Histamine can be Formed and Degraded in the Human and Mouse Heart

Histamine is metabolized by several enzymes in vitro and in vivo. The relevance of this metabolism in the mammalian heart in vivo is unclear. However, histamine can exert positive inotropic effects (PIE) and positive chronotropic effects (PCE) in humans via H2-histamine receptors. In transgenic mice (H2-TG) that overexpress the human H2 receptor in cardiomyocytes but not in wild-type littermate mice (WT), histamine induced PIE and PCE in isolated left or right atrial preparations. These H2-TG were used to investigate the putative relevance of histamine degrading enzymes in the mammalian heart. Histidine, the precursor of histamine, increased force of contraction (FOC) in human atrial preparations. Moreover, histamine increased the phosphorylation state of phospholamban in human atrium. Here, we could detect histidine decarboxylase (HDC) and histamine itself in cardiomyocytes of mouse hearts. Moreover, our data indicate that histamine is subject to degradation in the mammalian heart. Inhibition of the histamine metabolizing enzymes diamine oxidase (DAO) and monoamine oxidase (MAO) shifted the concentration response curves for the PIE in H2-TG atria to the left. Moreover, activity of histamine metabolizing enzymes was present in mouse cardiac samples as well as in human atrial samples. Thus, drugs used for other indication (e.g. antidepressants) can alter histamine levels in the heart. Our results deepen our understanding of the physiological role of histamine in the mouse and human heart. Our findings might be clinically relevant because we show enzyme targets for drugs to modify the beating rate and force of the human heart.


