Inhibition of mammalian target of rapamycin aggravates the respiratory burst defect of neutrophils from decompensated patients with cirrhosis


  • Potential conflict of interest: Nothing to report.


Cirrhosis is commonly accompanied by impaired defense functions of polymorphonuclear leucocytes (PMNs), increased patient susceptibility to infections, and hepatocellular carcinoma (HCC). PMN antimicrobial activity is dependent on a massive production of reactive oxygen species (ROS) by nicotinamide adenine dinucleotide phosphate (NADPH) 2 (NADPH oxidase 2; NOX2), termed respiratory burst (RB). Rapamycin, an antagonist of mammalian target of rapamycin (mTOR), may be used in the treatment of HCC and in transplanted patients. However, the effect of mTOR inhibition on the PMN RB of patients with cirrhosis remains unexplored and was studied here using the bacterial peptide, formyl-Met-Leu-Phe (fMLP), as an RB inducer. fMLP-induced RB of PMN from patients with decompensated alcoholic cirrhosis was strongly impaired (30%-35% of control) as a result of intracellular signaling alterations. Blocking mTOR activation (phospho-S2448-mTOR) with rapamycin further aggravated the RB defect. Rapamycin also inhibited the RB of healthy PMNs, which was associated with impaired phosphorylation of the NOX2 component, p47phox (phox: phagocyte oxidase), on its mitogen-activated protein kinase (MAPK) site (S345) as well as a preferential inhibition of p38-MAPK relative to p44/42-MAPK. However, rapamycin did not alter the fMLP-induced membrane association of p47phox and p38-MAPK in patients' PMNs, but did prevent their phosphorylation at the membranes. The mTOR contribution to fMLP-induced RB, phosphorylation of p47phox and p38-MAPK was further confirmed by mTOR knockdown in HL-60 cells. Finally, rapamycin impaired PMN bactericidal activity, but not bacterial uptake. Conclusion: mTOR significantly up-regulates the PMN RB of patients with cirrhosis by p38-MAPK activation. Consequently, mTOR inhibition by rapamycin dramatically aggravates their PMN RB defect, which may increase patients' susceptibility to infection. Thus, concerns should be raised about the use of rapamycin in immuno-depressed patients. (HEPATOLOGY 2013)

Reactive oxygen species (ROS) produced by polymorphonuclear leukocytes (PMNs), monocytes or macrophages, termed respiratory burst (RB) or oxidative stress (OS), play a key role in antimicrobial host-defense systems.1 The enzyme responsible for the phagocyte RB, nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 2 (NOX2), is a membrane multiprotein complex whose activation requires the phosphorylation and membrane translocation of cytosolic components, among which p47phox (phox: phagocyte oxidase) plays an important role.2 In pathological situations, ROS production becomes inappropriately regulated. An excessive production of ROS induces tissue damage, which has been implicated in various diseases,1 including hepatic fibrosis.3 A deficient production of ROS promotes patients' susceptibility to microbial infections.1

Cirrhosis is a typical example in which inappropriate ROS production induces both tissue damage and patient susceptibility to infections.4 PMNs have been shown to contribute to liver injury in animal models5 and patients with alcoholic hepatitis.6 In these patients, the level of intrahepatic expression of “neutrophil-attractant” CXC chemokines, interleukin-8 and ENA-78 (CXCL5), have been shown to correlate with poor survival.7 Direct evidence for the importance of ROS in PMN-induced liver injury is provided by the observation of an intracellular OS in hepatocytes during the PMN infiltration8 and in p47phox knockout mice.3 A common complication of liver fibrosis is the development of sepsis, a major cause of death,9 which is associated with impaired PMN RB, microbicidal activity, and phagocytosis.10 PMN dysfunctions were found to be reversible after endotoxin removal from patient plasma.11 In other studies, persistent cellular defects were also observed.12 An impaired RB of PMN was also reported in liver transplant recipients suffering from posthepatitic cirrhosis.13

