INTRODUCTION
Zoonotic diseases remain among the greatest overall threats to global public health. Current estimates indicate that 70% of all new emerging infectious diseases are of zoonotic origin (
1,
2). Historical examples of zoonotic pathogens include avian and swine influenza viruses, henipaviruses, and severe acute respiratory syndrome coronavirus (SARS-CoV). More recently, the emergence of Ebola virus and of Middle East respiratory syndrome coronavirus (MERS-CoV) has resulted in local and intercontinental pandemics, with animal reservoirs believed to play a critical role in their emergence (
3–6). Bats, which are among the most common reservoirs for these zoonotic pathogens, have been identified as reservoirs or potential reservoirs for a number of highly pathogenic viruses, including SARS-CoV (
7–9), MERS-CoV (
10), and Ebola virus (
4). A recent study of the viral diversity within
Pteropus giganteus identified 55 viruses spanning nine viral families. The authors extrapolated these findings to estimate that a minimum of 320,000 unknown viruses are currently circulating among wild-mammal populations globally (
11). Nevertheless, the prevalence of viruses in wild mammals, particularly among bats, remains grossly understudied. Thus, the surveillance of local and global bat populations can help identify potentially zoonotic or pandemic pathogens prior to their emergence.
The
Caliciviridae family consists of nonenveloped, positive-sense RNA viruses subclassified into five genera,
Vesivirus,
Lagovirus,
Norovirus,
Sapovirus, and
Nebovirus, along with two unclassified genera, "
Recovirus" and "
Valovirus." Caliciviruses infect a wide range of hosts, including humans and wildlife and domestic, companion, and agricultural animals, although the origins and radiation of these viruses through mammalian populations remain uncertain. For instance, the
Norovirus genus, which contains the most common and well-known caliciviruses, consists of six genogroups and has an expansive host range. Primarily identified in humans, noroviruses (NoVs) have also been identified in canine (
12–14), feline (
14,
15), swine (
16,
17), murine (
18,
19), ovine (
20), and bovine (
21) species. Other caliciviruses have been identified in sea lions (
22–24), minks (
25), rabbits (
26,
27), chickens (
28), geese (
29,
30), fish (
31), and nonhuman primates (
32,
33), indicating the broad host range of the
Caliciviridae.
The mechanisms by which the caliciviruses have expanded their host range and emerged to infect the human population are currently unknown. Some human noroviruses (HuNoVs), the most well-known and prevalent caliciviruses, undergo epochal evolution, with a new pandemic strain emerging every 2 to 5 years (
34); new strains emerge following herd immunity-induced evolution within antigenic and receptor binding epitopes (
35). The immunocompromised human population might also serve as a reservoir from which pandemic noroviruses might emerge, while zoonotic transmission remains likely but unsubstantiated (
36–39). However, antibodies against animal caliciviruses have been detected in the human population, suggesting the potential for cross-species transmission events, though clinical disease has yet to be confirmed (
40–44). Thus, the identification and study of animal caliciviruses, including bat caliciviruses (BtCalVs), and their potential role in zoonotic disease and cross-species transmission potential represent significant gaps in global health preparedness.
Several recent reports have identified
Caliciviridae sequences resident within several different bat species worldwide (
45–50). Phylogenetic analyses have revealed that these viruses are closely related to sapoviruses (
45,
46,
49) and "valoviruses" (
46). More importantly, bat noroviruses have been reported in two separate studies of microbats in China (
50,
51), including one nearly full-length
Norovirus genome (here referred to as bat norovirus [BtNoV]) (
50). However, in the absence of any biological data, the potential of these viruses to transmit into other mammals, including humans, is difficult to predict from genome-length sequences. These reports of novel caliciviruses identified in bats not only stress the need for surveillance but also emphasize the need for detailed biological and immunologic characterization of new caliciviruses identified in wild-animal populations, particularly bats, to provide potential insights into cross-species transmission potential and human health.
Here, we evaluate the antigenicity and receptor-binding profiles of two bat caliciviruses. Specifically, we generated virus-like particles (VLPs) from the BtNoV capsid sequence (
50) that were detected with HuNoV-derived hyperimmune serums, indicating antigenic relationships shared between HuNoVs and BtNoV. Further, we characterize a novel bat calicivirus capsid sequence isolated from
Perimyotis subflavus, the tri-colored bat, in the Mid-Atlantic region of the United States (
52). Phylogenetic analyses revealed that this bat capsid sequence likely represents a novel calicivirus that is most closely related to non-human caliciviruses, such as lagoviruses and "recoviruses." We used VLPs from the newly identified bat calicivirus and BtNoV to assess their potential carbohydrate patterns in comparison with those of human noroviruses. Here, we describe host carbohydrate ligand-receptor binding patterns that overlapped between bat and human caliciviruses, suggesting that bat caliciviruses have the potential to clear one barrier to cross-species movement. Our data also suggest that bat caliciviruses share antigenic epitopes with HuNoVs, indicating a potential linkage via evolutionary descent.
