CCR5-Delta32: The Mutation Behind HIV Resistance

A 32-letter deletion in CCR5 leaves about one northern European in a hundred nearly immune to the common form of HIV. Its history, its trade-offs, the retracted study that a genotyping probe fooled, and how it shows up in a raw file.

Through the 1980s and early 1990s, doctors kept meeting people who had been exposed to HIV many times and never became infected. In 1996 several groups found out why. A 32-base-pair deletion in the gene for CCR5, a receptor that most HIV strains need in order to enter a cell, removes the receptor entirely in people who carry two copies. The virus has nothing to hold on to. The variant is CCR5-Delta32, catalogued as rs333, and its story runs from the origins of European farming through the first person cured of HIV to one of the most instructive retractions in modern genetics.

What the deletion breaks

CCR5 is a chemokine receptor on the surface of T cells, macrophages and other immune cells, where it helps steer those cells toward sites of inflammation. HIV-1 enters a cell by binding CD4 and then a co-receptor, and the strains that cause nearly all new infections use CCR5. The Delta32 deletion removes 32 bases from the coding sequence, shifts the reading frame and produces a truncated protein that never reaches the cell surface.

  • Two copies confer strong resistance to CCR5-using strains. It is not absolute: a small number of infections with strains that use the alternative co-receptor CXCR4 have been documented in Delta32 homozygotes.
  • One copy gives partial protection. Carriers can be infected, but in the era before effective treatment they progressed to AIDS more slowly, by roughly two years on average.
  • Neither copy is the situation for the overwhelming majority of the world’s population.

People without any CCR5 are otherwise healthy, which is what made the receptor an attractive drug target. Maraviroc, an HIV drug that blocks it, was approved in 2007.

Who carries it, and since when

The deletion is essentially a European variant. Its frequency is around 10 percent across much of northern Europe, reaching the mid-teens in parts of Scandinavia and the Baltic, falling to a few percent in the Mediterranean, and close to zero in sub-Saharan Africa, East Asia and the Americas except where European ancestry has introduced it. Roughly one northern European in a hundred is homozygous.

That steep north-south gradient prompted a famous hypothesis: the deletion had been driven up by the Black Death, or perhaps smallpox, in medieval Europe. The story did not survive ancient DNA. In 2005 the deletion was found in Bronze Age skeletons nearly 3,000 years old. In 2025 a Copenhagen group screened 934 ancient genomes and traced every modern copy to a single origin at least 6,700 years ago in the Western Steppe near the Black Sea, with the allele rising quickly between roughly 8,000 and 2,000 years ago. Their suggestion is that a slightly dampened immune response was an advantage as people crowded into farming settlements and met new pathogens - a benefit that has nothing to do with HIV, which the allele predates by thousands of years.

The trade-off

A receptor that exists presumably does something, and the clearest cost of lacking it involves West Nile virus. CCR5 helps immune cells reach the brain during infection, and Delta32 homozygotes are heavily over-represented among people who develop symptomatic and severe West Nile disease - about 8 percent of symptomatic cases in one US cohort, against about 1 percent of the general population. Nobody should regard DD at this marker as an unqualified blessing.

The Berlin patient, and the CRISPR babies

In 2007 Timothy Ray Brown, an HIV-positive American living in Berlin, needed a bone marrow transplant for leukaemia. His doctor chose a donor who was Delta32 homozygous. Brown stopped antiretroviral therapy and the virus never returned; he lived HIV-free until his death from a recurrence of the leukaemia in 2020. A second case in London, reported in 2019, and several since have confirmed that the effect is real. Transplants are far too dangerous to be a treatment for HIV in general, but the cases proved a principle that gene-therapy programmes are now pursuing more safely.

The same principle was behind the most condemned experiment in recent genetics. In 2018 He Jiankui announced that he had edited CCR5 in human embryos, producing twin girls. The edits did not reproduce Delta32, their consequences are unknown, and He was sentenced to three years in prison. The episode is a reminder that a variant being harmless in the people who inherited it is not the same as an edit being safe.

The study that a probe fooled

In June 2019 a paper in Nature Medicine reported that Delta32 homozygotes in the UK Biobank - a cohort of about 410,000 people - had a 21 percent higher all-cause mortality rate, and that there were fewer homozygotes than expected. The conclusion was that the deletion is harmful in double dose, and the coverage connected it directly to the edited twins. Four months later the paper was retracted at the authors’ own request.

The problem was in the raw data. The UK Biobank array did not call rs333 itself, so the analysis used a nearby SNP, rs62625034, as a stand-in. The probe for that SNP binds across the very sequence that Delta32 deletes. In deletion carriers the probe misfires, so the marker is mis-called or missing precisely in the people it was meant to identify. The “missing” homozygotes had not died young; they had been mis-genotyped. When other groups repeated the analysis using genotypes imputed from a reference panel that handles the deletion properly, both the shortage of homozygotes and the mortality signal disappeared.

For anyone reading their own file, the lesson is concrete. An insertion or deletion can wreck a probe designed for a single letter nearby, and the result looks like an ordinary genotype rather than an error. If a marker that sits inside or next to a known indel gives a surprising result, suspect the chemistry before the biology - the same reasoning our article on no-calls applies to missing genotypes.

Finding it in your file

Because Delta32 is an indel, it does not appear as two letters. In 23andMe files it has been reported for years under the internal identifier i3003626 rather than rs333, using the indel convention: I for the intact allele and D for the deletion, so II means two intact copies, DI one deletion and DD two. On 23andMe’s website the same marker is shown as a dash versus the full 32-letter sequence. As far as we know, no other major consumer service includes enough at this position to call it, so a MyHeritage, FamilyTreeDNA or AncestryDNA file will usually simply lack it. A missing line is a missing marker, not a negative result.

grep -E '^(rs333|i3003626)' my_raw_dna.txt

If you find DD, you belong to a small group with a genuinely unusual immune system - one that resists most HIV and handles West Nile virus badly. If you find DI or II, you are with everyone else, and there is nothing to do.

This article is educational only and is not medical advice. Nobody should change their approach to HIV prevention on the basis of a consumer genotype.

References

  • Liu R, et al. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV-1 infection. Cell. 1996. PubMed 8756719
  • Samson M, et al. Resistance to HIV-1 infection in caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Nature. 1996. PubMed 8751444
  • Hummel S, et al. Detection of the CCR5-Delta32 HIV resistance gene in Bronze Age skeletons. Genes and Immunity. 2005. PubMed 15815693
  • Ravn K, et al. Tracing the evolutionary history of the CCR5delta32 deletion via ancient and modern genomes. Cell. 2025. PubMed 40328257
  • Glass WG, et al. CCR5 deficiency increases risk of symptomatic West Nile virus infection. Journal of Experimental Medicine. 2006. PubMed 16418398
  • Hütter G, et al. Long-term control of HIV by CCR5 Delta32/Delta32 stem-cell transplantation. NEJM. 2009. PubMed 19213682
  • Gupta RK, et al. HIV-1 remission following CCR5Δ32/Δ32 haematopoietic stem-cell transplantation. Nature. 2019. PubMed 30836379
  • Wei X, Nielsen R. CCR5-∆32 is deleterious in the homozygous state in humans. Nature Medicine. 2019. Retracted. PubMed 31160814
  • Maier R, et al. No statistical evidence for an effect of CCR5-∆32 on lifespan in the UK Biobank cohort. Nature Medicine. 2020. PubMed 31873311
  • dbSNP entry for rs333, NCBI.

Further reading