Rapamycin is used clinically for various purposes because of its ability to antagonize the kinase activity of mammalian target of rapamycin (mTOR). Inhibition of mTOR is under evaluation in patients with hepatocellular carcinoma (HCC).14, 15 Moreover, because there is some experimental evidence that mTOR is involved in portal hypertension (PH)-associated angiogenesis, it has been suggested that mTOR inhibtion could be a target for future therapies in PH.16, 17 Rapamycin is also used as an immunosuppressive drug to prevent rejection of transplanted organs.18 Rapamycin binds to the immunophilin, FKBP12, then inhibits mTOR, a protein kinase that plays an important role in protein synthesis, cell cycle, and cancer.19 mTOR exists at least in two multiprotein complexes.20 In one complex (mTORC1), mTOR is associated with Raptor and binds rapamycin. In the other complex (mTORC2), mTOR is associated with Rictor and cannot be directly inhibited by rapamycin.21 mTOR is activated by the protein kinase B (PKB or AKT) pathway22 and by phosphatidic acid generated by phospholipase D (PLD).19 We previously showed that AKT and PLD are two major signaling effectors in PMN and regulate NOX2 activity induced by the bacterial peptide, fMet-Leu-Phe (fMLP).23, 25, 26 However, whether mTOR up-regulates the RB of PMNs is unknown. If this were the case, rapamycin should aggravate the RB deficiency of PMNs from patients with cirrhosis, which may have clinical implications. To explore this hypothesis, the effect of mTOR inhibition was studied on RB and signaling events of PMNs from patients with decompensated alcoholic cirrhosis, using fMLP as an inducer.

This study shows that alcoholic cirrhosis strongly impaired the fMLP-induced RB of PMNs as a result of altered phosphorylation of a major NOX2 component, p47phox(S345), by mitogen-activated protein kinases (MAPKs). The results further show that mTOR is a novel effector of the PMN RB of control subjects and patients with cirrhosis. Consequently, mTOR inhibition by rapamycin dramatically aggravated the RB defect of PMNs of patients with cirrhosis through the inhibition of p38-MAPK signaling and phosphorylation of p47phox(S345). These results suggest that rapamycin should be used with caution in patients with cirrhosis.


AKT, protein kinase B; ERK1/2, extracellular signal regulated kinase 1/2; fMLP, formyl-Met-Leu-Phe; HCC, hepatocellular carcinoma; MAPK, mitogen-activated protein kinase; mTOR, mammalian target of rapamycin; NADPH, nicotinamide adenine dinucleotide phosphate; NOX2, NADPH oxidase 2; phox, phagocyte oxidase; OS, oxidant stress; PH, portal hypertension; PLC, phospholipase C; PLD, phospholipase D; PKC, protein kinase C; PMN, polymorphonuclear leucocyte; RB, respiratory burst; ROS, reactive oxygen species; S345, serine 345; siRNA, short interfering RNA.

Patients and Methods


Patients were hospitalized in the Liver Unit of Beaujon Hospital (Clichy, France). Inclusion criteria were age over 18 years, biopsy-proven cirrhosis, and Child-Pugh class B or C cirrhosis. Patients had a history of excessive alcohol ingestion (50 g/day), but no other causes of liver disease. Viral serologies for hepatitis B virus and hepatitis C virus were negative. Alcohol consumption was stopped for at least 3 days. Clinical characteristics of patients are shown in Table 1. Exclusion criteria were evidence of recent gastrointestinal bleeding, current bacterial infections, and treatment with corticosteroids, pentoxifylline, and other immunosuppressive drugs in the past 30 days, and presence of HCC, other cancer, or human immunodeficiency virus infection. Healthy subjects (controls) were hospital employee volunteers. The study was approved by our institutional review board, and written informed consent was obtained from all patients.

Table 1. Characteristics of Patients
  1. Plus minus values are means ± standard error of the mean.

  2. Abbreviation: MELD, Model of End-Stage Liver Disease.

No. of patients17
Age, years56 ± 3
Ascite, n (%)13 (76)
HCC, n (%)0 (0)
Encephalopathy, n (%)6 (36)
Serum albumin, g/L21.6 ± 1.6
Serum bilirubin, μmol/L168 ± 45
Prothrombin time, %45 ± 6
International normalized ratio2.06 ± 0.27
Serum creatinine, μmol/L68 ± 6
Child-Pugh score11 ± 1
Child-Pugh class C, n (%)13 (76)
MELD score20.4 ± 2.4
White blood cell count, per mm39,650 ± 2,090
Treatment with β-blockers, n (%)8 (47)

Because of space constraints, reagents and techniques used in this study are detailed in the Supporting Materials. These include cell-culture media, antibodies,24 mTOR short interfering RNA (siRNA) oligonucleotides,18 PMN isolation by sedimentation of Ficoll-Hypaque and Dextran,21 cell culture and transfection,25 PMN functional activities, such as RB studied by the cytochrome c reduction assay23 and chemiluminescence,25 phagocytosis of DsRed-conjugated Escherichia coli, PMN bactericidal activity, phosphorylation of signaling effectors, and membrane translocation of p47phox and p38-MAPK, which were studied by western blotting.26

Statistical Analysis.

