Branched chain amino acids enhance the maturation and function of myeloid dendritic cells ex vivo in patients with advanced cirrhosis

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

An imbalance of plasma amino acids is observed in patients with advanced cirrhosis. The aim of this study was to investigate the influence of the extracellular amino acid imbalance on the function of myeloid dendritic cells (DCs) in patients with advanced cirrhosis. We made a serum-free culture medium consistent with the average concentration of plasma amino acids from healthy controls (HC, n = 25) or patients with advanced cirrhosis (LC, n = 43) to reflect more closely the actual environment of the living body. We compared the phenotypical and biological functions of blood dendritic cells antigen-positive dendritic cells (BDCA+ DCs) and monocyte-derived dendritic cells (MoDCs) from LC and HC with these media. After adding stimulants, the CD83 and CD86 expressions of DCs from LC were lower than those from HC. In both HC and LC, both CD83 and CD86 expressions of DCs stimulated under the cirrhotic medium were lower than under the control medium. This phenomenon was accompanied by a suppression of the mammalian target of rapamycin (mTOR)/S6K-signaling pathways. The interleukin 12 (IL-12) production in the cirrhotic medium was significantly lower than in the control medium and increased when valine or leucine was added to the medium. In patients with advanced cirrhosis, peripheral blood mononuclear cells stimulated in the autologous plasma after oral administration of branched-chain amino acid (BCAA) granules had significantly increased interferon gamma production. Conclusion: In advanced cirrhosis, there is impairment of the function and maturation of DCs, which has been shown to be related to an imbalance in the extracellular amino acid profile. Elevating the extracellular concentration of BCAAs ex vivo in patients with advanced cirrhosis improved the function of DCs. (HEPATOLOGY 2009.)

Cirrhosis makes it increasingly difficult for the liver to carry out its essential functions, such as detoxifying harmful substances and manufacturing vital nutrients. Cirrhosis progresses to decompensated cirrhosis and ultimately liver failure because of a lack of suitable treatment. Not only hepatocellular carcinoma but also nosocomical infections, such as spontaneous bacterial peritonitis (SBP) or pneumonia, are frequent clinical complications in these immune-compromised patients.1 In patients with advanced cirrhosis, various metabolic disorders involving glucose, protein-amino acids, lipids, vitamins, and minerals might appear. Furthermore, an imbalance of plasma amino acids, with decreased levels of branched-chain amino acids (BCAAs) and increased levels of aromatic amino acids (AAAs), is commonly seen in patients with advanced cirrhosis.2 In clinical situations, long-term nutritional supplementation with oral BCAA has been shown to be useful to prevent progressive hepatic failure and to improve surrogate markers and the perceived health status.3, 4 Moreover, the oral administration of BCAA granules was reported to inhibit hepatic carcinogenesis in patients with compensated cirrhosis.5, 6

On the one hand, it has become clear that amino acids are not only important as substrates for various metabolic pathways but also activate a nutrient-sensitive signaling pathway in synergy with insulin.7–10 The mammalian target of rapamycin (mTOR) signaling pathway is one of the most representative pathways, and this pathway has been shown to act as a major effector of cell growth and proliferation by way of the regulation of protein synthesis.7–9 The phosphorylation of downstream effectors of mTOR is inhibited by rapamycin and activated by BCAA, especially by leucine,11–13 although little is known about the impact of changes in the extracellular amino acid levels on the immune system.14 Recently, we have shown that extracellular BCAAs, especially valine, regulate the maturation and function of monocyte-derived dendritic cells (MoDCs).15 Dendritic cells (DCs) are professional antigen-presenting cells (APCs) that stimulate innate and adaptive immune reactions by priming other types of blood cells. Typically, immature DCs migrate to lymphoid tissues and present antigenic peptides to naive T cells.16 The mature DCs, which characteristically express CD83,17 can rapidly activate other innate immune cells including natural killer (NK) cells and natural killer T (NKT) cells through the production of immunomodulatory cytokines such as interleukin (IL)-10 and IL-12. Several studies have reported that the immunological abnormalities occurring in cirrhosis,18, 19 such as a depressed reticuloendothelial system, neutrophil dysfunction, reduced serum complement, and low bactericidal function, account for the increased susceptibility of patients with cirrhosis to bacterial seeding and diffusion, and for the impaired functions of DCs in patients with liver cirrhosis.15, 20, 21 However, it is not clear why the responses of immune cells, particularly DCs, are suppressed in patients with cirrhosis.

