Isolation, Behavior and Cytogenetics

The dissonance mutant of courtship song was induced by chemical mutagenesis. This X-chromosoma1 mutation causes the D. melanogaster male’s acoustical output, resulting from his wing vibrations directed at a female, to include very long and loud tone “pulses.” Yet, a given train of pulses starts out as normal, with the signals in all but the shortest singing bouts eventually becoming polycyclic and high-amplitude. The aberrant songs caused by diss (map position, 1-52; cytological interval, 14C1-2 to 14C4-5) were quantitatively compared to those produced by mutant cacophony males, whose pulses are much more uniformly polycyclic (due to a mutation mapping elsewhere on the X chromosome). Males or females expressing diss are normal in several “general” behaviors. Yet diss males not only sing abnormally, but they also exhibit longer-than-normal mating latencies in their courtship of females. These decrements seem to be associated, at least in part, with visually aberrant behavior of diss flies-measured with regard to male courtships per se, and also in tests of more general visual responses. Such defects were found when testing diss males or females, and the genetic etiology of the visual impairments were provisionally mapped to the same locus to which the song abnormality has been localized. Neurogenetic connections between the control of courtship singing behavior and visual system functions are discussed with respect to the new song mutation (diss) and the older one (cac)-which also turned out to be genetically related to a mutation that causes abnormalities of light-induced behavior and physiology. T HE courtship song of a Drosophila male is produced by his wing vibrations that occur as he orients toward and follows a female (BENNETCLARK and EWING 1970). These sounds play an important role in the mating success of many species of fruit flies, each of which has its own distinct song (e.g., BENNET-CLARK and EWING 1969; SCHILCHER 1976b; COWLING and BURNET 1981; KYRIACOU and HALL 1986). The male courtship song of Drosophila melanogaster consists of tone pulse trains, interspersed with ca. 160 Hz hums (reviews: EWING 1977a; HALL 1982). The interpulse interval (IPI; ca. 35 ms for D. melanogaster) serves as a recognition signal for the conspecific females (BENNET-CLARK and EWING 1969; SCHILCHER 1976a, b; KYRIACOU and HALL 1982). Although a male’s mating success is dramatically reduced by the elimination of song through surgical removal of his wings, the courtship song is not a requirement for mating, in that wingless males do mate eventually (e.g., EWING, 1964; ROBERTSON 1982; SCHILCHER 1976b; KYRIACOU and HALL 1982, 1986). The first single-gene mutation that was found specifically to alter the pulse song is cacophony (cac, SCHILCHER 1976c, 1977). Individual pulses in the mutant song are polycyclic and larger in amplitude than the normal song. This pulse-song abnormality of cac appears to be specific, because courtship hums Genetics 118: 267-285 (February, 1988) and flight wing-beats (SCHILCHER 1977), plus flying ability (KULKARNI and HALL 1987), are normal. Recent genetic and behavioral analyses of this mutation (KULKARNI and HALL 1987) showed that it causes only a limited array of well defined defects: longer and louder tone pulses in the song, and marginally depressed locomotor activity. Latencies to initiation of mating with females (measured for cac males carrying X chromosomes that had undergone recombination with certain marker-bearing Xs) were normal, i.e., in spite of the aberrant sounds produced by the mutant. The neural control of the courtship song has been investigated in mosaics. Normal pulse song was found to be most closely associated with genotypically male tissue in the ventral thoracic ganglia (SCHILCHER and HALL 1979). Mosaic analysis has also suggested a thoracic focus for the cac mutant song (HALL et al. It seemed as if new mutations that might, with a reasonable degree of specificity, perturb the courtship song should be searched for. Thus, we have isolated the dissonance (diss) mutant, to augment the neurogenetic analysis of this element of reproductive behavior. We present the basic behavioral and genetic characterization of the new mutant. In order to determine if the gene’s action could be specific, we 1988). 