A Glu-496 to Ala polymorphism leads to loss of function of the human P2X7 receptor

P2X(7) receptor is a ligand-gated cation-selective channel that mediates ATP-induced apoptosis of cells of the immune system. We and others have shown that P2X(7) is nonfunctional both in lymphocytes and monocytes from some subjects. To study a possible genetic basis we sequenced DNA coding for the carboxyl-terminal tail of P2X(7). In 9 of 45 normal subjects a heterozygous nucleotide substitution (1513A-->C) was found, whereas 1 subject carried the homozygous substitution that codes for glutamic acid to alanine at amino acid position 496. Surface expression of P2X(7) on lymphocytes was not affected by this E496A polymorphism, demonstrated both by confocal microscopy and immunofluorescent staining. Monocytes and lymphocytes from the E496A homozygote subject expressed nonfunctional receptor, whereas heterozygotes showed P2X(7) function that was half that of germline P2X(7). Results of transfection experiments showed that the mutant P2X(7) receptor was nonfunctional when expressed at low receptor density but regained function at a high receptor density. This density dependence of mutant P2X(7) function was also seen on differentiation of fresh monocytes to macrophages with interferon-gamma, which up-regulated mutant P2X(7) and partially restored its function. P2X(7)-mediated apoptosis of lymphocytes was impaired in homozygous mutant P2X(7) compared with germline (8.6 versus 35.2%). The data suggest that the glutamic acid at position 496 is required for optimal assembly of the P2X(7) receptor.


ABSTRACT
The P2X 7 receptor is a ligand-gated cation selective channel which mediates ATP-induced apoptosis of cells of the immune system. We and others have shown that P2X 7 is nonfunctional both in lymphocytes and monocytes from some subjects. To study a possible genetic basis we sequenced DNA coding for the carboxyl terminal tail of P2X 7 . In 9 of 45 normal subjects a heterozygous nucleotide substitution (A1513C) was found while 1 subject carried the homozygous substitution which codes for glutamic acid to alanine at amino acid position 496. Surface expression of P2X 7 on lymphocytes was not affected by this Glu496Ala polymorphism demonstrated both by confocal microscopy and immunofluorescent staining.
Monocytes and lymphocytes from the Glu496Ala homozygote subject expressed non-functional receptor while heterozygotes showed P2X 7 function which was half that of germline P2X 7 . Results of transfection experiments showed the mutant P2X 7 receptor was non-functional when expressed at low receptor density but regained function at a high receptor density. This density-dependence of mutant P2X 7 function was also seen on differentiation of fresh monocytes to macrophages with interferon-γ which upregulated mutant P2X 7 and partially restored its function. P2X 7mediated apoptosis of lymphocytes was impaired in homozygous mutant P2X 7 compared with germline (8.6% versus 35.2%). The data suggest that the glutamic acid at position 496 is required for optimal assembly of the P2X 7 receptor.

