Mechanisms of Nonimmunological Histamine and Tryptase Release from Human Cutaneous Mast Cells

Human foreskins were obtained after circumcision. The skin was immediately placed in Ca²⁺–Mg²⁺-free Hanks’ balanced salt solution (CMF-HBSS; Sigma Chemical, St. Louis, MO) at 4°C and used within 24 h of circumcision. 4,6,12–16After removal of the subcutaneous fat the skin was weighed, chopped into 1- or 2-mm fragments, and washed twice with CMF-HBSS (10 ml/g) at room temperature. The tissue fragments were then incubated in CMF-HBSS with collagenase (20 mg/g), hyaluronidase (4 mg/g), and deoxyribonuclease (1,000 U/foreskin; Worthington Biochemical, Freehold, NJ) in a total volume of 5 ml/g wet weight of tissue for 3 h at 37°C, with constant stirring. After incubation the digested tissue was filtered through a 150-μm pore-sized Nytex cloth (Tetko, Elmsford, NJ) and washed three times with 3.5% bovine serum albumin made in CMF-HBSS (Sigma Chemical).

Dispersed cells were resuspended in RPMI-1640 (Roswel Park Memorial Institute) culture medium (10 ml/g skin) with 25 mM HEPES (Sigma Chemical), 2 mM glutamine, 5% fetal calf serum (GIBCO, Grand Island, NY), 1,000 U/ml penicillin G, and 1% gentamicin (vol/vol; Sigma Chemical). The suspension was inoculated into a Petri dish and incubated overnight at 37°C in humidified 95% air and 5% CO².

After overnight culture, the cells were harvested and washed three times with PAG buffer (5 ml/g), [Pipes, 25 mM piperazine-N,N′-bis(-ethenesulfonic acid) (Sigma Chemical), 110 mM NaCl, 6 mM potassium chloride (KCL), and 0.1% dextrose, pH 7.4 at room temperature. They then were counted in a hemocytometer after staining with toluidine blue. Viability was assessed by trypan blue exclusion. If the assessed cell viability was less than 80%, the cells were discarded. A Wright-stained Cytospin preparation (Shandon Inc., Pittsburgh, PA) was used to determine the percentage of mast cells in each sample.

One hundred-microliter aliquots of the cell suspension, containing between 2 × 10⁴and 5 × 10⁴mast cells/experimental tube were incubated in duplicate for 30 min at 37°C with the following drugs: 4.76 × 10⁻⁵M calcium ionophore A23187 (Sigma Chemical); 3 × 10⁻³M vancomycin (Vancocin HCl; Lilly, Indianapolis, IN); 10⁻³M atracurium (Tracurium; Burroughs Wellcome, Triangle Park, NC); and 2.63 × 10⁻³M morphine (Mallinckrodt, St. Louis, MO). The concentrations chosen were based on in vitro  data 3and data applied to clinically relevant doses. 17To determine the effects of calcium on histamine release, cells were incubated either in buffer from which Ca²⁺and Mg²⁺were omitted or in buffer containing 2.8 mM Ca²⁺and 1.0 mM Mg²⁺.

The release reaction was stopped by addition of 900 μl ice-cold CMF-HBSS and centrifugation at 1,000 g  for 5 min at 4°C. Collected supernatants were split into two aliquots and frozen at −70°C until analysis time. The total-release tubes had 900 μl of sterile ice water added instead. Total histamine–tryptase content of the skin mast cells was assessed by lysis of the cells by three cycles of freezing and thawing. The spontaneous release of histamine–tryptase was estimated from tubes to which buffer, instead of releasing agent, was added. Also, if the baseline release of histamine was greater than 15% of the total histamine release, the cells were deemed damaged during the preparation process and data were excluded. Histamine–tryptase release was calculated as percentage of total release. Net percentage of histamine–tryptase release is expressed as a percentage of total histamine–tryptase corrected for spontaneous release, according to the formula

Net %release = (T − B)/T · 100

where T is total histamine and B is spontaneous release.

