The authors have declared that no competing interests exist.
Conceived and designed the experiments: FL PG LRC LM AC AP EB MA HAB MIT VLS MRLA GAdE. Performed the experiments: FL PG LRC LM GAdE. Analyzed the data: FL PG LRC LM MRLA GAdE. Contributed reagents/materials/analysis tools: FL PG LRC LM AC AP EB MA HAB MIT VLS MRLA GAdE. Wrote the paper: FL PG HAB MRLA GAdE.
Influenza virus (InfV) infection during pregnancy is a known risk factor for neurodevelopment abnormalities in the offspring, including the risk of schizophrenia, and has been shown to result in an abnormal behavioral phenotype in mice. However, previous reports have concentrated on neuroadapted influenza strains, whereas increased schizophrenia risk is associated with common respiratory InfV. In addition, no specific mechanism has been proposed for the actions of maternal infection on the developing brain that could account for schizophrenia risk. We identified two common isolates from the community with antigenic configurations H3N2 and H1N1 and compared their effects on developing brain with a mouse modified-strain A/WSN/33 specifically on the developing of dopaminergic neurons. We found that H1N1 InfV have high affinity for dopaminergic neurons in vitro, leading to nuclear factor kappa B activation and apoptosis. Furthermore, prenatal infection of mothers with the same strains results in loss of dopaminergic neurons in the offspring, and in an abnormal behavioral phenotype. We propose that the well-known contribution of InfV to risk of schizophrenia during development may involve a similar specific mechanism and discuss evidence from the literature in relation to this hypothesis.
Influenza virus (InfV) infection during pregnancy is a known risk factor for neurodevelopment in the offspring, increasing the risk of schizophrenia
Experimental infection of pregnant mice with InfV results in pathological and physiological changes in the brain of offspring that resemble those observed in schizophrenia
Regardless of the molecular mechanism by which infection results in neurodevelopmental changes, deviant behavior in adult offspring should have a correlate in structural or functional changes in brain circuitry, and there are reasons to consider the dopaminergic system as a potential target. Indeed, intracerebral inoculation of InfV in adult rodents results in viral antigen accumulation in dopaminergic neurons
Is there any evidence that susceptibility of dopaminergic neurons to InfV A infection may modify risk of schizophrenia? First, the observation that parkinsonism can occur as a trait associated with schizophrenia suggests a dopaminergic deficit
Thus, we hypothesized that maternal infection with InfV during the critical period of risk for schizophrenia (the equivalent of the human second trimester, or embryonic days 11–16 in rodents), should result in preferential damage to developing dopaminergic neurons in the brainstem of offspring. We further hypothesized that such damage depends on the antigenic configuration of circulating InfV. In the present work we evaluated both of these hypotheses in primary cultures of ventral mesencephalon and in a mouse model of prenatal infection with circulating InfV strains.
All animal work was conducted according to relevant national and international guidelines and approved by the Animal Studies Committee of ANLIS Carlos G Malbran (Resolución 20/4/2004). Rat embryos were recovered at day 14 from timed pregant Wistar rats, and the ventral midbrain was dissected, mechanically dissociated, and plated on polyethylenimine (1 mg/ml)-coated culture wells, in DMEM/F12 medium, supplemented with 5% horse serum (US) and
Replication of viruses under liquid medium and plaque determinations are performed in MDCK cells (passage 59) obtained from the Centers for Disease Control and Prevention (CDC, Atlanta, GA).
We identified two common isolates from community samples of Buenos Aires from the Instituto Malbrán, with antigenic configurations H3N2 and H1N1. The results of characterization with ferret antisera are: A/New Caledonia/20/99-like (H1N1) (A/NC-L/99), and A/Sydney/5/97-like (H3N2) (A/Sy-L/97). Isolates were amplified in MDCK cells and frozen in lots of 100 cryotubes. We used a well characterized neurovirulent strain obtained from the CDC (Atlanta, GA) (A/WSN/33) as a positive control for toxicity to dopaminergic neurons
Titration of plaque forming units (PFU) was performed in MDCK cells and normalized for each lot to:
Lot 1 A/Sy-L/97 (H3N2) 3,000 PFU/ml.
