Impaired learning resulting from Respiratory Syncytial Virus infection

RSV Reaches the CNS After Intranasal Inoculation.

To evaluate infection of the CNS after viral pneumonia, BALB/c mice and Sprague-Dawley rats were instilled i.n. with either RSV or a noninfectious supernatant (i.e., mock), which was used as a negative control. After infection, disease progression was monitored by body weight loss, and airway inflammation was assessed by histopathology and flow cytometry (Fig. S1). At 1, 3, 7, and 10 days postinfection, the brain, lungs, olfactory epithelium, and blood were collected from infected and mock mice and analyzed by real-time RT-PCR for the detection of RSV N transcripts. As shown in Fig. 1, RSV-N transcripts were detected as soon as 3 d postinfection in the CNS of mice. Furthermore, viral spread occurred from the midbrain to the brainstem at day 7 (Fig. 1).

To clarify the possible entry pathway, we measured RSV RNA in the olfactory bulb, olfactory epithelium, and blood cells by real-time PCR. Viral genetic material was detected in olfactory epithelium as early as 24 h postinfection and in the olfactory bulb at 3 d postinfection. This result is consistent with previous studies showing RSV infection in human nasal epithelial cells (20, 21), which suggests that the olfactory epithelium is a primary target for RSV infection. In addition, similar experiments in RSV-infected rats were performed to evaluate disease parameters in airways and the CNS (Fig. S2). Conversely, RSV-N transcripts were detected in peripheral blood mononuclear cells (PBMCs) at 7 d postinfection, which is in accordance with previous reports (22, 23). Also, F protein expression was detected by flow cytometry in PBMCs isolated at days 3 and 7 postinfection (Fig. S3). Twenty-four hours after RSV challenge, both the F and N proteins of RSV were detected by immunofluorescence analyses in the olfactory bulb of infected, but not mock, mice (Fig. 2C). Starting at day 3 postinfection, disseminated staining for clustered viral proteins was observed in different anatomic areas of the brain (Fig. 2 A–D). Finally, at day 7 postinfection, RSV was detected in the brainstem (Fig. 1).

Total F-RSV-associated fluorescence in both whole brain [cortex, hippocampus, ventromedial hypothalamic nucleus (VMH), and olfactory bulb] and whole lung (Fig. 2E) revealed F-RSV presence in RSV-infected mice (Fig. 2E). However, F-RSV levels were lower in the brain compared to the lungs. To establish whether RSV was present in other tissues not related with the CNS, we performed immunofluorescence for RSV F protein. As shown in Fig. S4, RSV F protein could not be detected in other organs after RSV infection. Altogether, real-time and immunofluorescence data support the notion of RSV as a unique neurotropic virus. Notably, the most infected areas in the brain were those adjacent to ventricles, including the third ventricle and the contiguous VMH. In addition, RSV was detected in other zones of the brain, including the hippocampus and cortex (Fig. 3).

We also evaluated whether RSV could use a so-called “Trojan horse” mechanism via the hematogenous pathway to enter and spread within the CNS. The Trojan horse mechanism refers to CNS viral entry via RSV-infected leukocytes that cross the blood–brain barrier. To test this, infected and mock mice were administered with a blocking antibody for CD49d i.v. before the infection. CD49d is expressed by these cells and is involved in the transendothelial migration of leukocytes into tissues that are undergoing inflammation (24, 25). CD49d antibody prevents the interaction of immune cells and endothelial cells, thereby reducing their translocation (24). As shown in Fig. 4, 3 d postinfection, the viral load is reduced in brain tissues in animals treated with CD49d. These data suggest that the hematogenous pathway is involved in RSV entry into the CNS; however, an alternative-pathway CNS RSV entry cannot be ruled out.

RSV Infection Causes Long-Term Learning Impairment.

To evaluate the possible consequences of RSV infection in the CNS, the performance of control and RSV-infected BALB/c mice was evaluated by MB tests 30 d after viral challenge. Remarkably, mice infected with RSV 1 month before the test showed a significant impairment of MB behavior compared with control mice (Fig. 5 A and B). RSV-infected and control rats tested under similar experimental conditions (SI Materials and Methods) showed a similar impairment in MB behavior consistent with the effect observed in mice (Fig. S5).