Differences between means were identified using the Student paired t test or Mann-Whitney's U test, with a threshold of P < 0.05.


PMNs From Patients With Alcoholic Cirrhosis Exhibited an Impaired fMLP-Induced RB and p38-MAPK Activation.

PMNs from patients with decompensated alcoholic cirrhosis were stimulated with varying fMLP concentrations and their superoxide production by NOX2 was compared to that of PMNs from healthy volunteers (Fig. 1A). Suboptimal stimulation of PMNs with fMLP (25-50 nM) revealed a defective RB of PMNs from patients with cirrhosis. This dysfunction was aggravated under optimal PMN stimulation with 0.1-1.0 μM fMLP, resulting in a total superoxide production of only approximately 35% of control. This severe dysfunction was also confirmed in whole blood, in which fMLP-induced RB was measured by chemiluminescence (Supporting Fig. 1). This defective RB was associated with a significant decreased phosphorylation of the NOX2 component, p47phox, on its MAPK site, serine 345 (S345) (Fig. 1B,C). This site was previously shown to regulate NOX2 activity.24 Consistent with this observation, the fMLP-induced activation of p38-MAPK was also impaired in patients' PMNs (Fig. 1B,D; P < 0.05). The intensity and duration of NOX2 activity induced by fMLP in healthy PMNs can be potentiated by various agents, such as cytochalasin B,27 a fungal toxin that depolymerizes actin filaments. Interestingly, the deficient RB of patients' PMNs was not potentiated by cytochalasin B, unlike the RB of healthy PMNs (Fig. 1E), suggesting that biochemical alterations of cirrhotic PMNs may affect signaling events regulated by the cytoskeleton.

Figure 1.

PMN from patients with cirrhosis exhibited impaired fMLP-induced RB. (A) PMNs from healthy volunteers (control) and patients with cirrhosis were stimulated with fMLP (25-1,000 nM). RB is expressed in nmole/106 cells (n = 12 in each group; *P < 0.05 versus control). (B-D) PMNs were stimulated with fMLP (1 μM; 75 seconds). The phosphorylated forms of p47phox(S345) and p38-MAPK are expressed as the percentage of control values obtained with fMLP (n = 4 in each group; *P < 0.05). (E) PMNs were treated without (Buffer) or with cytochalasin B (5 μg/mL; 5 minutes) and were stimulated with fMLP (1 μM; 75 seconds), and RB is expressed in nmole superoxide/106 PMN (n = 8 in each group; *P < 0.05).

Blocking mTOR Activation by Rapamycin Inhibited fMLP-Induced PMN RB Parameters and MAPK Activation.

Human PMNs express mTOR, which has been previously shown to regulate chemotaxis.28 Whether mTOR regulates the PMN RB induced by proinflammatory agents remains unexplored. In resting PMNs of healthy donors, an active phosphorylated form of mTOR (phospho-S2448) was weakly detectable (Fig. 2A,B). PMN stimulation with fMLP greatly increased mTOR phosphorylation, which can be blocked with low rapamycin concentrations (10-20 nM), although these drug concentrations tended to increase basal phosphorylation of mTOR. The rapamycin concentration that reduced 50% of the fMLP-induced phosphorylation of mTOR (IC50) was approximately 3-5 nM. Rapamycin also inhibited fMLP-induced RB of healthy PMNs, reducing both total production of superoxide (Fig. 2C) and initial rate of RB (data not shown), which suggests a role of mTOR in the process of NOX2 activation. Rapamycin also inhibited fMLP-induced RB of cirrhotic PMNs (Fig. 2C), resulting in a dramatic aggravation of their RB defect (Fig. 2D). A rapamycin-inhibitory effect was also observed on RB measured in whole blood (Supporting Fig. 2).

Figure 2.

fMLP-induced phosphorylation of mTOR in PMNs and RB were inhibited by rapamycin. PMNs from healthy donors were pretreated with rapamycin (25 minutes) before stimulation with fMLP (1 μM; 75 seconds). The phosphorylated form of mTOR (S2448) (A) was quantified and is expressed as the percentage of control values obtained with fMLP (B) (n = 5; *P < 0.05). (C) PMNs from healthy donors (Control) and patients with cirrhosis were pretreated without (control) or with rapamycin (3-20 nM; 25 minutes) before stimulation with fMLP (1 μM). Results represent the amount of superoxide expressed as the percentage of that of the respective untreated cells (C) (*P < 0.05) or in nmole Omath image°/106 PMN (D) (n = 5 in each group; ΔP < 0.05 versus healthy donors).

