Molecular Virology

Molecular methods detect a virus by its genome rather than by growing it or finding its proteins. Because a nucleic acid sequence can be chosen to be unique to one virus or shared across a related group, and because amplification can find a handful of target copies in a clinical sample, these assays apply to almost any virus of diagnostic interest, including the many that grow poorly or not at all in culture.

Nucleic acid is also a robust analyte, detectable after transport has destroyed viable virus. For these reasons nucleic acid detection is now the dominant technology in diagnostic virology, having largely displaced culture and antigen detection for most acute diagnoses.

Every molecular assay follows the same logic:

• Extract the nucleic acid from the specimen.
• Amplify a chosen target sequence.
• Detect the product.
• Quantify it, when the question is how much virus is present rather than whether any is.

Everything that follows, the choice of amplification chemistry, real-time or conventional format, single-target or multiplex, benchtop or point-of-care, a yes or no or a full sequence, is a variation on that core.

Nucleic acid extraction and specimen quality

Extraction releases nucleic acid from the virion and the specimen and removes substances that would interfere with amplification.

Manual methods include phenol-chloroform extraction, effective but hazardous because phenol is caustic, and spin-column or silica methods that use safer reagents. They are inexpensive but laborious, and every extra manipulation adds a chance of cross-contamination and operator-to-operator variation.

Automated extraction, usually capture of nucleic acid on magnetic silica beads, has become the norm: recovery is consistent and reproducible, sample handling is minimal, and closed walk-away systems cut contamination risk, at the cost of instrument footprint and consumable expense.

Extraction quality feeds directly into the result. Carryover of ethanol, phenol or salts inhibits the downstream polymerase, and protein contamination degrades yield. Purity is checked spectrophotometrically by the ratio of absorbance at 260 nm to that at 280 nm: a ratio of ~1.8 is taken as clean for DNA and ~2.0 for RNA, and an appreciably lower value points to protein, phenol or other contaminants that absorb near 280 nm. RNA demands particular care because it is rapidly degraded by the ribonucleases present in clinical samples.

Detecting nucleic acid without amplification

Before amplification, viral nucleic acid was found by direct hybridisation of a labelled probe to the target. In situ hybridisation localises nucleic acid within cells or tissue and is the most sensitive of the non-amplified methods, detecting perhaps 10 to 100 molecules in a single cell; Southern blots and dot blots hybridise extracted nucleic acid to a probe.

These methods are still useful where localisation matters, for example showing which cell in a tissue harbours the virus, but as a route to diagnosis they are limited: direct hybridisation needs roughly 10^4 to 10^5 target copies to register. That sensitivity gap is exactly what amplification was invented to close.

Amplification strategies: target, signal and probe

Amplification assays fall into three families by what they multiply, a distinction worth holding because it organises an otherwise long list of named chemistries.

Target amplification makes more of the target sequence itself. The polymerase chain reaction (PCR) is the prototype, joined by the isothermal transcription-based methods nucleic acid sequence-based amplification (NASBA) and transcription-mediated amplification (TMA), by strand-displacement amplification (SDA), and by loop-mediated isothermal amplification (LAMP). These generate detectable product from very few starting copies and are the mainstay of modern diagnosis.

Signal amplification leaves the target untouched and instead multiplies the detection signal in proportion to the amount of target. The branched DNA (bDNA) assay captures the target, then builds a tree of branched probe molecules onto it, each branch carrying many enzyme-labelled probes, so a single target yields a large, quantifiable signal. Hybrid capture detects RNA-DNA hybrids with an antibody and a chemiluminescent readout. Because they do not copy the target, signal-amplification assays avoid amplicon carryover and give stable quantification.

Probe amplification multiplies a probe rather than the target, the ligase chain reaction (LCR) being the classic example, in which adjacent probes are joined only when they match the target and the ligated product is then amplified. Whichever family is used, an RNA target must first be reverse-transcribed into complementary DNA (cDNA) before a DNA-based method can copy it, unless the method is itself RNA-directed.