We further evaluated cognitive function using the MWM test with RSV-infected Sprague-Dawley rats (26). One month after RSV infection, mock and virus-challenged rats were subjected to the MWM test (for further details, see SI Materials and Methods). Our results showed that infected rats displayed a significant increase in latency compared with mock control rats (Fig. 6A and Table S1). On the first day, RSV and control rats had latencies between 70 and 80 s (Table S1), indicating that none of the groups had swimming impediments. Importantly, between days 1 and 3, the group infected with RSV had significantly higher P ≤ 0.05 latencies than the control group. However, by days 4 and 5 of the test, there were no significant differences between the latencies of RSV-infected and control animals. These data suggest that despite being infected 1 month before the assay, RSV-infected rats needed significantly more time and/or experience to learn the spatial cues required to find the hidden platform compared with control rats (Fig. 6A). It seems unlikely that the learning delay in RSV-infected rats was a result of visual perception defects because both RSV-infected and control rats showed equivalent latencies on MWM tests performed with a visible platform (cued-task training) (Fig. 6B). Therefore, normal visual perception of RSV-infected rats suggests that RSV infection may cause a long-lasting impairment in spatial learning capacity.

Learning Capacity Impairment Caused by Prior RSV Infection Is a Result of Reduced LTP.

To determine how prior RSV infection impairs MWM and MB performance, we analyzed function of the hippocampus because of its direct role in the spatial learning and memory processes. Specifically, we focused on synaptic plasticity events occurring in the hippocampus, such as LTP and LTD. These two types of synaptic plasticity are generally considered the closest neural model for the cellular mechanism responsible for learning and memory storage. LTP experiments consisted of recording stimulus-evoked field responses [field excitatory postsynaptic potentials (fEPSPs)] at the stratum radiatum of CA1 of hippocampal slices that were obtained from control and previously infected rats (27). A 20-min stable baseline transmission was obtained and tested every 15 s. After baseline acquisition, the TBS was applied as described in SI Materials and Methods. After stimulation, data were acquired once every 15 s for 1 h. Data from several experiments were compared according to the time of TBS stimulation for control and RSV-infected animals (Fig. 7B). For control rats, 176.4 ± 16% (Table S2) (SEM; n = 5), potentiation was achieved relative to the baseline. In contrast, for rats previously infected with RSV, potentiation was significantly reduced, at 138.4 ± 4% (Table S2) (SEM; n = 7). Representative traces of fEPSPs observed before and after LTP induction for control and RSV-infected animals are shown in Fig. 7C. These results indicate that RSV infection led to an impaired LTP response. Also, we evaluated LTP induction in the stratum pyramidale of CA1 of hippocampal slices and there were no differences between mock or infected animals (Fig. S6).

We also studied a second mechanism of synaptic plasticity known as LTD. To induce LTD, hippocampal slices received PBS stimulation 20 min after baseline (SI Materials and Methods). Our data show that RSV-infected animals demonstrated a significant impairment in the LTD response (Fig. 7D and Table S3). Impaired LTP and LTD responses were consistent with the increased latency shown by RSV-infected rats in MWM tests. These results support the hypothesis that rats previously infected with RSV are able to learn but require longer training periods compared with uninfected rats.

CD49d Blockade Prevents the Behavioral Impairment Caused by RSV.

As shown in Fig. 4, treatment with anti-CD49d caused a significant viral load decrease in the brain of infected mice. We evaluated whether CD49d blockade protected against CNS alterations caused by RSV infection. BALB/c mice were given one or two doses of CD49d (70 μg) and then infected with the virus, as described in SI Materials and Methods. Thirty days postchallenge, MB tests were performed on control and anti-CD49d–treated mice. As shown in Fig. 8, mice that received the blocking antibody displayed normal MB behavior, despite RSV infection (Fig. 8). These results suggest that CD49d antibody blockade precludes RSV from entering the CNS, thereby preventing the deleterious effects of the virus on MB performance.

rBCG-N Vaccine Prevented RSV-Caused Damage to Airways and CNS.

