In vitro effects of histamine receptor 1 antagonists on proliferation and histamine release in canine neoplastic mast cells

Abstract Canine mastocytomas (MCTs) are characterized by rapid proliferation of neoplastic mast cells (MCs) and clinical signs caused by MC‐derived mediators. In dogs suffering from MCT, histamine receptor 1 (HR1) antagonists are frequently used to control mediator‐related clinical symptoms. Previous studies have shown that the HR1 antagonists loratadine and terfenadine exert some growth‐inhibitory effects on neoplastic MCs. We examined whether other HR1 antagonists used in clinical practice (desloratadine, rupatadine, cyproheptadine, dimetindene, diphenhydramine) affect proliferation and survival of neoplastic MCs. Furthermore, we analysed whether these HR1 antagonists counteract IgE‐dependent histamine release from a MC line harbouring a functional IgE‐receptor. HR1 antagonists were applied on two canine MC lines, C2 and NI‐1, and on primary MCs obtained from three MCT samples. The HR1 antagonists desloratadine, rupatadine and cyproheptadine were found to be more potent in decreasing proliferation of C2 and NI‐1 cells when compared with dimetindene and diphenhydramine. Similar effects were seen in primary neoplastic MCs, except for diphenhydramine, which exerted more potent growth‐inhibitory effects than the other HR1 antagonists. Drug‐induced growth‐inhibition in C2 and NI‐1 cells was accompanied by apoptosis. Loratadine, desloratadine and rupatadine also suppressed IgE‐dependent histamine release in NI‐1 cells. However, drug concentrations required to elicit substantial effects on growth or histamine release were relatively high (>10 µM). Therefore, it remains unknown whether these drugs or similar, more potent, HR1‐targeting drugs can suppress growth or activation of canine neoplastic MCs in vivo.


Several previous studies have shown that HR1 antagonists exert growth-inhibitory effects on different tumour cells in vitro,
including melanoma cells and myeloid leukaemia cells (Aichberger et al., 2006;Jangi et al., 2006). In addition, it has been described that the HR1 antagonists terfenadine and loratadine inhibit spontaneous growth of human, feline and canine neoplastic MCs (Hadzijusufovic et al., 2010). However, terfenadine has been removed from the market and loratadine inhibits growth and survival of MCs only at higher concentrations.
Therefore, we were interested to test other HR1 antagonists used either in veterinary medicine (diphenhydramine) (Peters & Kovacic, 2009), or in human medicine (desloratadine, rupatadine, dimetindene, cyproheptadine) (Gimenez-Arnau, Izquierdo, & Maurer, 2009;Gunja, Collins, & Graudins, 2004;Horak, Unkauf, Beckers, & Mittermaier, 2011;Siebenhaar, Degener, Zuberbier, Martus, & Maurer, 2009) for their anti-neoplastic effects on canine neoplastic MCs. In addition, loratadine, the prodrug for desloratadine, was used as positive control. In this study we also tested multi-targeted tyrosine kinase inhibitors (TKIs), such as masitinib, toceranib and midostaurin, which are currently used to treat canine or human neoplastic MC disorders. Masitinib and toceranib are approved to treat recurrent or non-resectable high-grade MCT with or without regional lymph node involvement in dogs, whereas midostaurin has recently been approved for treatment of human patients with advanced systemic mastocytosis, including mast cell leukaemia (Gleixner et al., 2013;Gotlib et al., 2016;Hahn et al., 2008;London, 2009;London et al., 2009). We were interested to learn whether combinations of HR1 antagonists and TKIs could lead to cooperative drug effects and thus a reduction in the dose of the individual drug. Indeed, such an approach has also been followed in previous studies where different TKIs were combined with each other and were found to exert additive or synergistic anti-proliferative effects in human or canine neoplastic MCs (Gleixner et al., 2007(Gleixner et al., , 2013. In our studies we used in vitro methods employing two established canine MC lines, C2 and NI-1 (DeVinney & Gold, 1990;Hadzijusufovic et al., 2012) and primary MCs isolated from three MCT samples. We tested drug effects on proliferation and survival of neoplastic MCs as well as on IgE-dependent histamine release from NI-1 cells, known to express a functional IgE-receptor (Hadzijusufovic et al., 2012). C2 cells could not employed in these experiments due to their lack of a functional IgE-receptor (Brazís et al., 2002).
The aim of our study was to define growth-inhibitory and histamine release-suppressing effects of various HR1 antagonists in neoplastic canine MCs and to identify the most effective HR1-targeting drugs that could be further tested in clinical studies.