INTRODUCTION
Purinergic P2X 7 receptors are ligand-gated cation channels, present on cells of the immune and haemopoietic system, which have been shown to mediate the ATP-induced apoptotic death of monocytes (1) macrophages (2) and lymphocytes (3,4). The P2X 7 receptor family have two transmembrane domains with intracellular amino and carboxyl termini and an oligomeric structure in the plasma membrane based on trimeric or larger complexes of identical subunits (5). Moreover the P2X7 receptor does not appear to form heteropolymers with other P2X subtypes (6). The genes for both the rat and human P2X 7 receptors have now been cloned and show extensive homology (30 -40%) with the other members of the P2X receptor family, although P2X 7 differs in having a long carboxyl terminus of 240 amino acids from the inner membrane face (7). The genomic structure of P2X 7 consists of 13 exons with exon 12 and exon 13 coding for the carboxyl terminal tail of this molecule. There is strong evidence that this long carboxyl terminus is necessary for the permeability properties of the P2X 7 receptor since truncation of this tail abolishes ATP induced uptake of the fluorescent dye YoPro-1 (8). Studies of P2X 7 of monocytemacrophages or of lymphocytes as well as HEK-293 cells expressing the cDNA for P2X 7 have shown features which are most unusual for a channel. These include the slow further dilatation following channel opening (9) and the activation of various proteases including membrane metalloproteases (10) and intracellular caspases (2,11). The fully dilated state of the P2X 7 pore accepts ethidium cation (314 Da) as a permeant and since ethidium fluorescence is enhanced on binding to nucleic acids the technique of flow cytometry allows a sensitive measurement of the initial rate of permeant uptake which is essentially unidirectional (9). In normal leucocytes a close correlation has been found between ATP induced ethidium uptake and the surface expression of P2X 7 receptors measured by the binding of a FITC-conjugated antibody to the extracellular domain of this receptor (12).
There is increasing evidence that a genetic factor plays a role in the functional phenotype of the P2X 7 receptor. Thus, Lammas and colleagues (13) have shown that ATP-induced uptake of the dye lucifer yellow into monocytes was minimal in 2 out of 19 normal donors while our group has shown a lack of P2X 7 function in both lymphocytes and monocytes in 3 out of 12 patients with B-cell chronic lymphocytic leukaemia (CLL) despite strong expression of the P2X 7 protein (12). These results led us to search for non-functional P2X 7 receptors in a large cohort of normal subjects and study its possible genetic basis. The results show a single nucleotide polymorphism is present at low frequency in the Caucasian population and codes for a glutamic acid to alanine substitution at amino acid 496. Homozygosity for the polymorphism produces non-functional P2X 7 protein while the heterozygous state gives cells with half the function of cells with germline P2X 7 protein. has been reported in our previous study (12). Fixed cells (4% paraformaldehyde) were blocked with 20% horse serum/0.1% bovine serum albumin before incubating with anti-human P2X 7 receptor mAb or isotype control antibody and subsequent labeling with Cy2-conjugated donkey anti-murine IgG antibody.

Materials
Cells were visualised with a Leica TCS NT UV laser confocal microscope system as previously described (16). with FITC-anti-CD3 mAb and 7-AAD (20 µg/ml) for 20 min at room temperature. Viable and non-viable cells were measured by flow cytometry as previously described (17).

P2X 7 function in monocytes and lymphocytes
Our previous data has shown that P2X 7 receptor function in monocyte or lymphocyte subsets can be measured by the ATP-induced uptake of ethidium at 37 o C using timeresolved two-color flow cytometry (12). Mononuclear preparations from 32 normal subjects were pre-incubated with appropriate FITC-labeled monoclonal antibodies and ATP-induced uptake of ethidium into gated monocyte and lymphocyte subpopulations was measured. Ethidium uptake through the P2X 7 channel/pore was 5-fold greater for monocytes than for B-T-or NK-lymphocytes of normal origin but for all cell types there was variation in the functional response of the P2X 7 receptor (Fig.1) One subject showed complete lack of P2X 7 function in both monocytes and lymphocytes shown in Fig. 1 by the filled circles. Variability in ATP-induced dye uptake into monocytes has been observed by others (13).

Identification of a single nucleotide polymorphism in the C-terminal tail of P2X 7 gene
Since the long carboxyl-terminal tail of the P2X 7 receptor regulates its permeability properties, the sequence of genomic DNA corresponding to this region was analysed.
Thus a PCR product was amplified directly from DNA between nucleotide 1425 and 1780 of the coding region of the P2X 7 gene and the product was sequenced. In 9 of 45 subjects a heterozygous nucleotide substitution (adenine to cytosine) was found at position 1513 while in 1 out of 45 subjects a homozygous A1513C substitution was observed (Fig. 2).
Since the fractional frequency of the mutant allele was 11/90 (0.122) in the Caucasian population it fulfils the criterion for a single nucleotide polymorphism. The deduced amino acid change for this mutation is glutamic acid to alanine at amino acid 496 (Glu496Ala) of the P2X 7 protein.

The A1513C mutation is present in skin fibroblasts
Skin fibroblasts were cultured from a punch-biopsy of skin from the homozygous normal subject, DNA was extracted and a product amplified using primers for the C-terminal tail of the P2X 7 gene. Sequence analysis of the product showed only cytosine to be present at position 1513 (results not shown).

