morphine withdrawal dramatically reduces lymphocytes in morphine-dependent macaques

ORIGINAL ARTICLE

Morphine Withdrawal Dramatically Reduces Lymphocytesin Morphine-Dependent Macaques

Michael R. Weed & Lucy M. Carruth & Robert J. AdamsNancy A. Ator & Robert D. Hienz

Received: 19 February 2006 /Accepted: 31 May 2006 / Published online: 14 July 2006# Springer Science+Business Media, Inc. 2006

Abstract The immune effects of chronic opiate exposureand/or opiate withdrawal are not well understood. Theresults of human studies with opiate abusers are variableand may not be able to control for important factors such assubjects drug histories, health and nutritional status.Nonhuman primate models are necessary to control theseimportant factors. Amodel of opiate dependence in macaqueswas developed to study the effects of opiate dependence andwithdrawal on measures of immune function. Four pigtailedmacaques drank a mixture of morphine (20 mg/kg/session)and orange-flavored drink every 6 h for several months.During stable morphine dependence, absolute numbers ofneutrophils, monocytes and lymphocytes did not changerelative to pre-morphine levels. However, there was asignificant decrease in the absolute number and percentageof natural killer (NK) cells in morphine dependence. Eitherprecipitated withdrawal or abstinence for 24 h resulted inbehavioral withdrawal signs in all animals. Absolute lympho-cyte counts decreased and absolute netrophil counts increasedsignificantly in withdrawal, relative to levels during morphinedependence. Lymphocyte subset (CD4+, CD8+, CD20+)cells were also decreased in absolute numbers with littlechange in their percentage distributions. There was, however,a significant increase in the percentage of NK cells in

withdrawal relative to levels during morphine dependence.This study demonstrates the usefulness of voluntary oral self-dosing procedures for maintaining morphine dependence innonhuman primates and demonstrates that the morphinewithdrawal syndrome includes large alterations in bloodparameters of immune system function, including nearly50% reduction in numbers of CD4+, CD8+ and CD20+ cells.

Key words macaque . morphine dependence . withdrawal .

immune lymphocyte . monocyte . neutrophil

Introduction

Opiate abuse remains a large problem in the United Statesand around the world. In 2002, there were an estimated166,000 current heroin users in the United States. Addition-ally, 277,000 people were being treated for heroin abuse,many of whom chronically receive the -opiate receptoragonist methadone (SAMSHA 2002). Opiate abusers aretherefore exposed to both legal and illegal opiates for years.Unfortunately, the impact of chronic opiate exposure onimmune system function is not completely understood.Acutely, opiates have both immunostimulatory and immu-nosuppressive effects (Friedman et al. 1988; Kreek 1990;Donahoe and Vlahov 1998). However, the net effect ofchronic opiate abuse on the immune system remainscontroversial (Drucker 1986; Kreek 1990; Rouveix 1992;Donahoe and Vlahov 1998; Alonzo and Bayer 2002;Douglas et al. 2003). Human heroin users may have de-creased immune function (Kreek 1990; Govitrapong et al.1998), but tolerance to some aspects of immunosuppressionmay develop (Kreek 1990; Rouveix 1992; Donahoe 1993).

Morphine is the prototypical -opiate receptor agonistand the major active metabolite of heroin (Inturrisi et al.

J Neuroimmune Pharmacol (2006) 1: 250259DOI 10.1007/s11481-006-9029-z

M. R. Weed (*) :N. A. Ator :R. D. HienzDepartment of Psychiatry and Behavioral Sciences,Johns Hopkins University School of Medicine,BBRC Suite 3000, 5510 Nathan Shock Drive,Baltimore, MD 21224, USAe-mail: [email protected]

M. R. Weed : L. M. Carruth : R. J. AdamsDepartment of Molecular and Comparative Pathobiology,Johns Hopkins University School of Medicine,733 N. Broadway, Room 815,Baltimore, MD 21205, USA

1984). In morphine-dependent monkeys, overall reductionsin immunocompetence can occur following morphinedependence for more than 2 years (Carr and France1993). These effects include suppression of natural killer(NK) cell activity; selective regulation of T lymphocyte cells(increasing percentages of CD8+ T cells and decreasingpercentages of CD4+ T cells); and altered levels ofimmunoglobulins G and M (IgG and IgM), interleukin-2(IL-2), and IL-2 receptors in peripheral blood mononuclearcells (PBMC) of morphine-treated rhesus monkeys. Similar-ly, studies of rhesus monkeys dependent on either morphineor the longer-acting -opiate receptor agonist L--acetylme-thadol (LAAM) found an initial transient increase in T-cellactivity early in exposure to morphine that later turned intooverall immunosuppression of T helper functions, such asIL-2 secretion, as opiate exposure continued over severalmonths (Chuang et al. 1993). Shifting populations of restingand activated memory CD4+ T lymphocytes have also beenreported in long-term morphine treatment (Donahoe et al.2001). Studies in rats have suggested similar immunosup-pressive changes following both short- and long-termexposure to morphine (Bhargava et al. 1994, 1995; Westet al. 1997) or heroin (Weber et al. 2004), and have alsoindicated the differential development of tolerance to someof morphines immunomodulatory effects (i.e., tolerance tomorphines suppression of NK activity but not to mitogen-stimulated splenocyte proliferation and interferon- produc-tion; Shavit et al. 1986; West et al. 1997, 1998a).

