In the present study, we have demonstrated (i
) that morphine is present in human neutrophils and is colocalized with lactoferrin, an antimicrobial peptide released during infection 
from the secondary granules of neutrophils, (ii
) that LPS and IL-8 trigger the Ca2+
-dependent secretion of morphine and lactoferrin in vitro
) that endogenous morphine concentrations are significantly increased in the serum of patients with septic conditions, (iv
) that the concentration of morphine found in sepsis inhibits IL-8 release from neutrophils and finally (v
) that μ opioid receptor surface expression is increased after stimulation of white cells with LPS.
Our findings provide further evidence for the role of endogenous morphine 
as a signaling molecule that can act through μ opioid receptors expressed on immune cells. The activation of the morphinergic system as demonstrated by an increase in morphine secretion as well as μ opioid receptor expression by neutrophils, underlines the hypothesis that endogenous morphine is part of the innate immune response to stress, and in particular, infection. In addition, our data suggest a new link between endogenous morphine expression and host defense against infection.
During gram-negative sepsis bacterial LPS leads to the activation of PMN and triggers the synthesis and secretion of proteases, cytokines (e.g
., IL-6 and IL-8), and toxic radicals from immune cells 
. Among these cytokines, IL-8 acts as a potent stimulator of PMN and represents a major chemoattractant for these cells 
. PMN participate in the host defense via
the secretion of particular antimicrobial peptides including lactoferrin, which is present in secondary granules and is released upon LPS stimulation 
. Despite their important role in destroying the pathogen, compounds released by PMN inflict damage to host cells 
and contribute to organ dysfunction during systemic infections. Thus, regulatory mechanisms including the hypothalamic–pituitary–adrenal axis 
, as well as PMN autoregulatory mechanisms such as apoptosis 
, are necessary to prevent local tissue destruction 
. Recently, the μ opioid receptor agonist DAGO in concentrations as low as 10 nM was shown to significantly decrease the secretion of IL-8 from neutrophils after LPS stimulation 
. We now demonstrate that LPS and IL-8 trigger the secretion of morphine from human PMN in vitro
. Our results also indicate that low concentrations of morphine (8 nM coresponding to 2.25 ng/ml) consistent with the amount of morphine found in the serum of septic patients, significantly decrease the secretion of IL-8 from human neutrophils and that this inhibition can be reversed by naloxone. Therefore, it is tempting to speculate that morphine present in the serum of septic patients is secreted from neutrophils and is involved in the regulation of immunocyte activity. Interestingly, a recent paper reported that concentrations of morphine as low as 10–100 nM enhanced migration of primary microglial cells toward adenosine diphosphate. This effect was reversed by naloxone, as well as CTAP, indicating μ opioid receptor involvement 
. These data strongly suggest that release of endogenous morphine may affect immune cells either directly or via
central regulatory mechanisms.
The presence of morphine in human PMN and mononuclear cells is a matter of dispute: Whilst the presence of morphine was described by Zhu et al 
, Boettcher and colleagues 
failed to demonstrate morphine production within neutrophils. In the present study, we have unambiguously demonstrated the presence of morphine in PMN, however, the amount of morphine found was well below the concentration (0.32+/−0.2 pg/million PMN ± SD vs
12.33±5.64 pg/million cells ± SEM) described before 
. A very low amount of morphine was previously found in erythrocytes by Boettcher and colleagues 
, however, we failed to show the presence of morphine-like immunoreactivity in these cells in our experiments.
Atypical cholinergic regulation of morphine secretion from human white blood cells has been demonstrated before 
. Our present study shows that morphine is present in secondary granules and is secreted from neutrophils in response to LPS and IL-8 exposure, suggesting that at least part of the circulating morphine in the blood of septic patients stems from PMN. Interestingly, this is supported by the observation that LPS administration in animals dramatically increases the amount of circulating 
and cerebral endogenous morphine 
. However, the low amount of morphine produced and secreted by neutrophils seems not sufficient to reach the morphine concentrations observed during sepsis. Thus, during systemic infection morphine release from the adrenal gland 
, its delayed glucoronidation in the liver 
and morphine synthesis in the nervous system 
may potentially contribute to elevated morphine levels in the systemic circulation.
Morphine preferentially binds to μ opioid receptors, a large family of seven-transmembrane receptors derived from a single OPRM1 (opioid receptor μ) gene. However, endogenous opioid peptides such as β-endorphine, endomorphines and enkephalins are also able to bind to μ opioid receptors. Serum levels of endogenous opioid peptides are elevated in response to stress and also in septic patients 
. Treatment with opiate antagonists for the cardiovascular effects of septic shock was proposed more than 20 years ago 
, indicating a key role of opioid peptides in the stress response. It is likely that part of this response is mediated via
opioid receptors expressed in different tissues including endothelia, leukocytes and myocardium. Endogenous opioid peptides present in the serum of septic patients may compete with morphine for μ opiate receptor binding. However, no effects of these interactions have been described yet in septic patients. Thus, further studies are warranted to investigate how endogenous opioid peptides may influence or possibly regulate morphine-induced neutrophil inhibition in response to stress and infection.
