Skip to main content
. 2020 Dec 4;9:e54838. doi: 10.7554/eLife.54838

Figure 6. Specific deletion of Neurexin1α in thalamic nuclei does not reproduce choice abnormalities observed in constitutive KO.

(A) Neurexin1α was conditionally inactivated in thalamic progenitor cells by crossing the Neurexin1α-conditional knockout line onto the Olig3-Cre line. (B) Coronal section of Olig3Cre; Ai14 reporter cross showing expression of tdTOM broadly throughout thalamic nuclei. (C) RT-qPCR of RNA from adult mouse thalamus (n = 2 for Nrxn1αfl/fl;Olig3+/+ (gray); n = 3 for Nrxn1αfl/fl;Olig3Cre/+(orange)). Cre-mediated recombination results in reduced expression of Nrxn1α mRNA detected by exon 9 probe (two-sample t-test: p<0.0001) and moderate nonsense-mediated decay (two-sample t-test: p<0.001) (D) Nrxn1αfl/fl;Olig3Cre/+ mutant animals (orange; n = 10) do not exhibit changes in relative reward-stay in comparison with Nrxn1αfl/fl;Olig3+/+(gray; n = 8) control animals. (E–G) Nrxn1αfl/fl;Olig3Cre/+ mutant mice do not have a deficit in updating or representing choice values (two-way RM ANOVA). (H–J) Nrxn1αfl/fl;Olig3Cre/+ mutants exhibit a normal relationship between choice values and decision behavior. (K–M) Nrxn1αfl/fl;Olig3Cre/+ mutants do not exhibit a deficit in the allocation of choices guided by relative choice costs (K, two-way RM ANOVA, left; two-sample t-test, right, p>0.05). Mutants exhibit no difference in task engagement (L, p>0.05) or in choice latencies (M, p>0.05). All data represented as mean ± SEM.

Figure 6—source data 1. Source Data for Figure 6.

Figure 6.

Figure 6—figure supplement 1. Additional behavioral analysis of thalamic neuron Nrx1a deletion.

Figure 6—figure supplement 1.

(A) In-task adaptability. There is no effect for genotype on the adaptability measures of OligCre control and OligCre; Nrxn1αfl/fl mutant mice. (B) Relative initiation latency. We noted no significant differences in the relative latency to initiate trials after distinct reward outcomes. (C) Choice Latency. OligCre; Nrxn1αfl/fl mutant mice (n = 10) exhibit a trend toward extended choice latencies in comparison with Nrxn1αC/C (n = 8) controls across reward environments. (D) Model biases. There is no statistically significant effect for genotype or probability on the degree of bias generated (ΔReward = 12, 10, 6, 4 μL from left to right). (E) Initiation latency during choice paradigm. There is no statistically significant difference in the initiation latencies between genotypes as sessions progress. (F) OligCre; Nrxn1αfl/fl mutant animals exhibit extended choice latencies in comparison with their two wild-type counterparts. (A–C analyzed by three-way RM ANOVA, and D–F analyzed by two-way RM ANOVA).