Roswell Park Memorial Institute medium 1640 (RPMI 1640) with human or bovine serum is typically used to culture peripheral blood mononuclear cells (PBMCs) or DCs and examine the function. The concentrations of almost all the amino acids in RPMI 1640 are higher than those typically found in the plasma of healthy adult humans. Accordingly, there are large differences between the amino acids of living bodies and those of culture systems. The concentration of amino acids except BCAAs in the medium used in our previous study was higher than that of plasma in vivo.15 Furthermore, various types of amino acid imbalance actually appear in the plasma of patients with advanced cirrhosis. The aim of the study, therefore, was to investigate the influence of the extracellular amino acid imbalance observed in patients with advanced cirrhosis on the function of DCs using a serum-free culture medium consistent with the average concentration of plasma amino acids from healthy volunteers (healthy control media, HCM) or patients with advanced cirrhosis (advanced cirrhotic media, ACM) to reflect more closely the actual environment of the living body. Furthermore, we investigated whether oral administration of BCAA granules could enhance the responses of immune cells in patients with advanced cirrhosis.

Abbreviations

AAA, aromatic amino acid; ACM, advanced cirrhotic media; APC, antigen-presenting cell; BCAA, branched-chain amino acid; BDCA, blood dendritic cells antigen; DC, dendritic cell; HCM, healthy control media; IFN-γ, interferon gamma; IL, interleukin; MLR, mixed lymphocytes reaction; MoDC, monocyte-derived dendritic cell; mTOR, mammalian target of rapamycin; NKT, natural killer T; PBMC, peripheral blood mononuclear cell; SBP, spontaneous bacterial peritonitis.

Patients and Methods

Serum-Free Culture Media.

The concentrations of the plasma amino acids from fasting healthy volunteers (n = 25), chronic hepatitis (n = 14), and patients with cirrhosis (n = 60) were measured by high-performance liquid chromatography (HPLC) in the early morning (Table 1). Briefly, sulfosalicylic acid was added to plasma to a final concentration of 5%. The samples were then placed on ice for 15 minutes followed by centrifugation to remove precipitated proteins. The extracts were then analyzed for the amino acid content with a JLC-500/V (Japan Electron Optics Laboratories, Tokyo, Japan). Also, these patients with cirrhosis were classified according to the Child-Pugh classification. We defined as Child-Pugh grade B or C the patients with advanced cirrhosis (n = 43: hepatitis c virus [HCV] n = 22; primary biliary cirrhosis [PBC] n = 5; alcoholic n = 3; nonalcoholic steatohepatitis [NASH] n = 3; hepatitis b virus [HBV] n = 2; primary sclerosing cholangitis [PSC] n = 2; HCV+HBV n = 1; autoimmune hepatitis [AIH] n = 1; Wilson's disease n = 1; Budd-Chiari syndrome n = 1; cryptogenic n = 2). A serum-free culture medium consistent with the average concentration of plasma amino acids from healthy volunteers was defined as the HCM; whereas that from patients with advanced cirrhosis was defined as the ACM (Table 2). Other components except amino acids were identical among media. We verified that there was no difference between the theoretical value and actual value in HCM and ACM. We cultured PBMCs under the two media with stimulant for 48 hours and measured the amino acid concentrations of these media. There was no difference in the concentrations of amino acids before and after culture in these media. The viability of PBMCs was determined using Annexin VFITC, with dead cells identified by propidium iodide (PI) staining (Annexin V-FITC Apoptosis Detection Kit, BioVision, Mountain View, CA), according to the manufacturer's instructions. We confirmed the viability of PBMCs cultured in HCM and ACM equal to that of complete culture medium (CCM) and X-VIVO 10 (Cambrex Bio Science Walkersville, Walkersville, MD). The percentages of living cells were 78.7 ± 0.67, 77.7 ± 2.2, 71.7 ± 0.67, and 74.7 ± 0.33 for HCM, ACM, CCM, and X-VIVO10, respectively. The culture media, CCM, and other depleted media were made as described.15

Table 1. Aminogram for the Plasma in Chronic Hepatitis Patients and Patients with Cirrhosis
 HC (n=25)CH (n=14)Child A (n=17)Child B (n=19)Child C (n=24)
  • The concentrations of plasma amino acids from fasting healthy volunteers (n=25), chronic hepatitis (n=14) and patients with cirrhosis (n=60) were measured by HPLC in the early morning after fasting. Also, these patients with cirrhosis were classified according to the Child-Pugh classification. Amino acid concentrations are expressed in nmol/mL.

  • *

    P < 0.01 increased.

  • P < 0.01 decreased. Fischer's ratio means: Valine+Leucine+Isoleucine / Tyrosine+Phenylalanine †decrease *increase P < 0.01 vs. CH (the data were analyzed with ANOVA and Dunnett's post-hoc procedure).