Histamine concentrations were measured by the enzyme immunoassay method using commercially available kits (Immunotech, Westbrook, ME). Tryptase concentrations were measured using the UniCap 100 automated apparatus (Pharmacia and Upjohn, Kalamazoo, MI) and the UniCap 100 fluoroenzyme immunoassay.

For elucidating the mechanisms of histamine and tryptase release, two different types of enzymatic inhibitors were studied using a stimulus-secretion paradigm, with histamine and tryptase release from the mast cells as the endpoint.

3-(4-Octadecyl)-benzoylacrylic acid (OBAA; BIOMOL Research Laboratories Inc.; Plymouth Meeting, PA) is a potent inhibitor of PLA². Ten milligrams OBAA was dissolved in 9.34 ml absolute ethanol and split into 500-μl aliquots that were stored at −70°C. Just before use, 40 μl ethanol solution was further diluted with 1:1 solution (vol/vol) of ethanol–PAG to a final concentration of 24 μM. For the inhibition studies, 5-μl aliquots of the ethanol–PAG solution were added to 100 μl cell suspensions (resulting in a final concentration of 1 μM of the inhibitor). Tubes were preincubated for 30 min at 37°C before addition of vancomycin or calcium ionophore A23187, and the cell suspensions were incubated for an additional 30 min. Ethanol in PAG (2.08%) and OBAA without addition of drugs were used as controls.

Five milligrams U-73122 was dissolved in 5 ml methylene chloride and split into 100-μl aliquots, which were stored at −70°C. Just before use, methylene chloride was evaporated under nitrogen and U-73122 was redissolved in 150 μl absolute ethanol. This was further diluted with 927 μl PAG buffer, resulting in a 199.8-μM U-73122 solution. For the inhibition studies, 5 μl PAG–ethanol solution was added to 100 μl cell suspension, resulting in 8.36 μM U-73122. Tubes were preincubated for 30 min at 37°C before addition of vancomycin or calcium ionophore A23187, and the suspensions were incubated for an additional 30 min. Ethanol in PAG (0.58%) and U-73122 without addition of mediators were used as controls.

Results are presented as the mean and SD, and the number of experiments (n) is indicated. All two-group comparisons were performed using a paired or unpaired t  test, with the level of statistical significance set at P < 0.05. Multigroup comparisons were made using one-way analysis of variance, followed by a Scheffépost hoc  test for multiple comparisons, with P < 0.05 considered to be significant.

Each foreskin yielded between 2 × 10⁵and 17 × 10⁶cells, with 4–8% being mast cells. In the presence of Ca²⁺and Mg²⁺in the incubating buffer, we found that vancomycin, calcium ionophore A23187, morphine, and atracurium resulted in significant histamine release (31 ± 5.1, 60 ± 14.2, 17.4 ± 10.2, and 8.8 ± 5.2%, respectively), with an average baseline release of 6.9 ± 3.6% (fig. 1). Tryptase release was 22.7 ± 6.8, 18.1 ± 10.5, 22.2 ± 13.0, and 11.6 ± 6.3, respectively, with an average baseline release of 10.8 ± 5.9% (P < 0.005;fig. 2). Also, there was a direct correlation between the net percentage releases of histamine and tryptase by vancomycin, morphine, and atracurium (r²= 0.6;fig. 3), but no correlation between net percentage releases of histamine and tryptase for calcium ionophore A23287.

If calcium and magnesium were omitted from the incubation buffer, histamine release from neonatal skin mast cells was decreased by 90% for vancomycin (8.5 ± 5.9%; baseline 6.0 ± 4.0%) and virtually abolished for morphine (5.7 ± 2.5%; baseline 5.6 ± 3.5%;P < 0.05). Conversely, calcium had no effect on release reaction of A23187 (55.8 ± 15.1%; baseline 6.5 ± 3.7%;P < 0.0001) or atracurium (5.4 ± 2.5%; baseline 3.7 ± 2.0%;P = 0.0004;fig. 4). Furthermore, there was a significant difference in net percentage histamine release between Ca–Mg(+) and Ca–Mg(−) groups if vancomycin (P < 0.0001) or morphine (P = 0.0012) was used as the mast cell activator, but no difference was seen with calcium ionophore A23187 (P = 0.57) or atracurium (P = 0.85;fig. 5).

