The Antibiotic Polymyxin B Modulates P2X7 Receptor Function

Davide Ferrari, Cinzia Pizzirani, Elena Adinolfi, Sylvia Forchap, Barbara Sitta, Laura Turchet, Simonetta Falzoni, Mattia Minelli, Roberto Baricordi and Francesco Di Virgilio
J Immunol October 1, 2004, 173 (7) 4652-4660; DOI: https://doi.org/10.4049/jimmunol.173.7.4652

Abstract

The natural peptide polymyxin B (PMB) is a well-known and potent antibiotic that binds and neutralizes bacterial endotoxin (LPS), thus preventing its noxious effects among LPS-mediated endotoxin shock in animal models. We have investigated the effect of PMB on responses mediated by the P2X7R in HEK293 and K562 cells transfected with P2X7 cDNA and in mouse and human macrophages. In addition, in view of the potential exploitation of P2X7-directed agonists in antitumor therapy, we also investigated the effect of PMB in B lymphocytes from patients affected by chronic lymphocytic leukemia. PMB, at an optimal concentration dependent on the given cell type, greatly potentiated the effect of nucleotide-mediated P2X7 stimulation. In particular, ATP-mediated Ca2+ influx, plasma membrane permeabilization, and cytotoxicity were enhanced to an extent that, in the presence of PMB, cells were killed by otherwise ineffective nucleotide concentrations. The synergistic effect due to the combined application of ATP and PMB was prevented by incubation with the irreversible P2X blocker oxidized ATP (oATP), but not with the reversible antagonist 1-(N,O-bis(1,5-isoquinolinesulfonyl)-N-methyl-l-tyrosyl)-4-phenilpiperazine (KN-62). Cells lacking P2X7 were fully insensitive to the combined stimulation with PMB and ATP. Furthermore, PMB at the concentrations used had no untoward effects on cell viability. These results point to PMB as a useful tool for the modulation of P2X7R function and suggest that care should be used in the evaluation of ATP-stimulated immune cell responses in the presence of PMB as they may not solely be affected by removal of contaminating LPS.

Immune cells express receptors for extracellular nucleotides (P2R) (1, 2), among which the pore-forming receptor P2X7 (P2X7R) has attracted interest for its ability to elicit several responses typical of inflammation, such as activation of transcription factor NF-κB (3), release of the inflammatory mediators NO (4), IL-1β (5), IL-2 (6), IFN-γ (6), IL-6 (7), and TNF-α (8). Furthermore, an intriguing albeit controversial association between this receptor and the clinical course of chronic lymphocytic leukemia (CLL)4 has been disclosed by recently published studies (9, 10, 11, 12, 13, 14). Expression and function of P2X7 appear to correlate positively with an evolutive (aggressive) course of the disease and accordingly a loss of function polymorphism in the P2X7 molecule occurs with lower frequency in these same patients, thus pointing to P2X7 expression level as a potentially valuable prognostic index in CLL and prospective therapeutic target.

The P2X7R is a rather peculiar plasma membrane molecule in that it can exist in two functional states: as a cation-selective channel or as a nonselective pore (15). The channel-to-pore transition occurs upon either sustained stimulation with high ATP concentrations or repeated pulsed ATP applications (16). Several responses of potential physiological relevance elicited by the P2X7R are dependent on the opening of the pore and the ensuing perturbation in the intracellular ion homeostasis (2, 17). Opening of the P2X7R pore generates fast changes in the intracellular ion concentration that may function as signaling systems. In addition, further signal transduction pathways are likely to be stimulated because there is convincing evidence for interaction of P2X7R with several plasma membrane or cytoplasmic proteins (18, 19).

The physiological meaning of the in vitro P2X7R-mediated proinflammatory responses is so far uncertain because there is no strong in vivo proof to support an involvement of this receptor in inflammation. A recent study shows a deficit in the development of an inflammatory response, i.e., arthritis induced by anti-collagen Abs, in mice deleted of the p2x7 gene (20), but this is at the moment the only in vivo indication for a possible role of this receptor in inflammation. Therefore, it is fair to say that although circumstantial evidence suggests that P2X7 might play a role in inflammation and immunomodulation, strong proof is lacking.

A main obstacle to the elucidation of the role of P2X7 is the lack of selective receptor agonists or antagonists. Among agonists, 2′,3′-(4-benzoyl)-benzoyl-ATP (BzATP), the most widely used activator of P2X7, is known to also stimulate P2X1 and P2X3 (21). Among antagonists, pyridoxalphosphate-6-azophenyl 2′-4′-disulfonic acid also blocks other P2X receptors and may covalently modify other plasma membrane proteins unrelated to P2R. This may also be a problem with oxidized ATP (oATP) (22), since its aldehyde residues might form Schiff bases with any accessible lysines present in plasma membrane proteins and not just with those present in the P2X7 molecule (23, 24). Wiley and coworker (25) have introduced the compound KN-62 as a potent P2X7 blocker active in the nanomolar range. Several more potent KN-62 derivatives have been synthesized and shown to inhibit P2X7-mediated IL-1β production in human macrophages, but intrinsic characteristics of the lead compound (i.e., hydrophobicity and molecular mass) make its further development unlikely (26).

A so far neglected approach in this field is the search for compounds that are able to modulate ATP-dependent activation by acting as positive or negative effectors (allosteric modulators) at the P2X7R. A few years ago the compound tenidap was shown to synergize with ATP at the P2X7R (27). This molecule, originally developed at Pfizer as an anti-inflammatory, antiarthritic agent, is itself devoid of untoward effects on cell cultures in the absence of extracellular ATP, but when added along with this nucleotide causes a leftward shift of the ATP EC50 for cell killing of at least 1 log unit. Another molecule recently shown to potentiate P2X7-dependent responses is the antibiotic geldanamycin that has been proposed to relieve the repressor function of heat shock protein 90 on the P2X7R (28).