INTRODUCTION
Histamine was synthesized in a study to compare several monoamines by researchers in organic chemistry (Windaus and Vogt, 1907). Ackermann (at the same University of Freiburg, Germany) first described the conversion of histidine to histamine by bacteria (Ackermann, 1910), suggesting that histamine could be formed in vivo from histidine and thus could be of physiological relevance. It is now known that histamine is synthesized from histidine, for instance, in the human gut by bacteria (microbiome) but also in many organs of the human body (e.g., enterochromaffine-like cells, neurons or mast cells) and that histamine produced in such a way in the gut or ingested with food can enter the body. This is clinically relevant in some patients because they react e.g. with increased more rapid heartbeat (tachycardia) and fall in blood pressure (hypotension) to this orally applied histamine (Maintz et al., 2006;Haas et al., 2008;Schwelberger, 2010;Panula et al., 2015). The group of Sir Henry Hallet Dale in Oxford (Dale and Laidlaw, 1910) was apparently the first to publish on the positive inotropic effect (PIE, increase in force) and positive chronotropic effect (PCE, increase in heart beat) of histamine in the mammalian heart (e.g., rabbit) (Dale and Laidlaw, 1910). They also noted in rabbits a vasodilatory blood pressure reducing effect of histamine, speculated it might be involved in anaphylaxis and isolated histamine from the intestine of the rabbit, thus proving that it must be endogenously formed and were the first to show that infused histamine in animals is nearly completely metabolized (Dale and Laidlaw, 1910: isolated perfused liver). Human studies were initiated by a German gynecologist, who reported signs of allergic shock after injection of histamine including tachycardia and hypotension (Jäger, 1913a, Jäger, 1913b. Arrhythmia after injection of histamine in patients was already reported early on (Schenk, 1921) and confirmed ever after (in vitro, human atrium: Sanders et al., 1996). The direct cardiac effects of histamine are currently explained by stimulation of specific cardiac histamine receptors. Histamine is thought to act via only four receptors known as H 1 , H 2 , H 3 and H 4 histamine receptors (Jutel et al., 2009;Seifert et al., 2013;Panula et al., 2015). Regional differences in the actions of histamine or histamine receptor utilization by exogenous histamine in the mammalian heart are known: in rabbit left atrium, H 1 receptors are more prevalent but the PIE of histamine is mediated by H 2 receptors while in guinea pig left atrium a positive inotropic effect of histamine is mediated via H 1 receptors (Hattori et al., 1988;Sakuma et al., 1988;Hattori et al., 1990;Hattori et al., 1991). In humans, H 1 and H 2 receptors were detected early on in atrium and ventricle (radioligand binding: Baumann et al., 1982;Baumann et al., 1983;Baumann et al., 1984, antibody andRNA expression: Matsuda et al., 2004). Moreover, the human cardiac H 2 receptors were described to mediate the FIGURE 1 | A scheme of the putative signal transduction of the H 2 -receptor in the mammalian heart. Ca 2+ enters the mammalian heart cell via the L-type Ca 2+ channel (LTCC). This process can be enhanced by histamine and blocked by cimetidine via a cascade starting with the H 2 -receptor. Binding of histamine elevates the activity of adenylyl cyclase (AC) in the sarcolemma via stimulatory G-proteins (Gs), augmenting thereafter the cellular cAMP and thereby stimulates the activity of cAMPdependent protein kinase (PKA). PKA increases cardiac force generation and relaxation by increasing the phosphorylation state (P) of LTCC, phospholamban (PLB), and other regulatory proteins. Trigger Ca 2+ initiates release of Ca 2+ from the sarcoplasmic reticulum via ryanodine receptors (RYR) into the cytosol, where Ca 2+ activates myofilaments and leads to increased inotropy. In diastole, Ca 2+ is taken up into the sarcoplasmic reticulum (SR) via a sarcoplasmic reticulum Ca 2+ ATPase (SERCA) whose activity is higher when the phosphorylation state of PLB is elevated by PKA. Ca 2+ binds to calsequestrin (CSQ) in the SR. Moreover, the scheme depicts putative synthesis and degradation of histamine. More details are presented in Figure 2. Further, the putative storage and release of histamine is indicated.
PIE of exogenously applied histamine in isolated human cardiac preparations (atrium: Ginsburg et al., 1980;Levi et al., 1981;Zerkowski et al., 1993;Sanders et al., 1996). These PIE in the human heart were accompanied by and hence probably mediated by an increase in cAMP (human right atrial preparations: Sanders et al., 1996) and opening of L-type Ca 2+ channels (Eckel et al., 1982, compare Figure 1). However, phospholamban phosphorylation in the human heart in the presence of histamine was previously predicted (Sanders et al., 1996) but is now reported in the present work. Hence, the mode of action of H 2 -receptors mimics the ß-adrenoceptor system in the human heart.
Histamine is usually degraded by diamine oxidase (DAO) in the gut but higher doses of histamine can escape this pathway and can be ingested from food and thereafter histamine is delivered via blood cells to e.g. the heart (Smith and Garrett, 2005). Although peroral histamine intake is usually well tolerated, it has long been known to sometimes lead to tachycardia (humans, 500 mg; Weiss et al., 1932). A controversy exists whether histamine can also be synthesized and degraded in the heart and more specifically in cardiomyocytes, which will be addressed in the present report. We would make also the point here that histamine acts in many ways similar to serotonin (5-HT) in the heart. We reported previously that 5-HT can be formed and degraded in mouse and more importantly human hearts (Pönicke et al., 2012;Gergs et al., 2017b). Like histamine, 5-HT has no contractile effect in wild type (WT) mouse heart, but only in transgenic 5-HT 4 -overexpressing mice (5-HT 4 -TG: Gergs et al., 2013). We used this 5-HT 4 -TG to selectively monitor the contractile effect as a read out for 5-HT formation (Gergs et al., 2010;Buchwalow et al., 2013;Gergs et al., 2013). We histologically described the synthesizing and degrading enzymes in cardiomyocytes and used their inhibitors to raise or lower cardiac 5-HT levels (Gergs et al., 2017b). The same approach was used in the current communication for histamine metabolism in the heart. It was known before the start of our project that histidine via transporters can enter cells, that histidine is converted solely by histidine decarboxylase (HDC) to histamine (Figures 1, 2, for review: Haas et al., 2008;Panula et al., 2015). Moreover, HDC can be induced also in muscle cells, e. g. by stimulation with cytokines or hormones, leading to histamine synthesis followed by immediately release (Fogel, 2015). A HDC-knockout (HDC-KO) mouse has been generated and widely studied for that purpose (Ohtsu et al., 2001). Histamine can undergo two major pathways for degradation: N-methylation and oxidation via several chemical compounds until the degradation products leave the human body via the urine (Figure 2). For each of the enzymes involved, more or less specific small molecule inhibitors are available that were tested in vitro and in vivo to experimentally test these concepts ( Figure 2) but also have been tried in patients for therapeutic purposes. Histamine has been detected in human hearts (ventricle: 5 μg/g: Patella et al., 1995) and was in the same order of magnitude as noradrenaline (Petch and Nayäler, 1979;Wolff and Levi, 1986), and much less in mouse hearts (0.29 μg/g: Anton and Sayre, 1969, 1-4 pmol/mg;Zimmermann et al., 2011).
The present study was started to determine whether the inotropic and chronotropic effects of histamine can be due to histamine formed and metabolized in the mouse and human heart. Preliminary reports of this project have been published in abstract form (Grobe et al., 2015;Meister et al., 2015;Grobe et al., 2016).

