Author contributions: K.N.: conception and design, collection, assembly, analysis and interpretation of data, and manuscript writing; T.W., E.B., A.F., T.M., and S.K.: provision of study material or patients; B.M.: performing and evaluating of RT-PCR and in vitro experiments and collection and/or assembly of data; B.R.: performing and evaluating superoxide assays; A.P.: collection and/or assembly of data; E.M.: conception and design, data analysis and interpretation, and illustrations and manuscript writing.
Disclosure of potential conflicts of interest is found at the end of this article.
First published online in STEM CELLSEXPRESS November 1, 2011.
There are several clinical trials worldwide using bone marrow stromal cells (BMSCs) as a cellular therapy to modulate immune responses in patients suffering from various inflammatory conditions. A deeper understanding of the molecular mechanisms involved in this modulatory effect could help us design better, more effective protocols to treat immune mediated diseases. In this study, we demonstrated that human BMSCs express H1, H2, and H4 histamine receptors and they respond to histamine stimulation with an increased interleukin 6 (IL-6) production both in vitro and in vivo. Using different receptor antagonists, we pinpointed the importance of the H1 histamine receptor, while Western blot analysis and application of various mitogen-activated protein kinase inhibitors highlighted the role of p38, extracellular signal-regulated kinase, and c-Jun N-terminal kinase kinases in the observed effect. When BMSCs were pretreated with either histamine or degranulated human mast cells, they exhibited an enhanced IL-6-dependent antiapoptotic effect on neutrophil granulocytes. Based on these observations, it is likely that introduction of BMSCs into a histamine-rich environment (such as any allergic setting) or pretreatment of these cells with synthetic histamine could have a significant modulatory effect on the therapeutic potential of BMSCs. STEM CELLS 2012; 30:222–231.
Bone marrow stromal cells (BMSCs, also called mesenchymal stem cells or MSCs) are known to generate osteogeneic, adipogeneic, and chondrogeneic lineages and help to maintain the microenvironment necessary for hematopoiesis. In the last 5 years, it became evident that BMSCs—in addition to nursing hematopoietic stem cells—also have potent immunoregulatory functions; they appear to regulate the differentiation, survival, and activation of a wide variety of immune cells [1–4]. In recent studies, we have shown that when BMSCs are introduced into a pathological milieu, they can detect soluble disease-specific mediators (e.g., lipopolysaccharide and tumor necrosis factor α in sepsis or interleukin 4 [IL-4) and IL-13 in asthma) and respond to them in ways that are optimal for the host [5, 6].
Histamine is a biogenic amine that plays an essential role in several physiological and pathophysiological processes. It acts as a neurotransmitter in the central nervous system, regulates HCl synthesis in the stomach, and mediates anaphylaxis in allergic conditions. Released from mast cells (MCs) and basophil granulocytes, it also contributes to the pathological changes seen in sepsis and several autoimmune and alloimmune disorders [7, 8] and plays an important role in immunomodulation  through its effect on cytokines. Histamine appears to stimulate the production and release of cytokines including IL-1α and IL-6 in different cells, both known to be produced by BMSCs [10–12] and has proinflammatory, anti-inflammatory, and immunoregulatory functions [13, 14]. As in our previously published study , a G-protein–coupled receptor (GPCR) array based on multiplex polymerase chain reaction (PCR) suggested that BMSCs might express mRNAs encoding at least two of the histamine receptors, H1 and H2, we wondered whether BMSC-derived IL-6 production might be regulated by histamine in vitro as well as in vivo.
MATERIALS AND METHODS
BMSCs were isolated from bone marrow aspirates donated by healthy volunteers. Cells were grown in complete minimum essential medium α)MEM-α) medium (20% fetal bovine serum [FBS], 1% Pen/Strep, 1% glutamine). Characterization of the cells showed osteogeneic and adipogeneic differentiation potential in vitro and expression of various BMSC-specific cell surface markers (CD73, CD90, CD105, and CD146) but lack of hematopoietic markers (CD45, CD14, and CD34).