Given the very weak RB of PMNs from patients with cirrhosis, the biochemical alterations induced by rapamycin were further investigated using healthy PMNs. The RB of PMN is dependent on a rapid phosphorylation of p47phox on multiple sites, among which is S345.24 Rapamycin significantly reduced the phosphorylation of p47phox(S345) induced by fMLP, whereas basal phosphorylation of p47phox tended to increase (Fig. 3A,B). A rapamycin IC50 value of 3-5 nM was obtained for the fMLP-induced p47phox(S345) phosphorylation without considering basal phosphorylation values. The S345 of p47phox is phosphorylated by two families of MAPKs: p38-MAPK and p44/42-MAPK (extracellular signal regulated kinase 1/2; ERK1/2).29 Rapamycin partially inhibited the activation of both MAPKs induced by fMLP in a concentration-dependent manner (Fig. 3D-F). However, p38-MAPK was strongly inhibited (IC50 value: 3-5 nM), relative to ERK1/2 (IC50 of 20 nM), which suggests a preferential role of p38-MAPK in the activation of NOX2 mediated by mTOR.

Figure 3.

Rapamycin inhibited the fMLP-induced phosphorylation of p47phox(S345), p38-MAPK and p44/42MAPK in PMNs. Healthy PMNs pretreated without (control) or with rapamycin (1-20 nM; 25 minutes) were stimulated with fMLP (1 μM; 75 seconds). Phospho-p47phox(S345) (A and B), phospho-p38-MAPK, and phospho-p44/42-MAPK (P-ERK) (C) were quantified and are expressed as the percentage of control values obtained with fMLP (n = 5 in each group; *P < 0.05 versus control). In the particular case of phospho-p47phox, which was elevated at basal state, the net phosphorylation values induced by fMLP were calculated after substracting basal phosphorylation values (C) (n = 5; *P < 0.05 versus control). The phosphorylated form of p38-MAPK (P-p38-MAPK; left part of D) and p44/42-MAPK (P-ERK, right part of D) was quantified and are expressed as the percentage of control (E and F, respectively; n = 4 in each group; *P < 0.05 versus control).

Rapamycin Inhibited fMLP-Induced Phosphorylation of p47phox and p38-MAPKs at the Membranes, but Not Their Translocation From the Cytosol, in Patient PMNs.

The RB of PMNs is dependent on the translocation of cytosolic p47phox at the plasma membranes to form an active complex with NOX2.1, 2 Whether mTOR regulates the translocation of cytosolic components of NOX2 was studied by measuring the amount of phosphorylated p47phox and p38-MAPK at the membranes of PMNs of patients with cirrhosis. For this purpose, the patient's PMNs whose RB was strongly inhibited by rapamycin were selected (70% of patients). fMLP significantly increased the amount of both p47phox (Fig. 4A) and p38-MAPK (Fig. 4C) at the membranes of cirrhotic PMNs, consistent with a redistribution of both effectors. However, the translocation of both effectors was not altered by rapamycin. By contrast, their phosphorylation was almost completely inhibited (Fig. 4C,F).

Figure 4.

Rapamycin prevented fMLP-induced phosphorylation of membrane-associated p47phox and p38-MAPK in PMN, but not their membrane translocation. PMNs pretreated without (control) or with rapamycin (25 minutes) were stimulated with fMLP (1 μM; 90 seconds). The cell particulate fraction was prepared for the detection of phosphorylated p47phox (A) and p38-MAPK. Data are expressed as the percentage of control values obtained with fMLP (n = 4 in each group; *P < 0.05).

Depletion of mTOR in Neutrophil-Like HL-60 Cells Inhibited fMLP-Induced RB Parameters.

To further reinforce the possibility that mTOR is a novel effector of PMN RB, superoxide production was studied in mTOR-depleted cells. Treatment of neutrophil-like HL-60 cells with mTOR siRNA oligonucleotides reduced mTOR expression by approximately 50% (P < 0.05) (Fig. 5A,B). fMLP-induced RB was also impaired in the same proportion (Fig. 5C), whereas the phosphorylation of p38-MAPK and p47phox(S345) were markedly inhibited. These data confirm that mTOR is rapidly activated in fMLP-stimulated PMNs and contributes to NOX2 activation by the phosphorylation of p47phox(S345) by MAPKs.