The polymerase chain reaction

PCR remains the most widely used method because it is simple, sensitive and adaptable to almost any target, with an analytic sensitivity as low as ~1 to 10 copies.

Reagents and their roles

A PCR reaction needs a defined set of components:

• Primers (~18 to 30 nucleotides), two single strands flanking the target: the forward primer defines the 5’ end and the reverse the 3’ end of the region copied.
• A thermostable DNA polymerase such as Taq, which extends the primers in the 5’ to 3’ direction.
• The four deoxynucleotides (dNTPs), the building blocks of the new strand.
• Magnesium, an essential cofactor: too little starves the polymerase and lowers yield, while too much stabilises mismatched priming and undenatured double-stranded DNA, reducing specificity.
• Buffer, nuclease-free water and the extracted template, which complete the reaction.

The thermal cycle

Amplification runs as 20 to 40 repeated cycles, usually preceded by a hot-start initialisation (~95 °C) that activates the polymerase and suppresses non-specific priming during setup. Each cycle has three temperature steps:

• Denaturation (~94 to 98 °C), which melts the double strand into single strands.
• Annealing (~50 to 65 °C), which lets the primers bind.
• Extension (~72 °C), at which Taq polymerises roughly a thousand bases per minute.

A final elongation completes any partial products and a low-temperature hold stores the reaction. Under non-limiting conditions each cycle doubles the target, giving exponential (geometric) amplification.

Melting temperature and stringency

The melting temperature (Tm) is the temperature at which half of a double-stranded sequence has separated into single strands. It is set by base composition: guanine-cytosine pairs share three hydrogen bonds against two for adenine-thymine, so a higher GC content raises the Tm. Annealing is normally run about 3 to 5 °C below the primer Tm; above the Tm fewer primers stay bound, below it more do.

Stringency describes how exact a match hybridisation demands. Raising the temperature or lowering the salt (magnesium) concentration increases stringency, favouring perfectly matched duplexes; lowering the temperature or raising the salt does the reverse. High stringency is chosen for specificity, low stringency where efficient amplification of a variable target matters more. Touchdown PCR exploits both, starting a few degrees above the Tm for specificity and stepping down over successive cycles for efficiency.

Primer design and mismatches

A good primer pair is the single largest determinant of assay performance. Primers are designed to be:

• ~18 to 30 nucleotides long (optimally 20 to 25).
• ~40 to 60% GC content, with a Tm of ~55 to 65 °C matched between the two.
• Free of self- or cross-complementarity that would form primer dimers, and binding both sites with similar efficiency.

RNA viruses pose a particular problem because their sequences vary, so a fixed primer may bind some variants poorly; degenerate or mismatch primers, a mixture varying at defined positions, are used to capture divergent subtypes. Mismatch tolerance is not uniform along the primer: a mismatch at the 3’ end is poorly tolerated, because that is where the polymerase begins extension, while a mismatch at the 5’ end matters less.

Real-time (quantitative) PCR

In conventional PCR the product, or amplicon, is detected only at the end, by gel electrophoresis (with a dye such as ethidium bromide read under ultraviolet light), by labelled probes, or by sequencing. Most diagnostic PCR is now real-time PCR, which amplifies and detects simultaneously by reading a fluorescent signal every cycle. Because the tube is never reopened, real-time formats are fast (often under an hour), reduce contamination, run several targets at once, and report quantity.

Detection chemistries

Two categories of chemistry exist. The first uses a dye that binds any double-stranded DNA, principally SYBR Green, which fluoresces in proportion to product. It is cheap but non-specific, lighting up primer dimers and off-target products as well as the intended amplicon.