| Isolation of primary canine neoplastic MCs from mastocytoma specimens
Fresh MCT samples were obtained from three dogs undergoing surgery at the University of Veterinary Medicine Vienna (Vienna, Austria). Detailed characteristics of mastocytoma patients are listed in Table 2. Primary MCs were isolated using collagenase digestion as essentially described (Valent et al., 1989   24-well microtiter plates (5 × 10 5 cells/well) at 37°C for 24 or 48 hr. Apoptosis in drug-exposed cells was confirmed by the "TUNEL-In situ cell death detection kit-fluorescein" kit from Roche (Mannheim, Germany) following the manufacturer's instructions. The fluorescent stain 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) was used to visualize cell nuclei. Images were obtained using Zen imaging software and Axio Imager 1, both from Carl Zeiss (Oberkochen,

| Evaluation of apoptosis in drug exposed cells and confirmation by TUNEL assay
Germany).

| Effects of HR1 antagonists on proliferation of canine MC lines
As assessed by 3 H-thymidine uptake experiments, various HR1 blockers suppressed proliferation of C2 and NI-1 cells as well as of primary MCs. In MC lines, the following rank order of potency was found: desloratadine > rupatadine > cyproheptadine > dimetindene > diphenhydramine (Figure 1a,b). In primary MCs (n = 3), HR1 antagonists reduced proliferation according to following rank order of potency: diphenhydramine > desloratadine > rupatadine > cyproheptadine > dimetindene (Figure 1c). Responses of MCT cells to these drugs in the individual samples (patients, #1-3) are shown in the Supplementary Material ( Figure S2). Diphenhydramine was found to exert more potent growth-inhibitory effects in primary MCs compared with MC lines (Figure 1, Tables 2 and 3). IC 50 ranges are shown in Table 2 (primary MCs) and Table 3 (MC lines). Furthermore, HR1 antagonists were found to be more potent against NI-1 cells compared with C2 cells (Figure 1, Table 3).

F I G U R E 1
Effects of HR1 antagonists on proliferation of canine neoplastic MCs. C2 cells (a), NI-1 cells (b) and three primary MCT cells (c) were incubated in control medium (Co) or in medium containing various concentrations of HR1 antagonists (as indicated) at 37°C for 48 hr. Thereafter, 3 H-thymidine was added for 16 hr and then the uptake of 3 H-thymidine was measured.
Results show 3 H-thymidine uptake in percent of control (=100%, Co) and represent the mean ± SD of at least three independent experiments. Asterisk (*): p < .05 compared with control (Co)

| Effects of HR1 antagonists on survival of canine MC lines
We next examined the effects of HR1 antagonists on survival of C2 cells and NI-1 cells by measuring apoptosis and consecutive necro-  Table 3. Specificity of apoptosis induction in drug-exposed cells was confirmed by TUNEL assay (Figure 3).

| Effects of HR1 antagonists on IgE-mediated histamine release in NI-1 cells
To examine whether HR1 antagonists regulate IgE-dependent histamine release, we performed experiments using NI-1 cells, known to express a functional and cytokine-responsive IgE receptor (Bauer et al., 2017;Hadzijusufovic et al., 2012). Loratadine, desloratadine and rupatadine were found to inhibit anti-IgE-induced histamine release when applied at a concentration of 50 µM (Figure 4).
Desloratadine, a metabolite of loratadine, showed a stronger inhibitory effect on anti-IgE-induced histamine release when compared with loratadine and rupatadine. Cyproheptadine, dimetindene and

