RSV elicits the production of a wide variety of cytokines and chemokines during and after infection [6, 7, 20–23]. There is now mounting evidence that the type of immune response to RSV, specifically the spectrum of host cytokine and chemokine production, ultimately affects disease pathogenesis and chronicity . Cytokine expression likely affects the balance between viral clearance and progression of disease. Therefore, there is a critical need to identify viral factors that regulate cytokine expression during RSV infection .
The vast majority of studies into RSV pathogenesis have used laboratory strains or recombinant virus derived from these laboratory strains. These laboratory strains were isolated decades ago and the passage histories of these strains are not known or are poorly documented. Several studies suggest that sole investigation of laboratory strains may not be ideal to define and characterize RSV pathogenesis. For example, Line 19, a potential vaccine strain derived from a subgroup A isolate possesses biological features that differ from the reference strain A2 (isolated in Australia in 1961) . These phenotypic differences have been mapped to polymorphisms in the RSV F gene . Reference strain A2 differs from several clinical isolates in its ability to replicate on primary human bronchial epithelial cells and to induce interferon-inducible protein 10 and CCL5 suggesting that A2 alone may not be the ideal strain to study RSV pathogenesis . Recently, Stokes et al. demonstrated that several clinical isolates differ in their capacity to induce airway mucous production and virulence in mice . Therefore, the study of clinical isolates will likely yield important insights into the fundamental aspects of pathogenesis.
At least 2 RSV proteins have been implicated in the induction of IL-6, RSV F and G. In monocytes, RSV F stimulates IL-6 through interactions with CD14+/TLR4 complex by promoting nuclear translocation of NF-ĸB . IL-6, along with other cytokines, such as IL-1β and IL-8, play an important role in neutrophil and macrophage chemotaxis. The cellular inflammatory response during severe RSV infection is characterized by a preponderance of neutrophils and macrophages, suggesting that the expression of IL-6 at the site of infection is an essential feature in disease progression . Differences in IL-6 induction may explain, in part, differences in severity of illness seen with RSV infection. Indeed, clinical studies of both infants and adults have demonstrated that single nucleotide polymorphisms in the IL-6 promoter (position −174) were associated with a greater severity of illness suggesting a pivotal role of IL-6 in RSV pathogenesis [30, 31]. Our data suggests that variability in the induction of IL-6 may, at least, partially explain differences in the observed severity of human RSV disease associated with infection due to specific RSV genotypes [9, 12–15]. The RSV F protein of NH1067 and NH1125 differs at only 4 amino acid residues (positions 67, 292, 490 and 529). It did not appear that these amino acid changes can account for the differences in induction of IL-6 observed because the specific amino acid residues at these positions did not correlate to induction phenotype when other strains were examined.
Several lines of evidence indicate that the RSV G glycoprotein also plays a role in pathogenesis and can induce IL-6 among other cytokines [32–37]. Monocyte stimulation experiments using recombinant viruses with altered or absent G gene and synthetic peptides, corresponding to regions of the RSV G protein, revealed that portions of the highly conserved cysteine-rich region, specifically amino acids 164–176, inhibit the innate immune response . The CX3C motif (amino acid 182–186) of the RSV G glycoprotein mimics the CX3C chemokine fractalkine, competes for the chemokine-specific receptor and interferes with fractalkine-mediated leukocyte chemotaxis . RSV deficient in the soluble form of G (generated by initiation at the second AUG corresponding to amino acid 48) induced a pro-inflammatory response in A549 cells . Amino acids 164–176 and 182–186 were identical in NH1067 and NH1125. Both strains contained a methionine at position 48. Therefore, these regions of the RSV G gene, although likely important for cytokine induction pathways, cannot account for the differences observed in IL-6 induction between NH1067 and NH1125.
Our data suggest that the differences in the RSV F protein between NH1067 and NH1125 are unlikely to account for the differences in cytokine induction phenotype strongly suggesting that other viral factors, perhaps in other domains of the G protein, may play a role in the differences in cytokine induction observed in the clinical isolates. These factors may not necessarily be viral proteins. A single nucleotide substitution in the transcriptional start signal of the M2 gene, within a non-coding region of the RSV genome  is the major determinant of the temperature-sensitive, attenuated phenotype of a potential vaccine candidate. However, the consensus transcriptional start and stop signals for each gene were identical between NH1067 and NH1125. It appears as though viral transcriptional and/or replication activity rather than binding of virion-associated glycoproteins to cellular receptors is required as UV-inactivated virus failed to induce IL-6, an observation that has been reported elsewhere [40, 41].
RSV causes a wide-spectrum of disease and this may be due, in part, to the differences in biological properties of the infecting virus. This phenomenon has been observed with other viruses. For example, mutations in the HCV NS5A gene are associated with a sustained virological response to interferon therapy . A neurotropic variant of HIV, containing a mutation in the envelope gene, has been identified, which is present at a high frequency in brain tissue in AIDS patients with dementia . Mutations in several genes may account for the increased virulence of the 1918 influenza pandemic strain . Identification of the viral gene(s) involved in the induction of key cytokines, chemokines and inflammatory mediators would potentially lead to targeted antiviral therapies and would be a significant advance in the study of RSV pathogenesis.