Lot 1 A/NC-L/99 (H1N1) 3,000 PFU/ml.
Lot 1 A/WSN/33 (H1N1) 3,000 PFU/ml.
Viral RNA was purified from selected samples by QIAamp®Viral RNA Mini Kit.
RT PCR was used to amplify regions of the protein coding domains of viral RNA segment 7 to obtain finally an amplicon of 560 bp, using the QIAGEN®One Step RT-PCR Kit.
The antigen detection was performed by Indirect immunofluorescent assay (IFA) using the standard antibody for surveillance recommended by CDC (Atlanta, GA), namely mouse monoclonal anti influenza a nucleoprotein (immunized with Influenza A/Puerto Rico/8/34 (H1N1) and A/Bangkok/1/79 (H3N2) viruses), from spleen cells from BALB/c mice fused with cells of the P3 Ag8.653 mouse myeloma cell line (AbD Serotec, USA).
For neurotropism experiments we inoculated 50 µl of viral suspension (150 PFU/well) in primary cultures of rat mesencephalon, and examined the presence of viral infected cells at 24 h, 48 h and 72 h (after fixation with formalin), by indirect immunofluorescence with monoclonal anti-InfV A antibodies (CDC, Atlanta, GA) followed by fluorescein-conjugated anti mouse (Light Diagnostics) (
Culture dishes infected with New Caledonia-like (H1N1) (panels a–c) or Syndey-like (H2N3) strains (panels d–f), at 24 (a, d), 48 (b, e) and 72 hours (c, f) post inoculation. The two community isolates differed from each in that the New Caledonia-like strain (A/NC-L/99) showed greater affinity at earlier times and more toxic than Sydney-like (A/Sy-L/97) virus.
BALB-C mice were obtained from the Bioterio Central of the Instituto Nacional de Producción de Biológicos. Because infection entails discomfort to the animals, we minimized it by ensuring adequate hydration, temperature control, and cage conditions according to strict guidelines (Instituto Malbrán/CDC). Intranasal inoculation with 30 µl (3.000 PFU/ml of InfV, 90 PFU/mouse) strains or vehicle (PBS, pH 7.4) was performed at embrionic day 9–11 on the assumption that viremia peaks (and most likely invades CNS) three days later (and therefore ∼ embryonic day 14). Post-infection all animals were individually kept in filtered cages with ad lib access to food and water, at 25°C. Seroconversion was established by inhibition of hemagglutination of turkey red blood cells at day 14 following infection. We calculated the 50% infective dosing by measuring change in lung weights post infection in lots of 5 animals for each strain and for each of four dilutions. ID50 (expressed as dilutions of a normalized hemagglutinin unit) for each strain were: A/Sy-L/97 = 1.07E-02, A/NC-L/99 = 1.41E-02, and A/WSN/33 = 2.49E-02. We did not find lethal infections at any of the used viral loads and therefore LD50 could not be estimated for any strain. Institutional and ethical committee approval for this protocols was obtained at both centers. ID50 was calculated for each viral strain using the Reed and Muench Calculator
Pups from infected mothers were examined in a behavioral test battery including elevated plus maze, open field, novel object, and spontaneous activity before puberty (30 days) or in adulthood (90 days). In the latter we also carried out an object recognition task (90 days). Behavioral tests were carried out in an experimental room evenly illuminated provided with white noise. Animals were handled for 5 min daily for 3 days prior to testing. Sessions were recorded and later analyzed using a computerized video-tracking system (Ethovision XT, version 5, Noldus Information Technology, Wageningen, Netherlands) or ethological observation software (JWatcher V1.0).
Elevated plus maze testing was carried out in a standard mice apparatus elevated 37.5 cm from the floor. At designated times animals were placed onto the centrdal platform facing an open arm and allowed to freely explore the maze for 5 min. After each session the apparatus was cleaned. An arm entry was counted when all four paws were placed into an arm.
