Interleukin-10 stimulation of endogenous nitric oxide release from human saphenous veins diminishes immunocyte adherence

The Role of Interleukin-10 in the Pathophysiology of Preeclampsia

PURPOSE OF REVIEW

The pathophysiology of preeclampsia is complex and not entirely understood. A key feature in preeclampsia development is an immunological imbalance that shifts the maternal immune response from one of tolerance towards one promoting chronic inflammation and endothelial dysfunction. As a key regulator of immunity, IL-10 not only has immunomodulatory activity, but also directly benefits vasculature and promotes successful cellular interactions at the maternal-fetal interface. Here we focus on the mechanisms by which the dysregulation of IL-10 may contribute to the pathophysiology of preeclampsia.

RECENT FINDINGS

Dysregulation of IL-10 has been demonstrated in various animal models of preeclampsia. Decreased IL-10 production in both placenta and peripheral blood mononuclear cells has been reported in human studies, but with inconsistent results. The significance of IL-10 in preeclampsia has shifted from a key biomarker to one with therapeutic potential. As such, a better understanding of the role of this cytokine in the pathophysiology of preeclampsia is of paramount importance.

Evidence that scabies mites (Acari: Sarcoptidae) influence production of interleukin-10 and the function of T-regulatory cells (Tr1) in humans

We performed experiments to determine whether an extract of Sarcoptes scabiei (De Geer) influenced cytokine expression by human T-lymphocytes. Peripheral blood mononuclear cells from five sensitized donors and four donors without sensitization to scabies mites were challenged with a T-cell mitogen alone, with scabies extract (SS) alone, or with mitogen and SS together. Supernatants were analyzed for the cytokines interferon-gamma (IFNgamma), interleukin (IL)-2, IL-4, and IL-10. No IL-2 or IL-4 was produced in response to scabies extract. Cells from both naive and sensitized donors produced large amounts of IFNgamma and IL-10. The lack of IL-4 but high levels of IL-10 suggests that IL-10 was likely secreted by type 1 T-regulatory cells, which were activated by something in the scabies extract. IL-10 has anti-inflammatory and immune-suppressive effects. It may play a key role in depressing the inflammatory and immune responses in humans so that clinical symptoms are not seen until 4-6 wk after a person becomes infested with scabies mites.

Placental endothelial nitric oxide synthase localization and expression in normal human pregnancy and pre-eclampsia

1. The aim of the present study was to investigate whether pre-eclampisa, a state of placental hypoxia, is associated with placental abnormalities in the amount, distribution and expression of endothelial nitric oxide synthase (eNOS). 2. Localization and intensity of eNOS was determined by immunohistochemistry using an antibody specific for eNOS. The amount of eNOS mRNA expression was determined by reverse transcription-polymerase chain reaction (RT-PCR) and the densitometry of gel bands was expressed as a ratio of the band density of the housekeeping gene beta2-microglobulin. 3. Endothelial NOS staining was localized to syncytiotrophoblast cells within the villi and decidual trophoblast cells. It was not present in the endothelium of terminal villous vessels. There was no significant difference in eNOS villous or decidual staining intensity between normal pregnancy (NP; n = 12), pre-eclampsia (n = 14), or gestational hypertension (GH; n = 4). Staining for eNOS was not significantly different in the decidua compared with the villi in NP, GH or pre-eclampsia. Within the decidua, the depth of eNOS staining was similar in NP, pre-eclampisa and GH. 4. There was no significant difference in eNOS mRNA expression between NP (0.70 +/- 0.11), pre-eclampsia (0.5 +/- 0.07) or GH (0.69 +/- 0.26). 5. These findings suggest that the amount of eNOS in the placenta is not deficient in pre-eclampsia, excluding a possible pathogenic role for eNOS in this disease. Furthermore, placental hypoxia, which is associated with pre-eclampsia, did not induce an upregulation of eNOS

Evidence for a spontaneous nitric oxide release from the rat median eminence: influence on gonadotropin-releasing hormone release