DISCUSSION
Bats are important reservoir hosts for newly emerging viruses, many with pandemic potential. The reservoir hosts for caliciviruses and noroviruses are presently unknown; the existence of distinct genera and genogroups, coupled with multiple variable genotypes, suggests that multiple novel independent introductions from animal populations may have contributed to the genetic diversity of noroviruses and caliciviruses. Here, we reveal the potential for several bat caliciviruses to transmit across multiple species. Specifically, we found that a previously reported but uncharacterized virus, BtNoV (
50), is antigenically similar to HuNoVs and can bind HBGAs
in vitro. We built upon this finding through identification of a bat calicivirus sequence (BtCalV/A10) from
P. subflavus near an abandoned railroad tunnel outside Little Orleans, MD. We further characterized both bat caliciviruses through phylogenetic analyses, predictive structural homology modeling, VLP production, and identification of VLP-HBGA binding patterns
in vitro. Our report suggests that BtCalV/A10 is most closely related to RHDV, a highly virulent calicivirus that emerged suddenly in China from preexisting strains prior to spreading globally (
66,
67). While more sequences are needed for clarification, A10 likely represents a novel genus within
Caliciviridae, while BtNoV likely belongs to GV with MNV. Structural homology modeling indicated that both bat viruses retain ligand binding sites similar to those seen with HuNoVs, validating their capability to bind HBGAs. The capsid sequences of both bat viruses formed VLPs similar to those formed by HuNoVs and showed overlapping HBGA binding profiles with respect to several HuNoVs and MNV. Importantly, as HBGAs are encoded across numerous animal species, particularly among mammals (
68), similarities in attachment and entry cofactors suggest that these bat viruses have the potential to overcome a major roadblock to cross-species transmission across a variety of mammalian species.
Metagenomic sequencing has illuminated the great diversity and complexity of RNA viruses distributed throughout the natural world. However, new strategies are desperately needed to translate this information into meaningful predictions of biological processes, disease risk, and surveillance prioritization prior to epidemic or pandemic emergence. Previous studies on bat caliciviruses have focused on the use of phylogenetic analyses and structural modeling for classification of these novel viruses within the
Caliciviridae family (
45–49). While these techniques help contextualize viral relatedness within the framework of known and characterized viruses and their families, they do not necessarily aid in the understanding of biological functions or provide insight into the emergence potential of zoonotic pathogens. In this report, we utilized a novel platform combining metagenomics, phylogenetics, immunogenic comparisons, and homology modeling of two bat viruses; we showed that these viruses belong to
Caliciviridae and are structurally similar to known human and animal caliciviruses. We coupled these data with
in vitro techniques and immunologic assays to derive further relevant biological questions that can help to gauge the potential for cross-species transmission of two bat caliciviruses to other species prior to their emergence within the human population. A similar approach has identified the potential for MERS and SARS-like coronaviruses within Asian and African bat populations to replicate and emerge in human populations and to evade current therapeutics and has also identified bat coronaviruses that emerged more recently and caused outbreaks in swine (
8,
9,
69,
70).
Specifically, we showed that BtNoV VLPs maintain similarity to HuNoV VLPs with respect to antigenic and receptor-binding patterns (
Fig. 1 and
2). Additionally, BtNoV VLPs bound HBGA receptors across a variety of temperatures similarly to HuNoVs (
Fig. 2), indicating the potential for viral transmission under a variety of physiological conditions. Further, we utilized a second, unrelated bat calicivirus to demonstrate that these observations may be pertinent across multiple bat calicivirus species. For example, BtCalV/A10 rooted a phylogenetic clade that included RHDV and "recoviruses," which were the only viruses separating BtCalV/A10 from noroviruses (
Fig. 3), but likely represents a novel genus within the
Caliciviridae family. Next, we modeled the P domain structure of BtCalV/A10 and BtNoV and host carbohydrate receptor binding sites on their capsid structure similar to those seen with known HuNoV structures (
Fig. 4). Finally, we observed that BtCalV/A10 also formed VLPs of similar sizes and structures that bound to several HBGAs and/or sialic acid moieties and displayed binding patterns that overlapped those of HuNoVs
in vitro (
Fig. 5A and
B). However, despite the similarity in the predicted P domain structures, the bat viruses, similarly to GI and GII HuNoVs, displayed disparate HBGA binding patterns, indicating multiple potential paths for cross-species transmission (
Fig. 5B). Therefore, the utilization of the aforementioned platform identified antigenic relatedness of BtNoV and HuNoVs as well as the potential for cross-species transmission of two previously uncharacterized or unidentified bat caliciviruses.
