Establishing Methods to Monitor Influenza (A)H5N1 Virus in Dairy Cattle Milk, Massachusetts, USA

Abstract

Highly pathogenic avian influenza A(H5N1) virus has caused a multistate outbreak among US dairy cattle, spreading across 16 states and infecting hundreds of herds since its onset. We rapidly developed and optimized PCR-based detection assays and sequencing protocols to support H5N1 molecular surveillance. Using 214 retail milk samples from 20 states for methods development, we found that H5N1 virus concentrations by digital PCR strongly correlated with quantitative PCR cycle threshold values; digital PCR exhibited greater sensitivity. Metagenomic sequencing after hybrid selection was best for higher concentration samples, whereas amplicon sequencing performed best for lower concentrations. By establishing these methods, we were able to support the creation of a statewide surveillance program to perform monthly testing of bulk milk samples from all dairy cattle farms in Massachusetts, USA, which remain negative to date. The methods, workflow, and recommendations described provide a framework for others aiming to conduct H5N1 surveillance efforts.

Keywords: H5N1 subtype; PCR; United States; cattle; disease outbreak; farms; influenza; influenza A virus; milk; public health; public health surveillance; viral RNA; viruses; workflow.

Figure 1

Validation and characterization of dPCR and qPCR on synthetic spike-in samples in study of methods to monitor influenza A(H5N1) virus in dairy cattle milk, Massachusetts, USA. A, B) Limit of detection analysis for correlation of dPCR concentrations with qPCR Ct values (A) and measured concentrations compared to expected concentrations for both qPCR and dPCR (B). C, D) Detection of dPCR (H5_Taq and RP_Bov) (C) and qPCR (H5_Taq) (D) assays using serial dilutions of synthetic H5N1 RNA standard material. For qPCR data, we combined and jointly analyzed all standard curve data from runs during retail milk testing. Fitted lines in panels A and D represent simple linear regression lines of best fit. Error bars indicate +1 SD. Ct, cycle threshold; dPCR, digital PCR; qPCR, quantitative PCR; R², coefficient of determination.

Figure 2

Digital PCR detection of synthetic nucleic acid (top) and RNaseP Bovine (bottom) in study of methods to monitor influenza A(H5N1) virus in dairy cattle milk, Massachusetts, USA. For direct extraction, we extracted 200 μL of milk spiked with serial dilutions of H5N1 synthetic gene fragments. For precentrifugation, we centrifuged samples for 12,000 × g for 10 minutes after spike-in, after which we extracted 200 μL. Extractions were performed using the MagMAX CORE extraction kit (Thermo Fisher Scientific, https://www.thermofisher.com).

Figure 3

Comparison of dPCR and qPCR virus testing on retail milk samples in study of methods to monitor influenza A(H5N1) virus in dairy cattle milk, Massachusetts, USA. A) Agreement of positive and negative calls of milk samples between the 2 platforms; B) correlation of H5N1 measured by dPCR concentration compared with qPCR Ct value. For plotting purposes, samples not detected by dPCR were graphed with a dPCR concentration of 0 copies/μL, whereas samples not detected by qPCR were graphed with a Ct value of 40. Error bars indicate +1 SD. Ct, cycle threshold; dPCR, digital PCR; qPCR, quantitative PCR; R², coefficient of determination.

Figure 4

Virus and bovine ribonuclease P (RP_Bov) concentrations for all retail milk samples as measured by digital PCR in study of methods to monitor influenza A(H5N1) virus in dairy cattle milk, Massachusetts, USA. A) Concentration of H5N1 as a function of processing state and expiration date. B) RP_Bov data for all samples. The gray-shaded region corresponds to the average RP_Bov concentration of all data +1 SD. Error bars indicate +1 SD.

Figure 5

Virus genome assemblies from retail milk samples in study of methods to monitor influenza A(H5N1) virus in dairy cattle milk, Massachusetts, USA. A) Completeness of H5N1 genome assemblies generated by RNA-Seq, virus-enriched (hsRNA-Seq), and targeted H5N1 Amp-Seq as a function of H5N1 copies per milliliter of RNA. B) The most complete H5N1 assembly produced for each sample sorted by length and the underlying sequencing approach. Asterisks (*) above bars indicate ultrapasteurized samples. Amp-Seq, amplicon sequencing; hsRNA-Seq, hybrid-selected metagenomics; RNA-Seq, unbiased metagenomics.