Transgenic Mice
Transgenic mice (H 2 -TG) with cardiac myocyte-specific overexpression of the human H 2 -receptor and their littermate control mice (WT) were generated as described by Gergs et al. (2019a). Heart-specific expression was achieved via the α-myosin heavy-chain promoter. The age of the animals ranged from three to five months. Contraction experiments were performed on left and right atrial preparations as previously described by Gergs et al. (2013). All mice were housed under conditions of optimum light, temperature, and humidity with food and water provided ad libitum. The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the National Research Council (2011). Animals were handled and maintained according to approved protocols of the animal welfare committee of the University of Halle-Wittenberg, Halle, Germany (approval reference number 42502-02-691 MLU).

Clinical Patient Data
Right atrium samples were obtained from patients undergoing open-heart surgery with coronary artery bypass grafts, and electrically stimulated in organ baths as described previously (Gergs et al., 2013;Gergs et al., 2018;Boknik et al., 2019). Patients were treated with the following classes of drugs: ACEinhibitors, β-adrenoceptor-blockers, diuretics, antilipemic agents, cholinergic agents, antithyroid agents, hypoglycemic agents, gout suppressants, antibiotics, α-adrenoceptor-blockers, parasympatholytics, calcium channel blockers, proton-pump inhibitors, and analgesics. Patients were in CCS scale from III to IV and NYHA class II-III. Left ventricular ejection fraction ranged from 40 to 55%. This study complied with the Declaration of Helsinki and was approved by the local ethics committee (hmbü 04.08.2005). All patients gave informed consent.

Contractile Studies in Mice and Men
In brief, left atrial preparations from mouse or right atrial preparations from man were isolated and mounted in organ baths as described by Neumann et al. (2003); Gergs et al. (2013); Gergs et al. (2017a);Gergs et al. (2019b). The organ baths' bathing solution contained 119.8 mM NaCl 5.4 mM KCl, 1.8 mM CaCl 2 , 1.05 mM MgCl 2 , 0.42 mM NaH 2 PO 4 , 22.6 mM NaHCO 3 , 0.05 mM Na 2 EDTA, 0.28 mM ascorbic acid, and 5.05 mM glucose. It was continuously gassed with 95% O 2 and 5% CO 2 and maintained at 37°C and pH 7.4, as described by Neumann et al. (2003); Kirchhefer et al. (2004). Preparations were attached to a bipolar stimulating electrode and suspended individually in 10-ml glass tissue chambers for recording isometric contractions. Force of contraction (FOC) was measured with inductive force transducers connected to a PowerLab system (ADInstruments, Oxford, United Kingdom). Time parameters and first derivate of force of contraction of single contractions were evaluated from digitized recordings. Each muscle was stretched to the length of maximal FOC. The left atrial preparations from mice were electrically stimulated at 1 Hz with rectangular pulses of 5 m duration; the voltage was ∼10-20% greater than the threshold of initiation of contraction.

Real Time PCR
Total RNA isolation from mouse ventricular tissue and isolated cardiac myocytes followed by reverse transcription and real time PCR were performed as described previously (Gergs et al., 2019c). Relative gene expression was calculated according to the 2 −ΔΔCt method (Livak and Schmittgen, 2001) with the GAPDH signal for normalization. Following primer sequences were used: HDC, forward GGATGATGGTACACTTTGACTG, reverse TGTCTG ACATGTGCTTGAAGAT (Wiener et al., 2003); GAPDH, forward ATGCATCCTGCACCACCAAC, reverse, ATGCCT GCTTCACCACCTTC (Gergs et al., 2019c).

Histochemical and Immunohistochemical Staining of Mouse Samples
Tissue samples of mouse hearts (or human atrial samples for MAO-A or MAO-B detection) were fixed in buffered 4% formaldehyde and routinely embedded in paraffin. PBS was used for all washings and dilutions. 4-µm-thick paraffin tissue sections were deparaffinized with xylene and graded ethanols. For pathohistological analysis, tissue sections were routinely stained with hematoxylin-eosin (HE) and Masson-Goldner-Trichrome staining (MG). For immunohistochemical assays, deparaffinaized sections were subjected to antigen retrieval by heating in 10 mM sodium citrate buffer, pH 6.0, at 95°C for 30 min.
For histamine detection, tissue samples were fixed overnight in 4% 1-ethyl-3 (3-dimethylaminopropyl)-carbodiimide (EDAC) in PBS (pH 7.4), post-fixed for 30 min in 4% formaldehyde in PBS, cryoprotected with 20% sucrose in PBS, snap-frozen in liquid nitrogen and cryosectioned at 4 μm thickness. Sections were mounted on polysine microslides (Menzel Glaeser, Braunschweig, Germany), thoroughly air-dried and stored at −20°C until use. The sections were first washed in H 2 O for 5 min and then in PBS for 10 min.
Visualization and image processing. Immunostained sections were examined on a Zeiss microscope "Axio Imager Z1." Microscopy images were captured using AxioCam digital microscope camera and AxioVision image processing (Carl Zeiss Vision, Germany). The images were acquired at 96 DPI and submitted with the final revision of the manuscript at 300 DPI. Images shown are representative of 3 independent experiments which gave similar results.

