Overview of Molecular Epidemiological Methods for the Subtyping and Comparison of Viruses

1. Overview of Molecular Epidemiological Methods for the Subtyping and Comparison of Viruses An Overview

2. Uses of molecular epidemiological methods • Subtyping - in some viruses, different subtypes are associated with different clinical manifestations e.g. enteroviruses, adenoviruses, and human papillomaviruses. • General Epidemiology - by identifying the viral subtypes at different times and geographical locations, one can detect major changes in the epidemiological patterns of infection e.g. HIV and HCV. • Investigation of Outbreaks - to support or disprove a link between the donor and recipient viruses e.g. HIV, HBV, HCV, Norwalk virus.

3. Methods Used - Complete or Partial genome? • For greatest degree of accuracy, the complete genome should be used for the purpose of comparison. • However, since viral genomes ranges from 3500 bp to over 200,000 bp, it would be highly impractical to sequence the whole genome. • Certain simple methods are still used for the comparison of complete genomes e.g. RFLP for CMV, HSV, and Adenoviruses. • Nowadays in practice, a small part of the genome is amplified first by PCR and the product investigated by sequencing or other methods.

4. Strategies for identification of the PCR Product (Commonly used methods) • Sequencing of the PCR product • the gold standard but expensive and not widely available. • PCR product may be sequenced directly or cloned before sequencing. • However, it is the test of choice in outbreak situations where there are serious public health and/or medical-legal implications. • Sequencing can be used to confirm results of other molecular epidemiological assays. As a matter of fact, all other assays can be considered as simpler screening assays. • Restriction Fragment Length Polymorphism (RFLP) - very simple, rapid and economical technique but the result may be difficult to read. • Hybridization with a specific oligonucleotide probe - A wide variety of formats is available e.g. dot-blot, Southern blot, reverse hybridization, DNA enzyme immunoassay etc.

5. Principles behind Restriction Enzyme Analysis and Hybridization Probes REA EcoRI (GAATTC) Target 0 32 100 32 68 Hybridization Probes GAATTC Target

6. PCR-RFLP (PRA) • The gene target must be present in all viral strains. • It is amplified with primers directed against conserved areas in the target gene so that all subtypes can be amplified. • The PCR product is then digested with one or more restriction enzymes and on an agarose or polyacrylamide gel. • The species or genotype is identified from the restriction patterns seen. • Therefore PRA can be considered as probably the simplest DNA fingerprinting technique. • The principle of PRA is similar to that of RFLP of whole viral genomes and pulse field gel electrophoresis. • It is quick, simple and cheap and this is why it is preferred by many molecular biologists. • Examples include HCV genotyping and identification of mycobacteria.

7. Nature of Restriction Enzymes • 4-cutter Enzymes (frequency of cutting = 1/256) taq 1 TCGA Hae III GGCC Sau 96I GGNCC • 6-cutter Enzymes (frequency of cutting = 1/4096) Eco RI GAATTC Hind III AAGCTT • 8-cutter Enzymes (frequency of cutting = 1/65536) Not I GCGGCCGC

8. Specific Oligonucleotide Probe • Simple to carry out, particularly suitable for large scale testing • Results are usually easier to read than REA and requires less skill to interpret • Preferred strategy by commercial companies e.g. INNO-LIPA HCV, Sorin DEIA, Roche Amplicor and Taqman. • Can be made into a highly automated closed system e.g. Roche-Amplicor. • Therefore more attractive than PRA for the routine laboratory but the costs could be prohibitive. • Specific nucleic acid probe assays are available where the specimen is tested directly without amplification. However the sensitivity is much lower.

9. Choice of Genomic Region • The choice of genomic region to use for analysis is critical and could affect the outcome of results. • Too conserved – will not be able to demonstrate any differences between subtypes. • Too variable – may not be able to demonstrate a link between source and recipient viruses in outbreak studies because of the high mutation rate. • In general, RFLP is not suitable to highly conserved regions while nucleic acid probes are not suitable for highly variable regions. • It is often advisable to use more than one gene region, especially where there are serious medical-legal implications.

10. DNA Sequencing • DNA Sequencing is a gold standard of molecular virological investigation. • Automated Sequencing facilities are now used in many routine diagnostic laboratories. • The PCR product is directly sequenced without cloning. • A Blast search of the sequence is then carried out in GENBANK. This is normally good enough for identification. • For epidemiological investigations, a phylogentic analysis should be carried out.

11. GENBANK • GENBANK is the most important bioinformatics resource – may be accessed through Entrez or Blast searches. • However, there are a number of problems. • Sequences not refereed • Many sequences are very old and obtained by manual sequencing techniques that were not very accurate. • Many sequences are from strains of microorganisms that had not been well characterized. • A lot of the sequences deposited in Genbank are contaminated with plasmid vector sequences or PCR primer sequences. • Therefore there is a need for other well validated databases.

12. Phylogenetic Analysis • DNA sequences of outbreak strains are compared to those of the suspected source strains and reference strains. • A phylogenetic tree is drawn up with bootstrap resampling analysis. • Where the outbreak strains and source strains are similar, they should be close together on the tree with a high bootstrap value. • The actual gene that is used and the length that is sequenced is critical. • Too conserved – will not be able to demonstrate any differences between subtypes. • Too variable – may not be able to demonstrate a link between source and recipient strains in outbreak studies because of the high mutation rate. • Most useful for RNA viruses such as HIV, HCV and Norwalk where there is a lot of variation, but less useful for DNA viruses, bacteria, fungi and protozoa.

13. Investigation of Norovirus Outbreak Patients 1, 2, 3.

14. Summary • A wide variety of molecular epidemiological methods are available, of which DNA sequencing is the gold standard. • It is now usual to analyze a small part of the genome rather than the complete genome. The target fragment is first amplfied by PCR before analysis. • The most widely used screening methods involve either restriction enzyme analysis or hybridization with specific nucleic acid probes, or a combination of the two. • Other screening methods such as SSCP, dHPLC and other heteroduplex analysis techniques are rarely used outside a research setting because they often suffer from poor inter-laboratory reproducibility. • The choice of the genomic region to use is critical: it is often advisable to use more than one genomic region. • It is important to remember that all molecular epidemiological methods available for viruses can be applied to bacteria but not vice-versa.

15. Points to Consider • Molecular epidemiology techniques are can be used to good effect to disprove a link between donor and recipient strains of a particular infectious agent but they cannot prove a definitive link. • Therefore a negative result is much greater predictive value than the positive result. • The probability of a link depends on many factors including the prevalence of that particular genotype and the methods used. • Where the outbreak carries huge medical-legal implications e.g. HIV transmitted through blood factors, the case would have to be argued on an individual basis in court, preferably with the help of a statistician. • It is important to remember that molecular epidemiological investigation does not replace a good basic epidemiological investigation