F I G U R E 2 Induction of apoptosis by HR1 antagonists in C2 cells and NI-1 cells. C2 cells (a) and NI-1 cells (b)
were incubated in control medium (Co) or in medium containing various concentrations of HR1 antagonists (as indicated) at 37°C for 24 hr (upper panels) or 48 hr (lower panels). Thereafter, the cells were stained using the Hematek® Stain Pak (Modified Wright's Stain) and the numbers of viable, apoptotic and necrotic cells were counted using light microscopy. Results show the percentage (%) of viable (white open bars), apoptotic (black filled bars) and necrotic (grey filled bars) cells relative to the total cell number. Results represent the mean ± SD of at least three independent experiments. Asterisk (*): p < .05 compared with control (Co) diphenhydramine showed no significant inhibitory effects on histamine release from NI-1 cells (Figure 4).
Masitinib and toceranib are considered to be safe drugs and are effective in decreasing tumour progression in some dogs with highgrade or non-resectable MCT (Hahn et al., 2008;London, 2009;London et al., 2009). Midostaurin is well known to suppress the growth of human KIT D816V + MCs in vitro and in vivo (Gleixner et al., 2013;Gotlib et al., 2016). Since in initial experiments we found that the concentrations of HR1 antagonists required to block MCT cell growth are probably above the tolerable concentration in patients, we were interested to identify drug combinations in which the concentrations of the individual drugs could be reduced to a pharmacologically meaningful range. Therefore, we were interested to learn whether combinations of TKIs and HR1 antagonists can induce growth arrest in canine MCT cells. C2 cells and NI-1 cells were exposed to various combinations of HR1 antagonists (loratadine, desloratadine, rupatadine, cyproheptadine, dimetindene and diphenhydramine) and KIT-targeting drugs (masitinib, toceranib, midostaurin). The combination 'loratadine + midostaurin' was found to produce cooperative growth-inhibitory effects ( Figure 5). The other drug combinations tested showed no cooperative effects in C2 cells ( Figure S3) or in NI-1 cells ( Figure S4).

Several TKIs, including midostaurin, have been reported to inhibit
IgE-dependent histamine release from human basophils and MCs (Kneidinger et al., 2008;Krauth et al., 2009). In our study, we first asked whether the TKIs masitinib, toceranib and midostaurin alone are able to block IgE-dependent histamine release in NI-1 cells.
Among the TKIs tested, only midostaurin was able to counteract anti-IgE-induced histamine release in NI-1 cells when applied at a concentration of 10 µM (Figure 6a). Therefore, we combined midostaurin with selected HR antagonists (loratadine and desloratadine) in further IgE-dependent histamine release experiments. Only the combination 'desloratadine 50 µM + midostaurin 10 µM' was found to produce synergistic inhibitory effects on histamine release ( Figure 6b). The combination 'loratadine + midostaurin' showed no inhibitory effects on histamine release in NI-1 cells ( Figure S5).

F I G U R E 3 Confirmation of apoptosis-induction by HR1 antagonists in C2 cells and NI-1 cells using TUNEL assay. C2 cells (a) and NI-1 cells (b)
were cultured in control medium (Co) or in medium containing HR1 antagonist (as indicated) at 37°C. C2 cells were harvested and analysed after 48 hr. NI-1 cells were harvested and analysed after 20 hr. TUNEL-positive (apoptotic) cells are stained in green and TUNEL-negative/DAPI-positive cells are stained in blue F I G U R E 4 Effects of HR1 antagonists on IgE-dependent histamine release in NI-1 cells. NI-1 cells were pre-incubated with 5 µg/ml IgE at 37°C for 2 hr, followed by incubation with several HR1 antagonists (as indicated) at 37°C for 60 min. Thereafter, the histamine release was triggered by adding 5 µg/ml anti-IgE at 37°C for 30 min and afterwards histamine concentrations in the cell-free supernatants were determined. Histamine release was calculated as percentage of total histamine. Then, the calculated histamine percentages in the 'IgE + anti-IgE' condition were set to 100% and serve as positive control. Results show the percentage of positive control and represent the mean ± SD of at least three independent experiments. Asterisk (*): p < .05 compared with the 'IgE + anti-IgE' condition (positive control)

