Histamine H1 receptor antagonists selectively kill cisplatin-resistant human cancer cells

Cancer therapy is often hampered by the disease’s development of resistance to anticancer drugs. We previously showed that the autonomously upregulated product of fibroblast growth factor 13 gene (FGF13; also known as FGF homologous factor 2 (FHF2)) is responsible for the cisplatin resistance of HeLa cisR cells and that it is likely responsible for the poor prognosis of cervical cancer patients treated with cisplatin. Here we show that cloperastine and two other histamine H1 receptor antagonists selectively kill HeLa cisR cells at concentrations that little affect parental HeLa S cells. The sensitivity of HeLa cisR cells to cloperastine was abolished by knocking down FGF13 expression. Cisplatin-resistant A549 cisR cells were similarly susceptible to cloperastine. H2, H3, and H4 receptor antagonists showed less or no cytotoxicity toward HeLa cisR or A549 cisR cells. These results indicate that histamine H1 receptor antagonists selectively kill cisplatin-resistant human cancer cells and suggest that this effect is exerted through a molecular mechanism involving autocrine histamine activity and high-level expression of FGF13. We think this represents a potential opportunity to utilize H1 receptor antagonists in combination with anticancer agents to treat cancers in which emergent drug-resistance is preventing effective treatment.


Establishment of cisplatin-resistant A549 cisR cells.
Cisplatin-resistant A549 cisR cells were established as previously described for HeLa cisR cells. Briefly, parental human lung carcinoma A549 cells (RCB0098; freshly obtained from RIKEN Bio Resource Center) were maintained in growth medium composed of Dulbecco's modified eagle medium (DMEM; Wako Chemicals) with 10% FBS and 60 µg/ml kanamycin (Meiji Seika Pharma, Tokyo, Japan). By subculturing these cells in the presence of incrementally increasing concentrations of cisplatin, we established a highly resistant subline, A549 cisR cells, which were maintained in growth medium containing 3 µg/ml cisplatin. Measurement of cytotoxicity based on cell proliferation. Cytotoxicity was analyzed as described previously 1 . Cells (5 × 10 3 cells/100 µL/well) were seeded onto 96-well plates and incubated at 37 °C under 5% CO 2 . After 24 h, various compounds were added concomitantly in a volume of 2 µL/well to achieve the desired final concentrations, after which the cells were incubated for an additional 3 days. Cytotoxicity was measured based on cell numbers as follows. Cell Counting Kit (CCK)-8 reagent (a modified MTT assay system; Dojindo Laboratories, 5 µL/well) was added to each well, and the plates were incubated for 5 h, after which the absorbance at 450 nm was measured using a microplate reader. The ratios of the cell numbers in the drug-containing cultures to those in the control drug-free cultures were then calculated. To evaluate the cytotoxic effects exerted by two compounds, the compounds were added to the cultures concomitantly without delay.
Evaluation of cytotoxicity toward mixed cell populations. The cytotoxicity of various compounds was tested with HeLa S cells, HeLa cisR cells, and mixed populations of the two. For the separate cultures, 4 × 10 3 HeLa S cells/well or 1.5 × 10 4 HeLa cisR cells/well were analyzed. For the mixed population (combined culture), HeLa S cells (2 × 10 3 cells/well) and HeLa cisR cells (7.5 × 10 3 cells/well) were mixed and then seeded into 96-well plates. After incubation for 1 day to allow the cells to attach, selected compounds were added in 2 µL/well to achieve the desired final concentrations, and the cells were incubated for an additional 3 days. Cytotoxicity was then measured based on cell numbers as described above.
Measurement of histamine-induced cell proliferation. Cells (5 × 10 3 cells/100 µL/well) were seeded into 96-well plates and incubated at 37 °C under 5% CO 2 for 1 day. The attached cells were washed twice with serum-free medium and then serum-starved by incubation for 3 days in culture medium containing 0.1% serum. Thereafter, the cells were washed with serum free medium incubated for 7 days in 100 µL/well of culture medium containing selected concentrations of FBS and histamine (added in 2 µL/well). Cell numbers were then Scientific Reports | (2021) 11:1492 | https://doi.org/10.1038/s41598-021-81077-y www.nature.com/scientificreports/ measured as described above, and the ratios of the cell numbers in histamine-containing cultures to those in the control histamine-free cultures were calculated.
Evaluation of the effects of histamine and cloperastine on cell proliferation. Cells (5 × 10 3 cells/100 µL/well) were seeded into 96-well plates and incubated at 37 °C under 5% CO 2 . After 1 day, histamine was added in 2 µL/well, and the cells were incubated for an additional day. Cloperastine was then added in 2 µL/ well, and the cells were incubated for another 2 days. During the final 5 h of incubation, cell numbers and ratios were determined as described above.
