Ball python nidovirus: a candidate etiologic agent for severe respiratory disease in Python regius

Abstract

A severe, sometimes fatal respiratory disease has been observed in captive ball pythons (Python regius) since the late 1990s. In order to better understand this disease and its etiology, we collected case and control samples and performed pathological and diagnostic analyses. Electron micrographs revealed filamentous virus-like particles in lung epithelial cells of sick animals. Diagnostic testing for known pathogens did not identify an etiologic agent, so unbiased metagenomic sequencing was performed. Abundant nidovirus-like sequences were identified in cases and were used to assemble the genome of a previously unknown virus in the order Nidovirales. The nidoviruses, which were not previously known to infect nonavian reptiles, are a diverse order that includes important human and veterinary pathogens. The presence of the viral RNA was confirmed in all diseased animals (n = 8) but was not detected in healthy pythons or other snakes (n = 57). Viral RNA levels were generally highest in the lung and other respiratory tract tissues. The 33.5-kb viral genome is the largest RNA genome yet described and shares canonical characteristics with other nidovirus genomes, although several features distinguish this from related viruses. This virus, which we named ball python nidovirus (BPNV), will likely establish a new genus in Torovirinae subfamily. The identification of a novel nidovirus in reptiles contributes to our understanding of the biology and evolution of related viruses, and its association with lung disease in pythons is a promising step toward elucidating an etiology for this long-standing veterinary disease.

Importance: Ball pythons are popular pets because of their diverse coloration, generally nonaggressive behavior, and relatively small size. Since the 1990s, veterinarians have been aware of an infectious respiratory disease of unknown cause in ball pythons that can be fatal. We used unbiased shotgun sequencing to discover a novel virus in the order Nidovirales that was present in cases but not controls. While nidoviruses are known to infect a variety of animals, this is the first report of a nidovirus recovered from any reptile. This report will enable diagnostics that will assist in determining the role of this virus in the causation of disease, which would allow control of the disease in zoos and private collections. Given its evolutionary divergence from known nidoviruses and its unique host, the study of reptile nidoviruses may further our understanding of related diseases and the viruses that cause them in humans and other animals.

FIG 1

Representative macroscopic lesions. (A) Common wild-type ball python (no. 1), Python regius. The palatine mucosa is both thickened and necrotic, and there is an accumulation of caseous material in the internal choanae. (B) Mojave variant ball python (no. 11), Python regius. The lung is thickened and edematous with abundant mucoid to caseous material (arrow) in air passageways.

FIG 2

Representative photomicrographs of diseased lung section. (A) Ball python 1 (BP1). Photomicrograph of a cross section of the lung. The smooth muscle (SM) bundle at the luminal end of each septum is hyperplastic and surrounded by inflammatory cells. As the septa approach the base of each faveolus, the septal wall between adjacent faveoli is of normal thickness. No inflammatory cells are seen in air passageways. H&E stain; bar = 500 µm. (B) BP1. Photomicrograph of a cross section of lung. At a higher power, numerous round cells (RC) have infiltrated the submucosa and overlying epithelial cells are hyperplastic (arrows). H&E stain; bar = 100 µm. (C) BP11. Photomicrograph of a cross section of lung. Severe proliferative interstitial pneumonia. Inflammation and epithelial hyperplasia (arrows) markedly thicken the trabecular septa and faveolar walls. Abundant mucinous exudate fills the lumina. H&E stain; bar = 200 µm. (D) BP11. Photomicrograph of a cross section of the lung at a higher magnification. Faveolar septum, severe proliferative interstitial pneumonia. The epithelium (EP) is hypercellular, thick, and densely packed with pseudostratified columnar cells, the majority of which produce mucus. Macrophages, lymphocytes, and granulocytes infiltrate (IN) the epithelium and submucosa. Faveolar capillaries are below the epithelial layer. H&E stain; bar = 50 µm.

