Human Lymphocyte Metabolism. Effects of

Cyclic and Noncyclic Nucleotides

on- Stimulation by Phytohemagglutinin

JAY W. SMITH, ALTONL. STEINER, and CHARLESW. PARKER

From the Washington University School of Medicine, Department of Medicine,St. Louis, Missouri 63110

ABSTRACT The effects of extracellular nucleotidesand agents which elevate intracellular cyclic adenosine3',5'-monophosphate (cyclic AMP) concentrations onhuman lymphocyte metabolism have been studied. Ami-nophylline, isoproterenol, and prostaglandins, all of whichelevate lymphocyte cyclic AMP levels, inhibited incor-poration of 'H-labeled thymidine, uridine, and leucineinto the DNA, RNA, and protein of phytohemagglutinin(PHA)-stimulated lymphocytes. Aminophylline inhibi-tion was maximal only when the inhibitor was addedwithin 1 hr after exposure of cells to PHA, suggestingthat a relatively early step in the lymphocyte transforma-tion process may be affected.

The addition of various nucleotides to the culturemedium also inhibited incorporation of labeled precur-sors. The best inhibitor, dibutyryl cyclic AMP (DU cy-clic AMP), produced maximal inhibition only if presentduring the 1st hr after initial exposure to PHA. Amongthe various cyclic nucleotides derivatives of guanosineand adenine were the most effective inhibitors (sub-stantial inhibition at 0.1 mmconcentrations). However,the inhibition was not specific for nucleotides containingthe cyclic phosphodiester moiety since the tri-, di-, andmonophosphates of adenosine and guanosine were equallyeffective in diminishing thymidine uptake. The aboveinhibitions were not due to secondary effects of the in-hibitors on the interaction of PHA with lymphocytes asjudged by "25I-labeled PHAbinding studies.

Low concentrations (1-10 imoles/liter) of cyclicAMP produced slight stimulation of thymidine-BHuptake in resting lymphocytes (lymphocytes not stimu-lated with PHA). However, the effects were quite small

These studies have been presented in part to the 53rdMeeting of the Federation of American Societies for Ex-perimental Biology, April 1969 (Fed. Proc. 28: 566).

Received for publication 30 April 1970 and in revised form24 September 1970.

as compared with those produced by PHA itself. At-tempts to demonstrate increased thymidine uptake 48 hrafter pulsing lymphocytes with aminophylline or isopro-terenol were unsuccessful. The relationship of these ob-servations to a possible regulatory role for cyclic AMPin PHA-stimulated lymphocytes is discussed.

INTRODUCTIONIn the preceding paper we have presented evidence thatphytohemagglutinin (PHA) produces a defined se-quence of changes in the adenyl cyclase activity andcyclic adenosine 3',5'-monophosphate (cyclic AMP) con-centration of human peripheral blood lymphocytes.Changes in cyclic AMPconcentration occur very earlyraising the possibility that cyclic AMP fulfills the roleof a secondary messenger for PHA in lymphocytes,acting as an intracellular regulator in the complexchanges in intracellular metabolism which accompanylymphocyte transformation. In examining this possibilityfurther we have evaluated a number of pharmacologicagents which raise intracellular cyclic AMP levels(j8-adrenergic agents, prostaglandins, and methyl xan-thines) for inhibitory and stimulatory effects on pro-tein, RNA, and DNA synthesis in control and PHA-stimulated cells. Wehave made a similar evaluation ofthe effects of exogenous cyclic and noncyclic nucleotideson lymphocyte metabolism. In this paper we will demon-strate that exogenous cyclic and noncyclic nucleotides(1 X 10' to 1 X 10' mole/liter) inhibit the response ofhuman lymphocytes to PHAwhereas low concentrationsof cyclic AMP (1-10 X 10' mole/liter) are weaklystimulatory to cells unexposed to PHA. Evidence willalso be presented showing that pharmacological agentswhich raise the endogenous level of cyclic AMPmark-edly inhibit transformation in PHA-stimulated lympho-cytes. The relationship of these findings to a possible

442 The Journal of Clinical Investigation Volume 50 1971

messenger role of cyclic AMPin lymphocytes will bediscussed. Two brief resumes of this work have ap-peared previously (1, 2).

METHODSPreparation of lymphocytes. Purified lymphocytes were

prepared as described in the previous paper. In most ex-periments nylon column fractionation was not carried outand the preparation contained 40-70% lymphocytes.

