Diphenhydramine inhibits voltage-gated proton channels (Hv1) and induces acidification in leukemic Jurkat T cells- New insights into the pro-apoptotic effects of antihistaminic drugs

ABSTRACT An established characteristic of neoplastic cells is their metabolic reprogramming, known as the Warburg effect, with greater reliance on energetically less efficient pathways (such as glycolysis and pentose phosphate shunt) compared with oxidative phosphorylation. This results in an overproduction of acidic species that must be extruded to maintain intracellular homeostasis. We recently described that blocking the proton currents in leukemic cells mediated by Hv1 ion channels triggers a marked intracellular acidification and apoptosis induction. Moreover, histamine H1-receptor antagonists were found to induce apoptosis in tumoral cells but the mechanism is still unclear. By using Jurkat T cells, we now show how diphenhydramine inhibits Hv1 mediated currents, inducing a drop in intracellular pH and cellular viability. This provides evidence of a new target structure responsible of the known pro-apoptotic action of antihistaminic drugs.


Introduction
Diphenhydramine (DPH) is a well-known first-generation histamine H1-receptor antagonist, commonly used in treatment of allergic diseases. Its usefulness is principally related to a decrease of histamine effects produced during the hypersensitivity reaction. Furthermore, it has been reported that DPH has other molecular targets such as muscarinic and NMDA glutamatergic receptors which explain most of its adverse reactions. 1,2 On the other hand, Jangi et al. clearly described that DPH promoted leukemic Jurkat T cell death by apoptosis at higher doses than those needed for its antihistaminic action. 3 The authors later found similar results in other malignant cell lines. 4,5 Although these observations are reproduced with several antihistaminic drugs, other evidence contradicts the involvement of the H1 receptor pathway in the apoptosis mechanism as it is exerted even in presence of exogenous histamine, a-fluoromethylhistidine treatment or H1 receptor knockdown. Moreover, antihistaminic pro-apoptotic action affected neither cAMP nor cGMP intracellular levels, both widespread second messengers. 4,5 In addition, it is widely known that neoplastic cells show significant changes in glucose metabolism relying on glycolytic and pentose phosphate pathways instead of oxidative phosphorylation for energy production. 6,7 This fact leads to acidic species overproduction and therefore a strong need for proton extrusion to avoid intracellular acidification. 8 In this sense, several proton transporters were found overexpressed in tumor cells and they have been a matter of study in recent years. MCT, NHE, CA, NBC and V-ATPase (see abbreviations below) are the structures that most captured the attention in this field. 9,10 Although less studied, the Hv1 channel has a strong capacity of restoring intracellular pH after heavy acid loads; and as it is a passive transport pathway, its activity does not require any expense of energy. 11 This channel has been found upregulated in several tumor cell lines [12][13][14][15][16] and clinical isolates, correlating with poor prognosis in human breast and colorectal cancer. 17,18 In the same manner, other ion channels are emerging as key players in tumoral development. 19,20 Recently our group showed that Hv1 channel in Jurkat T cells is indispensable for pHi regulation as its inhibition exerted by Zn 2C or ClGBI, both proved Hv1 blockers, 21 induced an immediate drop in basal pHi along with significant impairment of the cell's capacity for restoring intracellular pH after heavy acid loads. Moreover, we observed that upon Hv1 channel blockade, acidification progressed from a pH around 7.2 to values below 6.8 and promoting cell death by apoptosis. 22 In addition, Kim et al. have recently reported that DPH and chlorpheniramine induce a significant inhibitory effect on Hv1 currents in murine BV12 microglial cells. 23 These authors also assert that Hv1 inhibition is independent of histamine receptor activity, as histamine does not modify any property of Hv1 currents.
Overall, we hypothesized that, if DPH is able to inhibit Hv1 in human Jurkat T cells (likewise it does in microglial cells), it could induce cell acidification as an early event of its known pro-apoptotic effect. 3 In this work we present electrophysiological data showing the inhibition by DPH on Hv1 currents in Jurkat T cells and its action on intracellular pH.

