Vigorous activation of monocytes in juvenile autoimmune liver disease escapes the control of regulatory T-cells^†

Patients and Controls.

Fifty-one patients with antinuclear antibody or smooth muscle antibody–positive AILD (AIH type 1 [AIH-1] or autoimmune sclerosing cholangitis) were studied; median age at diagnosis was 12.6 years (range, 3.6-16.6); aspartate aminotransferase (AST), 511 IU/L (range, 21-2,642); bilirubin, 54 μmol/L (range, 5-423); immunoglobulin G (IgG), 25.9 g/L (range, 9.55-68.9); antinuclear antibody titer, 1/320 (range, 1/20-1/10,240); and smooth muscle antibody titer 1/160 (range: 1/20-1/2,560). Clinical presentation was acute in 23 and chronic in 28. A liver biopsy performed at the time of or close to diagnosis showed histological features of interface hepatitis in all patients, with a median histological inflammatory activity index of 6 (range, 1-13), calculated as previously described.1, 15 Sixteen children with autoimmune sclerosing cholangitis had cholangiographic evidence of bile duct disease.15 All patients were studied while on immunosuppressive treatment (age at the time of study, 6.3-28 years; median, 14.5), having previously demonstrated that T-regs collected at disease presentation, before immunosuppressive treatment is started, are inefficient at suppressing the proliferation of target cells.5, 6 Thirty-two patients were studied while their disease was inactive (ID = normal levels of AST), and 22 while their disease was active (AD = elevated AST); three patients were studied on both occasions. Patients were receiving prednisolone (2.5-5 mg daily) with or without azathioprine (1-2 mg/kg/day). Twenty-seven healthy subjects served as normal controls. The study was approved by the Ethics Committee of King's College Hospital, London, United Kingdom. Demographic and laboratory data at the time of study are shown in Table 1.

Cell Separation.

Peripheral blood mononuclear cells (PBMCs) were obtained as previously described.3 Lymphocyte viability, determined by Trypan blue exclusion, exceeded 98%.

Flow Cytometry.

The frequency and phenotype of monocytes was determined by double-color flow cytometry, performed on fresh PBMCs. Unfractionated cells were stained with fluorescein isothiocyanate–conjugated anti-CD14 and phycoerythrin (PE)-conjugated anti–human leukocyte DR antigen (HLA)-DR, anti-CD86, anti-CD54 (all from BD Biosciences Discovery Labware, Oxford, UK), and anti-CD40 monoclonal antibodies (eBioscience, San Diego, CA). The expression of TLR4, whose engagement by its natural ligand LPS is key to the initiation of adaptive immune responses,16 was determined on purified monocytes (see Cell Purification) using PE-conjugated anti-TLR4 monoclonal antibodies (eBioscience). Monocyte TLR4 expression was assessed before and after LPS stimulation, in the absence or presence of T-regs (or CD4+CD25− T cells as control) (see Cell Purification). T-reg phenotype was assessed by triple-color flow cytometry using fluorescein isothiocyanate–conjugated anti-CD4, PE-conjugated anti-CD45RO, CD62L, and antigen-presenting cell–conjugated CD25 monoclonal antibodies (all from BD Biosciences) on purified CD4+CD25+ T cells. Because of recent evidence showing that absence or low expression levels of the activation markers CD127 (α chain of the IL-7 receptor) and CD154 (CD40L) define those CD4+CD25+ T cells with the strongest suppressor function (true T-regs),17-19 purified T-regs were further characterized using PE-conjugated anti-CD127 (BD Bioscience) and anti-CD154 monoclonal antibodies (eBioscience).

Cells were incubated at 4°C for 35 minutes in the dark, washed once with phosphate-buffered saline/1% fetal bovine serum, then resuspended and stored at 4°C in the dark until analysis. Expression of the forkhead winged/helix transcription factor box P3 (FOXP3), another T-reg marker,20, 21 was assessed by intracellular staining using PE-conjugated anti-human FOXP3 monoclonal antibodies (eBioscience), as previously described.22 Flow cytometry was performed on a Becton Dickinson FACScalibur (Becton Dickinson, Immunocytochemistry Systems, San José, CA), and CellQuest software (Becton Dickinson) was used for analysis. A minimum of 1 × 10⁴ gated events was acquired for each sample.

Immunohistochemistry.

Paraffin-embedded liver sections, available from five biopsy specimens, were stained with anti-CD68 monoclonal antibody, which stains both resident and infiltrating macrophages.23, 24 Sections were also stained using anti-CD3 and anti-CD79, markers for T and B lymphocytes, respectively. To unmask the antigen, samples were deparaffinized by placing them in xylene, then in ethanol. After rehydration with water, sections were placed in a jar containing Tris-ethylenediaminetetra-acetic acid buffer (pH: 9) and microwaved for 25 minutes. After cooling for 15 to 30 minutes, sections were placed in 1× phosphate-buffered saline for 5 minutes. Staining was carried out using a Novolink Polymer Detection System (Novocastra, Newcastle upon Tyne, UK). After endogenous peroxidase blocking and treatment with protein block solution, sections were washed with 1× phosphate-buffered saline for 5 minutes and stained with monoclonal antibodies to CD68, CD3, and CD79 (DakoCytomation Ltd., Ely, UK). Anti-CD68 monoclonal antibodies were prediluted by the manufacturer, whereas anti-CD3 and anti-CD79 monoclonal antibodies were used at the dilution of 1/10 and 1/50, respectively. After 1 hour incubation at room temperature, sections were washed, treated with a post-primary block solution, and incubated for 30 minutes with anti-mouse IgG polymer–horseradish peroxidase–labeled antibody. After washing, 3,3′-diaminobenzidene chromogen solution was applied at 100 μL per milliliter substrate.

