Modulation of Human Neutrophil Chemotactic Responses by Cyclic 3’,5’-guanosine Monophosphate and Cyclic 3’,5’-adenosine Monophosphate Harry

R. Hill, Richard

Nancy A. Hogan,

D. Estensen, and Nelson

Cyclic 3’,5’-guanosine monophosphate (cGMP) and cyclic 3’,5’-adenorine monophosphate (CAMP) and compounds known to effect the intracellular concentrations of these nucleotides were examined for their ability to effect human neutrophil responsiveness to chemotactic (PMN) stimulation. Incubation of neutrophils with agents recognized to promote increasesin intracellular CAMP in a variety of tissues (i.e., epinephrine, norepinephrine, isoproterenol, histamine, cholera toxin, and prostaglandin El and Es) or with CAMP inhibited the leukotactic response to a bacterial chemotactic factor. In contrast, cGMP and compounds which

Paul G. Quie,

D. Goldberg

have been shown to promote increases in (i.e., intracellular cGMP concentration carbamylcholine, phorbol acetylcholine, and prostaglandin myristate acetate, Fz,) markedly enhanced the neutrophil chemotactic response. The inhibitory or stimulatory influences on chemotactic responsiveness prowioted by several of the agents could be shown to be blocked by a specific pharmacologic antagonist of the particular compound tested. These data support the hypothesis that cGMP and CAMP can provide opposing regulatory influences on certain cellular functions; in this case, directed motility of leukocytes.

G

OLDBERG AND co-workers have recently suggested that cyclic 3’,5’guanosine monophosphate (cGMP)* and cyclic 3’5’-adenosine monophosphate (CAMP) serve as opposing regulatory effecters in a number of bidirectionally controlled cellular events. ‘J This concept of biologic regulation through opposing actions of the two cyclic nucleotides has been termed the “Yin-Ysing Hypothesis” of biologic control.’ The hypothesis predicts that in a variety of cell types, a function that would be modulated in one manner by an action of intracellular CAMP would be influenced in the opposite manner by an action of cGMP. In a preliminary report3 it was shown that two cholinergic agents and phorbol myristate acetate, which have been shown in other cell systems to increase cellular cGMP and cGMP itself, enhanced the chemotactic response of human neutrophils. The present report extends these original find*Abbreviations used in this paper: cGMP, cyclic guanosine monophosphate; CAMP, cyclic adenosine monophosphate; PMN, polymorphonuclear leukocyte. From the Departments of Pediatrics. Pharmacology. and Laboratory Medicine and Pathology. University of Minnesota Medical School, Minneapolis, Minn., and the Department of Pediatrics and Pathology, University of Utah, Salt Lake City, Ur. Receivedforpublication October 15.1974. Supported by VSPHS grants AI 08821, AI 06931, AI 08724, AM 18354-01, NS-05979, .and HL06314, Department of Pediatrics Training Grant HD &IO53-13. and a grant from the Twin Cities Diabetes Association. Part of these studies were conducted under the sponsorship of the Commission on Streptococcal and Staphylococcal Disease of the Armed Forces Epidemiological Board with support by the Medical Research and Development Command, U.S. Army, under Contract DADA17-70-C-0082. Reprint requests should be addressed to Harry R. Hill. Departments of Pediatrics and Pathology, Universiiy of Utah College of Medicine, Salt Lake City, Utah 84132. 0 1975 by Grune & Stratton, Inc. Metcrbolism, Vol. 24, No. 3 (March), 1975

447

HILL ET Al.

