Cellular receptors play an important role by controlling the first steps of infection, and, if they are sufficiently different in various hosts, they can provide effective species barriers. Corroborating their similarity in cell tropism and pathogenicity, various Morbilliviruses such as MV, CDV and RPV use the species orthologues of CD150 and nectin-4 as major uptake receptors 
. Our infection experiment of H358 cells with CDV-A75/17red
revealed that few passages were required for adaptation of this virus to human target cells, while the location of the detected mutations outside of H and F clearly indicated that there is no need for adaptation of H to human nectin-4. It is likely that the mutations outside of H and F are required for intracellular adaptation to human host factors, which, however, was not subject of this study and requires further investigations. Nectin-4 is highly conserved between species with almost identical extracellular domains of canine and human nectin-4. Consequently, as observed by us and others 
, no adaptive mutation is required for wild-type CDV to use human nectin-4 as receptor.
With 64% identity, canine and human CD150 are somewhat less conserved. This difference is reflected by our findings that an adaptation phase is required, and that one amino acid exchange at position 540 improves binding to the human receptor. It is astonishing how quickly the adaptation occurred (3 passages), although we started from a cloned recombinant virus with a defined sequence, which was confirmed by sequencing. This parental virus was propagated exclusively using Vero-cSLAM cells before. Given the swift adaptation, it is likely that in a sufficiently large pool of parental virus, as used for initial infection of Vero-hSLAM cells at a MOI of 1.0, the mutation emerges spontaneously. Three passages of “adaptation” are then sufficient to select for the functional H protein. Tatsuo and co-workers investigated earlier the usage of species-specific CD150 as cellular receptors for Morbilliviruses 
. They analyzed two CDV strains that were isolated using marmoset B95a cells and found that these viruses equally well were taken up by and formed syncytia on CHO cells expressing human or canine CD150. In the light of our results, it is likely that these two CDV strains adapted already during their isolation procedure to marmoset CD150, which is almost identical with human CD150, and therefore can utilize human CD150 very efficiently.
Our selection procedure reproducibly led to viruses bearing the same D540G mutation in the H gene and provided viruses fully adapted to the human receptor. That the observed adaptive mutation is required for a balanced charge distribution as consequence of the alteration in the human CD150 at position 71 appears to be conclusive. This interpretation is supported by the structural modelling data. Our modelling is based on structural analysis of wild-type MV-H interaction with CD150, which revealed the prerequisites for a functional interaction 
. As shown in , the D540G mutation confers good fusion capacity to the receptor interacting viral envelope proteins, which, however, is not completely as good as observed for the interaction of parental CDV with canine CD150. In spite of this slight difference in the cell-cell fusion, we did not observe a remarkable difference between the titres of these viruses in the single step growth curves. The titre of the adapted virus on Vero-hSLAM cells is even slightly higher at day 3 to 5 (). When syncytia become very large very quickly, membrane parts in the middle may detach, and the amount of newly synthesized virus decreases. Therefore, a slightly reduced fusogenicity and subsequent less detachment from the plastic dish may support a slight increase in virus titre. It was also interesting to observe that the adaptation of CDV-A75/17red
to human CD150 did not impair its growth capacity on Vero-cSLAM cells. Although improving the interaction with human CD150, the D540G mutation obviously does not affect the interaction of H with canine CD150. Further structural analyses may clarify this point.
Following the outbreaks of CDV infections in animal facilities in China 
, a CDV outbreak occurred in cynomolgus monkeys (Macaca fascicularis
) in Japan 2008 
. A corresponding CDV strain was isolated (CYN07-dV) and characterized, which efficiently utilized macaca CD150 as entry receptor 
. Since the amino acid sequence of Macaca mulatta
CD150 including position 70/71 is practically identical to the human sequence, one may suppose that the human CD150-positive cells should be well infected. In spite of this, the isolated virus did not very well infect Vero-hSLAM cells 
. The amino acid exchange D540G as observed by us for CDV-A75/17-adaptation was not found in CYN07-dV 
. Subtle differences between human and macaca CD150 might play a role, such as position 28 (R and H, respectively) and 49 (Y and H, respectively) and contribute to the observed effects.
Taken together, there is obviously no high hurdle for CDV to adapt and utilize human entry receptors. However, in spite of this swift adaptation, these data should not lead to the misunderstanding that other human target cells or humans will be infected so easily. It is clear that adaptation to human entry receptors is only the first step to full adaptation to human target cells, and that alterations in other viral genes are required for intracellular adaptation.