Fig 1.
On gestation day (GD) 15, 16 and 17, pregnant mice were subcutaneously injected with 25, 25 and 50 μg/Kg of LPS, respectively. Pregnant control mice were injected with phosphate-buffered saline (PBS). The open field test (OFT), light/dark box (LDB), wire hanging (WH) and elevated plus maze (EPM) behavioral tests were conducted when the offspring were 5 weeks old, 6 weeks old, 8 weeks old and 9 weeks old, respectively. Brains were dissected from fetuses on GD18, adolescent mice at 5 weeks old and adult mice at 8 weeks old. mRNA expression, immunohistochemistry and neurotransmitter levels (5-HT [serotonin] and DA [dopamine]) were examined.
Table 1.
Primers used in this study.
Table 2.
Primary antibodies used for the immunohistochemistry analyses in this study.
Fig 2.
Body weight, locomotive activity and muscle strength.
Mouse body weights from the ages of 5 to 8 weeks old were recorded in the PBS-treated controls (triangle, n = 30) and the LPS-induced offspring (circle, n = 30) (A). The locomotive ability and muscle strength of the PBS-treated controls (hollow, n = 30) and LPS-induced offspring (solid, n = 30) were observed at 5 and 8 weeks old, respectively. The total distance (m) of movement during the OFT (B) and the muscle strengths (scores) on the WH test (C) were recorded.
Fig 3.
At 5 weeks old, the time (sec) in the central area of the open-field test (A, OFT, n = 30), the number of times that the mice moved between the light and dark boxes during the light/dark box (LDB) test at 6 weeks old (B, n = 30) and the number of times that the mice moved between the open and closed arms during the EPM test at 9 weeks old (C, n = 30) were recorded in the offspring. Between PBS-induced (hollow) and LPS-induced (solid) offspring, the significant differences (*) at p<0.05 are indicated (n = 30; t-test; two-tailed).
Fig 4.
The brain tissues were dissected from fetuses at GD18 (A, n = 3), adolescent mice at 5 weeks old (B, n = 3) and adult mice at 8 weeks old (C, n = 3). The specific mRNA expression in the cerebrum was detected by real-time PCR. The tbp (TATA sequence binding protein) gene was used as a reference gene and an internal control because tbp expression is stable in the mouse brain across different stages of development and different phases of LPS stimulation [24, 25]. Each gene (independent samples, n = 3) was first normalized to the tbp reference gene using the following equation: ΔCq = Cq target gene−Cqtbp reference gene. The changes in gene expression between PBS-treated controls and LPS-induced groups were calculated as follows: ΔΔCq = (Cq target gene in LPS-induced group−Cqtbp reference gene in LPS-induced group)–(Cqtarget gene in PBS-treated control−Cqtbp reference gene in PBS-treated control), and the relative fold = 0.5ΔΔCq. The significant differences (*) at p<0.05 are indicated (one-way ANOVA; Tukey’s post hoc test).
Fig 5.
Anatomical position and histological examination of NeuN neurons.
As the representative tissue, brains were collected from the 8-week-old offspring. The anatomical position of the cerebral cortex (CTX), CA1 region of the hippocampus (CA1), dentate gyrus of the hippocampus (DG), cerebral aqueduct (AQ), periaqueductal gray (PAG), dorsal raphe nuclei (DR) of the midbrain, and substantia nigra (SN) are shown (A, H&E stain, 20X). In the DR region, the NeuN-expressing neurons are shown (B, IHC; upper left, PBS-treated controls, 100X; upper right, LPS-induced brains, 100X; lower left, PBS-treated controls, 400X; lower right; LPS-induced brains, 400X; NeuN neuron, arrow indicated). The number of NeuN-positive neurons was determined using 10 random microscopic fields (200X) for each brain tissue. The average numbers (cells/mm3) in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (C). The optical densities (units) of NeuN-positive areas were calculated using ImageJ software (open sources, https://imagej.net/Fiji). The average values in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced offspring (n = 6, independent tissues) are shown (D).
Fig 6.
Histological examination of tph2-expressing neurons.
Brain tissues were collected from the 8-week-old offspring. Using immunohistochemistry, the tph2-expressing neurons on the dorsal raphe nuclei (DR) are shown (A, upper left; PBS-treated controls;100X; upper right; LPS-induced brains; 100X; lower left; PBS-treated controls; 400X; lower right; LPS-induced brains; 400X; axon hillock, arrow indicated). The number of tph2-expressing neurons was determined using 10 random microscopic fields (200X) for each brain tissue. The average numbers (cells/mm3) in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (B). Based on the diameter (≥5 μm, 2–4 μm and ≤1 μm) of the axon hillock, the three categories of neurons are shown (C, upper). The average percentage (%) of 5-HTergic axons in each category was determined using 100 neurons and shown (C, lower; n = 6, independent tissues; each category and group). The optic densities (units) of the tph2-positive area were calculated using ImageJ software (open sources, https://imagej.net/Fiji). The average values in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (D). The significant differences (*) at p<0.05 are indicated (t-test).
Fig 7.
