Origin, distribution, and function of three frequent coding polymorphisms in the gene for the human P2X7 ion channel

Abstract

P2X7, an ion channel gated by extracellular ATP, is widely expressed on the plasma membrane of immune cells and plays important roles in inflammation and apoptosis. Several single nucleotide polymorphisms have been identified in the human P2RX7 gene. In contrast to other members of the P2X family, non-synonymous polymorphisms in P2X7 are common. Three of these occur at overall frequencies of more than 25% and affect residues in the extracellular "head"-domain of P2X7 (155 Y/H), its "lower body" (270 R/H), and its "tail" in the second transmembrane domain (348 T/A). Comparison of the P2X7 orthologues of human and other great apes indicates that the ancestral allele is Y-R-T (at 155-270-348). Interestingly, each single amino acid variant displays lower ATP-sensitivity than the ancestral allele. The originally published reference sequence of human P2X7, often referred to as "wildtype," differs from the ancestral allele at all three positions, i.e. H-H-A. The 1,000 Genome Project determined the sequences of both alleles of 2,500 human individuals, including roughly 500 persons from each of the five major continental regions. This rich resource shows that the ancestral alleles Y155, R270, and T348 occur in all analyzed human populations, albeit at strikingly different frequencies in various subpopulations (e.g., 25%-59% for Y155, 59%-77% for R270, and 13%-47% for T348). BLAST analyses of ancient human genome sequences uncovered several homozygous carriers of variant P2X7 alleles, possibly reflecting a high degree of inbreeding, e.g., H-R-T for a 50.000 year old Neanderthal, H-R-A for a 24.000 year old Siberian, and Y-R-A for a 7,000 year old mesolithic European. In contrast, most present-day individuals co-express two copies of P2X7 that differ in one or more amino acids at positions 155, 270, and 348. Our results improve the understanding of how P2X7 structure affects its function and suggest the importance of considering P2X7 variants of participants when designing clinical trials targeting P2X7.

Keywords: ATP; ATP-gated P2X ion channel; P2X7; adenosine triphosphate; ion channel; loss of function; purinergic receptor; single nucleotide polymorphism.

FIGURE 1

Distribution of coding SNPs in members of the human P2RX gene family. Schematic illustration of the exon/intron structures of the human P2RX gene family. The positions of coding SNPs are indicated above the exons. The color of the font indicates the frequency of the respective mutation in the overall human population: >25% red, 5%–20% black, 0.5%–5% blue among the ∼2.500 individuals from five continental regions analyzed by the 1,000 Genomes Project. The amino acid residues of the major and minor alleles are given in single letter code, before and after the position of the amino acid residue in the native protein, respectively. Numbers on the right indicate the chromosomal localization of the respective gene. P2RX4 and P2RX7 are immediate chromosomal neighbors on chromosome 12. P2RX1 and P2RX5 are separated by five other genes on Chromosome 17.

FIGURE 2

Distribution of minor and major alleles of the three most frequent coding SNPs in the human P2X7 gene (coding for amino acids 155, 270, and 348). The pie charts illustrate the allele frequencies of the major and minor alleles of P2RX7 encoding amino acid positions 155, 270, and 348 of human P2X7 in humans from the five major continental regions. The data were retrieved from the 1,000 Genomes Project. ALL: (n = 2.504 individuals) with about 500 from each region: AFR: African, AMR: American, ESN: East-Asian, SAS South-Asian, EUR: European. The nucleotide of the respective alleles is indicated next to the pie chart, the encoded amino acid residues is indicated on the left using the same color coding as in the pie charts. AFR: African (n = 247), AMR: American (n = 181), ASN: Asian (n = 286), EUR: European (n = 379). ALL: (n = 1.092 humans).

FIGURE 3

Amino acid sequence alignment of the P2X7 orthologues of the great apes. Residues highlighted in grey correspond to the two transmembrane domains. Residues highlighted in yellow indicate substitutions relative to the presumed ancestral version. The sequence of chimpanzee P2X7 corresponds completely to the presumed ancestral allele. Amino acid residues that deviate from the ancestral allele are indicated with their respective positions in the native protein above the alignment. hs: homo sapiens; pp: pan paniscus (bonobo); pt: pan troglodytes (chimpanzee); gg: gorilla gorilla.

FIGURE 4

Most modern-day humans carry two variants of the P2X7 gene. (A) Exons 5, 8, and 11 of the human P2X7 gene were PCR-amplified from genomic DNA of peripheral blood leukocytes. Amplification products were size fractionated by agarose gel electrophoresis and stained with ethidium bromide. (B) PCR amplification products were purified from gel fragments and sequenced with internal primers. Fluorograms illustrate homozygosity or heterozygosity of PCR amplification products at the positions of the three most frequent SNPs of human P2RX7. The deduced amino acids are indicated above the fluorographs.

FIGURE 5

Sensitivity of peripheral blood T cells to ATP-induced shedding of CD62L varies with the genotype at positions 155, 270, and 348 of P2X7. Human blood samples were incubated for 30 min at 37°C in the absence (solvent) or presence of 4 mM ATP and the additional absence (control) or presence of the P2X7-antagonizing monoclonal antibody L4 or the nanobody Dano1. Cells were then washed and co-stained with antibodies directed against CD4 and CD62L as well as with Annexin V to visualize externalization of phosphatidylserine).

FIGURE 6

3D structure models of “wildtype” (HHA) and ancestral (YRT) human P2X7. (A) The ribbon diagram highlights the positions of the three frequent coding SNPs in the different “body-parts” of the dolphin model of P2X7 (Karasawa and Kawate, 2016). Y155H is located in the head domain, R270H in the right flipper, and A348T in the tail of the dolphin. (B) 3D-structure models were generated using AlphaFold 2 and color-coded in pymol. The three divergent amino acid residues are color-coded as in Figures 2, 4.