Even less well understood is the effect of opiatewithdrawal on immune system function in opiate-dependentanimals. Rahim et al. (2004) recently noted that There areonly three studies in mice and one in humans on the effectsof abrupt withdrawal from morphine on immune responses(Bhargava et al. 1994; Govitrapong et al. 1998; West et al.1999; Rahim et al. 2002), and only two studies haveexamined precipitated withdrawal (Donahoe 1993; Rahimet al. 2002). Studies using rodents have suggested thatabstinence and withdrawal induces immunosuppression(Bhargava et al. 1994, 1995; West et al. 1998b, 1999).Morphine withdrawal in mice impairs aspects of immunesystem function within 4 6 h and the immunosuppressionmay persist for days (Rahim et al. 2003, 2004). In addition tothese rodent studies, an early study in dogs demonstrated noconsistent change in red or white blood cell counts duringmorphine dependence; however, during withdrawal, decreasesin red blood cells and increases in white blood cells occurred4 7 days following morphine cessation (Pierce and Plant1928). The mechanism of immunosuppression in opiatewithdrawal is likely related to increased activity in thehypothalamicpituitaryadrenal (HPA) axis and release ofsteroidal hormones (Pechnick 1993; West et al. 1999; Alonzoand Bayer 2002; Avila et al. 2004), although immunosup-pression has been reported in mice during opiate withdrawal

without changes in corticosterone levels (Kelschenbach et al.2005). Interestingly, the immune system has been linked toexpression of the opioid withdrawal syndrome in thatimmunosuppressed rats do not exhibit the typical behavioralsigns of withdrawal when opiate administration is discon-tinued (Dafny et al. 1989; Dougherty et al. 1990).

Given that most human opiate abusers experience with-drawal on a regular basis (Kreek 1987), studies in this area aresurprisingly few. In monkeys, Donahoe (1993) reported thatcell-mediated immune functions are not impaired in mor-phine-tolerant animals, but that acute withdrawal impairs cell-mediated immune function. The one study in humans reporteddecreases in CD4+ and NK cells and increases in B cells andCD8+ T cells in heroin-dependent patients and formerlydependent individuals abstaining from heroin (Govitrapong etal. 1998). Clearly, further clinical and preclinical studies arenecessary to fully characterize the effects of withdrawal onimmune system function in humans and nonhuman primates.Preclinical models of opiate abuse are useful because theycontrol for the many confounding factors in studies of humandrug abusers, including histories with many drugs, impurities/adulterants in street drugs, poor nutrition, reduced access tohealth care, and many other factors. The present study reportsdevelopment of a model of opiate dependence in nonhumanprimates and the use of this model to study the effects ofopiate withdrawal on measures of immune function.

Methods

Principles of laboratory animal care (Guide for the Care andUse of Laboratory Animals, National Academy Press,1996) were followed, and all protocols were approved bythe Institutional Animal Care and Use Committee of TheJohns Hopkins University School of Medicine.

Morphine self-dosing

Four male pig-tailed macaques (3 4 years old, 4 7 kg atthe start of the study) were individually housed in cagesequipped with the experimental apparatus described below.Lights were on in the housing room from 7 a.m. to 9 p.m..Water was continuously available throughout the study viaan automated drinking system on the rear wall of the cage(Edstrom Industries, Waterford, WI). Cages were equippedwith computer-controlled drinking equipment on one sideof the cage (BRS/LVE, Laurel, MD). Each cage had twospouts side by side, and each spout was connected to twofluid reservoirs. Fluid flowed from each of the tworeservoirs in succession, as described below. A buzzersounded for the duration of a drinking session, and yellowstimulus lights near either spout indicated availability offluid on that spout. Lip contact with the spout opened a

J Neuroimmune Pharmacol (2006) 1: 250259 251

solenoid valve for 1 s, delivering approximately 4 mL offluid for one drink delivery.

The monkeys readily learned to drink a 5% (w/v) orange-flavored Tang solution (Tang, General Foods Corp., WhitePlains, NY). The maximum number of 1-s drinks per sessionwas 25, for a total maximum intake of roughly 100 mL persession. The Tang was then adulterated with increasingamounts of quinine to habituate the monkeys to a bittersweetorange solution (up to 0.32 mg/mL quinine sulfate) in amanner derived from Turkkan et al. (1989) for oral drugdosing. Once monkeys were consuming all available Tangquinine solution at 0.32 mg/mL quinine, quinine wasremoved from the solution and Tang-only sessions wereconducted at four times per day at 9 a.m., 3 p.m., 9 p.m., and3 a.m., (7 days/week) and were a maximum of 15 min long.

Morphine sulfate was then added to the Tang solution inslowly increasing amounts from 4 to 20 mg/kg available persession (doses calculated as the salt). One fluid reservoircontained the morphineTang solution and the secondreservoir contained unadulterated Tang solution. The first13 drinks in a session were delivered from the morphineTang reservoir and the final 12 drinks in a session weredelivered from the Tang-only reservoir to provide a chaserto diminish any bitter taste and/or provide a reinforcer fordrinking all of the morphineTang solution available. Actualvolumes of the morphine solution delivered were measuredand individualized dosages were adjusted according to thevolume delivered and the monkeys weight.

