Extended Data Fig. 5: The input-output function of the habenulo-raphe pathway is largely invariant to the fraction of 5-HT neurons innervated by LHb, and this invariance depends on recurrent short-term plasticity. Related to network simulations depicted in Fig. 6.
From: Nonlinear recurrent inhibition through facilitating serotonin release in the raphe
Panels a-e depict results of simulations where short-term plasticity at 5-HT-5-HT connections is intact. a, Input-output relationship of 5-HT neurons during the transient (0–200 ms) phase of LHb input. Colors denote different fractions of 5-HT neurons receiving LHb input (shown in inset; 10-50%). b, Same as a, but during stationary phase of LHb input (200–400 ms). c, Same as a, but during the post-response period of LHb input (0-1 s after). d, Cumulative distribution function of GIRK conductance across simulated 5-HT neurons, for 5 Hz LHb input (red) and 20 Hz LHb input (blue), and for fractions of 5-HT neurons innervated by LHb at 10% (lighter shade) and 50% (darker shade) (all combinations depicted in inset). Note how GIRK conductance depends primarily on LHb input frequency, instead of on LHb input fraction. e, Relationship between LHb input frequency and GIRK conductance (population mean) as LHb input fraction is varied (colors denote distinct input fractions, shown in inset), demonstrating that population GIRK conductance is largely invariant to LHb input fraction. Panels D-H depict results of simulations where short-term plasticity at 5-HT-5-HT connections has been removed (that is linearized responses). f, Input-output relationship of 5-HT neurons during the transient (0–200 ms) phase of LHb input. Colors denote different fractions of 5-HT neurons receiving LHb input (shown in inset; 10-50%). g, Same as f, but during stationary phase of LHb input (200–400 ms). h, Same as f, but during the post-response period of LHb input (0-1 s after). i, Cumulative distribution function of GIRK conductance across simulated 5-HT neurons, for 5 Hz LHb input (red) and 20 Hz LHb input (blue), and for fractions of 5-HT neurons innervated by LHb at 10% (lighter shade) and 50% (darker shade) (all combinations depicted in inset). Note how without short-term plasticity, GIRK conductance depends on both LHb input frequency and LHb input fraction. j, Relationship between LHb input frequency and GIRK conductance (population mean) as LHb input fraction is varied (colors denote distinct input fractions, shown in inset), demonstrating that without short-term plasticity dynamics, the population GIRK conductance is not invariant to LHb input fraction.
