Here we used a mouse model to examine the influence of a maternal GI nematode infection on hippocampal gene expression, LTP and the neuroimmune system, as well as resistance to direct infection of three week-old juvenile offspring. We observed four key consequences of maternal infection: 1) earlier onset of the capacity to induce hippocampal LTP, with changes in gene expression suggesting this may be due to accelerated maturation of glutamatergic synapses; 2) higher levels of hippocampal gene expression in neurogenesis and gliogenesis pathways, and higher levels of gene expression associated with oligodendrocytes and myelination; 3) greater resistance to H. bakeri infection, evidenced by lower worm burden and parasite fecundity, consistent with maternal transfer of parasite-specific IgG1 to the serum of uninfected offspring; and 4) an increase in the number of hippocampal microglia and astrocytes, a higher percentage of CD206 positive microglia, and increased expression of the TGF- β signaling pathway involved in immune regulation. Together, the data suggest immune stimuli from the H. bakeri infected mother are transferred to the offspring, extend to their brain, and result in a Th2/Treg-biased neuroimmune response that underlies accelerated hippocampal maturation as evidenced by enhanced LTP, upregulation of neurogenesis, gliogenesis, and myelination genes, and enhanced spatial memory [10].
Maternal H. bakeri infection enhanced or accelerated the developmental capacity for LTP induction in offspring, which may be explained by the observed changes in gene expression. As the capacity for hippocampal LTP is believed to strongly correlate with successful spatial memory formation [18], this is consistent with the enhanced spatial memory we previously detected in offspring of H. bakeri infected dams[10]. To our knowledge, this has not been described in other maternal infection models, although similarly to our observations here, exposure to exercise during early life enhances LTP and spatial memory[36]. In contrast, maternal exposure to Escherichia coli lipopolysaccharides[5, 37] or stress[38], both of which are known to induce neuroinflammation in offspring[3, 39], impairs hippocampal LTP and spatial memory in offspring.
Consistent with altered LTP, we also found increased expression of genes associated with glutamatergic synapses and the MAPK signalling pathways associated with LTP, in addition to those associated with AMPAR and NMDAR subunits, CAMKII and PKC, each of which may provide key contributions to LTP induction and maintenance. AMPARs are composed of GluR1–4 subunits[40], while NMDARs contain an obligate NR1 and two or more NR2 subunits (NR2A-D) [41, 42]. The ratio of NR2A-D subunits in hippocampal NMDARs changes with age, with an early predominance of NR2B-containing NMDARs eventually maturing to NR2A-containing receptors[41, 42]. GluR1 AMPAR subunits are critical for LTP and memory retention[43] and exposure of pregnant mice to a bacterial mimic decreased GluR1 levels in the hippocampi of offspring and impaired their spatial memory[44], whereas increased levels of hippocampal GluR1 have been shown to improve spatial memory in rats[45, 46]. Further, hippocampal LTP also relies on NR2A-containing NMDARs[47], and exposure of pregnant mice to a viral mimic decreased NR2A levels in the hippocampus of offspring and impaired their spatial memory[48]. In contrast to maternal infections that are inflammatory in nature, maternal H. bakeri infection resulted in increased expression of the GluR1 and NR2A genes, Gria1 and Grin2a, whereas expression of the GluR4 gene, Gria4, was reduced. As Gria1 and Grin2a subunit expression increases during normal neurodevelopment, and Gria4 decreases[40–42], this suggests that maternal H. bakeri infection accelerates hippocampal glutamate synapse maturation.
CaMKII and PKC are calcium-dependent-kinases with important roles in LTP; both are linked to the positive regulation of GluR1 AMPAR subunit phosphorylation, which increases AMPAR conductance and postsynaptic responsiveness[19–21]. Injection or overexpression of CAMKII [49–51] or PKC [52, 53] enhances LTP and spatial memory. CaMKII has four distinct isoforms (α, β, γ, and δ)[54], and PKC consists of at least 10 isoforms, including the classical subfamily of PKC isozymes (PKCα, PKCβ and PKCγ) [55]. In response to maternal H. bakeri infection, we detected increased expression of CaMKII α and β genes (Camk2a and Camk2b, respectively), and the PKCα gene, Prkca. CaMKIIα and β and PKCα have distinct and crucial roles in neuronal plasticity and memory; loss of CaMKIIα, CaMKIIβ[56–59] or PKCα[55] result in severe LTP and spatial memory impairments. Taken together, higher expression of Gria1 and Grin2a, lower expression of Gria4, with higher expression of Camk2a, Camk2b and Prkca would be expected to promote LTP, as demonstrated here, and facilitate spatial memory in response to maternal infection, as we previously reported[10].