Morphine concentration was typically increased after 23 weeks of exposure. In rhesus monkeys, morphinedelivered orally has approximately 15% bioavailabilityrelative to intravenous administration (Rane et al. 1984).Intravenous, intramuscular, and subcutaneous (s.c.) admin-istration leads to near-complete bioavailability of morphine(Stanski et al. 1978; Glare and Walsh 1991; Lugo and Kern2002); therefore, assuming 15% bioavailability of oralmorphine, 20 mg/kg morphine, p.o., q.i.d., approximates3 mg/kg, s.c., q.i.d., a dose that has been used to maintain arobust morphine dependence (Seevers 1936). If a monkeyfailed to consume any morphine for three consecutivedrinking sessions, he was given an i.m. injection of 3 mg/kgmorphine sulfate (dissolved in saline 0.1 mL/kg); however,such supplemental injections were rarely needed.

The monkeys were periodically restrained with 10 mg/kgketamine hydrochloride i.m. to obtain blood samples forassessment of immunological parameters. Sampling proce-dures were separated by at least 2 weeks. Body weightswere also determined at these times. Before morphineexposure, blood was drawn on three occasions before themonkeys began their round-the-clock Tang drinking sched-ule (pre-Tang). To control for the effects of the schedulealone (i.e., potential immune effects of stress or sleepdisturbances), a fourth blood draw occurred after the

monkeys had been drinking Tang on the 3 a.m., 9 a.m.,3 p.m., and 9 p.m. drinking schedule for at least 3 months.There were no significant differences between any of theimmune parameters due to the effects of the drinkingschedule, as determined by paired t-tests between the meanof the three pre-Tang measurements and the fourth baselinedraw (data not shown). All four measurements were thenaveraged together to form the premorphine baseline mean.

Morphine withdrawal procedures

When the morphine dose had reached 20 mg/kg/session andmorphine consumption was stable for at least 2 weeks,withdrawal was initiated by either removal of morphinefrom the drink solution (abstinence) or precipitated by ad-ministration of the opiate antagonist naltrexone (0.1 mg/kgi.m.; precipitated withdrawal). At least 2 weeks of stablemorphine self-dosing separated the withdrawal procedures.

Morphine abstinence withdrawal procedures began afterthe 9 a.m. morphineTang drinking session by providingaccess to Tang with no morphine during the subsequentregular drinking sessions. Blood was drawn at 9 p.m., theevening of that same day (12 h after the last morphinedrinking session), and again the next morning at 9 a.m.(24 h after the last morphine drinking session). An injectionof 3 mg/kg morphine s.c. was administered at 11 a.m. toend the abstinence syndrome, and regular morphineTangdrinking sessions commenced at 3 p.m..

Precipitated withdrawal procedures began with administra-tion of 0.1 mg/kg naltrexone at 12 p.m., followed by removalof morphine from the drink solution for the next 24 hours.Blood draws occurred as described above at 9 p.m. thatevening and 9 a.m. the following morning.

To control for the effects of the blood draw schedule,monkeys were restrained with ketamine before a 9 p.m.drinking session and blood was drawn. The monkeys wereadministered 3 mg/kg morphine i.m. once they recoveredfrom the ketamine restraint. The 3 a.m. morphine drinkingsession occurred as usual and blood was drawn at 9 a.m. asdescribed above.

The abstinence and precipitated withdrawal procedureswere repeated once and each monkeys home cage behaviorwas videotaped after 12 and 24 h of abstinence for observerrating. For precipitated withdrawal, home cage behavior wasvideotaped for 15 min at 1 p.m., 1 h after administration ofnaltrexone, and again 24 h after the last drinking session withthe morphineTang drinking solution. Custom softwareallowed trained raters to score behaviors from the videotapesduring one 15-min time bin 12 and 24 h after the lastmorphine drinking session. Occurrences of 21 behaviorswere noted by two blind raters using a computerized check-list derived from Seevers (1936).

252 J Neuroimmune Pharmacol (2006) 1: 250259

Immune parameters

Complete blood counts with differentials were performed onblood samples drawn before morphine exposure, and duringmorphine dependence and withdrawal. The absolute numberof lymphocytes was determined by using a CellDyn 3200hematology analyzer (Abbott). Mononuclear cells wereseparated on Percoll discontinuous gradients and labeledwith fluorochrome-conjugated monoclonal antibodies toCD4+, CD20+, CD8+, and CD16+ (Becton Dickinson, SanJose, CA). Cells were analyzed on a FACSCalibur flowcytometer and data were analyzed using Cellquest software.

Data analysis

Observed behavioral ratings were compared using paired t-tests, adjusted for the number of comparisons for eachmeasure.

Blood parameters (neutrophils, monocytes, lymphocytesand subsets, etc.) were analyzed using repeated measuresANOVA with planned post hoc comparisons (TukeyKramer procedure, adjusted for the number of comparisons;Prism, GraphPad Software, San Diego, CA) between theconditions of premorphine vs morphine-dependent andmorphine-dependent vs morphine withdrawal. AllANOVAs had 2 degrees of freedom between subjects and3 degrees of freedom within subjects. Blood parameterswere also compared with t-tests between the first andsecond control blood draws in a 24-h period to determine ifthe bleeding procedure itself induced any changes in bloodparameters.