Polymyxin B (PMB) is a cationic cyclic decapeptide obtained from Bacillus polymyxa mainly active against Gram-negative rods that, due to its systemic toxicity, is mainly used as a topical agent (see Ref.29 for a recent review). The high affinity of PMB for lipid A is also currently exploited to remove contaminating LPS from solutions and reagents. In the course of an investigation on the effect of bacterial endotoxin (LPS) on immune cells, we noticed that PMB modified ATP-dependent cell responses in a fashion that could not be correlated to its LPS-binding activity. This effect of PMB was then thoroughly investigated in cells expressing the recombinant or the native P2X7R. Our data led us to conclude that PMB sensitizes cells to extracellular ATP, probably by direct interaction with the P2X7R.

Materials and Methods

Reagents

PMB, ATP, UTP, oATP, BzATP, digitonin, and ionomycin were purchased from Sigma-Aldrich (Milan, Italy), KN-62 was obtained from Calbiochem (Calbiochem-Novabiochem, La Jolla, CA), fura 2-acetoxy methyl ester (fura 2-AM) was from Molecular Probes (Leiden, The Netherlands), and Ficoll-Paque was obtained from Pharmacia Biotech (Uppsala, Sweden). The peptide used for inhibition of PMB binding (573-CRWRIRKEFPKSEGQYS-amide) was synthesized by MedProbe (Oslo, Norway). Human and rat P2X7R truncated (P2X7ΔC) at aa 415 were a kind gift from Dr. G. Buell (Ares-Serono Research Laboratories, Geneva, Switzerland).

Cells and solutions

HEK293 and K562 cells were cultured in DMEM-F12 (Sigma-Aldrich) and RPMI 1640 (Invitrogen Life Technologies, Gaithersburg, MD), respectively. Human monocytes were isolated from buffy coats as described previously (30) and cultured in RPMI 1640 (Invitrogen Life Technologies). The J774 mouse macrophage cell line and the B2 J774 variant were grown in DMEM. CLL cells were maintained in IMDM (Invitrogen Life Technologies). All media were complemented with 10% heat-inactivated FBS (5% human serum in the case of human monocytes), 100 U/ml penicillin, and 100 μg/ml streptomycin (all from Invitrogen Life Technologies). Fluorescence measurements and lactic dehydrogenase (LDH) release were performed in a saline solution containing 125 mM NaCl, 5 mM KCl, 1 mM MgSO4, 1 mM Na2HPO4, 5.5 mM glucose, 5 mM NaHCO3, 1 mM CaCl2, and 20 mM HEPES (pH 7.4 with NaOH), hereafter also referred to as standard saline solution. In some experiments, CaCl2 was omitted and 0.5 mM EGTA was added (Ca2+-free saline solution). Measurement of intracellular Ca2+ concentration ([Ca2+]i) of CLL lymphocytes was performed in a low ionic strength solution containing 300 mM sucrose, 1 mM MgCl2, 1 mM K2HPO4, 5 mM KHCO3, 5.5 mM glucose, 1 mM CaCl2, and 20 mM HEPES (pH adjusted to 7.4 with KOH).

Cytoplasmic-free Ca2+ measurements

Changes in the free [Ca2+]i were measured with the fluorescent indicator fura 2-AM using an LS50 PerkinElmer fluorometer (PerkinElmer, Beaconsfield, U.K.). For fura 2-AM loading, cells (1 × 107/ml) were resuspended in standard saline solution in the presence of 4 μM fura 2-AM and 250 μM sulfinpyrazone (Sigma-Aldrich). Incubation was performed at 37°C for 15 min. Cells were then washed in the same solution and [Ca2+]i changes were determined in a magnetically stirred cuvette with a thermostat, with the 340/380 excitation ratio at an emission wavelength of 505 nm. In some experiments, [Ca2+]i changes are reported as [Ca2+]i increases over basal (Δ[Ca2+]i).

Changes in plasma membrane permeability

ATP-dependent increases in plasma membrane permeability were measured by monitoring the uptake of the dye ethidium bromide. Briefly, 5 × 105 cells/ml were kept in a magnetically stirred fluorometric cuvette, with the thermostat at 37°C, and incubated in the presence of 20 μM ethidium bromide. To achieve complete permeabilization of the cells, 100 μM digitonin was added at the end of the experiment (100% fluorescence signal). Fluorescence was measured at an excitation wavelength of 360 nm; emission was 580 nm.

Measurement of enzymatic activity

LDH activity was measured according to standard methods. Briefly, cells (1 × 105/ml) were plated in 24-well plates and, after 24 h, cells were rinsed and incubated in standard saline solution. ATP was then added in the presence or absence of PMB. Supernatants were collected, cleared by centrifugation (10 min at 2500 × g), transferred into fresh tubes, and stored at −80°C. For measurement of activity, supernatants were thawed and added to a solution containing 0.63 mM pyruvate, 11.3 mM NADH, 44.4 mM K2HPO4, and 16.8 mM KH2PO4 (pH 7.5). Absorbance (340 nm) was measured in a spectrofluorometer (Ultrospec 3000; Pharmacia Biotech). Lysis of samples with 0.1% Triton X-100 (J.T. Baker, Milan, Italy) provided the total LDH cell content (100% LDH).

Measurement of plasma membrane potential

Changes in plasma membrane potential were measured with the fluorescent dye bis-1,3-diethylthiobarbiturate trimethineoxonal (bisoxonol; Molecular Probes) at the wavelength pair 450/580 as described previously (30).

Measurement of IL-1β release

IL-1β in the supernatant of LPS (Sigma-Aldrich)-primed J774 cells was measured with a mouse ELISA kit manufactured by R&D Systems (Minneapolis, MN).