Enzymes Activity Measurement
One part of the frozen tissue samples of the mouse and human cardia was analyzed with the most sensitive isotopic assays using 14 C labeled substrates: 14 C-serotonin for MAO A, 14 C-betaphenylethylamine for MAO-B, 14 C-putrescine for DAO and 14 C-methyl S-adenosyl-L-methionine for HMT as described elsewhere (Taylor and Snyder, 1972;Fowler and Tipton, 1982;Fogel et al., 1985). In more details, outer mitochondrial membrane bound MAO-A and MAO-B activities were estimated in the tissue homogenates using serotonin (final concentration 200 μM) and 14 C-beta-phenylethylamine (final concentration 20 μM), as well as specific inhibitors, i.e. clorgyline and deprenyl (0.3 μM each), respectively (Fowler and Tipton, 1982). Diamine oxidase, a soluble enzyme, was assayed in supernatants with 14 C-putrescine (300 μM) as substrate. Products of the oxidative deamination were separated by column chromatography on Dowex 50WX8 and measured by liquid scintillation counting (Fogel et al., 1985). Histamine N-methyltransferase activity (HMT) likewise was determined in the tissue supernatants using histamine as a substrate and adenosyl-L-methionine, S-[methyl-14 C] (Perkin-Elmer) as a donor of methyl group (100 μM each) with aminoguanidine and pargyline serving as inhibitors of DAO and MAO enzymes, after Taylor and Snyder (1972). The enzyme activities were expressed as pmol/min/mg protein.
Protein concentrations were analyzed according to Lowry's method (1951).

Mass Spectrometry
The amount of histamine in cardiac preparations (heart tissue, ventricular myocytes and plasma/serum) was quantified as follows: The tissue was weighed and subsequently homogenized using a FastPrep 24 device (MP Biomedicals). Therefore tissue was transferred into 2 ml screw caps filled with Lysing Matrix A (MP Biomedicals). 800 µl ethanol/water (70/30; v/v) was added and the samples were homogenized two times for 30 s with 5 m/s. After centrifugation 600 µl of supernatant was transferred to a reaction tube and evaporated to dryness under a gentle nitrogen stream at 40°C. The residual pellet was reconstituted in 50 µl 80/ 20 acetonitrile/water (v/v) containing 0.25 µM of d4-histamine (C/D/N isotopes, Pointe-Caire, Canada) as internal standard. This solution was transferred to an analysis vial for masspectrometric analysis. Cardiac myocyte suspension was treated with a 4 fold amount of 50/50 acetonitrile/water (v/v). After centrifugation supernatant was transferred to a reaction tube and prepared for analysis as described above. For analyte extraction of murine blood or serum, 50 µl sample was treated with 200 µl of 50/50 acetonitrile/water (v/v). After centrifugation supernatant was transferred to a reaction tube and prepared for analysis as described above. The calibrators (ranging from 0.0079 to 4 pmol/sample) were prepared in 50 mg/ml BSA and aliquoted to 50 µL. The calibrators were prepared in the same way as blood/ serum samples. Histamine was quantified using a liquid chromatography system coupled to a mass spectrometer (API2000; Sciex, Foster City, CA) as described previously (Zimmermann et al., 2011).

Data Analysis
Data shown are means ± SEM. Statistical significance was estimated by analysis of variance followed by Bonferroni's t-test or by student's t-test where appropriate. A p-value of less than 0.05 was considered significant. Experimental data for agonist-induced positive inotropic and chronotropic effects were analyzed by fitting sigmoidal curves to the experimental data with GraphPad Prism 5.0. All other statistical analyses were performed as indicated in the figures and tables. Statistical evaluation was conducted with GraphPad Prism 5.0 (GraphPad Software, San Diego, California, United States).

Drugs and Materials
Histidine, histamine, cimetidine, quinacrine, tranylcypromine, aminoguanidine, and compound 48/80 were purchased from Sigma-Aldrich (Taufkirchen, Germany). All other chemicals were of the highest purity grade commercially available. Deionized water was used throughout the experiments. Stock solutions were freshly prepared daily.