Results

Oral morphine dependence

Nearly all Tang solution was consumed in each session,including the 3 a.m. session (Table 1). During chronic con-

sumption of 20 mg/kg/session morphine, nearly all availablemorphine was consumed in the 9 a.m., 3 p.m., and 9 p.m.sessions. In the 3 a.m. session, the monkeys consumedroughly half of the available morphine, and there was signi-ficant individual variation in drinking in the 3 a.m. session.The four individual monkeys drank an average of 92%, 70%,38%, and 29% in the 3 a.m. session when 20 mg/kg/sessionwas available. No monkeys were observed spontaneouslyentering into withdrawal, for approximately 6 months, includ-ing the two monkeys drinking less often in the 3 a.m. session.

All four monkeys maintained morphine dependencethroughout the experiment and were never observed toexhibit characteristics of withdrawal (e.g., wet-dog shakes,repeated yawning, prostration, etc.) due to lack of drinking.Supplemental injections of morphine were given if themonkeys failed to consume morphine in three consecutivesessions; however, these injections were needed only sixtimes total for the group in the 6 months of drinking androughly 700 drinking sessions for each monkey.

Morphine withdrawal syndrome

Ratings of withdrawal behaviors were similar following24 h of abstinence from morphine or 1 h after injection of0.1 mg/kg naltrexone (Table 2). Signs previously charac-terized as moderate and severe (Seevers 1936) wereincreased relative to normal drinking or saline injection,respectively. Twenty-four-hour withdrawal significantly in-creased signs of wet-dog shakes and yawning, and decreasedoccurrences of cage shaking (restlessness) at the p < 0.05significance level. Precipitated withdrawal also increasedsigns of wet-dog shakes and yawning, and decreasedoccurrences of cage shaking (restlessness) at the p < 0.05significance level. Precipitated withdrawal, but not absti-nence, significantly increased the incidence of prostration, abehavior not evidenced by any animal before the with-drawal studies. The similarity of the withdrawal signsfollowing either precipitated withdrawal or 24 h abstinencesuggests that the severity of withdrawal was similarfollowing either procedure.

Table 1 Consumption of Tang and morphine, expressed as a per-centage of available solution consumed

Tang consumedwithout morphine

Morphine solution consumedat 20 mg/kg/session

9 a.m. (S.E.) 96.7% (2.2%) 98.5% (1.7%)3 p.m. (S.E.) 94.7% (6.1%) 99.9% (0.3%)9 p.m. (S.E.) 99.4% (0.7%) 96.8% (3.6%)3 a.m. (S.E.) 83.9% (6.3%) 57.3% (16.9%)

When monkeys consumed 100% of the available morphine solution,they received 20 mg/kg morphine. Monkeys consumed nearly all ofTang provided in all sessions. Monkeys consumed nearly all morphineduring the 9am, 3pm, and 9pm sessions.

Table 2 Withdrawal signs significantly changed during precipitatedwithdrawal or 24 h of abstinence

Saline(no withdrawal)

Precipitatedwithdrawal

24 hAbstinence

Wet dog shakes (S.E.) 0.00 (0.00) 1.13a (0.29) 1.25a (0.29)Yawning (S.E.) 0.27 (0.29) 2.88a (0.55) 2.25a (0.37)Cage shaking (S.E.) 1.21 (0.36) 0.13a (0.14) 0.25a (0.29)Prostration (S.E.) 0.00 (0.00) 1.88a (0.43) 0.25 (0.29)

Mean occurrences of each behavior in the sampling period.a Significant at the p < 0.05 level (paired t-test, adjusted for multiplecomparisons).


Numbers of neutrophils, monocytes, and lymphocytesare presented in Fig. 1 for the monkeys before morphineexposure, during chronic morphine dependence and duringwithdrawal at 24 hours. Both withdrawal procedures, either24 h of abstinence or naltrexone followed by abstinence for24 h, produced similar changes in blood parameters: Therewere no significant differences in numbers of neutrophils,monocytes, or lymphocytes in withdrawal between the twowithdrawal procedures. Therefore, these data were averagedtogether and presented as the mean of two replications.Additionally, there were no significant effects of eitherwithdrawal procedure at 12 h (data not shown); therefore,these data were omitted from the figures for simplicity.

The repeated-measures ANOVA revealed significantmain effects of morphine treatments for neutrophils (F =7.3; p < 0.05), monocytes (F = 5.2; p < 0.05), andlymphocytes (F = 13.0; p < 0.01). Planned comparisonsrevealed that there were no significant decreases in neutro-phils, monocytes, or lymphocytes between blood samplestaken before morphine exposure and blood samples takenduring stable morphine dependence (p > 0.05 for each).However, there were significant increases in neutrophils (p 0.05 for each). However, there was a significant reductionin the number of NK cells between the premorphine and

Fig. 1. Absolute numbers of neutrophils, monocytes, and lympho-cytes in pigtailed macaques. The y-axis represents cell count in K/L.The x-axis represents the conditions of premorphine (morphine nave),morphine dependence, and withdrawal (24 h of abstinence, mean ofwith and without naltrexone). An asterisk indicates that the withdrawalcondition differs significantly from the morphine-dependent conditionat the p < 0.05 level. and indicate that the withdrawal conditiondiffers from the premorphine condition at the p < 0.05 and p < 0.01levels, respectively.


