How HIV targets CD4 T-cells and writes itself into your DNA
Last reviewed November 16, 2026. OwieClear editorial team. Waiting for clinician review. Not medical advice. Talk with your doctor about your own care.
Most viruses borrow a cell, make copies of themselves and leave. HIV does something different and far harder to undo: it writes its own genetic code into the cell's chromosomes, so the cell itself becomes the instruction manual.
And it does this in the one cell the immune system can least afford to lose.
Related pages: HIV
The bottom line
- HIV enters cells that carry the CD4 receptor, mainly CD4+ T-helper cells.
- It needs a second receptor as well, usually CCR5 and sometimes CXCR4.
- Reverse transcriptase copies the viral RNA into DNA; integrase splices that DNA into a human chromosome.
- Once integrated, the viral DNA is part of that cell and is copied every time the cell divides.
- Why some infected cells make virus immediately and others go quiet.
- Exactly how much of the CD4 loss is direct killing by the virus and how much is immune activation and inflammation.
The story, step by step
- 1. The target ()
- 2. Docking ()
- 3. Rewriting ()
- 4. Integration ()
- 5. Budding ()
What the evidence labels mean
The cell that runs the immune response
What happens
CD4+ T-cells are not the soldiers of the immune system. They are the commanders.
What the research shows
- They recognize a threat and then signal other cells: B-cells to make antibodies, killer T-cells to destroy infected cells, macrophages to swallow debris.
- HIV binds the CD4 protein on the surface of those cells, which is where the cell gets its name.
- It also infects macrophages and some dendritic cells, which carry smaller amounts of CD4.
- A healthy adult usually has roughly 500 to 1,500 CD4 cells per cubic millimetre of blood.
Why it matters
Attacking the commanders is why one virus can eventually leave the body open to dozens of unrelated infections.
What's still uncertain
- Normal CD4 ranges vary between laboratories and between people, so a single number means less than the trend over time.
Read more
- CDC: About HIVGovernment
- HIVinfo (NIH): The HIV life cycleGovernment
Next: Docking
Two locks, not one
What happens
Binding CD4 alone is not enough to get in. HIV needs a second receptor, and which one it uses matters.
What the research shows
- The viral surface spike, gp120, grips CD4. That grip changes its shape and exposes a second binding site.
- That site latches onto a coreceptor, usually CCR5 and sometimes CXCR4.
- A second viral protein, gp41, then acts like a harpoon and pulls the viral envelope and the cell membrane together until they fuse.
- Almost all newly transmitted virus uses CCR5. Virus that can also use CXCR4 tends to appear later in untreated infection.
Why it matters
People born with two copies of a mutation that disables CCR5 are highly resistant to HIV. That single fact has shaped both drug design and every cure attempt so far.
What's still uncertain
- Why CXCR4-using virus emerges in some people and not others is not settled.
Read more
- NIAID: HIV replication cycleGovernment
- OwieClear: Why HIV has no cure yetOwieClear
Next: Rewriting
RNA becomes DNA, and the copying is sloppy on purpose
What happens
HIV carries its genes as RNA. Human chromosomes are DNA. Reverse transcriptase bridges that gap, and this backwards step is what makes HIV a retrovirus.
What the research shows
- Reverse transcriptase copies the viral RNA into double-stranded DNA inside the cell.
- Unlike the enzymes that copy human DNA, it has no proofreading, so it makes frequent mistakes.
- Roughly one error is introduced per genome copied, which means a person carries not one virus but a swarm of closely related variants.
- That swarm is why single-drug treatment fails quickly: a resistant variant is often already present before the drug is ever taken.
Why it matters
The sloppiness is not a flaw in the virus. It is the engine of drug resistance and the main reason an HIV vaccine has been so hard to make.
What's still uncertain
- How much of the variation seen in a person comes from new mutation versus reactivation of older stored virus is still debated.
Read more
- HIVinfo (NIH): The HIV life cycleGovernment
- OwieClear: What an enveloped retrovirus isOwieClear
Next: Integration
Spliced into a human chromosome
What happens
This is the step that makes HIV permanent.
What the research shows
- The viral enzyme integrase cuts open the host chromosome and stitches the viral DNA in. The inserted copy is called a provirus.
- Integration is not random: the virus favours genes that the cell is actively using.
- From then on, the viral DNA is copied along with the cell's own DNA every time that cell divides.
- No approved medicine removes integrated viral DNA. Integrase inhibitors prevent new integration; they do not undo an old one.
Why it matters
This is the single biological fact behind nearly everything else about HIV: why treatment is lifelong, why stopping medication leads to rebound, and why there is still no cure.
What's still uncertain
- Whether where in the chromosome the virus lands decides if it goes dormant is still being worked out.
Read more
- HIVinfo (NIH): The HIV life cycleGovernment
- NIAID: HIV replication cycleGovernment
Next: Budding
New virus leaves, and matures on the way out
What happens
An active cell reads the viral genes like any other genes and builds the parts.
What the research shows
- Host machinery makes long viral protein chains and new copies of the viral RNA.
- These gather at the inner surface of the cell membrane and push outward, taking a piece of the membrane with them as an envelope.
- The viral enzyme protease then cuts the long protein chains into working pieces. Without that cutting step, the particle is defective.
- Protease inhibitors work precisely here: the virus still buds, but what comes out cannot infect anything.
Why it matters
Each infected cell can release thousands of particles. Before treatment, roughly ten billion new virus particles are produced and cleared each day.
What's still uncertain
- Estimates of daily production come from mathematical models of viral decay, not direct counting.