Serotonin-dependent kinetics of feeding bursts underlie a graded response to food availability in C. elegans

Abstract

Animals integrate physiological and environmental signals to modulate their food uptake. The nematode C. elegans, whose food uptake consists of pumping bacteria from the environment into the gut, provides excellent opportunities for discovering principles of conserved regulatory mechanisms. Here we show that worms implement a graded feeding response to the concentration of environmental bacteria by modulating a commitment to bursts of fast pumping. Using long-term, high-resolution, longitudinal recordings of feeding dynamics under defined conditions, we find that the frequency and duration of pumping bursts increase and the duration of long pauses diminishes in environments richer in bacteria. The bioamine serotonin is required for food-dependent induction of bursts as well as for maintaining their high rate of pumping through two distinct mechanisms. We identify the differential roles of distinct families of serotonin receptors in this process and propose that regulation of bursts is a conserved mechanism of behaviour and motor control.

Figure 1. Automated detection of pumping dynamics.

(a) Bright-field image of a worm confined in a custom microfluidic device for longitudinal recording of pumping dynamics. (b) Images of a small region surrounding the pharyngeal grinder (red box in a) at two consecutive frames and the difference between the two. Pixels that turn darker and pixels that become brighter are coloured in red and blue, respectively. (c) Representative time courses of the total intensity of the red- and blue-marked pixels. Peaks in these measures are robust indicators of a pharyngeal motion, either contraction or relaxation. (d) Representative time courses of the displacement between the centrr of mass of the red- and blue-marked pixels. Positive values occur when the blue centre of mass is posterior to the red one, indicating a relaxation motion.

Figure 2. Pumping dynamics are modulated in response to food availability.

(a) Average pumping rate as a function of food concentration. The rates are averaged over all worms and all recorded periods at each concentration. Error bars are 95% confidence intervals, calculated from an ensemble of all 10-min subsets of the data. Sample sizes are given in Supplementary Table 2. (b) The distributions of pump duration Δt_PD. (c) Representative time series of pumping by individual worms at different food concentrations. Plotted is the total numbers of pumps counted since the beginning of the experiment (defined as time=0). (d) Histogram of pulse intervals, Δt_PI. These histograms show a narrow peak and a heavy tail. One of these peaks was fit to a Gaussian distribution (black dashed line) centred at 174 ms, with the s.d. of 20 ms (adjusted R²=0.985).

Figure 3. Bursts of fast pumping are a primary response to food availability.

(a) F(r), the fraction of time spent in intervals longer than 1/r at various food concentrations. Dashed lines for (OD₆₀₀=0.5, 1 and 8) are the best-fit lines between 100 pumps per min and 250 pumps per min. Sample sizes are given in Supplementary Table 2. (b) Survival probability of the durations of fast pumping bursts at OD₆₀₀=1 (black). Red curve is the burst durations calculated from the same pumping intervals in randomized order. (c,d) The average fraction of time spent by worms in long pauses (c) and fast pumping (d) at different food concentrations. Empty symbols, no added serotonin; filled symbols, in media supplemented with 5 mM serotonin. Error bars as in Fig. 2. (e) Average pumping rates plotted against the fraction of fast pumping at different food concentrations. Red line is a linear fit to the data.

Figure 4. Serotonin is essential for induction of fast pumping.

(a) The average fraction of time spent in fast pumping by tph-1 worms at various food concentrations. Empty symbols, no added serotonin; filled symbols, in media supplemented with 5 mM serotonin. Sample sizes are given in Supplementary Table 2. (b) F(r), the fraction of time spent in intervals longer than 1/r, in various food concentrations with no serotonin added to the media (red, green, blue) and in 5 mM serotonin in the absence of bacteria (orange). Dashed lines are the best-fit lines between 100 pumps per min and 250 pumps per min. (c) The average fraction of time spent in fast pumping by worms in which serotonin production or secretion has been perturbed in all or some of the serotonergic neurons (blue and grey) and in worms in which serotonergic neurons had been ablated by reconstituted caspase-3 (red). Food was supplied at OD₆₀₀=3. Error bars as in Fig. 2.

Figure 5. Distinct roles of serotonin receptors in food-driven pumping.

(a) Average pumping rates of wild-type (N2) worms, tph-1 mutants and a range of serotonin receptor mutants at different food concentration. (b) The average fraction of time spent by these worms in fast pumping at OD₆₀₀=3. Error bars as in Fig. 2, sample sizes in Supplementary Table 2. Here the threshold rate for fast pumping was taken to be 200 pumps per min for ser-4(ok512) worms and 250 pumps per min otherwise (based on c and on Supplementary Fig. 5A). The statistical significance of the difference from wild type is indicated (*P<0.05; **P<0.01; ****P<0.0001; NS, P>0.05). (c) F(r), the fraction of time spent in intervals longer than 1/r, for wild-type worms and ser-1(ok345), ser-4(ok512) and ser-4(opy024) mutants at OD₆₀₀=3, with the best fit lines (dashed) between 100 pumps per min and 250 pumps per min. The dotted vertical lines indicate the last point of the linear segment of F(r), taken as the threshold rate for fast pumping, demonstrating its slowdown in ser-4(ok512) mutants.

Figure 6. Distinct roles of serotonin receptors in pumping stimulated by exogenous serotonin.

(a) Average pumping rate of wild-type (N2) worms, tph-1 mutants and a range of serotonin receptor mutants at different concentrations of serotonin (5-HT) in bacteria-free media. Sample sizes are given in Supplementary Table 2. (b,c) The average fraction of time spent in long pauses (b) and fast pumping (c) by wild-type (N2) worms as a function of the concentration of serotonin in bacteria-free media. (d,e) The average fraction of time spent in long pauses (d) and fast pumping (e) by wild-type (N2) worms, tph-1 mutants and a range of serotonin receptor mutants in bacteria-free media supplemented with 5 mM serotonin. Here the threshold rate for fast pumping was taken to be 160 pumps per min for ser-4 worms and 200 pumps per min otherwise (based on Supplementary Fig. 6A,F). Error bars as in Fig. 2. Statistical significance indicated as in Fig. 5. (f) Summary of the distinct roles of serotonin receptors. In food-driven pumping, SER-1 is involved in promoting pumping bursts, presumably with contributions from SER-4. The stimulation of pumping by exogenous serotonin requires the SER-7 and SER-4 receptors.