Surface expression of P2X 7 is not affected by the polymorphism
Large amounts of P2X 7 protein are found in an intracellular location in monocytes and lymphocytes of all subtypes (12) and we studied whether the A1513C mutation may have reduced the surface expression of this receptor. Confocal microscopy showed strong surface expression of the P2X 7 receptor on lymphocytes from subjects who were germline or homozygous for this mutation (Fig.3) while monocytes showed a similar strong surface P2X 7 expression (data not shown). Flow cytometric measurement of P2X 7 expression using a monoclonal antibody to the extracellular domain of P2X 7 (14) showed that the surface expression of this receptor on either B-or T-lymphocytes from heterozygous or homozygous patients was not significantly different from B-or T-lymphocytes which were of germline sequence at position 1513 ( Table 1).

Correlation of P2X 7 function with the polymorphism
The function of P2X 7 receptors expressed on lymphocytes or monocytes was compared with the genotype at position 1513 of the P2X 7 gene. Typical ethidium uptake curves for monocytes, B-T-and NK-lymphocytes are shown in Fig. 4 for normal subjects each with germline, heterozygous and homozygous DNA at position 1513. A single patient with B-CLL from our previous study (12) and homozygous A1513C is included in Fig. 4 and Table 1 for comparison. Homozygosity for the mutation led to almost complete loss of function of the receptor while heterozygosity for the mutation gave a function approximately half that of the germline P2X 7 sequence (Fig. 4a, b, c, d). Measurement of P2X 7 function in a larger group of subjects (n = 20, Table 1) showed that the mean ATPinduced ethidium uptake was reduced in heterozygous subjects to half the uptake found in subjects with germline sequence, and this magnitude of reduction was found for the four cell types studied; monocytes (p<0.001) B-(p<0.002) T-(p<0.005) and NK-lymphocytes (p<0.03). ATP-induced uptake of Ba 2+ was also studied into lymphocytes prepared from the subject with homozygous mutant P2X 7 . These cells failed to respond to ATP (Fig. 5) indicating that the mutant P2X 7 channel was non-functional to small inorganic cations as well as ethidium + as permeants (12).

Function of A1513C mutated P2X 7 transfected into HEK293
cDNA for germline P2X 7 or P2X 7 carrying the A1513C mutation was transfected into HEK293 cells to study whether this mutation abolishes function in transfection experiments. At 40 hours after transfection the surface expression of the P2X 7 receptor was quantitated by binding of FITC-conjugated mAb (clone B2) and the ATP-induced uptake of ethidium was studied in the same cell population by two-color flow cytometry.
Preliminary experiments suggested that the function of the mutated P2X 7 depended on the density at which this receptor was expressed on the cell surface. For this reason a gating strategy was adopted in which cells expressing none, low or high density of P2X 7 receptors were analysed as three separate populations (Fig. 6a, b, c). The cohort of cells with negative P2X 7 expression showed no ATP-induced ethidium uptake with either germline or mutated cDNA (Fig. 6d). The cohort of cells with low expression of the P2X 7 receptor showed strong ATP-induced ethidium uptake in the germline P2X 7 but the mutant P2X 7 had no function (Fig. 6e). However in the cohort of cells with the highest P2X 7 surface expression, substantial ATP-induced ethidium uptake was observed both for the germline and to a lesser extent for the mutant P2X 7 (Fig. 6f). These data suggest that the impaired function of the P2X 7 receptor in cells carrying the Glu496Ala mutation could be reversed when the density of the mutant receptor was increased on the cell surface.

Homozygous Mutant P2X 7 regains partial function in macrophages
Differentiation of monocytes into macrophages greatly increases both the expression and function of the P2X 7 receptor (18,19). Peripheral blood monocytes were cultured with interferon-γ for 7 days and the function of P2X 7 receptor was measured in the CD14 + macrophage population. Macrophages from subjects with germline P2X 7 showed an ATPinduced ethidium uptake about 5-fold greater than their precursor monocytes (Fig. 7a).
Thus the area under the ATP-induced ethidium uptake curve increased from 28,920 units on day 0 to 160,000 units in day 6 macrophages. Macrophages from a subject homozygous for A1513C polymorphism developed partial P2X 7 function compared to the absent function in the precursor monocytes ( Fig. 7b; zero units on day 0 to 30,800 units on day 6).
Although the P2X 7 expression (mean channels fluorescence intensity) on germline monocyte/macrophages increased from 48 to 284 this increase was less in the homozygous mutant cells (from 51 to 96). Thus the functional defect associated with the Glu496Ala polymorphism in monocyte/macrophages could be partially reversed when the abundance of native P2X 7 was increased on the cell surface.

