Recombination of the epsilon determinant and corneal tropism: Human adenovirus species D types 15, 29, 56, and 69

Abstract

Viruses within human adenovirus species D (HAdV-D) infect epithelia at essentially every mucosal site. Hypervariable loops 1 and 2 of the hexon capsid protein contain epitopes that together form the epsilon determinant for serum neutralization. We report our analyses comparing HAdV-D15, 29, 56, and the recently identified type 69, each with highly similar hexons and the same serum neutralization profile, but otherwise disparate genomes. Of these, only HAdV-D type 56 is associated with epidemic keratoconjunctivitis (EKC), a severe infection of ocular surface epithelium and underlying corneal stroma. In the mouse adenovirus keratitis model, all four viruses induced inflammation. However, HAdV-D56 entry into human corneal epithelial cells and fibroblasts in vitro dramatically exceeded that of the other three viruses. We conclude that the hexon epsilon determinant is not a prime contributor to corneal tropism.

Keywords: Adenovirus; Epidemic keratoconjunctivitis; Homologous recombination; Keratitis.

Fig. 1

(A) Global pairwise whole-genome sequence comparison between HAdV-D69 and HAdV-D15 and HAdV-D28 (negative control). (B) Global pairwise alignment of HAdV-D69 fiber sequence with fiber sequence from ATCC HAdV-D15 lots “Ad15-VR661”, “Ad15-VR16” and “Ad15-VR1092-CH38-V215”. (C) SimPlot analysis comparing HAdV-D69 with whole genomes of the other 43 HAdV-Ds in GenBank.

Fig. 2

Bootstrap-confirmed (500 replicates) neighbor-joining phylogenetic tree of (A) whole genome, (B) penton base, (C) hexon, (D) 100K gene, (E) E3 transcription unit, and (F) fiber, from 44 viruses within HAdV-D. Bootstrap values below 80 are indicative of low confidence.

Fig. 3

(A) SimPlot analyses for the hexon regions of HAdV-D15, 29, 56, 69 and 28 (negative control). (B) Serum neutralization results using antiserum against HAdV-D15, 22, 29, and 28, for the viruses HAdV-D15, 29, 56, 69, 22 and 28 (later two as negative controls). (C) Bootscan analysis of HAdV-D69 penton base gene with all HAdV-Ds (excluding viruses with high similarity) shows recombination with HAdV-D53. (D) Bootscan analysis of HAdV-D69 fiber with all HAdV-Ds except type 22, shows recombination with HAdV-D42.

Fig. 4

(A) Clinical photographs of keratitis in C57Bl/6j mice corneas injected intrastromally with dialysis buffer control (Mock) or 10⁵ TCID of HAdV-D15, 29, 56 and 69 at 4 days post infection. (B) Flow cytometric analysis of CD45+ cell infiltration into infected corneas at 4 days post infection. (C) ELISA analysis of infected corneas at 16 hours post infection. (D) Comparison of infection by Cy3-labeled HAdV-D15, 29, 56 and 69 in A549 cells, Tert-immortalized human corneal epithelial cells (THE) and primary human corneal fibroblasts (HCF), as seen by confocal microscopy at 30 minutes post infection. Cy3 labels virus red, DAPI stains nuclei blue, and phalloidin stains cellular actin green. (E) Quantification of images was performed using Amira software. See text for statistical results.