Assessment of mBDNF/proBDNF ratio in the SCN
First, we established whether there was a reduction in mBDNF levels in the SCN of tPA−/− mice. For these experiments, SCN tissue was isolated from tPA+/+ and tPA−/− mice at ZT 4 and ZT 12 and immediately transferred into extraction buffer for protein analysis. SCN content of pro- and mBDNF from tPA−/− and tPA+/+ mice was quantified and normalized to α-tubulin. Then an mBDNF to proBDNF ratio was computed as the index for relative mBDNF quantity. There was a significant main effect of genotype (F1,8 = 7.88, p = 0.023) (Fig. 1), but not ZT (F1,8 = 1.29, p = 0.29) or an interaction (F1,8 = 0.42, p = 0.53), indicating reduced conversion of proBDNF to mBDNF in tPA−/− mice.
Reeintrainment to an advance of the LD cycle
We previously reported that tPA−/− mice took longer to adjust to a 12-h shift of the LD cycle than did tPA+/+ mice [12]. However, these shifts represent only the phase delaying effects of light. To measure the impact of the loss of tPA on phase advances, we measured the time to adjust to a 6-h advance of the LD cycle (Fig. 2). tPA−/− mice took significantly longer (8.1 ± 0.7 days) than tPA+/+ (5.9 ± 0.5 days) to reentrain to the shifted LD cycle (t17 = 2.57, p = 0.02). We also exposed the animals to a 6-h phase delay, but due to suppression of activity by light (masking) an accurate assessment of reentrainment time could not be performed.
Locomotor activity during timed restricted feeding
During baseline measurements both tPA+/+ and tPA−/− mice exhibit typical patterns of nocturnal locomotor activity. Nocturnal activity was divided into two discrete bouts of locomotor activity, a high level of activity in early to mid-night ending in a drop of locomotor activity followed by a brief increase in activity ending gradually at ZT 24. However, the level of activity was reduced in tPA−/− mice during the first part of the dark phase in LD (Fig. 3a and b) from ZT12-17 (F23,575 = 2.63, p < 0.001). Food availability had an effect on both the pattern and level of locomotor activity in tPA+/+ and tPA−/− mice. Food deprivation led to increased diurnal activity across genotypes on both days. When food was removed at ZT 12 locomotor activity was suppressed compared to baseline activity during the first portion of the dark phase. tPA−/− mice had decreased activity compared to tPA+/+ mice on LD fast day one (F1,22 = 4.57, p = 0.044)(Fig. 3c). During fast day two locomotor activity increased significantly over tPA+/+ (Fig. 3d) during both night (ZT 15-18) and day (ZT 5-7, 9, 10) (F23,529 = 2.23, p < 0.001). There was no difference in weight loss between genotypes (Fig. 4a) (tPA−/−: − 20.6% ± .008 and tPA+/+: − 21.7% ± .009, t29 = 0.937, p = 0.399). During restricted feeding the baseline differences in raw locomotor activity between genotypes disappeared (Fig. 5) and there was no difference in nocturnal or food anticipatory activity levels (F23,547 = 1.18, p = 0.253). Following restricted feeding tPA−/− mice gained less weight than tPA+/+ mice, but this difference was not statistically significant (Fig. 4b)(tPA−/−: −.9% ± .01 and tPA+/+: 2.91% ± .02, t28 = − 1.65, p = 0.109).
This experiment was repeated, except that the light cycle utilized was a skeleton photoperiod, which is 15 min of light only at the beginning and end of a 12 h “day”. This design examines whether activity during the day is being suppressed, or “masked”, by the presence of light. The results from this experiment did not substantially differ from those obtained in standard light/dark conditions (Additional file 1).
Assessment of circadian phase during food restriction
The effect of timed restricted feeding on the SCN can be masked by light. When released to constant darkness following restricted or ad libitum feeding there was no genotypic effect on free-running period (F1,23 = 0.60, p = 0.447) (Fig. 6a and b). However, RF treatment had an aftereffect on free-running period, shortening the free-running period of RF groups (tPA−/− RF: 23.77 ± 0.036, tPA+/+ RF: 23.75 ± 0.019) compared to AL groups (tPA−/− AL: 23.84 ± 0.073, tPA+/+ AL: 23.91 ± 0.026) (F1,23 = 8.49, p = 0.008). Activity data were also analyzed to determine if the underlying nocturnal activity rhythm was advanced in food restricted mice, which would not be observable in LD due to the masking effect of light on activity but which could subsequently be predicted by the onsets of activity upon release into DD. There was no evidence of an underlying shift in the phase angle of entrainment toward food presentation in LDsk across genotypes (F1,23 = 2.24, p = 0.148) or treatment (F1,23 = 0.024, p = 0.631) (tPA−/− RF: −.33 ± 0.339, tPA+/+ RF: .24 ± 0.194, tPA−/− AL: −.40 ± 0.188, tPA+/+ AL: −.01 ± 0.203) (Fig. 6c). Additionally, no difference in phase angle of entrainment was seen between genotypes following release from RF in LD to constant conditions (tPA−/− RF = − 0.24 ± 0.154, tPA+/+ RF = 0.15 ± 0.262)(t14 = 1.383, p = 0.1882) (Fig. 6d).
Food intake analysis
Since differences in FAA might reflect differences in the motivation for feeding, we compared food intake in tPA−/− and tPA+/+ mice. There was no difference in food intake (tPA−/−: 4.65 g ± .09 g tPA+/+: 4.79 g ± .07 g) during standard LD conditions with food available ad libitum (t18 = 0.045, p = 0.964). During ad libitum feeding following food deprivation there was no difference in food intake (tPA−/−: 5.76 g ± .19 g tPA+/+: 5.70 g ± .15 g) (t18 = 0.21, p = 0.83). During restricted feeding food intake in tPA−/− mice was reduced compared to tPA+/+ (tPA−/−: 2.14 g ± .06 g, tPA+/+ 2.73 g ± .11 g) (t18 = 4.656, p < 0.001) (Fig. 7a). There are no significant genotypic differences in weight at baseline (tPA−/− = 29.21 g ± .58 g, tPA+/+ = 28.43 g ± .44 g) (t18 = 1.073, p = 0.297 or following RF (tPA−/− = 26.27 g ± .41 g, tPA+/+ = 26.91 g ± .47 g) (t18 = 1.031, p = 0.316) (Fig. 7b). While tPA−/− mice weighed slightly more than tPA+/+ after 48 h of food deprivation, the difference was not statistically significant (tPA−/− = 23.62 ± .51, tPA+/+ = 22.22 ± .44) (t18 = 2.081, p = 0.0519). Changes in weight following food deprivation were due to changes in lean mass regardless of genotype. tPA−/− mouse lean mass change was less than tPA+/+ (tPA−/− = − 3.265 g ± .23 g, tPA+/+ = − 3.7076 g ± .27 g) (t18 = 3.919, p = 0.001). There was no difference in the loss of fat mass between genotypes (tPA−/− = 1.243 g ± .11 g, tPA+/+ = − 1.1872 g ± .20 g) (t18 = − 0.453, p = 0.657). Following RF there was no genotypic difference in lean mass change (tPA−/− = − 2.753 g ± .10 g, tPA+/+ = − 2.714 g ± .45 g)(t18 = 0.223, p = 0.823) but fat mass change differed between genotypes (tPA−/− = −.78 g ± .26 g, tPA+/+ = .3533 g ± .10 g) (t18 = 4.097, p < 0.001) (Fig. 7c).