Results

PMB potentiates the ATP-dependent [Ca2+]i rise in HEK293 cells transfected with the P2X7R

Fig. 1⇓A (trace a) shows that stimulation with 1 mM ATP of wild-type (wt) HEK293 cells induced a modest and transient [Ca2+]i rise, likely due to activation of P2Y1R or P2Y2R (31). A strikingly different response was elicited by ATP in HEK293 cells overexpressing the P2X7R (HEK293-P2X7) (Fig. 1⇓A, trace b). In this case, the Ca2+ peak was 3- to 4-fold higher, and the fast initial rise in the Ca2+ signal was followed by a very slow decrease (sustained plateau). Preincubation (3 min) of HEK293-P2X7 with PMB (10 μg/ml) enhanced the Ca2+ response elicited by ATP (Fig. 1⇓B, cf traces c and b). Peak increase was slightly affected, while the sustained, delayed plateau was potentiated to such an extent that within the length of a typical experiment (20 min), [Ca2+]i never returned to basal level (data not shown). PMB treatment did not affect the ATP-induced Ca2+ response in wt HEK293 cells (Fig. 1⇓B, trace a).

FIGURE 1.
FIGURE 1.

PMB potentiates the ATP-dependent [Ca2+]i rise in HEK293 cells transfected with the P2X7R. The fura-2-loaded cells (1 × 106/ml) were incubated at 37°C in standard saline solution and challenged with ATP (1 mM) in the absence or presence of PMB (10 μg/ml). PMB was added 3 min before ATP. A, Trace a, wt HEK293 cells; trace b, HEK-P2X7. B, Trace a, wt HEK293 in the presence of PMB; trace b, HEK293-P2X7; trace c, HEK293-P2X7 cells incubated in the presence of PMB. C, Dependency on the PMB concentration of the ATP-stimulated [Ca2+]i rise in HEK293-P2X7. •, Peak [Ca2+]i measured 30 s after ATP addition; ○, plateau [Ca2+]i measured 6 min after ATP addition. D, ATP dose dependency of the [Ca2+]i rise in HEK293-P2X7 cells measured 30 s after ATP addition (peak [Ca2+]i); ○, without PMB; •, with PMB (10 μg/ml). E, ATP dose dependency of the [Ca2+]i rise in HEK293-P2X7 cells measured 6 min after ATP (plateau [Ca2+]i); ○, without PMB; •, plus PMB (10 μg/ml). Data are averages ± SD of triplicate determinations. Where not shown, error bars are smaller than the size of the symbol.

Effective PMB concentrations were within the 0.30–10 μg/ml range, with a maximal effective dose of 10 μg/ml (Fig. 1⇑C). All subsequent experiments were performed with this PMB dose, unless otherwise indicated. Fig. 1⇑, D and E, shows an ATP dose dependency of the [Ca2+]i increase in the absence or presence of PMB. The [Ca2+]i was measured either at the peak increase (30 s after ATP addition, Fig. 1⇑D) or during the sustained plateau (6 min after ATP addition, Fig. 1⇑E). PMB caused a leftward shift of the dose-dependency curve and a modest increase in potency at plateau, where at the optimal concentration of 1 mM ATP the [Ca2+]i was almost doubled.

PMB potentiates the ATP-dependent ethidium uptake: effect of P2X7 blockers

A hallmark of P2X7 activation is plasma membrane permeabilization. To test this response, HEK293-P2X7 cells were incubated in the presence of the impermeant fluorescent dye ethidium bromide (20 μM) and then stimulated with ATP in the presence or absence of PMB. Preincubation with PMB substantially increased the permeabilizing activity of ATP (Fig. 2⇓A, cf traces a and b). PMB had a strong effect on the initial rate of ethidium bromide uptake calculated as rate of fluorescence increase in arbitrary units (Fig. 2⇓B). PMB also enhanced the response to 300 μM BzATP, which under these conditions induced a near complete permeabilization (Fig. 2⇓C).

FIGURE 2.
FIGURE 2.

P2X7 pore opening is increased by PMB: effect of inhibitors. HEK293-P2X7 cells (5 × 105/ml) were incubated in the presence of 20 μM ethidium bromide and then stimulated with ATP (1 mM) or BzATP (300 μM) in the presence or absence of PMB (10 μg/ml). Fluorescence increase was measured as reported in Materials and Methods. Digitonin (Dig), when added, was 100 μM. A, Trace a, without PMB; trace b, with PMB. B, ATP dose dependency of ethidium bromide uptake in the absence (○) or presence (•) of PMB; data are shown as rate of fluorescence increase in arbitrary units. C, Trace a, without PMB; trace b, with PMB. D, Solid line, cells stimulated with BzATP plus PMB; dotted line, cells were pretreated with 600 μM oATP at 37°C for 2 h, then stimulated with PMB immediately followed by ATP; dashed line, cells were pretreated with 600 μM oATP at 37°C for 2 h, then stimulated with BzATP alone. E, Solid line, cells were stimulated with PMB immediately followed by BzATP; dashed line, cells were treated with KN-62 (50 nM) and immediately after stimulated with BzATP; dotted line, cells were treated with KN-62 (50 nM) and immediately after stimulated with PMB plus BzATP. Data are averages ± SD of triplicate determinations. Where not shown, error bars are smaller than the size of the symbol.

In the presence of PMB, HEK293-P2X7 cells were differentially sensitive to oATP and KN-62, respectively. Preincubation with oATP completely prevented BzATP-mediated permeabilization whether in the presence or absence of PMB (Fig. 2⇑D). On the contrary, PMB overcame the inhibitory effect of KN-62, as shown by its ability to re-establish a near maximal sensitivity to BzATP (Fig. 2⇑E).

