(Hypertension. 1999;34:997-1001.)
© 1999 American Heart Association, Inc.
Scientific Contributions |
Influence of Verapamil and Diclofenac on Leukocyte Migration in Rats
From the Department of Pharmacology, Institute of Biomedical Science, University of São Paulo, São Paulo, Brazil.
Correspondence to Zuleica Bruno Fortes, Departamento de Farmacologia, Instituto de Ciências Biomédicas, Universidade de São Paulo, Ave Prof Lineu Prestes, 1524, Cidade Universitária, CEP:05508-900, São Paulo, Brasil. E-mail zbfortes{at}icb.usp.br
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract—Nonsteroidal anti-inflammatory drugs and calcium channel blockers can reduce inflammatory responses. Leukocytes play an important role in these responses. An increased expression of adhesion molecules may increase leukocyte migration. Verapamil and diclofenac are known to reduce leukocyte-endothelium interaction. To investigate a possible synergism between these drugs that could be beneficial in cardiovascular diseases, we studied leukocyte behavior by using intravital microscopy. Venules of the spermatic fascia of anesthetized Wistar rats were observed with a closed-circuit TV coupled to an optical microscope. The number of leukocytes rolling along the venular endothelium ("rollers"), sticking after application of a stimulus such as leukotriene B4 or zymozan-activated plasma ("stickers"), or migrating after a carrageenan stimulus was reduced by verapamil at the dose of 10 mg/kg IP and by diclofenac at the dose of 2.5 mg/kg IP. The combination of both did not augment the effect of each agent alone. Verapamil, diclofenac, or their combination did not interfere with vessel diameter, number of circulating leukocytes, blood pressure levels, or heart rate. Verapamil alone or together with diclofenac reduced venular blood flow velocity and in consequence, the venular shear rate. Our data allow us to suggest that these drugs might interfere with the expression of adhesion cell molecules to reduce cell migration in inflammation. The lack of synergism between the drugs might be explained by the reduction in venular shear rate induced by verapamil, which might not be sufficient to hinder the effect of verapamil alone but hindered the summation effects of both.
Key Words: anti-inflammatory agents, nonsteroidal • verapamil • cell adhesion molecules • endothelium • cell movement • leukocytes • microcirculation
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Leukocyte extravasation is essential in the inflammatory response and can be divided into 3 steps: initial interaction of leukocytes with the activated endothelium (rolling), leukocyte activation with firm adhesion to endothelial cells (sticking), and leukocyte extravasation into the surrounding tissues.1 2 Several adhesion molecules are involved in the interaction between leukocytes and vascular endothelial cells, which include E-selectin, vascular cell adhesion molecule-1 (VCAM-1), and intercellular adhesion molecule-1 (ICAM-1).3 An increased expression of adhesion molecules may increase leukocyte migration and contribute to end-organ damage in cardiovascular diseases.3 Calcium channel blockers4 and nonsteroidal anti-inflammatory drugs5 are known to reduce the expression of several adhesion molecules and reduce leukocyte-endothelium interaction.
Diclofenac, a nonsteroidal anti-inflammatory drug, reduces inflammation, swelling, and arthritic pain by inhibiting the production of prostaglandins.6 7 8 Diclofenac affects polymorphonuclear leukocyte function in vitro, reducing chemotaxis, superoxide radical generation, and neutral protease production.9 Diclofenac also reduces the expression of several adhesion molecules such as L-selectin (ELAM),5 E-selectin, ICAM-1, and VCAM-1.10
Verapamil, a calcium channel blocker, acts on the cardiovascular system and has antianginal, antiarrhythmic, antihypertensive, and cardiovascular-protective effects.11 Verapamil has been reported to inhibit superoxide production in human neutrophils12 13 14 and to reduce the expression of adhesion molecules such VCAM-14 and ICAM-115 in endothelial cells.