Effects of Histidine and Histamine
As seen in the original recording ( Figure 3A), histidine exerted a PIE in isolated electrically stimulated (1 Hz) human right atrial preparations: 10 mM histidine increased force of contraction by 27% from 6.02 ± 0.63 mN to 7.66 ± 0.43 to mN (p < 0.05, n 3).
In the presence of NSD-1055 (100 µM), a decarboxylase inhibitor, histidine failed to exert a PIE in human atrial preparations (data not shown). For comparison, the effects of histidine (cumulatively applied) and histamine (10 µM) on a H 2 -TG mouse left atrium are shown in Supplementary Figure S1. In mouse atrial preparations, histidine was less effective than in human atrium. Similarly, histamine exerted a concentration-dependent PIE in human right atrial preparations (original recording, Figure 3B) when histamine was cumulatively applied. In a different set of patient samples, a PIE to single additions of histamine (1, 10, or 100 µM) was recorded which could be blocked by 10 µM cimetidine ( Figure 3C). Histamine was more potent than histidine to elicit a PIE. Moreover, as a new finding, histamine in the set of patient samples where it was not cumulatively applied increased the phosphorylation state of PLB at serine 16 ( Figure 3D), which was likewise blocked by cimetidine. This PIE of histamine at human atrium was also accompanied by shortened relaxation time and elevated maximum rate of relaxation, consistent with the known functional consequences of phospholamban phosphorylation ( Table 1).
In mouse atrial preparations, the histamine releasing agent compound 48/80 exerted a PIE on left atria from H 2 -TG and this effect was attenuated by cimetidine ( Figure 4). Initial experiments detected a PIE of compound 48/80 also in WT atria, which was absent in the additional presence of 50 µM propranolol. Therefore, the experiments for compound 48/80 in H 2 -TG and WT all contained 50 µM propranolol. Under these conditions compound 48/80 was unable to exert a PIE in WT atria (data not shown).

Effects of Enzyme Inhibitors: Aminoguanidine, Quinacrine and Tranylcypromine
Aminoguanidine (AG) alone (1 mM), an inhibitor of diamine oxidase ( Figure 5A), reduced FOC in H 2 -TG and shifted the concentration response curve of histamine on FOC to the left in H 2 -TG but was ineffective in WT ( Figure 5A). More specifically, aminoguanidine (1 mM) shifted the log concentration response curves for the PIE in H 2 -TG left atrium to the left from EC 50values ( half maximum effective drug concentrations) of 110 nM to 37 nM (n 4-6; p < 0.05).  Quinacrine (QAN), an inhibitor of histamine-Nmethyltransferase (HMT) that inactivates histamine ( Figure 5B), did not shift the concentration response curve of exogenously administered histamine on FOC in H 2 -TG atrium and was ineffective in WT ( Figure 5B). Now, one can ask what would happen if one inhibited both metabolizing pathways: diamine oxidase and histamine-N-methyltransferase: aminoguanidine combined with quinacrine ( Figure 5C) increased per se FOC in H 2 -TG, suggesting an action of endogenous histamine and shifted the concentration response curve of histamine on FOC ( Figure 5C) to the left in H 2 -TG but was ineffective in WT (data not shown).
Similar effects of aminoguanidine and quinacrine were noted on the beating rate of right atrial preparations from H 2 -TG. Regrettably, only a few right atria per group remained in sinus rhythm till the end of the histamine concentration response curves in the additional presence of inhibitors of metabolism (Supplementary Figure S2). Therefore, it was difficult to properly analyze the effects of the inhibitors on the beating rate, that is to say that the noted differences lack statistically significance (Supplementary Figure S2). In WT right atrial preparations, aminoguanidine, quinacrine and of course histamine were ineffective (Supplementary Figure S2).
Furthermore, the inhibitor of both monoamine oxidase A and B (MAO-A and MAO-B), tranylcypromine (10 µM), shifted the PIE of histamine in H 2 -TG atrium from EC 50 -values of 70 nM to 38 nM and increased the efficacy of histamine for the PIE in H 2 -TG atrium from a maximum response of 9.3 ± 0.8 mN to 12.4 ± 0.2 mN (n 3-4; p < 0.05, Figure 5D). Experiments were performed in the presence of 50 µM of the ß-adrenoceptor antagonist propranolol because we noted in initial experiments a profound PIE of tranylcypromine alone (in H 2 -TG and WT). This PIE was blocked by previously applied propranolol suggesting that the PIE of tranylcypromine (in the absence of propranolol) resulted from elevated noradrenaline near the β-adrenoceptor in the sarcolemma, due to inhibition of noradrenaline uptake into cells.