Time course and dose dependency of ATP-mediated HEK293-P2X7 cell killing in the presence of PMB

Activation of P2X7R causes cell death in many different cell systems (15). Given the potentiating effect of PMB on Ca2+ influx and plasma membrane permeabilization, we hypothesized that ATP-dependent cytotoxicity would also be enhanced. Fig. 3⇓, A and B, shows that ATP-induced cell death, as measured by release of the cytosolic enzyme LDH was time- and dose-dependently increased by the presence of PMB, which by itself was not toxic. For the time course, we chose an ATP dose (600 μM) that was below threshold for cytotoxicity. In the presence of PMB, this ATP concentration was cytotoxic even after as little as 2 h of incubation.

FIGURE 3.
FIGURE 3.

Effect of PMB on time course and dose dependency of ATP-induced killing. HEK293-P2X7 cells (4 × 105/ml) were incubated in standard saline solution at 37°C in the presence of ATP with or without PMB (10 μg/ml). A, Cells were incubated for different times in the presence of 600 μM ATP, with (•) or without (○) PMB. B, Cells were incubated for 2 h in the presence of increasing ATP concentrations with (•) or without (○) PMB. At the end of the incubation, supernatants were collected and analyzed for LDH content as specified in Materials and Methods. Data are averages ± SD of triplicate determinations. Where not shown, error bars are smaller than the size of the symbol.

The 2-h incubation was also chosen for the ATP dose dependency (Fig. 3⇑B). At this incubation time ATP by itself had little cytotoxic effect (<10% LDH release), even at the high concentration of 3 mM. In the presence of PMB, both affinity and potency were increased, as indicated by the leftward shift in the dose-dependency curve and by the higher cytotoxic response.

Site of action of PMB

PMB had no effect on the viability of wt HEK293 cells and was able to revert the inhibitory effect of KN-62 in HEK293-P2X7 transfectants. These observations support the hypothesis that 1) PMB effect depends on the expression of P2X7R and 2) PMB directly interacts with P2X7. PMB is well known for its high affinity for LPS. It has been recently suggested that LPS might interact with aa 573–590 of the P2X7R COOH tail (32); thus, we tested the hypothesis that PMB too might bind to this receptor domain. To this aim, we performed two kinds of experiments: on one hand we investigated the effect of PMB on HEK293 cells transfected with a rat P2X7R lacking the C-terminal 180 aa (P2X7ΔC) (16), and on the other we used a peptide mimicking the putative LPS binding site of the P2X7R COOH tail in an attempt to inhibit the PMB effect, assuming that, if this amino acid sequence was involved in the binding to PMB, the peptide should compete with PMB and block its effect. We used rat P2X7ΔCR as the human truncated receptor showed a very low level of plasma membrane expression (our unpublished observation).

However, as shown in Fig. 4⇓A, PMB also potentiated Ca2+ influx in response to ATP in cells transfected with rat P2X7ΔCR. Plasma membrane depolarization was also potentiated by PMB (Fig. 4⇓B). Treatment with an optimal (10 μg/ml) PMB concentration for HEK293 cells shifted slightly the dose-response curve to the left and increased the potency, with a higher effect on the plateau rather than on the peak [Ca2+]i level, as expected (Fig. 4⇓, C and D). Finally, preincubation in the presence of the 573–590 peptide was unable to prevent the PMB effect (data not shown).

FIGURE 4.
FIGURE 4.

Lack of the C-terminal tail does not prevent PMB-dependent potentiation of ATP stimulation. A, The fura-2-loaded HEK293 cells (1 × 106/ml) transfected with the P2X7ΔCR were incubated in standard saline solution at 37°C in the absence (trace a) or presence (trace b) of PMB (10 μg/ml) and stimulated with 1 mM ATP. B, HEK293 cells (2 × 105/ml) transfected with the P2X7ΔCR were incubated in bisoxonol (100 nM) containing standard saline solution at 37°C in the absence (trace a) or presence (trace b) of PMB (10 μg/ml) and stimulated with 1 mM ATP. After the ATP pulse, sequential additions of KCl (15 mM each) were performed (arrowheads). C, Dependency on the ATP concentration of the fast [Ca2+]i rise (measured 30 s after ATP addition) in the absence (○) or presence (•) of PMB (10 μg/ml). D, Dependency on the ATP concentration of the delayed [Ca2+]i rise (measured 6 min after ATP addition) in the absence (○) or presence (•) of PMB (10 μg/ml). Data are averages ± SD of triplicate determinations.

Time course and dose dependency of ATP-mediated K562-P2X7 cell killing in the presence of PMB

PMB potentiation of ATP stimulation was not restricted to HEK293 cells but also occurred in other cell lines transfected with P2X7R, among which was the K562 erythroid cells (K562-P2X7) (Fig. 5⇓). K562-P2X7 cells are comparatively less susceptible to ATP-mediated cytotoxicity than HEK293-P2X7 or other P2X7 transfectants, such as, for example, the LG14 lymphoblastoid cells, in that high ATP doses and long incubation times are needed to induce cell death. Nonetheless, in the presence of PMB, ATP had a clear-cut lytic effect even after an incubation as short as 30 min (Fig. 5⇓A) and at doses similar to those active in the vast majority of the other cell types (2).

FIGURE 5.
FIGURE 5.

Time course and dose dependency of ATP-mediated K562-P2X7 cell killing in the presence of PMB. K562-P2X7 cells (4 × 105/ml) were incubated in standard saline solution at 37°C with (•) or without 10 μg/ml PMB (○). A, Time course, ATP concentration was 600 μM and B, dose dependency, incubation time was 2 h. Supernatants were collected and analyzed for LDH content as detailed in Materials and Methods. Data are averages ± SD of triplicate determinations. Where not shown, error bars are smaller than the size of the symbol.