Preincubation of the dispersed mast cells with the secretory PLA²inhibitor OBAA attenuated vancomycin-induced histamine release from 19.4 ± 8.5% to 9.3 ± 6.3% (baseline: 6.0 ± 2.7%; control: 5.8 ± 2.3%; OBAA: 6.4 ± 2.9%); whereas the calcium ionophore A23187–induced histamine release was not affected (77.0 ± 28.7 vs.  73.3 ± 17.4%; baseline 7.3 ± 2.0%; control: 7.1 ± 1.1%; OBAA: 7.6 ± 1.6%;fig. 6).

For the PLC inhibitor U-73122 we found a similar pattern as for the PLA²inhibitor: It abolished vancomycin-induced histamine release but not calcium ionophore A23187–induced histamine release. For the vancomycin group, data were as follows: baseline: 5.2 ± 3.4%; control: 5.5 ± 2.8%; U-73122: 8.9 ± 3.8%; U-73122 with vancomycin: 9.3 ± 3.8%; and vancomycin: 35.3 ± 18.7%. For the calcium ionophore A23187 group, values were as follows: baseline: 7.3 ± 3.0%; control: 5.7 ± 2.3%; U-73122: 8.2 ± 3.4%; U73122 and A23187: 80.4 ± 18.9%; and A23187: 77.3 ± 19.4%. The groups were compared using analysis of variance, and we found a significant release for vancomycin alone, for U73122 with A23187, and for A23187 (P = 0.05), compared with the baseline and the control (fig. 7).

We found that tryptase is released in addition to histamine during nonimmunologic processes by known mast cell degranulators. Also, vancomycin, a prototypical mast cell degranulator, appears to be partially dependent on calcium and releases histamine and tryptase via  PLC- and PLA²-dependent mechanisms. This may be a result of either direct action on PLC and PLA²by vancomycin or stimulation of an unknown receptor mechanism. Our finding of significant tryptase release by vancomycin differs from the result described in a recent publication by Renz et al.  18Those authors looked at plasma levels of tryptase and histamine in patients receiving rapid vancomycin infusion and saw increases in plasma histamine levels only. The difference in the results may be caused by either direct activation of mast cells in vitro  or the fact that only in vivo  basophil activation occurred. Conversely, mast cell degranulation induced by calcium ionophore A23187, which causes Ca²⁺influx and releases Ca²⁺from intracellular sites, such as the sarcoplasmic reticulum, 19,20was not inhibited by PLC and PLA²inhibitors or decreased in the absence of extracellular calcium. This extracellular calcium independence on the histamine release by A23187 is opposite to the response seen in the dispersed mast cells from human lungs. Church et al.  21reported that histamine release induced by the calcium ionophore A23187 was decreased by more than 90% if calcium was omitted from the incubation medium. Because mast cells are functionally heterogeneous and possess different characteristics, 7,22previously described pathways from animal experiments and nonconnective tissue mast cells may not be valid for human cutaneous infant mast cells, the cells used in this study.

