The impact of response latency on working memory assessment in mice

In rodent tasks of working memory, increased response latencies are often accompanied by decreased task accuracy. As response latency can be altered by interventions such as drugs, interpretation for the direct effect of these interventions on working memory may be clouded. Here we investigated the relationships between working memory and response latency in a rodent task of working memory, attempting to isolate concurrent effects of interventions on these outcomes.

C57BL/J mice (male: n = 18; female; n = 10) were trained on the Trial Unique Non-Match to Location task of working memory. Mice received saline, 0.1 mg/kg, or 0.3 mg/kg MK-801. Some mice were then connected to an electrophysiological tether before testing. Response latencies were extracted for correct and incorrect trials from all sessions. Binomial logistic regression assessed latency and intervention as independent predictors of performance.

MK-801 dose-dependently affected task accuracy in both sexes (p < 0.0001), but whereas MK-801 increased latency in males, this was decreased in females (p < 0.0001 for both). Drug (p < 0.001) and response latency (p < 0.001) were significant independent predictors of task accuracy in both sexes. In males treated with MK-801, a high proportion of responses >10 s were observed, but this was not present in treated females. Head tethering also increased response latency (p < 0.0001), but did not impact task accuracy (p = 0.991).

MK-801 impairs working memory and alters response latency, effects which are independent from each other, which isolates a true effect of this drug on working memory performance. We recommend similar regression analyses are conducted in future studies to rule out confounding effects of response latencies on task performance.