Nipah virus (NiV) and Hendra virus (HeV)- The first confirmation that a G-F protein complex was required for henipavirus entry was demonstrated by Bossart et al [
69], where cell lines were transfected with plasmids encoding for HeV F-protein, G-protein, or a combination of both. It was observed that syncytial formation (i.e. the multinucleated cells) only occurred when both plasmids were transfected. Furthermore, recombinant vaccinia virus expressing these 3 constructs also showed that syncytial formation in HeLa cells required both F- and G-proteins [
69]. This study also demonstrated that syncytium formation could be observed using cells from bats, horses, cats, pigs, rabbits, monkeys, mice and ducks indicating a broad host spectrum.
EphrinB2 and –B3 Are Host Receptors for NiV and HeV
Upon the observation that syncytium formation was detected after transient expression of F- and G-proteins, experiments to determine the precise host-receptors that interact with HeV and NiV envelope proteins were performed by Bonaparte et al [
71]. Having previously observed that a HeLa cell-line derivate (HeLa-USU) was non-permissive to syncytium formation following NiV and HeV expression, this cell-line was compared to NiV and HeV permissive cell-lines by microarray analysis; these included HeLa-CCL2, the human glioblastoma cell line U373, and the human head and neck carcinoma PCI-13. Potential gene candidates were identified based on known or predictive plasma membrane surface-expressed proteins. Of the 10 that were identified, ephrinB2 (EFNB2) was found to permit NiV and HeV infection, based on the ability of EFNB2 to bind to soluble HeV and NiV glycoprotein in ELISA assays, as well as permitting NiV and HeV infection when expressed in HeLa-USU cells [
71].
Experiments using the closely related ephrin-B1 receptor demonstrated that this receptor did not mediate entry into entry of NiV F/G pseudotype viruses [
72], indicating receptor specificity. However, following studies using recombinant proteins encoding for all known ephrins (i.e. ephrinA1-A5 and ephrinB1-B3) found that ephrinB2 and ephrinB3 interacted with NiV-G, as detected by surface plasmon resonance [
73]. Furthermore, experiments using pseudotype VSV (expressing NiV envelope proteins) showed that soluble ephrinB2 and ephrinB3 competitively inhibited interactions with CHO cells expressing either ephrinB2 or ephrinB3, suggesting that NiV-G interacts with both receptors through an overlapping site [
73]. Mutational analysis further located this site to two residues, Leu124-Trp125, located in the G-H loop of both ephrinB2 and ephrinB3 that are critical for NiV binding and entry [
73].
Guillaume et al provided further insight into the critical residues of NiV G which were required for binding to ephrinB2 [
74]. In that study, plasmids encoding for NiV G were constructed with point mutations of fifteen charged residues within the globular head of NiV G, and then analyzed in cell fusion assays. Mutation of seven residues, W504, E505, N557, Q530, T531, A532, and E533, reduced fusion, with the E533 mutation being responsible for the highest decrease in fusion capacity. The critical relevance of E533 for receptor binding was further supported by the finding that E533 was replaced in an escape mutant.
Studies using soluble NiV and HeV G proteins found that while both proteins interacted with ephrinB2 with similar affinities, soluble NiV G protein bound to ephrinB3 with an approximately 30-fold higher affinity than HeV G [
74]. As the globular domain of both proteins differ only at amino acid residue 507 (i.e. HeV Ser507, NiV Thr507), a recombinant soluble HeV G protein was constructed, substituting Ser507 for Thr507 (i.e. S507T). This mutation conferred HeV G-binding affinity to ephrinB3 to a level comparable to NiV G [
75]. As these results suggested that Thr507 plays an important role in the affinity of NiV G for ephrinB3, additional experiments were performed using NiV G with amino acid substitutions previously implicated by Guillaume et al [
73] for ephrinB2. While most of the mutants (with the notable exception of E533Q) did not show decreased affinity for ephrinB2, all of the assayed mutants did show decreased affinity for ephrinB3, suggesting that the interaction of NiV G with both receptors is distinct [
75].
As ephrinB2 is expressed on endothelial cells, neurons, and smooth muscle cells [
76,
77], while ephrinB3 is expressed in the central nervous system [
78], the tissue distribution of these molecules overlaps strongly with the organ tropism associated with NiV infection.