For immunostaining, human BMSCs in chamber slides (eight-well chambers from Lab-Tek) were used (5,000 cells per well fixed with 4% paraformaldehyde). The slides were blocked with 1× Universal Blocking Reagent (Biogenex, San Ramon, CA) for 10 minutes at room temperature (RT). Primary antibodies (rabbit anti-human H1R, H2R, H3R, H4R from Alpha Diagnostic) were applied overnight at 4°C, diluted 1:100 in 1% BSA containing 0.25% Triton X100. After washing three times in phosphate-buffered saline (PBS), Alexa-488-conjugated anti-rabbit secondary antibody was used at a 1:1,000 dilution for 1 hour at RT. 4′,6-Diamidino-2-phenylindole was used to visualize cell nuclei. The fluorescent staining was observed with a Leica DMI600 inverted fluorescent microscope using the FITC filter set.
Reverse Transcriptase PCR
Total RNA was isolated from human BMSCs using the Stratagene Absolutely RNA Microprep Kit (Stratagene, La Jolla, CA) with modification of the DNase treatment step. Briefly, on-column DNA digestion was performed using Qiagen DNase (RNase-free DNase Set, Qiagen, Valencia, CA) at RT for 30 minutes. Isolated RNA (40 ng) was amplified with the one-step QuantiTect SYBR Green RT-PCR Kit (Qiagen) using previously published primers for H1R, H2R, H3R (16), and H4R (17).
The one-step reverse transcriptase PCR (RT-PCR) conditions were as follows: 50°C, 30-minute reverse transcription, 95°C, 15-minute initial activation of the Taq polymerase and denaturation of the reverse transcriptase, and then 40 cycles of 94°C, 15-second denaturation, 58°C, 30-second annealing, and 72°C 30-second extension. For the H4R primers, the annealing temperature was 55°C. PCR product was run on 2% agarose gel and visualized with ethidium bromide.
Enzyme-Linked Immunosorbent Assay (ELISA)
IL-6 and IL-8 levels were measured by ELISA following the DuoSet ELISA Development Kit Instructions (R&D Systems).
BMSCs were collected following trypsin digestion, and total protein concentrations were determined using the Micro BCA Kit (Pierce, www.piercenet.com). Total protein lysates (20 μg) were run on 4%–12% Bis-Tris gels and blotted onto nitrocellulose membranes (Invitrogen). The membranes were blocked with 4% nonfat dry milk in TBS (1× TBS, 0.01% Tween 20) and incubated overnight at 4°C with antibodies against phosphorylated p38 mitogen-activated protein (MAP) kinase (p38), phosphorylated p44/42 MAP kinase (extracellular signal-regulated kinase [ERK]), and phosphorylated c-jun terminal NH2 kinase at 1:500, 1:1,000, and 1:500 dilution, respectively, or as a control with antibodies against total p38, total ERK, and total c-Jun N-terminal kinase (JNK) at 1:1,000 dilution each. All antibodies were from Cell Signaling (www.cellsignal.com).
Anti-rabbit horseradish peroxidase (Jackson Immunoresearch, www.jacksonimmuno.com) was used as a secondary antibody for 1 hour at RT. The blots were visualized with Western lightning enhanced chemiluminescence reagents (Perkin Elmer, www.perkinelmer.com).
Histamine Stimulation with or Without Antagonists, Inhibitors
For in vitro IL-6 measurements, BMSCs were cultured in 96-well flat bottom cell culture plates (Corning, catalog2.corning.com/Lifesciences/) at a density of 20,000 cells per well per 200 microliters in complete MEM-alpha medium. Histamine was applied at a concentration of 10−4, 10−5, 10−6, 10−7, or 10−8 M, and histamine receptor antagonists were applied at a concentration of 10−5 M (diphenhydramine hydrochloride [H1 antagonist], cimetidine [H2 receptor antagonist], ciproxifan hydrochloride [H3 antagonist], JNJ7777120 [H4 antagonist]). MAP kinase pathway inhibitors (SP203580 a specific inhibitor of p38, PD98059 a specific inhibitor of ERK phosphorylation, SP600125 a specific inhibitor of JNK) and the phospholipase C (PLC) inhibitor U73122 were used at 10−5 or 10−6 M concentrations. Antagonists or inhibitors were added 1 hour prior to addition of histamine. IL-6 was measured from the supernatants after 6-, 12-, or 24-hour histamine stimulation using DuoSet ELISA Kit following the manufacturer's instructions (R&D Systems, www.rndsystems.com).