Animals were placed in the center of a standard open field apparatus and total distance moved, number of rearings and time spent in the central area were recorded for 20 minutes. A mouse was considered to be into the central area (arbitrarily defined as a square of 30×30 cm) when its four paws were in it. The apparatus was cleaned between sessions.
Upon completion of the 20-min open field session, a novel object (a metal cup), was placed in approximately 20 cm away from the mouse and the latency to contact the object and time spent in contact with it were recorded for 10 min.
Two days after the open field and novel object tests, the spontaneous locomotor activity was monitored using a webcam and quantified using a simple criteria for number of squares crossed per unit time (24 h).
For 3 consecutive days prior to this task mice were handled once a day for 5 min and placed 10 min in the open field to allow habituation. On the fourth day each mouse was observed during two 5 min trials separated by an interval of 1 h during which animals were returned to their cage. In the sample trial (T1) mice were faced with two identical objects placed in a symmetrical position and the time exploring each object was recorded. In the retention trial (T2) mice one of the two objects was replaced by a novel, non-familiar object and the time exploring each was recorded. Sets composed of three copies of the same object were used to prevent odor cues and all combinations and location of objects were used to prevent bias due to preference for a particular object or location. Exploration time was computed when the snout pointed to the object at a distance ≤2 cm.
After behavioral testing mice were sacrificed and fixed, brain sections (15 µm) were obtained in stereological series (every 8th section), and processed for histopathological analysis that included gross morphology with Nissl staining, stereological quantitation of immunohistochemistry for dopaminergic neurons and reactive astroglia in the substantia nigra and ventral tegmental area, and electron microscopy of dopaminergic neurons stained with tyrosine hydroxylase. Post-fixed brains stored in 1× PBS containing 0.05% sodium azide were embedded in 3% low-melting agarose, sectioned cornally (40 µm) on a vibratome (Leica VT1200) and stored in cryoprotectant (30% glycerol, 30% ethylene glycol in 1× PBS) at −20°C until processing. Immunohistochemistry was performed on every 6th section. For immunohistochemistry sections were washed in PBS endogenous peroxidases were blocked by incubation in 3% H2O2, and blocking of non-specific binding was achieved by incubating in PBS containing 0.25% (w/v) Triton-X 100 and 10% normal serum. Sections were incubated overnight at 4°C with one of the following primary antibodies: Tyrosine Hydroxylase (TH, Pelfreeze Biologicals, Rogers, AR; or Boehringer GmbH, Mannheim, Germany), Nuclear Factor kappa B (NFkBp65, Abcam, Cambridge, MA), Anti-GFAP antibody - Astrocyte Marker (ab4674, Abcam, Cambridge, MA) and Anti-InfV A antibodies provided by the Instituto Malbrán (Buenos Aires, Argentina). For microglia activation we used lectin staining (IsoB4) and antibody staining (ED-1). Following washing, sections were then incubated with appropriate biotinylated secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 hr followed by incubation with avidin-biotin-peroxidase complex (ABC Elite Kit; Vector Laboratories) detection system with diaminobenzidine (Ultratech HRP Streptavidin-biotin Universal Detection System and DAB Chromogen Kit, Inmunotech Co, Marseille, France). For immunofluorescence labeling, sections were stained with streptavidin conjugated Alexa Fluor 568 (working solution 1–5 µg/ml; Molecular Probes, Eugene, OR). For detection and quantification of apoptosis (programmed cell death) at single cell level, we used a protocol based on labeling of DNA strand breaks (TUNEL technology) carried out following manufacturer’s recommendations (DeadEnd™ colorimetric TUNEL System, Promega, Madison, WI).
Embryonic brains 3 days after maternal infection were fixed with 3% v/v glutaraldehyde in PBS overnight, washed, postfixed in 1.5% w/v osmium tetroxide in the same buffer for 2 hs, contrasted with 2% w/v acuose uranyl acetate, and embedded in Fluka Poly-Bed 812 (Sigma-Aldrich, St Louis, MO). Ultrathin sections (90–100 nm) were collected in cupper grids and stained with uranyl acetate and Reynolds solution, and imaged using a Philips EM300 transmission electron microscope.