Flow cytometric analyses. Flow cytometric analyses were conducted using a Model SH-800 flow cytometer (SONY, Tokyo, Japan) following the manufacturer's instructions. For flow cytometric analysis of the cell cycle, growing cells in 100-mm dishes were incubated for 3 days in their respective growth media (control) or in medium containing 1 µg/ml cisplatin or 100 µM cloperastine. The cells were then washed, fixed with 70% ethanol, stained with propidium iodide (PI) and subjected to flow cytometry. For flow cytometric analysis of apoptosis/cell death, growing cells in 100-mm dishes were incubated for 1 day with selected compounds and washed. The unfixed cells were then stained with Annexin V-FITC and PI using an ApoAlert Annexin V-FITC Apoptosis Kit (Clonetech Laboratories, Tokyo, Japan), after which the intensities of the FITC and PI fluorescences were analyzed using a flow cytometer and plotted.

Statistical analyses.
Comparisons of the results expressed a percentages of control (e.g., HeLa cisR cells vs. HeLa S cells at particular cisplatin concentrations in Fig. 1) actually means comparison among raw data from samples with two factors and two levels. We therefore used ANOVA for the statistical analysis of these data. Dunnett's post hoc test was used to verify the significance because the analyses described above always compare the no-drug (zero concentration) measurements with measurements at a particular drug concentration. The significance of difference(s) are shown as * for p < 0.05, ** for p < 0.01, or *** for p < 0.001 upon verification with a post hoc test.

Results
Identification of cloperastine and other H1 receptor antagonists as selective and strong cytocidal drugs for HeLa cisR cells. HeLa cisR cells show greater resistance to cisplatin than the parental HeLa S cells (Fig. 1a,c). However, by screening nearly the entire approved drug library (1492 drugs currently approved by PMDA, Japan), we found that cloperastine, an antitussive drug with H1 histamine receptor antagonist activity 5 , killed 50% of HeLa cisR cells at a concentration of 40 µM and 100% at 80 µM (Fig. 1b,c). Cloperastine was much less cytotoxic toward cisplatin-sensitive HeLa S cells, with IC 50 higher than 120 µM. Even at 80 µM, 90% of HeLa S cells remained viable in the presence of cloperastine (Fig. 1b). Two other H1 receptor antagonists, desloratadine and clemastine, exhibited similar cytotoxicity toward HeLa cisR cells, though desloratadine was somewhat less selective. Desloratadine and clemastine killed nearly 100% of HeLa cisR cells at 50 µM and 20 µM, respectively (Fig. 1d,e). By contrast, the H2 receptor antagonist nizatidine showed little cytotoxicity toward either HeLa cisR or HeLa S cells (Fig. 1f), and while the H3 receptor antagonist pitolisant and H4 receptor antagonist JNJ-7777120 were both selectively cytotoxic toward HeLa cisR cells, killing was incomplete (55% and 70% viable, respectively) at the highest achievable concentrations tested (100 µM; Fig. 1g,h). HeLa cisR cells and HeLa S cells were treated without or with H1 receptor antagonists at their approximate IC 50 , and the effect of cisplatin was evaluated to see if there is any synergism between these drugs. We found that cisplatin cytotoxicity toward HeLa cisR cells is not synergistically enhanced by cloperastine, desloratadine, or clemastine (Fig. 2a-c). However, when an H1 receptor antagonist combined with cisplatin was applied to a mixed culture of HeLa S and HeLa cisR cells, the combination of cloperastine and cisplatin killed all the cells in the mixed cell population (Fig. 3a,b).
Cytotoxicity of H1 receptor antagonists positively correlates with resistance to cisplatin and FGF13 expression. Interestingly, an FGF13 knock-down line of HeLa cisR cells (FGF13kd, Fig. 4a), which are cisplatin-sensitive (Fig. 4b), survived in the presence of cloperastine at concentrations cytotoxic to HeLa cisR cells, in which FGF13 is upregulated (Fig. 4a,c). Desloratadine (Fig. 4d)  Note that only H1 receptor antagonists resulted in 0% cell viability. Also note that error bars are not visible in many places due to their small size. The significance of the difference between the results from HeLa S and HeLa cisR cells at the same concentration of the indicated compound was determined with ANOVA (***p < 0.001; **p < 0.01; *p < 0.05). (c) HeLa cisR and HeLa S cells cultured for 3 days with 10 µg/ml cisplatin or 100 µM cloperastine were photographed under a phase contrast microscope. Bar, 50 µm. Consistent with that activity, histamine at 2.7 mM (500 µg/ml) improved the viability of HeLa cisR cells but not HeLa S cells in the presence of cloperastine ( Fig. 5f).