Effects of PMB in macrophages

Macrophages are the cell type in which P2X7R expression and function have been more thoroughly investigated; thus, this cell type was an interesting model to test the effects of PMB. In both human monocyte-derived macrophages and the murine cell line J774, the effect of PMB differed from that observed in HEK293 cells in that although at low concentrations there was a potentiation of both the fast and the delayed ATP-dependent Ca2+ increase, at higher PMB concentrations the response was inhibited (Fig. 6⇓, A and B). The fact that the fast Ca2+ increase rise was modified by PMB might suggest that the target in macrophages was mainly the P2YR, but control experiments performed with the P2X7R-less cells that lack the P2X7R but express P2YR, PMB had no effect on the [Ca2+]i increase (Fig. 6⇓C).

FIGURE 6.
FIGURE 6.

Modulation of ATP-stimulated [Ca2+]i rise by PMB in macrophages. The fura-2-loaded cells (1 × 106/ml) were incubated at 37°C in standard saline solution and challenged with ATP (1 mM) in the absence or presence of increasing PMB concentrations. PMB was added 3 min before ATP. A, Dependency on the PMB concentration of the fast (measured 30 s after ATP addition, •) and delayed [Ca2+]i rise (measured 6 min after ATP addition, ○) in J774 macrophages. B, Dependency on the PMB concentration of the fast (measured 30 s after ATP addition, •) and delayed [Ca2+]i rise (measured 6 min after ATP addition, ○) in monocyte-derived human macrophages. C, P2X7-less B2 J774 macrophages were stimulated with 1 mM ATP in the absence (dashed line) or presence (solid line) of PMB (10 μg/ml).

One of the most striking responses elicited by P2X7R stimulation in macrophages is IL-1β secretion, thus we wondered whether this functional response was also potentiated by PMB. Fig. 7⇓A shows that in LPS-primed J774 cells IL-1β release was strongly enhanced by PMB at the optimal concentration of 10 μg/ml. At the lower concentration of 100 ng/ml, PMB had no effect (Fig. 7⇓B).

FIGURE 7.
FIGURE 7.

Potentiation of ATP-dependent IL-1β release by PMB. J774 macrophages (5 × 105) were plated in 24-well plates in 500 μl of standard saline and primed for 2 h with LPS (1 μg/ml) in a CO2 incubator at 37°C. A, At the end of this incubation, PMB was added (10 μg/ml) and soon after cells were stimulated with increasing ATP concentrations. After a 2-h incubation at 37°C, supernatants were withdrawn and assayed for IL-1β content. B, Macrophages were primed for 2 h with LPS (1 μg/ml) and stimulated with PMB, ATP (1 mM), or both PMB and ATP. After a 2-h incubation at 37°C, supernatants were withdrawn and assayed for IL-1β content. Data are averages ± SD of triplicate determinations. Contr, Control.

PMB potentiation of ATP-mediated killing in a tumor model: CLL

B cell CLL is a hemopoietic tumor originated by expansion of small CD5+ B lymphocytes. Leukemic lymphocytes express P2X7R, whose expression and function was shown to be increased in lymphocytes from CLL patients with the aggressive variant compared with patients with the indolent variant of the disease (9). The unusually high P2X7R expression and function in CLL lymphocytes suggest the possible administration of ATP (or ATP analogues) as antitumor therapy in these patients. However, ATP has itself local and systemic side effects that make it advisable to administer doses as low as possible. Therefore, we tested the effect of the combined application of PMB and ATP to lymphocytes from three different CLL patients affected by the aggressive variant. In these patients, PMB profoundly modified the ATP-stimulated [Ca2+]i rise; by increasing the concentration of this antibiotic, the fast initial transient was progressively obliterated and replaced by a delayed and long-lasting increase that did not reach a plateau, even several minutes after ATP addition (Fig. 8⇓). Furthermore, albeit the three subjects showed some individual variability in their susceptibility to ATP, yet in all three cases PMB was able to reduce the EC50 and increase the potency of the nucleotide in a typical cytotoxicity test (Fig. 9⇓).

FIGURE 8.
FIGURE 8.

Modulation of ATP-stimulated [Ca2+]i rise by PMB in CLL lymphocytes. B lymphocytes from three CLL patients (A–C) were loaded with fura 2-AM as described in Materials and Methods, incubated (1 × 106/ml) in sucrose-containing saline at 37°C, and stimulated with 1 mM ATP with or without PMB. Dotted line, no PMB; dashed line, 100 ng/ml PMB; solid line, 10 μg/ml PMB.

FIGURE 9.
FIGURE 9.

PMB potentiates ATP-mediated CLL lymphocyte killing. Cells (1 × 106/ml) from the three CLL subjects (A–C) shown in Fig. 8⇑ were plated in 24-well plates and incubated in culture medium containing increasing ATP concentrations at 37°C for 16 h in the presence (•) or absence (○) of 10 μg/ml PMB. Surviving cells were then counted by using an Olympus IMT-2 phase-contrast microscope (Olympus, Melville, NY). Three fields from three different wells for each condition were counted. Data are averages ± SD. Where not shown, error bars are smaller than the size of the symbol.

Discussion

The P2X7R is a most intriguing plasma membrane ion channel because it has the ability to undergo a transition to a nonselective pore that causes a fast plasma membrane permeabilization to normally impermeant aqueous solutes of molecular mass up to 900 kDa (15, 17, 33). This receptor mediates several responses relevant in inflammation, such as transcription factor activation, cytokine and NO release, superoxide production, proliferation, and cytotoxicity, that make it appealing in different biomedical fields spanning from immunology to infectious diseases, from oncology to neurodegeneration. However, affinity of P2X7 for its physiological ligand (ATP) is very low (EC50 in the hundred micromolar range) and deeply affected by several factors such as divalent ion concentrations, pH, and ionic strength (16). These factors, in addition to the paucity of potent and selective agonists or antagonists, have made it difficult to exploit the peculiar features of P2X7 in therapy.