Specificity is recovered by melting-curve (dissociation) analysis after amplification: as the temperature rises, each product releases the dye at a melting point fixed by its length and sequence, so artefacts resolve from the true amplicon, and two targets with different melting points can be told apart in one reaction, as when separating herpes simplex virus types 1 and 2.

The second category uses sequence-specific fluorogenic probes, which confirm the amplicon’s sequence and so add specificity beyond the primers. The dominant format is the TaqMan (5’ hydrolysis) probe, carrying a reporter dye and a quencher; while both are on the intact probe the quencher suppresses fluorescence, but during extension the 5’ exonuclease activity of Taq cleaves the reporter free, releasing signal in proportion to product. Hybridisation (FRET) probes and molecular beacons are the other main designs.

Fluorescence resonance energy transfer (FRET) is the non-radioactive transfer of energy between two fluorophores held within ~10 to 100 ångström of each other with overlapping spectra, so signal arises only when probe binding brings the dyes into proximity.

Advantages and limitations

Real-time PCR is fast, highly sensitive and specific, automated, and quantitative, and its closed-tube operation reduces contamination. Against this, the instrument and probe reagents are expensive, and setup, optimisation and the interpretation of borderline or discordant results demand more technical skill than a simple qualitative assay.

Quantification and viral load

Real-time PCR quantifies because the cycle at which fluorescence crosses a set threshold, the cycle threshold (Ct), falls as starting target rises: a difference of three cycles corresponds to roughly an eightfold (2^3) difference in template. Reading an absolute concentration from a Ct requires a standard curve of known quantities, on which the log of concentration is plotted against Ct and linear regression gives the line used to interpolate unknowns. Digital PCR sidesteps the standard curve by partitioning the reaction into thousands of compartments and counting the positive fraction to give an absolute count directly.

The slope of the standard curve also gives the PCR efficiency, ideally ~90 to 110%, corresponding to near-perfect doubling each cycle; a value outside that band signals a problem, and exact efficiency is harder to reach in multiplex assays where targets compete. Efficiency is degraded by inhibitors, nucleic acid degradation, primer-template mismatch and suboptimal chemistry or cycling.

To make loads comparable between laboratories and platforms, results are reported against World Health Organization international standards in international units per millilitre (IU/mL) rather than raw copies, because the copy-to-IU relationship differs by assay. Quantitative testing underpins viral load monitoring in human immunodeficiency virus (HIV), hepatitis B and hepatitis C, and surveillance of cytomegalovirus (CMV) and other viruses after transplantation, where a change in load is interpretable only against this shared scale and the assay’s own limit of quantification.

Controls, contamination and inhibition

The sensitivity that lets PCR find ten copies also lets it find contaminating copies and be defeated by inhibitors, so controls and clean workflow are part of the method.

Controls

Every run carries several controls, each guarding a different failure:

• Positive control, which mimics the specimen and is set just above the limit of detection, so it challenges the assay where it is least sensitive and flags loss of efficiency.
• Negative (extraction) control, which detects contamination introduced during extraction.
• No-template control, which detects contamination of the reagents themselves.
• Internal control, a separate target co-amplified in the reaction to detect inhibition (and, when added before extraction, extraction failure). It is endogenous (a housekeeping gene such as albumin or β-globin) or exogenous, the latter either homologous (same primers, different probe) or heterologous (its own primers and probe).

Contamination control

The characteristic danger of a highly sensitive amplification assay is the false positive from carryover contamination, either amplicon from a previous run or high-titre target from another specimen. The core defence is a unidirectional workflow through physically separated areas, ideally separate rooms, for reagent preparation, specimen addition and post-amplification steps, with dedicated equipment and pipettes that never move upstream. Supporting measures include:

• Aerosol-resistant filter tips and small single-use reagent aliquots.
• Ultraviolet decontamination of the work cabinet.
• Enzymatic carryover control: incorporate uracil in place of thymine, then destroy any carried-over uracil-containing amplicon with uracil-N-glycosylase before the next reaction.
• Routine negative controls, and especially strict pre- and post-amplification separation for nested PCR.