448

ings to demonstrate that other agents, which have recently been shown to be capable of increasing intracellular cGMP levels, also enhance the chemotactic response of neutrophils. In contrast, CAMP and a number of agents known to enhance its intracellular accumulation had an inhibitory influence on chemotactic responsiveness. Specific antagonists of certain of the compounds which are believed to act through an enhanced generation of CAMP or cGMP prevent the effect usually produced by the agonist. These blocking agents were not effective, however, in altering the effect of cGMP or CAMP on human neutrophil chemotactic responses. MATERIALS

AND METHODS

Leukocyte Function Tests Chemotaxis. The in vitro method used for studying chemotaxis was a modification of the procedures of Boyden. Leukocyte-rich plasma was obtained by allowing the erythrocytes in 10 ml of heparinized blood from normal adult donors to sediment over a l-hr period. The concentration of polymorphonuclear leukocytes (PMN) per millileter of leukocyte-rich plasma was then determined by quantitative and differential cell counts. Following this, 0.1 ml of suspension was diluted to 0.4 ml with tissue culture medium 199 (Microbial. Associates, Inc., Bethesda, Md.). The PMN in this suspension were then deposited on one side of a 5-r pore size Millipore filter (Millipore Corp., Bedford, Mass.) utilizing a cytocentrifuge (Shandon Southern Instruments Inc., Sewickley, Pa.). The filters were immediately placed in a modified Boyden chamber (Neuroprobe Corp., Bethesda, Md.), and a chemotactic stimulus was added to the attractant side. Tissue culture medium 199 with or without a predetermined concentration of the agent being tested was added to the top or cell side of the modified Boyden chamber. In a previous study it was shown that a brief 3%5-min exposure of leukocytes to cGMP altered chemotactic responsiveness. In the present study, the pharmacologic agents remained in contact with the cells during the 3-hr incubation period carried out at 37’C. The filters were then removed, fixed in methanol, stained with hematoxylin, dried in ethanol, and cleared with xylene. The number of cells that had migrated completely through the filter within ten random fields bounded by a photographic reticule was determined by visual counting (using a IO x ocular, 45 x objective and 5 x 5-mm photographic reticule). A chemotactic index was calculated by dividing the number of PMNs that had migrated completely through the filter within the reticule in ten random fields by the total number of PMNs (in millions) delivered to the starting side of the filter.4 As discussed by Keller and co-workers,’ the use of a short incubation time (3 hr) and less than a maximal chemotactic stimulus does not lead to a significant number of the neutrophils falling off of the Millipore filter into the fluid on the attractant side of the chamber. Quantitation of the cells in the fluid after 3 hr incubation revealed that between 0.1% and 1% of PMN delivered to the filter had fallen off. The number of cells in the attractant fluid did not vary significantly when CAMP or cGMP were present in the system. All chemotactic assays were conducted in duplicate or triplicate and on two to three separate occasions. The results are expressed as the mean and standard deviation of such sampling. Chemotactic facror. A bacterial chemotactic factor was prepared from a culture filtrate of E. co/i grown in medium 199 for 24 hr at 37’C6 After passage through a 0.22-p pore size Millipore filter, the bacterial chemotactic factor (BF) was frozen at -70°C in I-ml aliquots. Each day an aliquot of frozen BF was thawed and diluted in medium 199 so that each millileter of the solution contained 50 pi of the factor. Two millileters of the solution were added to the bottom or attractant side of the chemotactic chamber. The bacterial chemotactic factor was selected because it produces very reproducible effects and can be used to provide less than a maximal stimulus to the PMN. Other physiologically active materials, such as complementassociated chemotactic factors, were not used in this study, since Becker’ has indicated that they probably all have similar mechanisms of action and the present studies were intended to study the modulation of chemotaxis and not its initiation. Random mobility. The effects of exogenous cyclic nucleotides on neutrophil random mobility

MODULATION

OF HUMAN

NEUTROPHIL

CHEMOTACTIC

449

RESPONSES

was determined by a modification of the method of McCall and co-workers. 478Glass capillary tubes were filled with a solution containing I x IO’ PMN/ml. After centrifugation, the tubes were cut at the cell-fluid interface and placed in Sykes-Moore Chambers (Bellco Glass, Inc., Vineland, N.J.). The chambers were then filled with a predetermined concentration of the agent being tested diluted in tissue culture medium and 10% fetal calf serum. After incubation for 4 hr, the leading edge of migrating granulocytes was mapped by projection microscopy. The effect of the cyclic nucleotides on random mobility as measured by the Boyden chamber technique without a chemotactic gradient was also determined.