The involvement of nitric oxide (NO) as a gaseous neurotransmitter in the hypothalamic control of pituitary LH secretion has been demonstrated. NO, as a diffusible signaling gas, has the ability to control and synchronize the activity of the neighboring cells. NO is secreted at the median eminence (ME), the common termination field for the antehypophysiotropic neurons, under the stimulation of other signaling substances. At the ME, NO stimulates GnRH release from neuroendocrine terminals. The present studies were undertaken to determine whether NO is secreted spontaneously from ME fragments ex vivo and whether its secretion is correlated to GnRH release. To accomplish this, female rats were killed at different time points of the day and/or of the estrous cycle. The spontaneous NO release was monitored in real time, with an amperometric probe, during 4 periods of 30 min, from individual ME fragments (for each time point, n = 4). GnRH levels were measured in parallel for each incubation-period by RIA. The results revealed that NO was released in a pulsatile manner from female ME fragments and, unambiguously, that the amplitude of NO secretion varied markedly across the estrous cycle. Indeed, though the NO pulse period (32 +/- 1 min, n = 36) and duration (21 +/- 2 min, n = 36) did not vary significantly across the estrous cycle, the amplitude of this secretion pulse was significantly higher on proestrus (Pro; 39 +/- 3 nM, n = 20), compared with diestrus (16 +/- 1 nM, n = 8) or estrus (23 +/- 3 nM, n = 8, P < 0.05). The GnRH levels in the incubation medium were positively correlated to NO secretion across the estrous cycle (r = 0.86, P < 0.003, n = 9), confirming that NO and GnRH release are coupled. Furthermore, 5 x 10(-7) M L-N(5)-(1-iminoethyl)ornithine (L-NIO), a NO synthase inhibitor, succeeded in inhibiting the strong NO-GnRH secretory coupling and GnRH release on PRO: Because at this concentration, L-NIO selectively inhibits endothelial NO synthase, the results further demonstrate that the major source of NO involved in GnRH release at the ME is endothelial in origin. Additionally, the induction of a massive NO/GnRH release in 15-day ovariectomized rat treated with estradiol benzoate strongly suggested that estradiol is participating in the stimulation of NO release activity between diestrus II and PRO: The present study is the first demonstrating that ME can spontaneously release NO and that NO's rhythm of secretion varies markedly across the estrous cycle. This pulsatile/cyclic ME NO release may constitute the synchronizing link to anatomically scattered GnRH neurons.

Endogenous morphine is produced in response to cardiopulmonary bypass in neonatal pigs

BACKGROUND

Cardiopulmonary bypass (CPB) is associated with a systemic inflammatory response. Endogenous morphine production has previously been demonstrated in humans after cardiac surgery with CPB. It has been hypothesized that morphine plays a role as an anti-inflammatory mediator in the systemic inflammatory response. The aim of this study was to investigate if the CPB procedure in itself elicits an endogenous morphine production in neonatal pigs.

METHODS

Endogenous morphine production was measured in arterial blood in piglets exposed to sternotomy alone (sham group, n=10) or sternotomy and CPB (n=10). Blood samples were obtained immediately after the induction of anaesthesia, at the end of CPB and 4 h later. Morphine in arterial blood was detected by radioimmunoassay and confirmed by gas chromatography mass spectrometry.

RESULTS

Animals undergoing CPB showed detectable endogenous morphine concentrations immediately after CPB, with increased concentrations postoperatively. There was no measurable morphine production in the sham operated pigs.

CONCLUSION

The CPB procedures elicits an endogenous morphine production in neonatal pigs. This morphine response is analogous to the previously demonstrated response in patients subjected to cardiac surgery and CPB.

The dual personality of NO

In the body, nitric oxide (NO) is an important physiological regulator of functions such as vasodilatation and neurotransmission. Under pathological conditions, high concentrations of NO can be either beneficial(e.g. anti-bacterial, anti-parasitic and anti-viral) or detrimental; NO can therefore be considered a double-edged sword. When manipulating NO levels clinically, attention should be paid to minimize the negative effects and maximize the beneficial effects of NO. This article highlights recent evidence that supports the complexity of the regulatory mechanisms that lead to sophisticated endogenous NO production.