Stimulation of BMSCs with Additional Mast Cell-Derived Factors
To test the hypothesis of additional MC-derived factors affecting the BMSCs IL-6 production, we repeated the in vitro experiments and stimulated the BMSCs with additional factors released by MCs , such as leukotriene C4 (LTC4), platelet-activating factor (PAF), serotonin (5HT), leukotriene B4 (LTB4), and prostaglandin D2 (PGD2). We chose these compounds, because based on our GPCR array  BMSCs bear the appropriate receptors to respond (Supporting Information Table 1). A total of 10,000 BMSCs per well were plated in 96-well plates in MEM-alpha medium with 5% FBS. After overnight incubation, the medium was changed and the cells were stimulated with PGD2 at a concentration of 10−5, 10−6, and 10−7 M for 6, 12, or 24 hours. The same volume of medium and vehicle (ethanol) was added to the control wells as in the wells with PDG2 stimulation. Other mediators we used to stimulate BMSCs for 6 hours in these experiments were: serotonin (concentration range: 10−4–10−10 M), platelet activating factor (10−6–10−12 M), LTB4 (10−6–10−10 M), and LTC4 (10−6–10−10 M). All of the mediators were purchased from Cayman Chemical (Ann Arbor, MI, www.caymanchem.com), except serotonin that was from Tocris Bioscience (Ellisville, MO, www.tocris.com). Cell culture supernatants were measured in quadruplicates using the human IL-6 DuoSet ELISA Kit (R&D Systems). Nonstimulated cells (medium alone) and histamine-stimulated cells were included in each assay (data not shown). Experiments were done using BMSCs derived from four different donors.
Culture and Assays Involving Human Mast Cells
Human-derived MCs (HuMC) were cultured as described. In brief, peripheral blood CD34+ progenitor cells were collected from healthy donors after informed consent and affinity column apheresis. CD34+ cells were cultured in StemPro-34 serum-free medium (Life Technologies, Grand Island, NY, www.lifetechnologies.com/us/en/home.html) in the presence of recombinant IL-3 (first week only), recombinant IL-6, and recombinant human stem cell factor (rhSCF; Peprotech, www.peprotech.com). HuMC cultures were maintained up to 10 weeks at 37°C and 5% CO2.
HuMCs were sensitized overnight with 100 ng/ml biotinylated IgE (Sigma-Aldrich, www.sigmaaldrich.com). The next day, 50,000 HuMCs were added to the upper chamber of a 96-well transwell system (Corning) in complete MEM-alpha medium and degranulated by the addition of 100 ng/ml streptavidin (Sigma-Aldrich). After 6 hours, the transwell inserts were removed and the lower chamber was gently washed with cell culture medium twice. After the last wash, 200 μl fresh medium was added to each wells. After 6 hours, supernatants were collected and assayed for IL-6.