Fig. 2. Relative levels of CD4+, CD8+, CD20+, and NK cells inpigtailed macaques. The y-axis represents cell count in K/L. The x-axis represents the conditions of premorphine (morphine nave),morphine dependence, and withdrawal (24 h of abstinence, mean ofwith and without naltrexone). and indicate that the morphine-dependent condition differs from the premorphine condition at the

p < 0.05 and p < 0.01 levels, respectively; * and ** indicate that thewithdrawal condition differs from the morphine-dependent group atthe p < 0.05 and p < 0.01 levels, respectively. and indicate thatthe withdrawal condition differs from the premorphine condition at thep < 0.05 and p < 0.01 levels, respectively.


morphine-dependent conditions (p < 0.01). Morphinewithdrawal produced significant reductions in the numberof CD4+, CD8+, and CD20+ T cells relative to levels duringstable morphine dependence (p < 0.05 for each). Thenumber of NK cells did not differ between morphinedependence and withdrawal (p > 0.05). Numbers of alllymphocyte subsets were significantly reduced duringwithdrawal compared with premorphine levels (CD4+p 0.05 for each). Group comparisonsrevealed a small but significant decrease in the percentage ofCD4+ T cells between the morphine-dependent and mor-phine-withdrawal conditions (p < 0.05). Group compar-isons also confirmed significant decrease in the percentageof NK cells between the premorphine and morphine-dependent conditions (p < 0.05). There was also anincrease in the percentage of NK cells during withdrawalrelative to stable morphine dependence (p < 0.05).

Discussion

The results of this study clearly show that the oral self-dosing procedures produced reliable morphine intake and arobust opiate dependence. An opiate withdrawal syndromewas clearly evidenced following either abstinence orprecipitated withdrawal. Chronic morphine dependenceproduced only a modest change in the blood parametersmeasured, including a decline in absolute NK cell counts.The major finding in this study is the demonstration that themorphine withdrawal syndrome included both an increasein absolute numbers of neutrophils and a dramatic reductionin the number of lymphocytes, including CD4+, CD8+,CD20+, and NK cell counts, relative to either morphine-nave or morphine-dependent levels.

The oral self-dosing procedure described here delivered98% of the 20 mg/kg morphine p.o. on average over the 9 a.m.,3 p.m., and 9 p.m. sessions and 57% in the 3 a.m. sessions.Although there is some variation inherent in the oral self-dosing model, this is actually consistent with human drugtaking as variation in day-to-day drug intake occurs in mostdrug abusers (Kreek 1987). The model presented here extendsearlier work with oral morphine self-dosing in rodents(Khavari et al. 1975; McMillan et al. 1976; Gellert andHoltzman, 1978) and oral opiate self-dosing in nonhumanprimates with the potent -agonists etonitazene (Tang 1982),

methadone (Crowley et al. 1975), and LAAM (Crowley et al.1985). Morphine is often used chronically in the treatment ofpain, and is the principal active metabolite of the major opiateof abuseheroin; therefore, the development of an automat-ed model of morphine dependence is an important method-ological step in this area of research. Automated self-dosingprocedures allow round-the-clock drug dosing withouttechnical supervision for each time point. Other methods,such as adulteration of water supply or adulteration of foodsources, have potential confounds due to alterations inhydration levels or nutritional status. Because most in vitroimmunological studies utilize morphine, the development of amodel of automated, voluntary, oral morphine dependence invivo allows for direct comparison of the same opiate in vivoand in vitro.

The present study assessed opiate withdrawal usingvideotaped behavior in the home cage and a checklistmodified from Seevers (1936). Withdrawal signs ofmoderate and severe status were evident during themorphine withdrawal syndrome at 24 h, but not at 12 h(Seevers 1936). Overall, the method of videotaping homecage behavior used here is likely to be less sensitive thanthe methods described by Seevers (1936), which requireinteraction and physical contact between the monkey andthe raters. However, the current study used these methodsonly to establish that the monkeys were indeed dependenton morphine. Therefore, the relatively insensitive video-taped measures were sufficient for the qualitative demon-stration of withdrawal that verifies physical dependence onmorphine. Using these procedures, the lack of withdrawalsigns at 12 h is not strong evidence that withdrawal had notyet begun; more sensitive measures may have producedindications of withdrawal at that time. Opiate withdrawal inmacaques tends to peak between 24 and 48 h after the lastopiate exposure, and it is therefore likely that thewithdrawal syndrome reported here underestimates theseverity of the opiate dependence produced by this dosingregimen. As mentioned above, the withdrawal proceduresdescribed here were included to determine whether thisdosing regimen produced morphine dependence rather thanto fully characterize the nature of that dependence. Theresults are clear that this procedure is useful for producingmorphine dependence in macaques.

The results of the present study are consistent with thosefrom earlier studies reporting only small changes in immunefunction in opiate-dependent monkeys, and larger changesduring opiate withdrawal (Donahoe 1993; Donahoe et al.2001). Donahoe et al. (2001) reported relative differences intotal lymphocyte numbers and numbers of lymphocytesubsets between morphine-treated and saline-treatedmonkeys receiving injections every 6 h. These effects werenoted between control and morphine-dependent groups;however, the differences were largely due to changes in


control lymphocyte levels over time. Comparisons oflymphoctye levels within subjects indicated that there wasno change from baseline in the morphine-dependentmonkeys, similar to the results of the present study.However, there was a change from baseline in the saline-treated control group in the Donahoe et al. (2001) study.These results were interpreted as being produced byalterations in stress-related hormones from the experimentalprocedure. Morphine apparently normalized the effects ofthe experimental stress in the morphine-treated monkeys,whereas the control group was not protected. In the presentstudy, there was not a separate control group, but beforemorphine exposure the monkeys drank Tang for severalmonths on the same 3 a.m., 9 a.m., 3 p.m., and 9 p.m.schedule. Unlike the monkeys injected with saline in theDonahoe et al. (2001) study, in the present study there wereno changes in immune parameters when the monkeysreceived the Tang vehicle every 6 h. Given the expecteddifference in stress levels between voluntary oral self-dosing and s.c. injections, this difference is not surprisingand supports the interpretation in Donahoe et al. (2001).Similar to that study, there were no changes in lymphocytenumbers after months of stable morphine dependence in themorphine-dependent monkeys, relative to the premorphinebaseline levels.