Basal nitric oxide limits immune, nervous and cardiovascular excitation: human endothelia express a mu opiate receptor

Nitric oxide (NO) is a major signaling molecule in the immune, cardiovascular and nervous systems. The synthesizing enzyme, nitric oxide synthase (NOS) occurs in three forms: endothelial (e), neuronal (n) and inducible (i) NOS. The first two are constitutively expressed. We surmise that in many tissues there is a basal level of NO and that the actions of several signaling molecules initiate increases in cNOS-derived NO to enhance momentary basal levels that exerts inhibitory cellular actions, via cellular conformational changes. It is our contention that much of the literature concerning the actions of NO really deal with i-NOS-derived NO. We make the case that cNOS is responsible for a basal or 'tonal' level of NO; that this NO keeps particular types of cells in a state of inhibition and that activation of these cells occurs through disinhibition. Furthermore, naturally occurring signaling molecules such as morphine, anandamide, interleukin-10 and 17-beta-estradiol appear to exert, in part, their beneficial physiological actions, i.e., immune and endothelial down regulation by the stimulation of cNOS. In regard to opiates, we demonstrate the presence of a human endothelial mu opiate receptor by RT-PCR and sequence determination, further substantiating the role of opiates in vascular coupling to NO release. Taken together, cNOS derived NO enhances basal NO actions, i.e., cellular activation state, and these actions are further enhanced by iNOS derived NO.

Essential role of nitric oxide and interferon-gamma for tumor immunotherapy with interleukin-10

Several laboratories have reported anti-tumor activity for high levels of interleukin-10 (IL-10) expressed as a transgene or administered as recombinant protein. The authors have reported a positive correlation for nitric oxide production and anti-tumor activity of IL-10 in a murine model of breast cancer. In the current study, they sought evidence of a mechanistic role for nitric oxide in IL-10-mediated tumor growth inhibition. They wanted to determine whether pharmacologic inhibition of nitric oxide synthase (NOS) activity reverses the therapeutic effect of IL-10. Administration of either of two NOS inhibitors, aminoguanidine (AG) or L-lysine,N6-1-iminoethyl-dihydrochloride, appears to abrogate in part the tumor growth inhibition observed when IL-10 is overexpressed as a transgene in two murine mammary tumor cell lines. Nitric oxide levels were assessed at the tumor site by measuring nitrosylated heme levels by electron spin resonance spectroscopy. Nitric oxide hemoglobin levels were lower in tumors from aminoguanidine-treated mice, indicating that effective inhibition of nitric oxide production occurred at the tumor site. Previous investigations showed that the inducible form of NOS protein (iNOS), but not constitutive NOS, was expressed at higher levels in IL-10-expressing tumors. Because iNOS is regulated at the transcriptional level, the authors compared iNOS mRNA levels in IL-10 and control tumors. Northern analysis revealed strong iNOS message expression in all six IL-10-expressing tumors examined, whereas message was faintly detected in parental or 66-neo tumors. The inducible form of NOS is responsive to induction by interferon-gamma (IFN-gamma). The role of IFN-gamma in IL-10-mediated tumor inhibition and iNOS mRNA induction was determined. When tumors were transplanted to IFN-gamma mutant mice, the tumor-inhibitory activity of IL-10 was lost. Furthermore, iNOS mRNA was no longer induced in the absence of host expression of IFN-gamma. These data indicate that nitric oxide contributes to the anti-tumor activity of IL-10 and that expression of iNOS in this context depends on IFN-gamma.

NF-kappaB, nitric oxide and opiate signaling

NF-kappaB, a DNA binding factor, has been implicated in inflammatory cytokine activation. NF-kappaB is activated by IkappaBalpha, its inhibitor, which is phosphorylated and proteolytically degraded. In this regard, NF-kappaB is also responsive to reactive oxygen intermediates and calcium. Reports also have emerged that demonstrate that nitric oxide inhibits NF-kappaB transcriptional activation in a variety of cells, including monocytes and endothelial cells. Recently, we have demonstrated that morphine, not opioid peptides, via the mu3 opiate receptor is coupled to constitutive nitric oxide release in these same cells. In this regard, we provide a scenario whereby morphine modulates NF-kappaB activation via nitric oxide. This pathway appears to be the key step in regulating inducible nitric oxide synthase expression, controlling the balance between constitutive nitric oxide synthase and the inducible form.

Mu3 opiate receptor expression in lung and lung carcinoma: ligand binding and coupling to nitric oxide release

The mu3 opiate receptor subtype is expressed in human surgical specimens of both normal lung and non-small-cell lung carcinoma. Nitric oxide (NO) release is mediated through the mu3 receptor, and in lung carcinoma, morphine-stimulated NO release is significantly higher and prolonged than in normal lung. Using reverse transcriptase-polymerase chain reaction (RT-PCR) and Southern blot analysis we show that specific mu opioid receptor transcripts are present in lung carcinoma and other cells with the mu3 profile. Our findings identify a unique role for the mu3 opiate receptor in opiate-mediated NO release and suggest that endogenous opiates, through their release of NO, may play a role in cancer progression.