Glycine225250205234313
Alanine391400311317339
Serine119135139137169
Threonine142139137135165
Cystine3854636273
Methionine2931406068
Glutamine564585616642739
Asparagine5157625877*
Glutamic acid4270626547
Aspartic acid33543
Valine249243222195164
Leucine13214112011093
Isoleucine7671635651
Phenylalanine6370808999*
Tyrosine6581111112151*
Tryptophan6252524347
Lysine183223219199179
Arginine78799493100
Histidine8390778193
Proline204163142165202
Fischer's ratio3.573.012.361.951.27
Table 2. Serum-Free Culture Media Used in This Study (nmol/mL)
 CCMHCMACM
  1. Complete culture medium (CCM) contains 20 amino acids that are relevant to the make-up of mammalian proteins. HCM (healthy control medium): consistent with the average concentration of plasma amino acids from healthy volunteers (n=25). ACM (advanced cirrhotic medium): consistent with the average concentration of plasma amino acids from patients with advanced cirrhosis (Child-Pugh grade B or C, n=43). The amino acid concentrations are expressed in nmol/mL. Fischer's ratio means: Valine+Leucine+Isoleucine / Tyrosine+Phenylalanine.

Glycine400225280
L-Alanine400391307
L-Serine400119151
L-Threonine800142138
L-Cystine 2HCl2003867
L-Methionine2002975
L-Glutamine4000564689
L-Asparagine4005164
L-Glutamic Acid4004253
L-Aspartic Acid40034
L-Valine800249175
L-Leucine800132100
L-Isoleucine8007653
L-Phenylalanine4006399
L-Tyrosine40065133
L-Tryptophan806245
L-Lysine-HCl800183184
L-Arginine-HCl4007892
L-Histidine HCl-H2O2008385
L-Proline400204176
Fischer's ratio3.003.571.42

Patients and Healthy Volunteers.

We selected 15 patients with cirrhosis for in vitro or ex vivo studies (Table 3). All of these patients were inpatients. There were no significant differences on clinical and laboratory findings in this population compared to the 43 patients with advanced cirrhosis (Table 1): age 60.4 ± 12.8 versus 59.1 ± 11.3; aspartate aminotransferase (AST) 78.8 ± 45.4 IU/L versus 96.3 ± 65.0 IU/L; alanine aminotransferase (ALT) 47.6 ± 25.2 IU/L versus 54.3 ± 36.7 IU/L; total bilirubin 4.5 ± 5.36 mg/dL versus 3.94 ± 3.70 mg/dL; albumin 2.80 ± 0.51 g/dL versus 2.85 ± 0.55 g/dL; prothombin time / international normalized ratio (PT-INR) 1.54 ± 0.39 versus 1.37 ± 0.29; PLT 93.9 ± 68.7 × 103/μL versus 113.1 ± 54.2 × 103/μL; Child Pugh score 9.0 ± 1.77 versus 8.6 ± 2.10; Model for End-Stage Liver Disease (MELD) score 11.9 ± 5.55 versus 11.2 ± 4.23; plasma Fischer's ratio 1.56 ± 0.77 versus 1.65 ± 0.57. The MELD score22 was calculated by an online worksheet available on the Internet at www.mayoclinic.org/meld/mayomodel5.html. None of the patients had clinical or laboratory findings compatible with bacterial infection when we collected PBMCs from the patients. Written informed consent was obtained from each individual and the study protocol was approved by the Ethics Committee of Tohoku University School of Medicine (2003–326, 2008–337).

Table 3. Characteristics of Study Participants
Patient NumberDiseaseSexAge (years)AST/ALTTotal BilirubinAlbuminPT-INRPLTChild-Pugh ClassificationMELD ScorePlasma Fischer's RatioBCAA Medication
  1. LC-C, liver cirrhosis due to HCV; HCC, hepatocellular carcinoma; PBC, primary biliary cirrhosis; NASH, nonalcoholic steatohepatitis; NA, not available; PLT, platelet counts (x103/μL); PT-INR, prothrombin time-international normalized ratio; AST/ALT, aspartate aminotransferase / alanine aminotransferase (IU/L); total bilirubin (mg/dL); albumin (g/dL); Fischer's ratio: Valine+Leucine+Isoleucine / Tyrosine+Phenylalanine.

