Tissue distribution of influenza virus receptors in mammalian hosts using tissue microarrays

Abstract

Susceptibility of a host to a virus requires the expression of virus-specific receptors. These are host molecules present on the cell surface that support the binding of specific viral attachment proteins. The receptors for many viruses, including influenza A virus, have previously been elucidated. However, their tissue and host distribution contributing to the susceptibility of species is often not well understood. In order to broaden our knowledge on the expression of these host molecules, we aimed at creating tissue microarrays (TMA) of various mammalian species. In particular, TMAs represent the true complexity of cell surface molecules present in a host and are therefore excellent tools to study virus-host interactions. This technique still faced several technical drawbacks. Here, we first technically optimized the TMA construction method such that tissue-slides were easier easier and more intact generated. Subsequently, a variety of tissues of pig, dog and horse of animals that were submitted to the necropsy hall of the Pathobiology department. Formalin-fixed, paraffin embedded tissues were analyzed and healthy tissues from these animals were selected based on microscopic evaluation of tissue slides. Tissue cores of each of the tissues were combined in one tissue block to generate (monospecies or multispecies) porcine, canine and equine TMAs, respectively. From these TMA blocks, tissue-slides were generated to study viral receptor expression. First, the tissue distribution of the sialic acids Neu5Acα2-3Gal (α2-3 linked sialic acid) and Neu5Acα2-6Gal (α2-6 linked sialic acid), was studied using the plant-derived lectins; Maackia amurensis lectins I & II and the Sambucus nigra lectin. These sialic acids constitute the viral receptors of avian and human influenza A viruses, respectively. Next, we studied the binding of the viral attachment proteins HA of various influenza A subtypes using protein histochemistry assays. To study interspecies variation of these different IVA receptors, a monospecies TMA composed of tissues of six anatomic locations of the respiratory tracts from six animals were compared. We observed that the variability of sialic acid expression in the same regions was low, indicating that TMAs composed of tissues of one animal already provide sufficient information in order to draw valid conclusions for larger animal groups. Interestingly, in the porcine respiratory tract, co-expression of Neu5Acα2-3Gal and Neu5Acα2-6Gal was only detected in the lower regions of the respiratory tract, and was absent in the trachea, suggesting that the current hypothesis that pigs can function as a ‘mixing vessel’ for IAVs, is disputable. In order to gain insight to the interspecies transmission of equine IAVs to the canine population, we analyzed the presence of Neu5Acα2-3Gal & Neu5Gcα2-3Gal receptor expression in equine and canine tissues. Lectin staining revealed that Neu5Acα2-3Gal was expressed in various canine and equine tissues, which was confirmed by the specific binding of HA-H3, of equine IAV subtype, to respiratory tract tissue of both species. In contrast, Neu5Gcα2-3Gal expression could only be detected in equine tissues upon staining of TMA slides with a specific antibody, which was in agreement with the protein binding of HA-H7 of equine IAV subtype to equine, but not canine tissues. These results might explain that H3 IAVs of equine origin have crossed the species barrier to dogs, while this has not been observed for H7 IAVs. Taken together, the generated TMAs of various mammalian species provide an excellent tool for the analysis of the distribution of viral receptors in multiple host species, thereby increasing our understanding of host susceptibility and interspecies transmission for particular viruses.