Proper direction of male genitalia is prerequisite for copulation in Drosophila, implying cooperative evolution between genitalia rotation and mating behavior

Abstract

Animal morphology and behavior often appear to evolve cooperatively. However, it is difficult to assess how strictly these two traits depend on each other. The genitalia morphologies and courtship behaviors in insects, which vary widely, may be a good model for addressing this issue. In Diptera, phylogenetic analyses of mating positions suggested that the male-above position evolved from an end-to-end one. However, with this change in mating position, the dorsoventral direction of the male genitalia became upside down with respect to that of the female genitalia. It was proposed that to compensate for this incompatibility, the male genitalia rotated an additional 180° during evolution, implying evolutionary cooperativity between the mating position and genitalia direction. According to this scenario, the proper direction of male genitalia is critical for successful mating. Here, we tested this hypothesis using a Drosophila Myosin31DF (Myo31DF) mutant, in which the rotation of the male genitalia terminates prematurely, resulting in various deviations in genitalia direction. We found that the proper dorsoventral direction of the male genitalia was a prerequisite for successful copulation, but it did not affect the other courtship behaviors. Therefore, our results suggested that the male genitalia rotation and mating position evolved cooperatively in Drosophila.

Figure 1

Evolutionary relationship between mating position and male genitalia rotation in dipterans. (a) Phylogenetic tree showing the evolution of mating positions in Diptera. Illustrations show the mating of mosquitos (upper, end-to-end position) and flies (bottom, male-above position). (b) Hypothesis for the cooperative evolution between mating position and male genitalia rotation. Red-blue triangles show male genitalia, and black-grey triangles represent female genitalia. If the male genitalia rotate 180°, their dorsoventral direction becomes upside down. Under this condition, the dorsoventral directions of the male and female genitalia are in accordance in the end-to-end position of mosquitoes (left). However, in the male-above position, this relative dorsoventral direction of the male genitalia with respect to those of the female would be upside down. It was previously proposed that this inconsistency in the dorsoventral direction of the female and male genitalia could be overcome by an additional 180° rotation of the male genitalia in flies (right). This hypothesis implies that cooperative evolution occurs between the mating position and the rotation of male genitalia.

Figure 2

Classification of male genitalia with angle deviation. (a and b) Definition of the angle deviation in male genitalia. We measured the angle (shown in red) between the midline of the abdomen (solid black line) and the line connecting the anus and penis (broken magenta line). A photograph of male genitalia with the angle deviation (a) and its schematic representation (b) are shown. cl, clasper; ap, anal plate. (c) Eight classes of angle deviation in male genitalia: 0° (Left 22°-Right 22°), Right 45° (Right 22°-Right 67°), Right 90° (Right 67°-Right 112°), Right 135° (Right 112°-Right 157°), 180° (Right 157°- Left 157°), Left 135° (Left 157°-Left 112°), Left 90° (Left 112°-Left 67°), Left 45° (Left 67°-Left 22°). (d–g) Male genitalia of wild-type (d) and of Myo31DF^K2 mutant (Myo31DF) flies with 0° (e), Right 45° (f), and 180° (g) angle deviation. The head of the magenta arrow shows the penis, and the tail indicates the anus in (d)–(g).

Figure 3

Males with genitalia exhibiting less angle deviation had an advantage in reproduction. (a,b) Reproduction success rates (%) after mating for 1, 2, 3, and 4 days. A single wild-type male (a) or Myo31DF^K2 male with genitalia in the normal dorsoventral direction (b) was mated with three wild-type virgin females. (c and d) Reproduction success rates (%) for males with 8 classes of genitalia angle deviation, after mating for 4 days. Virgin females used for the mating were wild type (c) or Myo31DF^K2 homozygotes (d).

Figure 4

Analysis of courtship behavior using a novel image-processing program. (a) A snapshot of the video data of mating behavior between a virgin wild-type female and a male with genitalia with angle deviation or a control male with genitalia in the normal dorsoventral direction. (b and c) Snapshot of the original video (b) and its corresponding binarized picture (c) from the video recording of mating behavior in a single chamber. (d) Processed image in which the binarized images were automatically arranged in 3D space along a time line. A moving bar (yellow line) that was synchronized with the 2D image in b, c was inserted into the 3D image. Black triangles show time points.

Figure 5

Males with genitalia exhibiting less angle deviation had more successful copulation. (a) Copulation success rates (%) of males with genitalia with angle deviation classified as 0° (0), Right 45° (45), and 180° (180) and of control males (control). (b) Reproduction success rates (%) of males with genitalia with angle deviation classified as 0° (0), Right 45° (45), and 180° (180) and of control males (control). (c) Courtship latency (seconds) of males with genitalia with angle deviation classified as 0° (0), Right 45° (45), and 180° (180) and of control males (control), shown as boxplots. In each boxplot, the middle band is the median (50^th percentile), and the length of the box shows the 1^st and 3^rd quartile (25^th and 75^th percentile). (d) CI of males with genitalia with angle deviation classified as 0° (0), Right 45° (45), and 180° (180) and of control males (control). The mean and standard error (error bars) are shown. In a-d, all the females were wild type, and the number of examined males is shown in parentheses as n. ***P-value < 0.01. N.S., not significant statistically.