1LC-CM71116/610.83.31.09149A62.49-
2LC-C+HCCM7073/461.52.31.1575B62.26-
3LC-C+HCCF8072/551.32.81.19144B9NA+
4LC-CM4252/384.21.81.7979C160.99+
5LC-C+HCCF61238/986.32.91.6576B182.74+
6PBCF43241/14412.32.81.32152C181.57-
7LC-CM5671/452.23.71.2481B101.90+
8LC-CM48111/1091.63.71.0881A8NA-
9LC-CF6025/511.63.22.0583C150.88+
10LC-C+HCCF6968/401.32.81.17132B71.81-
11non B non CF4428/182.42.61.54122C81.31+
12PBCF62130/496.82.01.33120C81.43+
13PBCF6283/302.32.51.11207B131.29+
14AlcoholicM5453/242.53.11.60219C141.24+
15LC-C+HCCM6583/532.03.21.2996B121.52+

BDCA+ DCs Maturation and MoDCs Generation.

PBMCs were separated from the peripheral blood of HC and LC by centrifugation on a density gradient. The blood dendritic cells antigen-positive dendritic cells (BDCA+ DCs) and the CD14-positive monocytes were isolated from PBMCs using magnetic microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). BDCA1+ DCs were cultured at a density of 2.5 × 105 cells/well in 96-well flat-bottom plates (Corning, NY) for 48 hours with 1,000 U/mL GM-CSF (PreproTech, London, UK), 500 U/mL (hu) IL-4 in each media. At 24 hours culture, DCs were stimulated by 500 ng/mL lipopolysaccharide (LPS; Escherichia coli 026:B6; Sigma, St. Louis, MO) or polyinosinic:polycytidylic acid (polyI:C) (30 μg/mL). Monocytes were cultured at a density of 3.0 × 105 cells/well with granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-4 for 6 days in CCM. On day 6 we changed the medium from CCM to HCM or ACM with poly(I:C) and the culture was continued for an additional 48 hours.

Surface Marker Analysis.

DCs were harvested and labeled with fluorescein isothiocyanate (FITC)- or phycoerythrin (PE)-labeled monoclonal antibodies (mAbs) (antihuman CD14, CD40, CD83, CD86, CD98, HLA-DR, or the relevant isotype controls; BD PharMingen, San Diego, CA) according to the manufacturer's instructions. Using a FACS Calibur (BD Immunocytometry Systems, San Diego, CA) flow cytometer, surface marker expressions were analyzed using the CellQuest (BD Immunocytometry Systems) program.

Phagocytosis Assay with Dextran.

To evaluate the endocytosis potential of DCs, 1 mg/mL of FITC-dextran was supplied to 2.5 × 105 DCs that were then incubated for 30 minutes at 37°C. As a control, the DCs were given the same doses of FITC-dextran and stored for 30 minutes at 4°C. After the incubation the DCs were washed and subjected to FACS analysis.

Cytokine Analysis.

BDCA1+ DCs were cultured at a density of 2.5 × 105 cells/well in 96-well flat-bottom plates for 48 hours with 1,000 U/mL GM-CSF, 500 U/mL (hu) IL-4 in each of the media. At 24 hours, 500 ng/mL LPS or poly(I:C) (30μg/mL) were added. The supernatants were collected after 48 hours and immediately IL-12 (p40+p70) and IL-10 were determined by specific cytokine enzyme-linked immunosorbent assay (ELISA) kits (Bender MedSystems) according to the manufacturer's instructions. For the interferon gamma (IFN-γ) production of PBMCs, PBMCs were cultured at a density of 2.5 × 105 cells/well in HCM or ACM for 48 hours, and at 5.0 × 105 cells/well in autologous plasma for 12 hours. IFN-γ was determined by specific cytokine ELISA kits (Bender MedSystems).

Mixed Lymphocytes Reaction (MLR).

BDCA+ DCs were cultured at a density of 1.0 × 105 cells/well in 96-well round-bottom plates (Falcon) containing HCM or ACM with GM-CSF and IL-4 for 48 hours. At 24 hours culture, immature DCs were induced to mature using LPS or poly(I:C) for an additional 24 hours. The allostimulatory capacity of irradiated DCs (3,000 Rad) was tested in a one-way MLR with normal 2 × 105cells/well allogeneic CD4+ lymphocytes (isolated from PBMCs using magnetic beads) under CCM. Cocultured cells were maintained for 7 days and the proliferation rate of the cells was measured using an 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) Assay (CellTiter 96 aqueous one-solution cell proliferation assay; Promega, Madison, WI) according to the manufacturer's instructions. On carboxyfluorescein succinimidyl ester (CFSE) staining, cells were analyzed using a CellTrace CFSE Cell Proliferation Kit (Molecular Probes, Eugene, OR). The staining methods followed the manufacturer's protocol.

Immunoblotting.