The increased gene expression associated with neurogenesis pathways may also contribute to enhanced spatial memory in juvenile offspring[10]. Neurogenesis in the hippocampal dentate gyrus contributes to spatial memory[60], and maternal physical exercise elevates hippocampal neurogenesis and spatial memory of offspring, via brain derived neurotrophic factor (BDNF) [61–64]. BDNF is a key ‘memory molecule’ as a positive modulator of LTP and neurogenesis[65]. Other data from our lab indicates that whole brain Bdnf mRNA is significantly up-regulated in neonatal brain in response to maternal H. bakeri infection (not shown; p = 6.8E-05), and here we found increased hippocampal Bdnf expression in response to maternal infection (p = 0.015, adjusted p value = 0.067). Whether enhanced hippocampal neurogenesis is an additional mechanism underlying enhanced spatial memory in juvenile offspring[10] remains to be further explored.
How maternal GI nematode infection alters hippocampal function and spatial memory in developing uninfected offspring may be explained by transfer of maternal immunity, as offspring immunity strongly influences neurodevelopment and behaviour[8]. Maternal immune antibodies and cells are transferred to young offspring via nursing, offering protection from pathogens and shaping the development of the immune system[25]. For example, via nursing, H. bakeri-specific IgG1 antibody is transferred from infected mothers to pre-weaned 10-day old neonates, protecting them against H. bakeri [26], and maternally derived Th2 CD4 + T cells are transferred from Nippostrongylus brasiliensis infected mothers to offspring, providing long-lasting cellular immunity against direct infection with this nematode[27]. Our hypothesis that the Th2/Treg immune response in the H. bakeri infected dam would be mimicked in 3-week-old uninfected weaned offspring was supported by the presence of H. bakeri-specific IgG1 in the serum of the uninfected offspring of infected dams. Further, the lower numbers of adult worms and lowered fecundity in H. bakeri infected offspring of infected dams, compared with infected offspring of uninfected dams, indicates heightened resistance and that the functional immunity induced by maternal infection parallels that seen during a secondary challenge in this mouse model[9].
In addition, we have evidence that the H. bakeri-specific antibodies in the offspring serum parallel a Th2/Treg biased neuroimmune response in offspring hippocampus, which may alter differentiation, development and function of glia. This might support LTP, neurogenesis, gliogenesis, and myelination, and contribute to improved spatial memory[10]. We previously reported that whole brain samples of uninfected neonates of H. bakeri infected dams had up-regulated expression of IL4 and TGF-β genes[13, 14]. These cytokines are hallmarks of an H. bakeri infection[9], and vital for the regulation of brain immunity with measurable downstream effects on LTP, neurogenesis, and spatial memory[66–70]. Consistent with this, we detected higher expression levels of the TGF-β signaling pathway and up-regulated expression of both Tgfb2 and Tgfbr1 genes but downregulated expression of the Tgfb3 gene in the hippocampus of uninfected juvenile offspring. Of the three isoforms of TGF-β, encoded by the genes Tgfb1-3, TGF-β2 regulates hippocampal synaptic plasticity[67] and neurogenesis[68], and TGF-β2 knockout mice are characterized by synaptic and cognitive dysfunction[71, 72]. Further, gene expression and protein levels of TGF-β2 and its receptor, TGF-β receptor 1 (TGF-βR1), are higher in IL-4 induced microglia[73], which are associated with immune regulation and memory[74]. In contrast, TGF-β3 has been found to promote Th17 cell differentiation and plays a role in the pathogenesis of autoimmune diseases[75, 76]. Thus, the up-regulated expression of both Tgfb2 and Tgfbr1 genes, and downregulation of the Tgfb3 gene suggests a TGF-β immunoregulatory response in the hippocampus of the uninfected offspring.
TGF-β is also critical for differentiation, development and function of neurons and glia[68, 77]. The higher density of microglia and higher expression levels of microglia-specific genes (Hexb, Sall1, Tgfbr1, Mef2a, Golm1, Tmsb4x, and Tppp), higher density of astrocytes and higher expression levels of astrocyte-specific genes (Nfia, Nfib, Gfap, S100B and Aqp4), and greater percentage of microglia with the CD206 marker, in the hippocampus of offspring of H. bakeri-infected dams, all indicate a hippocampus responding to Th2/Treg factors. CD206, like TGF-β2 and TGF-βR1, is typically increased on microglia responding to IL-4, and this cell surface marker suggests a microglia phenotype associated with immune regulation[30, 31, 35, 73]. Expression of Tgfbr1, Hexb, Golm1, and Sall1 increase with microglia maturity, and maturation is dependent on TGF-β signaling [77, 78] which was up-regulated by maternal infection. Nfia and Nfib play central roles in astrocyte development, Gfap, S100B and Aqp4 are markers of mature astrocytes[79], and Aqp4 is important in TGF-β-associated immunoregulation[80].
