Molecular characterization of hemagglutination domains on the fibers of subgenus D adenoviruses

Abstract

The adenovirus fiber mediates the agglutination of erythrocytes. Based on differential hemagglutinating properties, subgenus D adenoviruses can be subdivided into clusters DI, DII, and DIII. While subgenus DI adenoviruses agglutinate rat and human erythrocytes, DII adenoviruses simply agglutinate rat erythrocytes and DIII adenoviruses display no or only weak rat erythrocyte agglutination. Amino acid sequence comparisons revealed distinct domains on the fiber knob which could be involved in hemagglutination. In order to localize and characterize the domains responsible for the interaction with rat and human erythrocytes, potential hemagglutination domains of the adenovirus type 9 (Ad9) (subgenus DI) fiber knob were introduced into Ad17 (subgenus DII) and Ad28 (subgenus DIII) fiber knobs by primer-directed mutagenesis. Furthermore, rat erythrocyte hemagglutination domains were also introduced into the Ad3 (subgenus B) fiber knob, which only agglutinated monkey erythrocytes. Altogether, 27 chimeric and mutated fiber proteins were expressed in Escherichia coli and subsequently tested for hemagglutination activity. The hemagglutination tests revealed that at least two domains can mediate the agglutination of rat erythrocytes. While one domain is located on the GH loop, the other domain extends from the C beta strand to the CD loop. The domain on the GH loop was partially conserved in all adenoviruses showing an incomplete hemagglutination pattern with rat erythrocytes. The domains involved in the agglutination of human erythrocytes are located on the CD and HI loops of the subgenus DI fiber knob.

FIG. 1

Comparison of the predicted amino acid sequences of the Ad9, Ad19, Ad15, Ad17, and Ad28 fiber polypeptides. The numbering on the left takes into account the deletions which resulted from the alignment. The structural domains tail, shaft, and knob and the eight repeating motifs in the shaft region are marked by arrows. Amino acids identical to the amino acids presented in the preceding sequence are marked by asterisks; plus signs indicate homology to Ad9; number signs indicate homology to Ad19; and deletions are represented by dashes. Subg., subgenus. The amino acid residues which were exchanged between Ad9, Ad3, Ad17, and Ad28 are underlined and written in bold (the Ad3 knob sequence is shown in Fig. 2).

FIG. 2

Fiber knob amino acid alignment of human adenoviruses representing subgenera (Subg.) A to F. Deletions are represented by dashes. Potential HA domains in the Ad9 fiber knob are underlined and written in bold. Amino acid residues identical to Ad9 amino acid residues are underlined. FK3V₁V₂ to FK17S represent mutated fiber knob proteins. The boxes indicate the β strands (A to J) corresponding to the published structure of the Ad5 knob domain. The regions (loops; not indicated) between the β strands were labeled by Xia et al. (40, 41) after the β strands that they connect, e.g., AB loop and CD loop; the DG loop also includes the short β strands E and F.

FIG. 3

Coomassie blue-stained SDS-PAGE (15% polyacrylamide) gel of expressed recombinant fiber knob proteins. Lanes: M, molecular mass markers (the sizes of the markers are indicated on the left in kilodaltons); 1, uninduced E. coli M15 cells (transformed with pQE); 2, fiber knob protein FK17GCE (sample boiled prior to loading); 3, fiber knob protein FK17GCE (sample not boiled); 4, fiber knob protein FK17S (sample boiled prior to loading); 5, fiber knob protein FK17S (sample not boiled). The arrows indicate the positions of the monomeric and trimeric fiber knob proteins.

FIG. 4

Western blot (of 15% polyacrylamide gel). Lanes: 1, uninduced E. coli M15 cells (transformed with pQE); 2, fiber knob protein FK17GCE (sample boiled prior to loading); 3, fiber knob protein FK17GCE (sample not boiled); 4, fiber knob protein FK17S (sample boiled prior to loading); 5, fiber knob protein FK17S (sample not boiled). The sizes of the markers are indicated on the left in kilodaltons. The arrows indicate the positions of the monomeric and trimeric fiber knob proteins.

FIG. 5

Localization of the HA domains within the polypeptide chain of the knob monomer. The positions of the HA domains were determined by comparing the subgenus D fiber knob sequences to the Ad5 knob structure published by Xia et al. (41). Loops, β strands (A to J), and chain termini are labeled. Arrows 1 and 2 indicate the domains involved in the agglutination of rat erythrocytes, and arrows 3 and 4 indicate the domains involved in the agglutination of human erythrocytes.