4 Discussion

Overall, Mammalia, Aves, and Reptilia formed a distinct cluster, indicating the TMPRSS2 was more likely to be used for recognizing and processing S-protein of SARS-CoV-2 in Aves and Reptilia rather than Amphibia and fish. Human, pangolin, cat, dog and hamster are proved to be be infected by SARS-CoV-2 in previous studies[5,6,7,8,9,10,11], especially for human, the TMPRSS2 was shown to be used for priming S-protein on ACE2 to help SARS-CoV-2 to enter into cells[13]. We then focused on these species to predict the utilizing capability of TMPRSS2 for S-protein priming. Primates closely clustered together with human (Fig S1, Fig S2), indicating the TMPRSS2 could be effectively used in primates for S-protein priming. Cat, dog clustered with bat (Fig 2), which was inconsistent with the phylogenetic result of ACE2[14], and also showed the high possibility of TMPRSS2 utilizing capability. The brown bear was placed with pangolin, bat and dog forming a distinct clade (Fig S1), showing yet another possible wild animal of high probability to use TMPRSS2 to cleave S-protein of SARS-CoV-2. Considering the lack of direct contact with the human, brown bear was less likely to be the intermediate host to transfer the virus to human. Interestingly, the TMPRSS2 of horse closely clustered with hamster, but not with its much closer relatives (Fig S1), showing a potentially effective function of TMPRSS2 in horse to process the S-protein for priming. When focused on animals that have a close relationship with humans; the pig, sheep, cattle and pangolin formed a distinct clade (Fig 2). Although the incubation test showed the pig was not susceptible with SARS-CoV-2 infection[7], our result indicated the probable availability of TMPRSS2 for S-protein priming in the pig. A previous study showed that the SARS-CoV-2 was not able to use ACE2 to enter into the mouse cell[4]. In our study, however, rat and mouse TMPRSS2 were closely clustered with human and hamster TMPRSS2, showing that the TMPRSS2 in mouse and rat should be able to activate the S-protein of SARS-CoV-2.

ACE2 can tolerate a variety of amino acid changes among animals, making the low species barrier of SARS-CoV-2[20]. We expected the same situation for TMPRSS2, so we further used the PolyPhen-2 to evaluate the possible impact on the function of TMPRSS2 at single amino level, but not at the whole sequence level, because many amino acid changes will not cause change in function. By scoring the TMPRSS2 of all 164 species, we found that the prediction result was highly similar to prediction from the phylogenetic tree, with Mammalia having the highest scores, and then Aves, Reptilia, Amphibia and fish. However, the most important three amines (His296, Ser441 and Asp435)[21] located at the catalytic triad of TMPRSS2 were ultra-conserved (Fig 1, Fig S4). Only four out of 164 species were found to have mutations at these three amines, indicating a conserved utilizing capability of TMPRSS2 across species, although other amino acid changes may also influence the efficiency of TMPRSS2 to cleave the S-protein of SARS-CoV-2. Animals that have been proven to be affected by the SARS-CoV-2, like pangolin, cat, dog, and hamster, exhibited very high scores (Table 1), which showed the rationality of this scoring method. Both high score and less damaging amino acid changes were found in rat, mouse and pig than pangolin and cat, which in turn supported the prediction of the phylogenetic tree, showing the effective utilizing capability of TMPRSS2 in rat, mouse and pig. In addition, both the phylogenetic tree and PolyPhen-2 scores showed relatively low TMPRSS2 capability in poultry.

COVID19 is still progressing and SARS-COV-2 strains are constantly evolving. It needs to be emphasized again that we need to pay more attention to mammals, especially pigs and rats, to prevent these animals from becoming intermediate hosts of a future pandemic, considering the more effective utilizing capability of TMPRSS2 in mammals than in other animals.