Materials Cyclic nucleotides, atropine, histamine, adrenergic, and cholinergic agents were obtained from Sigma Chemical Co., St. Louis, MO.; and propranolol from Ayerst Laboratories, New York. Cholera enterotoxin was supplied by Dr. R. A. Finkelstein, The University of Texas Southwestern Medical School, Dallas, Texas; burimamide by Dr. Paul Black, Smith Kline and French, England. Prostaglandins were obtained from Merck and Company, Rahway, N.J., and phorhol myristate acetate from Consolidated Midland, Brewster, N.Y. RESULTS

Eflect of Adrenergic Agents on Neutrophil Chemotactic Responses

The catecholamines epinephrine, norepinephrine, and isoproterenol, which promote adrenergic effects of the beta type, when tested as effecters of PMN function, all produced a suppressive effect on the chemotactic response in concentrations of 1O-5-1O-3 A4 (Fig. 1). In contrast, the a-adrenergic agent, phenylephrine, produced the opposite effect, i.e., a marked enhancement of the neutrophil response to the bacterial chemotactic factor. Maximal stimulation of almost 300% was obtained with 10e4 phenylephrine, the intermediate concentration tested.

T----

Fig. 1. Effect of adrenergic agents on neutrophil chemotactic responses to the bacterial chemotactic factor.

I

,

I

to-5 IO” 10-3 Epinaphrine

Norapinephrine

lsoproterenol

(Ml

(M)

CM)

Ptmylephrine (M)

450

HILL ET AL.

Table 1. Effect of Adrenergic Receptor Blockade on the Modulation of Chemotactic Responsesby Epinephrine or Phenylephrine Agent

BlockingAgent

Chemotoctic Index

Per Cent of Control

TC 199*

TC 199*

113

Epinephrinet

TC 199*

44

100 39

Epinephrine

Propranolol~

125

110

Epinephrine

Phentolaminet

40

35

Phenylephrinet

TC 199*

190

168

Phenylephrine

Phentolomine

125

110

*Tissue culture medium 199 alone. t Final concentration war 10m5 M.

.Eflect of Adrenergic

Blocking

Agents

In the presence of the beta-adrenergic blocking agent propranalol, epinephrine no longer exhibited a suppressive influence (Table 1). In contrast, the inhibitory effect of epinephrine was unaffected by the a-adrenergic blocking agent, phentolamine. Phentolamine, however, did exhibit an effect to partially block the stimulatory actions of phenylephrine on chemotactic responsiveness (Table 1). E#ect of Cholinergic

Agents

Acetylcholine in a concentration range of 10-7-10-5 M could be shown to stimulate chemotactic responsiveness in a dose-dependent fashion to over 200% of the control value (Fig. 2). Addition of 10m5M atropine to the leukocyte sus-

t 260 1

7

I I

t-

Acetyfchdiine (M)

Acetycholine +Atropine lwwvl)

Carbamylchdine

Carbamylcholine +Atropine IO’(M)

Fig. 2. Effect of choliner;jiti siitnubtl@h and blockode on neutrophil chemotoctic responses to the bocteriol ehemotactic factor.

MODULATION

OF HUMAN

NEUTROPHIL

CHEMOTACTIC

RESPONSES

T

260 .” t

220

E E

180

I

Fig. 3. Effect af prostaglandins on neutrophil chemotact(c twponses to the kcterial chemotoctic (actor.

I

I

wa lLi7 IO6 F,, (M) Prosfoglondin

pension 10 min prior to the addition of acetylcholine markedly attenuated the facilitory effect on chemotaxis produced by exposure of the cells to a&@ choline. Carbamylcholine had an effect similar to that of aeetylcheline in stimulating the ehemotactic responsive, and partial blockade of the response could be demonstrated by atropine with the combinations of antagonist and agonist employed (Fig. 2). E’ect of Prostaglandins The prostaglandins El, Ez, and Fr., were found to influence the leukotactic process in a rather selective manner (Fig. 3). Prostaglandin Ei had the effect of inhibiting chemotactic responsivgngsg at each of the concentrations tested (i.e., 10-*-10-6 M). P rostaglandin & had an inhibitory effect at 10V6 M, but did not significantly affect chemotaxjs at lower concentrations (i.e., 10e7 and lo-’ M). Prostaglandin Fza had p suppressive effect at 10e6 M which was the highest concentration used, but mar4edly enhanced chemotactic responsiveness when the concentration was reduced to 10e7 or 10e8 M. Efect

