RSV Takes Control of Neonatal Breg Cells: Two Hands on the Wheel

Abstract

The viral attachment protein of RSV has many surprising features, especially its mimicry of fractalkine (CX3CL1). Zhivaki et al. (2017) now show that, in addition to using this homology to attach to ciliated cells, it activates human neonatal regulatory B cells, thereby inhibiting immunological responses.

Respiratory syncytial virus (RSV) is common pathogen that infects the ciliated respiratory epithelium, causing variable disease that ranges from severe infantile bronchiolitis to a mild (or unapparent) common cold. It infects virtually all children by the age of 3 and repeatedly re-infects them throughout life. It has two major surface proteins, one mediating fusion (F) and the other attachment (G). The latter has a structure similar to the chemokine fractalkine (CX3CL1), binding CX3CR1 on the apical surface of ciliated cells in the lung. In a new twist, Zhivaki et al. (2017) now show that RSV G also binds to a novel subset of neonatal regulatory B (nBreg) cells that make the immunosuppressive cytokine interleukin 10 (IL-10), thus inhibiting protective immune responses.

Rather than depending on rapid evolution to escape immune pressure, the success of RSV seems to depend on a partial but specific immunological amnesia (Openshaw et al., 2017). RSV’s attachment protein G has few parallels in the viral kingdom. It is known to inhibit Toll-like receptor (TLR)-induced type I interferon host responses. One of the several interesting properties of RSV G is that vaccinating mice with it can induce enhanced and eosinophilic lung disease during subsequent RSV infection, a phenomenon that requires an intact central conserved domain (Sparer et al., 1998).

When the structure of CX3CL1 was first described in 1997, RSV aficionados immediately appreciated its close resemblance to RSV G. Both G and CX3CL1 have heparin binding domains, and both have membrane bound and soluble forms and both have “mucinoid” proximal regions on extended serine-threonine rich stalk that ends in a cysteine-rich chemokine domain (Melero et al., 2017). News of this remarkable similarity spread quickly in the RSV community and by 2001, Tripp et al. (2001) showed that the attachment protein G does indeed bind to the receptor CX3CR1. RSV exploits this property to infect the respiratory epithelium by binding of G to CX3CR1 on ciliated cells (Jeong et al., 2015; Johnson et al., 2015), although it remains a mystery why CX3CL1 should be present at this site.

CX3CL1 is chemoattractive for both lymphocytes and monocytes and is normally made by activated endothelial cells. Its receptor is naturally present on cells with high cytotoxic potential (such as natural killer [NK] cells, cytotoxic T cells, and γδ T cells). Presumably, the soluble form of RSV’s attachment protein recruits such cells into sites of infection, but the benefit to the virus of such an effect is not clear: RSV bronchiolitis is characterized by an abundance of inflammation, but does cell ingress benefit the virus? Again, a mystery.

Seeking to better understand why RSV has such a special ability to infect children early and often, Zhivaki et al. (2017) focused on regulatory B (Breg) cells. These cells have been recently described as an immunosuppressive subset of B cells (Rosser and Mauri, 2015). Their functions have been well characterized in mice, in which they promote tolerance via the production of IL-10, IL-35, and TGF-β and thereby inhibit inflammation by actions on monocytes, IL-12-producing dendritic cells, invariant NKT cells, T helper 17 (Th17) cells, Th1 cells, and cytotoxic CD8⁺ T cells. Breg cells can also induce the differentiation of regulatory T cells, Foxp3⁺ T cells, and T regulatory 1 (Tr1) cells.

Zhivaki et al. (2017) now identify a novel subset of CD5^hi B cells in neonates that they term neonatal Breg (nBreg) cells, which they have not been able to identify in adults. They show that the RSV fusion protein (F) binds to the nBreg cells via polyreactive (IGHJ4/D6) cell surface immunoglobulin M (IgM), causing upregulation of CX3CR1 on the nBreg cells and making them a target for engagement of G. This double-handed grip leads to non-lytic but activating RSV infection and the production of IL-10, which in turn inhibits Th1 and Th0 cells that would normally have an antiviral effect (Figure 1).

Figure 1. RSV Targets Neonatal Regulatory B Cells.

RSV fusion protein engages polyreactive IgM on neonatal regulatory B (nBreg) cells. This causes increased expression of CX3CR1, which is also expressed on resting ciliated respiratory epithelial cells and monocytes. CX3CR1 is the natural ligand for fractalkine (CX3CL1), but also binds the RSV attachment protein G. This engagement leads to non-lytic infection of nBreg cells, cell activation, and IL-10 production.

Looking for nBreg cells in nasopharyngeal aspirates (NPAs) of hospitalized RSV-infected children requiring respiratory assistance, Zhivaki et al. (2017) found that in 6 out of the 13 children, RSV-infected nBreg cells were found in NPAs, whereas two patients showed RSV-infected mature naive B cells (Zhivaki et al., 2017). The frequency of nBreg cells correlated with the severity of the disease as assessed by the duration of oxygen support and hospitalization in the intensive care unit. They also found a higher frequency of nBreg cells in the blood of RSV-infected children suffering from acute bronchiolitis compared to non-infected children and a positive correlation between the percentage of nBreg cells with the disease severity and the RSV viral load. The frequency of CXCR3⁺ Tem cells was significantly lower in the peripheral blood when nBreg cells were infected in the NPA. The proposal that RSV infects B cells is given added plausibility by the finding that RSV persistently and non-lytically infects bovine B cells after experimental infection of young calves (Valarcher et al., 2001).

Although the observations of Zhivaki et al. (2017) are intriguing and create an attractive model that gives new depth to our understanding of RSV immunopathogenesis, there are questions left open. Are the observed changes in Breg cells caused by RSV infection, or are they an effect of disease? Zhivaki et al. (2017) find that nBreg cells wane rapidly in the postnatal period, so what happens in older children or adults? If nBreg cells play an essential role in immune evasion by RSV, how is it that older children and adults also show a deficit in B cell memory to RSV infection? Clearly studies of cord blood (or of peripheral blood) may not fully reveal what happens in relevant tissue sites: accessing those sites remains a major challenge and there are yet more mysteries to solve.

In conclusion, Zhivaki et al. (2017) show that B cells have a more complex and interesting role in RSV disease than was hitherto appreciated. They identify a novel subset of neonatal regulatory B cells that become infected with RSV and may inhibit many of the responses that normally play a part in restricting viral inflammation, favoring viral replication and possibly inhibiting immunological memory, thus allowing reinfection. If this hypothesis is correct, nBreg cells may represent a novel target for the manipulation of lower respiratory tract viral infections and their pathological consequences later in life. At the dawn of RSV vaccination, this is a new and exciting twist in understanding immune responses to this fascinating virus.