DCs were cultured at a density of 3.0 × 105 cells/well in 96-well flat-bottom plates (Corning) containing 200 μL medium supplemented with GM-CSF and IL-4 for 24 hours and the DCs were stimulated by poly(I:C) for 1 hour. The DCs were harvested and lysed using CelLyticTM-M Mammalian Cell Lysis/Extraction Reagent (Sigma). The lysed cells were centrifuged to pellet the cellular debris. Thereafter, these protein concentrations were determined by a Modified Lowry Protein Assay Kit (Pierce, Rockford, IL). Equal amounts of protein were loaded onto sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and electrotransferred to PVDF (Immun-Blot PVDF Membrane; Bio-Rad, Hercules CA). After washing and blocking, immunostaining was performed with rabbit polyclonal primary antibody (PI3K, phospho-PI3K, mTOR, p70 S6K, phospho-p70 S6K; Cell Signaling Technology, Beverly, MA), followed by incubation with a secondary antibody conjugated to horseradish peroxidase (HRP) (Sigma). Immunoreactive proteins were revealed with an ECL reagent (ECL advance; Amersham Biosciences, Little Chalfont, UK).

Oral Administration of BCAA to Patients with Advanced Cirrhosis and Ex Vivo Cytokine Production Assay.

In the early morning we measured the fasting concentration of the plasma amino acids before and after oral administration of BCAA granules (30, 60, 120, 180 minutes) from healthy volunteers and patients with advanced cirrhosis. The BCAA granules: LIVACT (Ajinomoto Pharma, Tokyo, Japan) were composed of a mixture of valine, 1.144 g, leucine, 1.904 g, and isoleucine, 0.952 g. The concentrations of the plasma amino acids were measured by HPLC. We stimulated PBMCs from patients for 12 hours by LPS or poly(I:C) under autologous plasma, which was collected both before and after oral administration. After 12 hours we recovered the plasma and measured the IFN-γ by ELISA.

Statistical Analysis.

The data were analyzed with analysis of variance (ANOVA) and multiple comparisons were performed with Dunnett's post-hoc procedure for the plasma aminogram. When two groups were analyzed, the differences between media were analyzed by the Wilcoxon t test. Frequencies of BDCA1+ DCs were compared between patient groups by the Mann-Whitney U test. All statistical analyses were performed with standard statistical software (SPSS 13.0 for Windows, Chicago, IL).

Results

Amino Acid Concentrations Similar to Those in Plasma of Patients with Advanced Cirrhosis Impaired the Maturation of Myeloid DCs from Healthy Controls.

First we measured the cytokine production from PBMCs both under HCM and ACM. The IFN-γ production of PBMCs stimulated by poly(I:C) under ACM was significantly impaired (28.1 ± 7.3 pg/mL versus 16.7 ± 3.9 pg/mL; P = 0.04). Next, we cultured the BDCA+ DCs (purity >90%) for 48 hours under HCM and ACM and evaluated the phenotypes of DCs by flow cytometry. In ACM, the CD83 and CD86 expression of DCs was significantly impaired compared to that in HCM (Table 4). The HLA-DR expression had a tendency to decrease in ACM. This phenomenon was observed in MoDCs (Supporting Fig. 1). Next, The IL-12 production of BDCA+ DCs stimulated under ACM was significantly impaired (110.7 ± 8.6 pg/mL versus 79.9 ± 12.5 pg/mL; P = 0.04), although the IL-10 production of DCs was not different between HCM and ACM (31.0 ± 4.0 versus 32.4 ± 8.2; P = 0.59). Flow cytometric analysis revealed that the amount of FITC-dextran taken up by BDCA+ DC and MoDC did not differ between HCM and ACM (data not shown). The allostimulatory capacity of BDCA+ DCs cultured under ACM was significantly decreased as shown by the MTS assay (1.00 ± 0.15 versus 0.82 ± 0.13; P = 0.04; absorbance 490 nm), and this tendency was also confirmed by the CFSE assay.

Table 4. Phenotypic Difference of BDCA1+DCs Derived from Patients with Cirrhosis and Healthy Volunteers
   CD40CD83CD86HLA-DR
  • The MFI are presented for each marker as the mean ± SD of healthy controls and patients with cirrhosis (isolated DC: Patients 6, 7, 8, 10 / mature DC: Patients 8, 9, 10, 11, 12).

  • *

    Value of P < 0.05 vs. DCs of healthy control cultured under HCM (Wilcoxon t test).

  • Value of P < 0.05 vs. DCs of healthy control cultured under HCM (Mann-Whitney U test).