Estradiol coupling to endothelial nitric oxide stimulates gonadotropin-releasing hormone release from rat median eminence via a membrane receptor

The median eminence (ME), which is the common termination field for adenohypophysiotropic systems, has been shown to produce nitric oxide (NO), a signaling molecule involved in neuroendocrine secretion. Using an ex vivo technique, 17beta-estradiol exposure to ME fragments, including vascular tissues, stimulated NO release within seconds in a concentration-dependent manner, whereas 17alpha-estradiol or testosterone had no effect. 17Beta-estradiol conjugated to BSA (E2-BSA) also stimulated NO release, suggesting mediation by a membrane surface receptor. Tamoxifen, an estrogen receptor inhibitor, antagonized the action of both 17beta-estradiol and E2-BSA. Furthermore, estradiol-stimulated NO stimulates GnRH release. This was demonstrated by hemoglobin (a NO scavenger), N(omega)-nitro-L-arginine methyl ester, and L-N5-(1-iminoethyl)ornithine (nitric oxide synthase inhibitors) inhibition of estradiol stimulated NO and GnRH release. In this regard, L-N5-(1-iminoethyl)ornithine, specific for endotheliol constitutive nitric oxide synthase, was significantly more potent, suggesting that the estradiol-stimulated NO release arose from vascular endothelial cells. Additionally, the NO-stimulated GnRH release occurs via guanylyl cyclase activation in GnRH nerve terminals, as ODQ, a potent and selective inhibitor of NO-sensitive guanylyl cyclase, abolished the estradiol-stimulated GnRH release. The results suggest that at physiological concentrations, 17beta-estradiol may have immediate actions on ME endothelial cells via nongenomic signaling pathways leading to NO-stimulated GnRH release.

Interleukin-10 stimulation of corticotrophin releasing factor median eminence in rats: evidence for dependence upon nitric oxide production

Following treatment with interleukin-10 (IL-10), basal corticotrophin releasing factor (CRF) levels from rat hypothalamic median eminence (ME) were found to be increased. Our data show: (1) the specificity of stimulation of CRF through the use of recombinant IL-10 and its blockage by monoclonal anti-IL-10 antibody; (2) the requirement of NO in this process through the use of N(omega)-nitro-L-arginine methyl ester (L-NAME), a nitric oxide synthase inhibitor; (3) the blockage of IL-10 stimulated NO production by anti-IL-10; and, (4) the presence of IL-10 transcripts in hypothalamic poly A+ mRNA. These results provide the first evidence of IL-10 acting in the ME to influence CRF levels and further support our earlier findings of a potential for IL-10 in the hypothalamic-pituitary axis.

Leukocyte-suppressing influences of interleukin (IL)-10 in cardiac allografts: insights from IL-10 knockout mice

To investigate the role of interleukin (IL)-10 in late graft outcomes, we compared BALB/c donor hearts transplanted into immunosuppressed wild-type or IL-10 gene-deficient (-/-) C57BL recipients (n = 49) at 50 +/- 5 days. There was prominent leukocyte infiltration and parenchymal destruction with more severe vascular occlusion in grafts from IL-10 -/- recipients. An occlusive CD45+ arteritis with medial necrosis occurred with IL-10 deficiency instead of the a-smooth muscle actin-rich arteriosclerosis seen in wild-type recipients. Increased interferon (IFN)-gamma as well as Mac-1, inducible nitric oxide synthase, and allograft inflammatory factor-1 (but not CD3 and IL-4) transcript levels were seen in allografts from IL-10 -/- recipients as assessed by 32p reverse transcription polymerase chain reaction. We then evaluated the contribution of IFN-gamma-mediated responses by neutralizing IFN-gamma. Anti-IFN-gamma monoclonal antibody (MAb) treatment of IL-10 -/- recipients did not improve graft survival, parenchymal rejection, or occlusive arteritis, indicating that these processes are IFN-gamma independent. However, medial smooth muscle cell loss in IL-10 -/- recipients was attenuated by anti-IFN-gamma MAb. Hence, in this transplant model, IL-10 suppresses T cell and macrophage responses in the parenchyma and vasculature and confers a protective effect against late rejection.