Our mast cell degranulation results are in accordance with other studies. Levy et al.  3showed that vancomycin induced 22.9% histamine release. Benyon et al.  23showed that calcium ionophore A23187 released 28.6% histamine. Stellato et al.  5showed a significant positive correlation between the percentages of histamine and tryptase release induced by morphine from human skin mast cells, and the release was calcium dependent, and, as in our study, no release was observed if calcium was removed from the incubation buffer. 13Atracurium is a known histamine releaser, both in vivo  24and in vitro  (net histamine release 12% with 10⁻³M), 6although another study reports that atracurium causes modest histamine release. 25We also found only a modest release of histamine and tryptase after incubation with atracurium. Tryptase has been shown to be released with histamine from human heart and synovial mast cells, if stimulated immunologically in vitro , 26,27and others have shown that it is released with histamine after an immunoglobulin (Ig) E–mediated event in vivo . 28,29Stellato et al.  5showed a positive correlation between the percentage of histamine and tryptase release induced by buprenorphine or by anti-IgE from human lung mast cells. Watkins 30previously defined the nonimmune mechanisms as absence of plasma tryptase but presence of increased urinary methylhistamine excretion. According to a study by Fisher and Baldo, 31increased concentrations of mast cell tryptase are not associated invariably with immunologic reactions. In a recent study, Edston and van Hage-Hamsten 32found elevated levels of tryptase in heroin addicts who died suddenly after injection, indicating a possible mast cell degranulation. There was no correlation between total IgE and tryptase levels, suggesting that heroin and morphine may have activated mast cells nonimmunologically. 32Data from our own in  vitro  study shows that tryptase also can be released, along with histamine, by nonimmunologic mast cell activation.

There was marked variability in histamine release from different foreskin preparations (all prepared using the same cell-dispersion protocol), suggesting that the ability to release influences the effect of the drugs on cutaneous mast cells. Furthermore, the number of mast cells isolated per gram of skin varied markedly, which could be another factor that influences drug sensitivity or other nonimmunologic mast cell activators.

The mechanisms for nonimmunologic mast cell activation are not well-described but suggest a dose-dependent and reproducible noncytotoxic reaction. 33Some authors describe the reaction in vitro  to be independent of extracellular calcium 1; others describe it as partially calcium-independent 23,34; and still others find it to be dependent on extracellular calcium, with maximum release if incubated with 3-mM extracellular calcium. 4,35–36,37Our studies with neonatal, partially purified, mast cells showed that the release was partially dependent on extracellular calcium, except in the case of the calcium ionophore A23187, which can release calcium from intracellular stores. 19,20 

Studies with peritoneal rat mast cells show that PLA²inhibitors can suppress histamine release but not prostaglandin release, if challenged with IgE or nonimmunochemical stimuli, such as calcium ionophore A23187, concanavalin A, compound 48/80, and substance P. 38Murakami et al.  39also showed that if rat mast cells are exposed to type II PLA², a significant release of histamine is observed. Our studies with the human cutaneous mast cell showed decreased histamine release if PLA²or PLC was inhibited (figs. 6 and 7). Elucidating pathways responsible for the histamine and tryptase release requires further studies to obtain a better understanding of in vivo  and in vitro  mast cell activation.

Laroche et al.  40measured concentrations of plasma histamine and tryptase after exposure to radiocontrast media in 20 patients with mild to severe reactions. Increased plasma histamine levels developed in 14 patients and increased tryptase levels developed in 16. 40Histamine and tryptase concentrations correlated with the severity of the reaction. The authors also reported the presence of IgE against the radiocontrast agents, although radiocontrast reactions traditionally have been considered to be nonimmunologic anaphylactoid reactions. Tests to determine IgE antibodies against radiocontrast agents are not validated; the authors did not perform inhibition studies to further validate their assay; and radiocontrast reactions occurred without previous exposure, suggesting that the reactions they noted were anaphylactoid.

Based on our in vitro  study, we conclude that tryptase is released with histamine during nonimmunologic mast cell activation and also can be an indicator for anaphylactoid reactions because it is only present in the cytoplasmic granules of mast cells. 41The mechanism of nonimmunologic histamine and tryptase release relies on membrane associated enzymes, such as PLA²and PLC. Although the physiologic role of tryptase is not clear, it is a neutral protease that can both activate and inactivate multiple inflammatory mediators. Further elucidation of direct molecular effects of mast cell activation may be important in understanding mechanisms of anaphylactoid reactions and other pathologic states of mast cell activation.

The authors thank Dr. Maria E. Arias and the staff at Georgia Baptist Hospital, Atlanta, Georgia, and Dr. Raphael Gershon and the staff at Grady Memorial Hospital, Atlanta, Georgia, for collecting the foreskins.