A Multiplex PCR/LDR Assay for Viral Agents of Diarrhea with the Capacity to Genotype Rotavirus

Abstract

Rotavirus and noroviruses are major causes of diarrhea. Variable rotavirus vaccination efficacy in Africa and Asia is multifactorial, including the diversity of circulating strains and viral co-infection. We describe a multiplexed assay that detects and genotypes viruses from stool specimens. It includes a one-step reverse transcriptase PCR reaction, a ligase detection reaction (LDR), then hybridization of fluorescent products to micro-beads. In clinical samples it detects rotavirus, caliciviruses (sapovirus and norovirus), mixed infections, and genotypes or genogroups of rotaviruses and noroviruses, respectively. The assay also has the capacity to detect hepatitis A. The assay was validated on reference isolates and 296 stool specimens from the US and Ghana. The assay was 97% sensitive and 100% specific. The genogroup was concordant in 100% of norovirus, and the genotype in 91% and 89% of rotavirus G- and P-types, respectively. Two rare rotavirus strains, G6P[6] and G6P[8], were detected in stool specimens from Ghana. The high-throughput assay is sensitive, specific, and may be of utility in the epidemiological surveillance for rare and emerging viral strains post-rotavirus vaccine implementation.

Figure 1

Phylogenetic tree of VP7 gene sequences from G6 rotavirus isolates. *Indicates isolates identified in the current study. Both of these isolates cluster with those recently reported from the geographically proximal country of Burkina Faso in the study of Nordgren et al.. The EF554087.1BEL/B1711/2002 isolate is that reported from Mali.

Figure 2

Schematic of the PCR/LDR assay followed by hybridization to the VeraCode™ micro-beads and detection using the BeadXpress platform (Illumina Inc., San Diego, CA). Virus-specific PCR primer pairs are used to amplify distinct genetic targets (one for Sapovirus and 2 each for Rotavirus and Norovirus) in a multiplex reaction. Each PCR amplicon ranging in size from 300 to 500 base pairs is then subjected to LDR using primer pairs to identify SNPs at multiple locations along the amplicon. This allows the identification of the viruses and the differentiation of the P and G genotypes of rotavirus as well as genogroup I and II of norovirus. At any given SNP, the allele specific upstream LDR primers are designed to ligate to the downstream primer only if there is a perfect match at the junction point. The upstream LDR primers bear zipcode complement sequences and amino blocking groups on the 5′-end, while the downstream LDR primers have a Cy3 fluorescent label at the 3′end. Ligation of the LDR primers results in fluorescently labeled products that are subsequently hybridized to zipcode addresses attached to VeraCode™ micro-beads. Each micro-bead contains a unique barcode that can be scanned by the BeadXpress reader. Detection of the LDR products hybridized to the specific micro-beads is accomplished by the BeadXpress reader that identifies the micro-beads by their barcode and the scores the associated fluorescent signal thus identifying the pathogen.

Figure 3

Scheme for the detection, identification, and genotyping of enteric viruses. A pattern of positive zipcode signals are used to detect and identify viruses as well as determine the genotypes of noroviruses and rotaviruses. Representative patterns of zipcode signals from 6 samples and their interpretation are shown. The normalized signal to background fluorescent signal (S/B) for each zipcode target for a given sample is indicated. An S/B value >3 fold higher than the median signal of negative controls for a given zipcode target was defined as a positive signal. The assay has built-in redundancy with at least 3 zipcode addresses assigned to identify each virus. Only 2 (for calicviruses) or 3 (for rotaviruses) positive signals are required for the identification of a virus or genotype. For example, sample 3 was identified as norovirus GGII based on 2 positive zipcode signals (9 and 12) out of the 4 designated zipcode addresses for norovirus (9, 10, 11 and 12). Similarly, samples 5 and 6 were determined as rotavirus P[6] based on a unique zipcode signature of 2 out of 3 zipcode addresses assigned (See Supplementary Fig. 1 for details of genotype determination of rotavirus). ^†Designated zipcodes did not provide positive signals as indicated in red; however, identification was achieved successfully since the criteria of 2 out of 4 positive zipcodes was met.