*2.2. AG Improved Short-Term Working Memory*

The effect of AG on short-term working memory was measured using the Y-maze test (Figure 2). TBI did not have a significant effect on the alternation behaviors. However, AG increased the alternation behavior in both TBI and normal mice (Figure 2A). Similarly, CMS did not affect the alternation behavior in CMS mice. In contrast, AG treatment increased the alternation behavior (Figure 2B)

**Figure 2.** Short-term working memory by AG in TBI and CMS mice, measured by the percent alternation in the Y-maze. (**A**) TBI model. There were before–after (within) effects [F(1,28) = 3.57, *p* = 0.012], treatment group (between) effects [F(3,28) = 3.23, *p* = 0.081], and within–between interactions [F(3,28) = 3.28, *p* = 0.036]. There were no before–after changes in the control (*p* = 0.6) and TBI (*p* = 0.38), but improvements in the AG (*p* = 0.008) and TBI + AG (*p* = 0.047). (**B**) CMS model. There were before–after effects [F(1,36) = 6.2, *p* = 0.018] and treatment group effects [F(3,36) = 11.8, *p* < 0.001], but no significant within–between interaction differences [F(3,36) = 1.5, *p* = 0.24]. There was no before–after

change in the control group (*p* = 031), AG (*p* = 0.97) or CMS (*p* = 0.33) groups. However, there was an increase of % alternation in the CMS + AG group (*p* = 0.006). All data were normally distributed and are represented as means ± S.E.M. Control: vehicle (DW) treated; AG: *Angelica gigas* 1 mg/kg; TBI: vehicle treated + traumatic brain injury; TBI+AG: *Angelica gigas* 1 mg/kg + traumatic brain injury; CMS: vehicle treated + chronic mild stress; CMS+AG: *Angelica gigas* 1 mg/kg + chronic mild stress. Repeated measures ANOVA, Tukey's HSD post-hoc test.
