alpha-methyldopa). Although a large number of induced lethal alleles (50 for Ddc and 33 for amd) have been recovered (WRIGHT et al. 1982), none of these have yet been identified as alterations in noncoding regulatory elements. We have observed natural activity variants that lead to both elevated dopa decarboxylase activity and to increased resistance to dietary alpha-methyldopa relative to Oregon-R controls. In this report we examine the possible regulatory nature of these variants and document their affects on the level of activity of the Drosophila Ddc and amd genes. Specifically, we determined that the increases in DDC activity are completely accounted for by increased levels of DDC crossreacting material and not from a greater catalytic efficiency of the enzyme. Genetic reconstruction experiments suggest that the increased resistance to dietary alpha-methyldopa most likely results from increased expression of the amd+ gene product rather than from the increase in DDC activity. Cytogenetic and molecular analysis indicates that the altered regulation in these variants results from accumulated changes in a genetic element (or elements) which affect the expression of both the Ddc and amd genes. MATERIALS AND METHODS Drosophila strains: The variant in the R” strain was originally identified in a survey of 17 laboratory stocks with altered adult pigmentation. Young adults of a speck’ REGULATORY VARIANTS OF O d C 25 1 blistered' stock (sf?, 2-107 and bs', 2-107.3; LINDSLEY and GRELL 1968) had higher DDC activity and a higher LD50 to dietary alpha MD than Oregon-R-derived controls (SHERALD and WRIGHT 1974). The variant in the R" strain that also increases both DDC activity and resistance to alpha MD was originally recovered as an alpha MD resistant line (RM1) from an EMS mutagenesis screen using a strain that was presumed to be isogenic for the Oregon-R6 second chromosome (SHERALD and WRIGHT 1974). The chromosomes used in this study were placed in a common genetic background to minimize variation. The RS, RE and C1A (Oregon-R-derived control) strains are derived, respectively, from the S, R and C strains of SHERALD and WRIGHT (1974). The S, R and C strains had their first and third chromosomes replaced by chromosomes from an isogenic Oregon-R strain. The C strain was also made isogenic for a lethal free second chromosome from the Oregon-R strain. For this study, the RS, RE and C1A strains were constructed by recovering exchange events which replaced all of the second chromosome, except the 1.5 map units between rdo and pr, with the lethal-free second chromosome from the Oregon-R-derived C strain. Analysis of restriction site polymorphisms by blotting of genomic DNA from the different strains indicated that both the RL nd RS strains exhibit restriction site polymorphisms which are similar to the CantonS strain and are unlike the Oregon-R-derived strains. The Oregon-R and Canton-S lines derive from the Yale stock collection and have been maintained for 15 yr at the University of Virginia. The wild-type strains of D. melanogaster and one strain of D. simulans were obtained from the Bowling Green Drosophila Stock Center. These strains originate from widely separated geographical locations, but have been maintained as laboratory cultures for many years. Chromosomal aberrations used include the following: Df(2L)TW130 = Df(2L)37Bg-C1;37D1-2, Df(2L)TW158 = Df(2L)37B2-8;37E2-F4. Dp(2;1)C239 = Dp(2;1)7A-B;36C;39E. Dp(2;I)AT: New order = lA-7A/36D1,2-37D1,-2/5A-20. Dp(2;Y)H1,36B4-37F/39C40F. For a description of other genetic markers see LINDSLEY and GRELL (1968). Resistance to dietary alpha-methyldopa: Food containing