rs35947132 — PRF1 A91V

PRF1 A91V — The Perforin Pore Puncher That Falls Short

Perforin is the immune system's master assassin molecule. When a cytotoxic T lymphocyte (CTL) or natural killer (NK) cell locks onto a virus-infected cell or tumor cell, it releases perforin from secretory granules. Perforin polymerizes on the target cell membrane, punching transmembrane pores¹¹ transmembrane pores
Perforin monomers oligomerize into ring-shaped channels ~10 nm in diameter, allowing granzymes to enter and trigger apoptosis that let granzymes flood in and trigger programmed cell death. Without effective perforin, the immune system cannot terminate its own killing response — and that failure can spiral into a life-threatening inflammatory storm called hemophagocytic lymphohistiocytosis (HLH).

The A91V variant (c.272C>T; rs35947132) substitutes valine for alanine at position 91 of the perforin precursor protein. It is one of the most common functional variants in the PRF1 gene, occurring in approximately 4–9% of chromosomes in European populations and far less often in East Asian and African populations.

The Mechanism

Position 91 sits in the EGF-like domain of the perforin precursor, a region critical for protein folding and post-translational maturation²² protein folding and post-translational maturation
Perforin is synthesized as a 67-kDa precursor that must undergo calcium-dependent conformational change and N-glycosylation before becoming lytically active. The A91V substitution disrupts an antigenic epitope recognized by the monoclonal antibody deltaG9, a sign that the local tertiary structure is altered. More importantly, the mutation impairs cleavage of the precursor to the active form³³ impairs cleavage of the precursor to the active form
Patient cells expressing A91V show only immature and intermediate perforin species; the mature active form is absent or barely detectable on Western blot. This processing defect translates into a 10-fold reduction in target-cell lysis when purified A91V protein is tested on cells, and roughly a 50% reduction in intact CTL cytotoxicity — because wild-type perforin on the second allele partially compensates in heterozygotes.

The functional deficit is bimodal. Voskoboinik et al. (Blood, 2007)⁴⁴ Voskoboinik et al. (Blood, 2007) showed defects at both the presynaptic level (reduced perforin secretion into the immunological synapse) and the postsynaptic level (reduced membrane permeabilization even when secretion occurs). The 2015 study by House et al.⁵⁵ House et al. confirmed these findings translate to primary human NK cells from healthy A91V heterozygous volunteers — a ≥35% reduction in cytotoxicity relative to wild-type individuals, measured in cells that were never exposed to inflammatory conditions.

The Evidence

Functional evidence is strong and replicated across multiple independent laboratories. A91V reduces perforin lytic activity, reduces expression, and causes measurable NK cell cytotoxicity deficits even in healthy heterozygous carriers.

Clinical evidence focuses on compound heterozygosity. In a landmark case series by Clementi et al. (2002)⁶⁶ Clementi et al. (2002), siblings with adult-onset HLH were found to carry A91V on one allele and a null mutation (W374X) on the other — classic compound heterozygosity. Zhang et al. (Blood, 2011)⁷⁷ Zhang et al. (Blood, 2011) found A91V in 48% of 25 adult-onset familial HLH patients, many presenting in middle age or later. Documented A91V combinations with W374X, G149S, R104C, and I125T all resulted in HLH with later and milder presentations than pediatric null-allele homozygotes.

Homozygous A91V is an unusual presentation. Mancebo et al. (Haematologica, 2006)⁸⁸ Mancebo et al. (Haematologica, 2006) described an adult patient who was homozygous for A91V and developed FHL2 triggered by active tuberculosis infection — the first documented A91V homozygous FHL case. Crucially, the patient's monozygotic twin, with identical PRF1 genotype, remained completely healthy, demonstrating that homozygous A91V does not cause disease on its own: environmental triggers matter.

In FHL patient cohorts, Busiello et al. (2006)⁹⁹ Busiello et al. (2006) found A91V in 26.2% of FHL patients carrying other PRF1 mutations, versus 3.7% in healthy controls (P=0.0002), firmly establishing its role as a disease-modifying susceptibility factor rather than a primary pathogenic mutation.

Beyond HLH, A91V has been detected at elevated frequency in NK/T-cell lymphomas — particularly nasal-origin cases (25% prevalence)¹⁰¹⁰ NK/T-cell lymphomas — particularly nasal-origin cases (25% prevalence) — and in young severe COVID-19 patients who developed HLH-like hyperinflammatory syndrome. A small study found A91V in 2 of 22 young critical COVID-19 patients (both died), with markedly elevated ferritin suggesting a virally triggered HLH phenotype.

Practical Implications

For heterozygous carriers with a single A91V allele and no known second PRF1 variant, the primary practical implication is awareness: persistent high fever, unexplained cytopenia, and dramatically elevated ferritin are warning signs that warrant urgent HLH evaluation. Standard genetic panels for HLH should include full PRF1 sequencing to identify compound heterozygosity.

For confirmed compound heterozygotes (A91V plus any other PRF1 pathogenic variant) or homozygous A91V individuals, surveillance with serum ferritin during febrile illnesses and avoidance of immunosuppression without specialist oversight are critical. Early recognition and treatment of HLH are the main determinants of survival.

Interactions

The A91V allele interacts with any second loss-of-function PRF1 allele to produce compound heterozygosity — the most clinically significant interaction. Documented combinations include A91V/W374X, A91V/G149S, A91V/R104C, and A91V/I125T, all causing late-onset HLH with milder phenotype than pediatric biallelic null mutations.

A91V also interacts with genes in the exocytosis pathway. Zhang et al. (2011)¹¹¹¹ Zhang et al. (2011) documented adult HLH patients carrying A91V in PRF1 alongside pathogenic variants in MUNC13-4 (UNC13D) or STXBP2, suggesting that cumulative functional deficits across the cytotoxic killing pathway can produce HLH even without biallelic PRF1 mutations.