Western Blotting and Histology
Given the effects of the various inhibitors described above, we asked if the different enzymes presumably inhibited by these compounds are present in mouse and human heart tissue. We could detect the mRNA of HDC in mouse ventricular samples and in isolated mouse cardiac myocytes ( Figure 6A) and the HDC protein in its active 53 kDa form in isolated mouse cardiac myocytes ( Figure 6B). Here, a mouse stomach sample was used as positive control and a cardiac sample of a HDC knockout mouse as negative control ( Figure 6B). By comparison with the knockout sample, HDC was detected as monomer and dimer as indicated. The detection of cardiac calsequestrin (CSQ) was used as loading control especially for cardiac samples ( Figure 6B). Moreover, we could detect MAO-A, MAO-B, DAO (extremely low amounts) and HMT on Western blots of mouse tissue homogenates (exemplary blots are shown in Figure 6C). In addition, we were able to detect and localize the HMT, MAO- A, MAO-B and, to considerably lesser extent, DAO in mouse cardiac tissue, specifically in cardiomyocytes (Figure 7), using immunohistology. Likewise, we could detect HMT and trace amounts of DAO in human atrial tissue by Western blotting ( Figure 8A), and HMT, MAO-A and MAO-B but not DAO in human atrium by immunohistological staining (Figures 8B,C). Classically, the mast cells are a possible source of histamine. Identification of mast cells in the myocardium by Giemsa stain (Figures 9A,B) as well as by an enzymohistochemical reaction for visualization of specific ASD-chloracetate-esterase activity ( Figures 9C,D), permitted to reveal only extremely rare mast cells exclusively in the close vicinity of blood vessels both in WT- (Figures 9A,C) and in H 2 -TG-samples ( Figures 9B,D). Control stainings for mast cells (positive controls) were performed on samples of a human mastocytosis case and a gastrointestinal stromal tumor case (data not shown). Interestingly, in all WTand H 2 -TG-samples from mice, positive immunostaining was observed for histidine decarboxylase (Figure 10) in atria. Histidine decarboxylase was immunostained in cardiomyocytes as well as in intima and media of blood vessels (Figures 10C-F). The histamine produced can be made visible histologically in the mouse heart. In all WT-and H 2 -TG-samples, positive immunostaining for histamine was observed. Of note, histamine was immunostained also in cardiomyocytes ( Figures  10G,H).
Moreover, by mass spectrometry we could measure histamine in mouse ventricular myocytes at a concentration range of 0.0128-0.0190 pmol/mg protein. In mouse hearts, freshly taken out from mice (where blood was not washed out) the histamine content amounted to 0.177 pmol/mg heart tissue and was much lower after Langendorff perfusion with Tyrode's solution namely 0.038 pmol/mg heart tissue. In blood of mice, we typically measure histamine concentrations about 39 pmol/ml and in serum about 16.5 pmol/ml.