An approach that has been largely neglected in P2X7R studies is the search of positive or negative effectors (allosteric modulators) that potentiate the effect of ATP. The combined administration of ATP in combination with any such compounds might have a substantial advantage over ATP alone since, by decreasing the ATP dose, it might also reduce the side effects of ATP breakdown products or of ATP itself. In a previous work (27), we described the potentiating effect of the anti-inflammatory drug tenidap on ATP-mediated cytotoxicity. This drug caused both an increase in the affinity of the P2X7R for ATP and a potentiation of the effect. Synergism between ATP and tenidap was fully dependent on the expression of P2X7 since macrophage clones that lacked P2X7 were insensitive to treatment with either tenidap or tenidap and ATP together. Like tenidap, the cationic antibiotic PMB sensitizes several cells to ATP, whether expressing the native or recombinant P2X7.

PMB is widely used in in vitro studies as a “trap” for LPS. The linear hydrophobic N-terminal region of PMB strongly binds lipid A of LPS and this leads to toxin neutralization (34, 35). The peculiar high affinity of PMB for LPS has prompted the use of this cationic antibiotic to remove endotoxin from laboratory material and solutions. ATP is well known to trigger IL-1β release from mononuclear phagocytes via activation of the P2X7R (5). The most potent stimulus for IL-1β secretion is currently thought to be LPS, which is active even at concentrations of a few nanograms per liter. Therefore, to exclude artifacts attributable to the potential contamination by LPS, we routinely ran control experiments in the presence of PMB. This brought to our attention that PMB affected the ATP response in a fashion that could not be simply explained on the basis of its known LPS-binding activity. In particular, we noticed that short-term incubations in the presence of PMB caused a potentiation of short-term ATP-dependent effects. Thus, we performed an in depth investigation of PMB effects in different cell types.

PMB by itself had no effect on [Ca2+]i changes, plasma membrane permeability or cell viability up to a concentration of 20 μg/ml, and incubation times up to 24 h (D. Ferrari and F. Di Virgilio, unpublished results). On the contrary, in the presence of ATP, Ca2+ influx, ethidium bromide uptake, and release of LDH were strongly potentiated by PMB both in cells transfected with P2X7 and in lymphocytes from CLL patients. The effect of PMB in macrophages (whether human or mouse) was somewhat different because in these cells potentiation of the ATP response was observed at low antibiotic concentrations, while high concentrations were inhibitory. PMB strongly potentiated ATP-mediated cytotoxicity in HEK293-P2X7, K562-P2X7, and CLL cells. Macrophages were less sensitive to this potentiating effect. We do not have an explanation for this different behavior, but it might be due to modulatory effects of accessory proteins differentially expressed in the HEK293 and K562 cells, B lymphocytes, and macrophages. Finally, it has to be stressed that, although is has been reported to increase the plasma membrane conductance of mammalian cells to small ions (35), in our hands PMB alone neither affected [Ca2+]i nor had any cytotoxic effects; furthermore, the delayed [Ca2+]i increase typically observed in CLL lymphocytes treated with PMB always required the simultaneous presence of ATP and could be blocked by oATP. It is of interest that although in macrophages ATP-dependent cytotoxicity was minimally potentiated by PMB, another typical P2X7R-dependent proinflammatory response, i.e., IL-1β secretion, was strongly enhanced, although at a PMB dose (10 μg/ml) well above the optimal concentration for potentiation of the [Ca2+]i changes.

Our observations suggest that P2X7R function is differently modulated in native or heterologous expression systems. Although several proteins putatively interacting with P2X7 have been identified in HEK293-P2X7 cells (18), how the native P2X7R is regulated by plasma membrane or cytoplasmic proteins is still an open question. Anyway, although a formal proof is lacking, we think that both early and delayed increases in plasma membrane permeability are attributable to a P2X7 pore opening.

Cells lacking the P2X7R were fully resistant to the application of PMB, whether alone or in combination with ATP at all concentrations tested. Furthermore, PMB added after KN-62 relieved the inhibitory effect of this antagonist. Taken together, these two observations suggest that PMB directly interacts with the P2X7R. The site of action of PMB cannot be identified at the present stage, but it is unlikely to reside in the carboxyl-terminal tail of the P2X7R since the sensitivity to ATP of cells expressing P2X7ΔC, truncated at residue 415, is also enhanced by PMB treatment. This conclusion is also supported by the lack of inhibitory activity of the 573–590 P2X7 peptide. Ability to relieve the inhibitory action of KN-62 might suggest that PMB acts near or at the KN-62 binding site that presumably lies within the extracellular domain, but whose precise location is at present unknown. However, we cannot exclude as an alternative explanation that PMB binds and sequesters KN-62 due to nonspecific hydrophobic interactions.

The potentiating effect of PMB may have practical applications in the development of ATP-based therapy. We have recently shown that lymphocytes from patients affected by the aggressive variant of CLL have a higher level of expression and function of the P2X7R (9). This makes the CLL lymphocytes eminently sensitive to the cytotoxic effect of ATP, while lymphocytes from healthy subjects are resistant. Therefore, it would be possible in principle to administer ATP as an anticancer drug in patients affected by the aggressive form of CLL. However, ATP and ATP degradation products have several local and systemic side effects that make it advisable to keep the concentration of this nucleotide in the extracellular space as low as possible. As shown by our data, PMB alone has no effect on cell viability even during prolonged incubations, but at the same time strongly synergized with ATP. Our data show that in the presence of PMB, an otherwise ineffective ATP concentration (600 μM) causes the deletion of 50–80% of target cells. In addition, PMB might be the lead to identify additional compounds that synergize with ATP at the level of the P2X7R and might have a use in therapy (36).