ATP-induced cytotoxicity is impaired by the homozygous P2X 7 polymorphism
P2X 7 mediated cytotoxicity was studied in lymphocytes from subjects who were germline or homozygous for the Glu496Ala polymorphism. A mononuclear preparation of peripheral blood was exposed to BzATP for 15 min, washed and incubated a further 24 hr prior to assay by two-color flow cytometry using (a) 7-amino-actinomycin D (7-AAD) as a viability dye and (b) FITC-conjugated CD3 mAb to gate on the predominant Tlymphocyte subpopulation. The fluorescent dot-plots (Fig 8) identify two distinct

populations of viable cells (lower region) and non-viable cells (upper region) after 24 hr
incubation. The % of non-viable cells was markedly reduced in the homozygote P2X 7 mutant compared with germline T-cells (Fig 8a,b). In control lymphocytes not exposed to BzATP, the percentage of non-viable cell was 3.2% for germline and 6.6% for homozygote after 24 hr incubation.

DISCUSSION
The data in this study shows that the function of the human P2X 7  Clearly A1513C is only one of several genetic changes which can inhibit the function of the P2X 7 receptor. It has been previously reported that truncation of the long carboxylterminal tail of the rat P2X 7 receptor abolishes ATP-induced uptake of large fluorescent dyes such as Yo-Pro 2+ (8) and truncation of the carboxyl terminal tail of the human P2X 7 receptor also abolishes ATP-induced channel/pore formation (Gu and Wiley, unpublished).
For this reason we sought a loss-of-function mutation in the carboxyl terminus of P2X 7 .
Truncation of a receptor often leads to failure of its surface expression such as shown for the 10 amino acid deletion in the chemokine receptor CCR5 gene (21) or the 49 amino acid deletion from the carboxyl terminus of the sulfonylurea receptor which prevents trafficking of this receptor and its associated ATP-sensitive K + channel to the surface of the pancreatic β-cell (22). However the Glu496Ala polymorphism in P2X 7 allows full expression of the mutant receptor on the cell surface as shown in Fig. 3 and Table 1 for both B-and Tlymphocytes.
The polymorphic A1513C mutation of P2X 7 changes glutamic acid to alanine at amino acid 496 (Glu496Ala) and the present data suggests that this glutamic acid residue at position 496 is centrally involved in the interactions which lead to formation of the P2X 7 channel/pore. The molecular mechanisms underlying the opening of the cation channel and its transition to a fully dilated pore are not resolved. The simplest model for pore dilation is that it is an intrinsic property of the P2X 7 receptor that involves a small scale structural change, perhaps in the selectivity filter of the channel (23). Alternative views suggest that pore dilation involves a large scale structural change such as that induced by the dynamic addition of subunits to the existing oligomeric structure (24) or by an interaction with a protein partner, or the activation of a molecularly distinct pore protein (25) by ligated P2X 7 receptor. The inability of some oocyte expression systems to display BzATP activated pore formation (26,27) also provides evidence for regulation of the pore dilation. The finding that the homozygous mutant P2X 7 receptor is non-functional for both a small cation, permeant Ba 2+ as well as the larger ethidium + emphasizes the importance of this glutamic acid at position 496 both for immediate channel opening as well as its dilatation to a pore. The simplest explanation of the present data is that the glutamic to alanine substitution in the mutant P2X 7 weakens the electrostatic interactions governing the assembly of the P2X 7 channel complex in the plasma membrane.
Transfection of A1513C mutant P2X 7 into HEK293 demonstrated that the loss of channel function in mutant P2X 7 could be reversed at high levels of surface expression of the mutant receptor. Two-color flow cytometry (Fig. 6) was used to directly compare ethidium influx through the P2X 7 pore in transfected cells gated into three subpopulations, those expressing none, those with low and those expressing high levels of this receptor.