DISCUSSION
Histidine is taken up in the gut from food by a transporter. At least in the central nervous system it enters the nerve cells, is converted to histamine by HDC, and this histamine is stored and released by the nerve cells where it acts on histamine receptors in an autocrine or paracrine fashion (Review: Haas et al., 2008;Panula et al., 2015). A similar mechanism was found in a linage of hematopoietic stem cells, which produce histamine by HDC and where histamine acts in an auto/paracrine fashion via H 2 receptors to protect them from myelotoxic injury (Chen et al., 2017a). We assume that also in cardiomyocytes this transporter passes histidine trough the sarcolemma and, in the cytosol, it can be a substrate for HDC. Possibly, this transporter is differentially expressed in human and mouse cardiac tissues or HDC expression itself is different. Moreover, one could argue that only little histidine (a polar molecule) might pass via diffusion in the cell and the transport of histidine might be too weak to transport substantial histidine amounts into the cell. Further, a loss of the histamine formed in the cell may occur because it may be rapidly degraded to inotropically inactive metabolites or the outward transport systems of histamine (which exist) might be insufficient to transport sufficiently high concentrations of histamine out of the cell where is can bind to the ligand binding site on the outer surface of the sarcolemma. That all would be an explanation for the small effect of histidine compared to exogenously applied histamine and for the smaller effect of histidine in mouse left atrium compared to human right atrium. But this is currently speculative. Much larger concentrations of exogenously applied histidine are required to achieve a PIE than for histamine. That is plausible as first HDC has to form histamine from histidine and the formed histamine has to exit the sarcolemma, as the ligand binding site of histamine receptors is located on the outer surface of the sarcolemma. These findings are reminiscent of our previous work where we noted for another monoamine, 5-HT, that it can be formed by decarboxylation from its direct precursor (5-hydroxytryptophane) in the heart and can exert a PIE in the isolated human atrium (Gergs et al., 2017b). HDC is certainly the main histamine producing enzyme at least in mice, because HDC-KO mice have very low levels of histamine in tissue and plasma (Ohtsu et al., 2001; review:; Jutel  (Jutel et al., 2009). HDC is classically found in mast cells, basophils and enterochromaffin-like cells where it is constitutively expressed (Jutel et al., 2009;Fogel, 2015). During inflammation, initiated by LPS, HDC expression seems to be induced in the lung and stomach but in this experimental model, the HDC promoter was hardly active in the heart. These expression patterns were identified by green fluorescence protein expression under control of the HDC promoter in transgenic mice (Takai et al., 2019). In the human heart, HDC was imunohistologically detected in half of the ganglions of the right atrium (Singh et al., 1999) and in isolated rat cardiomyocytes, HDC was detected by Western blotting (Costiniti et al., 2018). In the present paper, we present the detection of HDC in mouse cardiomyocytes by immunohistology, Western blotting und real time RT-PCR. Cardiac HDC expression in mice may be strain-dependent but, furthermore, an important role for histamine in the heart was demonstrated by a mouse model of myocardial infarction using HDC-KO mice. In these mice, histamine deficiency worsened cardiac dysfunction (Deng et al., 2015;Chen et al., 2017b). But by use of HDC-EGFP transgenic mice, it was shown that CD11b and Gr-1 positive immature myeloid cells seem to be the predominant HDC-expressing sites in acute myocardial infarction (Deng et al., 2015). This source of HDC, more precisely, HDC-expressing macrophages located in the mouse heart, may also apply in our model, but we investigated the basal expression pattern of HDC. Nevertheless, this is a limitation of our study. Therefore, it would be interesting to investigate by use of HDC-EGFP transgenic mice if EGFP expression can be found in cardiomyocytes under basal conditions. Moreover, it would be interesting to study whether any HDC activity is present in isolated cardiomyocytes. Unfortunately, these experiments are beyond the scope of our current study. Inactive mutants of HDC have been reported and linked to neurological diseases like Tourette`s syndrome (Jones and Kearns, 2011). These patients should have less histamine in the heart but apparently this has not been reported so far. HDC was detected by Western blotting in rabbit atrium (Matsuda et al., 2002). HDC is also expressed by some bacteria in the gut (Chen et al., 2019) but also some  fermented food (cheese, red wine) and fish may contain substantial levels of histamine and can lead to adverse reactions (Maintz et al., 2006). Once histamine activates the H 2 -receptor, a cascade leading to phospholamban (PLB) phosphorylation is initiated (Figure 1), which we could detect in human atrial preparations and which can explain at least in part the inotropic as well as the lusitropic effects of histamine in the human heart in agreement with our findings in H 2 -TG atrium (Gergs et al., 2019a). Moreover, histamine can regulate the beating rate via H 2 -receptors of the vertebrate heart (reptile) but the origin of the histamine was not evaluated (Skovgaard et al., 2009). We present tentative evidence that histamine can be released from the H 2 -TG atrium, as compound 48/80 exerts a cimetidine sensitive effect (Figure 4). It is well known that compound 48/80 can extrude histamine from the heart (guinea pig heart: Gomez and Antonio, 1981; human heart: Patella et al., 1995). Unfortunately, it is currently unclear from which cell type histamine is released in H 2 -TG atrium by compound 48/80. On the one hand, it is assumed that muscle cells, likewise including cardiac muscle cells, cannot store histamine (Fogel, 2015) but on the other hand, mast cells, a main source of histamine, are extremely rare in the mouse atrium as we noted here. So far, this question remains unanswered.