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • ↵1 This work was supported by grants from the Italian Ministry of Education, University and Scientific Research, the National Research Council of Italy, the Italian Association for Cancer Research, the Italian Space Agency and by local funds from the University of Ferrara.

  • ↵2 D.F. and C.P. contributed equally to this work.

  • ↵3 Address correspondence and reprint requests to Dr. Davide Ferrari, Department of Experimental and Diagnostic Medicine, Section of General Pathology, University of Ferrara, Via Borsari 46, I-44100 Ferrara, Italy. E-mail address: dfr{at}unife.it

  • ↵4 Abbreviations used in this paper: CLL, chronic lymphocytic leukemia; BzATP, 2′,3′-(4-benzoyl)-benzoyl-ATP; oATP, oxidized ATP; PMB, polymyxin B; fura 2-AM, fura 2-acetoxymethyl ester; LDH, lactic dehydrogenase; [Ca2+]i, intracellular Ca2+ concentration; wt, wild type.

  • Received November 24, 2003.
  • Accepted July 26, 2004.

References

  1. ↵
    Dubyak, G. R., C. El-Moatassim. 1993. Signal transduction via P2-purinergic receptors for extracellular ATP and other nucleotides. Am. J. Physiol. 265:C577.
    OpenUrlPubMed
  2. ↵
    Di Virgilio, F., P. Chiozzi, D. Ferrari, S. Falzoni, J. M. Sanz, A. Morelli, M. Torboli, G. Bolognesi, O. R. Baricordi. 2001. Nucleotide receptors: an emerging family of regulatory molecules in blood cells. Blood 97:587.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Ferrari, D., S. Wesselborg, M. K. A. Bauer, K. Schulze-Osthoff. 1997. Extracellular ATP activates transcription factor NF-κB through the P2Z purinoceptor by selectively targeting NF-κB p65. J. Cell Biol. 139:1635.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    Denlinger, L. C., P. L. Fisette, K. A. Garis, G. Vasquez-Torres, A. Kwom., A. D. Simon, B. Nguyen, R. A. Proctor, P. J. Bertics, J. A. Corbett. 1996. Regulation of inducible nitric oxide synthase expression by macrophage purinoceptors and calcium. J. Biol. Chem. 271:337.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Ferrari, D., P. Chiozzi, S. Falzoni, M. Dal Susino, L. Melchiorri, O. R. Baricordi, F. Di Virgilio. 1997. Extracellular ATP triggers IL-1β release by activating the purinergic P2Z receptor of human macrophages. J. Immunol. 159:1451.
    OpenUrlAbstract
  6. ↵
    Langston, H. P., Y. Ke, A. T. Gewirtz, K. E. Dombrowski, J. A. Kapp. 2003. Secretion of IL-2 and IFN-γ, but not IL-4, by antigen-specific T cells requires extracellular ATP. J. Immunol. 170:2962.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Solini, A., P. Chiozzi, A. Morelli, R. Fellin, F. Di Virgilio. 1999. Human primary fibroblasts in vitro express a purinergic P2X7 receptor coupled to ion fluxes, microvesicle formation and IL-6 release. J. Cell Sci. 112:297.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Hide, I., M. Tanaka, A. Inoue, K. Nakajima, S. Kohsaka, K. Inoue, Y. Nakata. 2000. Extracellular ATP triggers tumor necrosis factor-α release from rat microglia. J. Neurochem. 75:965.
    OpenUrlCrossRefPubMed
  9. ↵
    Adinolfi, E., L. Melchiorri, S. Falzoni, P. Chiozzi, A. Morelli, A. Tieghi, A. Cuneo, G. Castoldi, F. Di Virgilio, O. R. Baricordi. 2002. P2X7 receptor expression in evolutive and indolent forms of chronic B lymphocytic leukemia. Blood 99:706.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Wiley, J. S., L. P. Dao-Ung, B. J. Gu, R. Sluyter, A. N. Shemon, C. Li, J. Taper, J. Gallo, A. Manoharan. 2002. A loss-of-function polymorphic mutation in the cytolytic P2X7 receptor gene and chronic lymphocytic leukaemia: a molecular study. Lancet 359:1114.
    OpenUrlCrossRefPubMed
  11. ↵
    Thunberg, U., G. Tobin, A. Johnson, O. Söderberg, L. Padyukov, M. Hultdin, L. Klareskog, G. Enblad, C. Sundström, G. Roos, R. Rosenquist. 2002. Polymorphism in the P2X7 receptor gene and survival in chronic lymphocytic leukaemia. Lancet 360:1935.
    OpenUrlCrossRefPubMed
  12. ↵
    Zhang, L. Y., R. E. Ibbotson, J. A. Orchard, A. C. Gardiner, R. V. Seear, A. J. Chase, D. G. Oscier, N. C. Cross. 2003. P2X7 polymorphism and chronic leukemia: lack of correlation with incidence, survival and abnormalities of chromosome 12. Leukemia 17:2097.
    OpenUrlCrossRefPubMed
  13. ↵
    Starczynski, J., C. Pepper, G. Pratt, L. Hooper, A. Thomas, T. Hoy, D. Milligan, P. Bentley, C. Fegan. 2003. The P2X7 receptor gene polymorphism 1513 AC has no effect on clinical prognostic markers, in vitro sensitivity to fludarabine, Bcl-2 protein expression or survival in B-cell chronic lymphocytic leukemia. Br. J. Haematol. 123:66.
    OpenUrlCrossRefPubMed
  14. ↵
    Nuckel, H., U. H. Frey, J. During, U. Duhrsen, W. Siffert. 2004. 1513A/C polymorphism in the P2X7 receptor gene in chronic lymphocytic leukemia: absence of correlation with clinical outcome. Eur. J. Haematol. 72:259.
    OpenUrlCrossRefPubMed
  15. ↵
    Di Virgilio, F.. 1995. The P2Z purinoceptor: an intriguing role in immunity, inflammation and cell death. Immunol. Today 16:524.