The present study did not find changes in the percen-tages of CD4+ and CD8+ T cells previously reported inmonkeys treated once daily with morphine (Carr andFrance 1993); however, the reductions in NK cell numberreported here are consistent with decreases in NK activityfrom Carr and France (1993). The Carr and France (1993)study administered once daily injections of 3.2 mg/kg,which is roughly equivalent to one of the doses adminis-tered by Donahoe et al. (2001) or one of the four drinkingsessions used in the present study. Carr and France (1993)also report that monkeys receiving morphine less frequently(once to twice per week) had similar reductions in CD4+

and increases in CD8+ T cells. Both dose and duration ofopiate exposure are important factors in severity of opiatedependence (Seevers 1936). Results from the present studyand those from Donahoe et al. (2001) suggest that littlealteration in levels of CD4+ or CD8+ T cells occur over along period of frequent exposure to morphine in monkeys(i.e., months of four times per day morphine), suggesting aform of tolerance develops. Results of the Carr and France(1993) study are consistent with the notion that infrequentor once-daily exposure in monkeys is not sufficient toengender tolerance to alterations in CD4+ or CD8+ T-celllevels. However, little is known about the development oftolerance to the immune effects of opiates, making itdifficult to resolve differences between these studies.

The results of these studies are consistent with others(Kreek 1990; Rouveix 1992; Donahoe 1993), showing that

chronic opiate administration results in a state of immunetolerance. Immune tolerance would explain why chronic,stable, opiate dependence appears to have fewer immunesequelae than acute administration of opiates. The level ofimmune tolerance is not complete, as indicated byreductions in the number of NK cells in the present studies.Although a number of studies demonstrate immunetolerance, tolerance is not complete and some level ofdysfunction continues in heroin abusers (Kreek et al. 1990;Zajicova et al. 2004). For instance, Govitrapong et al. (1998)reported that heroin users had increased numbers of CD3+

and CD8+ lymphocytes along with decreased numbers ofCD4+ and NK cells. It is possible that the stable morphinedependence by q.i.d. dosing reduced immune compromise,and this notion is consistent with the reduced effects ofmorphine dependence on lymphocytes in Donahoe et al.(2001) and in the present study, compared with effectshuman heroin abusers (Kreek et al. 1990; Govitrapong et al.1998). The idea that consistent opiate dependence sup-presses the immune system less than dependence inter-spersed with withdrawal is also supported by evidence ofimproved immune function in former heroin abusers under-going methadone maintenance therapy (Kreek et al. 1990;Zajicova et al. 2004). This hypothesis is an easily testableone, and such experiments were part of the motivation todevelop automated self-dosing methods.

The results of the Govitrapong et al. (1998) study alsodiffered from those of the present study in that theyreported increased numbers of lymphocytes in heroin usersabstaining for between 15 days and 6 months. The differ-ences between that study and the present study may reflectdifferences due to stability of drug exposure, as describedabove, or may simply be due to the different period ofwithdrawal studied. The present study examined 1 day ofwithdrawal, whereas the Govitrapong et al. (1998) studybegan at least 15 days into withdrawal and this differencemakes comparisons of the results difficult. Further con-trolled studies in human abusers combined with furtherstudies in primate models are needed to characterize thenature, time course, and mechanisms of immune tolerance.

The mechanism behind the increase in neutrophils andreduction in the number of lymphocytes during morphinewithdrawal has not been completely elucidated; however,activation of the HPA axis is likely to be involved. Opiatewithdrawal increases activity throughout the HPA and highHPA activity has been shown to be immunosuppressive(Bryant et al. 1988, 1991; Alonzo and Bayer 2002). Thetime course of HPA activity and glutocorticoid/corticoste-rone release coincides with immunosuppression of opiatewithdrawal. Indeed, given that acute but not chronic opiatesresult in activation of the HPA (Avila et al. 2004), it islikely that tolerance to the HPA-stimulating effects ofopiates produces the immune tolerance seen in chronic


opiate dependence. The oral dosing methods developedhere would be of great use for the study of thesemechanisms in primates.

Other studies of the time course of withdrawal in rodentshave suggested effects of abstinence on immune parameterswithin 46 h (Rahim et al. 2003; Rahim et al. 2004). In thecurrent study, there were no significant effects on immunemeasures after 12 h of abstinence (with or withoutnaltrexone administration). A full study of the time coursewould be needed to determine if this difference is due to thedifferent measures employed by the different studies orwhether there is a species difference in the timing ofimmune system compromise in withdrawal. For instance,the current study measured the absolute number andrelative percentage of NK cells but did not include ameasure of their functional activity. It is possible that theabsolute number of NK cells is unchanged after 12 h ofwithdrawal, but that the functional activity of the NK cellsis diminished.