We have previously shown that immunization with live, attenuated BCG-N can protect against RSV-induced lung disease (28) (Fig. S7). To evaluate whether immunization with BCG-N could also induce an immune response capable of protecting against CNS infection by RSV, BALB/c mice were immunized with either BCG-N or BCG-WT and inoculated with RSV (Materials and Methods). Detection of RSV nucleoprotein transcripts was performed 3 d postinfection by real-time RT-PCR in the lungs and brains of infected mice. BCG-N-immunized mice showed significantly lower viral loads in both lungs and brain than did mice immunized with BCG-WT (Fig. 9A). We then performed MB tests 30 days postchallenge in vaccinated and control mice to evaluate whether immunization could prevent behavior impairment. BCG-N-immunized mice showed normal MB behavior despite RSV infection (Fig. 9B). These data suggest that immunization with BCG-N can confer an effective immune protection against RSV infection and respiratory disease, as well as preventing virus spread to the brain and RSV-induced CNS alteration.

Animals and RSV Challenge.

Four- to 6-wk-old female Balb/cJ WT mice and 3-wk-old female Sprague-Dawley rats were obtained from Jackson Laboratory and the central nurse-pond of the Pontificia Universidad Católica de Chile, respectively. Animals were maintained at the pathogen-free animal facility at the Pontificia Universidad Católica de Chile. All animal work was performed according to the respective biosafety committees of the Pontificia Universidad Católica de Chile and Universidad Andrés Bello. RSV was propagated over human epithelial cell line 2 (HEp-2) cells, as described in SI Materials and Methods. The RSV challenge was performed in a special room of the Public Health Institute of Chile. Animals were anesthetized with ketamine/xylazine (20 mg/kg and 1 mg/kg, respectively) and challenged i.n. with 1 × 10⁷ pfu of the RSV serogroup A, strain 13018–8. RSV is a clinical isolate provided by the Public Health Institute of Chile and was propagated over HEp-2 cells (SI Materials and Methods). As a control, noninfectious HEp-2 supernatant (mock) was instilled i.n. Body weight was determined daily after infection.

Mouse Immunization.

rBCG-N was obtained as previously described (3). Five- to 6-wk-old BALB/c mice received a dorsal-flank subdermal injection with 1 × 10⁸ cfu of either a Calmette–Guérin-WT bacillus or rBCG-N. After 10 d, mice received a booster with the respective BCG strain, as indicated earlier. Twenty-one days postimmunization, mice were challenged as described above. Body weight was determined daily after vaccination and during the days after infection.

Sample Collection.

Mice and rats were killed at 1, 3, 7, and 10 days postinfection, and the lungs, brain, olfactory epithelium, and blood were obtained for quantitative real-time RT-PCR and immunofluorescence analyses. Specifically, the brain was dissected for separate analyses of the olfactory bulb, midbrain, and brainstem (medulla oblongata). The samples were analyzed by real-time RT-PCR and immunofluorescence, as described in SI Materials and Methods.

MWM Test for Spatial Learning and Memory.

We performed the MWM test on 2-mo-old rats 30 d after RSV or mock challenge to assess neurological alterations in infected rats, as described in SI Materials and Methods.

MB Test for Limbic Behavior.

To evaluate the object-hiding behavior of infected and control mice, we performed the MB test as described in SI Materials and Methods.

Extracellular Field Recording and LTP Induction.

LTP and LTD assays from transverse hippocampal slides for RSV and mock rats was performed as described in SI Materials and Methods.

Inhibition of Leukocyte Transendothelial Migration.

BALB/c mice were anesthetized with ketamine/xylazine and challenged i.n. with 1 × 10⁷ pfu of RSV (mock animals received supernatants of uninfected HEp-2 cells). After infection, mice received 70 µg of anti-CD49d blocking antibody intravenously. Three days later, mice were killed, and the lungs and CNS tissues were collected for real-time PCR analyses. In addition, animals received either one dose at day 0 or two doses of anti-CD49d at day 0 and day 3 after challenge with RSV or mock. Thirty days later, mice were evaluated by MB tests.