Homing test
PNS cohort. We found sex-specific effects revealing that PNS affected females only, improving the ability to reach the nest zone. Interestingly, prenatal NAC prevented this effect (see Supplementary Fig. 2 for details).
mHFD cohort. No significant effects were found upon mHFD or prenatal NAC in both sexes.
Emergence Test
PNS cohort. PNS affected female offspring only, reducing the latency to emerge from the shelter compared to the CTRL group (χ2 = 8.364, p = 0.0038, Fig. 2A). Sex-dependent effects were also found when the time spent in the different zones of the apparatus was evaluated: PNS females reduced the time spent inside the shelter but increased the time in the periphery of the arena compared to CTRL subjects (stress x zones: F(3,126) = 9.585, p < 0.0001; post hoc comparisons: p < 0.01 PNS vs CTRL), see Fig. 2B. Furthermore, PNS females increased speed and distance travelled (stress - mean speed: F(1,42) = 5.675, p = 0.0218; distance: F(1,42) = 5.485, p = 0.0240; Supplementary Table 2). The maternal administration of NAC did not affect the parameters assessed.
mHFD cohort. No difference between groups was observed when assessing the latency to emerge from the shelter. A preference for the peripheral zone was observed for all subjects (zones - males: F(3,93) = 519.339, p < 0.0001; females: F(3,87) = 374.592, p < 0.0001, see Fig. 2A and Supplementary Table 2). Overall, neither the mHFD nor the prenatal NAC affected the behaviors assessed in the Emergence test in both sexes.
Figure 2
Elevated Plus Maze - Epm
PNS cohort. All adolescent mice showed a preference for the open arms (arm - females: F(1,37) = 9.401, p = 0.0040; males: F(1,43) = 3.612, p = 0.0641), a behavior often observed during adolescence. Moreover, PNS females spent more time in the open arms (stress x arm: F(1,37) = 7.867, p = 0.0080; post hoc comparisons: p < 0.05 PNS vs CTRL; Fig. 3A). Exploratory activity was increased by PNS in both sexes (females: F(1,37) = 5.136, p = 0.0294; males: F(1,43) = 9.290, p = 0.0039), while the frequency of risk assessment behaviors was reduced (females: F(1,37) = 4.947, p = 0.0323; males: F(1,43) = 5.015, p = 0.0304; Fig. 3B). In this test, the prenatal NAC did not affect the parameters assessed.
mHFD cohort. Also in this cohort, all adolescent mice preferred the open arms (arm - females: F(1,43) = 12.891, p = 0.0008; males: F(1,49) = 12.848, p = 0.0008). Mirroring the PNS, mHFD increased the time spent in the open arms specifically in females (diet x arm: F(1,43) = 5.827, p = 0.0201; post hoc comparisons: p < 0.05 HFD vs CD; Fig. 3A). However, mHFD decreased exploratory activity in female mice (F(1,43) = 4.611, p = 0.0374), while no changes were observed in risk assessment behaviors (Fig. 3B). In this test, the prenatal NAC did not affect the parameters assessed.
Figure 3
Social Interaction Test
PNS cohort. PNS decreased social behaviors in female subjects, while prenatal NAC administration increased exploratory behaviors both in females and males (see Supplementary Fig. 3 for details).
mHFD cohort. Prenatal NAC was able to boost social behaviors in males exposed to mHFD (see Supplementary Fig. 3).
Forced Swim Test And Coping Stress Strategy
PNS cohort. Dealing with an acute inescapable stress, CTRL male offspring showed a clear preference towards an active coping strategy (stress x strategy: F(1,36) = 5.341, p = 0.0267; post hoc comparisons: *p < 0.05 CTRL-Veh active vs passive). By contrast, PNS males spent an equal amount of time displaying active as well as passive strategies (Fig. 4). No significant differences were observed in female offspring exposed to different prenatal conditions.
mHFD cohort. Exposure to mHFD increased the time spent performing an active coping strategy at the expense of a passive strategy in both sexes (diet x strategy - males: F(1,33) = 3.901, p = 0.0567; females: F(1,31) = 3.997, p = 0.0544; post hoc comparisons: **p < 0.01 HFD-Veh active vs passive, see Fig. 4). No significant changes were observed as a result of prenatal NAC administration.