Isolation of Granulocytes from Whole Blood
Granulocytes were isolated using Lymphocyte Poly(R) density gradient (Cedarlane Labs, www.cedarlanelabs.com) following the manufacturers instructions. Briefly, anticoagulated whole blood is layered on top of the density solution. After centrifugation, the granulocyte layer was removed, and contaminating red blood cells were lysed using ACK lysing buffer. Granulocyte purity was routinely more than 90% as confirmed by FACS using CD66b-FITC antibody (BD Biosciences, www.bdbiosciences.com/home.jsp). For the superoxide production experiments, granulocytes were isolated using a different method. In these experiments, anticoagulated (heparin) blood was layered on top of Histopaque 1119 (Sigma, St. Louis, MO, www.sigmaaldrich.com/united-states.html) and centrifuged (800g, 30 minutes RT). Upper serum layer was aspirated; the middle phase containing white blood cells was collected. Pellets were resuspended in PBS and layered on top of a Percoll gradient (65%, 70%, 75%, 80%, and 85%, Sigma, St. Louis, MO). After centrifugation (800g, 20 minutes RT), the 70%–75%–80% Percoll layers containing neutrophils were collected, washed, and resuspended in PBS. Viability of the cells was determined by Trypan Blue dye extrusion and resulted in >99% viable neutrophils. The purity of the preparations was determined by Wright-Giemsa staining and yielded >95% neutrophil granulocytes (PMNs).
Granulocyte/BMSC Cocultures and Assay for Apoptosis
A total of 200,000 granulocytes were added in 200 μl complete medium to 96-well U-bottom plates either alone or together with 20,000 BMSCs. After 18 hours of culture, percentage of apoptotic granulocytes were assessed by FACS using Annexin-V–allophycocyanin (APC) and 7-aminoactinomycin D (7-AAD) (both from BD Biosciences). Annexin-V single-positive cells were considered early apoptotic, whereas Annexin-V/7-AAD double-positive cells were late apoptotic.
In case of histamine or HuMC prestimulation of stem cells, BMSCs were pretreated for 24 hours with 10−5 M histamine or cocultured for 6 hours with 50,000 degranulated HuMCs. Granulocytes were given thereafter.
Granulocyte/BMSC Cocultures and Measurement of Neutrophil Superoxide Production
Granulocytes were cultured with or without BMSCs in Roswell Park Memorial Institute 1640 complete medium on six-well plates at a 500:1 granulocyte/BMSC ratio for 24 hours. Neutrophils were collected, washed, resuspended in Hanks' balanced salt and transferred to 96-well white opaque plates. Cells were stimulated in triplicates at a final concentration of 1.2 × 106 per milliliter with 1 μM formyl-methionyl-leucyl-phenylalanine (fMLP; Sigma, St. Louis, MO) or 10 nM phorbol myristate acetate (Sigma, St. Louis, MO). Superoxide production was measured by the Diogenes cellular luminescence enhancement system (National Diagnostics, Atlanta, GA, www.nationaldiagnostics.com). Luminescence was followed for 15 minutes on a Luminoskan Ascent microplate luminometer (ThermoScientific, Hudson, NH, www.thermoscientific.com). Cumulative superoxide production was calculated as the area under the measured curve and it is presented as integrated relative luminescence units.
In Vivo Experiments
Assaying the Specificity of Histamine in the Effect
6–8 week-old male WT C57BL/6 mice (JAX, jaxmice.jax.org), WT BALB/c, and HDC (histidine-decarboxylase) KO mice (provided by Dr. Andras Falus, Department of Genetics, Cell- and Immunobiology, Semmelweis University, Hungary) or MC-deficient KitW-sh/W-sh mice (provided by Dr. David M. Segal, NCI, NIH) were sensitized i.p. with 1 μg monoclonal IgE anti–2,4-dinitrophenol (anti-DNP) (Sigma) in 200 μl PBS followed 24 hours later by i.p. challenge with 0.5 mg/ml DNP–human serum albumin in 200 μl PBS. Immediately following challenge, 4 million human BMSCs were injected into the peritoneal cavity. After 6 hours, peritoneal cavity was washed with 2 ml PBS. After centrifugation of the lavage fluid to remove cellular components, the supernatants were frozen and stored at −80° until assaying for human IL-6.