An interaction between diclofenac and verapamil has been demonstrated. Verapamil enhanced the anti-inflammatory effect of diclofenac in vivo (paw edema) and potentiated the diclofenac inhibitory effect on the chemiluminescence response of isolated human polymorphonuclear leukocytes in vitro.16
To investigate a possible synergism between diclofenac and verapamil on leukocyte migration that could be beneficial in cardiovascular diseases, we studied leukocyte behavior with intravital microscopy.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Animals
Male Wistar rats that weighed 180 to 200 g were used. All rats were derived from breeding stock maintained at our institute and were randomized into 4 groups that were matched for age and weight, with at least 8 animals per group. The groups consisted of the following: untreated control rats; rats treated with verapamil 10 mg/kg IP; rats treated with diclofenac 2.5 mg/kg IP; and rats treated with both verapamil 10 mg/kg IP and diclofenac 2.5 mg/kg IP. All animal procedures were in accordance with the Ethics Committee of the Institute of Biomedical Sciences, University of São Paulo.
Leukocyte Counts
Leukocyte counts were performed on blood samples collected at
the time of the animals were killed. Total leukocyte counts were made
in a Neubauer chamber. Stained blood films were used for differential
leukocyte counts. Blood samples for these measurements were collected
from the abdominal aorta while the rats were under
anesthesia.
Direct Vital Microscopy of the Microcirculation: Surgical
Preparation
The animals were anesthetized with an injection of 40
mg/kg IP sodium pentobarbital, and the internal spermatic fascia of the
wall of the scrotal chamber was exteriorized for microscopic
examination in situ. This was done through a longitudinal incision of
the skin and dartos muscle in the midline over the ventral aspect of
the scrotum and opening of the cremaster muscle to expose the internal
fascia. This procedure does not require extensive surgical manipulation
for observation of the vascular network and provides a valuable means
for transilluminating a tissue for quantitative studies of the
microcirculation. In addition, the preparation is not affected by
respiratory movements of the animal, and its microcirculatory
characteristics remain basically invariant throughout the course of the
experiment. The animals were maintained on a special board
thermostatically controlled at 37°C, which included a transparent
platform on which the tissue to be transilluminated was placed. The
preparation was kept moist and warmed by irrigating the tissue with
warmed (37°C) Locke-Ringer's solution, pH 7.20 to 7.40,
containing 1% gelatin. The composition of the solution (in
mmol/L) was as follows: 154 NaCl, 5.6 KCl, 2
CaCl2 · 2H2O, 6
NaHCO3, and 5 glucose. The rate of outflow of the
solution onto the exposed tissue was controlled to maintain the
preparation in continuous contact with a thin film of liquid. A
television camera (500-line, Samsung Digital, SHC 410 NAD
Aerospace Samsung Industries Ltd, Korea) was incorporated with a
trinocular microscope (model 420, Reichert Diastar, Cambridge
Instruments Inc, Buffalo, NY) to facilitate observation of the enlarged
image (x2500) on the video screen (model KV2173S, Sony Trinitron,
Brazil). Images were recorded on a video recorder (model M-X41
M, Toshiba Recorder, Brazil) with a x40 longitudinal distance
objective with a 0.65 numerical aperture. Measurements of vessel
diameter were realized through an image-shearing monitor (model 908,
PTM, San Diego, Calif) incorporated into the system. Vessels selected
for study were postcapillary venules, and their diameters ranged from
12 to 16 µm. In another series of experiments, the left carotid
artery of each anesthetized (sodium pentobarbital, 40 mg/kg IP)
rat was catheterized, and mean arterial blood pressure and
heart rate were measured. The catheter was filled with heparinized
saline (20 IU/mL). Direct blood pressure recordings were
obtained by connecting the arterial cannula to a
physiograph (MK-III, Narco Bio System, Houston, Tex). Indirect heart
rate recordings were obtained by counting wave forms generated
on the physiograph tracings. Centerline red blood cell velocity was
measured with an optical Doppler velocimeter
(Microcirculation Research Institute, Texas A&M University, College
Station, Tex) that was calibrated against a rotating glass disk coated
with red blood cells. Arteriolar and venular blood flows were
calculated from the product of mean red blood cell velocity
(Vmean=centerline velocity/1.6) and microvascular
cross-sectional area, with cylindrical geometry assumed. Arteriolar and
venular shear rates (
) were calculated from the newtonian
definition:
=8(Vmean/Dv).17 18
Experimental Protocols
Leukocyte Rolling
In a series of experiments, interaction of leukocytes with the
luminal surface of the venular endothelium was studied
in a segment of the vessel. Leukocytes moving in the periphery of the
axial stream in contact with the endothelium were
considered to be "rollers,"19 and their number was
determined in 10-minute periods. These leukocytes moved slow enough to
be individually visible and were counted as they rolled past a
100-µm-length venule.20
Chemoattractant-Induced Leukocyte Adhesion
Leukocytes adhering to the endothelium were
quantified after the application of an irritant stimulus such as
leukotriene B4
(LTB4) or zymozan-activated plasma (ZAP).