All experiments were carried out in triplicates. For stereology, cells were counted using every 8th consecutive section, 400 µm apart from each other, throughout the entire ventral mesencephalon (substantia nigra compacta and ventral tegmental area) using a Nikon 80i Eclipse microscope (Tokyo, Japan) equiped with Stereologer software (SRC, Tampa, FL) and a motorized stage. Digital photographs were taken using a Zeiss camera with AxioVision software (Carl Zeiss Microscopy GmbH, Göttingen, Germany). Digital images were then processed with ImageJ (NIH). The cross sectional area of each section was determined with Stereologer and the data were expressed as number of cells per hemibrain. Statistical analyses were performed with SPSS software (IBM Corporation, Somers, NY). Data are presented as mean ± standard error of the mean (SEM). Comparisons among groups were performed using a one or two way ANOVA as indicated, followed by a Kruskal–Wallis post hoc test as appropriate. For significant differences α was set at 0.05. For multivariate analysis of behavioral performance, discriminant analysis was performed using prenatal exposure as the classification criterion (SPSS, IBM Corporation, Somers, NY).
Epidemiological reports show an association between respiratory infection by InfV A in pregnancy and risk of schizophrenia to the offspring; we investigated the effects of common respiratory InfV A strains on development of dopaminergic neurons in primary mesencephalic cultures to test the hypothesis that InfV A could mediate its effects on schizophrenia risk through developmental damage to specific dopaminergic projections
Because control cultures undergo progressive attrition of neurons in serum-free conditions we compared the effect of infection to spontaneous neuronal loss over time in PBS inoculated wells (
Panels (a–c) display characteristic morphology changes of TH stained neurons at baseline (a), 48 h (b) and 72 h (c) post infection; dopaminergic neurons are clearly damaged by the virus, with loss of dendrites and vacuolation of the cytosol. Time curves of TH stained neuronal counts for each viral strain regarding of their morphological state, and therefore probably representing an overestimation of surviving neurons. The A/WSN/33 strain was most toxic at 24 and 48 h. Points represent means of 6–8 independent experiments (two wells per experiment). Standard errors are two small to be displayed on this scale (range 11 to 86). Comparisons were performed by two way ANOVA (treatment and time). ** p<0.001 compared to vehicle.
Panels a and b show TUNEL staining in primary cultures following inoculation of cultures (11 day in vitro) with either PBS solution (a) or A/WSN/33 (d) at 200X. Cultures were fixed, stained for TUNEL and counted at times specified on the x axis of panel (c). Cell counts bars represent means of 6–8 independent experiments (two wells per experiment). Error bars are SEM. Comparisons were performed by two way ANOVA (treatment and time). ** p<0.001 compared to control (PBS solution).
We have previously described that programmed cell death of dopaminergic neurons in this culture system is mediated by NFkB (51), and others have shown that NFkB signaling is an early and necessary step in InfV-triggered apoptosis pathway
Panels (a) and (b) show NFκB staining in primary cultures following inoculation of A/WSN/33 (a) or PBS solution (b). Cultures (11 day in vitro) were inoculated with PBS, A/WSN/33, A/NC-L/99 (H1N1), A/Sy-L/97 (H2N3) influenza strains as described in the text, fixed, immunostained for NFκB and neurons were counted at times specified in panel (c). Cell bars represent means of 6–8 independent experiments (two wells per experiment). Error bars are SEM. Comparisons were performed by two way ANOVA (treatment and time). ** p<0.001 compared to control (PBS solution).