Culture medium. The culture medium was a fortifiedEagle's minimum essential medium as described in theprevious paper.

Reagents. Phytohemagglutinin P (PHA-P ), lot Nos.52650, 52936, and 545526 from Difco Labs (Detroit, Mich.)was reconstituted in 5.0 cc of sterile H20 and diluted to1: 5 in Gey's balanced salt solution. In most experiments0.05 cc of the reconstituted and diluted PHA-P was addedto 1.5 cc of culture medium. This concentration of PHA-Pgave maximal uptake of thymidine-3H into DNA as mea-sured with a 4 hr thymidine pulse beginning 48 hr afterthe addition of PHA.

Cyclic AMP, Ns, O'2 dibutyryl cyclic AMP (DU-cyclicAMP), adenosine 5'-mono-, di-, and triphosphates (AMP,ADP, ATP), cyclic 3',5'-guanosine monophosphate (cyclicGMP, guanosine-5'-mono- and triphosphates (GMP, GTP),cyclic uridine 5'-monophosphate (cyclic UMP), cyclic cyti-dine 5'-monophosphate (cyclic CMP), cyclic inosine 5'-monophosphate (cyclic IMP), aminophylline, norepinephrine,and isoproterenol were purchased from the Sigma ChemicalCompany, St. Louis, Mo. Guanosine 5'-diphosphate (GDP)was purchased from the Schwarz BioResearch, Orange-burg, N. Y. These reagents were dissolved in Gey's bal-anced salt solution and filtered through Swinnex 0.45 /Lfilters (Millipore Corp., Bedford, Mass.).

Prostaglandins were dissolved in a 10% ethanol solutionas described in the previous paper. Final ethanol concen-trations were 1% or below and control solutions containingcomparable levels of ethanol were included in each experi-ment.

Radioisotopes included thymidine-methyl-'H (New Eng-land Nuclear, specific activity 6.7 Ci/mmole), leucine-4,5-3H(New England Nuclear, 59.1 Ci/mmole), and uridine-5-3H(Amersham/Searle, 22.8 Ci/mmole).

Culture conditions. Peripheral blood leukocytes (or puri-fied lymphocytes where indicated) were suspended in forti-fied Eagle's minimum essential medium at a concentrationof 106 lymphocytes per ml; 1.5 ml of the cell suspensionwas added to each culture tube (No. 2005, Falcon Plastics,Div. of Bioquest, Los Angeles, Calif.). The pH of the in-cubation medium remained within 0.1 pH U of 7.3 afterthe addition of reagents. The cells were incubated at 37°Cin a 5% CO2 atmosphere.

Measurement of incorporation of thymidine-3H into DNA.At 48 hr 1 ,Ci thymidine-3H was added, and 4 hr later thecells were harvested. In some experiments thymidine-3Hwas added at 48 hr and the cells were harvested at 72 hr.Using 4-hr pulses the uptake of radioactivity into DNAwas linear over the time period f rom 48 to 60 hr. Afterincubation cells were washed three times with cold phos-phate-saline. DNA was precipitated by incubating cells in2.0 ml of cold 10% trichloroacetic acid (TCA) for 20 min.The precipitates were washed once with 2.0 cc of cold 5%oTCA and allowed to air dry at room temperature beforedissolving in 0.5 cc of hydroxide of Hyamine 10-X (Pack-ard Instrument Co., Downers Grove, Ill.). Hyamine solu-

tions were quantitatively transferred to counting vials with10 cc of Bray's solution and measured for radioactivity ina Packard liquid scintillation counter. 95%o of the radio-activity has been shown to be in DNAusing this procedure(3). All experimental conditions were performed in trip-

licate.Measurement of incorporation of leucine-3H into cellular

protein. At 24 hr 1 gCi of leucine-3H was added to eachculture tube and 12 hr later the cells were harvested usingthe procedure for measuring thymidine-3H incorporation intoDNA. Under these conditions leucine uptake into TCA-pre-cipitable protein was stimulated 3-fold by PHA as com-pared with control cells.

Measurement of incorporation of uridine-3H into nucleicacid. At 24 hr 1 /ACi of uridine-3H was added to each cul-ture tube and 12 hr later the cells were harvested usingthe procedure for measuring DNA synthesis as describedabove. Under these conditions the amount of TCA-pre-cipitable radioactivity was stimulated 12-fold by PHA ascompared with control cells.