A leukocyte was considered to be adherent to the venular
endothelium if it remained stationary for >30
seconds.21 Adherent cells ("stickers") were expressed
as the number per 100-µm-length of venule. Adhesion was investigated
under 2 conditions. In 1, the internal spermatic fascia, after a
suitable control period of normal circulation, was exposed to 0.1 mL of
a solution containing 10% homologous ZAP in
physiological saline. To obtain activated
plasma, zymosan, an enzyme from Saccharomyces cerevisiae,
was incubated (1 mg/mL) with plasma from normal animals for 1 hour at
37°C. After centrifugation at 1600g for 10
minutes, the supernatant fraction ZAP was collected and diluted 1:10
with physiological saline and topically added to
the preparation. Adhesion of leukocytes was assessed after 10 minutes
of addition of zymosan. Plasma treated identically, except for the
addition of zymosan, was used as a control. Leukocyte adhesion was also
quantitated with the same protocol in animals given a local application
of LTB4 (1 ng/mL, 0.1 mL). Each section of the
vascular bed was tested only once, and no more than 2 determinations
were performed on a single animal. The 2 measurements were averaged for
each animal.
Carrageenan-Induced Leukocyte Transmigration
In another series of experiments, the number of leukocytes that
accumulated in a 2500-µm2 standard area of
connective tissue adjacent to a postcapillary venule was determined
after the induction of a local inflammatory response. Cells were
counted on the recorded image. Five different fields were evaluated
for each animal to avoid variability on the basis of sampling. Data
were then averaged for each animal. The inflammatory reaction was
evoked by injecting 100 µg of carrageenan in 0.1 mL of saline into
the scrotum of the animals, and the number of migrated cells was
counted after 2 hours of carrageenan injection. At the end of the
experiments, the preparations were stained with toluidine blue for 15
minutes to check mast cell degranulation.
Drugs and Reagents
The following reagents were used: LTB4,
zymosan, carrageenan (all from Sigma Chemical Co, St. Louis, Mo);
diclofenac, potassium salt (Cataflan-Geigy); verapamil
(Dilacoron–Knoll); sodium pentobarbital (Hypnol–Cristália);
heparin (Liquemine–Roche S/A, RJ, Brazil); toluidine blue (ECIBRA
Brazil S/A); and NaCl, KCl, CaCl2 ·
2H2O, NaHCO3, and glucose
(all from Merck S/A, RJ, Brazil).
Statistical Analysis
Data are given as mean±SEM. One-way ANOVA followed by the
Tukey-Kramer multiple comparisons test and Student's t test
were used, when pertinent. The minimum acceptable level of significance
was P at a value 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Leukocyte Counts
Animals treated with verapamil, diclofenac, or a combination of both showed similar total and differential leukocyte counts relative to control animals (Table 1).
|
Direct Vital Microscopy of the Microcirculation
The number of leukocytes rolling along the venular
endothelium (rollers), sticking after stimulus with
LTB4 or ZAP (stickers), and migrating after
carrageenan stimulation was reduced by verapamil at the
dose of 10 mg/kg IP, diclofenac at the dose of 2.5 mg/kg IP, or the
combination of verapamil (10 mg/kg IP) and diclofenac (2.5
mg/kg IP) in comparison with control rats (Table 2). The drug combination did not augment
the effect of each agent alone (Table 2). Neither treatment
tested interfered with vessel diameters (Table 2).