Figure 5.

mTOR depletion in neutrophil-like HL-60 cells inhibited fMLP-induced RB, phosphorylation of p38-MAPK, and p47phox. Differentiated HL-60 cells were transfected with mTOR siRNA oligonucleotides (100 nM; 4 days) or scramble RNA (control). mTOR expression was quantified (A) and is expressed as the percentage of control values (B) (n = 5 in each group; *P < 0.05). P47phox was quantified to compare the amount of protein loaded. The RB of PMNs induced by fMLP (1 μM) is expressed as the percentage of control values (C) (n = 5; *P < 0.05; 100% = 6 nmole Omath image°/106 cells). The fMLP-induced (1 μM; 90 seconds) phosphorylation of p38-MAPK (D and E) and p47phox (D and F) is expressed as the percentage of control values obtained with fMLP (n = 4 in each group; *P < 0.05 versus control).

Consequences of mTOR Inhibition on PMN Bacterial Uptake and Killing.

Inhibition of RB by rapamycin suggests that it may affect PMN antibacterial activities. To explore this possibility, the effects of rapamycin were studied on bacterial engulfment and killing by PMNs. Interestingly, rapamycin concentrations that blocked mTOR activation (10-20 nM) did not alter uptake of DsRed-conjugated E. coli (Fig. 6A,B). However, PMN killing activity was impaired (Fig. 6C), which is consistent with a defective RB, which we also observed when a particulate inducer of RB, such as zymosan, was used (Supporting Fig. 3B).

Figure 6.

Rapamycin inhibited PMN bactericidal activity, but not bacterial uptake. PMNs were treated at 37°C without (control) or with rapamycin (1-20 nM 25 minutes), then in the presence of DsRed-conjugated E. coli for 25 minutes. To estimate the amount of bacteria bound at the surface of PMNs, bacteria were incubated with PMN at 4°C. A representative quantification of the fluorescence intensity of PMN is shown in (A) (fluorescence-activated cell sorting analysis) and is expressed as the percentage of control values (B) (n = 6 in each group). The number of viable bacteria after incubation with PMN pretreated or not (control) with rapamycin (C) is expressed as the percentage of control values (n = 6-7 in each group; *P < 0.05 versus control).


This study provides new insights into the RB deficiency of PMNs of patients with alcoholic liver cirrhosis and reveals a rapamycin-aggravating effect on NOX2 activity as a consequence of the feedback inhibition of mTOR. NOX2, the motor system of phagocyte RB, is a potent source of ROS and plays a key role in phagocyte microbicidal activity. A deficient RB increased patients' susceptibility to bacterial infection.1 ROS have also been involved in collagen synthesis,30 liver injury, and fibrosis.31 However, during the progression of liver cirrhosis, the patient's susceptibility to microbial infection increases, which represents a main cause of death in alcoholic cirrhosis.32 In some reports, the defective microbicidal activity was associated with impaired RB,12 but not in other studies in which RB was not changed or paradoxically increased.33 These discrapencies may be the result of differences in the severity of the liver disease and/or methodological approaches (e.g., the use of the indirect methods to assess NOX2 activity, such as luminescence). In this study, RB of cirrhotic PMNs was studied using a specific assay for superoxide (cytochrome c) and revealed a severe dysfunction of NOX2 activity (Fig. 1), consistent with the weak RB observed in whole blood (Supporting Fig. 1). These data confirmed previous works,10, 12 although our healthy donors were younger (42 ± 15 years; female/male: 11:10) than patients (53 ± 3 years; female/male: 8:9; Table 1). This difference in age might contribute, in part, to the difference in neutrophil function between the two groups. However, the fifties of our controls (25%) did not show significant differences in their RB in whole blood, compared to younger subjects (data not shown). The RB defect was primarily associated with an impaired intracellular signaling toward NOX2 activation, because the fMLP-induced phosphorylation of p47phox(S345) and its effector, p38-MAPK,24, 29 was strongly decreased, whereas the amount of both proteins was unchanged. The biochemical origin of this deficiency is not known. However, a major upstream signaling effector leading to the activation of p38-MAP kinase by protein kinase C (PKC) (inositol-specific phospholipase C; PLC) was also impaired in fMLP-stimulated PMNs of patients with cirrhosis.34 In PMNs, this PLC (PLCβ2) is directly activated by the βγ subunits of a trimeric G protein (Gi) coupled to fPR1. The inability of cytochalasin B to potentiate the RB of cirrhotic PMNs (Fig. 1E) strongly suggests that biochemical alterations may affect cytoskeleton structures and alter early signaling events proximal to fPR1.

