We describe here the replication features of CHIKV, a virus responsible for recent epidemic outbreaks in India and other countries in the Indian Ocean region. So far, this virus species has not been scrutinized using modern-day techniques. We have designed a panel of assays in order to follow viral replication and to describe the cellular tropism of CHIKV in culture experiments. We report that CHIKV replicates in various human adherent cells, including epithelial and endothelial cells, and primary fibroblasts and macrophages. In contrast, T and B lymphocytes and monocyte-derived DCs are not susceptible. Viral entry occurs through a pH-dependent, endocytic pathway. The life cycle of this virus is short: as soon as 8–16 h pi, numerous newly infected cells can be detected, which release high levels of progeny virions. Viral titers in supernatants reach 105
TCID50/ml, depending on the cell type. CHIKV is highly cytopathic for mammalian cells, inducing apoptosis of infected cells. Moreover, CHIKV replication is significantly inhibited by type I and II IFNs. These characteristics generally correspond to those of other alphaviruses, among which the most studied are SINV, SFV, and RRV [14
]. The alphavirus genus contains over 20 members, which likely diverged 2,000 to 3,000 years ago [40
]. Some alphaviruses are nonpathogenic in humans, whereas others cause different diseases with various intensities, and can be broadly divided into the American encephalitis viruses and the globally distributed arthritogenic viruses [2
]. Each virus type has thus likely evolved its own way of interacting with the host.
So far, alphaviruses have been mostly studied in murine and other animal cells. In particular, the interaction of CHIKV with human cells has not been characterized. We have reported here that some human epithelial and endothelial cells and fibroblasts are sensitive to CHIKV. Together with the recent finding that CHIKV replicates in human muscle satellite cells, and not in differentiated myotubes [42
], our results indicate that CHIKV displays a rather wide tropism for adherent cells. However, CHIKV neither efficiently replicated in hCMEC/D3 endothelial cells nor in A549 epithelial cells. The identification of such CHIKV-resistant cells will be useful for further studies aimed at deciphering entry or post-entry viral events. For instance, the cellular receptors for CHIKV are not known. We generally observed a direct correlation between viral binding and infection of target cells (). Our results suggest that hCMEC/D3 cells are not infected, because of the absence of surface binding receptors. The situation may be different with A549 cells, which are refractory to infection but still able to bind viral particles. A549 cells may either nonspecifically bind incoming virions (through receptors not involved in productive entry), or restrict infection at a post-binding step.
As for other alphaviruses [14
], CHIKV entry is pH-dependent. Viral replication is blocked by compounds inhibiting endosomal acidification and likely requires the clathrin machinery, as demonstrated by the requirement of Dyn-2 for viral replication. Dyn-2 also controls clathrin-independent uptake via caveolae [38
], and it will be thus of interest to examine whether caveolae are also involved in CHIKV entry. Although chloroquine blocked CHIKV replication, the therapeutic (antiviral) index of chloroquine in cell cultures is rather narrow (); thus, one should be cautious when proposing the use of chloroquine as an antiviral treatment in infected individuals.
CHIKV is highly cytopathic in human cell cultures, and infected cells rapidly undergo apoptosis. Alphavirus replication strongly affects fundamental cell physiology processes, with an inhibition of cellular transcription and translation, and a redirection of cellular resources towards the synthesis of viral proteins and genomes [45
]. Induction of apoptosis by SINV occurs at the level of cell entry, without requiring virus replication [46
]. Determining whether this is also the case for CHIKV, and whether apoptosis of CHIKV-infected cells is involved in pathogenesis, will require further investigations.
IFNs are essential components of the innate immune system, protecting against alphaviral disease [14
]. A 2005 study proposed a correlation between alphavirus virulence and resistance to type I IFNs, as reported for the Eastern equine encephalitis virus [49
]. We show here that the situation may be different for CHIKV, since this pathogenic virus retains full sensitivity to type I and II IFNs in cell cultures. This strongly suggests that the innate immune system controls the virus and is responsible for the rapid decline (a few days) of viremia observed during the acute phase of infection. It will be useful to determine which IFN-induced proteins mediate the inhibition of CHIKV replication. An interesting candidate is ISG15, which was recently shown to function as a critical antiviral molecule against SINV in mouse [50
In contrast to adherent cells, primary lymphocytes, T cell, B cell, and monocytoid lines did not allow CHIKV replication. Similar results have been described for RRV and SFV [51
]. Retroviral vectors pseudotyped with envelope glycoproteins from these two alphaviruses efficiently transduced adherent cells, but failed to infect lymphocytes and monocytes [51
]. This is probably due to the lack of adequate receptor expression in lymphocytes and monocytes. One can speculate that CHIKV receptors are also absent in these cells, as supported by the lack of virus binding in our assay. Since a high peak of viremia occurs during the acute phase of chikungunya disease (P. Laurent, K. Le Roux, P. Grivard, G. Bertil, F. Naze, et al., unpublished data), we sought to determine whether PBMCs from acutely infected individuals harbor CHIKV. We did not observe detectable levels of viral RNA in the blood cell fraction from three individuals with a plasmatic viral load ranging from 105
RNA copies/ml (not shown). Therefore, PBMCs are not sensitive to CHIKV in vitro, and are probably not infected in vivo.