  • Value of P < 0.05 vs. DCs of LC patients cultured under HCM (Wilcoxon t test).

Isolated DCHealthy control (n=4) 5 ± 1.46 ± 2.214 ± 3.1166 ± 52.2
 LC patients (n=4) 12 ± 16.14 ± 1.412 ± 3.4195 ± 79.3
Mature DCHealthy control (n=5)HCM131 ± 54240 ± 25201 ± 67910 ± 121
 ACM121 ± 37190 ± 33*170 ± 53*783 ± 90
 LC patients (n=5)HCM139 ± 44154 ± 48169 ± 37691 ± 112
 ACM124 ± 47125 ± 45122 ± 11625 ± 160

Amino Acid Concentrations Similar to Those in Plasma of Patients with Advanced Cirrhosis Also Impaired the Maturation of Myeloid DCs from Patients with Cirrhosis.

We first evaluated the frequency of BDCA+ DCs between HC and LC (Fig. 1). The frequencies of DCs were significantly lower in the peripheral blood from patients with advanced cirrhosis compared to those from HC or patients with early cirrhosis. Second, we determined the phenotype of BDCA1+ DCs from the LC before and after adding the stimulants. There was no difference regarding the mean fluorescence intensity (MFI) of isolated immature DCs expressing CD40, CD83, CD86, and HLA-DR between the HC and LC (Table 4). After adding the stimulants, the expressions of CD83 and HLA-DR by DCs from the LC were significantly decreased compared to those from the HC in both HCM and ACM (Table 4). The CD83 and CD86 expression of DCs was significantly impaired in ACM compared to that in HCM (Table 4).

Figure 1.

The frequencies of DCs were significantly lower in the peripheral blood from patients with advanced cirrhosis compared with those from HC or early patients with cirrhosis. Percentages of BDCA+ DC in PBMCs were determined by flow cytometry. Significant differences in the percentages of DCs were observed between patients with advanced cirrhosis (Child-Pugh grade B or C: n = 10) and HC (n = 7). There was no difference between patients with Child-Pugh grade A (n = 7) and HC. Data are expressed as mean ± standard error of the mean (SEM).

Elevating the Concentration of BCAA Enhanced the IL-12 Production in BDCA+ DCs.

As in the in vivo study, we confirmed that the plasma concentrations of BCAAs were significantly decreased and AAAs (except tryptophan) were increased along with the Child-Pugh grade (Table 1). Based on these data, to investigate which amino acid especially influenced the function of BDCA1+ DCs, we measured the cytokine production of DCs under HCM, ACM, and ACM supplemented with 800 nmol/mL of a single amino acid: valine, leucine, isoleucine, or AAAs. Interestingly, the IL-12 production of DCs stimulated under ACM plus valine or leucine was more increased than that under ACM, although there was no difference among ACM plus isoleucine, ACM plus AAAs, and ACM (Fig. 2A). Similar to the cytokine production, the allostimulatory capacity of DCs cultured under ACM plus valine or leucine had a tendency to be increased, as shown by the MTS assay (ACM: 0.71 ± 0.07, ACM plus valine: 0.88 ± 0.06; ACM plus leucine: 0.83 ± 0.03; absorbance 490 nm). Next, we determined the BDCA1+ DCs phenotype (CD14 and CD83) in CCM, BCAA-depleted, valine-depleted, leucine-depleted, and isoleucine-depleted media. In CCM, leucine-depleted and isoleucine-depleted media the DC phenotypes were similar (the percentages of CD83-positive cells were 33.7 ± 7.2%, 31.5 ± 5.4%, and 35.5 ± 7.9% for CCM, leucine-depleted, and isoleucine-depleted media, respectively). However, in BCAA-depleted and valine-depleted media, the CD83 expression of DCs was significantly impaired compared to that in CCM (BCAA-depleted media: 19.6 ± 3.0% and valine-depleted media 14.6 ± 1.8%; P = 0.04 versus CCM). After we cultured the DCs under depletion of valine for 2 days, we added valine to the medium and cultured the cells for an additional 24 hours. Then, the percentage of mature DCs was higher than that of valine-depleted media. Furthermore, to reflect more closely the actual environment of the living body, we induced DCs from LC to mature with either autologous plasma or autologous plasma supplemented with 100 nmol/mL valine for 12 hours. In all cases the DCs matured in the autologous plasma with valine had enhanced allostimulatory capacity and IL-12 production (Fig. 2B).

Figure 2.