In summary, the present study demonstrates the success-ful development of morphine dependence in nonhumanprimates using oral self-dosing procedures. This model wasthen used to study the effects of withdrawal on immunesystem function. Whether precipitated by naltrexone orabstinence from morphine, withdrawal induced an absti-nence syndrome with clear behavioral signs and clearimpact on blood parameters of immune system function,including a reduction in lymphocyte counts of >50%.

Acknowledgements The observer rating software was developed byBarbara Kaminski, Ph.D., and we are grateful for her contribution tothese studies. Kristin M. Wilcox, Ph.D., provided valuable commentson an earlier version of the manuscript. Rachel Gray, Chris Pyle, andApril Hatfield provided expert technical assistance in the performanceof these studies. We would like to acknowledge the support of theRetrovirus Laboratory in the Department of Comparative Medicine fortheir support. This research was supported by The National Institute onDrug Abuse grants DA013343 (R.D.H., N.A.A., and R.J.A.),DA05831 (M.R.W.), MH070306, NS047984, and HL075840 (L.M.C.).

References

Alonzo NC, Bayer BM (2002) Opioids, immunology, and host de-fenses of intravenous drug abusers. Infect Dis Clin North Am 16(3):553569

Avila AH, Alonzo NC, Bayer BM (2004) Immune cell activity duringthe initial stages of withdrawal from chronic exposure to cocaineor morphine. J Neuroimmunol 147(12):109113

Bhargava HN, Thomas PT, Thorat S, House RV (1994) Effects ofmorphine tolerance and abstinence on cellular immune function.Brain Res 642(12):110

Bhargava HN, House RV, Thorat SN, Thomas PT (1995) Effects ofnaltrexone on morphine-induced tolerance and physical depen-dence and changes in cellular immune function in mice. Brain Res690(1):121126

Bryant HU, Bernton EW, Holaday JW (1988) Morphine pellet-induced immunomodulation in mice: temporal relationships. JPharmacol Exp Ther 245(3):913920

Bryant HU, Bernton EW, Kenner JR, Holaday JW, (1991) Role ofadrenal cortical activation in the immunosuppressive effects ofchronic morphine treatment. Endocrinology 128(6):32533258

Carr DJ, France CP (1993) Immune alterations in morphine-treatedrhesus monkeys. J Pharmacol Exp Ther 267(1):915

Chuang LF, Killam KF Jr., Chuang RY (1993) Opioid dependency andT-helper cell functions in rhesus monkey. In Vivo 7(2):159166

Crowley TJ, Macdonald MJ, Zerbe G (1985) Variability in simianmotor and social behavior with alternating-day acetylmethadol.Psychopharmacology (Berl) 85(3):353360

Crowley TJ, Hydinger M, Stynes AJ, Feiger A (1975) Monkey motorstimulation and altered social behavior during chronic methadoneadministration. Psychopharmacologia 43(2):135144

Dafny N, Dougherty PM, Pellis NR (1989) The immune system andopiate withdrawal. Int J Immunopharmacol 11(4):371375

Donahoe RM (1993) Neuroimmunomodulation by opiates: relation-ship to HIV-1 infection and AIDS. Adv Neuroimmunol 3:31 46

DonahoeRM,VlahovD (1998)Opiates as potential cofactors in progressionof HIV-1 infections to AIDS. J Neuroimmunol 83(12):7787

Donahoe RM, Byrd LD, McCLure HM, Brantley M, Wenzel D,Ansari AA, Marsteller F (2001) Effects of morphine on T-cellrecirculation in rhesus monkeys. Adv Exp Med Biol 493:89101

Dougherty PM, Pellis NR, Dafny N (1990) The brain and the immunesystem: an intact immune system is essential for the manifestationof withdrawal in opiate addicted rats. Neuroscience 36(2):285289

Douglas SD, Camarca M, Xu J, Durako S, Murphy D, Moscicki B,Wilson CM (2003) The relationships between substance abuse,psychosocial variables, and natural killer cell enumeration andfunction in HIV-infected and high-risk uninfected adolescents.AIDS Res Hum Retroviruses 19(5):399 408

Drucker E (1986) AIDS and addiction in New York City. Am J DrugAlcohol Abuse 12(12):165181

Friedman H, Klein T, Specter S, Pross S, Newton C, Blanchard DK,Widen R (1988). Drugs of Abuse and Virus Susceptibility.Psychological, Neuropsychiatric and Substance Abuse Aspects ofAIDS T. P. e. a. Bridge. New York, Raven Press

Gellert VF, Holtzman SG (1978) Development and maintenance ofmorphine tolerance and dependence in the rat by scheduledaccess to morphine drinking solutions. J Pharmacol Exp Ther205(3):536546

Glare PA, Walsh TD (1991) Clinical pharmacokinetics of morphine.Ther Drug Monit 13(1):123

Govitrapong P, Suttitum T, Kotchabhakdi N, Uneklabh T (1998)Alterations of immune functions in heroin addicts and heroinwithdrawal subjects. J Pharmacol Exp Ther 286(2):883889

Inturrisi CE, Max MB, Foley KM, Schultz M, Shin SU, Houde RW(1984) The pharmacokinetics of heroin in patients with chronicpain. N Engl J Med 310(19):12131217

