Molecular Detection and Phylogenetic Characterization of Bat and Human Adenoviruses in Southern China

Introduction

Adenoviruses (AdVs) are members of the family Adenoviridae, which can be divided into five genera: Mastadenovirus, Aviadenovirus, Atadenovirus, Siadenovirus, and Ichtadenovirus. These icosahedral, nonenveloped, double-stranded DNA viruses have been found to infect almost all vertebrate classes (Knowles 2011). In humans, 57 distinct adenoviral serotypes have been found; these cause a wide range of illnesses, from infections that produce no symptoms to those resulting in severe complications, including respiratory disease, conjunctivitis, and gastroenteritis (Kojaoghlanian et al. 2003). Similar to infections in humans, AdV-infected animals present a variety of symptoms and outcomes, from subclinical to lethal.

It is well acknowledged that AdVs are species specific; however, the cross-species transmission of new AdVs between humans and other animals could be observed (Chiu et al. 2013, Yu et al. 2013). In 2009, an outbreak of AdV infection among captive titi monkeys in California resulted in the deaths of most animals in a closely clustered colony (Yu et al. 2013). Meanwhile, serological evidence of AdV infection was found in a researcher working closely with these monkeys. These findings raised concerns regarding the potential of animal-to-human AdV transmission. It has been well recognized that bats might serve as reservoirs of a variety of zoonotic diseases, such as severe acute respiratory syndrome, rabies, and Nipah, Hendra, and ebola virus infections (Calisher et al. 2006). Recently, several studies have documented the genetic diversity of AdVs in bats originating from Kenya, Brazil, China, Japan, Germany, and Hungary (Maeda et al. 2008, Sonntag et al. 2009, Li et al. 2010, Janoska et al. 2011, Lima et al. 2013, Conrardy et al. 2014).

These studies motivated us to evaluate whether there is zoonotic transmission of AdV between bats and humans. Below, we report the molecular detection and phylogenetic characteristics of bat and human AdVs in southern China in samples that were collected at similar times.

Materials and Methods

Results

We sampled 520 fecal specimens from eight bat species (four families), including Hipposideros larvatus, Hipposideros pomona, Myotis ricketti, Miniopterus schreibersii, Scotophilus kuhlii, Taphozous melanopogon, Rhinolophus blythi, and Cynopterus sphinx. Thirty-six (6.9%) of the fecal samples were positive for AdVs. Notably, AdVs were detected in all examined samples from bats of the family Vespertilionidae, including the following species: M. ricketti (28.6%, 2/7), M. schreibersii (0.7%, 1/144), S. kuhlii (13.6%, 24/177), and T. melanopogon (33.3%, 1/3). C. sphinx was tested for the presence of AdV for the first time, yielding a positive rate of 13.3% (8/60). No positive results were found in either Hipposideridae or Rhinolophidae (Table 1).

A phylogenetic tree was constructed based on deduced amino acid sequences of partial DPOL gene reads. Mastadenovirus strains were subdivided into five groups: Group 1 contained AdVs from Vespertilionid bats, Group 3 contained human AdVs, and Group 5 contained six AdVs from C. sphinx. Groups 2 and 4 both harbored AdVs from ovine, bovine, tree shrew, porcine, and sea lion. Phylogenetic analysis revealed a long-term coevolution between most of the identified bat AdVs and their hosts. AdVs detected in S. kuhlii in China occupied a distant branch from the main cluster of AdVs, which were found in Vespertilionid bats (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Phylogenetic analysis based on 82 amino acids of the DPOL of adenovirus detected from bats in southern China. The phylogenetic tree was generated by using the NJ algorithm with the p-distance model. A bootstrap test was replicated for 1000 times. The numbers above the branches indicated the NJ bootstrap values. Bold triangles indicated the adenoviruses detected in this study. CS, Cynopterus sphinx; DPOL, DNA polymerase; GZ, Guangzhou; HL, Hipposideros larvatus; HN, Hainan province; HP, Hipposideros pomona; HZ, Huizhou; MR, Myotis ricketti; MS, Miniopterus schreibersii; NJ, neighbor-joining; RB, Rhinolophus blythi; SK, Scotophilus kuhlii; TM, Taphozous melanopogon; YF, Yunfu.

All identified bat AdVs belonged to the genus Mastadenovirus, but were divergent from other Mastadenoviruses. Low amino acid similarities were observed when comparing the AdVs identified in the current study with those from other animal species, including bovine (55.5–63.4%), ovine (52.3–66.0%), porcine (50.7–65.0%), tree shrew (53.2–67.7%), Tursiops (55.0–63.4%), and sea lion (48.3–62.9%). Notably, two canine AdVs (CaAdV 1 and CaAdV 2) and equine AdV (EqAdV-1) fell into the Vespertilionidae clade of AdVs, sharing 72.5–80.6% and 70.9–72.5% amino acid similarities with S. kuhlii. The bat AdVs identified in this study had 41.9–100.0% amino acid identity with the bat AdVs retrieved from GenBank. Within the same bat species, amino acid identities were conserved in M. ricketti and S. kuhlii, but not with some previously unknown AdVs from C. sphinx (61.4–100.0%, with 57.3% of similarity being less than 90.0%). These viruses were genetically divergent from those previously described because they shared only 41.9–69.3% amino acid similarity (Supplementary Table S2).