Blast cell changes in PHA-stimulated lymphocytes. Lym-phocyte morphology was evaluated in some experiments bymaking thin smears of cultured cells and staining withWright's stain. In the presence of PHA typical blast celltransformation was demonstrated in up to 70% of cells.Under conditions in which there was a decrease in thymi-dine-3H uptake decreased blast cell transformation wasobserved.

Binding of 'I-labeled PHA to purified lymphocytes.Chromatographically purified erythroagglutinating PHA(4) labeled with "2I was generously provided by Dr. StuartKornfeld, Washington University School of Medicine. Puri-fied (99%o) lymphocytes in which the erythrocytes had beenlysed by hypotonic shock (5) were incubated with labeledPHA (specific activity 6 X 104 cpm/flg) under the condi-tions described in the legend to Table IV. The hypotoniclysis has been shown not to alter PHA binding to lympho-cytes (5) whereas it essentially eliminates PHA bindingto erythrocytes. Lymphocytes treated in this way respondedto PHA (thymidine-3H uptake) and exhibited the expectedinhibition of thymidine uptake by the substances listed inTable IV. Binding of radioiodinated PHA to lymphocytestakes place largely at the external cell membrane as judgedby the results of elution studies (S. Kornfeld, personalcommunication). Within the time period of the experimentpresented in Table IV approximately 90% of the cell-bound radioactivity can be removed by fetuin or fetalcalf serum. Moreover, a purified glycopeptide isolated fromhuman erythrocytes markedly inhibits the PHA-5I lym-phocyte interaction providing strong evidence that a specificbinding process is involved (6).

RESULTSEffect of reagents which elevate endogenous cyclic

AMP levels on PHA-stimulated thymidine-'H uptake.Aminophylline, isoproterenol, norepinephrine, and pro-staglandins inhibited thymidine-3H uptake into PHA-P-stimulated lymphocytes (Fig. 1). These agents raisethe level of intracellular cyclic AMPin lymphocytes asdemonstrated in the previous paper. Interestingly, theprostaglandin which was the least potent stimulator ofcyclic AMP, PGF1, (See Table IV, preceding paper)was also the least effective inhibitor of lymphocyte trans-formation (Fig. 2). The degree of inhibition at specified

Human Lymphocyte Metabolism 443

o80 ~~~~Aminophy//me/080

't 60 Isoproterenol

ICI

20c0z X*'

3 5 7 9IxIU4M 3 5 7 91X10-3M 3CONCENTRATION

FIGURE 1 Peripheral blood leukocytes were incubated for72 hr with PHAwith thymidine-'H present during the final24 hr of the culture. The inhibitors (final concentrationsindicated on the abscissa) were added at 0 time just beforethe addition of PHA.

levels of isoproterenol, aminophylline, and prostaglandinvaried as much as 15-20% with different lymphocytedonors, but marked inhibition was consistently demon-strated.

The inhibition by isoproterenol and prostaglandinswas not readily reversible, whereas cells which had beenexposed to aminophylline for 72 hr and carefully washedwere fully responsive to PHA. At the concentrationsused none of the inhibitors appeared to have adverseeffects on cell viability (as defined by uptake of vitaldyes).

Inhibition by aminophylline was most marked whenthe reagent was present in the culture medium duringthe 1st hour of incubation; isoproterenol and prosta-glandin were effective inhibitors when added at 6 and24 hr after PHA-P stimulation (Fig. 3). The inhibitionby aminophylline, isoproterenol, and prostaglandins wasequally striking when purified lymphocytes (> 99%pure) were used instead of mixed cell populations.

100

~80-60- PGA,

PGACr

I-20-- AGA2

00 2 4 6 8 10CONCENTRATION1if5 M

FIGURE 2 Inhibition of DNA synthesis by prostaglandins.Peripheral blood leukocytes were incubated for 52 hr withPHA with thymidine-3H present during the final 4 hr ofthe culture. The prostaglandins were added at 0 time, justbefore PHA.