Kelschenbach J, Barke RA, Roy S (2005) Morphine withdrawalcontributes to Th cell differentiation by biasing cells toward theTh2 lineage. J Immunol 175(4):26552665

Khavari KA, Peters TC, Baity PL, Wilson AS (1975) Voluntarymorphine ingestion, morphine dependence, and recovery fromwithdrawal signs. Pharmacol Biochem Behav 3(6):10931096

Kreek MJ (1987) Tolerance and dependence: implications for thepharmacological treatment of addiction. NIDA Res Monogr76:5362

Kreek MJ (1990) Immune function in heroin addicts and formerheroin addicts in treatment: pre- and post-AIDS epidemic. NIDARes Monogr 96:192219

Kreek MJ, Khuri E, Flomenberg N, Albeck H, Ochshorn M (1990)Immune status of unselected methadone maintained formerheroin addicts. Prog Clin Biol Res 328:445 448


Lugo RA, Kern SE (2002) Clinical pharmacokinetics of morphine. JPain Palliat Care Pharmacother 16(4):518

McMillan DE, Leander JD, Wilson TW, Wallace SC, Fix T, ReddingS, Turk RT (1976) Oral ingestion of narcotic analgesics by rats. JPharmacol Exp Ther 196(2):269279

Pechnick RN (1993) Effects of opioids on the hypothalamopituitaryadrenal axis. Annu Rev Pharmacol Toxicol 33:353382

Pierce H, Plant OH (1928) Studies in chronic morphine poisoning indogs II. Changes in blood cells and hemoglobin during addictionand withdrawal. J Pharmacol Exp Ther 33(3):359370

Rahim RT, Adler MW, Meissler JJ Jr, Cowan A, Rogers TJ, Geller EB,Eisenstein TK (2002 Abrupt of precipitated withdrawal frommorphine induces immunosuppression. J Neuroimmunol 127(12):8895

Rahim RT, Meissler JJ, Zhang L, Adler MW, Rogers TJ, Eisenstein TK(2003) Withdrawal from morphine in mice suppresses splenicmacrophage function, cytokine production, and costimulatorymolecules. J Neuroimmunol 144(12):1627

Rahim RT, Feng P, Meissler JJ, Rogers TJ, Zhang L, Adler MW,Eisenstein TK (2004) Paradoxes of immunosuppression in mousemodels of withdrawal. J Neuroimmunol 147(12):114 120

Rane A, Sawe J, Lindberg B, Svensson JO, Garle M, Erwald R, JorulfH (1984) Morphine glucuronidation in the rhesus monkey: acomparative in vivo and in vitro study. J Pharmacol Exp Ther 229(2):571576

Rouveix B (1992) Opiates and immune function. Consequences on in-fectious diseases with special reference to AIDS. Therapie 47(6):503512

SAMSHA (2002) National Survey on Drug Use and Health. www.samsha.gov

Seevers MH (1936) Opiate addiction in the monkey. I. Methods ofstudy. J Pharmacol Exp Ther 56(2):147156

Shavit Y, Terman GW, Lewis JW, Zane CJ, Gale RP, Liebeskind JC(1986) Effects of footshock stress and morphine on natural killerlymphocytes in rats: studies of tolerance and cross-tolerance.Brain Res 372(2):382385

Stanski DR, Greenblatt DJ, Lowenstein E (1978) Kinetics of intravenousand intramuscular morphine. Clin Pharmacol Ther 24(1):5259

Tang AH (1982) The etonitazene-dependent rhesus monkey as amodel to study narcotic agonist and antagonist activities. NIDARes Monogr 41:200207

Turkkan JS, Ator NA, Brady JV, Craven KA (1989) Beyond chroniccatheterization in laboratory primates. In: Segal EF (ed) Housing,care and psychological wellbeing of captive and laboratoryprimates. Noyes Publications, Park Ridge, NY, pp. 305324

Weber RJ, Gomez-Flores R, Smith JE, Martin TJ (2004) Immune,neuroendocrine, and somatic alterations in animal models ofhuman heroin abuse. J Neuroimmunol 147(12):134 137

West JP, Lysle DT, Dykstra LA (1997) Tolerance development tomorphine-induced alterations of immune status. Drug AlcoholDepend 46(3):147157

West JP, Dykstra LA, Lysle DT (1998a) Differential tolerance tomorphines immunomodulatory effects following continuousadministration. Drug Alcohol Depend 53:3138

West JP, Lysle DT, Dykstra LA (1998b) Time-dependent immuno-modulatory effects following natural opoiod withdrawal. NIDARes Monogr 178:201

West JP, Dykstra LA, Lysle DT (1999) Immunomodulatory effects ofmorphine withdrawal in the rat are time dependent and reversibleby clonidine. Psychopharmacology (Berl) 146(3):320327

Zajicova A, Wilczek H, Holan V (2004) The alterations of immunol-ogical reactivity in heroin addicts and their normalization inpatients maintained on methadone. Folia Biol (Praha) 50(1):24 28


Morphine Withdrawal Dramatically Reduces Lymphocytes in Morphine-Dependent MacaquesAbstractIntroductionMethodsMorphine self-dosingMorphine withdrawal proceduresImmune parametersData analysis

ResultsOral morphine dependenceMorphine withdrawal syndrome

DiscussionReferences

morphine withdrawal dramatically reduces lymphocytes in morphine-dependent macaques

Documents