Reverse Transcriptase in HIV: Structure, Function, and Drug Targeting

HIV reverse transcriptase (RT) is the enzyme that converts the virus’s single-stranded RNA genome into double-stranded DNA, a step without which HIV cannot integrate into a host cell’s chromosomes or establish a lasting infection. Because no human cell needs to perform this RNA-to-DNA trick during normal operation, RT has been one of the most intensively targeted proteins in antiretroviral therapy since the epidemic began. Its unusual architecture, its notorious sloppiness when copying genetic material, and the clever ways drugs exploit its vulnerabilities all help explain both the successes and the ongoing challenges of treating HIV.

A Lopsided Pair of Subunits

Mature HIV-1 RT is a heterodimer, meaning it is built from two protein chains of different sizes that lock together. The larger chain, called p66, contains all the catalytic machinery the enzyme needs. The smaller chain, p51, is actually made from the same gene but is processed into a shorter form. Despite sharing much of its amino acid sequence with p66, p51 folds into a distinctly different shape and contributes no enzymatic activity of its own. Biophysical studies using electron spin-resonance techniques confirm that this structural asymmetry is already present even in precursor forms of the enzyme before the final processing step occurs, with p51 serving purely as a structural scaffold that stabilizes the active p66 subunit.¹

The p66 subunit is often described using a right-hand analogy: its polymerase domain has “fingers,” “palm,” and “thumb” subdomains that together cradle the nucleic acid template during copying. The palm houses the catalytic residues that actually join new DNA building blocks to the growing chain, while the fingers help position incoming nucleotides and the thumb grips the double-stranded product as it slides through. A separate connection subdomain links the polymerase region to a second catalytic site at the far end of p66, the RNase H domain. This elongated shape means that when RT is sitting on a stretch of RNA-DNA hybrid, the polymerase end and the RNase H end contact the nucleic acid at points roughly 17 to 18 base pairs apart.

Two Catalytic Activities in One Enzyme

RT is unusual because it performs two chemically distinct jobs. The polymerase active site reads the viral RNA template and assembles a complementary DNA strand, while the RNase H active site chews up the RNA strand of the resulting RNA-DNA hybrid so that a second DNA strand can be synthesized. Both activities are essential for converting the single-stranded RNA genome into the double-stranded DNA that eventually integrates into the host chromosome.²

The polymerase active site sits in the palm subdomain of p66 and is defined by a trio of aspartic acid residues at positions 110, 185, and 186, which are strictly conserved across retroviral reverse transcriptases.³ These residues coordinate metal ions (typically magnesium) that catalyze the chemical bond formation between an incoming nucleotide and the primer strand. Nearby, a highly conserved four-amino-acid sequence known as the YMDD motif (tyrosine-methionine-aspartate-aspartate at positions 183 through 186) plays a structural and functional role in recognizing the template. The tyrosine at position 183 makes hydrogen bonds with nucleotide bases in the template-primer duplex, helping the enzyme “read” which nucleotide should be added next.⁴

The RNase H domain, located at the opposite end of p66, is a nuclease that specifically cleaves the RNA strand in an RNA-DNA hybrid. Not all RNA is degraded equally, though. A short stretch of the viral genome called the polypurine tract (PPT) resists RNase H cleavage because its rigid structure is accommodated outside the nuclease active site rather than being threaded into it.² The PPT survives degradation deliberately: it serves as the primer for second-strand DNA synthesis. This selective preservation is one of many reminders that HIV’s replication machinery, while sloppy in some respects, is precisely tuned in others.

How Reverse Transcription Gets Started

RT cannot begin copying the viral RNA from scratch. It needs a primer, a short piece of nucleic acid already base-paired to the template to give the enzyme a starting point. HIV hijacks a host cell molecule for this purpose: a transfer RNA called tRNA-Lys3, which is packaged into the virus particle during assembly. This tRNA anneals to a specific sequence on the viral RNA called the primer binding site, and RT extends from there.⁵

The process that follows is intricate. RT first copies a short stretch near the 5′ end of the RNA, producing what is called “minus-strand strong-stop DNA.” A strand transfer then repositions this short DNA to the 3′ end of the genome, where synthesis continues along the full length. RNase H activity degrades the RNA template behind the advancing polymerase, and eventually the PPT primes second-strand (“plus-strand”) DNA synthesis in the opposite direction. A second strand transfer completes the process, yielding a linear double-stranded DNA molecule flanked by long terminal repeats that will be recognized by the viral integrase enzyme.

An Error-Prone Copying Machine

One of RT’s most consequential features is its poor accuracy. Unlike the DNA polymerases that replicate your own genome, RT has no proofreading ability, meaning it cannot back up and correct a mismatch after incorporating the wrong nucleotide. Early measurements using laboratory fidelity assays found that HIV-1 RT made roughly one error for every 1,700 nucleotides incorporated, making it exceptionally inaccurate even compared to other retroviral reverse transcriptases.⁶

Given that the HIV genome is about 9,700 nucleotides long, this error rate means that virtually every new copy of the genome carries at least a few mutations. Most of these mutations are harmful to the virus and produce nonviable offspring, but the sheer volume of viral replication (billions of new particles per day in an untreated person) means that a small fraction of these errors land on residues that confer drug resistance or immune evasion. This built-in diversity is the fundamental reason HIV is so difficult to cure: the virus is constantly exploring new genetic territory, and any drug or immune response that fails to suppress replication completely will select for resistant variants.