Results
We first examined the effect of DPH on whole-cell currents mediated by Hv1 channels in Jurkat T cells, depicted in Fig. 1. Extracellular perfusion of 0.1 mM DPH significantly inhibited the Hv1 current (Fig. 1a), achieving its maximal effect between 3 and 5 minutes. Moreover, 0.1 mM histamine affected neither the Hv1 current nor its inhibition by DPH. The effect of DPH was partially reversible (around 80%) in all tested cells after 15 min. (One Way ANOVA and Holm-Sidak post hoc analysis, n D 4-8, p < 0.05).
We next evaluated the effect of long-term DPH mediated Hv1 inhibition on pHi. Jurkat T cells were incubated with 0.1, 0.5 and 1.0 mM DPH for 24 hours alone or in presence of 0.1 mM histamine (experimental conditions used by Jangi et. al. 3 to demonstrate DPH pro-apoptotic effects). Then pHi, cell size and complexity were simultaneously analyzed by flow cytometry and the results are presented in Fig. 2.
As it can be seen, in 0.5 and 1 mM DPH conditions a clear change in cell size and complexity (a typical phenomenon of the apoptotic process) occurred in  Simultaneously, fluorescence emission of the pH sensitive ratiometric probe BCECF was measured in the same cells, where a decrease in the values of FL1/ FL3 ratio means intracellular acidification (Fig. 2b). Moreover, using the calibration curve performed in each experiment, FL1/FL3 ratio values were used to calculate the mean pHi value for each population  (Fig 2d). The latter figure shows that DPH, in parallel with viability impairment, induced a significant dose dependent acidification that was also unaffected by 0.1 mM histamine (Two Way ANOVA and Holm-Sidak post hoc analysis, n D 5-6, p<0.05). Neither the viability nor the pHi showed differences in control conditions (0 mM DPH) with or without 0.1 mM histamine (Student's t-test, p<0.05). These results are coherent with our patch-clamp recordings, where the presence of 100 mM of histamine did not modify Hv1 currents or the inhibition exerted by DPH.

Discussion
We previously demonstrated that Zn 2C and ClGBI, two Hv1 channel blockers, induced apoptosis of Jurkat T cells following a profound intracellular acidification. 22 Now, in this work, we describe that DPH also inhibited proton currents in these leukemic cells, resulting in a drop on pHi in the same conditions where apoptosis was described previously. 3 Our initial hypothesis has thus been reinforced by the results obtained with another compound that inhibits Hv1 channels, and the observed effect on pHi and apoptosis were similar to the ones observed with Zn 2C and ClGBI.
Our results and those reported by Jangi et al. showed the same concentration-dependent behavior in a range of concentration where the fractional occupancy of the H1-receptor is constant (remaining above the 99,9%, calculated assuming a pK i D 7.9, IUPHAR DB) suggesting that DPH effects were not a consequence of H1 histamine receptor blockade.
In addition, the facts that both human Jurkat T cells and murine BV12 microglial cells 23 exhibit the same percentage of proton current DPH inhibition with the same concentrations, as well as the short period of time action needed (3-5 min.) let us speculate that a direct interaction Hv1-DPH is involved in DPH pro-apoptotic effects.
Altogether, Hv1 inhibition is reinforced as a pro-apoptotic stimulus and a mechanism plausible to clarify the unexplained anti-tumorigenic properties of certain drugs.

Cell culture
Jurkat T cells were grown in DMEM High Glucose (25 mM) medium supplemented with 10% (vol/vol) heat-inactivated fetal bovine serum (Internegocios), in 5% CO2/95% humidified air at 37 C at an average density of 10 6 cells per ml.

Patch clamp experiments
The cells were observed with a mechanically stabilized, inverted microscope (Telaval 3, Carl Zeiss, Jena) equipped with a 40x objective lens. The standard tight-seal whole-cell configuration of the patch-clamp technique was used to record macroscopic whole-cell currents. 24  For Hv1 blockade experiments, DPH solutions were made adding appropriate amounts of 100 mM aqueous stock solution to the ESS the same day of the experiment.

Flow cytometry pHi determinations
Cells were incubated in 96-well plates (200 ml/well) at a starting concentration of 0.5 £ 10 6 cells/ml and cultured in the conditions abovementioned (see "Cell culture" section). Cells were exposed for 24 h to DPH 0.1, 0.5 and 1.0 mM before flow cytometry measurements. As DPH and histamine have been dissolved in aqueous media, control condition is simply medium addition in the same volume as stimulus.
The protocol described in Current Protocols in Cytometry (1997) 25 was used for pH measurement with BCECF using Pseudo Null Calibration (also depicted by P. Frank et al 26 and D. A. Eisner et al., 27 among others). Briefly, after incubation cells were centrifuged 5 min at 500 rpm and loaded with 2 mg/ml BCECF-AM 15 min at 37 C, centrifuged and resuspended in 10% FBS-HEPES solution. Prior measurement every batch of cells were exposed to the corresponding treatment condition at the same concentration of incubation to prevent eventual pHi recovery. Pseudo Null Calibration curve was performed according to Chow et al 25

Viability determinations
Simultaneous to FL1 and FL3, FSC and SSC intensity was acquired from the same batch of cells of pHi determinations; these parameters reflect cell size and granularity, respectively. Typically, during apoptosis progression cells reduce its volume and increase its intracellular complexity, so the percentage viable cells has been estimated as those that remains in the FSC/ SSC gate were control cells commonly reside.

Statistics
The results are expressed as mean § standard error of the mean (SEM). Paired or unpaired Student's t tests were used to compare 2 groups. ANOVA (One Way or Two Way) test was used to compare 3 or more groups. In all cases, a P value lower than 0.05 was considered for establishing statistically significant differences.