Figure 4. pC1d/e neurons drive female shoving and chasing, but do not affect receptivity.

(A) pC1-A activation did not affect copulation rates in either the d0 or d3 conditions (n = 40, 38, 40 for control/d0/d3; p=0.79 or 0.29 for control vs d0 or d3; Cox proportional hazards regression, see Materials and methods). (B, C) Shoving (B) and chasing (C) probabilities (control/do/d3: 0.02/0.17/0.20 and 0.018/0.11/0.07 for shoving and chasing) were significantly higher in both the d0 and d3 conditions compared to control (two-sample t-test; *p<0.05). (D–E) same as (A–C), but for pC1-S activation. pC1-S activation did not affect either copulation rate (D) or shoving or chasing (E) probabilities (control/do: 0.02/0.02 and 0.01/0.01 for shoving and chasing). (F) Probability distribution of the fraction of time the female spent shoving (left) or chasing (right) following pC1-A activation (solid line) or pC1-Int activation (dashed line). Arrows indicate the difference in peak shoving (at d3; two-sample t-test, p=0.11) or chasing (at d0; two-sample t-test, *p=4.5*10⁻⁴) probability, between pC1-Int and pC1-A activation. Bonferroni correction was used for multiple comparisons. (G) Fraction of frames with shoving for pairs in which females copulated (R - Receptive) or did not copulate (NR – Not Receptive) for all experimental conditions taken together (d0–d6). Each dot is a single pair, and the bar value is the mean over all pairs (p=0.92 and 0.13 for control and d0-d6). (H) Same as (G), for female chasing (p=0.13 and p<10⁻⁵ for control and d0-d6). (I) Female pC1d/e cells drive persistent shoving and chasing, but do not affect female receptivity. Female receptivity, controlled by a separate pC1 subset labeled in the pC1-Int line, suppresses female chasing, while possibly enhancing female shoving. (J) Shoving and chasing probabilities (control/d0: 0.077/0.015 and 0.039/0.006 for shoving and chasing; p=0.026 and 0.035; n = 20/20 for control/d0) were significantly higher in the d0 condition compared to control following 2 min of pC1-A activation, but not following 30 s (n = 20/21 for control/d0) of pC1-A activation (p=0.52 and p=0.24 for shoving and chasing).

Figure 4—figure supplement 1. Auditory responses in pC1d/e neurons recorded via whole-cell patch clamp.

(A) Left: Experimental setup. A sound tube (similar to Deutsch et al., 2019) was used to deliver sounds during patch clamp experiments. Middle: Image of the pC1-A driver line expressing GFP (see Key Resources Table), used to target pC1d/e cells in patch clamp experiments (arrows point to pC1d/e somas). Right: EM reconstructions of a single pC1d (blue) and a single pC1e cell (red) from FlyWire. (B) Example membrane voltage (Vm) traces of pC1d/e cells responding to pulse trains with different inter-pulse intervals (IPIs). Each trace is a recorded cell in a different individual. The stimulus pattern is shown below Vm traces. (C) Example response (top) to natural song (bottom). (D) Tuning of pC1d/e cells for IPI (left, n = 6 cells), sine frequency (middle, n = 3 cells), and white noise intensity (right, n = 3 cells) - each cell recording comes from a separate fly. The response was calculated as the baseline subtracted Vm integrated over the stimulus duration. Each individual curve corresponds to the trial-averaged responses of a pC1d/e cell from one individual and was normalized to peak at 1.0. IPI tuning is more variable across flies. (E) Comparison of response magnitudes to pulse, sine, and noise stimuli. Responses are normalized to the stimulus that drove the strongest response (n = 3 cells).