The possibility that lymphocyte transformation mightbe stimulated by isoproterenol, aminophylline, and theprostaglandins under altered experimental conditionswas evaluated. In one set of experiments cells were ex-posed to various concentrations (3 X 10O- to 8 X 10-7mole/liter) of these agents for periods of 1-6 hr, washed,and reincubated under standard conditions in the ab-sence of PHA and inhibitor. In other experiments lowconcentrations of the above agents (1 X 10-', 2.5 X 10-,1 X 10-5, 2.5 X 10-, and 1 X 10' mole/liter) were main-tained for the entire 72 hr of the culture. Unequivocalstimulation of thymidine uptake was not observed ineither set of experiments. In one experiment norepineph-rine present throughout the culture period (at 1.6 and5.0 X 10-5 mole/liter) produced low grade stimulation.In two other experiments the same concentrations ofnorepinephrine failed to produce stimulation.

Effects of exogenous nucleotides on PHA-stimulatedDNA synthesis. PHA-stimulated uptake of thymidine-3H into the DNA of lymphocytes was inhibited bycyclic GMP, cyclic AMP, cyclic UMP, cyclic CMP,and cyclic IMP (Fig. 4). The dibutyryl derivative ofcyclic AMPwas the most effective inhibitor followedby cyclic GMPand cyclic AMP. Noncyclic mono-, di-,and trinucleotides of guanosine and adenine inhibitedthymidine-3H uptake just as effectively as the respec-tive cyclic nucleotides (Fig. 4). Adenosine inhibited

100

80I ~~~~ /~soprtenyup-- /°29*PO^0'M

Aminophy//ine -_

601PGE - -

----C %it __"P 2.9xlt-4M

K '\ Dibutyryl Cyclic AMP40- 1.4z 4M

20-

C

t 2 4 6 2420min TI ME (Hours)

FIGURE 3 Peripheral blood leukocytes were incubated for72 hr with PHA with thymidine-3H present during thefinal 24 hr. Inhibitors were added at various times afterPHA (time indicated on the abscissa).

444 J. W. Smith, A. L. Steiner, and C. W. Parker

thymidine-3H uptake just as effectively as the adeninenucleotides.

The possibility that inhibition of lymphocyte trans-formation by cyclic AMPmight be secondary to overtcellular damage was evaluated. After 72 hr of incubationwith cyclic AMP, lymphocytes exhibited normal cellularmorphology, excluded trypan blue, and were not de-creased in number as compared with the control culture.Moreover, the inhibition by dibutyryl cyclic AMPwasreversible; lymphocytes incubated with inhibitor for 72hr, washed, and exposed to PHA-P incorporated thymi-dine-'H to the same extent as control lymphocytes.Thus there was no evidence that irreversible lympho-cyte damage had occurred.

The inhibition by dibutyryl cyclic AMPwas maximalonly if the nucleotide was present during the 1st hourof incubation. When it was added after 24 hr of incuba-tion the inhibition of PHA-P stimulation was much lessstriking (Fig. 3). It would therefore appear that theinhibitor was acting at a relatively early phase in thetransformation process.

The possibility that the inhibition is indirect, beingmediated through some cell other than the lymphocyte,was also considered. However, in experiments utilizinghighly purified peripheral blood lymphocytes ratherthan a mixed cell population the inhibitors were equallyeffective.

The possibility that low concentrations of cyclic AMPmight stimulate DNA synthesis was evaluated. At exo-genous cyclic AMPlevels of 1-10 X 106 moles/liter lowgrade stimulation of thymidine-'H uptake was demon-strable (Table I). Stimulation of similar magnitude of'H-labeled thymidine, uridine, and leucine uptake wasobtained with 1-2 X 108 M concentrations of DU cyclicAMP.

OC PHA lbtr i

S80 2 -.AMP - CAMP

60 2 D4AP 8at0time.cAMPGMP1 cUMP, c cUMP

tcMP~20 --

C~~~~~~CM0 2 4 6 8

CONCENTRATION-10,4MFIGURE 4 Inhibition of DNA synthesis in PHA-stimulatedlymphocytes. Peripheral blood leukocytes were incubatedwith PHA for 72 hr with thymidine-'H present during thefinal 24 hr. The cyclic and noncyclic nucleotides were addedat 0 time. cAMP, CGMP, cUMP, cCMP, and cIMP referto the respective cyclic nucleotides.