alpha-methyldopa was prepared by autoclaving dried yeast-agar-dextrose medium (CARPENTER 1950) for 30 min at 15 p.s.i., then adding 10 ml of Tegosept M (methyl p-hydroxybenzoic acid, 10% in 95% ethanol) and 10 ml propionic acid (0.5%) to 2 liters of food, followed by 40 ml ascorbic acid neutralized with sodium bicarbonate (5 mg ascorbate + 2.5 g bicarbonatel 100 ml H'O) to prevent breakdown of alpha MD in the food. When the food temperature had dropped to ca. 50°, the inhibitor was added as an aqueous solution, and the food was poured immediately (SPARROW and WRIGHT 1974). T o determine the LD50 of alpha MD, 50 or 100 eggs from flies of the appropriate genotype were placed on a piece of moist, dark blotting paper in a dairy creamer containing fresly made inhibitor-bearing food. The results are expressed as the percentage of hatched eggs surviving to eclosion relative to the number surviving in the absence of inhibitor. The data for amdH', RE, and C1A (Figure 1) represent the means of four determinations, each involving over 300 eggs per inhibitor concentration. The data for RS and d2 represent the means of two such determinations. Animals with varying doses of Ddc+ (Figure 3) were obtained by collecting eggs from a mating of rdo hk Ddcn8pr cnlCyO females to Dp(2;Y)Hl; rdo hk Ddc"'pr/ Df(2L)TW130,rdo pr cn males, and resistance to various concentrations of alpha-methyldopa was determined as described above. Survival of each genotype on the various concentrations of alpha-methyldopa is reported as percentage of survival relative to 100% survival of each genotype on standard food without inhibitor. Dp(2;Y)HI (HODGETTS 1980) is a duplication carrying a normal copy of Ddc+ and all the genes in the deficiency region [Df(2L)TWI30] and is present in all males from this cross. DflCyO heterozygotes have curly wings and bright orange eyes resulting from the interaction of the two eye-color mutations, pr and cn. The duplication covers the entire deficiency region plus rdo and hk, but not p r or cn. The gene dose ratios for this cross are given with the genotypes and phenotypes in the legend to Figure 3. T o obtain animals with varying doses of amd+ as shown in Figure 4, eggs were 252 J. L. MARSH AND T. R. F. WRIGHT collected from a mating of 1(2)amdH"'cn bw/CyO females to Dp(2;Y)Hl; dp b amdHLZ1 p r / Df (2L)TW130,rdo p r cn males, and resistance to various concentrations of alphamethyldopa was determined as described above. The gene dose ratios, phenotypes and genotypes for the cross in Figure 4 are given in the legend. For the data shown in Figure 5, eggs were collected from matings of rdo hk Ddc"'pr cn/CyO males to both rdo hk Ddcn5pr/Cy0 and rdo hk Ddc"'pr/CyO females. The progeny of these crosses include partially complementing heterozygotes with <5% DDC activity. Control genotypes tested in parallel on the same food include Ddc"'lCy0, Ddc"'/CyO and Ddc"'/CyO, each with <50% activity (data not shown), as well as the double heterozygotes of two fully complementing lethals in nearby genes; namely rdo hk 1(2)37Ca' prlrdo hk 1(2)37Cb' p r (shown as +/+ in the figure) and the parental genotypes rdo hk 1(2)37Ca'pr/CyO and rdo hk 1(2)37Cb' pr/CyO (data not shown). 