Abstract
Influenza A is a common respiratory infection of horses, and rapid diagnosis is important for its detection and control. Sensitive detection of influenza currently requires viral culture and is not always feasible. The polymerase chain reaction (PCR) was used to detect DNA produced by reverse transcription of equine influenza in stored nasal secretions, vaccines, and allantoic fluids. Primers directed at a target of 212 bp on conserved segment 7 (matrix gene) of human influenza A/Bangkok/ 1/79 (H3N2) produced amplification products of appropriate size with influenza A/Equine/Prague/ 1/56 (H7N7), A/Equine/Miami/63 (H3N8), A/Equine/ Kentucky/79 (H3N8), and A/Equine/Kentucky/2/91 (H3N8) in infected frozen allantoic fluids and in frozen extracts of nasal swabs of 2 horses with naturally acquired influenza. The products bound a 32P-labeled hybridization probe to an inner region of the target. Control samples, including nasal secretions from a horse infected with herpesvirus, were negative. In a prospective study, 2 ponies inhaled aerosols of influenza A/Equine/ Kentucky/2/91 (H3N8), and thereafter supernatants of nasal swabs in transport medium were obtained daily for 10 days for culture and PCR. Amplification products were evaluated by size and binding of a 32P-labeled probe and also by dotblotting and binding of a biotin-labeled probe. Culture detected influenza more consistently than did PCR in the first 2 days of infection, but PCR detected virus more often later in infection. Gels were the most sensitive, but radiometric and biotin-labeled probes gave specific results and were consistently positive from days 3–6. PCR is suitable for detection of equine influenza in clinical samples.