Pregnant (embryonic day 9–11) mice received intranasal infusions containing A/NC-L/99 (H1N1), A/WSN/33 (H1N1), or vehicle alone (control). Gross inspection of mesencephalon of mouse embryos 72 h after mothers were infected with InfV with Nissl staining at low magnification revealed dystrophic appearance of neurons, with significant cytosolic vacuolation and increased intercellular space, both of which were more apparent on offspring of mothers infected with A/WSN/33 strains (
dams | offspring | young offspring | adult offspring | ||||
|
|
|
|
|
|
|
|
Control | 38 | 0 | 29 | 20 | 4 | 16 | 4 |
A/NC-L/99 | 24 | 17 | 23 | 13 | 4 | 9 | 4 |
A/WSN/33 | 39 | 22 | 34 | 25 | 4 | 18 | 4 |
A/Sy-L/97 | 9 | 0 | 12 | 0 | 0 | 0 | 0 |
Test | Control | A/WSN/33 | A/NC-L/99 | ANOVA for Strain |
|
||||
Open Field | ||||
|
17.44+/−2.30 | 24.39+/−1.43 | 19.78+/−2.62 |
|
|
972.86+/−184.67 | 1791+/−232.43 | 944.24+/−196.17 |
|
|
1.34+/−0.95 | 3.09+/−0.93 | 1.22+/−0.56 |
|
Novel Object | ||||
|
448.05+/−40.47 | 284.64+/−38.21 | 318.69+/−59.1 |
|
|
14.43+/−5.68 | 32.84+/−6.66 | 42.15+/−16.52 |
|
Elevated plus maze | ||||
|
10.05+/−0.99 | 15.36+/−1.93 | 8.46+/−1.86 |
|
|
||||
Open Field | ||||
|
14.13+/−2.22 | 28.56+/−2.80 | 26.00+/−3.29 |
|
|
879.79+/−174.46 | 1329.77+/−487.70 | 947.97+/−285.45 |
|
|
0.70+/−0.53 | 27.50+/−8.16 | 30.53+/−14.53 |
|
Novel Object | ||||
|
421.63+/−45.06 | 199.78+/−34.85 | 186.89+/−27.33 |
|
|
28.44+/−8.56 | 75.83+/−18.53 | 74.00+/−15.55 |
|
Object Recognition | ||||
|
5.45+/−1.63 | 0.80+/−3.68 | −1.70+/−6.16 |
|
Elevated plus maze | ||||
|
6.13+/−0.97 | 14.11+/−1.67 | 5.13+/−1.34 |
|
Means are provided for for exploratory behavior, rearing and and anxiety levels in the Open Field, exploratory behavior of a novel object, and performance on an elevated cross maze test of young (post natal day 30) and adult (post natal day 90) offspring of mothers inoculated with control solution (PBS) or infected with influenza A strains A/WSN/33 or A/NC-L/99. A separate cohort of adult offspring was tested for an object recognition task; time spent exploring the familiar and non-familiar object was computed and the difference (Δ) is reported. Values represent means ± SEM. Bold values in grayed cells represent p<0.01 compared to control. Italic values in grayed cells represent p<0.05 compared to control. Statistics represent ANOVA followed by Kruskal Wallis post hoc comparisons for means.
Top panels display low magnification (X200) of the area of the substantia nigra stained with Nissl. Bottom panels show electron micrographs (X24,000) obtained from the same specimens. Infection with influenza results in significant swelling of midbrain embryonic neurons. Arrowheads indicate vacuolar dilatations.
Open field behavior in young mice revealed a significant effect of prenatal exposure to InfV A, whether considering total trajectories, number of rearings, or time spent in the central area. Young offspring of mothers infected with A/WSN/33 had longer trajectories, higher number of rearings and spent more time in the central area of the open field than animals exposed to A/NC-L/99 or controls. The latter two were not distinguishable on this test. Likewise, adult offspring from mothers exposed to A/NC-L/99 or A/WSN/33 had significantly more rearings and spent more time in the central area than offspring of control mothers, but the increase in total trajectory did not reach significance.
When a novel object was introduced to young mice, exposure to InfV A during pregnancy resulted in significantly reduced latency to the first contact with the novel object, and more time spent in contact with it during the session. However, even though the mean values for offspring of mothers infected with A/NC-L/99 were indistinguishable from those of mothers infected with A/WSN/33, only the latter were significantly different from control offspring, suggesting a type 2 error. Indeed, when adult mice were exposed to the novel object offspring from mothers exposed to both strains showed significantly reduced latency and increased contact with the novel object. During the object recognition task exploration times during the first session (T1) were unaffected by prenatal exposure, even though offspring of mothers infected InfV A strains showed a trend towards longer exploration times. On the other hand, exploration times of the familiar and non-familiar objects during T2 revealed a significant effect of prenatal exposure to InfV A such that offspring of control mothers spent significantly more time exploring the novel, non-familiar object while no differences were observed in offspring of mothers infected with A/NC-L/99 or A/WSN/33, suggesting working memory impairment.