Elevating the concentration of BCAAs enhanced the IL-12 production in BDCA1+ DCs. Isolated BDCA1+ DCs were cultured under HCM, ACM, and ACM supplemented with 800 nmol/mL single amino acid: valine, leucine, isoleucine, or AAAs. (A) After 48 hours the supernatants were assayed for cytokine concentrations. Mean ± SEM values from five different donors. (B) We induced BDCA1+ DCs from LC patients (Patients 1-5) to mature with either autologous plasma or autologous plasma supplemented with 100 nmol/mL valine for 12 hours. Supernatants were measured by ELISA. P < 0.05 (paired Student's t test, two-tailed).

Amino Acid Concentration of Plasma in Patients with Advanced Cirrhosis Down-regulated the mTOR/S6K Signaling Pathway of BDCA1+ DCs.

We hypothesized that the amino acid imbalance of the plasma in patients with advanced cirrhosis influence the mTOR/S6K signaling pathway of DCs and impaired their maturation. Under HCM with rapamycin, the percentage of CD14-/CD83+ mature DCs was higher than under HCM without rapamycin (Fig. 3A). BDCA+ DCs expressed similar levels of total PI3K, phospho-PI3K, mTOR, p70 S6K, and β-actin among all media. Interestingly, DCs cultured in ACM expressed lower levels of phospho-p70 S6K than those cultured in HCM (Fig. 3B). The expression of phospho-p70 S6K by DCs in ACM was partially recovered by adding 400 nmol/mL BCAA to the medium during stimulation. Isolated immature BDCA+ DCs expressed moderate levels of CD98 which modulate the amino acid transport functions and, after adding the stimulants, mature DC showed the up-regulation of CD98. There was no difference regarding the expression of CD98 between HCM and ACM (data not shown).

Figure 3.

Amino acid imbalance in plasma of patients with advanced cirrhosis downregulated the mTOR/S6K signaling pathway of BDCA1+ DCs. (A) We stimulated BDCA1+ DCs under HCM, ACM, and HCM plus rapamycin (500 nM) for 24 hours with GM-CSF and IL-4, and exposed them to poly(I:C) for an additional 24 hours. We evaluated the phenotypes of DCs by flow cytometry. The percentages indicate the proportion of cells adopting the DC immunophenotype (CD14−/CD83+). (B) We cultured BDCA1+ DCs under HCM and ACM for 24 hours with GM-CSF and IL-4 and stimulated them with poly(I:C) for 1 hour. We also evaluated HCM plus rapamycin, and ACM plus BCAA. Equal amounts of protein were loaded and the levels of PI3K, phospho-PI3K, mTOR, p70 S6K, and phosho-p70 S6K were determined by Western blot analysis. (A,B) Data shown are representative of four independent experiments with cells from different donors.

Oral Administration of BCAAs Enhanced the Production of IFN-γ by PBMCs from Patients with Advanced Cirrhosis Ex Vivo.

Finally, we evaluated whether BCAAs have an effect on the immune response ex vivo. In healthy volunteers the concentration BCAAs of plasma was maximum 30 minutes after oral administration (Fig. 4A). Fischer's ratio increased from 4.78 ± 1.41 (standard deviation [SD]) to 13.39 ± 2.41 (SD). On the other hand, in the patients with advanced cirrhosis (Table 3: Patients 10–13), the concentration BCAAs of plasma was maximum 60 minutes after oral administration. Fischer's ratio increased from 1.37 ± 0.98 (SD) to 4.94 ± 0.99 (SD). AAAs decreased slowly during the following 3 hours. We stimulated PBMCs from the patients with advanced cirrhosis (Table 3: Patients 11–15) using either autologous plasma before and after 60 minutes oral administration. Interestingly, in all cases PBMCs stimulated by LPS in the latter had more IFN-γ production than the former (Fig. 4B).

Figure 4.

Oral administration of BCAA granules enhanced the production of inflammatory cytokines from PBMCs stimulated by LPS ex vivo. (A) We analyzed the kinetics of the plasma amino acids after oral administration of BCAA granules. In the early morning while fasting, the concentrations of plasma amino acids were measured before and after oral administration of BCAA (30, 60, 120, 180 minutes). Mean ± SD values from three different HC and four patients with advanced cirrhosis (Patients 10–13). (B) We stimulated PBMCs from the patients using either autologous plasma before or after 60 minutes oral administration. After 12 hours we recovered the plasma and measured the IFN-γ by ELISA (Patients 11–15). P < 0.05 (paired Student's t test, two-tailed).