Real-time and closed-cartridge formats reduce the risk structurally by never reopening the tube.

Inhibition and interpretation

The mirror failure is the false negative, whose commonest cause after a mistimed or wrong specimen is inhibition by substances in the sample such as blood, heparin, salts, detergents or residual ethanol and phenol. The internal control catches it: if the internal control fails while the viral target is negative, the result is uninterpretable (inhibited), not a true negative, and the sample must be re-extracted, diluted or repeated, sometimes with additives such as bovine serum albumin or a more inhibitor-resistant polymerase.

Other causes of a false-negative result, worth rehearsing for a low-volume specimen such as cerebrospinal fluid (CSF), include sampling too early or too late, low target load, primer-site sequence variation, and RNA degradation.

Interpretation extends beyond the controls to meaning: detecting viral nucleic acid is not always evidence of disease. Viral DNA may not separate latent from active infection unless a productively expressed messenger RNA is targeted, and some findings mislead outright, such as the persistently very high blood DNA of chromosomally integrated human herpesvirus 6 (HHV-6), which is not active infection. A significant or unexpected result is confirmed on a fresh sample before it is acted on.

Multiplex and syndromic panels

Multiplex PCR puts several primer sets in one reaction to detect several targets at once. Its clinical expression is the syndromic panel: a single respiratory, meningitis-encephalitis or gastrointestinal cartridge that tests one specimen for a dozen or more viral and often bacterial causes of that syndrome in about an hour. The appeal is broad, fast coverage from one sample when the differential is wide and empirical treatment or isolation cannot wait.

The challenges are equally real. Packing many reactions into one tube risks competition between targets and reduced sensitivity for any single one compared with a dedicated assay. Broad panels detect organisms of uncertain clinical significance, prolonged shedding or incidental carriage that may not be causing the presentation, so a positive needs interpreting against the clinical picture rather than treated as automatically causal.

Cost per test is high. The practical resolutions are to match the panel to the clinical question, to confirm or quantify a key positive with a targeted assay where it changes management, and to report results with interpretative comment rather than as a bare list.

Point-of-care and integrated platforms

Integrated, sample-to-answer platforms carry out extraction, amplification and detection inside a single closed cartridge with minimal hands-on time. The widely used example is the GeneXpert cartridge system, simple enough to run outside a specialised molecular laboratory and sometimes near the patient. Larger integrated analysers extend the same principle to high throughput, returning hundreds to a thousand results a day for viruses such as HIV, hepatitis B and C, human papillomavirus and respiratory panels.

Taking testing closer to the patient is helped by isothermal amplification, which works at a single temperature and needs no thermal cycler. Transcription-based methods (TMA and NASBA) mimic retroviral replication to amplify RNA, while loop-mediated isothermal amplification (LAMP) uses several primers and a strand-displacing polymerase to make a large amount of product very quickly, read by turbidity or a dye. Coupling amplification to CRISPR-based detection adds a further layer of sequence-specific confirmation and is an active area of point-of-care development.

Sequencing, genotyping and resistance testing

Reading the actual sequence, rather than just detecting or counting it, answers questions amplification alone cannot.

Sanger sequencing

Sanger (chain-termination) sequencing remains the reference for a defined region. It runs a cycling reaction with a single primer and a mix of normal dNTPs and fluorescently labelled dideoxynucleotides (ddNTPs); whenever a ddNTP is incorporated, extension stops because it lacks the 3’ hydroxyl needed for the next bond. The result is a set of fragments of every length, each ending in a colour-coded base, which are separated by size on capillary electrophoresis and read past a laser and detector to reconstruct the sequence.

A single run reads roughly 750 to 800 bases, so longer genes are covered with internal primers and confirmed in both directions.