Scientific Reports
Regarding the concentrations of histamine used, according to one study 6 , the pathophysiological concentrations of histamine in circulating blood may vary from 40 ng/mL in a healthy individual to 121 ng/ml in an ischemic heart disease patient. Another study, focusing on the effects of histamine on mast cell degranulation 7 , used histamine at concentrations ranging from 0.3 mM (33 µg/mL) to 32 mM (3556 µg/mL). Thus, the precise concentration of histamine in the cellular microenvironment may be much higher than that in the circulating blood. In the present study, therefore, to identify the potential involvement of histamine in the regulation of cancer cells, we experimentally determined the concentrations of histamine to be used.

Effects of cisplatin and cloperastine on cell cycling and apoptosis in HeLa cisR cells. The effects
of cisplatin and cloperastine on HeLa cisR and HeLa S cell cycling were examined using flow cytometry. The cells were cultured in FBS-containing growth medium in the presence of a compound for 3 days prior to fixation and staining with PI for flow cytometric analysis. Serum-containing medium was used because the cell cycles of HeLa cisR and HeLa S cells could not be synchronized, even in the absence of serum (Matsumoto, unpublished observation). Among control HeLa S cells, G1 phase cells (2n) predominated ( Fig. 6a,b). When cells were treated with 1 µg/ml cisplatin, there was a dramatic decrease in cells in G1 phase and increases in cells in sub-G1 phase (sb) and G2/M phase (4n) (Fig. 6a,b). Treatment with 100 µM cloperastine weakly increased numbers of sub-G1 and G2/M phase cells but did not significantly reduce G1 phase cells (Fig. 6a,b). In contrast to HeLa S cells, a relatively small fraction of control HeLa cisR cells were in G1 phase ( Fig. 6a,b). When cells were treated with 1 µg/ml cisplatin, the cell fractions at each cell cycle phase were largely unchanged ( Fig. 6a,b). Treating HeLa cisR cells with 100 µM cloperastine dramatically increased the sub-G1 phase fraction and reduced the S phase and G2/M phase fractions (Fig. 6a,b), suggesting the occurrence of apoptosis.
We then used Annexin V-PI staining assays to examine the effects of cisplatin and cloperastine on apoptosis/ cell death among HeLa cisR and HeLa S cells. For these assays, growing cells in heterogenous cell-cycle phases were cultured with cisplatin or cloperastine for 1 day before the unfixed cells were stained with Annexin V and PI prior to cytometric analysis. Cells undergoing early phase apoptosis are shown as Annexin V HIGH PI LOW cells, while those undergoing late phase apoptosis are shown as Annexin V HIGH PI HIGH cells. Cells undergoing plasma membrane disintegration are shown as Annexin V LOW PI HIGH cells. The majority of control HeLa cisR and HeLa S cells were Annexin V LOW PI LOW , which indicates that there had been little artificial damage to the analyzed cells (Fig. 7a,b). When HeLa cisR cells were treated with cloperastine, the incidences of both early phase and late phase apoptosis increased markedly (Fig. 7a,b). When HeLa cisR cells were treated with 1 µg/ml cisplatin, the cell distribution did not change significantly as compared to control cells (Fig. 7b). By contrast, when HeLa S cells were treated with 1 µg/ml cisplatin, there were dramatic increases in cells in the late phase of apoptosis or exhibiting plasma membrane disintegration (Fig. 7a,b). Treating HeLa S cells with 100 µM cloperastine did not greatly affect the numbers of cells falling into either of those groups (Fig. 7a,b).
Signal transduction pathways downstream of FGF receptor tyrosine kinases are not involved in the cloperastine cytotoxicity toward HeLa cisR cells. FGF13 reportedly exerts its effects through direct binding to multiple intracellular binding partners, but not FGF receptor tyrosine kinases 8 . For instance, FGF13 enhances p38 MAP kinase (MAPK) signaling by binding to its scaffold protein, Id2. To determine whether MAPK signaling is involved in FGF13′s apparent activity enhancing the cloperastine sensitivity of HeLa   Fig. 8, the cytotoxic effects of cisplatin (Fig. 8a,c) and cloperastine (Fig. 8b,d) toward HeLa S and HeLa cisR cells were not significantly affected by inhibition of MEK or p38 MAPK (Fig. 8a-d).
