A mutation-mediated host shift drives Avian paramyxovirus type 1 evolution

PPMV-1, an antigenic variant of APMV-1, associated with specific pigeon host species. However, its evolutionary strategy and underlying drivers of host specificity remain unknown. In this study, we collect the outbreak data on a global scale to investigate its evolutionary dynamics, and provide an evidence-supported analysis the host shift of PPMV-1 from chickens to pigeons, and this shift is driven by the P protein. Our data indicated that the viruses in the United States and China have undergone convergent evolution. We find that three mutations of P protein, especially R163G, can significantly affect the adaptation of APMV-1 in pigeons. Mechanistically, sensor LSm14A inhibits the replication APMV-1 in DF-1 cells, and R163G substitutionon P protein increase LSm14A degradation. We propose the host shift drive the evolution of PPMV-1 and the underlying mechanism, offering new insights into the adaptive evolutionary process of the virus.

1 in chickens result in host adaptation driven by mutations in the polymerase complex, 81 suggesting that changes in this complex can drive host shifts (Dortmans et al, 2011;82 Olszewska-Tomczyk et al, 2018). Paramyxovirus P protein is essential for viral RNA 83 synthesis and other biological processes (Lamb and Kolakofsky 1996). The V protein 84 affects the host range of the virus via its species-specific IFN antagonist activity (Park 85 et al, 2003). Therefore, the role of P protein in host shift appears to be particularly 86 important, yet it has been rarely addressed in previous studies. 87 In the present study, we collect global data and used a phylogeographic Bayesian  its introduction from the APMV-1 through the three mutations on P protein. 205 We next focused on the three P protein amino acid substitutions of the virus in China.

206
A phylogenetic analysis based on the complete P protein was conducted and found that 207 Site 93 settled as K after multiple changes (T, N, R), Site 136 changed from W to R/K, 208 while site 163 changed from R/K to G (supplementary fig. 5). Convergent evolution 209 can be used to differentiate adaptation from neutral genetic variation on the basis of 210 sequence data (Pepin et al, 2010). To verify whether the virus has undergone adaptive 211 evolution, a convergent evolution analysis was performed in the United States. We 212 found the three mutations are similar to those in China (I93K, R163G, beisides, 136R 213 is consistent with the mutated site of the Chinese strain) (supplementary fig. 6).

215
Since R163G is the most sufficient site to improve the replication capacity in PEF cells, 216 we thus compared the contact transmission potential of APMV-1, APMV-1 R163G, and 217 PPMV-1 viruses in 4 weeks old pigeons (the PPMV-1 isolate was used as positive 218 control). The viral load of APMV-1 R163G group in the ranges of 1.15-3.91 lg TCID50, 219 higher than those inoculated with APMV-1, with the viral load in the ranges of 1.25-220 2.83 (figs. 6B and 6C). In contact pigeons, APMV-1 R163G viral load in the ranges of 221 0.34-1.23 lg TCID50, higher than those exposed to APMV-1, with viral load in the 222 ranges of 0.15-0.25 lg TCID50 (figs 6B and 6C). These results suggest that APMV-1 223 R163G virus showed better replication in inoculated donors and transmitted more 224 rapidly to contact pigeons than the APMV-1 wild type isolates.

258
The PEF cells were isolated from 10-day-old pigeon embryos, and maintained in streptomycin.

262
The complete F gene (1662bp) sequences from PPMV-1 isolates were downloaded 263 from Genebank and aligned using MEGA version 7. We selected the best fitting model  Sampling/Stepping-stone analysis in order to estimate the best-fitting demographic 287 model for our dataset. We then used BF comparisons from the resulting marginal-288 likelihood estimates to select among the models. We found that the uncorrelated clock 289 with constant model provided the best fit to our data (Supplementary Table1). 290 We inferred the global and within-China geographic origins of PPMV-1 and its 291 significant dispersal routes between affected countries or regions using discrete-state 292 ancestral reconstruction methods implemented in BEAST SPREAD3 version 0.9.6. We 293 used BF > 3 as the significance threshold.

298
To model transmission between host species, chicken, pigeon, and wild birds were used 299 as discrete traits. We combined the best fitting coalescent tree model and branch-rate 300 prior as described above. The BSSVS method was used to identify significant migration 301 routes and their directionality between hosts. Using the Markov-jump (MJ) method, the 302 intensity of backward and forward transitions within discrete trait matrices was inferred 303 as a proxy for the mean number of viral jumps between hosts.          The significance thresholds above the red line indicate significant associations.