Discussion
This is the first in vivo that studies the evolution of
productive RSV infection in monocytes and other immune cells during the
course of a primary acute infection. It is also the first to demonstrate
persistence of the virus in a subset of pulmonary DCs after the clinical
symptoms of infection have resolved. While Infection of the regulatory B
cells in vitro has been described previously26 this is the first
to demonstrate in vivo productive infection of a small proportion
of B-cells in an animal model. These results provide further evidence
that RSV has evolved an approach to subverting the host’s immune
response which appears to be unique amongst common respiratory viruses.
It seems likely that its ability to productively infect monocytes
contributes to the characteristic poor long-term immunity generated by
its human host and hence poor herd immunity while persistence in
pulmonary DCs following the acute infection would provide a niche in
which the virus can persist between the characteristic winter epidemics.
This model of primary RSV infection has been used extensively and,
reassuringly, the time course of the acute illness and associated
neutrophil dominated airways inflammatory response were consistent with
that seen in human infants in whom lavage samples from both the upper
and lower airways indicate an intense inflammatory response dominated by
neutrophils27,28. The neutrophil response peaked at day 5 and had
resolved by 21 days post-infection corresponding to the timing of
changes in weight, clinical symptom severity scoring and RSV viral load
in the lung.
In contrast to the neutrophil dominated inflammatory response observed
within the airways flow cytometry analysis showed that antigen
presenting cells (APC) were the dominant cells observed in the lung
tissue during the acute infection. Again, this is consistent with
findings in humans with monocytes being the most common interstitial
inflammatory cell found in postmortem specimens from infants dying of
RSV29. In contrast to the inflammatory cells recovered from the airways,
which peaked at day 5, the lung APC populations increased until day 7
and most remained elevated until 21 dpi. Both cDCs and macrophages
numbers remained significantly above baseline at 21 dpi. This is
consistent with a previous mouse study that simply assessed lung
dendritic cell numbers for 21 days after infection with APC number
remaining elevated well beyond the resolution of the acute infection30.
A particular strength of this study is that it was possible to follow
RSV replication using the fluorescent labelled RSV as fluorescence is
only produced when the virus replicates. The intensity of fluorescence
has been shown to directly mirror viral replication in a mouse model of
RSV infection31. Unsurprisingly, the data clearly demonstrated RSV
replication in lung APC populations though analysis demonstrated that
that this was bi-phasic, peaking at 5 and 21 dpi in cDC and B cells. At
21 dpi, when symptoms had completely resolved, the data suggest that RSV
replication, and hence persistence, was present in 80% of cDCs. Further
subpopulation analysis of cDC showed that CD103 expression was dominant
at 5 dpi with co-expression of CD11b observed at 21 dpi. In contrast,
pDC numbers did not change significantly and negligible RFP signal
(indicating RSV replication) was detected in pDCs. These results are in
line with a previous in vitro study which found that pDCs are
significantly less susceptible to infection that cDCs33 and a study
involving intubated infants in which analysis of BAL samples indicated
that cDCs but not pDCs were significantly elevated when compared to
samples from controls. The data obtained in this study indicates that
RSV preferentially infects and replicates in a subset of inflammatory
cDCs and, to lesser extent, B cells, and that RSV replication persisted
in these cells well after the clinical symptoms of infection had
subsided.
cDCs are known to exist in close proximity to the airway lumen13-15,34
and are likely to be exposed to virus infecting the airways epithelium.
Further studies have demonstrated a propensity for CD103 cDCs in
carrying and cross-presenting viral antigens, with a likely significant
role in induction of antibody memory responses13,35. Given the
importance of these cDCs and B cells in certain aspects of immunological
memory, productive infection of and persistence within these cells may
well prove to be a key factor in the virus’s ability to suppress
long-term production of protective antibodies levels which in turn
contributes to the annual epidemics of respiratory infection. A number
of viruses are known to infect and replicate in DCs, with a suggestion
some may have evolved to act like a ‘Trojan horse’ in order to subvert
normal immune responses36. Data generated in this and previous studies22
suggests that RSV may be one of those viruses.
The observed replication of RSV in lung cDCs over a prolonged period
extending well beyond resolution of clinical symptoms is consistent with
a previous in vitro study showing that the virus can continue to
replicate within human monocyte derived DCs for many months after the
initial infection22. After a period of several months of replication the
virus appeared to enter a latent state, which lasted for several months,
during which no fluorescence and hence replication was evident before
replication re-commence following exposure to nitric oxide22. During the
latent period RSV RNA could be identified by PCR even though it was not
replicating. In this study replication within a subset of cDCs and some
B cells was still ongoing at 21 days post infection. Further in
vivo studies with significant longer follow up periods are required to
confirm that the virus becomes latent and can be reactivated in an
animal model. Previous mouse studies have indicated that this is likely
given that in one study RSV RNA could be detected in murine lungs up to
77 days after inoculation in the absence of an evidence of replication37
while in another this period was extended to over 100 days38. Of
particular note was the observation that low levels of replicating virus
was isolated in a subset of mice who were immunosuppressed with
monoclonal antibodies to CD4 and CD8 at 150 days post infection38.
In summary, the mouse model for RSV infection used in this study
appeared to be robust in that the most severe clinical symptoms
coincided with the peak in RSV viral load and airways neutrophilia as
has been observed in human. As has been observed in vitro , the
virus infected and replicated in antigen presenting cells and was shown
to occur predominantly in CD103+ cDCs and, to a lesser
extent, B cells. Importantly, persisting RSV replication was observed in
the majority of lung CD103+CD11b+cDCs beyond resolution of the clinical infection and associated acute
airways inflammation. Understanding how RSV activity is able to persist
in CD103+CD11b+ cDCs without causing
an inflammatory response indicative of viral infection, and the role of
this subpopulation in antibody production, are likely to assist in novel
therapeutic approaches and effective vaccines.