|
Mean Arterial Blood Pressure, Heart Rate, Blood Flow
Velocity, and Wall Shear Rate
Under baseline conditions, untreated and treated animals had
similar arterial blood pressure levels and heart rates
(Table 3). Arterioles of treated animals
had a similar centerline red blood cell velocity (Table 4) and wall shear rate relative to
untreated controls (data not shown). However, in venules,
verapamil alone or in combination with diclofenac reduced
blood flow velocity and consequently the venular wall shear rate when
compared with control animals (Table 4). On the other hand, in
venules no differences were observed between diclofenac-treated and
untreated control animals (Table 4).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Postcapillary venules in particular were chosen for observation of leukocyte-endothelial interactions because they are considered to be the major site for leukocyte adhesion to the vascular wall in response to noxious stimuli. A reduced number of leukocytes rolling along the venular endothelium, sticking, and migrating was observed in rats treated with verapamil, diclofenac, and the combination of both. The findings were not paralleled by any significant decrease in the number of circulating leukocytes, thereby suggesting that adhesion changes were induced by the treatments affecting the leukocyte-endothelial interaction.
Adhesion of leukocytes induced by both LTB4 and ZAP was reduced by the different treatments tested. No differences were observed between the effects of the drugs on the magnitude of the reduction. During an acute inflammatory reaction induced locally by the injection of carrageenan, animals treated with verapamil, diclofenac, or the combination of both exhibited a similar pattern: only a few cells accumulated in the perivascular tissue. In controls, however, cells emerged massively into the connective tissue around the vessel under the influence of the inflammatory stimulus.
It is well known that during the inflammatory process, leukocytes are rapidly transported via the circulatory system to areas of tissue injury, in which they adhere to the endothelium and emigrate to the perivascular space.21 These events depend on the interaction between hemodynamic parameters (flow and resistance) that affect the transport of leukocytes to an injury site as well as a balance of forces to sweep them away.22 23 24 To evaluate the possible interference of hemodynamic changes on leukocyte behavior (rolling, sticking, and migration) studied, we measured arterial blood pressure (to estimate vascular resistance), blood flow velocity (to estimate blood volume), and venular diameter. We also calculated the wall shear rate, because the dependence of leukocyte adhesion on shear rate has been demonstrated in vivo25 26 27 and in vitro.28 Low shear rates promote leukocyte adherence to the microvascular endothelium in postcapillary venules.25 Treatment of the animals with verapamil, diclofenac, and the combination of both did not interfere with the blood pressure levels or venular diameters of the animals, leading us to exclude the interference of these parameters on cell behavior. Diclofenac reduced cell migration without any interference on wall shear rate. On the other hand, verapamil and the combination of this agent with diclofenac reduced cell migration and the venular shear rate. Because reduction of venular shear rate should promote leukocyte adherence,27 the lack of synergism observed when verapamil and diclofenac were tested in combination might be explained by the verapamil-induced reduction of venular shear rate, which might not be sufficient to hinder its effect when tested alone but hindered the summation of effects of verapamil and diclofenac when together.
The mechanisms underlying leukocyte accumulation in a tissue depend on the interaction between the cells and the vascular endothelium. During the development of inflammatory responses, leukocytes roll along the lining endothelium of postcapillary venules and eventually become firmly attached to the vascular wall before migrating into tissues. Specific adhesion glycoproteins expressed on the surface of leukocytes and endothelial cells play a relevant role in the adhesion phenomenon.25 Members of the selectin family of cell adhesion molecules are thought to mediate leukocyte rolling along the walls of the microvasculature.26 Glycoproteins of the CD11/CD18 complex (ß2 integrins) expressed on leukocytes interact with ligands such as ICAM-1 on endothelial cells to mediate leukocyte adhesion and emigration.29 Blockade of adhesion molecules on leukocytes, endothelial cells, or both can effectively inhibit leukocyte adhesion. Therefore, another explanation for the abnormal leukocyte function observed after treatment with verapamil, diclofenac, and the combination of both might be an interference of the drugs used on the adhesion molecule expression in vivo. In fact, in in vitro studies, diclofenac reduced ELAM expression on the leukocyte surface5 and ICAM-1, VCAM-1, and E-selectin on endothelial cells.10 Similarly, verapamil inhibited VCAM-1 expression on endothelial cells.4 However, Hailer et al30 could not find any inhibition of ICAM-1, VCAM-1, and ELAM-1 on these cells. On the contrary, in higher concentrations, increased expression of ICAM-1 and ELAM-1 was found.30 Therefore, the lack of synergism between verapamil and diclofenac might be explained by an increased expression of these adhesion molecules that compensates for the reduction in expression induced by diclofenac. Studies are in progress to identify the adhesion molecules involved and altered by the treatments tested.