Figure 4
Hpa Axis Reactivity
PNS cohort. When CORT levels were assessed under basal conditions, we found that PNS exposure reduced basal CORT levels in male offspring only (stress x treatment: F(1,36) = 6.353, p = 0.0163; post hoc comparisons: *p < 0.05 PNS-Veh vs CTRL-Veh), see Fig. 5. Next, the reactivity of the HPA axis was evaluated 30 and 180 minutes following an acute stress (FST). Overall, PNS males reacted to an acute stress with enhanced CORT release after 30 minutes and they still showed higher circulating levels after 180 minutes from the stress, when CORT is expected to return to a baseline (stress x treatment: F(1,36) = 5.585, p = 0.0236; post hoc comparisons: *p < 0.05 PNS-Veh vs CTRL-Veh, see the inset in Fig. 5). Interestingly, prenatal NAC buffered CORT rise in PNS males (*p < 0.05 PNS-NAC vs PNS-Veh, Fig. 5). We did not observe differences in CORT levels of female offspring.
mHFD cohort. Mirroring the PNS cohort, exposure to mHFD led to reduced CORT levels under basal conditions in male mice (diet x treatment: F(1,29) = 15.49, p = 0.0005; post hoc comparisons: *p < 0.05 HFD-Veh vs CD-Veh, Fig. 5). Prenatal NAC was able to restore basal CORT levels (*p < 0.05 PNS-NAC vs PNS-Veh, Fig. 5). As for HPA axis reactivity after acute stress, no changes were observed in both male and female offspring (Fig. 5).
Figure 5
Gene Expression Analysis In The Hippocampus
To identify molecular changes underlying PNS and mHFD, we investigated the hippocampal gene expression of some prototype genes that indicate the functional status of different systems, including the neurotrophins Igf-1 and Bdnf for neuronal plasticity; the transcription factor Nrf-2 and its chaperone Keap-1 for redox balance; the microglia/macrophage markers Cd68, Tmem119 and Trem2; the neuroinflammatory markers iNOS, Arg-1, Ucp-2, and the cytokines Tnf-α, Il-6 and Tgf-β.
PNS cohort. A decrease in total Bdnf mRNA levels in PNS females was found (stress x treatment F(1,40) = 24.36, p < 0.0001); post hoc comparisons: **p < 0.01 PNS-Veh vs CTRL-Veh). Interestingly, prenatal NAC was able to prevent this effect, restoring Bdnf levels (**p < 0.01 PNS-NAC vs PNS-Veh, Fig. 6A). The evaluation of Igf-1 levels revealed an overall decrease in PNS females (stress: F(1,25) = 46.79, p < 0.0001; Fig. 6A) that was not prevented by NAC administration. As for the redox regulations, PNS greatly decreased hippocampal Nrf-2 expression, regardless of sex (stress - females: F(1,42) = 46.05, p < 0.0001; males: F(1,43) = 10.10, p = 0.0027, Fig. 6A). No difference was observed in Keap-1 mRNA levels in female offspring. In general, less pronounced effects were observed in male offspring. Prenatal NAC increased total Bdnf levels in PNS subjects only (stress x treatment F(1,43) = 5.240, p = 0.027; post hoc comparisons: *p < 0.05 PNS-NAC vs PNS-Veh). When evaluating the expression levels of Keap-1, an increase in PNS-NAC group compared to PNS-Vehicle was found in male offspring (stress x treatment: F(1,44) = 6.450, p = 0.0147; post hoc comparisons: *p < 0.05 PNS-NAC vs PNS-Veh, Fig. 6A).
When assessing macrophages/microglia related markers, again we found sex-dependent effects, which were magnified in female offspring. Hippocampal levels of Cd68 and Tmem119 RNAs were decreased by PNS specifically in females (stress - Cd68: F(1,26) = 8.349, p = 0.0077; Tmem119: F(1,24) = 31.46, p < 0.0001, Fig. 6B) with no effect of prenatal NAC.