Calcium channel blocker verapamil weakly enhances the cytotoxicity of H1 receptor antagonists, and the effect is not selective to HeLa cisR cells. The stem-like properties of tumor side population cells are reportedly modulated by the calcium channel blocker verapamil, which enhances the cytotoxic effects of chemotherapeutic drugs and suppresses multidrug resistance by targeting the transport function of the P-glycoprotein 9 . We found that verapamil weakly enhanced the cytotoxicity of cloperastine toward both HeLa cisR and HeLa S cells (Fig. 9a)  www.nature.com/scientificreports/ nists tested, desloratadine and clemastine (Fig. 9b,c). Moreover, these effects of verapamil were not selective for either HeLa cisR or HeLa S cells (Fig. 9a-c).

Like HeLa cisR cells, cisplatin-resistant A549 cisR cells are selectively killed by H1 receptor antagonists.
Following the same experimental protocol used to establish HeLa cisR cells, we established cisplatin-resistant A549 cisR cells from parental A549 lung cancer cells (Fig. 10a). The A549 cisR cells were subcultured in the presence of 3 µg/ml cisplatin. As with HeLa cisR and HeLa S cells, cloperastine killed A549 cisR www.nature.com/scientificreports/ cells at lower concentrations than were needed to kill A549 cells (Fig. 10b), though the difference in sensitivity to cloperastine between A549 cisR and A549 cells was smaller than between HeLa cisR cells and HeLa S cells. Similar effects were seen with desloratadine and clemastine (Fig. 10c,d). Importantly, nearly 100% of the parental A549 cells were killed by 50 µM cloperastine, 40 µM desloratadine, or 40 µM clemastine (Fig. 10b-d), which is in contrast with > 120 µM, 100 µM, or 60 µM for the parental HeLa S cells, respectively (Fig. 1b,d,e). The H2 receptor antagonist nizatidine showed little cytotoxicity toward A549 cisR or A549 cells (Fig. 10e), and both the H3 receptor antagonist pitolisant and the H4 receptor antagonist JNJ-7777120 were selectively but less strongly cytotoxic toward A549 cisR cells (Fig. 10f,g).  Practical utility of this finding. Interestingly, cloperastine is currently used as an approved antitussive drug to treat various types of cough, including the severe cough experienced by lung cancer patients. The regular clinical dose of cloperastine as a cough suppressant (10-20 mg three times daily for adults) is calculated to be approximately 3 µM. However, chronic toxicity tests performed in rats treated for 3 months with cloperastine hydrochloride at 15 mg/kg (45 µM) or 45 mg/kg (135 µM), and in dogs treated for 3 months with cloperastine hydrochloride at 20 mg/kg (60 µM) revealed no particular symptomatology or changes in hematochemical, hematological, or urinary parameters as compared to control animals and baseline values 10 . Thus, given the priority of cancer therapy, application of cloperastine to kill cisplatin-resistant cancer cells would be well within the scope of clinical testing. The two other H1 receptor antagonists tested in this study, desloratadine and clemastine, are also promising compounds with which to kill cisplatin-resistant cancer cells.
We showed previously and confirmed here that the cisplatin resistance of HeLa cisR cells is mediated by upregulation of FGF13 expression 1 . We therefore suggest that detection of upregulated FGF13 expression in biopsy samples could be utilized as an important biomarker, as we previously reported 1 . If high-level expression of FGF13 is detected in a cancer, then use of H1 receptor antagonists may be an effective approach to therapy.
Our results also indicate that clemastine suppresses the viability of both HeLa cisR and HeLa S cells at lower concentrations than desloratadine or cloperastine (Fig. 4c-e). However, as one of the aims of this study was to understand the molecular mechanism underlying cisplatin resistance, we chose to study cloperastine, which selectively exerts a strong cytotoxic effect on HeLa cisR cells but a weaker effect on cisplatin-sensitive HeLa S cells ( Fig. 4c; LD 50 of cloperastine was 40 µM for HeLa cisR cells and > 120 µM for HeLa S cells). A second reason for using cloperastine in this study was that earlier reports had established that the concentrations required to kill HeLa cisR cells may be safely used in patients (described above). In the future, however, desloratadine or clemastine may also prove useful once their safety has been established.