In the elevated cross maze both young and adult offspring from A/WSN/33 infected mothers spent significantly more time in the open arms, whereas offspring of A/NC-L/99 infected mothers were undistinguishable from controls.
Multivariate analysis of the behavioral data by discriminant analysis revealed that the net effect of prenatal infection with InfV virus on behavior could be described by two statistically significant discriminant functions accounting for more than 90% of the variance (
Eigenvalues | ||||
Discriminant Function | Eigenvalue | % of variance | Cumulative % | Canonical Correlation |
1 | 63.842 | 91.9 | 91.9 | 0.992 |
2 | 5.632 | 8.1 | 100 | 0.922 |
|
||||
|
|
|
|
|
1 through 2 | 0.02 | 22 | 36.383 | 0.28 |
2 | 0.151 | 11.351 | 10 | 0.331 |
|
||||
|
|
|||
1 | 2 | |||
Control | −4.595 | 2.470 | ||
A/NC-L/99 | 8.020 | −0.475 | ||
A/WSN/33 | −8.382 | −3.167 | ||
|
||||
|
|
|||
|
|
|
||
Control | 100 | 0 | 0 | |
A/NC-L/99 | 0 | 100 | 0 | |
A/WSN/33 | 0 | 0 | 100 |
Two significant functions were identified which allowed correct classification of all animals, suggesting that neuroadapted (A/WSN/33) and respiratory (A/NC-L/99) strains of H1N1 influenza result in significant differences in severity of abnormal behavior in the offspring of mothers infected during pregnancy.
In summary, maternal infection with both strains resulted in abnormal behavior in the offspring, including time spent in the open arm of the elevated plus maze, changes in exploratory behavior, reduced latency to contact and increased time in contact with a novel object, and working memory impairment when compared to offspring of mothers exposed to vehicle. Additionally, the behavioral alterations appear earlier in the offspring of mothers infected with A/WSN/33.
Histopathological abnormalities in the offspring of mice infected with InfV strains during pregnancy was carried out in two sets of samples obtained at p30 (n = 4 per group), and at p90 (n = 4 per group). Both viral strains caused significant dystrophic alteration and loss of dopaminergic neurons in the mesencephalon of adult (post-natal day 90) offspring of infected mothers (
From left to right, panels show increasing magnification of tyrosine hydroxilase immunostaining (100X, 400X, 1000X). There is a clear loss of stained neurons in the exposed animals, and at higher magnification a dystrophic appearance of surviving neurons is evident. Panel (j) displays mean ± SEM of stereological counts of dopaminergic (TH positive) neurons throughout the entire brainstem (n = 4/group). * p<0.05 compared to control. The insert displays the results discriminating between substantia nigra pars compacta (SNpc) and ventral tegmental area (VTA) in a separate set of 4 animals for each viral strain.
Sections consecutive to the ones used for
The three major findings of the present work provide evidence in favor of the hypotheses tested. First, we found that circulating strains of InfV infect cultured mesencephalic neurons resulting in selective loss of dopaminergic neurons; this effect was dependent on the antigenic configuration of the strain, it was preceded by NFkB activation and mediated by apoptosis. Both H1N1 strains of InfV had the greatest affinity for dopaminergic neurons, whereas an H3N2 strain induced apoptosis preferentially in other cell types, and did not result in NFkB activation.
Second, following maternal infection with H1N1 (but not H3N2) InfV strains we found a persistent and a selective loss of dopaminergic neurons in substantia nigra pars compacta and ventral tegmental area of the offspring (
Third, offspring of mother infected with both InfV strains showed marked behavioral abnormalities in exploration, anxiety and working memory. Additionally, behavioral alterations emerge in different neurodevelopmental stages depending on the InfV strain, appearing in adult life in offspring of mother infected with A/NC-L/99.