Borrelia burgdorferi and interleukin-1 promote the transendothelial migration of monocytes in vitro by different mechanisms

A prominent feature of Lyme disease is the perivascular accumulation of mononuclear leukocytes. Incubation of human umbilical vein endothelial cells (HUVEC) cultured on amniotic tissue with either interleukin-1 (IL-1) or Borrelia burgdorferi, the spirochetal agent of Lyme disease, increased the rate at which human monocytes migrated across the endothelial monolayers. Very late antigen 4 (VLA-4) and CD11/CD18 integrins mediated migration of monocytes across HUVEC exposed to either B. burgdorferi or IL-1 in similar manners. Neutralizing antibodies to the chemokine monocyte chemoattractant protein 1 (MCP-1) inhibited the migration of monocytes across unstimulated, IL-1-treated, or B. burgdorferi-stimulated HUVEC by 91% +/- 3%, 65% +/- 2%, or 25% +/- 22%, respectively. Stimulation of HUVEC with B. burgdorferi also promoted a 6-fold +/- 2-fold increase in the migration of human CD4(+) T lymphocytes. Although MCP-1 played only a limited role in the migration of monocytes across B. burgdorferi-treated HUVEC, migration of CD4(+) T lymphocytes across HUVEC exposed to spirochetes was highly dependent on this chemokine. The anti-inflammatory cytokine IL-10 reduced both migration of monocytes and endothelial production of MCP-1 in response to B. burgdorferi by approximately 50%, yet IL-10 inhibited neither migration nor secretion of MCP-1 when HUVEC were stimulated with IL-1. Our results suggest that activation of endothelium by B. burgdorferi may contribute to formation of the chronic inflammatory infiltrates associated with Lyme disease. The transendothelial migration of monocytes that is induced by B. burgdorferi is significantly less dependent on MCP-1 than is migration induced by IL-1. Selective inhibition by IL-10 further indicates that B. burgdorferi and IL-1 employ distinct mechanisms to activate endothelial cells.

Long-term exposure of human blood vessels to HIV gp120, morphine, and anandamide increases endothelial adhesion of monocytes: uncoupling of nitric oxide release

Acute exposure of human saphenous vein or internal thoracic artery endothelium to either morphine [27.4 +/- 3.7 and 35.4 +/- 4.1 nM nitric oxide (NO), respectively] or anandamide (18.3 +/- 2.2 and 24.3 +/- 3.0 nM, respectively) results in NO release, whereas exposure to the human immunodeficiency virus envelope protein gp120 does not. After the short-term exposure of the vessel endothelium, monocyte adherence is diminished with morphine and anandamide treatment (jointly by -80%), whereas it is enhanced with that of gp120 (approximately 40%), indicating that gp120 enhances the ability of the endothelium to adhere monocytes. Long-term or continuous exposure of the endothelia to all agents results in a significant enhancement of monocyte adherence (p < 0.05), which is further increased when exposed to either morphine and anandamide plus gp120. This is caused by a desensitization of the endothelium to further NO release after the initial exposure to either anandamide or morphine. The results serve to indicate that in individuals abusing opiates and or cannabinoids, a tissue [i.e., central nervous system (CNS)] viral load may be higher, and acquired immunodeficiency syndrome (AIDS) may progress more rapidly because monocyte adherence and mobility is significantly increased, indicating a higher level of transmembrane migration.

Delta2 opioid receptor subtype on human vascular endothelium uncouples morphine stimulated nitric oxide release

We demonstrate the presence of both delta and mu opioid receptors on the endothelium of human saphenous vein and internal thoracic artery. Displacement analysis revealed that a variety of opioid peptides were found to be ineffective in displacing specifically bound 3H dihydromorphine and only delta2 ligands were effective in regard to 3H Ala2-met5 enkephalinamide (DAMA), indicating the presence of mu3 and delta2 opioid receptor sites, respectively. Confirming the presence of both mu and delta sites we demonstrated positive immunostaining with anti-delta and anti-mu receptor antibodies. Exposure of these vessels to DAMA significantly enhances granulocyte adherence (P<0.01) even in vessels 5 min later exposed to 10(-6) M morphine. Unlike morphine, DAMA did not stimulate nitric oxide from either blood vessel and human granulocytes. Additionally, DAMA preadministered before morphine exposure to the endothelium or granulocytes, inhibited the morphine-stimulated release of NO in a dose-dependent manner. The data indicate that opioid peptides and opiate alkaloids regulate endothelial function in an antagonistic manner thereby influencing the microvascular environment.

