Both epidemiological and animal experimental studies have demonstrated that during early pregnancy, maternal immune response mediated by pro-inflammatory cytokines is associated with higher risk for neuropsychiatric disorders in the offspring [31, 32]. It was reported that the NF-κB inhibitor, which blocked the NF-κB signaling pathway and reduced cytokine release, was effective in many related diseases [33–35]. In this study, PolyI:C was administered to rats in early pregnancy to stimulate the release of pro-inflammatory cytokines. Maternal cytokine levels and the behavior of adult offspring were measured to explore the role of the cytokine-mediated immune response during pregnancy in the development of psychiatric disorders. We also examined the effects of intervention with an NF-κB inhibitor, PDTC.
In this study, serum levels of IL-10 and TNF-α in the model group (pregnant rats treated with PolyI:C) increased significantly compared to levels in the control group. This imitated inflammatory reactions mediated by cytokines in maternal hosts after infection. Gayle et al.  reported that PolyI:C, as well as LPS, could increase the cytokine level in the amniotic fluid and the placenta of the maternal host. Increased cytokines could enter the circulatory system of the fetus.
In our previous study, schizophrenic patients showed activation of NF-κB and elevated levels of cytokines . In the present study, activated NF-κB was below the detection limit in our assays. In the intervention group, NF-κB activation was inhibited through injection of PDTC and serum levels of IL-10 and TNF-α were suppressed relative to the model group. These results provide indirect evidence that NF-κB activation was successfully reduced in the intervention group. All the above indicated that PDTC, a kind of NF-κB inhibitor, can interfere with the inflammatory reactions mediated by cytokines.
Behavioral deficits occurred in offspring from mother rats that had an immune response induced by PolyI:C treatment. The offspring of the model group showed weakened PPI and weakened latent inhibition. These findings are consistent with Meyer , who reported that administration of PolyI:C to pregnant mice led to a loss of PPI, loss of latent inhibition, and multiple schizophrenia-like neuropathologic manifestations in the offspring. Behavioral abnormalities were less severe in offspring from the intervention group compared to offspring from the model group, demonstrating that inhibition of NF-κB during pregnancy reduced neurodevelopmental disorders in the offspring.
Latent inhibition exists in all classical and instrumental conditioned reflexes, such as passive and active avoidance. Baruch et al.  first reported latent inhibition loss in schizophrenia patients, finding that acute schizophrenic patients lost latent inhibition, while chronic patients treated with antipsychotics presented with normal latent inhibition. Several clinical studies [38–40] further supported this result. Salgado  reported that the dopamine antagonist amphetamine could cause loss of latent inhibition in normal healthy people, and conversely, antipsychotics could enhance latent inhibition. Similar results occurred in animals . In addition, individuals from schizophrenic parents showed abnormal latent inhibition. These studies showed that abnormal latent inhibition in patients with schizophrenia could be regarded as a stable manifestation and a cognitive deficit in behavior. In active avoidance tests, the total conditioned reflex time in offspring from the control group was significantly higher than in the offspring from model group (PolyI:C-treated), implying that latent inhibition abnormality, impaired learning, and impaired memory occurred in offspring from model group. In passive avoidance tests, the T1 and T2 in model group offspring were also significantly different from those in control group offspring, indicating that memory was impaired. These results are consistent with previous studies [13–15]. In the active avoidance test, a significant effect of the treatment was found. However, for total conditioned response times, the performance of offspring from the intervention group and model group were not significantly different, demonstrating that NF-κB inhibition did not improve all behavioral outcomes in offspring from the intervention group.
The neurodevelopmental hypothesis of schizophrenia posits a correlation between the disease and neurodevelopmental disorders. It has been suggested that the maternal immune response to viral infections in pregnancy may interfere with normal fetal brain development. Motivated by this hypothesis, researchers have created many animal models to study the effects of prenatal and perinatal environments on schizophrenia. Meyer  et al. reported that the PolyI:C treatment model in rats shared a wide range of characteristics with humans, and PolyI:C treatment effects manifested in post-pubescent offspring were consistent with the neural development hypothesis. We demonstrated that prenatal treatment with PolyI:C could elevate maternal cytokines and cause reduced PPI and reduced latent inhibition in adult offspring, confirming Meyer's results. Current studies concentrate on cytokines as a neurodevelopmental disorder trigger in maternal hosts after infection. Pro-inflammatory cytokines released by the maternal immune system may disrupt fetal brain development. Transfer of maternal cytokines to fetuses is not the only means of elevating cytokine levels in fetal brains ; the response of fetal immune systems to increased maternal cytokines might be an alternate mechanism . The influence of enhanced anti-inflammatory cytokine signaling on early brain development should be also emphasized . Disruption of the balance between pro- and anti-inflammatory cytokine signaling in fetal brains may be a key mechanism precipitating schizophrenia-related pathology following prenatal maternal infection . In our study, the NF-κB inhibitor in the intervention group evidently suppressed cytokine release induced by PolyI:C and improved behavioral outcomes in adult offspring. The results suggest cytokines play an important role in neurodevelopmental disorders in this model, and provide evidence for the correlation between increased cytokines in maternal hosts and abnormal behavior of offspring. Activated NF-κB was not detected in this study. Thus, there is no direct evidence of a relationship between NF-κB activity and abnormal behavior of offspring. The exact mechanism needs further study.