Could the specific findings with each viral strain could perhaps be explained by differences in LD50 or TCID50, rather than because of antigen configurations? We specifically assessed this question and found that TCID50 of both H1N1 strains in brain primary cultures was similar, but one order of magnitude higher than the TCID50 for the H3N2 isolate. Yet, the H3N2 strain caused more apoptosis than either H1N1 strain, indicating that neurotoxicity is not limited by TCID50. On the other hand, the effect of both H1N1 strains was selectively higher on dopaminergic neurons, again indicating specificity of this effect. In vivo, all three strains resulted in effective infection judged by increased lung weight, such that A/NC-L/99 = A/WSN/33> A/Sy-L/97. Indeed, no seroconversions were found after infection of the dams with the latter strain and thus no further studies were carried out in their offspring. In summary, differences in infectivity between strains do not explain the specificity of their effects on different neuronal populations (i.e., dopaminergic vs non-dopaminergic)
Schizophrenia is a heterogeneous disorder with genetic, epigenetic and environmental factors contributing to its causation
In 1916, an epidemic of encephalitis lethargica suggested an association between viral infections, parkinsonism and psychosis, and a recent case series confirmed an encephalitis of the midbrain and basal ganglia
Thus, the net effect of InfV infections on brain pathology may depend on genetic susceptibility, antigenic strain characteristics, and neurodevelopmental stage. Our data support the view that in susceptible individuals, prenatal infection may result in selective destruction of key dopaminergic projections increasing the risk of neuropsychiatric disorders.
Extracellular cleavage of the hemagglutinin by host trypsin-like proteases is a prerequisite for the infectivity and pathogenicity of human InfV A, and tissue specific infectivity is largely dependent on the availability of proteases
Neurovirulence in mice is a unique property of some InfV strains. We included a neurovirulent strain (A/WSN/33) as a positive control in our experiments. A brief review of the published effects of direct infection with this InfV strain in brain is warranted. Indeed, A/WSN/33 in culture strictly infect neurons
Mice survive and the viral infection is cleared from the brain, but a variety of neuronal changes occur over a period of weeks. Infected neurons up regulate Fas ligand molecules leading to activation of JNK signal transduction pathway followed by DNA fragmentation and activation of caspase-3
Viral gene products are eliminated from brainstem dopaminergic neurons by a mechanism dependent on Transporter associated with Antigen Presentation 1 (TAP1)
A/WSN/33 injection resulted in behavioral changes months after infection, including increased exploration in the open arms of an elevated plus-maze
Lastly, after intranasal inoculation of A/WSN/33 invasion of brain also occurs, once again with greatest affinity for catecholaminergic neurons
In addition to direct neuronal infection and apoptosis, several other mechanisms have been invoked that could result in neuronal damage. Secondary production of autoantibodies against neuronal populations in hippocampus, cerebral cortex and cerebellum, also may occur after infection with H1N1 InfV viruses, including A/WSN/33
Experimental infection with InfV with human virus isolates without adaptation to an animal host has been also shown to result in neuropathological outcomes and changes in brain function associated with areas of virus replication in neurons
In brains exposed prenatally to A/WSN/33, glial fibrillary acidic protein (GFAP), an important marker of gliosis, neuron migration, and reactive injury increases in cortical and hippocampal cells; GFAP-positive cells have ‘hypertrophy’ and more stellate morphology
Exposure to the lipopolysaccharide during the critical developmental window in rats (embryonic day 10.5), leads to the birth of animals with fewer than normal dopaminergic neurons, reduction in striatal dopamine, and increased TNF-α
In summary, there is at least one strain of circulating InfV (not pandemic) in Argentina able to selectively decrease of dopaminérgic neurons in the offspring of dams infected during a critical developmental period. Selective loss of dopaminergic neurons follows activation of NFkB and apoptosis, and results in profound behavioral abnormalities when the animals reach adulthood.
GdE is Sydney R. Baer and Stephen and Constance Lieber Investigator and Roskamp Chair of Biological Psychiatry at University of South Florida.