Human monocyte–derived DCs were not sensitive to CHIKV replication. Multiple parameters regulate the ability of alphaviruses to infect DCs; for instance, RRV envelope glycoproteins allow infection of murine, and not human DCs [52
]. Infection of human DCs by a SINV vector is determined by a single amino acid substitution in E2 [53
]. In a mouse model, the in vivo targeting of Venezuelan equine encephalitis virus to skin DCs is required for pathogenesis, and is also regulated by amino acids in E2 [54
]. On the other hand, mosquito cell–derived RRV and Venezuelan equine encephalitis virus exhibit enhanced infection of murine myeloid DCs, compared to mammalian cell–derived preparations [55
]. This is due to a better induction of type I IFN by viruses produced in mammalian cells [55
]. In our hands, monocyte-derived DCs were insensitive to CHIKV produced in either mosquito (C6/36) or human (HeLa) cells. Similar results were observed with immature and mature DCs, and exposure of immature DCs to CHIKV (from C6/36 or HeLa cells) did not induce their maturation, nor promote type I IFN production (not shown). Our results strongly suggest that monocyte-derived DCs are intrinsically resistant to CHIKV. Further work should examine the sensitivity of other DC subsets, like plasmacytoid DCs and Langerhans cells.
We demonstrate here that in contrast to DCs, human primary macrophages are susceptible to CHIKV. Infection of macrophages is associated with release of infectious viral progeny in the supernatants. This process is less efficient than in HeLa cells, with only 5%–50% of the cell population being positive, and viral titers plateauing at 104
pfu/ml, depending on cell donors. This restricted replication may be due to the secretion of IFN or other cytokines by infected macrophages. RRV, another arthritogenic virus, also infects macrophages [41
]. Macrophages have been implicated in the pathogenesis of RRV disease, at least in a mouse model of infection. In this model, infiltrates of inflammatory macrophages are observed in muscles and joints [41
], and treatment of mice with macrophage-toxic agents abrogated symptoms [60
]. Moreover, RRV can induce persistent productive infection of macrophages in cultures [56
]. Antibody-dependent enhancement of RRV infection in monocyte/macrophage cell lines has been reported, but whether this process plays a role in vivo remains to be proven [41
There are, however, noticeable differences between RRV and CHIKV interactions with macrophages. First, RRV has been mostly studied using murine monocytic/macrophage cells, or human monocytoid cells, and to our knowledge, there is no evidence for RRV infecting primary human macrophages. Moreover, RRV [56
], and not CHIKV (B), infects monocytoid cells such as U937 cells. CHIKV induces a sudden onset of severe arthritis and fever, whereas RRV-induced symptoms are generally mild and more gradual [40
]. Whether these differences are linked to a differential interaction of the viruses with macrophages in vivo remains to be determined.
During the recent outbreak of CHIKV in La Réunion, an evolution of the viral genome was reported when comparing initial and later circulating strains. In particular, the selection of a mutant of the E1 glycoprotein (A226V) was noted [5
]. In our studies, we have used four clinical CHIKV strains with different sequences. All the variants behave similarly in our assays, suggesting that the reported evolution is not due to the selection of viruses with a modified tropism for human cells.
Overall, our results provide fresh insights into the replication of alphaviruses and the pathogenesis of CHIKV. We identified a panel of human cell types sensitive to CHIKV, and our next aim is to further characterize how CHIKV interacts with these cells. It will be of interest to determine how DCs, macrophages, and other cell types sense CHIKV, and which cytokines are produced during the encounter. It will be also useful to correlate our results with the situation in vivo, for example, by determining in biological samples from CHIK-infected individuals which cells harbor the virus during both acute and chronic phases of the disease.