TABLE I

Effect of Low Concentrations of Exogenous CyclicAMPon DNASynthesis in Unstimulated

(No PHA) Lymphocytes

Control(no cyclic

AMPadded),

Exogenous cyclic uptake ofFinal cyclic AMP AMP, uptake of thymi-

concentration thymidine-3H dine-3H

cPm cPm60 X 10-6 mole/liter 280 215

292 301* 259331 367

20 X 10-6 mole/liter 285 229118 228* 171281 246

6 X 10-6 mole/liter 302 163301 307* 156318 209

224*

Results are expressed in counts per minute after a 4 hr pulsebeginning at 44 hr.Peripheral blood leucocytes were incubated for 52 hr withthymidine-3H present during the final 4 hr of the incubation.Cyclic AMPwas added at 0 time.* Numbers represent averages.

PHA-Stimulated protein synthesis. Table II sum-marizes the effect of various exogenous nucleotides andreagents which elevate endogenous lymphocyte cyclicAMP levels on PHA stimulated uptake of leucine-2H

TABLE I IPer Cent Inhibition of PHA-Stimulated Protein

Synthesis at 48 hr

6.5 X 10-4 1.6 X 10-4Reagent mole/liter mole/liter

Dibutyryl cyclic AMP 80 61Cyclic AMP 57 13AMP 56 20ATP 66 54Cyclic GMP 67 61Isoproterenol 88 37Aminophylline 100 72PGA, 100

Peripheral blood leucocytes were incubated for 36 hr with PHAwith leucine-'H present during the final 12 hr. Inhibitors wereadded at 0 time, before PHA. Unstimulated control cells in-corporated an average of 8840 counts per 10 min whereasPHA-stimulated cells incorporated an average of 26,760counts per 10 min.

Human Lymphocyte Metabolism 445

TABLE I IIPer Cent Inhibition of PHA-Stimulated RNA

Synthesis at 48 hr

6.5 X 10-4 1.6 X 10-4Reagent mole/liter mole/liter

Dibutyryl cyclic AMP 87 66Cyclic AMP 10 0AMP 28 0ATP 26 15Cyclic GMP 69 30GMP 84 52GTP 70 52Cyclic CMP 95 89Cyclic TMP 18 17Cyclic IMP 0 0Isoproterenol 59 29Aminophylline 97 61PGA, - 100

Peripheral blood leucocytes were incubated for 36 hr withuridine-'H present during the final 12 hr of the culture.Inhibi-tors were added at 0 time, before PHA. Unstimulated controlcells incorporated an average of 2445 cpm whereas PHA-stimulated cells incorporated an average of 28,120 cpm.

into lymphocyte protein. The inhibitory effects of thesereagents on protein synthesis largely parallel their in-hibitory effects on DNA synthesis.

PHA-stimulated RNA synthesis. In Table III asimilar evaluation of inhibitory effects on PHA-stimu-lated uptake of uridine-'H into lymphocyte RNA ispresented. The adenosine nucleotides were less effectivein inhibiting RNA synthesis than DNA synthesis.Contrariwise, cyclic CMPwas a much better inhibitorof uridine than of thymidine uptake. Otherwise, thepattern of inhibition is similar to that obtained in thestudy of DNA synthesis.

Binding of labeled PHA to lymphocytes. Table IVsummarizes the results of a study of the binding of 'I-labeled erythroagglutinating PHA to purified humanperipheral blood lymphocytes in the presence and ab-sence of several inhibitors of lymphocyte transformation.There is no indication that the inhibitors interfere withcellular uptake of PHA.

DISCUSSION

The results of this study indicate that isoproterenol,norepinephrine, aminophylline, and members of theprostaglandin family (PGE1, PGE2, PGA1, PGA2 andPGFla) interfere with the incorporation of labeled pre-cursors into the protein, DNA, and RNA of PHA-stimulated human peripheral blood lymphocytes. Studieswith radiolabeled PHA largely excluded the possibilitythat these agents interfere with PHA-lymphocyte bind-

ing. As shown in the preceding paper each of the aboveagents raises the level of endogenous cyclic AMP inhuman lymphocytes suggesting a possible basis for theinhibition of transformation. In accord with this possi-bility PGF1., which over the concentration range ex-amined is the least potent of the five prostaglandins inraising lymphocyte cyclic AMPlevels, is also the poorestinhibitor of PHA-stimulated incorporation of thymidine-'H into DNA. Admittedly each of the inhibitors mightaffect intracellular metabolites other than cyclic AMP.For example, phosphodiesterase inhibition by theophyl-line might be expected to raise intracellular levels ofcyclic GMP and other cyclic nucleotides. Taken to-gether, however, the common effects of these agents onintracellular cyclic AMPlevels would appear to be themost likely explanation for the inhibition.