of Other Agents

Phorbol myristate acetate, a compound which has previously been shown to jncrease cellular cQMP in fibroblasts and platelets,’ was also shown to enhance the chemotactic response to neutrophils when introduced at concentrations of Q,lO~lO ng/ml (Fig. 4). Imidazole, a compound which exhibits cholinergic-like gffects in a number of systems, lo also had the effect of enhancing neutrophil chemotaxis. In contrast, histamine and cholera toxin, which have been shown to increase intracellular CAMP in leukocytes,“*‘2 markedly inhibited chemotactic activity, and this effect could be at least partially blocked with an antihistamine, burimamide, and cholera antitoxin, respectively. (Fig. 5) Cyclic Nucleotide Modulation of Chemotaxis

Leukocytes were incubated with CAMP or 5’-q&lP to determine the effectiveness and specificity of exogenous nucleotides to mimic the action of agents

HILL ET Al.

452

260

t

20

c I_

Fig. 4. Meet of phorbol myrirtote acetate and imidazole on neutrophil chemotactic responses to the bacterial chemotactic factor.

-Ll_i__

IO.’ 10-3IO,2

0 IO 1.0 IO nanograms per ml Phorbol Myristate Acetate

lmidazole

(M)

3oo 7

-l

c

I A-

60

L IO

Histamine (Ml

Effect of histamine Fig. 5. terial chemotactic factor.

6 lo-510.'

Histamine (ml + Burimamide IO-3(M)

I

IO

m?o~~~To~irn

and cholera toxin on neutrophil

Iol

1

I

I

IO

1

100

nanograms per ml Chdera Toxin +Antitoxin

chemotactic

responses to the bac-

MODULATION

OF HUMAN

NEUTROPHIL CHEMOTACTIC

RESPONSES

453

r

260

240

1

220 200 180 160 140 120 I

e

3

ti a

III! 10.7 B.6

,0-s K)”

CAMP(M)

Fig. 6. Effect of CAMP and chemotactic factor.

5’ AMP

I

IIII

lo-’ 10-6Ill‘~ IV

CAMP(M) + Propronolol IO-‘(M)

5’AMP(M)

on neutrophil

I

chemotactic

responses

to the

bacterial

whose suppressive effects on chemotactic activity are believed to be associated with their ability to elevate the cellular concentration of this cyclic nucleotide. As shown in Fig, 6, CAMP inhibited the chemotactic response at concentrations of 10S4 and lo-’ M. No significant inhibitory effect on chemotaxis was observed with concentrations of 5’-AMP in a similar range. The beta-adrenergic blocking agent propranolol was not effective in preventing the inhibitory action

7. Effect of cGMP Fig. and 5’ GMP on neutrophil chemotactic responses to the bacterial chemotactic factor.

-L1

IO-’D-6 IO’”

10-l 10-6 10-S cGMP(M)

S’GMP (M )

cGMP(M) +Atropine IO-? Ml

HILL ET At.

454

Table 2, Effect of cGMP op PMN Chemotoctic Bacterial Chamotactic Molar Concentration of cGMP

Response to the

Pactor MeanPer Centof Control Chemotoxis

1o-5

130 f 29”

1o-6

214 zk 26

lo-’

223 f

1o-8

263 f 56

1o-9

223 & 15

10 -10

97&

12

10

‘Standard deviation of three replicates.

of CAMP and, for reasons which are not explainable at this time, seemed to potentiate the effect of the cyclic nucleotide. The effects produced by cGMP were opposite to those that were brought about by CAMP; incubation of leukocytes with cGMP markedly enhanced the chemotactic response (Fig. 7) while 5’-GMP had no significant effect. The effect of cGMP to enhance chemotactic activity was not prevented by addition of atropine to the incubation solution. As shown in Table 2, significant stimulation of chemotactic responsiveness was observed with a cGMP concentration as low as 1O-9 M. Particularly noteworthy is the observation that the stimulatory effects of cGMP are greatest between 10m9 and lo-’ but minimal at 10-5 M. DISCUSSION

