Preservation of Essential Odor-Guided Behaviors and Odor-Based Reversal Learning after Targeting Adult Brain Serotonin Synthesis

Experimental timeline for all mouse cohorts. Three separate cohorts of Tph2^fl/fl male mice were used. A, Cohort 1 was used solely for immunohistochemistry. B, Cohorts 2 and 3 were used for olfactory go/no-go behavioral experiments and, after behavioral data collection, perfused for subsequent qPCR and HPLC analyses of postmortem brain tissue (see Materials and Methods). Two mice from the behavioral group (cohorts 2 and 3) were euthanized or died during the late stages of task acquisition after displaying signs of illness (indicated by stars).

Targeting of Tph2 in the brains of adult mice. A, Schematic of wild-type Tph2 allele (top), floxed allele with exon V flanked by loxP sites (middle), and targeted allele after Cre-mediated deletion of exon V (bottom). B, Representative images of Tph2^fl/fl mice confirming near-complete loss of 5-HT and Tph2 immunoreactivity within the MRN and DRN of AAV-Cre–treated mice. C, Representative images of Tph2^fl/fl mice illustrating the preservation of 5-HT– and Tph2-immunostained neurons in the medullary raphe, including the raphe magnus (RMg) and raphe pallidus (RPa) in AAV-Cre–treated mice. The raphe magnus does not send many fibers into the forebrain or olfactory structures specifically, but instead largely projects into the spinal cord (Bowker et al., 1981; Skagerberg and Björklund, 1985). D, Representative images of Tph2^fl/fl mice confirming near-complete loss of 5-HT immunoreactivity within both the main olfactory bulb and piriform cortex in AAV-Cre–treated mice. Images are from stacks of 1.2-µm-thick confocal images. Dashed lines in olfactory bulb images represent borders of glomeruli identified with nuclear counterstain (not shown), and in piriform lines, the border of the lateral olfactory track (lot) form layer i of the piriform. Images are from coronal sections, 20 µm thick. Images have been converted to monochrome and inverted. Contrast/brightness adjustments were equally applied to images in B and C to optimally display immunostained cell bodies. Contrast/brightness adjustments were equally applied to images in D to optimally display immunostained fibers.

Go/no-go olfactory task design and stimulus control. A, Outline of the go/no-go task structure as described in detail in Materials and Methods. In the final stage of the task, mice nose-poke in the right port (600-ms hold duration required), receive either CS+ or CS– (400-ms minimum hold [50-ms for self-regulation paradigm]; 2-s maximum hold), and retrieve a reward in the left port (S+ trial) or withhold their response (S– trial). Green shaded circle, presentation of CS+; red shaded circle, presentation of CS–; blue waterdrop icon, brief 3-µL water reward delivery. B, Averaged voltage trace from a photoionization detector (PID) to illustrate the rapid odor stimulus dynamics as controlled by the go/no-go olfactometers. Fifteen trials of the odorant, heptanal (see Materials and Methods for intensity used and flow rate), were delivered by each olfactometer while the PID sampling port was positioned in the center of the odor sampling port. Data are normalized to the maximum value acquired by each olfactometer’s averaged PID output (over the 15 trials) and plotted as the average across all three olfactometers. Time 0 equals odorant valve onset. Time points of 50% (T₅₀) and 90% rise times (T₉₀) are indicated. Gray-shaded area indicates SEM. Although these dynamics may vary slightly across odors, this measure illustrates the precision and stability of the odor presentation methods used.

Adult Tph2-targeted mice learn the olfactory go/no-go task and display similar levels of odor acuity and odor sampling durations. A–D, Learning curves (left) and total number of blocks required to complete each phase (right) across phases 1–4. **p < 0.005. E, Average block percent correct performance during the odor discrimination task when odor sampling time was fixed (left, 2000 trials/mouse) and self-regulated (right, 300 trials/mouse). F, Average sampling durations (nose poking during odor on) for CS+ and CS– odors during phase 4 odor discrimination (blocks ≥85%, 300 trials/mouse, left), and during self-regulation of odor discrimination (blocks ≥85%, 300 trials/mouse, right). G, Odor-removal control experiment to demonstrate the reliance of the mice on the odor stimuli to engage in the go/no-go task. Data are mean ± SEM of all mice/block. Dots: individual mouse data.

Gross motor performance, body weights, water intake, and water motivation of AAV-GFP– and AAV-Cre–treated Tph2^fl/fl mice during the go/no-go task. A, Average duration of withdrawal from odor port to nose poke in the reward port during phase 4 odor discrimination (15 blocks ≥85%, 150 CS+ trials/mouse; Ai) and the self-regulation odor discrimination testing (15 blocks ≥85%, 150 CS+ trials/mouse; Aii). B, Baseline body weights of mice used in olfactory go/no-go testing before water deprivation. C, Mean body weights of all mice (averaged across all days of behavioral testing, range 21–24 days) expressed as percentage of weight (during water restriction) as a function of baseline weight (A). D, The mean of supplemental water for each mouse across all testing days. E, Water motivation test results. Histograms of the average block duration (11–30 blocks/mouse; Ei) and the number of blocks completed in a single 1-h session of the water motivation test (Eii). Data are mean ± SEM of all mice. Dots: individual mouse data. **p < 0.005.

Adult brain 5-HT synthesis is not required for olfactory reversal learning. A, Learning curves during odor-pair reversal, plotted with a three-block average sliding window until each mouse reached or surpassed 85% correct responses. B, Average number of sliding blocks to reach ≥85% correct. Circles: values for individual mice.