If intracellular cyclic AMP is directly involved inthe inhibitory effects of prostaglandin, aminophylline,and catecholamines on lymphocyte transformation itmight be possible to obtain verification by the additionof cyclic AMP (7) to lymphocyte cultures. However,experiments in which lymphocytes were incubated withvarious cyclic and noncyclic nucleotides have failed toelucidate this question. To be sure, relatively low con-centrations (0.1 mmole/liter) of cyclic AMP and its2'0, N6 dibutyryl derivative (DU cyclic AMP) didproduce significant inhibition of thymidine uptake intonuclear DNA. However, further investigation cast con-siderable doubt as to the significance of this inhibition interms of the intracellular effects of cyclic AMP. Non-cyclic adenine nucleotides (ATP, ADP, and AMP)and adenosine were equivalent to cyclic AMP as in-hibitors. Moreover, members of the guanosine series(cyclic GMP, GTP, and GMP) were shown to be evenmore effective inhibitors of thymidine uptake than the

TABLE IVEffect of Cyclic Nucleotides, Aminophylline, and Isoproterenol

on PHA-Binding to Lymphocytes

Reagent Bound PHA

cpm*None (control) 6670Dibutyryl cyclic AMP(1 X 10-4 mole/liter) 6650Cyclic GMP(1 X 10-4 mole/liter) 7014Isoproterenol (7 X 10-4 mole/liter) 6709Aminophylline (1 X 10-4 mole/liter) 6927

1 X 106 purified lymphocytes were incubated with Il'I-labelederythroagglutinating PHA(105 cpm) for 40 min at R.T., washedfour times with phosphate-saline containing 0. 1% bovineserum albumin and counted. In the absence of cells, tubes con-tained an average of 160 cpm.* Each value is corrected for background and is the averageof four determinations.

446 J. W. Smith, A. L. Steiner, and C. W. Parker

various adenosine derivatives, and as in the adenineseries, cyclic GMPinhibited no better than the noncyclicguanine 5'-nucleotides. The inhibitory effects of non-cyclic nucleotides did not appear to be due to secondaryincreases in intracellular cyclic AMP concentrationssince guanosine and 5'GMP failed to alter lymphocytecyclic AMPlevels.' While it is possible that with moreextended studies direct effects of extracellular guanosineon lymphocyte adenyl cyclase might be demonstrable itseems more probable that the inhibition of thymidine-3Huptake into DNA by extracellular nucleotides is unre-lated to the inhibition observed when intracellular cyclicAMPlevels are elevated.

The basis for the inhibition of lymphocyte transfor-mation by extracellular nucleotides is uncertain. Inter-ference with uridine and thymidine transport into thecell by competition for transport receptors might explaindecreased isotope uptake into RNA and DNA. How-ever, the decrease in leucine incorporation into lympho-cyte protein is more difficult to explain on this basis.Another point to consider is that cyclic AMP addedat 24 hr is relatively ineffective in preventing thymidineuptake into DNA. Since the thymidine-3H is not addeduntil another 24 hr have elapsed interference of cyclicAMPwith thymidine transport cannot be the explana-tion. Whatever explanation is given, the near equiva-lence of the mono-, di-, and trinucleotides both in theguanosine and adenosine series (Fig. 4) must be con-sidered. Possibly the various adenosine and guanosinephosphates are broken down to common derivatives,most likely adenosine and guanosine, during transportinto the cell. In human red cells ATP and ADP arebroken down to adenosine by enzymes present on thecell surface. Adenosine is rapidly transported into thecell where it is quantitatively rephosphorylated to 5'-AMP(8).

The diminution in lymphocyte transformation in asso-ciation with increased lymphocyte cyclic AMP levelsmust be reconciled with the observation that one ofthe initial effects of PHA, a potent stimulator of lym-phocyte transformation, is to elevate the intracellularcyclic AMPconcentration. A possible explanation mightbe that isoproterenol, aminophylline, and prostaglandinsarrest cellular differentiation at a later point in the cellcycle, at a time when cyclic AMP levels in PHA-stimulated cells have fallen below those in control cells.As shown in the preceding paper this fall occurs within6 hr after exposure to PHA. Inhibition at this pointmight be nonspecific in the sense that it involves anintracellular effect of cyclic AMPnot normally exertedduring lymphocyte transformation.