Histological examination of th-expressing neurons.
Brain tissues were collected from the 8-week-old offspring. Using immunohistochemistry, representative th-expressing neurons in the dorsal raphe nuclei (DR) (A, left; PBS-treated controls; right; LPS-induced brains; 400X, arrow indicated). The number of th-expressing neurons on DR was determined using 10 random microscopic fields (200X) for each brain tissue. The average numbers (cells/mm3) of th-positive cells in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (B). On substantia nigra (SN), the th-expressing neurons and fibers are shown (C, upper, PBS-treated controls; lower, LPS-induced brains; 400X; th-expressing fibers, arrow indicated). The number of th-expressing neurons on SN was determined using 10 random microscopic fields (200X) for each brain tissue. The average numbers (cells/mm3) of th-expressing neurons in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (D). The optic densities (units) of the th-expressing areas on SN were calculated using ImageJ software (open sources, https://imagej.net/Fiji). The average values in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (E).
Fig 8.
Histological examination of glial cells.
Brains were collected from the 8-week-old offspring. Using immunohistochemistry, representative Iba1-positive microglia in the cerebral cortex are shown (A, left; PBS-treated controls; right; LPS-induced brains; arrow indicated; 400X). The number of Iba1-positive microglia was determined using 10 random microscopic fields (200X) for each brain tissue. The average numbers (cells/mm3) of Iba1-positive microglia in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (B). The optic density (units) of the Iba1-expressing areas was calculated using ImageJ software (open sources, https://imagej.net/Fiji). The average values in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (C). GFAP-positive astrocytes in the surrounding perivascular cuffs of the hippocampus are shown (D, left; PBS-treated controls; right; LPS-induced brains; arrow indicated; 400X). The number of GFAP-expressing astrocytes was determined using 10 random microscopic fields (200X) for each brain tissue. The average numbers (cells/mm3) of GFAP-expressing astrocytes in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (E). The optic densities (units) of the GFAP-expressing areas were calculated using ImageJ software. The average values in the PBS-treated controls (hollow; n = 6, independent tissues) and LPS-induced brains (solid; n = 6, independent tissues) are shown (F).
Fig 9.
Expression of cerebral serotonin.
The serotonin (5-HT) levels in homogenized brains from the offspring were detected. The cerebral 5-HT levels (ng/g) in fetuses at GD18 (A, n = 26, PBS-treated controls; n = 22, LPS-induced offspring), adolescent mice at 5 weeks old (B, n = 23, PBS-treated controls; n = 22, LPS-induced offspring) and adult mice at 8 weeks old (C, n = 23, PBS-treated controls; n = 22, LPS-induced offspring) are shown. The means are indicated by the dotted line. The p-value was determined by two-tailed t-test.
Fig 10.
Expression of cerebral dopamine.
The dopamine (DA) levels in homogenized brains from the offspring were detected. The cerebral DA levels (ng/g) in fetuses at GD18 (A, n = 26, PBS-treated controls; n = 22, LPS-induced offspring), adolescent mice at 5 weeks old (B, n = 23, PBS-treated controls; n = 22, LPS-induced offspring) and adult mice at 8 weeks old (C, n = 20, PBS-treated controls; n = 20, LPS-induced offspring) are shown. The means are indicated by the dotted line. The p value was determined by two-tailed t-test.
Fig 11.
Summary of the effect of maternal immune activation on serotonin and dopamine.
Pregnant mice were subcutaneously injected with lipopolysaccharide (LPS) at GD15 (25 μg/Kg), GD16 (25 μg/Kg) and GD17 (50 μg/Kg). Compared to the PBS-treated controls, the mRNA level of the tph1 gene (an isoform of serotonin synthesizing enzyme expressed on pineal gland) was down-regulated and the DA levels were elevated in the LPS-induced female fetal brains at GD18. In the female offspring, anxiety-like behaviors were observed in the OFT at 5 weeks old, in the LDB test at 6 weeks old, and in the EPM at 9 weeks old. At 5 weeks old, the mRNA level of the tph1 gene was up-regulated, whereas that of the tph2 gene (another serotonin synthesizing enzyme isoform mainly expressed in the raphe nuclei of the midbrain), slc6a3 (a DA transporter), slc6a4 (a 5-HT transporter) and slc6a5 (a glycine transporter) as well as the 5-HT levels were down-regulated in the cerebrum of the female offspring. At 8 weeks old, the tph1 and slc6a5 genes were up-regulated, and the tph2, slc6a3 and slc6a5 genes were down-regulated in the brains of the female offspring. The cerebral 5-HT levels were significantly decreased at 5 and 8 weeks old (solid arrow), whereas the DA levels exhibited a trend towards a decrease (dotted arrow). Moreover, the total number and distribution of tph2-expressing 5-HTergic neurons, and the diameter of the 5-HTergic axon hillocks were decreased at the age of 8 weeks. We concluded that maternal immune activation induced by exposure to a low dose of LPS decreased cerebral 5-HT in parallel with the down-regulation of 5-HT-related genes (tph2 and slc6a4), resulting in anxiety-like behavior in the female offspring.