Discussion
In the present study, we conducted both in vitro and in vivo experiments that characterize immunomodulatory and antiviral effects of a CpG-A molecule in the domestic cat. We show that ODN 2216, the first described CpG-A[12], can upregulate the expression of a series of genes in feline immune cells that play important roles in early responses to viruses. Soluble molecules produced by feline PBMCs upon ODN 2216 stimulation significantly increased resistance of various feline target cell lines to propagation of viruses from five different families, namely FCV, FPV, FHV, FCoV and FeLV. The observed repression of viral replication highly correlated with the mRNA expression of type I IFN genes in stimulated PBMCs as well as the induction, prior to inoculation, of antiviral mechanisms in the target cells. Furthermore, a single administration of ODN 2216 significantly increased the expression of Mx in the blood of treated cats. Plasma from these animals could also inhibit replication of FCV in vitro. Our data underline the prophylactic potential of ODN 2216 in an outbred species as a stand-alone agent and against a large range of viral pathogens simultaneously. Cats may highly benefit from such a molecule when placed in environments with strong infectious pressure such as catteries, shelters or pet shows.
Cells of the feline immune system were strongly influenced when cultured in the presence of ODN 2216. First, feline PBMCs significantly proliferated in presence of ODN 2216, indicating a direct and /or indirect stimulation of one or more immune cell subpopulations by this molecule. Although in contrast to other classes of ODN CpG-A does not generally induce lymphocyte proliferation in mice[12], similar effects have previously been observed in ovine cells, where higher concentrations of CpG-A could induce minimal levels of proliferation[68]. Additionally, ODN 2216 increased the expression of cell surface co-stimulatory molecules in PBMCs, reinforcing possible interactions between various immune cell populations in subsequent specific responses. While B7.1 molecules were upregulated on both lymphocytic and non-lymphocytic cells after ODN 2216 stimulation, MHCII expression could only be increased on lymphocytic cells. This observation may be linked to the time point of our measurements, or to the necessity of transport to the cell surface of MHCII molecules coupled with antigen for presentation to other immune cells. Finally, the transcription of a series of markers of innate immune responses was considerably influenced in feline PBMCs by stimulation with ODN 2216. The most astonishing effects of this molecule thereby remain the potent induction of IFNα and IFNω, with mRNA expression of these genes increased by up to 12 000 and 35 000-fold respectively in PBMCs of certain animals. In line with observations published recently, the induction of IFNγ by this CpG ODN remained moderate[34]; nevertheless, the measurements carried out 24 h after stimulation of purified feline cells support induction of Th1-oriented immune responses and enhanced NK cell activity, both highly desired in the contexts of viral infection.
Variability in the immunomodulatory and antiviral effects of ODN 2216 on feline cells was mainly related to the age of the cats. Among middle-aged adult cats, consistency in the stimulatory potential of this molecule was reflected through the highly similar patterns in gene expression profiles induced in immune cells after stimulation, as well as through the moderate deviations in those experiments aiming to characterize the immunomodulatory properties of ODN 2216 (Figure1). The cells of several animals however, in particular cat c08, but also c09 and c12, indicated particularly strong responsiveness to stimulation throughout the study. Such observations are most likely linked to the genetic variability of individuals of an outbred species, as inherited factors are known to play an important role in the magnitude of innate immune responses[69]. With respect to this observation, it should be noted that the SPF origin and maintenance in a barrier facility of the animals included in this study may slightly lower the variability in strength and breadth of innate immune responses, and studies with cells obtained from field cats would give further information on divergence in response to stimulation of the innate immune cells of individual animals. The immune cells of cats of 14 years of age seemed to respond slightly less well to ODN 2216 than those of younger adult animals, while PBMCs of kittens indicated much more reticence to stimulation, with either limited or absent upregulation of both type I IFN and other genes measured after incubation with ODN 2216. Stimulated PBMCs from this group of young animals moreover indicated a tendency to develop an immunologic environment with a Th2 orientation, including higher production of IL-4 and impaired induction of IL-12 compared to cells from adult cats. Although the kittens included in this study were already of 10 weeks of age, these observations strongly corroborate the immature IFN system of neonatal mice[70], the impaired immune cell activation via TLR9 in human neonatal mononuclear cells[71, 72] and the bias towards Th2 rather than Th1 responses in human fetuses and neonates[73, 74]. Furthermore, concordantly to findings in human neonatal blood[75, 76], basal TLR9 transcription in kittens indicated levels similar to those of adults. However, ODN 2216 stimulation increased mRNA expression of this gene only in the young animals (Figure2C). These results open new perspectives on possible explanations for the qualitative discrepancy between innate immune responses in newborns and adults. The availability of specific antibodies to feline TLR9 would greatly support further investigations in this direction.