As for the neuroinflammatory mediators, we found that PNS overall increased iNOS/Arg-1 ratio in females (stress: F (1,25) = 34.44, p < 0.0001) by increasing iNOS and reducing Arg-1 expression (stress – iNOS: F(1,25) = 40.97, p < 0.0001; Arg-1: F(1,25) = 8.211, p = 0.0083); the prenatal NAC increased Arg-1 (treatment: F(1, 25) = 7.472, p = 0.0113) although the iNOS/Arg-1 ratio was not significantly reduced. PNS did not alter Ucp-2 expression in females, while prenatal NAC overall reduced it (treatment: F(1,25) = 12.76, p = 0.0015, Fig. 6C). No changes of iNOS/Arg-1 expression were observed in male offspring, while Ucp-2 was upregulated in the PNS-NAC group (stress x treatment: F(1,27) = 33.75, p < 0.0001, post hoc comparisons: **p < 0.01 PNS-NAC vs PNS-Veh, Fig. 6C).
In addition, the analysis of the cytokines Tnf-α, Il-6, Tgf-β revealed no significant PNS- or NAC- induced changes in either sex, except for an overall increase of Tgf-β in NAC-treated females (see Supplementary Fig. 4).
mHFD cohort. As in the PNS cohort, we found decreased total Bdnf levels as a result of mHFD, specifically in females (diet x treatment (F(1,41) = 34.69, p < 0.0001; post hoc comparisons: **p < 0.01 HFD-Veh vs CTRL-Veh), an effect prevented by prenatal NAC administration (**p < 0.01 HFD-NAC vs HFD-Veh, Fig. 6A). In parallel, mHFD female offspring were characterized by lower levels of Nrf-2 (diet: F(1,47) = 10.99, p = 0.0018), NAC treatment, also in this case, preventing these effects (treatment: F(1,47) = 27.04, p < 0.0001, Fig. 6A). Keap-1 levels were reduced specifically in mHFD female subjects (diet: F(1,45) = 4.673, p = 0.0360). As for male offspring, while no changes were found in total Bdnf, Igf-1, or Keap-1, a general increase of Nrf-2 mRNA levels was observed upon the prenatal exposure to NAC (treatment: F(1,51) = 7.715, p = 0.0076, Fig. 6A).
Concerning macrophage/microglial specific markers we found an increase in Cd68 levels both in males and females exposed to mHFD (diet - males: F(1,20) = 20.27, p = 0.0002); females: F(1,20) = 6.699, p = 0.0176), see Fig. 6B. Prenatal NAC was able to prevent this effects this effect, only in females (treatment: F(1,20) = 6.004, p = 0.0236). No significant changes were observed in mRNA levels of Tmem119 or Trem2.
Furthermore, mHFD while upregulating Arg-1 expression, did not decrease the iNOS/Arg-1 ratio in females (diet: F(1,27) = 10.33, p = 0.0034). mHFD upregulated also Ucp-2 expression in females (diet: F(1,28) = 11.96, p = 0.0018; see Fig. 6C). In male offspring, prenatal NAC reduced iNOS expression (F(1,24) = 7.522, p = 0.0113) as well as iNOS/Arg-1 ratio (F(1,24) = 9.709, p = 0.0047). Similarly to the PNS cohort, an increase of Ucp-2 expression was found in the HFD-NAC male group (stress x treatment: F(1,29) = 12.47, p = 0.0014, post hoc comparisons: **p < 0.01 HFD-NAC vs HFD-Veh, Fig. 6C).
In addition, a reduction of Il-6 by both mHFD and NAC was found in females only (Supplementary Fig. 4).
Figure 6
Iba1-positive Cell Number And Morphology In The Hippocampal Ca1
We estimated the number of microglia in the CA1 region of the hippocampus and analyzed their cellular morphology as well as dorso-ventral distribution.
PNS cohort. There were no differences in the total number of microglial cells in the CA1 between groups. In males, the exposure to PNS combined with NAC treatment reduced the number of ramified microglia cells (“surveilling” phenotype) in the ventral CA1, compared to prenatal NAC alone (F(1,21) = 4.680, p = 0.0422; *p < 0.05 Tukey’s test PNS-NAC vs CTRL-NAC, Supplementary Fig. 5).
mHFD cohort. Also in the mHFD cohort no changes were found in the total number of microglial cells in CA1. Again, in males the mHFD, combined with NAC treatment, reduced the number of ramified “surveilling” microglia compared to prenatal NAC alone (**p < 0.01 Tukey’s test HFD-NAC vs CD-NAC, see Supplementary Fig. 5).