In conclusion, our data allow us to suggest that verapamil and diclofenac, by interfering with adhesion molecule expression, reduce leukocyte migration in vivo. The lack of synergism between the drugs might be explained by the reduction of venular shear rate induced by verapamil that might not be sufficient to hinder the effect of verapamil alone but hindered the summation of effect of both.
|
Acknowledgments |
|---|
L.L.M. is a recipient of a CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) scholarship. The authors wish to thank Luciana Santana for technical assistance.
Received May 10, 1999; first decision June 22, 1999; accepted July 2, 1999.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Butcher EC. Leukocyte endothelial cell migration: three (or more) steps to specificity and diversity. Cell. 1991;67:1033–1036.[Medline] [Order article via Infotrieve]
2. Shimizu Y, Newman W, Tanaka Y, Shaw S. Lymphocyte interactions with endothelial cells. Immunol Today. 1992;13:106–112.[Medline] [Order article via Infotrieve]
3.
Carlos TM, Harlan JM.
Leukocyte-endothelial adhesion molecules.
Blood. 1994;84:2068–2101.
4. Yamaguchi M, Suwa H, Miyasaka M, Kumada K. Selective inhibition of vascular cell adhesion molecule-1 expression by verapamil in human vascular endothelial cells. Transplantation. 1997;63:759–764.[Medline] [Order article via Infotrieve]
5. Díaz-González F, Gonzáles-Alvaro I, Campanero MR, Mollinedo F, Pozo MA, Muñoz C, Pivel JP, Sánchez-Madrid F. Prevention of in vitro neutrophil-endothelial attachment through shedding of L-selectin by nonsteroidal antiinflammatory drugs. J Clin Invest. 1995;95:1756–1765.
6. Brooks PM, Hill W, Geddes R. Diclofenac and ibuprofen in rheumatoid arthritis and osteoarthritis. Med J Aust. 1980;1:29–30.[Medline] [Order article via Infotrieve]
7. Todd PA, Sorkin EM. Diclofenac sodium, a reappraisal of its pharmacodynamic and pharmacokinetics properties, and therapeutic efficacy. Drugs. 1988;35:244–285.[Medline] [Order article via Infotrieve]
8. Skoutakis VA, Carter CA, Mickle TR, Smith VH, Arkin CR, Alissantratos J, Petty DE. Review of diclofenac and evaluation of its place in therapy as a non-steroidal anti-inflammatory agent. Drug Intell Clin Pharm. 1988;22:850–859.[Abstract]
9. Friman C, Johnston C, Chew C, Davis P. Effect of diclofenac sodium, tolfenamic acid and indomethacin on the production of superoxide induced by N-formyl-methionyl-leucyl-phenylalanine in normal human polymorphonuclear leukocytes. Scand J Rheumatol. 1986;15:41–46.[Medline] [Order article via Infotrieve]
10. Sakai A. Diclofenac inhibits endothelial cell adhesion molecule expression induced with lipopolysaccharide. Life Sci. 1996;58:2377–2387.[Medline] [Order article via Infotrieve]
11. Millard RW, Grupp G, Grupp IL, Di Salvo J, De Pover A, Schwartz A. Chronotropic, inotropic and vasodilator actions of diltiazem, nifedipine and verapamil: a comparative study of physiological responses and membrane receptor activity. Circ Res. 1983;52:129–139.
12. Simchowitz L, Spilberg I. Generation of superoxide radicals by human peripheral neutrophils activated by chemotactic factor: evidence for the role of calcium. J Lab Clin Med. 1979;93:583–593.[Medline] [Order article via Infotrieve]
13. Steiner RD, Allan P, William WB. Cytochalasin B facilitates the inhibition of human polymorphonuclear leukocyte generation of superoxide by verapamil. J Lab Clin Med.. 1984;103:1949–1958.