Even with the inhibition exerted by prostaglandins,

1 Purified lymphocytes were incubated for 5 and 60 minwith 5'-GMP and guanosine (400, 80, and 20 ,umoles/liter).

aminophylline, and catecholamines on late metabolicevents in PHA-stimulated lymphocytes, an early stimu-latory effect of cyclic AMP needed to be considered.Such an effect might be masked if elevated cyclic AMPlevels were maintained over too long a period of time.With this possibility in mind we attempted to manipu-late lymphocyte cyclic AMP levels with isoproterenoland other agents so as to appropriate the absolute valuesand time course of changes produced by PHA itself.However, exposure of cells to various concentrations ofisoproterenol for periods of a few minutes up to 6 hrfailed to produce late changes in thymidine-3H uptakeinto nuclear DNA. Similar results were obtained withaminophylline and PGE1. The failure to stimulate lym-phocyte differentiation with these agents despite the pro-duction of alterations in lymphocyte cyclic AMP levelssimilar to those produced by PHA has three possibleinterpretations. (a) While an increase in cyclic AMPconcentration may be a necessary early effect of PHA,other early metabolic alterations directly attributable toPHA (changes in cyclic GMP levels for example)may well be equally important. (b) PHA may have toact over a period of many hours, and in the absence oflate PHA effects the final events in lymphocyte trans-formation cannot take place. In point of fact studieswith antibodies to PHA suggest a need for continuedPHA stimulation over a period of at least 6 hr. (c)As a less likely possibility lymphocytes could containtwo separate adenyl cyclase systems, one responsive toPHA and one responsive to isoproterenol and the pros-taglandins. If each of the two adenyl cyclases were toregulate its own intracellular cyclic AMP pool, dis-tinctive metabolic effects depending on the cyclase in-volved could occur. Under these circumstances iso-proterenol and prostaglandins would be unable to pro-duce the major intracellular effects of PHA. However,one might still expect that brief or prolonged exposureof the cells to dibutyryl cyclic AMP (which presumablywould have access to both compartments) would resultin lymphocyte transformation. DU cyclic AMP didstimulate lymphocyte transformation but the effect wassmall in comparison with that of PHA. The weaknessof this argument is that results obtained with a DUcyclicAMPdo not necessarily correspond to what is observedwith nonacylated cyclic AMPin cells permeable to bothnucleotides (9, 10). Nonetheless, separate pools of cyclicAMPin lymphocytes would appear to be an unlikely ex-planation for our failure to stimulate lymphocyte trans-formation during these experiments.

Taken as a whole our data fail to provide evidencethat cyclic AMPalone, increased as a result of stimula-tion of lymphocyte adenyl cyclase by PHA, can initiatethe complex series of metabolic alterations which cul-minate in lymphocyte transformation. Indeed the pre-

Human Lymphocyte Metabolism 447

dominant effect of a sustained elevation of lymphocytecyclic AMP levels is an inhibition of macromoleculesynthesis both in PHA-stimulated and control cells.The only data which might be considered to be incon-sistent with this interpretation is the weak stimulatoryeffect of low concentrations of added cyclic AMP inlymphocytes not exposed to PHA (Table I). Howeverthe increase in thymidine uptake in this circumstanceis quite small relative to what is observed with PHAitself.

The influence of exogenous nucleotides on lympho-cyte transformation differs in certain details from whatwas observed by Ryan and Hedrick (11) with malig-nant mammalian cells in tissue culture. In cultures ofmouse L-cell fibroblasts and Hela cells low concentra-tions of cyclic AMP (11) produced marked inhibitionof cell replication. DU cyclic AMP and various non-cyclic nucleotides were substantially less inhibitory.With PHA-stimulated lymphocytes DU cyclic AMPwas a considerably better inhibitor than cyclic AMP.The basis for this difference is uncertain but it mayhave to do with an alteration in cell membrane perme-ability in the neoplastic cells.