Where Reverse Transcription Happens Inside the Cell

For decades, the standard textbook picture placed reverse transcription in the cytoplasm: the virus was thought to shed its capsid shell shortly after entering the cell, releasing RT and the RNA genome to work in the cytoplasm before the resulting DNA was imported into the nucleus. Recent experiments have overturned that model. Studies using inducible blockade of nuclear pore complexes showed that the intact or nearly intact viral core is imported into the nucleus relatively quickly, within about five hours, and that reverse transcription is actually completed inside the nucleus rather than in the cytoplasm.⁷

Complementary work confirmed that reverse transcription intermediates are heavily enriched in nuclear fractions compared to cytosolic ones, consistent with the idea that most of the copying work occurs after nuclear entry.⁸ This revised understanding matters for drug design because it changes where and when RT-targeting drugs need to be active. It also raises questions about how the large capsid structure navigates through nuclear pores and what signals trigger its eventual disassembly once inside.

NRTIs and Chain Termination

The first class of drugs approved against HIV targeted reverse transcriptase, and nucleoside reverse transcriptase inhibitors (NRTIs) remain a backbone of modern antiretroviral regimens. NRTIs are synthetic analogs of the natural nucleotide building blocks that RT uses to assemble DNA, but they are missing a critical chemical group: the 3′-hydroxyl on the sugar ring. This hydroxyl is the attachment point for the next nucleotide in the growing chain. When RT incorporates an NRTI instead of a genuine nucleotide, DNA synthesis stalls because there is no hydroxyl for the next building block to bond to.⁹

NRTIs must be activated inside the cell before they work. They enter as inactive prodrugs and undergo phosphorylation by host cell enzymes, ultimately being converted into their triphosphate forms, which are the versions that RT recognizes and incorporates. This activation requirement is one reason NRTIs can have different potencies in different cell types: cells with lower levels of the necessary kinases may activate the drug less efficiently.

NNRTIs and Allosteric Inhibition

Non-nucleoside reverse transcriptase inhibitors (NNRTIs) work by an entirely different mechanism. Instead of masquerading as building blocks, they bind to a hydrophobic pocket in the p66 subunit located roughly ten angstroms from the polymerase active site. This pocket does not exist in the unbound enzyme in a fully formed state; it opens up when the inhibitor binds, which is part of why it was not immediately obvious as a drug target. By lodging in this pocket, NNRTIs act as noncompetitive inhibitors: they do not compete with incoming nucleotides for space but instead distort the enzyme’s structure so that it can no longer catalyze the chemical reaction efficiently.¹⁰

Hydrogen-exchange experiments have revealed that the effects of NNRTI binding ripple far beyond the binding pocket itself. When efavirenz, one of the most widely used NNRTIs, binds to RT, molecular flexibility is reduced across multiple regions of the enzyme, including four distinct areas of allosteric stabilization and one area that paradoxically becomes more flexible.¹⁰ This widespread stiffening essentially freezes RT in a catalytically incompetent conformation. Six NNRTIs have been approved for clinical use: nevirapine, delavirdine, efavirenz, etravirine, rilpivirine, and doravirine.¹¹ The newer members of this group, particularly etravirine and rilpivirine, were specifically designed with flexible chemical scaffolds that can adapt to mutations in the binding pocket, addressing the vulnerability of earlier NNRTIs to resistance.

How Resistance Develops Against RT-Targeting Drugs

Because RT is so error-prone, resistance mutations arise frequently. The virus uses two main biochemical strategies to evade NRTIs. The first is excision: certain mutations, most famously a cluster called thymidine-analogue mutations, give RT the ability to use ATP as a chemical tool to remove an already-incorporated NRTI from the end of the DNA chain, essentially reversing the chain-termination event and allowing synthesis to resume.¹²

The second strategy is discrimination: mutations at or near the nucleotide-binding site alter the shape of the pocket so that RT can better distinguish between genuine nucleotides and the drug analogs, preferentially incorporating the real thing. The M184V mutation, one of the most commonly selected by the drugs lamivudine and emtricitabine, is a textbook example. Replacing methionine with valine at position 184 introduces steric hindrance that either blocks the drug’s triphosphate from entering the binding site entirely or allows it in only in an orientation that prevents incorporation.¹³ Interestingly, M184V also reduces RT’s overall copying efficiency, which is why some treatment strategies deliberately allow this mutation to persist: the fitness cost it imposes on the virus can partly compensate for the loss of one drug’s activity.