Testing Histamine Prestimulation on BMSCs Effect in Zymosan-Induced Peritonitis
Zymosan A (2 mg/ml; 500 μl [Sigma]) was injected intraperitoneally into 10-week-old male C57/B6 mice (n = 4 per group). Human BMSCs (0.5 million in 200 μl PBS) were injected 30 minutes later intraperitoneally. In some cases, BMSCs were prestimulated overnight with 10−4 histamine or cocultured with degranulated human MCs with transwell separation for 6 hours. After 6 hours, peritoneal cavity was washed with 5 ml PBS. Peritoneal cells were analyzed using FACS. To rule out possible contamination of human BMSCs, first we gated on mouse CD45-positive cells (CD45-PerCP from BD Biosciences). PMNs were identified as Gr-1 (high) CD11b (high) double-positive cells (Gr-1 FITC, CD11b-PE both from BD Biosciences). Apoptotic cells were detected using APC-conjugated Annexin-V (eBioscience). The readout of the experiment was the absolute number of living (functional) peritoneal neutrophils.
The actions of histamine are mediated by four G-protein-coupled receptor subtypes (H1–H4), each of which has a tissue- and cell-specific distribution in the human body. To find out which histamine receptors are present in BMSCs, we used RT-PCR and immunostaining. We observed a relatively strong expression of the H1 receptor, a less pronounced expression of the H2 and H4 receptors, and a complete lack of H3 receptors (Fig. 1).
Next, we measured the amount of IL-6 in the media of human BMSC cultures after exposing the cells to different concentrations of histamine, and found a time-and dose-dependent increase in the IL-6 output. The dose/response curve was bell-shaped with a maximum IL-6 response (a fourfold to fivefold increase over the baseline) at 10−5 M histamine. This trend can be seen at all time points examined—6, 12, and 24 hours after introduction of histamine into the culture medium (Fig. 2A). This effect of histamine could be completely blocked using a specific H1 receptor antagonist but was not affected by H2–H4 receptor antagonists (Fig. 2B).
To see if histamine released from human MCs could also stimulate IL-6 production and secretion by BMSCs, we cocultured the two cell types in a transwell system. Preincubation of BMSCs with activated (degranulated) MCs resulted in a significant increase in their IL-6 secretion. Unlike when only histamine was used, this effect was blunted but not eliminated by the addition of a histamine H1 antagonist to the culture medium. The effect of H1 antagonist studied on IL-6 secretion (Fig. 3A) was incomplete suggesting that MC-derived factors in addition to histamine might mediate the effect on cytokine release. When we tested additional factors released by MCs that had the appropriate receptors present in BMSCs (Supporting Information Table 1), we found that PGD2 was the only compound that induced BMSC-derived IL-6 release (Fig. 3B).
Next, after inducing MC degranulation, we introduced human BMSCs into the peritoneal cavities of control and histamine-deficient mice—animals lacking HDC, an enzyme required for histamine biosynthesis . After 6 hours, IL-6 increased in peritoneal lavage fluid from control animals but significantly reduced in fluid from the histamine-deficient mice. It was not absent, however, suggesting that IL-6 production by the BMSCs in the presence of activated MCs is not only due to histamine but also due to other factors. When BMSCs were introduced into MC-deficient mice, the same sensitization/challenge protocol resulted in no IL-6 increase (Fig. 3C) suggesting that the MCs are indeed responsible for driving the release of IL-6.
Histamine mobilizes calcium in several cell types [8, 9]. To test if it has the same effect on BMSCs, we measured calcium concentrations in individual cells after addition of histamine at a dose of 10−5 M. Immediately after the amine was added, cytoplasmic calcium concentrations started to increase and peaked after 30 seconds and reached a plateau thereafter (Fig. 4A).