Possible mechanism of histamine-dependent cell proliferation. Histamine reportedly stimulates proliferation of some types of cancer cells 11 . Indeed, we found that histamine dose-dependently enhanced the proliferation and viability of HeLa cisR cells, but not HeLa S cells, even in the presence of cloperastine (Fig. 5bf). These results, together with the finding that H1 receptor antagonists reduce the viability of HeLa cisR cells, strongly suggest that histamine or basal signaling of the histamine receptor acts as an autocrine stimulator of HeLa cisR cell proliferation. HeLa cisR cells express histidine decarboxylase (data not shown), the key enzyme in the synthesis of histamine, though its significant enhancement as compared to HeLa S cells has not been confirmed.
It is thus likely that there is an autocrine loop between histamine production and histamine recognition by H1 receptors that supports HeLa cisR cell proliferation. However, in the reported case of colorectal cancer cells, histamine increased cell proliferation via H2/H4 receptors 11 . In the present study, by contrast, it appears to be the H1 receptor that mediates transduction of intracellular signals to induce/maintain HeLa cisR and A549 cisR cell proliferation in the presence of cisplatin. In addition, we found that parental A549 cells are also susceptible to H1 receptor antagonist-induced cytotoxicity, though less so than A549 cisR cells. Thus, the importance of histamine and the histamine receptor signaling pathway for survival and proliferation may vary depending on the cancer cell type. Indeed, the authors of a recent review concluded that histamine regulates cancer-associated biological processes during cancer development in multiple cell types, and the outcome will depend on the tumor cell type, the level of histamine receptor expression, and the associated signal transduction pathways 12 . Although those authors suggested that H4 receptors may be the most promising therapeutic target for cancer treatment 12 , our results show that the H4 receptor antagonist JNJ7777120 does not effectively decrease the viability of HeLa S, HeLa cisR, A549, or A549 cisR cells. Moreover, we observed that in addition to H1 receptor antagonists, the H3 receptor antagonist pitolisant decreased the viability of A549 cisR and A549 cells (Fig. 9). We suggest that the cytocidal efficacy of H1 receptor antagonists and other histamine receptor antagonists toward various cancers should be evaluated.
What are the links among FGF13 activity, cisplatin resistance, and histamine signaling that enhance cell proliferation? The actions of FGF13 include binding to and modulating the function of the voltage-gated sodium channel (VGSC) 8 ; binding to Id2, a MAP kinase scaffold protein, to augment p38-modulated signaling 13 ; and binding to microtubules to stabilize their structure 14 . The involvement of one or more of these activities in mediating HeLa cisR cells' susceptibility to H1 receptor antagonists is not yet clear. Nonetheless, the positive correlation between the level of FGF13 expression and susceptibility to cloperastine was clearly  (Fig. 4c,f). Taken together, these findings strongly suggest that there is a molecular link between FGF13 activity, cisplatin resistance, and histamine signaling to enhance cell proliferation. Our findings that cloperastine does not affect FGF13 expression and that FGF13 knock-down abolishes HeLa cisR cell sensitivity to cloperastine strongly suggest that FGF13 is situated upstream of the cloperastine sensitivity mechanism, and that the reverse is not the case. It was recently reported that Apigenin, a major plant flavone, inhibits histamine-induced cervical cancer tumor growth in vivo 15 . The authors of the study reported that histamine stimulates cervical tumor growth in vivo and in vitro by altering estrogen receptor (ER; ER-α and ER-β) expression levels and signaling. It was speculated that apigenin inhibited cervical tumor growth by reversing this histamine-induced abnormal ER signaling 15 . If this mechanism applies to HeLa cisR cells, then their growth in the presence of cisplatin may be more dependent on histamine-mediated abnormal ER signaling than is the growth of HeLa S cells. Although no direct linkage between FGF13 and ER signaling has yet been reported, it was recently found that FGF13 promotes metastasis of triple-negative breast cancer, which tests negative for ERs 4 . This suggests FGF13 may provide a highly metastatic nature to these cancer cells that is not positively dependent on signaling by ERs. A fuller understanding of the mechanism by which FGF13 supports the activity of histamine receptor antagonists to effectively kill HeLa cisR cells and A549 cells will require further investigation.
Additional implications of the present findings. It was recently reported that FGF13 is expressed in cancer cells and that its high level contributes to the cells' invasive 16 and metastatic 4 activity. We also found that HeLa cisR cells expressing high levels of FGF13 show greater mobility than HeLa S cells in in vitro scratch wound healing assays (data not shown). Together with our finding of a positive correlation between high FGF13 expression and cloperastine sensitivity, we anticipate that highly metastatic cancer cells expressing high levels of FGF13 can be selectively killed with cloperastine and other H1 receptor antagonists, providing a new approach to treating metastatic cancers. This possibility will need to be examined in future studies.