14. Irita K, Fujita I, Takeshige K, Minakami S, Yoshitake J. Calcium channel antagonists induced inhibition of superoxide production in human neutrophils: mechanisms independent of antagonizing calcium influx. Biochem Pharmacol. 1986;35:3465–3471.[Medline] [Order article via Infotrieve]
15. Haller H, Schaper D, Philipp S. LDL induces surface expression of adhesion molecules ICAM-1 and VCAM on endothelial cells via the activation of protein kinase C. J Am Soc Nephol. 1992;3:456.
16. Al-Tuwaljri A, Mustafa AA. Verapamil enhances the inhibitory effect of diclofenac on the chemiluminescence of human polymorphonuclear leukocytes and carrageenan-induced rat's paw oedema. J Immuno-pharmacol. 1992;14:83–91.
17. Davis MJ. Determination of volumetric flow in capillary tubes using an optical Doppler velocimeter. Microvasc Res. 1987;34:223–230.[Medline] [Order article via Infotrieve]
18.
Panés J, Kurose I, Rodriguez-Vaca MD, Anderson
DC, Miyasaka M, Tso P, Granger DN. Diabetes exacerbates inflammatory
responses to ischemia-reperfusion. Circulation. 1996;93:161–167.
19.
Dahlén SE, Björk J, Hedqvist P,
Arfors KE, Hammarström S, Lindgren JA, Samuelson B.
Leukotrienes promote plasma leakage and leukocyte adhesion
in postcapillary venules: in vivo effects with relevance to acute
inflammatory response. Proc Natl Acad Sci U S A. 1981;78:3887–3891.
20. Fortes ZB, Farsky SP, Oliveira MA, Garcia-Leme J. Direct vital microscopic study of defective leukocyte-endothelial interaction in diabetes mellitus. Diabetes. 1991;40:1267–1273.[Abstract]
21. Grant L. The sticking and emigration of white blood cells in inflammation. In: Zweifach BW, Grant L, McCluskey L, eds. The Inflammatory Process. Orlando, Fla: Academic Press; 1973;2:205–249.
22.
Atherton A, Born GVR. Quantitative investigations of
the adhesiveness of circulating polymorphonuclear leukocytes to
blood vessel walls. J Physiol. 1972;222:447–474.
23.
Schmid-Schoenbein GW, Fung YC, Zweifach BW. Vascular
endothelium-leukocyte interaction, sticking shear force
in venules. Circ Res. 1975;36:173–184.
24. Mayrovitz HN, Wiedeman MP, Tuma RF. Experimental and clinical-thrombosis: factors influencing leukocytes adherence in microvessels. Thromb Haemost. 1977;38:823–830.[Medline] [Order article via Infotrieve]
25. Perry MA, Granger DN. Role of CD11/CD18 in shear rate-dependent leukocyte-endothelial cell interactions in cat mesenteric venules. J Clin Invest. 1991;87:1798–1804.
26.
Ley K, Gaehtgens P, Fennie C, Singer MS, Lasky LA,
Rosen SD. Lectin-like cell adhesion molecule 1 mediates leukocyte
rolling in mesenteric venules in vivo. Blood. 1991;77:2553–2555.
27. Granger DN, Benoit JN, Suzuki M, Grisham MB. Leukocyte adherence to venular endothelium during ischemia-reperfusion. Am J Physiol. 1989;257:683–688.
28.
Lawrence MB, Smith CW, Eskin SG, McIntire LV. Effect of
venous shear stress on CD18-mediated neutrophil adhesion to cultured
endothelium. Blood. 1990;75:227–237.
29.
von Adrian UH, Chambers JD, McEvoy LM, Bargatze RF,
Arfors K, Butcher EC. Two-step model of
leukocyte-endothelial cell interaction in inflammation:
distinct roles for LECAM-1 and the leukocyte
ß2 integrins in vivo. Proc Natl Acad Sci
USA. 1991;88:7538–7542.
30. Hailer NP, Blaheta RA, Harder S, Scholz M, Encke A, Markus BH. Modulation of adhesion molecule expression on endothelial cells by verapamil and other calcium channel blockers. Immunobiology. 1992;191:38–51.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||