High level estradiol impairs and low level estradiol facilitates non-spatial working memory.

The present study investigated the effect of different doses of estradiol treatment on performance in the non-spatial delayed alternation T-maze a task in which performance is mediated by the integrity of the prefrontal cortex (PFC). Ovariectomized (OVX) female rats were injected with estradiol benzoate (0.3 microg/0.1 ml sesame oil (EB0.3), 5 microg/0.1 ml sesame oil (EB5) or 10 microg/0.1 ml sesame oil (EB 10)) or vehicle (sesame oil, 0.1 ml). Approximately 2 h after each injection, animals were trained daily on the T-maze with an initial delay of 10 s (short delay). Following a month with no treatment animals were re-trained at a 40 s delay (long delay). Days to reach criterion (one error per day for three consecutive days), mean total errors, errors across days, change in performance across training (short subtracted from long delay), and latency to reach goal arm, were scored. At the short delay, there was a weak effect for the low dose of estradiol (EB0.3 low-to-medium physiological) to significantly decrease the number of working memory errors compared to controls. However at the longer delay the higher doses of estradiol EB5 (high physiological) and to a lesser extent EB10 (suupraphysiological) significantly increased the number of working memory errors compared to controls. These data demonstrate the differential effect of estradiol during short and long delays on prefrontal cortex dependent working memory. High levels of estradiol impair PFC-dependent working memory at longer delays, while low level estradiol weakly facilitates PFC-dependent working memory at a shorter delay. These data suggest that estradiol's facilitatory effects on working memory may not be mediated through the PFC, while estradiols inhibitory effects on working memory may be mediated at least in part through the PFC.