Evaluation of positive Rift Valley fever virus formalin-fixed paraffin embedded samples as a source of sequence data for retrospective phylogenetic analysis

Highlights

• The aim of the study was to evaluate the potential of RVFV positive FFPE samples as a source of sequence data for retrospective analysis.

• A RT-PCR targeting a 490 nt portion of the G_N glycoprotein gene of RVFV was applied to total RNA extracted from RVFV positive FFPE samples.

• Attempts to obtain target amplicons were unsuccessful. FFPE samples were then analysed using next generation sequencing.

• Using reference mapping, gapped virus sequence data of varying degrees of shallow depth was aligned to a reference sequence.

• However, the NGS did not yield long enough contigs that consistently covered the same genome regions in all samples to allow phylogenetic analysis.

Abstract

Rift Valley fever (RVF), caused by an arthropod borne Phlebovirus in the family Bunyaviridae, is a haemorrhagic disease that affects ruminants and humans. Due to the zoonotic nature of the virus, a biosafety level 3 laboratory is required for isolation of the virus. Fresh and frozen samples are the preferred sample type for isolation and acquisition of sequence data. However, these samples are scarce in addition to posing a health risk to laboratory personnel. Archived formalin-fixed, paraffin-embedded (FFPE) tissue samples are safe and readily available, however FFPE derived RNA is in most cases degraded and cross-linked in peptide bonds and it is unknown whether the sample type would be suitable as reference material for retrospective phylogenetic studies. A RT-PCR assay targeting a 490 nt portion of the structural G_N glycoprotein encoding gene of the RVFV M-segment was applied to total RNA extracted from archived RVFV positive FFPE samples. Several attempts to obtain target amplicons were unsuccessful. FFPE samples were then analysed using next generation sequencing (NGS), i.e. Truseq^® (Illumina) and sequenced on the Miseq^® genome analyser (Illumina). Using reference mapping, gapped virus sequence data of varying degrees of shallow depth was aligned to a reference sequence. However, the NGS did not yield long enough contigs that consistently covered the same genome regions in all samples to allow phylogenetic analysis.

Introduction

Rift Valley fever virus (RVFV) is a member of the genus Phlebovirus in the family Bunyaviridae with a tripartite single stranded RNA (ssRNA) genome (Ikegami et al., 2005). It is a zoonotic pathogen that causes potentially severe disease in humans and animals and therefore requires biosafety level (BSL) 3 facilities for isolation (Pepin et al., 2010). The ssRNA genome of the virus is vulnerable to environmental degradation. Fresh and frozen tissues are therefore in general the preferred sample source for the extraction of intact RNA for molecular analysis. However, there is a high cost associated with storage and preservation of large numbers of frozen samples for a prolonged period (Joon-Yong et al., 2006) resulting in limited availability of such samples for retrospective studies. With the exception of cases where the virus needs to be isolated, most laboratories receiving clinical or post-mortem samples routinely have such samples fixed in 10% neutral buffered formalin (McKinney et al., 2009).

Although formalin-fixed and paraffin-embedded (FFPE) material still remains the optimal sample material for microscopic tissue-based techniques such as immunohistochemistry and histopathology, archived FFPE samples are potentially an alternative source of nucleic acids for molecular analysis and pathogen identification (Adam, 2012). However, the use of molecular diagnostic techniques to identify pathogens and obtain sequences sufficient enough for phylogeny in these samples is still challenging (Bhoopat et al., 1996). Formalin fixation has detrimental effects on the quality and yield of RNA recovered from FFPE material and there is a need for optimization of nucleic acid extraction protocols (Sharma et al., 2012). The effects of formalin on sample material include RNA fragmentation, methylation of nucleotides through the addition of mono-methylol groups (–CH₂OH), and formalin-induced protein-nucleic acid cross linkages (Joon-Yong et al., 2006, Masuda et al., 1999).

Whilst the challenges in obtaining sequences from FFPE material through RT-PCR and Sanger sequencing exist, shotgun or direct sequencing using next generation sequencing (NGS) platforms potentially offer an alternative approach. The NGS platform allows for deep sequencing that potentially increases the likelihood of obtaining sequences long enough for phylogenetic comparison. However, a challenge associated with a shotgun NGS approach is host nucleic acid contamination and viral RNA being present in very low amounts compared to the host nucleic acids leading to insufficient coverage depth for the virus of interest (Marston et al., 2013, Pop, 2009).

The objective of this study was to obtain a 490 nt region of the G_N encoding gene on the M-segment of RVFV previously described (Grobbelaar et al., 2011) by RT-PCR and NGS. This was mainly aimed at evaluating the potential use of FFPE archived samples as a source of sequence data for use in retrospective phylogenetic analysis. A large number of RVF samples representative of previous outbreaks in South Africa are readily available at the pathology laboratory at the University of Pretoria. If sequence data of sufficient length are obtained from these samples, then phylogeography analyses of all previous outbreaks would be feasible in the effort to elucidate the transmission dynamics of the virus in country.