The effect of mTOR inhibition on the RB of PMNs of patients with cirrhosis is not known and may have clinical implications. Here, we show that rapamycin further aggravated the RB defect of PMNs from patients with cirrhosis (Fig. 2). The residual PMN production of ROS not inhibitable by rapamycin was very low, approximately 35% only of that of healthy PMNs (Fig. 2C). This amount of ROS is similar to that produced by PMNs from patients with chronic granulomatous diseases,35 which may contribute to the increased sensitivity of the patient to bacterial infections. Although one cannot exclude that the rapamycin-inhibitory effects described here could result from nonspecific effects, a role of mTOR to the NOX2 activation process was further confirmed by depletion of mTOR with siRNA (Fig. 5C) or antisense oligonucleotides (data not shown). Together, these results indicate that fMLP induced a rapid activation of mTOR (Fig. 2) involved in RB, which represents a novel function of mTOR in phagocyte oxygen-dependent defense systems, as supported here by the impaired bacterial killing induced by rapamycin (Fig. 6C).

The mechanism by which mTOR contributes to the PMN RB is not known. However, a possible model can be proposed based on the degree of inhibition of internal effectors induced by rapamycin (Fig. 7). Indeed, the rapamycin concentration that inhibits 50% of the fMLP-induced mTOR phosphorylation (IC50 of 3-5 nM; Fig. 2) was similar to that obtained for the phospho-p38-MAPK (Fig. 3E) and phospho-p47phox (Fig. 3C), whereas that obtained for phospho-ERK was much higher (IC50 of 20 nM; Fig. 3F). These results suggest that mTOR preferentially induces the activation of p38-MAPK, which may, in turn, phosphorylate p47phox(S345). Whether mTOR activates p38-MAPK directly or indirectly by its upstream effector MEK3/6 remains to be elucidated. Interestingly, mTOR does not appear to regulate the translocation of both p38-MAPK and p47phox from the cytosol to the membranes (Figs. 4A,B and 5D,E). Thus, mTOR regulates the activation of NOX2 through the phosphorylation of its effectors, but not by the assembly process of the NADPH oxidase complex. Our data do not exclude the possibility that mTOR directly phosphorylated p47phox, NOX2, or other partners (p67phox and p40phox). Alternatively, mTOR may regulate other signaling effectors of the RB, such as PKC. Consistent with this hypothesis, we observed that the PKC-dependent phosphorylation site of p47phox (S320) induced by fMLP was also inhibited by rapamycin (data not shown). However, when PKCs were directly activated by phorbol myristate acetate, RB was not altered by rapamycin (Supporting Fig. 3C). Thus, mTOR antagonists should be useful to dissect the activation mechanism of NOX2 and to attenuate its hyperactivity in pathological situations. However, rapamycin may have detrimental effects in PMNs from immuno-depressed patients.

Figure 7.

Proposed model for the stimulation of RB by fMLP by the mTOR/MAPK-signaling pathway in human healthy and cirrhotic PMNs. Stimulation of PMNs by fMLP by its specific G-protein-coupled receptor (fPR1) induces a rapid activation of mTOR on Ser2448 through the AKT pathway. mTOR mediates the activation of p38-MAPK, which, in turn, phosphorylates p47phox on its Ser345 at the membrane and thus contributes to superoxide production by NOX2. In patients with cirrhosis, NOX2 activity is decreased as a result of a deficient activation of p38-MAPK and p47phox.

In conclusion, the use of the bacterial peptide, fMLP, as an inducer of RB of PMNs from patients with decompensated alcoholic cirrhosis reveals a strongly impaired NOX2 activity resulting from deficient signaling activities, leading to the phosphorylation of the NOX2 component, p47phox(S345), by p38-MAP kinases. The results further identify mTOR as a novel effector of RB of PMNs. Consequently, mTOR inhibition by rapamycin dramatically aggravated the RB defect of patients' PMNs. This rapamycin-induced inhibition of NOX2 activity in PMNs from patients with cirrhosis was mediated through inhibition of p38-MAPK signaling and phosphorylation of p47phox(S345). Therefore, the use of mTOR inhibitors may increase the susceptibility of patients with cirrhosis to bacterial infections. These results suggest that rapamycin or rapalogs should be used with caution in immuno-depressed patients.


The authors thank Margarita Hurtado-Nedelec, Anh Cung, Michèle Fay, and Célia Madjene for their help with flow cytometry and imaging studies.