Sixteen (4.9%) of the 328 samples from patients with diarrhea tested positive for AdV, including 14 positives for human AdV 41, 1 positive for human AdV 5, and 1 positive for human AdV 7. High similarity (100.0%) was observed in the DPOL gene among 14 human AdV 41. AdV-GZ700, the representative human AdV 41 in this study, shared 100.0% amino acid similarity with human AdV 41 isolates from different regions, including Beijing, New York city, and Japan. AdV-GZ927 and AdV-GZ1030 were identical to previously reported AdV types 7 and 5. Only 68.2–69.8% amino acid similarity was found among AdV-700, AdV-927, and AdV-1030. Compared with other animal species, human AdVs shared amino acid similarities of 58.7–63.4%, 66.6–69.8%, 66.6–71.4%, 60.3–68.2%, 61.9–63.4%, 61.0–65%, and 58.7–66.6% with bovine, ovine, porcine, tree shrew, Tursiops, sea lion, and canine, respectively (Supplementary Table S2).

All bat and human AdVs were clustered in Mastadenovirus. Only 57.1–69.3% amino acid similarity could be found between human and bat AdVs in this study.

Discussion

To assess the possibility of zoonotic transmission from bats to humans, we conducted a comparative analysis of bat and human AdVs collected during the same time period and at the same locations. The results showed a 7.1% (37/520) and 4.9% (16/328) AdV detection rate in bats and humans, respectively. In addition, we identified diverse and novel AdVs in bats as well as classical human AdVs, including AdV types 5, 7, and 41.

Bat AdV was initially isolated from a megachiropteran species in Japan in 2008 (Maeda et al. 2008). Subsequently, multiple AdVs have been identified in different megachiropteran and microchiropteran species worldwide (Sonntag et al. 2009, Li et al. 2010, Janoska et al. 2011, Lima et al. 2013, Conrardy et al. 2014). Consistent with previously reported studies (Maeda et al. 2008), AdVs detected from megachiropterans were distinct from those from microchiropterans in this study. The newly identified AdVs in the megachiropteran species C. sphinx had significant sequence variation (61.4–100.0%, with 57.3% of similarities being <90%), forming five genetically related yet distinct clusters.

The topology of the phylogenetic tree suggested the long-term coevolution of most of the identified bat AdVs and their hosts, except for S. kuhlii AdVs. Distant from those detected in other Vespertilionid bats, S. kuhlii AdVs shared higher DPOL amino acid similarity with CaAdV-1 and CaAdV-2 derived from canine. A previous full genomic analysis demonstrated that BtAdV-TJM, a novel bat AdV, isolated from bat fecal samples in China was closely related to tree shrew and canine AdVs (Li et al. 2010). Moreover, CaAdV-1 and CaAdV-2 might have a broad spectrum of hosts, as they have been detected in coyotes, bears, jackals, pandas, skunks, and wolves (Philippa et al. 2004). These findings implied that although AdVs generally are species specific, some AdVs might circulate in various animal populations through cross-species transmission (Philippa et al. 2004, Li et al. 2010, Yu et al. 2013). However, further studies are warranted to verify the hypothesis.

Low sequence variation was found between human AdVs in this study. AdV-GZ700 was identical to human AdV 41, which has been detected in Beijing, Japan, and New York, with amino acid identities of 100.0%. AdV-GZ927 and AdV-GZ1030 were identical to AdV types 7 and 5. In addition, human AdVs shared low similarities with bat AdVs in deduced amino acid sequence of the DPOL region. It is notable that only 57.1–69.3% amino acid similarity could be found between human and bat AdVs, including in the DPOL gene alignment. This suggests there might have been no possibility of zoonotic transmission of AdV from bat to human. It must be highlighted that the DPOL fragment analyzed in this study is quite short and conserved, which might have led to the low resolution of the phylogenetic tree. Additional genomic sequence data is required to examine the potential pathogenic role of AdVs on other bat species, providing insights into the ecology and evolution of AdVs in different bat species.

Conclusions

In conclusion, we found diverse and novel AdVs in bats based on molecular detection and phylogenetic analysis. Our study results further indicated that bat AdVs and human AdVs are species specific. There is no evidence of cross-species transmission of AdV between bats and humans based on the current data. Further studies that include more detailed genomic sequence information from bat and human AdVs, coupled with additional epidemiological investigation, might offer greater insights.