    OpenUrlCrossRefPubMed
  16. ↵
    Surprenant, A., F. Rassendren, E. Kawashima, R. A. North, G. Buell. 1996. The cytolytic P2Z receptor for extracellular ATP identified as a P2X receptor (P2X7). Science 272:735.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    North, R. A.. 2002. The molecular physiology of P2X receptors. Physiol. Rev. 82:1013.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Kim, M., L. H. Jiang, H. L. Wilson, R. A. North, A. Surprenant. 2001. Proteomic and functional evidence for a P2X7 receptor signalling complex. EMBO J. 20:6347.
    OpenUrlAbstract
  19. ↵
    Wilson, H. L., S. A. Wilson, A. Surprenant, R. A. North. 2002. Epithelial membrane proteins induce membrane blebbing and interact with the P2X7 receptor C terminus. J. Biol. Chem. 277:34017.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Labasi, M., N. Petrushova, C. Donovan, S. McCurdy, P. Lira, M. M. Payette, W. Brissette, J. R. Wicks, L. Audoly, C. A. Gabel. 2002. Absence of the P2X7 receptor alters leukocyte function and attenuates an inflammatory response. J. Immunol. 168:6436.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Evans, R. J., C. Lewis, G. Buell, R. A. North, A. Surprenant. 1995. Pharmacological characterization of heterologously expressed ATP-gated cation channels (P2X purinoceptors). Mol. Pharmacol. 48:178.
    OpenUrlAbstract
  22. ↵
    Murgia, M., S. Hanau, P. Pizzo, M. Rippa, F. Di Virgilio. 1993. Oxidized ATP: an irreversible inhibitor of the macrophage purinergic P2Z receptor. J. Biol. Chem. 268:8199.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    Beigi, R., S. Kertesy, G. Aquilina, G. R. Dubyak. 2003. Oxidized ATP (oATP) attenuates proinflammatory signaling via P2 receptor-independent mechanisms. Br. J. Pharmacol. 140:507.
    OpenUrlCrossRefPubMed
  24. ↵
    Di Virgilio, F.. 2003. Novel data point to a broader mechanism of action of oxidized ATP: the P2X7 receptor is not the only target. Br. J. Pharmacol. 140:441.
    OpenUrlCrossRefPubMed
  25. ↵
    Gargett, C. E., J. S. Wiley. 1997. The isoquinoline derivative KN-62: a potent antagonist of the P2Z-receptor of human lymphocytes. Br. J. Pharmacol. 120:1483.
    OpenUrlCrossRefPubMed
  26. ↵
    Baraldi, P. G., M. del Carmen Nunez, A. Morelli, S. Falzoni, F. Di Virgilio, R. Romagnoli. 2003. Synthesis and biological activity of N-arylpiperazine-modified analogues of KN-62, a potent antagonist of the purinergic P2X7 receptor. J. Med. Chem. 46:1318.
    OpenUrlCrossRefPubMed
  27. ↵
    Sanz, J. M., P. Chiozzi, F. Di Virgilio. 1998. Tenidap enhances P2Z/P2X7 receptor signalling in macrophages. Eur. J. Pharmacol. 355:235.
    OpenUrlCrossRefPubMed
  28. ↵
    Adinolfi, E., M. Kim, M. T. Young, F. Di Virgilio, A. Surprenant. 2003. Tyrosine phosphorylation of Hsp90 within the P2X7 receptor complex negatively regulates P2X7 receptors. J. Biol. Chem. 278:37344.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    Hancock, R. E.. 2001. Cationic peptides: effectors in innate immunity and novel antimicrobials. Lancet Infect. Dis. 1:156I.
    OpenUrlCrossRef
  30. ↵
    Falzoni, S., M. Munerati, D. Ferrari, S. Spisani, S. Moretti, F. Di Virgilio. 1995. The purinergic P2Z receptor of human macrophage cells: characterization and possible physiological role. J. Clin. Invest. 95:1207.
    OpenUrlCrossRefPubMed
  31. ↵
    Morelli, A., P. Chiozzi, A. Chiesa, D. Ferrari, J. M. Sanz, S. Falzoni, P. Pinton, R. Rizzato, M. F. Olson, F. Di Virgilio. 2003. Extracellular ATP causes ROCK 1-dependent bleb formation in P2X7-transfected HEK293 cells. Mol. Biol. Cell 14:2655.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    Denlinger, L. C., P. L. Fisette, J. A. Sommer, J. J. Watters, U. Prabhu, G. R. Dubyak, R. A. Proctor, P. J. Bertics. 2001. Cutting edge: the nucleotide receptor P2X7 contains multiple protein- and lipid-interaction motifs including a potential binding site for bacterial lipopolysaccharide. J. Immunol. 167:1871.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Rassendren, F., G. N. Buell, C. Virginio, G. Collo, R. A. North, A. Surprenant. 1997. The permeabilizing ATP receptor, P2X7: cloning and expression of a human cDNA. J. Biol. Chem. 272:5482.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    Tsubery, H., I. Ofek, S. Cohen, M. Fridkin. 2000. The functional association of polymyxin B with bacterial lipopolysaccharide is stereospecific: studies on polymyxin B nonapeptide. Biochemistry 39:11837.
    OpenUrlCrossRefPubMed
  35. ↵
    Berg, J. R., C. M. Spilker, S. A. Lewis. 1996. Effects of polymyxin B on the mammalian urinary bladder. J. Membr. Biol. 154:119.
    OpenUrlCrossRefPubMed
  36. ↵
    Daugelavicius, R., E. Bakiene, D. H. Bamford. 2000. Stages of polymyxin B interaction with the Escherichia coli cell envelope. Antimicrob. Agents Chemother. 44:2969.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section