Phasic Activation of Dorsal Raphe Serotonergic Neurons Increases Pupil Size

Abstract

Transient variations in pupil size (PS) under constant luminance are coupled to rapid changes in arousal state,^1-3 which have been interpreted as vigilance,⁴ salience,⁵ or a surprise signal.^6-8 Neural control of such fluctuations presumably involves multiple brain regions^5,9-11 and neuromodulatory systems,^3,12,13 but it is often associated with phasic activity of the noradrenergic system.^9,12,14,15 Serotonin (5-HT), a neuromodulator also implicated in aspects of arousal¹⁶ such as sleep-wake transitions,¹⁷ motivational state regulation,¹⁸ and signaling of unexpected events,¹⁹ seems to affect PS,^20-24 but these effects have not been investigated in detail. Here we show that phasic 5-HT neuron stimulation causes transient PS changes. We used optogenetic activation of 5-HT neurons in the dorsal raphe nucleus (DRN) of head-fixed mice performing a foraging task. 5-HT-driven modulations of PS were maintained throughout the photostimulation period and sustained for a few seconds after the end of stimulation. We found no evidence that the increase in PS with activation of 5-HT neurons resulted from interactions of photostimulation with behavioral variables, such as locomotion or licking. Furthermore, we observed that the effect of 5-HT on PS depended on the level of environmental uncertainty, consistent with the idea that 5-HT could report a surprise signal.¹⁹ These results advance our understanding of the neuromodulatory control of PS, revealing a tight relationship between phasic activation of 5-HT neurons and changes in PS.

Keywords: arousal; behavioral states; dorsal raphe; foraging; mouse; neuromodulation; optogenetics; pupil; serotonin; uncertainty.

Graphical abstract

Figure 1

Tracking PS during a Foraging Task for Head-Fixed Mice (A) A mouse placed on a treadmill exploits (i.e., licked for a water reward) one of two resource sites materialized by movable arms. To explore a new site, the mouse switched sides by running on the treadmill, during which time the site in front moved away, and the distal one moved into place. Licks and the pupil were tracked by two different cameras, and locomotion was monitored by the rotary encoder of the treadmill. (B) The task consisted of periods of locomotion, during which the mice switched sites, and stillness, during which the mice licked. A behavioral bout was defined as the time spent at a given site. (C) Example PS (major axis in gray), treadmill activity (pink), and lick time (green) from one experimental session. (D and E) Locomotion speed and pupil response aligned to locomotion onset and offset. (F) Lick-triggered average pupil response. (D–F) Mean ± SEM (across sessions, n = 96 sessions from 9 mice). See also Figure S1.

Figure 2

Optogenetic Activation of DRN 5-HT Neurons Increases PS (A) Sagittal view of a mouse brain, illustrating the optogenetics approach. DRN neurons are infected with AAV2/9-Dio-ChR2-eYFP. In transgenic SERT-Cre mice (n = 6), 5-HT neurons (red) express ChR2-YFP (green) and can be photoactivated with blue light (473 nm) delivered through an optical fiber implant. The same approach is used for WT littermates (n = 3), which do not express ChR2-YFP. (B) Fluorescence image of a parasagittal section, showing part of the DRN, aqueduct (Aq.), and cerebellum (Cb.). 5-HT neurons are labeled in red (rabbit anti-5HT), and ChR2-eYFP expression (green) is localized to the DRN (DAPI, blue). See also Figure S2. (C) Examples of infected 5-HT neurons, circled in white. (D) Photostimulation (10-ms pulses, 25 s⁻¹ at 5 mW for 30% of bouts) starts at bout onset (i.e., offset of locomotion) and ends at running initiation or after 5 s of stillness. See also Figure S3. (E) Pupil responses of non-stimulated bouts aligned to bout onset (n = 96 sessions). (F) Time course of pupil responses aligned to the first lick of stimulated (blue) and non-stimulated (black) bouts lasting at least 4 s for two example mice (one SERT-Cre mouse and one WT mouse). The insets show the difference between pupil responses in stimulated and non-stimulated bouts (i.e., the difference between the blue and black traces [Δ PS]). (G) Left: summary across mice of Δ PS. Right: median ± MAD across mice of Δ PS, estimated in a 0.5-s window (yellow). See also Figure S3. (H) Time course of pupil responses aligned with stimulation offset-stimulated (blue) and -non-stimulated (black) bouts lasting at least 7 s for the same animals as in (F). Because the duration of stimulation was variable, aligning with the offset of stimulation was necessary to visualize the lasting effect of stimulation. The insets show Δ PS. (I) Left: summary across mice of Δ PS. Right: median ± MAD across mice of Δ PS, estimated in the first 2 s after stimulation (yellow).

Figure 3

The Effects of DRN 5-HT Photostimulation Are Not Specific to Behavioral States (A) Schematic of the linear mixed model. Four predictors are weighted and summed linearly to predict PS at the end of the stimulation epoch. For each session, the method finds the set of coefficients (β weights) applied to each predictor that best explains the PS at the end of the stimulation epoch. (B) Linear mixed model weights for SERT-Cre and WT animals. Bars show fixed effects estimates, and error bars represent 95% confidence intervals. See also Figure S4. Gray dots indicate the estimated weights for each session.

Figure 4

The Effects of DRN 5-HT Photostimulation Depend on the Level of Uncertainty (A–C) BPS, locomotor speed, and number of licks across protocols (mean ± SEM across sessions; low uncertainty, n = 29; medium uncertainty, n = 24; high uncertainty, n = 18). (D) Average change in PS of SERT-Cre mice between stimulated and non-stimulated bouts lasting at least 4 s, aligned to the first lick, and summary statistic (i.e., mean ± SEM across sessions of Δ PS estimated in a 0.5-s window, yellow). Colors represent protocols with different levels of uncertainty (low: P_REW = 90%, P_DPL = 30%; medium: P_REW = 60%, P_DPL = 20%; high: P_REW = 45%, P_DPL = 15%; mean ± SEM across sessions). See also Figure S4. (E) Average change in PS of SERT-Cre mice between stimulated and non-stimulated bouts lasting at least 7 s, aligned to the end of photostimulation, and summary statistics (i.e., mean ± SEM across sessions of Δ PS estimated in a 2-s window, yellow). (D and E) The blue trace represents stimulation duration. (F) Linear mixed model weights for SERT-Cre mice run independently for the three different levels of uncertainty. Bars show fixed effects estimates, and error bars represent 95% confidence intervals. Gray dots indicate the estimated weights for each session.