Specificity of other Henipavirus Envelope Proteins in Cellular Tropism
M74 (otherwise known as Ghana virus)- While NiV and HeV are the most extensively studied, the characterization of other henipavirus genomes have led to the study of cell tropism, using a strategy of expressing the putatively encoded envelope proteins in cells. An example of this was described by Krüger et al [
79], where the envelope proteins of M74 were cloned into an expression plasmid and assayed for syncytial formation. When transiently expressed in BHK 21, Vero76 and HypNi/1.1 cells (a kidney cell line derived from the hammer-headed fruit bat Hypsignathus monstrosus), syncytium formation was only observed in HypNi/1.1 cells; although it should be noted that the number of nuclei per syncytium was considerably less than that observed for HeV and NiV. This suggested that M74 has a narrower range of cell tropism compared to HeV and NiV. However, this same study also demonstrated that M74 G protein can interact with ephrinB2, and that syncytium formation was pH dependent most likely due to the requirement of an acidic pH in the endosomal compartment for henipavirus fusion protein proteolytic cleavage [
80,
81,
82]. These findings suggested that despite its more limited range of cell tropism, M74 envelope proteins may interact with the host cell in a manner similar to its NiV and HeV counterparts. These findings were further expanded upon, with another study on the host-restrictive properties of M74 envelope proteins [
83]. In this study, two other cell lines derived from bats (i.e. chiropteran) were transfected with plasmids encoding for M74 F and G proteins, with all chiropteran cell lines showing increased surface expression of G-proteins (compared to non-chiropteran cell lines). Interestingly, when the G proteins of both M74 and NiV were expressed in the same chiropteran cell lines, NiV G protein exhibited a significantly greater level of surface expression compared M74 G protein, suggesting that these differences in surface expression may explain the more limited syncytium formation observed by M74 envelope proteins [
83].
The host cell specificity associated with M74 envelope proteins was also expanded upon by a study from Weis et al [
84]. Here, heterotypic expression of the envelope protein (i.e. the co-expression of NiV G with M74 F-protein) in Vero cells did not result in syncytium formation; however, when M74 G-protein was co-expressed with NiV F-protein, syncytium formation was observed (albeit with a reduced number of nuclei/syncytium), thereby suggesting that the limited host range of M74 may be due to the F protein. Furthermore, when pulse-chase experiments were performed on both M74 and NiV F proteins expressed in MDCK cells, it was found that relatively little M74 F protein was cleaved into its biologically active form (i.e. F
1 and F
2) compared to NiV, suggesting a possible mechanism behind M74 F protein’s limited host range [
84].
These results were also confirmed by Lawrence et al [
85] when comparing M74 and NiV envelope proteins in another chiropteran cell line EidNi (a kidney cell line derived from the straw-colored fruit bat Eidolon helvum). As was the case in the study of Weis et al, M74 envelope proteins induced syncytium formation in EidNi, but not in 293T, HeLa or Vero cells. Heterotypic assays using combinations of NiV and M74 G- and F-proteins also yielded identical results to those presented by Weis et al.
Mojiang virus (MojV)- Rissanen et al demonstrated that MojV G protein is antigenically distinct from NiV and HeV; however, recombinant expression of MojV F and G protein did induce syncytium formation in human cells lines A549, U87 cells, along BHK and HEK293T cells [
86]. This same study also demonstrated (through heterotypic expression of MojV F and G proteins with their NiV counterparts) that MojV envelope proteins do not interact with human ephrinB2 or B3.
Cedar virus (CedV)- A study by Pryce et al using HEK293T cells found that CedV G protein displayed high affinity for human ephrin B2, while showing little affinity for ephrin B3. More interestingly, CedV G protein also displayed an affinity for human ephrin B1 [
87]. These results were further confirmed when CHO cells expressing recombinant ephrin B1 or B2 enabled entry of CedV pseudotyped virus, while CHO cells expressing ephrin B3 did not [
87]. Furthermore, Laing et al demonstrated that CedV glycoproteins displayed a unique affinity for ephrin receptors [
88]. Using cell-cell fusion assays between CHO745 cells expressing CedV F/G proteins and ephrin-expressing target cells, they found that fusion could be observed with target cells expressing ephrinA1, A2, A5, B1 and B2 (with the latter two showing the greatest degree of fusion). It should also be noted that CedV glycoproteins had species specific affinity for mouse ephrin A1 [
88].