Stimulation of μ opioid receptors (for review: 
), triggers various effects including direct inhibition of immunocyte function. Human granulocytes and monocytes express μ opioid-like receptors as demonstrated by specific and saturable binding of dihydromorphine 
. μ opioid receptors were also found on human and monkey lymphocytes 
. A specific μ opioid receptor isoform, the μ3 receptor (morphine selective but opioid peptide insensitive) has also been characterized in human monocytes and granulocytes as well as in endothelial cells 
. The existence of numerous splicing variants underpins the complexity of the morphinergic system 
. To date, fourteen human μ opioid receptor isoforms have been identified 
with various affinities for morphine and its derivatives (Ki in the nanomolar range with low affinity for the recently described isoform mMOR-1B4 
), suggesting that different morphine concentrations are needed for specific receptor isoform activation. In addition, homo- or hetero-dimerisation of μ opioid receptor subtypes was recently found to occur and to induce switch signaling 
Although it has been shown before that LPS increases μ opioid receptor expression in the rat mesenteries 
, to our knowledge our results demonstrate for the first time that μ opioid receptor surface expression on neutrophils is increased after stimulation with LPS. The expression of μ opioid receptors is well documented for immune and endothelial cells 
. Amongst other opioid receptor subtypes, PMN express the μ3 opioid receptor subtype coupled to constitutive nitric oxide (NO) release 
. Due to its immunoinhibitory nature 
, it has been postulated that endogenous morphine down-regulates immune function to maintain a balance between pro- and anti-inflammatory processes. In this regard, morphine is known to dramatically affect innate and adaptive immune responses 
. For example, morphine affects lymphocyte proliferation 
, natural killer T-cell activities 
, antibody production 
, and the number of circulating leukocytes 
. Inhibition of phagocyte function, such as chemotaxis, phagocytosis, and surface receptor expression after morphine exposure in vitro
is, in part, a consequence of NO synthesis 
and involves the modulation of the transcription factors NF-κB and AP-1 
. In addition, recent data on rat microglial cells treated with adenosine diphosphate have shown that low concentrations of morphine induce a biphasic effect, with an initial phase involving PI3K/Akt pathway activation that leads to an increase of cell migration 
, followed by a longer-term phase with increased expression of Iba1 and P2X4 receptor protein, leading to a greater cell migration.
Furthermore, different μ opioid receptor splicing variants are expressed by PMN, suggesting that these μ receptor subtypes may mediate neutrophil responses during infection. Interestingly, the μ opioid receptor possesses an affinity for morphine with an IC50 around 10nM 
compatible with the concentration of morphine observed in the patients with sepsis (8nM or 2.5ng/ml) suggesting that auto- and/or paracrine regulation involving immune cells and possibly the endothelium exists.
Several studies have linked endogenous morphine with inflammatory conditions. Endogenous morphine was found in the plasma of patients after cardiovascular bypass which induces a well-described systemic inflammatory response 
. Interestingly, our study shows that morphine levels found in SIRS patients (mean between 0,39 and 0.57 ng/ml) were comparable to those described after cardiovascular bypass (mean of 0.75 ng/ml) 
, wheras morphine was not detected in healthy controls.
, which is a condition associated with a systemic inflammatory response caused by bacterial, fungal or viral infection, is one of the main causes of mortality in intensive care units 
. The diagnosis of sepsis is based on clinical symptoms 
, although several endogenously produced proteins, such as procalcitonin or C-reactive protein (CRP), have been proposed as markers of infection and are used in clinical practice 
Our study demonstrates a significant increase of serum morphine levels in patients with severe sepsis or septic shock, while in critically ill patients suffering from SIRS and controls, morphine levels were significantly lower. No significant difference was observed between patients from the three septic groups, but morphine was undetectable in the serum of healthy donors. Our results indicate that endogenous morphine might represent a new biomarker for the diagnosis of sepsis, however, an extended cohort of patients is needed to validate this hypothesis.
Opioid analgesia is a standard component of sedation in critical care, with synthetic opiates such as fentanyl and its derivatives being most frequently used. The clinical significance of sedation-induced immunosppression is still under debate. So far, no direct effect of fentanyl in clinically relevant concentrations could be demonstrated on neutrophil functions and fentanyl does not bind to opioid receptors expressed on neutrophils 
. It remains unclear whether the application of exogenous opiates interferes with the signaling of endogenous morphine via
direct or centrally mediated mechanisms.
Interestingly, our data reveal that concentrations of morphine found in the serum of septic patients are in range with the affinity of μ opioid receptors 
. This correlation suggests a putative physiological role during sepsis. Based on morphine's immunoinhibitory activity, it is tempting to propose that during sepsis, increased morphine levels may down-regulate immune and vascular tissue responses to prevent organ damage caused by excessive inflammation. However, recent data suggest that low concentrations of morphine (10–100 nM), synergistically with other factors increase immune activity (i.e.,
. In addition, several studies in rodents have shown that morphine in clinically relevant doses exacerbates inflammation and increases sepsis mortality 
. Thus, we can not exclude that the increase of endogenous morphine may also act in favor of proinflammatory responses in certain cells. As demonstrated in our study, endogenous morphine concentrations are in the nanomolar range while morphine concentrations after administration of clinically relevant doses for analgosedation in intensive care are 10 to 100-fold higher 
. It is unclear so far whether the different concentration range of exogenous and endogenous morphine affects neutrophil function in contrasting ways. Morphine levels can be related, in part to PMN secretion, but also to secretion from other organs including adrenal gland and liver. Therefore, direct monitoring of neutrophil function in relation to the plasma morphine concentrations is required to eventually answer this question.
Taken together, our present observations together with previously published data, support the hypothesis that endogenous morphine represents a modulator of immune cell activity. Our data indicate for the first time that endogenous morphine is secreted from human neutrophils after LPS or IL-8 stimulation. In patients with generalized infection such as sepsis, severe sepsis, and septic shock serum levels of endogenous morphine were found to be dramatically increased, whereas inflammatory states without infection such as SIRS did not lead to profound endogenous morphine rises. In addition, our results indicated that low concentrations of morphine inhibit LPS-induced IL-8 secretion by neutrophils. Further investigations, including clinical studies, are required to fully understand the effects of circulating morphine on immune function in sepsis.