268 S. J. Kulkarni, A. F. Steinlauf and J. C. Hall have carried out a detailed analysis of the mutation’s effects on singing plus other behaviors, in conjunction with mapping the genetic etiology of these phenotypic changes to a narrowly defined interval of the X chromosome, which is separate from the cuc locus (4. KULKARNI and HALL 1987). MATERIALS AND METHODS General: Flies were raised on a cornmeal, agar, molasses, yeast medium, at room temperature (ca. 22-23”) and under natural lighting conditions, unless stated otherwise. Mutagenesis: The preparation of a feeding solution containing N-nitroso-N-ethyl urea (ENU) and the feeding method have been described by VOCEL and NAIARA-JAN (l979a, b) and VOCEL and LUERS (1975), respectively. However, we used an unpublished procedure first developed by M. A. CROSSBY and E. B. LEWIS, later modified by H. LIPSHITZ (personal communications). Briefly, 1 g of the mutagen was dissolved by injecting 100 ml of 0.01 M acetic acid into a sealed 150 ml bottle (Isopac) containing the ENU in powdered form. One ml of this solution was mixed thoroughly with 49 ml of I % (w/v) sucrose in distilled water. This ENU-containing sucrose solution (final concentration of the muta en: 0.2 mg/ml) was added to mutagenesis “chambers” ( Y ’ L2 pmt glass bottles), i x . , enough to dampen a cushion made of Kimwipe paper. Then, 200 Canton3 wild-type males, that had just been starved overnight, were gently transferred into mutagenesis chambers (ca. 100 males in each bottle) and allowed to feed for 24 hr. Since ENU is extremely carcinogenic, all procedures were carried out inside a large container that had a 1-2cm layer of 1 M NaOH as a decontaminant. Each mutagenesis bottle was kept in a polypropylene container, modified with inlet and outlet connectors and a silicon-rubber gasket. The air inside the sealed container was continuously aspirated over a 0.1 M ammonium hydroxide decontaminant solution. Mutagenized males were mass-mated (10 males plus 25 females/bottle) with attached-X, y / virgin females. These females were allowed to lay eggs for 5 days and later discarded. Approximately 2500 lines were established by mating FI males singly to further attached-X, y f virgin females. Isolation of putative song mutants: The courtship song produced by the wing vibrations of one FP male from each line, in the presence of a virgin female, was monitored through earphones and on an oscilloscope as described previously (KULKARNI and HALL 1987; GORCZYCA and HALL 1988). We looked for any departures from the normal song, in terms of shape and amplitude of pulses, numbers of cycledpulse (cf. typical wild-type numbers: 1-3), or the basic characteristics of sine song (cf. SCHILCHEK 197613). If a male from a particular line showed an abnormal song, 5 additional males from it had their songs recorded, to retest the possibly mutant phenotype. Males from such a line were crossed to females homozygous for the Zn(1 )FM7a balancer chromosome to establish a balanced stock and, eventually, a true-breeding subline containing hemizygous mutant males plus homozygous mutant females. From these procedures one new “solid” mutant was isolated (see RESUL.I.S), which was characterized genetically as follows: Recombination mapping: T o determine the approximate location of the new dissonance mutation (symbol: diss) on the X chromosome, males from its balanced stock were crossed to a strain carrying markers yellow (y, O.O), chocolate (cho, 5.5), crossveinless (cv , 13.7), vermilion ( v , 33.0) andforked (f, 56.7). FI progeny were self crossed, and the songs produced by recombinant F2 males in the presence of virgin females were observed on the oscilloscope (see above). Song analysis: Cycles/pulse values were computed as in KULKARNI and HALL (1987) for ca. 150 pulses obtained from several complete pulse trains in a given song. A pulse train is a series of characteristic signa1s;separated from another train by either a sine song bout (courtship hum) or an interval of at least 65 ms of “silence” (e.g. , BENNETCLARK and EWINC 1969; KYRIACOU and HALL 1980). Two pulses is the minimum number in a train. Single pulses separated from other signals by 265 ms were not analyzed. These “blips” account for less than 5% of the total signals in a song record; more often than not, they are caused by nonspecific wing flicks produced by the male or the female when they are not courting (cf. EWINC; 1977a). Each cycles/pulse value was determined as in Figure 2A of KULKARNI and HALL (1987) and plotted with respect to its position (pulse number) in the train (see below, Figure 3). A scatter plot was obtained by determining 150-250 cycles/pulse values from several complete trains (usually 15 to 30 per fly), and a regression of pulse number (within a given train) on cyclesipulse values was obtained with the help of a “Statworks” program run on an Apple MacIntosh computer. We found that ca. 80% of the trains containing 2-5 pulsedtrain did not show any aberrant pulses; and only ca. 10% of the trains contained more than 20 pulses (see below, Figure IA). Therefore, the regression analyses were restricted to only those trains which contained 5-20 pulses. Slopes of the regression lines were diagnostic of the variation in cyclesipulse values for diss songs, as a function of increasing pulse number in the train (see RESLII;I.S). For each genotype, pulse songs from 3-5 flies were analyzed. Mean slope values for the different mutant us. normal genotypes were compared using the statistics noted in Amplitudes of pulses were measured directly from the patches of songs displayed on the oscilloscope (4. KULKAKNI and HALL 1987). Instrument settings were the same for songs recorded from all different genotypes. The overall amplification (cf. Go~r:zuc:.s and HAIL 1988) was such that pulse amplitudes ranged from 1-4.5 V (on the oscilloscope tracings or on oscillographed hard copy records), depending on the genotype. Complementation tests: Homozygous g 2 sd diss f or diss / females were crossed to males carrying a series of translocations (FALK et a1. 1984), in which deletions had been induced in a segment of the X chromosome (13F116A2) translocated to the fourth chromosome (see Table 1 and Figure 8). Thus, the “starting” Dp(l;4)r+f+ covers a region of the X corresponding approxinlately to the sd f , dissflanking interval. All deletions induced in the Dp( I;4)rfff weref+ because the most proximal breaks created by such deletions extended only up to 15B (FALK et al. 1984), thus leaving the f+ containing region (l5Fl-3) intact. For our complementation tests, the presence of these “deleted duplications” in males marked as specified above could therefore be followed. We applied several interstitial deletions, involving regions 14 and 15, to diss mapping (see Table 1 and Figure 8); they had been induced by B. H. JLWI) (unpublished data) and were subsequently characterized (e .g . , with regard to breakpoints) by S. BANC~A and J. B. Bovu (personal communication). The latter two investigators also have made a slight correction of breakpoint determinations for one of FALK et d ’ s chromosome aberrations (see above and legend to Table I ) . KESCII.~I’S. dissonance Courtship Song Mutant 269 A semilethal interaction was observed between the 14B15A deletions and the third chromosomal tra mutation of S.I.LIR.L.EVAN-I (1945). That is, when dissfmales heterozygous for tra and a third chromosomal balancer (In(?LR)TM6B, dominant marker = Dzchaete) were crossed to Of( 1 )l32; trul TM6B; T(1;4)r+ff (Of = 14B17-CI to 14F4-6) or to Df(1 )El50lFM7; tralTM6B; T(1;4)r+f+ (Of = 14B3-4 to 14F) females, very few Dfldissf; traitru pseudomales were obtained (frequency, ca. 0.2%). Although these pseudomales seemed weak in their general movements, which precluded their “overall” behavioral testing (e .g . , mating success, phototaxis; see below), they nevertheless sang robust songs. More generally, all genotypes involving two X chromosomes and homozygosity for tra led to normal courtship songs (in terms of overt appearance of the pulses, cycleslpulse values, and “song-slope” computations, cf. Figure 4, below), with the proviso that diss was not being expressed. Thus, application of this sex-transforming mutation did not interfere with interpretation of our complementation tests. Response to mechanical shock: Mechanical shock-induced paralysis of flies was observed as described by GANE~ZKY and Wu (1982). Briefly, 3-5-day-old flies ( X = 5-10 of each genotype) were placed in empty culture vials and vibrated on a vortex mixer (Scientific Products, model ~8223) at its top speed for 15 s. The length of paralysis was measured as the time required after vortexing until the first f ly regained the ability to stand upright. The test was repeated at least 5 timedgenotype using different sets of flies each time. Measurements of male courtship: Before being tested behaviorally, newly emerged flies were collected under ether anesthesia. The flies were then stored for 5 days at room temperature: males 1 per unyeasted food vial, virgin females 5/vial. Five-day-old flies of each sex were paired singly by introducing them gently into a plastic “mating wheel” with its ten observation chambers (Ho-ITA and BENZER 1976) at ca. 25”. The times elapsing between the moments of pairings and the initiations of copulations were recorded. The courtship response of the test male was quantified as a courtship index (CI; e.g., SIEGEL and H.41.1. 