FIG 3

Transmission electron photomicrographs of pulmonary epithelial cells of ball pythons, Python regius, with proliferative pneumonia. (A) BP6. Multiple bacilliform (arrows) and spherical nucleocapsid cores of virions are seen within the cytoplasm. Bar = 200 nm. (B) BP6. At a higher magnification, bacillary nucleocapsids are surrounded by granular cytoplasmic material (white arrows), presumed to be a component of the envelope. Bar = 100 nm. (C) BP6. Progression of immature forms in the cytoplasmic matrix to more mature forms (arrow) within a cytoplasmic vesicle. Bar = 200 nm. (D) BP2. Mature enveloped spherical virions (arrows) within a cytoplasmic vesicle. Spikes can be seen on the surface of several virions (arrowhead). Bar = 200 nm. (E) BP6. Spherical and filamentous mature virions (arrows) can be seen within cytoplasmic vesicles that are subjacent to microvilli projecting from the cell membrane Bar = 200 nm. (F) BP6. Spherical and bacillary forms of mature virions (arrows) can be seen in between microvilli and cilia of a pulmonary epithelial cell. Bacillary bacteria lacking a trilaminar cell wall (arrowheads) are also present. Scale bar corresponds to 1 µm.

FIG 4

Ball python nidovirus genome and replicase polyprotein organization. (A) A cartoon showing the genome organization of ball python nidovirus. Major open reading frames are indicated. RFS, position of putative ribosomal frameshift “slippery” sequence. (B) Position of Sanger sequencing PCR and RACE amplicons used to validate the NGS-based assembly (C) Concordance plot of read pairs mapping to assembly. Read pairs from snake no. 1 data set were mapped to the genome assembly using the Bowtie2 aligner. The inferred size of each library molecule is plotted. (D) Read coverage. Reads were aligned as in panel C, and the average number of bases supporting each position in 50-nt sliding windows of the alignment is indicated. The dip at approximately nt 5000 corresponds to a repeat-containing region with decreased mappability. (E) Conserved domains present in the replicase polyproteins (pp1ab). Domains were identified by searching the PFam database using the HMMER3 tool as described in the text and Table 2. Scale bars are indicated.

FIG 5

Analysis of subgenomic RNA species. (A) A cartoon showing the location of primers used to amplify regions of sgRNAs. Twenty-three kilobases have been removed for clarity. (B) Agarose gel electrophoresis analysis of PCR products obtained from reactions using the indicated primer pairs. Positions of molecular size standards (in bp) are indicated. (C) Sequences of PCR products spanning the junctions between upstream ORF sequences and 5′ UTR (leader) sequence. The top sequence in each triplet is the sequence of the genomic RNA (gRNA) beginning at base 150; the middle sequence is from the amplicons in panel B, corresponding to the indicated subgenomic mRNAs; the bottom sequence is that of the genomic RNA beginning at the indicated base. Regions of homology and putative start codons are underlined. For the ORF5a/b triplet, the start codons of both ORFs are indicated, and 180 nt of intervening sequence has been removed.

FIG 6

Detection of ball python nidovirus RNA is associated with clinical manifestation. (A) Relative levels of viral RNA were measured by qRT-PCR in case and control lung samples. Primers targeted ORF1a or ORF2 (S) sequences. Each row represents values from a single case or control animal of the indicated species. Levels were normalized to levels of cellular GAPDH mRNA and are represented as a heat map. Corallus annulatus and boa constrictor samples have been described previously (59). A polymorphism in the S sequence primer-binding site of snake 8 decreased amplification efficiency. (B) Relative levels of viral RNA in different tissues from infected snakes. For each snake, values were calculated and quantified as in panel A. Lung from a healthy control ball python served as a negative control.

FIG 7

Phylogeny of conserved replicase subunits of representative nidoviruses. Unrooted Bayesian phylogeny based on concatenated multiple sequence alignments of conserved regions of RdRp and helicase domains of replicase polyproteins. At select nodes, Bayesian posterior probabilities and ML bootstrap values for clusters are given to the left and right of forward slashes. Nodes with 100% support are marked with asterisks. The five major nidovirus families and subfamilies are indicated, as are select genera and the currently uncategorized BPNV and possum nidovirus taxa. Known host phyla are indicated in parentheses. The branch leading to BPNV is highlighted red for emphasis.