Macrophage behavior associated with acute and chronic exposure to HIV GP120, morphine and anandamide: endothelial implications

We demonstrate that immediate exposure to gp120 (5 min; 0.1 microg/ml) results in a significant shift of the macrophage population to an amoeboid and motile category (P<0.01; 91.7+/-5.5 vs. a control value of 42.4+/-4.2) and prior exposure with anti-gp120 antagonizes this shift. Acute exposure of the macrophages to morphine (10(-6) M) or anandamide (10(-6) M) resulted in the cells rounding up (shape factors of 0.84 and 0.87 respectively) and becoming non-motile. The action is blocked by prior treatment with the specific antagonists naloxone and SR 141716A. Chronic exposure (6 h) of the cells to all three agents resulted in a random migration pattern. Further, all agents blocked chemotaxis induced by DAMA and IL-1. Observation of the cells behavior during chronic exposure revealed a sporadic activity pattern with gp120 whereas morphine and anandamide first induced a period of inactivity which is followed by a period of activity (chemokinesis). Recent work from our laboratory has demonstrated that both morphine and anandamide acutely stimulate constitutive macrophage nitric oxide (NO) release, which then induces macrophage rounding and inactivity. It was therefore of interest to examine their behavior by exposing macrophages to the NO-donor SNAP. In a concentration dependent manner SNAP exhibited the same behavioral actions as both substances of abuse. Given this, we next determined if macrophages exposed to gp120 would release NO. We demonstrated that NO was released only when exposed to morphine and anandamide not gp120. Thus. the chemokinetic inducing activities of these agents may be the basis for excitotoxin liberation in neural tissues and/or a higher viral load in various organ systems since cellular adherence and random migration are stimulated.

Morphine and anandamide coupling to nitric oxide stimulates GnRH and CRF release from rat median eminence: neurovascular regulation

Nitric oxide (NO) is involved in neurohormonal secretion from median eminence neuroendocrine nerve terminals. We report that stimulation of NO release from median eminence fragments including vascular tissues occurs by mu3 receptor activation by morphine, or by cannabinoid type 1 receptor activation by anandamide. The released levels of NO are lower after anandamide than after morphine stimulation. These processes can be blocked by L-NAME, a specific nitric oxide synthase inhibitor, by naloxone for the morphine-stimulated NO release, or SR 141716A, a specific CB1 receptor inhibitor, for the anandamide-stimulated NO release. Furthermore, morphine and anandamide, by this NO dependent process, influences neurohormonal release from median eminence nerve terminals within 10 min. Via this NO dependent process, morphine stimulates both GnRH and CRF release, whereas anandamide selectively stimulates GnRH release. These observations together with previous data suggest that morphine and the anandamide-stimulated NO originates from the vascular endothelium of the portal plexus. These results indicate that endothelial cells of the median eminence may be involved in the release of neurohormones.

Endogenous morphine levels increase following cardiac surgery as part of the antiinflammatory response?

Exogenous morphine downregulates the activity of immunocompetent cells such as lymphocytes, granulocytes and macrophages. Furthermore, morphine increases the secretion of CRH, ACTH and glucocorticoids, i.e. substances with inhibitory effects on the immune system. In the present study we tested the hypothesis that endogenous morphine production is increased as part of the antiinflammatory response to cardiac surgery. Sixteen patients submitted to elective coronary artery bypass grafting (CABG) surgery were randomized to either thoracic epidural analgesia combined with general anaesthesia (group I) or high-dose fentanyl anaesthesia (group II). Patients in group I did not receive morphine while patients in group II received systemic morphine for postoperative pain relief. From each patient 18 blood samples were taken perioperatively and tested for morphine. Furthermore, monocyte function with respect to motility and shape was determined by computer-assisted image analysis. A steep increase in plasma morphine concentrations was demonstrated on the first postoperative day in patients in group I (not given morphine). Plasma morphine levels remained significantly elevated during the following five postoperative days. Patients in group II given morphine as pain treatment showed a larger and earlier morphine peak related to the morphine administration. Computer-assisted image analysis of leukocyte behaviour revealed a biphasic increase in cell motility. In conclusion, we demonstrate for the first time that endogenous morphine levels increase after the trauma of surgery. We surmise that morphine is part of the antiinflammatory response to cardiac surgery.