The results of the present experiments support the concept that human neutrophil responsiveness to chemotactic stimulation may be modulated by neurohumores and other biologically active agents whose regulatory influences may be mediated in part through cellular actions of cGMP and CAMP. One group of substances, cholinergic agents, cY-adrenergic agents, and phorbol myristate acetate promote an enhancement of the leukotactic response, an action which is also produced by cGMP. The other group, ,f3-adrenergic agents, cholera toxin, and histamine inhibit the process, and this effect can also be produced with CAMP. The latter results agree with those of othersI who have found that agents which increase cellular CAMP concentration inhibit chemotactic responsiveness of PMNs. The fact that both the stimulatory and inhibitory modulating effects can be produced by cyclic GMP and cyclic AMP, respectively, coupled with the results with agonists and specific antagonists, would argue in favor of the cyclic nucleotides acting at a step beyond that involving hormone-receptor interaction. The effect of exogenous cyclic nucleotides to mimic the action of the agents thought to promote their selective accumulation in cells is worthy of additional consideration. In earlier investigations dealing with the role that cGMP may play as a biologic regulator, the addition of exogenous cGMP was found to produce either no effect or to have an effect similar to that of cAMP’*‘~‘~ Only recently have effects opposite to those promoted by CAMP been observed following exposure of intact cells to cGMP. Such a response is, in fact, more consistent with changes that most often occur in the cellular levels of cGMP in association with the response of a cell to a number of biologically active sub-

MODULATION

OF HUMAN

NEUTROPHIL

CHEMOTACTIC

RESPONSES

455

stances that promote events opposite to those that occur in association with an elevation of cellular cAMP.‘*14 Upon inspection of the present results, it is clear that the concentration range in which cGMP is most effective in enhancing chemotactic activity is 10-9-10-6 M, while at higher concentrations the effect is diminished. It would seem, therefore, that the effectiveness of cGMP may be confined to a rather narrow, discrete concentration range below or beyond which little effect may be observed. A similar phenomenon has been uncovered with regard to the stimulatory effect cGMP has on nuclear RNA synthesis.* The characteristics of a discrete and low-concentration effectiveness of cGMP seems also to be related to the effectiveness (i.e., chemotactic facilitation) of agents which presumably promote cellular cGMP accumulation. Over the concentration range tested, intermediate or low levels of such compounds as phenylephrine and prostaglandin FZa produced maximal effects. In the highest concentration tested (10m6 M), prostaglandin FZa actually produced an effect opposite to that seen with the two lower concentrations. The reason for this behavior with cyclic GMP and agents whose actions seem to be associated with cellular cyclic GMP accumulation is not explainable at this time, but the phenomenon points out what would constitute a major difference between the latter group of agents (and cyclic GMP) and those in which cyclic AMP has been implicated as the effector molecule; the concentration-respopse curves in this case are usually sigmoidal. In our earlier studies3 of leukotaxis, we employed short (i.e., 55lo-min) pulse exposure of PMNs to cyclic nucleotides, hormones, or other agents before the cells were placed in the assay system to test leukotactic responsiveness. In the present study the cells were exposed to the test agents continuously (i.e., even during the chemotactic testing procedure). Because the same results were obtained with both procedures, it would seem that the modulating influence on chemotaxis of the cyclic nucleotides and the other substances tested is imposed during a relatively short period of time and that the regulatory effect is not magnified with time of exposure. While the modulation of chemotaxis in leukocytes seems to fit the Yin-Yang model of biologic control,’ there is no understanding as yet of the role, if any, that cyclic nucleotides may play in initiating chemotaxis. Preliminary experiments have shown that cyclic GMP alone on the attractant side of the chamber at concentrations effective in modulating chemotaxis does not serve as a chemotactic attractant. This observation in conjunction with the negative results with cyclic nucleotides on random motility reported here suggests that exposure to cyclic GMP alone does not serve to initiate leukotaxis. Another function of PMNs, degranulation, has also been shown to be modulated by opposing influences of these two cyclic nucleotides”*“j; cyclic GMP has been shown to promote and cyclic AMP to suppress lysosomal enzyme release in PMNs. REFERENCES 1. Goldberg ND, Haddox MK, Hartle DK, Hadden JW: The biological role of cyclic 3’5’guanosine monophosphate, in Proceedings of

the Fifth International Congress on Pharmacology, Basel, Karger, 5:146-169, 1973 2. Goldberg ND, Haddox MK, Dunham E,