| D ISCUSS I ON
Antihistamines are often applied in human and canine MC neoplasms to control mediator-related clinical symptoms (Arock, Akin, Hermine, & Valent, 2015;Blackwood et al., 2012;Peters & Kovacic, 2009;Sader, Cai, Muller, & Wu, 2017;Thamm & Vail, 2007;Welle et al., 2008;Willmann et al., 2018). However, only few reports on the effects of HR1 antagonists on the proliferation of neoplastic cells are available (Aichberger et al., 2006;Hadzijusufovic et al., 2010;Jangi et al., 2006). We have previously shown that the HR1 antagonists terfenadine and loratadine can suppress the in vitro growth of neoplastic MCs (Hadzijusufovic et al., 2010). Due to the removal of terfenadine from the market and the relatively weak anti-neoplastic effects of loratadine (Hadzijusufovic et al., 2010), we were interested in analysing additional HR1 antagonists for their potential anti-neoplastic effects.
We found that several HR1 antagonists produce growth-inhibitory and anti-survival effects in canine neoplastic MCs. In addition, we found that HR1 antagonists loratadine, desloratadine and rupatadine, and the multi-targeted protein kinase inhibitor midostaurin suppress the IgE-dependent histamine release from canine neoplastic MCs.

F I G U R E 5
Effects of drug-combinations on proliferation in canine neoplastic MCs. C2 cells (a) and NI-1 cells (b) were incubated in control medium (Co) or in medium containing various concentrations of loratadine (blue line), midostaurin (red line) or combinations of these two drugs (black line) at 37°C for 48 hr. Thereafter, 3 H-thymidine was added for 16 hr and then the uptake of 3 H-thymidine was measured. Results show the 3 H-thymidine uptake in percent of control (=100%, Co) and represent the mean ± SD of at least three independent experiments. Asterisk (*): p < .05 compared with control (Co) F I G U R E 6 Effects of TKIs and TKI/HR1 antagonistcombinations on IgE-dependent histamine release in NI-1 cells. NI-1 cells were pre-incubated with 5 µg/ml IgE at 37°C for 2 hr, followed by an incubation with control medium or medium containing TKIs (depicted in a) or an incubation with control medium, medium containing various concentrations of desloratadine, midostaurin or drug combinations composed of 'desloratadine + midostaurin' (depicted in b) at 37°C for 60 min. Thereafter, the histamine release was triggered by adding 5 µg/ml anti-IgE at 37°C for 30 min and histamine concentrations in the cell-free supernatants were determined. Histamine release was calculated as percentage of total histamine. Then, the calculated histamine percentages in the 'IgE + anti-IgE' condition were set to 100% and serve as positive control. Results show the percentage of positive control and represent the mean ± SD of at least three independent experiments. Asterisk (*): p < .05 compared with the 'IgE + anti-IgE' condition (positive control). The combination index (CI) values for the drug-combinations 'desloratadine 25 µM + midostaurin 5 µM' and 'desloratadine 50 µM + midostaurin 10 µM' are depicted in b (CI < 1) In our 3 H-thymidine uptake experiments, several HR1 antagonists showed dose-dependent growth-inhibitory effects in canine MC lines and primary MCs. Diphenhydramine was found to exert more potent growth-inhibitory effects in primary MCs compared with MC lines. We also observed that HR1 antagonists have slightly stronger effects in NI-1 cells compared with C2 cells. This is consistent with previously reported data, showing NI-1 cells to be more sensitive to various TKIs than C2 cells (Hadzijusufovic et al., 2012). The reason for this phenomenon remains unknown.
A possible explanation for higher sensitivity of NI-1 cells to anti-proliferative drugs may be their faster proliferation when compared with C2 cells (Hadzijusufovic et al., 2012). Both cell lines, NI-1, which were derived from a MC leukaemia and C2, which Previously, we have shown that terfenadine and loratadine not only inhibit proliferation, but also reduce survival of neoplastic MCs (Hadzijusufovic et al., 2010). Using morphological assessment and TUNEL assay, we found that desloratadine, rupatadine and cyproheptadine induced significant dose-and time-dependent apoptosis in both C2 cells and NI-cells. These data suggest that the HR1 antagonists tested do not exert toxic effects, but rather initiate apoptosis in canine neoplastic MC lines. In contrast to our experiments on cell proliferation, higher concentrations were necessary to achieve a comparable effect on apoptosis (e.g. incubation with 50 µM of desloratadine results in a total inhibition of proliferation in C2 cells, whereas in our apoptosis-induction experiments approximately 50% of cells remain viable when the same concentration is applied, at least considering morphological criteria). This is consistent with studies on human melanoma and hepatocellular carcinoma, in which diphenhydramine and cyproheptadine (applied in a range of 40-100 µM) induced apoptosis, but no toxic necrosis (Feng et al., 2015;Or et al., 2016).