Noroviruses have been historically classified on the basis of capsid sequence homology (
59); based on this strategy, BtNoV would likely be classified within
Norovirus GV alongside MNV. Enzyme immunoassays have been powerful tools for understanding antigenic relationships among noroviruses in spite of sequence homology, especially in elucidating how antigenic drift contributes to the evasion of herd immunity and emergence of novel HuNoV strains (
35). In the present study, EIA performed with polyclonal serums derived from historic and current strains of HuNoVs showed that BtNoV VLPs are antigenically similar (
Fig. 1) despite the parental virus’s phylogenetic relatedness to MNV (
Fig. 3). Other bat noroviruses have shown extensive homology with GIV noroviruses (
51), whose antibodies have been detected in humans and canine species (
71,
72). These results further illustrate the importance of our proposed platform: capsid sequence homology alone would not have predicted that bat caliciviruses could bind synthetic carbohydrates or be detected by serums against human viruses;
in vitro assays revealed novel information regarding the caliciviruses circulating in bat populations.
The
FUT2 gene, whose protein product promotes expression of HBGAs on mucosal surfaces, has a well-established role in HuNoV infection and RHDV disease (
73–75). However, the receptors for animal caliciviruses are variable and less clear. For instance, GIII bovine noroviruses bind α-galactose (
76,
77), MNVs bind sialic acids and CD300lf (
78–80), porcine sapoviruses bind sialic acids (
81), and both feline calicivirus and the zoonotic San Miguel sea lion virus Hom-1 bind junctional adhesion molecule-1 (
82,
83). While protein receptors seem quite variable, there is significant overlap in the identified carbohydrate binding patterns of all caliciviruses (
Fig. 5), demonstrating multiple potential routes of cross-species transmission of these viruses.
Recent genomics analysis has further indicated the potential for animal caliciviruses, including bat caliciviruses, to cross species barriers. For instance, HBGAs have been identified or predicted across numerous animal species (
68). Importantly, the A and B fucosyltransferase enzymes of
Myotis lucifugus, a microbat similar to
P. subflavus, contain amino acid sequences, amino acid motifs, and levels of phylogenetic relatedness similar to those of their human orthologs (
68). Similarly, CD300lf orthologs have been predicted in numerous animal species, including several Asian bats, such as
Hipposideros armiger and
Rhinolophus sinicus. In addition to CD300lf, sialic acids have previously been implicated as receptors or cellular cofactors for binding and entry by MNVs and, more recently, by HuNoVs (
78,
84–86). Future studies will reveal whether the BtCalV/A10 or BtNoV VLPs also bind bat or mammalian CD300lf proteins. Though numerous factors mediate the cross-species transmission, replication, and pathogenesis of viruses, our results, coupled with those revealing the analogous host receptor structures, suggest that these bat caliciviruses have the potential to overcome host receptor barriers as a roadblock to cross-species transmission.
Our findings, along with previous reports (
45,
46,
48–50), support the hypothesis that bats are a potential reservoir species for noroviruses and other caliciviruses, such as RHDV. While the BtCalV/A10 strain is very distant from RHDV, our data strongly suggest that more bat caliciviruses should be identified and sequenced, primarily in China, and that the results may reveal precursor strains that are more akin to this highly virulent pathogen of rabbits. The present study did not support the conclusion that bat caliciviruses will emerge in human or other mammalian populations but did suggest their potential for cross-species transmission. Our results further illustrate the importance of conducting continued surveillance of bat populations for viral pathogens prior to their emergence in human and other mammalian populations (
11). If viral discovery is partnered with developing and answering meaningful biological questions to identify high-risk viral strains prior to epidemic or pandemic emergence, the cost of such combined scientific investigations will pale in comparison to the cost of treatment and eradication of these pathogens as they emerge in the human population.