PLC plays a critical role in mobilization of intracellular calcium. Upon cleavage of phosphatidylinositol 4,5-bisphosphate, phospholipase C releases inositol trisphosphate, which triggers calcium release from the smooth endoplasmic reticulum. PLC inhibitors can prevent cells from mobilizing calcium in response to histamine, and we used U73122, a specific inhibitor of PLC, to determine whether this enzyme mediated the effect of histamine on BMSCs described above. When the cells were pretreated with this agent, histamine could no longer elicit the same level of IL-6 increase seen earlier, pointing to the importance of PLC-triggered cellular events (Fig. 4B)
Histamine has been shown to trigger the activation of MAP kinase pathways in different cell types . To determine whether histamine can stimulate MAP kinases in BMSCs, we performed Western blot analysis measuring the amount of phosphorylated p38, ERK, and JNK. Addition of histamine elicited a time-dependent activation of all of the kinases examined. Maximal phosphorylation of p38, ERK, and JNK occurred after 15 minutes and declined to basal level by 60 minutes (Fig. 4C). Levels of the nonphosphorylated MAP kinases studied remained unchanged at all time points.
To explore how much individual MAP kinases contribute to the histamine-induced IL-6 increase, we looked at the effects of specific MAP kinase inhibitors. SB203580, a p38 inhibitor, PD98059, an ERK inhibitor, and SP600125, a JNK inhibitor all significantly impaired histamine's ability to induce IL-6 production by the stem cells (Fig. 4D).
PMNs are professional phagocytes that rapidly engulf invading organisms that are coated with antibodies and complement. PMNs are inherently short-lived cells, with a half-life of approximately 6–12 hours in the circulation. At the end of their lives, they undergo spontaneous apoptosis. Recently Pistoya's group reported that BMSCs can significantly prolong survival of PMNs in vitro in an IL-6-dependent manner . As histamine seemed to increase IL-6 production by BMSCs, we asked whether histamine pretreatment could enhance the antiapoptotic effect of BMSCs on PMNs. As expected, BMSCs suppressed spontaneous apoptosis of neutrophils as measured by Annexin-V binding and concurrent 7-AAD staining using FACS. Histamine prestimulation of the BMSCs increased their efficacy; the number of apoptotic PMNs was even smaller than it had been when unstimulated BMSCs were cocultured with neutrophils (Fig. 5A–5C). To find out if histamine prestimulation of BMSCs can also help preserve the reactive oxygen species producing activity of neutrophils, we measured superoxide production of PMNs following coculture with unstimulated or histamine-pretreated BMSCs. Although histamine prestimulation of BMSCs did not cause a measurable effect on the spontaneous superoxide production of neutrophils, we were able to detect a significant increase in fMLP-triggered superoxide synthesis (Fig. 5D). To test if histamine pretreatment of BMSCs will result in an increase of neutrophil survival in vivo in an infectious environment, we used a model of zymosan-induced peritonitis . When the mice were injected with histamine-pretreated BMSCs intraperitoneally, we found that the peritoneal wash contained significantly higher numbers of live neutrophils than if the BMSCs were not pretreated or were cultured with activated MCs (Fig. 6).
To explore the molecular mechanisms through which BMSCs acted, we repeated the apoptosis experiments in the presence of an IL-6 blocking antibody or using an H1 receptor inhibitor. Potentiation of the antiapoptotic effect of BMSCs was eliminated when neutralizing anti-IL-6 antibodies were added to the culture medium, suggesting that IL-6 mediates the phenomenon. In another set of experiments, instead of adding histamine we used degranulated human MCs to prestimulate the stromal cells. BMSCs treated this way exhibited a significantly greater antiapoptotic effect than those treated with histamine alone, and the increase in PMN survival in response to degranulated MCs was only partly countered by a histamine H1 receptor antagonist. As stated earlier, the most likely reason for this is that factors in addition to histamine are secreted by MCs to modulate BMSCs. Based on our findings, these other factors are likely to contribute to the increased production and secretion of IL-6, which appears to be the principal mediator of the antiapoptotic effect of the BMSCs on PMNs.