1(2)37Ca and 1(2)37Cb are cO.01 cM proximal to Ddc. All five mutations were induced in the same parental chromosomes in the same screen. DDC activities are n5/n8 = 3.7 -+ 1.5%; n8/nl = 2.7 k 2.3% (WRIGHT et al. 1982) and n8/CyO = 40%; n5/CyO = 33%; nl/CyO = 33%; 1(2)37Ca' or 1(2)37Cb1/Cy0 = 100% (WRIGHT, BEWLEY and SHERALD 1976). DDC determinations: Crude extracts were prepared at about 3 mg/ml protein ( i .e . , about ten 0-2-hr adults or white prepupae/ml) in 0.1 M phosphate buffer, pH 7.1, 0.3 M sucrose, and 0.2 mM phenylthiourea to prevent melanin formation. DDC enzyme activity was determined in triplicate by the micro-liquid-cation exchange assay of MCCAMAN, MCCAMAN and LEES (1972), with slight modifications. Protein was assayed by the method of LOWRY et al. (1951) using four replicates per homogenate. Typical activity in the Oregon-R control strain CIA was 12 pmol dopamine/30 min/pg protein for a signal of ca. 4600 cpm as assayed. In control experiments for Figures 3 and 4, DDC-specific activity was determined in homogenates of flies bearing 1, 2 and 3 copies of the Ddc+ gene. One dose flies were Df (2L)TW130, rdo p r cn/CyO and Df (2L)TW158/CyO. Two doses were Canton-S; Dp(2;l)C239; Df (2L)130,rdo p r cn/CyO; Dp(2 ; l )AT;Df (ZL)TWl3O,rdo p r cn/CyO; Dp(2;Y)Hl ;Df (2L)TW130,rdo p r cn/CyO, and three doses were Dp(2;l)C239;+/CyO and Dp(2;qHl;+/CyO. In all cases, DDC activity is a quantal function of gene dose. Preparation of antibodies: DDC was purified from larvae approximately 200-fold by slight modifications of the method of CLARK et al. (1978). The most purified fractions were electrophoresed on preparative nondenaturing 7.5% acrylamide gels, stained briefly (5 min) in Coomassie blue, 50% methanol, 7% acetic acid at 37", and destained 1 h at 37" in distilled HnO. Portions of longitudinal slices were assayed for DDC activity. A single band giving >85% of all activity detected on the gel and containing an estimated 50 pg or less of DDC was lyophilized, crushed in a mortar and resuspended in PBS (0.15 M NaCI, 10 m M phosphate, pH 7.1). Preimmune serum was collected on day 0 for controls. On day 3, rabbits were injected intramuscularly with a 1: 1 mixture of acrylamide suspension and Freund's adjuvant, and a sample of serum (5 ml) was collected on day 25. On day 27, a booster of acrylamide suspension'without adjuvant was administered intradermally, and 50 ml of serum was collected from fasted animals 9 days later, on day 35. This regimen gave a high titer of anti-DDC activity. Subsequent boosts were administered on day 66 and 388, followed by bleeding 7 and 8 days later, respectively. IgC was partially purified by 50% ammonium sulfate precipitation. Precipitation assay: IgC was diluted in 0.5 M phosphate buffer, pH 7.1, and 5 pl of diluted IgG mixed with 20 PI DDC supernatant, incubated 2 h at 37" and centrifuged 5 min in an Eppendorf microfuge to precipitate antibody-antigen complexes. Supernatant (3 p l ) was assayed in a 10 fil reaction mix, as above, in triplicate to determine the amount of DDC remaining in solution. Control experiments in which the pellet was resuspended and assayed revealed >99% of the DDC activity remaining in the pellet. Thus, the antibodies precipitate but do not inactivate the enzyme. DNA analysis: DNA was prepared by a scaled-up version of the method of BENDER, SPIERER and HOGNESS (1 983). Restriction digests employed a core buffer suggested by REGULATORY VARIANTS OF DdC 253 P. O'FARRELL that gives final concentrations of 33 mM Tris-acetate, pH 7.9, 66 mM potassium acetate, 10 mM MgC12, 0.5 m M DTT (dithiothreitol) and 100 rg/ml BSA. Electrophoresis and blotting were performed as described by SOUTHERN (1975). Nicktranslation was performed as described by RIGBY et al. (1977). Hybridization was in 50% formamide (MCB), 5X SSC, 1X Denhardt's solution, 10% dextran sulfate, 100 pg/ml salmon sperm DNA and 40 mM sodium phosphate, pH 6.8, at 42" for at least 16 h. Filters were washed at 65" in successive 500-ml 30-min washes of 0.1 SDS in 3X SSC, lx SSC, 0.3X and, finally, 0.1X SSC, followed by drying and exposure to film.