Histamine N-methyltransferase (HMT) converts and inactivates histamine by forming 1-methyl-histamine (Ganellin, 1982). HMT was found within liver cells and is thought to degrade intracellular histamine (Brown et al., 1959;Yoshikawa et al., 2013). At least in the guinea pig heart and pig hearts, HMT was detected earlier (Tahara et al., 2000;Klocker et al., 2005). HMT can be inhibited by diphenhydramine and other histamine receptor antagonists which might pose problems in data interpretation (Adachi et al., 2011). In the rat brain, HMT inhibitors could increase the levels of histamine in vivo (Duch et al., 1978), but which is difficult to compare with our present findings. It is suggested that HMT is localized intracellular and histamine has to be transported into the cytosol before it can be methylated by HMT (Ogasawara et al., 2006). Therefore, one could speculate that the role of HMT for degradation of exogenously administered histamine in the mouse heart in vitro, like in our experiments, depends also on the activity of the cardiac monoamine transporters. Moreover, the uptake of exogenously applied histamine into the cells might be slower in the heart than in the brain and species differences should be considered. All this might explain the missing effect of HMT inhibition in isolated atria. This needs to be clarified. There are mutations of HMT in patients with reduced enzymatic activity and reduced stability (Heidari et al., 2015). These patients should have higher levels of histamine in the heart leading to arrhythmias and hypotension (higher histamine levels in vessels) judged from our present work. But this hypothesis requires testing. In all regions of the mouse heart and (as a novel finding) in human atrium, we noted HMT using enzymatic assays, Western blotting or immunohistochemistry. No differences in activities between H 2 -TG and WT were noticeable.
Typically, extracellular histamine will be degraded by DAO. At least in the intestinal tract, DAO is released by enterocytes and can destroy histamine present in food, possibly a protective mechanism (Mainz et al., 2006). DAO is thought not to be expressed in the central nervous system but in the periphery (Yoshikawa et al., 2013). Nevertheless, it was shown that DAO can be present under physiological conditions in the plasma of mouse and man (Schwelberger, 2010) and DAO was detected earlier in the mouse heart (Bono et al., 1999) but not e.g. in pig hearts (Klocker et al., 2005) indicating relevant species differences. Exogenous histamine increased activity of DAO in the plasma, supposedly by a negative feedback loop (Wollin et al., 1998). One might argue that we detected only minute amounts of DAO in the heart that might be insignificant for histamine inactivation. Therefore, the effect of DAO might be indirect and in the future it might be possible to resolve this issue by use of DAO knock-out mice. Aminoguanidine is a classical inhibitor of the activity of DAO (Sattler et al., 1985). However, clinically utilized drugs (with totally different chemical structures) like the antihypertensive drug dihydralazine, the anti-malarial drug chloroquine, the antibiotic cycloserine and the muscle relaxant tubocurine (Sattler et al., 1985), but also cimetidine, diphenhydramine and ethanol (Wantke et al., 1998) can inhibit the activity of DAO. Thus, in some patients these drugs might inadvertently increase histamine levels in the human heart and might thus likewise lead to cardiac arrhythmias (as histamine, like all cAMP increasing drugs, is known to induce arrhythmias in some patients: Wolff and Levi, 1986;atrial fibrillation: Layritz et al., 2014). At least in pigs, drastic hypotension (as a result of histamine induced vasodilation) was noted when animals were given 60 mg histamine per os in the additional presence of aminoguanidine (Sattler et al., 1988). Mutations of DAO have been reported which are implied in gastrointestinal diseases (Maintz et al., 2006;Jones and Kearns, 2011). It would be interesting to study cardiac function after histamine infusion in such patients to find out whether these mutations might induce arrhythmias by increasing histamine levels in the heart. Moreover, in clinical studies an increase in cardiac levels of histamine correlated with the presence of heart failure (e.g. Patella et al., 1998). This might be a compensatory effect like the increase in noradrenaline in heart failure patients, in order to sustain cardiac force of contraction. High levels of histamine in an animal model increased the incidence of myocardial infarction (Dai, 1976). Of note, the activity of DAO was very small in mouse cardiac tissue and human atrial tissue whereas the effect of aminoguanidine on the PIE of histamine is quite pronounced ( Figure 5). This apparent discrepancy might indicate that the localization of DAO on the cell surface (in this case the sarcolemma) might be strategically important to diminish the level of histamine on the crucial histamine receptor outer surface. But this remains speculative.
The degradation product of histamine called 1methylhistamine is oxidized in mitochondria by MAO-B (Waldmeier et al., 1977). In human hearts, about similar levels of MAO-A and MAO-B were immunnologically detected (Duicu et al., 2015). Rat hearts contain MAO-A and -B in a ratio of 100:1 (Saura et al., 1992). Functionally, the rat heart mainly expresses MAO-A whereas the mouse heart mainly expresses MAO-B (Dorris, 1982). That indicates that we probably have inhibited mainly MAO-B in our mouse model with tranylcypromine and indicates that the contractile alterations in the presence of tranylcypromine might be due to inhibition of MAO-B in H 2 -TG atrium. Using Western blotting, enzyme activity measurement and immunohistology we could detect high expression and activity of MAO-B in all regions of the mouse heart investigated as well as in human atrial samples. However, one might argue that the importance of MAO-B remains unclear as inhibition of the activity of MAO-B would only increase 1methylhistamine levels but not necessarily histamine levels in the heart ( Figure 2). Thus, we cannot rule out indirect effects of tranylcypromine.

CONCLUSION
In conclusion, the data in this study indicate that exogenously applied histidine can be converted to inotropically active histamine, which is then subject to degradation in the mammalian heart. But further studies are necessary to clearly localize the source of histamine and the pathway of degradation in the heart. Negative controls (Ctr) of WT and H 2 -TG samples. HDC could be localized in the myocardium of WT mice (C,E) and in the myocardium of H 2 -TG mice (D,F). Positive immunostaining was observed in the cardiomyocytes as well as in the endothelium (arrows) and in the vascular smooth muscle cells (tailed arrows) of blood vessels (G,H) Immunohistochemical localization of histamine in the myocardium of WT (G) and H 2 -TG (H) mice. Along with cardiomyocytes, positive immunostaining can also be seen in the endothelium (arrow). Immunocomplexes were detected by DAB-HRP staining (brown staining) followed by counterstaining of nuclei with haematoxylin. Horizontal bars indicate 50 µm.