Our data suggest that pharmacologically active concentrations needed for inhibiting MC proliferation and survival are higher than can be achieved in vivo in mastocytoma patients, even if applied in up to four-time higher doses, which are well tolerated by humans (Gimenez-Arnau et al., 2009;Gupta et al., 2002;Powell et al., 2007;Siebenhaar et al., 2009;Zuberbier et al., 1996).
Therefore, further efforts to identify more effective and less toxic drug derivatives are needed. However, it seems that the observed effects are at least in part mediated through HR1-binding, as several chemically diverse HR1 antagonists showed anti-neoplastic effects, whereas none of the HR2 antagonists applied showed effects on neoplastic cells in previous studies (Aichberger et al., 2006;Hadzijusufovic et al., 2010). Nevertheless, as not all HR1 antagonists exhibit the same anti-proliferative and pro-apoptotic effects in C2 and NI-1 cells, it remains unclear if the effects observed in this study, are indeed all mediated via HR1. Potential other targets of HR1 antagonists might be intracellular histamine receptors, such as cytochrome P450 isoenzymes (Dy & Schneider, 2004). Certain cytochrome P450 isoenzymes have been suggested to play a role as targets of loratadine and terfenadine by mediating growth-inhibitory effects in chronic myeloid leukaemia cells (Aichberger et al., 2006).
Previous studies have also shown that histamine and histamine-metabolizing enzymes, both known to interact with cytochrome P450 isoenzymes, regulate the growth and survival of various tumour cells (Darvas et al., 2003;Falus et al., 2001;Malaviya & Uckun, 2000;Radvány et al., 2000;Rivera et al., 2000). Some of these interactions may be associated with autocrine or intercrine growth regulation in neoplastic cells (Darvas et al., 2003;Falus et al., 2001;Malaviya & Uckun, 2000;Radvány et al., 2000;Rivera et al., 2000). Since MCs produce and contain substantial amounts of histamine it also might be that certain HR1 antagonists are able to disrupt autocrine or in- to proliferation experiments where these drugs were applied for 24 or 48 hr (Hadzijusufovic et al., 2012). In these experiments, midostaurin was found to counteract anti-IgE-induced histamine release at relatively high concentrations, exceeding those that are effective in human MCs (Krauth et al., 2009). Masitinib and toceranib showed no effects on IgE-mediated histamine release.
We also combined midostaurin with loratadine and its active metabolite desloratadine. However, only the combination 'desloratadine + midostaurin' produced synergistic inhibitory effects on IgE-dependent histamine release in NI-1 cells. Since midostaurin not only blocks mediator release but also inhibits proliferation of canine neoplastic MCs (Gotlib et al., 2016;Hadzijusufovic et al., 2012), this agent may be a promising drug for the treatment of canine mastocytoma.
In summary, our data show that the HR1 antagonists desloratadine, rupatadine and cyproheptadine reduce proliferation and survival of canine MC lines more effectively than dimetindene and diphenhydramine. In contrast, diphenhydramine was found to exert more potent growth-inhibitory effects in primary neoplastic MCs when compared to the other HR1 antagonists tested. Why the HR1 antagonists exert different biologic activities in MC lines and in primary MCs remains unknown. Moreover, we found that loratadine, desloratadine and rupatadine inhibit IgE-dependent histamine release in NI-1 cells. However, concentrations needed for these effects are higher than those usually applied in patients. Whether more potent derivatives combining anti-mediator with growth-inhibitory effects can be developed, remains to be elucidated.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest. Conflict of interest unrelated to this study: PV received honoraria from Novartis, Blueprint, Deciphera and Incyte.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/vms3.336.