Next-generation sequencing

Next-generation sequencing (NGS) reads millions of fragments in parallel and has become central to the field. Its depth lets it detect minority variants that make up a small fraction of the viral population, resolve mixed infections, and sequence whole genomes for surveillance. The trade-off is a more demanding workflow and bioinformatic analysis, and validating an NGS assay to clinical standard, for example for HIV resistance, is correspondingly harder than validating a single-target PCR.

Genotyping and resistance

Sequence data identifies a virus precisely, genotypes it, and detects antiviral resistance mutations. Resistance testing is genotypic or phenotypic.

Genotypic assays sequence the target gene and read off known mutations: the HIV polymerase, protease and sometimes integrase genes; the CMV UL97, UL54 and UL56 genes; the hepatitis B polymerase. They are fast and widely available but interpretable only where each mutation’s effect is known, and a single mutation may confer full, partial or no resistance.

Phenotypic assays instead measure viral replication across drug concentrations; they are slower, less standardised and mostly confined to reference laboratories, but they can assess resistance when the genetic basis is unknown or when multiple interacting mutations make a genotype hard to read.

Bioinformatics and phylogenetics

Making sense of sequence is a bioinformatics task. Alignment tools such as BLAST compare reads to reference databases to identify them, and phylogenetic analysis places sequences on a tree, the basis of molecular epidemiology. The genetic distance between sequences and the bootstrap support for a branch are used to judge whether isolates form a genuine transmission cluster, to investigate unexpected relatedness, and to separate true infection from laboratory cross-contamination in a low-prevalence setting where a single positive may need sequence confirmation.

Virus discovery and metagenomics

Classic virus discovery relied on electron microscopy, cell culture and antibody-based detection, each limited: electron microscopy needs high titres and rarely identifies a virus unambiguously, many viruses cannot be cultured and no single cell line supports all of them, and antibody methods need a reagent and a candidate chosen in advance. Molecular methods developed to circumvent these, including degenerate and consensus PCR and representational difference analysis, led to the discovery of clinically important viruses such as hepatitis C virus and Kaposi sarcoma herpesvirus.

The genomic era has since transformed the field through metagenomics: largely unbiased sequencing of everything in a specimen, then computational sorting of which reads are viral. The central difficulty is that viral nucleic acid is a tiny fraction against a vast excess of host genome, so enrichment is used to raise the ratio, by culturing the virus, physically purifying virions, subtractive hybridisation to reduce host sequence, or random sequence-independent amplification to boost scarce material.

The output is handled by high-throughput bioinformatic pipelines that align reads against databases with tools such as BLAST and classify them. This approach identified the SARS coronavirus as a novel coronavirus and discovered agents such as human bocavirus and Merkel cell polyomavirus, and it now supports outbreak investigation where the cause is unknown.

References and recommended reading

• Whiley DM, Sloots TP; Revill PA, Bowden DS; Hoffman NG, Roshal M. Molecular Amplification Methods in Diagnostic Virology; Viral Genotyping and the Sequencing Revolution; Design of Molecular Virologic Tests. In: Jerome KR, editor. Lennette’s Laboratory Diagnosis of Viral Infections, 4th edition, Chapters 3-5. Informa Healthcare; 2010. The conceptual scaffold for nucleic acid extraction, amplification chemistry, real-time detection, quantification, sequencing and molecular test design.
• Greninger AL, Wang D, Storch GA, Jerome KR. Diagnostic Virology. In: Fields Virology, 7th edition, Volume 4 (Fundamentals), Chapter 16. Wolters Kluwer; 2023. The current account of molecular platforms, real-time and multiplex formats, point-of-care and integrated systems, resistance testing, sequencing, metagenomics and virus discovery.
• Jeffery K, Aarons E. Diagnostic Approaches. In: Principles and Practice of Clinical Virology, 6th edition, Chapter 1. Wiley-Blackwell; 2009. The reference for how molecular methods are selected and interpreted alongside the other diagnostic approaches.