Histamine was discovered a century ago by Dale and Laidlaw , and it soon became obvious that the amine has a variety of biological functions. With regard to BMSCs, histamine has been suggested to play a role in osteogenesis, and the H1 receptor was identified as one of the downstream targets for the vitamin D receptor and was suggested to play a role in osteogenic differentiation of BMSCs . It soon became clear that histamine was also a very important proinflammatory mediator. Its role in the immediate type allergic reactions was quickly recognized, but its other immunomodulatory effects were only appreciated in the last decade. Histamine is now known to alter cytokine and chemokine production by a variety of cells including MCs, eosinophils, dendritic cells, and T cells [9, 24-26]. We wondered if and how histamine might affect the immunomodulatory functions of human BMSCs [8, 11, 27-30]. Histamine was described to induce mouse bone marrow cells to produce IL-1α, a proinflammatory cytokine that is needed in small amounts to combat infection and initiate healing processes [31, 32] but one that can be harmful when its levels are excessive. A human disease, mastocytosis is a condition characterized by infiltration of MCs into the tissues of the body, including the bone marrow. MCs release chemicals including large amounts of histamine during a variety of inflammatory and allergic conditions. In patients with mastocytosis, the plasma histamine levels are usually high and increased IL-6 levels were reported . In patients with elevated IL-6 levels, a significant increase in circulating neutrophil counts was also observed . We hypothesized that human BMSCs might respond to histamine stimulation by increasing their IL-6 production just as other immune cells do. Using cultured human BMSCs, we found that histamine stimulation results in an increase of BMSC-derived IL-6 secretion in a dose- and time-dependent manner. Next, we asked which of the four histamine receptors might be responsible for this effect and found that in the in vitro conditions it is the H1 receptor that is responsible for the majority of the increase in BMSC IL-6 production in response to histamine. Histamine is a major mediator of acute inflammatory and immediate hypersensitivity responses and MCs are known to release large amounts of histamine during degranulation, thus they might be a natural source of histamine when interacting with BMSCs in vivo. This is why we used a coculture of human BMSCs and human MCs as the source of histamine, and found that—unlike when histamine alone was used—the H1 receptor blockade only partially eliminated the IL-6 increase in the medium. To determine the contribution of MC-derived histamine to the effect seen, we performed an in vivo experiment when we introduced human BMSCs into the peritoneal cavity of control and histamine-deficient mice  after inducing MC degranulation. Although the increase in IL-6 content of the cell-free peritoneal wash was significantly reduced in the histamine-deficient mice, the IL-6 content of the peritoneal wash was still significantly higher than that without inducing MC degranulation. The most likely explanation for this result is that degranulating MCs release a number of mediators in addition to histamine, such as monoamines, leukotrienes, proteases, and SCF [7, 18]. We focused on a few MC mediators that affect receptors that we have found to be present in human BMSCs based on our GPCR array . These included LTB4, LTC4, PAF, serotonin, and PGD2. We found that in our experimental conditions only PGD2 was able to induce IL-6 release from BMSCs at the 6-hour time point. In fact, as PGD2 is one of the major lipid mediators in MCs , it is possible that the IL-6 increase that we could not block using H1 blockers is due—at least in part—to released PGD2.
Using kinase inhibitors, we established that the signaling molecules responsible for IL-6 induction belong to the MAP kinase family; p38, ERK, and JNK are all phosphorylated in BMSCs in response to histamine. This result is in agreement with data showing that histamine receptors trigger the MAPK pathways in other cell types . In addition to MAP kinases, Ca signaling has also been suggested to play a role in histamine receptor-induced cellular effects  and we indeed found that addition of histamine induces increase in intracellular calcium concentrations and blocking calcium mobilization eliminates the IL-6 increase.