456

Lopez C, Hadden JW: The yin yang hypothesis of biological control: opposing influences of cyclic GMP and cyclic AMP in the regulation of cell proliferation and other biological processes, in Clarkson B, Beserga R (eds): The Cold Spring Harbor Symposium on the Regulation of Proliferation in Animal Cells. New York, Academic Press, 1974, pp 609-625 3. Estensen RD. Hill HR. Quie PG. Hogan N, Goldberg ND: Cyclic GMP and cell movement. Nature 245:458-460, 1973 4. Hill HR. Gerrard JM. Hogan NA, Quie PG: Hyperactivity of neutrophil leukotactic response during active bacterial infection. J Clin Invest 53:996-1002, 1974 5. Keller HV, Bore1 JF, Wilkinson PC, Hess MW, Cottier H: Reassessment of Boyden’s technique for measuring chemotaxis. J Immunol Methods 1:165-168, 1972 6. Ward PA, Lepow IH, Newman LJ: Bacterial factors chemotactic for polymorphonuclear leukocytes. Am J Pathol 52:725-736, 1968 7. Becker EL: The relationship of the chemotactic behavior of the complement derived factors, Cs,, Csa, and Cs67, and a bacterial chemotactic factor to their ability to activate the proesterase I of rabbit polymorphonuclear leukocytes. J Exp Med 135:376-387, 1972 8. McCall CE, Caves J, Cooper R, DeChatelet L: Functional characteristics of human toxic neutrophils. J Infect Dis 124:68-75, 1971 9. Estensen RD. Hadden JW, Hadden EM, Touraine F, Touraine JL, Haddox MK, Goldberg ND: Phorbol myristate acetate: Effects of a tumor promotor on intracellular cyclic GMP in mouse fibroblasts and as a mitogen on

HILL ET At.

human lymphocytes, in Clarkson B, Beserga R (eds): Cold Spring Harbor Symposium on the Regulation of Proliferation in Animal Cells. New York, Academic Press, 1974. pp 6277634 10. Goldberg ND, Lust WD, O’Dea RF, Wei S, O’TooIe AG: A role of cyclic nucleotides in brain metabolism, in Costa E, Greengard P (eds): Advances in Biochemical Psychopharmacology, vol. 3. New York, Raven Press, 1970, pp 67.-74 II. Bourne HR. Melmon KL. Lichtenstein L: Histamine augments leukocyte adenosine antigenic 3’.5’-monophosphate and blocks histamine release. Science 173:743-745, 197 I 12. Bourne HR. Lehrer RI, Lichtenstein LM. Weissmann G, Zurier R: Effects of cholera enterotoxin on adenosine 3’,5’-monophosphate and neutrophil function. J Clin Invest 52:698708, 1973 13. Rivkin I. Becker EL: Possible implication of cyclic 3’5’-adenosine monophosphate in the chemotaxis of rabbit peritoneal polymorphonuclear leukocytes. Fed Proc 3 1:657, I972 14. Goldberg ND, O’Dea RF. Haddox MK: Cyclic GMP, in Greengard P, Robinson GA (eds): Advances in Cyclic Nucleotide Research, vol. 3. New York, Raven Press, 1973, pp 155223 15. Ignarro LJ. George WJ: Hormonal control of lysosomal enzyme release from human neutrophils: elevation of cyclic nucleotide levels by autonomic neurohormones. Proc Natl Acad Sci USA 7 I :20277203 I. 1974 16. Zurier RB, Hoffstein S, Weissmann G: Mechanisms of lysosomal enzyme release from human leukocytes. I. Effect of cyclic nucleotides and colchicine. J Cell Biol 58:27-41, 1973