TGF-β signaling is also critical for oligodendrogenesis and developmental myelination[69]. In response to maternal H. bakeri infection, 13 key genes associated with oligodendrocytes and myelination had higher expression levels. Olig1, Olig2, Sox10, Nkx2.2, Myrf, Zfp488 and Cldn11, expressed in all cells of the oligodendrocyte lineage, are necessary for oligodendrocyte development and differentiation and important during developmental myelination[32, 81]. Plp1, Foxo4 and Cnp are expressed during transition from progenitor to differentiated oligodendrocytes, and differentiated, axon-myelinating oligodendrocytes are characterised by Mbp, Mag and Mog[32]. Delayed development of myelination in the hippocampus leads to impaired excitatory synaptic transmission and cognitive dysfunction[4, 82, 83], whereas early development of hippocampal myelination promotes excitatory synaptic transmission and cognitive function, including spatial memory[84]. Thus, earlier oligodendrogenesis and myelination in the developing hippocampus via TGF-β signaling could further contribute to the enhanced LTP and spatial memory[10] we observed in juvenile offspring in response to maternal H. bakeri infection.
IL-4 plays a critical role in cognitive function as administration of IL-4 improves LTP[24], neurogenesis[31] and spatial memory[85] and stimulates microglia and astrocytes to produce high levels of BDNF[31, 74, 86]. Conversely, the absence or reduction of IL-4 activates inflammatory microglia, reduces BDNF production by microglia and astrocytes, and impairs LTP, neurogenesis and spatial memory[31, 74, 86]. Although IL-4 expression was not altered in the hippocampus in response to maternal H. bakeri infection, IL-4 expression was up-regulated in the whole brains of neonates[13, 14], and as noted above, the higher expression levels of Bdnf, Tgfb2 and Tgfbr1, and a greater percentage of CD206 positive microglia are an indication of IL-4 induced microglia[30, 31, 35, 73]. Taken together, our data suggests a Th2/Treg biased neuroimmune response that would promote earlier LTP and spatial memory.
It is worth highlighting the contrasting consequences of maternal infection with bacteria and viruses, compared with intestinal nematodes. Maternal bacterial or viral infections induce a strong Th1/Th17 immune response in the mother which extends to the offspring brain, and is associated with microglia- and astrocyte-mediated neuroinflammation. If prolonged, neuroinflammation drives apoptosis of oligodendrocytes, delays myelination, and impairs LTP and neurogenesis, ultimately resulting in the emergence of ASD-like behaviours and cognitive impairments in the offspring[3–5, 82, 83, 87]. Conversely, our data demonstrate that maternal intestinal nematode infection induced a Th2/Treg immunoregulatory environment in the developing hippocampus, which was associated with upregulated genetic markers of neurogenesis, gliogenesis and myelination, and enhanced LTP and spatial memory[10]. We hypothesize that these differences result from the transfer of Th2/Treg-specific immune molecules from the nematode-infected mother but cannot exclude other pathogen-related differences in the maternal microbiome that may affect microbial colonization of the offspring[88, 89] with consequences for the developing immune system and brain. Regardless, if the observed changes induced by maternal nematode infection persist as pups grow, and if damaging neurological changes do not occur, this raises the possibility that GI helminths might be important for proper brain development and function, and may also provide a promising avenue for preventing inflammation-associated neurodevelopmental disorders.
We acknowledge the following limitations. Our electrophysiology experiments recorded LTP in male but not female offspring. We hypothesize that similar results would have been seen in females as sex did not affect spatial memory of juvenile offspring of H. bakeri infected mothers[10] or hippocampal gene expression or microglia/astrocyte numbers. Also, data from only the first slice tested per animal was used in our LTP experiment as subsequent slices were of lower quality, limiting our sample size, however, this avoided pseudoreplication. Finally, caution is needed in interpreting gene expression results as well as function of microglia and astrocytes given their high plasticity.
To the best of our knowledge, this is the first study to show that a maternal GI helminth infection promotes hippocampal LTP and upregulates genetic markers associated with neurogenesis, gliogenesis and myelination in the uninfected juvenile offspring, possibly through transfer of a Th2/Treg immune phenotype from the infected dam that protects the offspring from direct infection and extends to their developing brain. These findings identify mechanisms underlying our previous observation of enhanced spatial memory in two and three week-old offspring exposed to H. bakeri maternal infection[10]. These positive effects on neurodevelopment and cognition identify a potential unappreciated benefit of maternal GI helminth infection. Given the immunoregulatory nature of this parasite, which extends to the offspring, our findings may be valuable in efforts to prevent the development of inflammation-associated neurodevelopmental disorders, like ASD.