After we established that both histamine and MCs induce human BMSCs to produce and release IL-6, we wondered about the physiological significance of this effect. A number of studies suggested that IL-6 can block apoptosis of a variety of cell types including immune cells and neural stem cells [11, 12, 29, 34, 35]. As our previous study in septic mice  suggested that increasing the number of circulating PMNs has a significant effect in fighting infection, we wanted to see whether BMSCs might release IL-6 to decrease PMN apoptosis. Our results showed that in a coculture system of human PMNs and BMSCs, the BMSCs effectively suppressed spontaneous apoptosis of neutrophils. Prestimulation of BMSCs with histamine further decreased the number of apoptotic PMNs while increasing fMLP-induced superoxide production of these phagocytic cells. Coculturing the BMSCs with degranulated human MCs before they were added to PMN cultures significantly increased the antiapoptotic effect. This effect that could only be partially blocked by H1 blockade is suggesting a role for other MC-derived factors in the process. To examine the significance of IL-6 in such effect, we also used neutralizing antibodies to IL-6 in the media, and this treatment eliminated the antiapoptotic function. We set up an in vivo experiment using a mouse model to test if histamine pretreatment would increase the efficiency of BMSCs in improving PMN survival. In this model, zymosan, a polysaccharide cell wall component derived from Saccharomyces cerevisiae, causes an acute inflammation, and the zymosan-induced peritonitis model is extensively used to quantify the recruitment of monocytes and neutrophils into the peritoneal cavity . We found that the BMSCs were significantly more efficient if they were pretreated with histamine prior to their intraperitoneal injection. In the peritoneal wash, we found a significant increase in the number of live neutrophils if these histamine-pretreated BMSCs were applied. Interestingly, although, if instead of pretreatment with histamine, the BMSCs were cocultured with human MCs prior to injection, we did not find more PMNs in the peritoneal wash. This might be due to a variety of factors, such as using human MCs for pretreatment (unlike our previous in vivo experiment, when we inject the human BMSCs and the responding cells are naturally mouse MCs) or the likelihood that MCs release many additional moderators other than histamine, and some of these might have opposing effects on neutrophil survival.
Most of the in vivo studies suggesting the immunomodulatory actions of BMSCs are based on i.v. injection of BMSCs. After introducing the BMSCs into the circulation, the cells will end up in small vessels (arterioles, capillaries, and postcapillary venules) by following the natural path of blood circulation. In these vascular beds, some of the BMSCs might participate in clot formation that attracts large number of platelets (that might also release histamine) while others might engage in direct interactions with endothelial cells and eventually enter the perivascular space. During these processes, the BMSCs are exposed to platelet and endothelial cell-derived factors both of which might release histamine [24, 36] that will activate the BMSCs. This activation can be further enhanced when BMSCs encounter MCs in the perivascular space (where most of the MCs reside), especially when administered in an allergic setting or in a tissue that previously recruited large number of MCs (as seen in mastocytosis or in tolerated allogeneic organ grafts). In these in vivo settings, the BMSCs could be bathed in histamine that would cause them (through their histamine receptors) to make and release large amounts of IL-8 a potent neutrophil chemoattractant and a cytokine known to potentiate the oxidative burst of neutrophils  as well as IL-6 a known antiapoptotic cytokine (Fig. 7). From the four receptors examined, the H1 receptor had the highest relative expression in BMSCs, and using receptor-specific antagonists, we also found that histamine-induced IL-6 production in vitro is solely dependent on H1 binding. Our experiments using human BMSCs and histamine-deficient mice confirmed that MC-derived histamine works in concert with other MC-specific factors to induce the observed IL-6 increase.
Our studies revealed another new mechanism underlying the anti-inflammatory/anti-infectious effects of human BMSCs. We established the presence of three of the histamine receptors on these cells and pinpointed the H1 receptor as the major target receptor for the effect. As H1 receptor antagonists are available and widely used, their effects on the BMSCs function could be significant and this possible side effect should be considered. Conversely, as histamine treatment seems to induce the ability of BMSCs to attract PMNs and block their apoptosis, one could also imagine an advantage of using pretreatment of donor BMSCs before their therapeutic use to enhance their efficacy in improving survival in septic settings or in other inflammatory disorders in patients.
We thank Dr. Dean Metcalfe for his help with the human mast cells and Dr. Tamas Balla for his guidance and help with calcium measurements. This work was supported by the DIR, NIDCR, of the IRP, NIH, DHHS.