Work in progress indicates that agents which elevateintralymphocyte cyclic AMP levels also interfere withantigen-induced transformation of human lymphocytesin vitro. Thus with lymphocytes from subjects withtuberculin sensitivity, prostaglandin El, isoproterenol,and aminophylline at concentrations of < 0.5 mmole/liter markedly inhibited purified protein derivative(PPD)-stimulated uptake of thymidine-3H into lympho-cyte DNA. The physiological significance of the pros-taglandin inhibition is uncertain since the levels re-quired appear to be well above the circulating pros-taglandin concentrations normally observed in vivo.Whether the clinical use of theophylline and 9-adrener-gic agents in bronchial asthma and related diseases everresults in immunosuppression at the level of the sec-ondary cellular response to antigen remains to beestablished. Benner, Enta, Lackey, Makino, and Reed(12) have studied the effects of epinephrine on theprimary and secondary immune response in rats to 2,4-dinitrophenyl-egg albumin. The early administrationof epinephrine in a dose of 600 tg/kg for 10 days afterinitial exposure to antigen reduced antibody formationabout 10-fold. Repetition of the epinephrine injectionsafter booster injections of antigen also suppressed thesecondary response. On the basis of the in vivo studiesand our own results in vitro, drugs capable of pro-ducing sustained elevation of intracellular cyclic AMPlevels deserve further evaluation as immunosuppressiveagents.

Note added in proof. We have recently received acommunication from Dr. Margaret Cross, Department of

Biochemistry, University of Oxford, Oxford, England,dealing with her own as yet unpublished studies on thetransformation of pig peripheral blood lymphocytes. Inthis system DU cyclic AMP (10-7-10' mole/liter)caused an immediate burst of RNAsynthesis, later DNAsynthesis, and eventual transformation indistinguishablein timing and magnitude from that produced by PHA.Her results would appear to strengthen the possibilitythat cyclic AMPmay indeed be an important intracellu-lar messenger in PHA-stimulated lymphocytes.

ACKNOWLEDGMENTSWe wish to thank Miss Mary C. Johnson for valuabletechnical assistance.

Dr. Parker is the recipient of Research Career AwardAI09881. This work was supported by U. S. Public HealthService Grants A100219 and A104646.

REFERENCES1. Smith, J. W., A. Steiner, W. M. Newberry, Jr., and C.

W. Parker. 1969. The effect of dibutyryl cyclic adeno-sine 3',-5' monophosphate (DC-AMP) on human lym-phocyte stimulation by phytohemagglutinin. Fed. Proc.28: 566.

2. Smith, J. W., A. L. Steiner, W. M. Newberry, Jr., andC. W. Parker. 1969. Cyclic nucleotide inhibition of lym-phocyte transformation. Clin. Res. 17: 549.

3. Dutton, R. W., and J. D. Eady. 1964. An in vitro systemfor the study of the mechanism of antigenic stimulationin the secondary response. Immunology. 7: 40.

4. Weber, T., C. T. Nordman, and R. Grasbeck. 1967.Separation of lymphocyte-stimulating and agglutinatingactivities in phytohaemagglutinin (PHA) from Phaseo-lus vulgaris. Scand. J. Haematol. 4: 77.

5. Kornfeld, S. 1969. Decreased phytohemagglutinin recep-tor sites in chronic lymphocytic leukemia. Biochim. Bio-phys. Acta. 192: 542.

6. Kornfeld, R., and S. Kornfeld. 1970. The structure of aphytohemagglutinin receptor site from human erythro-cytes. J. Biol. Chem. 245: 2536.

7. Sutherland, E. W., G. A. Robison, and R. W. Butcher.1968. Some aspects of the biological role of adenosine3',5' monophosphate (cyclic AMP). Circulation. 37: 279.

8. Parker, J. C. 1970. Metbolism of external adenine nu-cleotides by human red blood cells. Amer. J. Physiol.218: 1568.

9. Kitabchi, A. E., S. S. Solomon, and J. S. Brush. 1970.The insulin-like activity of cyclic nucleotides and theirinhibition by caffeine on the isolated fat cells. Biochem.Biophys. Res. Commun. 39: 1065.

10. Solomon, S. S., J. S. Brush, and A. E. Kitabchi. 1970.Divergent biological effects of adenosine and dibutyryladenosine 3',5'-monophosphate on the isolated fat cell.Science (Washington). 169: 387.

11. Ryan, W. L., and M. L. Heidrick. 1968. Inhibition ofcell growth in vitro by adenosine 3',5'-monophosphate.Science (Washington). 162: 1484.

12. Benner, M. H., T. Enta, S. Lockey, Jr., S. Makino, andC. E. Reed. 1968. The immunosuppressive effect of epi-nephrine and the adjuvant effect of beta-adrenergicblockade. J. Allergy. 41: 110.

448 J. W. Smith, A. L. Steiner, and C. W. Parker