The experiments conducted both in vitro and in vivo in this study underline the feasibility of utilizing Mx as a marker for induction of resistance to viral infections in the domestic cat. Admittedly, both the supernatants of PBMCs stimulated with CpG-A and serum of treated cats contain a mixture of molecules that could play a role in the observed antiviral effects. However, the viral inhibition in our in vitro assays was similar to that obtained after treatment of the cells with rfeIFNα and highly correlated with the induction of Mx transcription in the cell lines incubated with the supernatants. After ODN 2216 injection in vivo, Mx expression in plasma also strongly correlated with its antiviral potential. It is however important to note that Mx mRNA and protein cannot be mechanistically linked to inhibition of the viruses tested. Although the expression of other antiviral molecules such as 2'5'oligoadenylate synthetase (OAS) and the RNA-dependent protein kinase (PKR) was not measured, their induction has been reported following CpG ODN stimulation of PBMCs in other species[77, 78] and differential interplay between several effector mechanisms most likely leads to the inhibition of individual viruses.
In comparison to ODN of other groups, CpG-A have been used in much fewer studies concerning innate antiviral responses. This is most likely the result of the more time consuming and costly manufacturing process of these molecules linked to the synthesis of the flanking poly (G) strings necessary for optimal immune effects[31]. Also, CpG-A exhibit only weak stimulation of B cells, which produce antibodies that are valuable in later antiviral responses. Stability of CpG-A in vivo has also been questioned due to the only partial phosphorothioate backbone of these molecules. This can be partially compensated however, by the formation, through association of their poly (G) stretches, of highly ordered G-tetrad structures with enhanced stability[12]. Their smaller content in synthetic backbones could moreover help reduce possible long-term side effects[79, 80]. Most importantly, CpG-A remain the most potent inducers of type I IFN, themselves the most biologically active antiviral molecules known to date[19]. With regard to this, a recombinant feline type I IFN marketed in both Japan (Intercat®) and Europe (Virbagen Omega®) has made its way into therapeutic protocols for FCV, FHV, FeLV and canine parvovirus infections[67, 81–86] and has demonstrated preventive capacities in a cattery developing an outbreak of FPV[87]. However, when compared to direct initiation of antiviral mechanisms by a recombinant IFNα protein, administration of CpG ODN holds the advantage of inducing the production of all type I IFN and their subtypes, which have been shown to possess differential antiviral properties and kinetics[88]. Our data demonstrate that five viral species belonging to the Calicivirus, Herpesvirus, Parvovirus, Coronavirus and Retrovirus families were highly sensitive to the immunologic effects of ODN 2216 in feline cells. Strength and breadth of antiviral defense mechanisms induced in vivo by this molecule have the potential to outweigh those observed with recombinant IFN.
In frame with the results of our in vitro experiments, the administration of ODN 2216 in vivo could promote an antiviral state in the domestic cat in the absence of adverse reactions. Just as stimulated PBMCs exhibited high mRNA expression of type I IFN and only a marginal increase in mRNA levels of the proinflammatory cytokines IL-6 and TNFα, treated cats showed significantly increased Mx expression in their blood and lacked flu-like symptoms generally linked with administration of immunostimulatory molecules in vivo. Additionally, although our in vivo studies included a limited number of cats, the strength of responses to ODN 2216 stimulation in vitro and in vivo for individual animals strongly correlated, indicating that individual sensitivity to stimulation most likely can be predicted with in vitro experiments. Our data further show that a single subcutaneous injection of ODN 2216 sufficed to induce the systemic presence of antiviral molecules for 24 to 36 h in the plasma of treated cats; during this time, an increase of Mx expression of at least 4-fold in the blood of treated cats was indicative of significant inhibition of viral replication in vitro. These observations support possible resistance to viral infection by ODN 2216 for several days, a phenomenon described already in mouse models[89].
Altogether, this study underlines the strong potential for ODN 2216 in the prevention of a large variety of viral diseases. CpG-A molecules should thus not be left aside when stimulating innate immunity for prophylactic purposes. Future studies should aim at optimizing the kinetics of these molecules in vivo, while addressing safety, stability and manufacturing issues. Combinations with adjuvants or other TLR agonists as well as optimized administration protocols should support research in this direction. Field studies with larger cohorts of animals will provide further insights on the utilization of CpG-A molecules for the prevention of viral infections in a clinical setting.