This gating strategy employed a FITC-conjugated mAb (clone B2) that binds to an extracellular epitope of P2X 7 but does not inhibit the function of the receptor. Thus ATPinduced ethidium uptake was measured on the red (570 nm) channel into two cell populations defined by high and low (R3 and R2 respectively) fluorescence on the green FITC (525 nm) channel. Germline P2X 7 showed function at both high and low receptor numbers at the plasma membrane. In contrast the mutant P2X 7 was non-functional at low numbers but regained partial function at higher density of expressed receptors. This important finding was confirmed for the native P2X 7 receptor which is upregulated when monocytes from peripheral blood are cultured with interferon-γ to produce macrophages (Fig. 7). The function of germline P2X 7 was stimulated about 5-fold in macrophages compared with their precursor monocytes, but the mutant P2X 7 only regained partial function in macrophages compared with its zero function in precursor monocytes.
Increased receptor abundance may explain the partial restoration of mutant P2X 7 channel function since raising the receptor numbers in the membrane of either HEK293 cells ( Fig.   6) or human macrophages (Fig. 7) would tend to compensate for weakened self associations and promote receptor assembly by a direct mass action effect. Whatever the mechanism of P2X 7 assembly in the membrane, the data in Table 1 shows that much of the person-to-person variation in P2X 7 function can be explained by the genetic polymorphism at amino acid position 496 of the P2X 7 receptor molecule.
Both gain-of-function as well as loss-of-function mutations can affect genes encoding ion channel proteins (23). Thus an asparagine to lysine polymorphism in the third intracellular loop of the human α2 A -adrenergic receptor enhances coupling to G i in the presence of agonist (28). Three loss-of-function mutations have been identified in the human K 1R 6.2 gene which encodes the two-transmembrane protein subunit of the pancreatic β-cell ATP sensitive K + channel (22). Loss-of-function mutations occur in the nompC gene, a sixtransmembrane ion channel in Drosophila responsible for mechanosensory signalling(29).
However, few if any genetic polymorphisms have been previously described in which one allele encodes a non-functional channel.
Extracellular ATP has an emerging role in the immune system since P2X 7 activation leads to apoptotic death of thymocytes (30,31), B-lymphocytes (4), macrophages (2) and dendritic cells (32,33). Thus incubation of mononuclear cells from peripheral blood with ATP gave substantial apoptotic death of T-lymphocytes but cell death was greatly attenuated in T-lymphocytes from the subject with homozygous Glu496Ala P2X 7 protein (Fig. 8). There is good evidence that activation of macrophage P2X 7 receptors by ATP can produce killing of intracellular Mycobacteria tuberculosis by these cells (13,34).
Stimulation of phospholipase D appears to be involved in the killing mechanism (35) and one of the consequences of P2X 7 activation is stimulation of the activity of phosphatidyl choline-specific phospholipase D (36)(37)(38)(39)(40). Other downstream effects of the P2X 7 receptor activation may also occur such as the generation of reactive oxygen intermediates and the stimulation of intracellular caspases which not only kill the organism but also lead to the apoptotic death or cytolysis of the host cell. It is possible that the polymorphism described above may be one of the susceptibility factors predisposing individuals to Mycobacterial infections. Thus study of the P2X 7 knockout mouse (41) and its resistance to certain infectious requiring competent macrophages for control will be important in defining a role for this receptor. Regardless of the clinical associations of the polymorphism at amino acid 496, this loss-of-function mutation affecting the carboxyl terminal tail of P2X 7 may help unravel the molecular events leading to channel/pore formation.      Monocytes from mononuclear preparations were allowed to adhere to plastic culture flasks overnight and were cultured for another 6 days in medium plus 100 ng/ml interferon-γ.

Figure Legend
Cells were collected by gentle mechanical scraping and labeled with FITC-anti-CD14 mAb. Linear mean channel fluorescence intensity was measured in each 5 s interval on the gated CD14 + population after 25 µM ethidium and 1 mM ATP were added.