Discussion

In this study we started by making two serum-free media (HCM and ACM) to be more representative of the human physiological environment and quantitatively measured the plasma amino acid profiles. First, we found that the amino acid imbalance of plasma in patients with advanced cirrhosis impaired the production of IFN-γ from PBMCs. IFN-γ is a dimerized soluble cytokine that is the only member of the type II class of interferons.23 IFN-γ is secreted by Th1 cells, DCs, and NK cells. Although the commitment toward either the Th1 or the Th2 phenotype can be influenced by many signals active at the moment of naive Th cell priming, the levels of IL-12p70 (IL-12) produced by APC, especially DCs, are of major importance.24, 25 Therefore, we hypothesized that the impaired production of IFN-γ from PBMCs caused the dysfunction of DCs. Expectedly, the maturation and the IL-12 production of DCs were impaired in ACM. Furthermore, we confirmed that the allostimulatory capacity of DCs stimulated in ACM was impaired by MTS and CFSE assays. Previous studies have suggested an increase in IL-10 in cirrhosis and a potential link between high IL-10 and low HLA-DR expression in relation to immune dysfunction,26 but in this study there was no difference in IL-10 secretion between DCs from ACM compared with HCM. Such differences were probably caused by (1) differences in the stimulation period of the immune cells (the former was ex vivo, this study was in vitro); (2) differences in the cell sources (the former was monocytes, this study was DCs); (3) other factors besides amino acids influence IL-10 production. Also in patients with cirrhosis, the CD83 and CD86 expression of DCs stimulated under ACM was lower than that under HCM. When compared under the same medium, the CD83, CD86, and HLA-DR expressions of DCs from LC were lower than those from DCs of HC. To summarize these results, in advanced cirrhosis not only the DCs themselves but also the extracellular environments tend to impair the maturation of DCs.

Second, we examined which amino acids more strongly influences the function of DCs between HCM and ACM. We found that BCAA, especially valine and leucine, increased the BDCA+ DC allostimulatory capacity and IL-12 production. This confirms the findings of our previous study,15 although the enhancement by a single amino acid was very subtle. To obtain greater enhancements, we may need to use combinations of other amino acids.

Concerning the mechanism that underlies these phenomena, we confirmed that the CD98 expression of DCs were not different between HCM and ACM. CD98 can regulate the expression and distribution of the light chains to modulate the amino acid transport functions. CD98hc is highly expressed on proliferating lymphocytes and on other rapidly growing cells.27 Next, we examined whether the amino acid imbalance in the plasma of patients with advanced cirrhosis influenced the mTOR/S6K signaling pathway of the DCs. Recently, some studies reported the PI3K-mediated negative feedback regulation of IL-12 production in DCs,28 and rapamycin-enhanced IL-12 production in LPS-stimulated DC.29, 30 In the present study, BDCA+ DCs stimulated in ACM impaired IL-12 production, even though the mTOR signaling was decreased. This paradox raises the possibility that the amino acid imbalance influences not only mTOR signaling but also other types of signaling such as GSK3 or NF-κB signaling. This issue should be evaluated in future studies.

Finally, we investigated whether elevating the level of plasma BCAAs enhances the immune response ex vivo in patients with advanced cirrhosis. BCAA granules have been used to effectively reverse the hypoalbuminemia and hepatic encephalopathy in patients with advanced cirrhosis.31 In the preliminary investigation, we analyzed the kinetics of plasma amino acids after oral administration of BCAA granules. After oral administration, the BCAA concentration in plasma was maximal at 30 minutes in healthy volunteers. This was in contrast to patients with advanced cirrhosis, who had a slow increase in BCAA plasma concentrations that was maximal at 60 minutes. This difference was probably caused by the malabsorption of amino acids in the patients. In the ex vivo study, we could not use the medium to analyze the function of DCs of PBMCs because the concentration of the amino acids in medium influences the function. Thus, we stimulated cells in autologous plasma and analyzed the function over a short period of time. We found that oral administration of BCAAs enhanced the production of IFN-γ from PBMCs ex vivo in patients with advanced cirrhosis.

The results of this study still cannot be construed as conclusive evidence of a change in the functional clinical state in terms of lowering the risk of sepsis in cirrhosis or enabling consideration of such treatment for viral hepatitis. We need to perform a prospective, randomized, controlled trial in a well-characterized group of patients with appropriate immune mechanistic evaluation and determine the effects on the risk of sepsis in a longitudinal follow-up. In the present study we demonstrated at least that extracellular amino acids, especially BCAAs, influence the function of the immune system, and the amino acid imbalance in the plasma of patients with advanced cirrhosis impaired the maturation of DCs and the production of inflammatory cytokines from PBMCs or DCs.

In conclusion, the data from this study provide a rationale for future studies utilizing nutrition therapies that could be beneficial to immune function in patients with advanced cirrhosis.

Ancillary