1979; TOMPKINS, HALL and HALL 1980; TOMPKINS et ul. 1982), which is the percentage of an observation period during which the male performs any of the courtship behaviors. CIS were measured for 5 min or until initiation o f copulation, whichever occurred first. The ability of a male to “track’ the female visually during courtship (cf. COOK 1981) was estimated by scoring the number of times the male reoriented towards a female, after breaking away from her for at least 2 s, within a total observation period of 5 min (c f . TOMPKINS et ul. 1982). Phototaxis: Five-day-old males and females were tested separately in a Y-tube apparatus (QUINN, HARRIS and BENZER 1974), which had been adapted for phototaxis tests as in KULKARNI and HALL (1987). From 15 to 30 flies were placed in a dark “start tube” and allowed to choose between the “light” and “dark” arms of the Y-tube during a 120-s test interval. At the end of this time, the numbers of flies distributed in each arm, and those remaining in the start tube, were counted. Each test of a given genotype was repeated with five or more separate groups of flies. The controls (see Figure 13) included flies blinded by norpA mutations (PAK 1979). Electroretinograms (ERG): These ERGs (cf. PAK and GRABOWSKI 1978: HEISENBERC and WOLF 1984) were recorded extracellularly from the adult eyes as follows: all test flies were dark adapted for 2 min. Each f ly was immobilized on ice and quickly anchored to a glass coverslip by securing all of its moving body parts with Elmer’s glue. The cornea of the illuminated eye was penetrated slightly with a recording electrode (glass, ca. 30 megaohms resistance) filled with 0.8% saline. The reference electrode, filled with 3 M KCI, was placed dorsally in the thorax: and light-evoked voltage changes were recorded on a Could 2400 chart recorder. The illumination at the fly’s eye level was 6 foot-candles (f.c.). For each fly, ERGs in response to light pulses of 1, 2, 3, 5, and 10 s durations with 30 s rest in dark before each consecutive stimulus, were obtained. Amplitudes of the “light on” and “light off” transient spikes, plus that of the “maintained component” (PAK and GRABOWSKI 1978), were measured for three flies of each pertinent genotype (see Table 3). Walking optomotor tests: These tests were conducted based on a method previously described by GREENSPAN, FINN and HALL (1980). Individually stored flies, 3-5 days old, were starved for 3-4 hr in empty vials and tested (at room temperature) for their turning behavior in a visual field. For this, an individual was placed under a 25-mm diameter watch glass in the middle of a plexiglass rotating drum (diameter 15 cm; height 22.5 cm), which had alternating black and white vertical stripes. One black-white pair of stripes subtended a 19’ of arc, and the drum was rotated at 12 rpm. White fluorescent light (Sylvania, FC 12T10 CW RS) illuminated the center of the drum during the tests. Behavior was scored by counting the number of times the f ly ran across a quadrant line in the same direction as the rotating stripes, us. the number of times it ran in the opposite direction (4. GREENSPAN, FINN and HMI. 1980). Flies were tested in three successive trials, each consisting of a 1-min clockwise run followed by 1 min of rest, and finally a I-min run in the counterclockwise direction. Results of all three trials, for a given fly, were pooled separately for the clockwise and counterclockwise runs. Three individuals o f each sex were tested for each genotype. The results were expressed as the fraction of total lines which were crossed in the Same direction as that in which the stripes were moving. Locomotor activity: These measurements were made on single flies at c r ~ . 25“ in a cylindrical plastic chamber divided across the diameter by a straight line (4. KULKAKNI and HALL 1987). A given individual was transferred to the chamber; after a 5-min “accommodation” period, the number of times the fly crossed the line in the next 5 min was recorded. Circadian rhythms: Circadian rhythms of locomotor activity were monitored automatically and analyzed as described by HAMBLEN et al. (1986). Flight: A transparent plastic cylinder (BENZER 1973; KCLKARNI and HALL 1987) with its insides coated with paraffin oil was placed upright in a petri dish. From 25 to 50 5-day-old flies were gently introduced into the top of the cylinder. The numbers of flies stuck to the sides along the length of the cylinder, at 2.5-cm intervals, were counted. Each test was replicated with five groups of flies of a given genotype.