NNRTI resistance tends to develop through mutations in or around the hydrophobic binding pocket. Because the pocket is not essential for the enzyme’s normal catalytic function, mutations there can disrupt drug binding without greatly impairing RT’s ability to copy DNA. Single-point mutations like K103N or Y181C can cause high-level resistance to first-generation NNRTIs like nevirapine and efavirenz. The newer NNRTIs were engineered with enough conformational flexibility to maintain contact with the pocket even when one or two residues change, but accumulation of multiple mutations can still erode their effectiveness.

A Newer Twist on Chain Termination

A more recent class of RT inhibitor does something no earlier drug achieved: it terminates DNA synthesis even though it retains the 3′-hydroxyl group that NRTIs lack. Islatravir (also known by its chemical name EFdA, 4′-ethynyl-2-fluoro-2′-deoxyadenosine) belongs to a class called nucleoside reverse transcriptase translocation inhibitors (NRTTIs). After RT incorporates islatravir into the growing DNA chain, a small chemical group on the sugar, a 4′-ethynyl substituent, wedges into a hydrophobic pocket on the enzyme and physically prevents it from sliding forward along the template to the next position.¹⁴ Without this translocation step, no new nucleotide can be loaded, and DNA synthesis halts.

Structural studies have clarified the mechanism in detail. Islatravir retains its 3′-hydroxyl, which means it looks even more like a natural nucleotide than traditional NRTIs do, contributing to its extremely high potency. The key is that after incorporation, the 4′-ethynyl group locks RT in the pre-translocation state, preventing the shift from the nucleotide-binding site to the primer-binding site that normally occurs after each addition.¹⁵ Because the mechanism of inhibition is fundamentally different from standard chain termination, islatravir shows a high barrier to resistance and retains activity against many NRTI-resistant viral strains.¹⁶

Targeting Both Active Sites at Once

Current approved drugs hit either the polymerase activity (NRTIs, NNRTIs, NRTTIs) or neither active site directly (integrase inhibitors, protease inhibitors target other viral enzymes). The RNase H domain, despite being essential for viral replication, has no approved inhibitor. That gap has motivated a search for dual inhibitors: single molecules designed to block both the polymerase and RNase H functions of RT simultaneously.¹⁷

The rationale is straightforward. A drug that hits two targets on the same enzyme should be harder for the virus to evade through mutation, because a single change that rescues one function may compromise the other. Laboratory studies on an isatin-based compound called RMNC6 have shown that this approach is feasible: RMNC6 inhibits both the polymerase and RNase H activities through what appears to be a double-site binding mode distinct from the classic NNRTI pocket, and it retains full inhibitory potency against common NNRTI-resistant mutants like K103N, Y181C, and Y188L.¹⁸ None of these dual inhibitors has yet reached clinical approval, but they represent one of the more promising directions in RT drug development.

The Host Strikes Back

RT does not operate in a vacuum. Human cells have their own antiviral defenses that directly interfere with reverse transcription. One of the best-studied is APOBEC3G, a cytosine deaminase enzyme that the host cell can package into newly formed virus particles. When an APOBEC3G-containing virus infects a new cell and RT begins copying the RNA genome, APOBEC3G chemically modifies the single-stranded DNA intermediate by converting cytosine bases to uracil. This introduces catastrophic levels of mutation, a process called hypermutation, that can inactivate the viral genome.¹⁹

Beyond hypermutation, APOBEC3G also physically interferes with RT’s ability to initiate DNA synthesis, adding a deaminase-independent layer of restriction. HIV has evolved a countermeasure: its Vif protein tags APOBEC3G for destruction by the host cell’s own protein-disposal machinery, preventing it from being incorporated into new virus particles. The Vif-APOBEC3G battle is one of the clearest examples of a molecular arms race between HIV and its host, and it has attracted interest as a potential therapeutic angle. If a drug could block Vif’s ability to neutralize APOBEC3G, the host’s own defenses would be unleashed against the virus.

Watching RT Work in Real Time

Much of what we know about RT comes from crystal structures, which capture the enzyme frozen in a single pose. But RT is a dynamic machine that undergoes large-scale conformational changes as it copies DNA, and understanding those motions has required newer biophysical techniques. Single-molecule fluorescence approaches have made it possible to watch individual RT molecules bind to nucleic acid substrates, flip orientation between polymerase-competent and RNase-H-competent binding modes, and carry out DNA synthesis in real time.²⁰

These experiments have revealed that RT’s binding orientation on its substrate is not fixed. The enzyme can slide onto a nucleic acid duplex in either direction, and which orientation it adopts determines whether the polymerase or RNase H active site engages the substrate. The balance between these orientations, and the kinetics of flipping between them, turns out to be an important regulator of which enzymatic activity predominates at any given moment during reverse transcription. This kind of dynamic information is invisible in a static crystal structure but can change how researchers think about designing drugs that trap the enzyme in a nonfunctional conformation.

Single-molecule tools have also been useful for understanding how RT pauses, stalls, or dissociates from the template at difficult sequences. RT is not very processive compared to cellular DNA polymerases, meaning it tends to fall off the template after copying a few hundred bases and must rebind to continue. The nucleocapsid protein, another component of the virus particle, helps RT stay engaged and re-anneal to the template after dissociation, functioning as a kind of molecular chaperone for the entire reverse transcription process.