Cell Purification.

Monocytes were isolated from PBMCs by positive selection using microbeads conjugated with anti-human CD14 antibodies (Miltenyi Biotec, Bergisch-Gladbach, Germany). T-regs were obtained from PBMCs by CD4-negative selection followed by CD25-positive selection (Dynal Invitrogen, Oslo, Norway),7 and purified CD4+CD25+ T cells localizing in the CD4+CD25^high cell gated area by cell sorting (model BD Vantage SE/DiVa, Immunocytochemistry Systems, San José, CA). CD4+CD25− T cells used in control experiments were obtained by negative selection (Dynal Invitrogen) as previously described.7 The purity of the three cell populations, assessed by flow cytometry, exceeded 95% for monocyte and T-reg subsets and 90% (range, 92%-97%) for CD4+CD25− T cells.

True T-regs (tT-regs), i.e., CD4+CD25+CD127− T cells, were further purified from total T-regs. Because of the large number of cells required, this additional purification step was limited to those subjects (11 AILD patients, six ID and five AD, and six healthy subjects) from whom sufficient numbers of CD4+CD25+ T cells were obtained. In brief, purified CD4+CD25+ cells were incubated with PE-conjugated anti-CD127 monoclonal antibodies for 30 minutes, then with microbeads conjugated to monoclonal anti-PE antibodies (Miltenyi Biotec) for a further 15 minutes at 4°C. CD4+CD25+CD127− and CD4+CD25+CD127+ T cell populations were purified by negative and positive selection, respectively, using MS columns (Miltenyi Biotec) according to the manufacturer's instructions. Their purity, assessed by flow cytometry, exceeded 90%. Unfractionated CD4+CD25+ T cells are henceforth referred to as conventional CD4+CD25+ T-regs (cT-regs) and CD4+CD25+CD127− T cells as true T-regs (tT-regs).

Chemotaxis Assay.

A series of transwell experiments was performed to assess the migration of monocytes in response to a chemoattractant as an in vitro readout of the in vivo ability of these cells to migrate and be recruited to the sites of inflammation. Purified monocytes (2 × 10⁵) were resuspended in chemotaxis medium (Roswell Park Memorial Institute 1640 containing 1% bovine serum albumin) and placed in the upper chamber of a transwell system, separated from the lower chamber by an 8-μm pore membrane (BD Biosciences Discovery Labware). Monocyte migration was assessed in both the presence and the absence of the chemoattractant stromal cell–derived factor 1α (SDF-1α), added to the lower chamber at a concentration of 100 ng/mL.25 After incubation for 1 hour at 37°C and 5% CO₂, the number of monocytes that had migrated through the membrane was determined by 1-minute flow cytometry acquisition,26 using anti-CD14 monoclonal antibody. To test whether the migration of monocytes through the pore membrane was affected by T-regs, cT-regs, tT-regs, or CD4+CD25+CD127+ T cells, these cells were added at a ratio of 1/8 to autologous monocytes either in direct contact in the upper chamber or in the lower chamber, separated from them by the membrane. The 1/8 ratio was selected because it is capable of exerting a detectable regulatory function in preliminary experiments in which ratios of 1/16, 1/8, and 1/4 were compared. Control experiments were performed using CD4+CD25− T cells.

Chemotaxis in the Presence of Matrigel Matrix.

The chemotaxis assay was repeated after matrigel matrix was layered onto the transwell membrane, because matrigel matrix is a basement membrane preparation whose structure mimics that of blood vessel walls. Cell migration across it, occurring on enzymatic (for example, metalloproteinases) digestion,27 reflects the in vivo capability of the cells to cross the blood vessel barrier. Matrigel (BD Biosciences Discovery Labware) was used at a final concentration of 100 μg/cm² according to the manufacturer's instructions. After thawing, matrigel was diluted in chemotaxis medium, layered onto 8-μm polycarbonate tissue culture inserts (Nunc, Roskilde, Denmark), and maintained at 37°C and 5% CO₂ for 1 hour to gel. Cells were then added and incubated at 37°C for 2 hours; the number of monocytes that had migrated through the matrigel and transwell membrane was tested by flow cytometry. Monocyte migration was assessed in the presence and absence of chemoattractant and after addition of cT-regs, tT-regs, CD4+CD25+CD127+, or CD4+CD25− T cells as described previously.

Results of chemotaxis assays are presented as a percentage of monocyte migration calculated as the ratio of the number of monocytes from the lower chamber over the total number of monocytes seeded initially in the upper chamber. Changes between the percentage of monocyte migration in the presence and absence of the chemoattractant SDF-1α and before and after T-reg addition were calculated for each individual as fold increase/decrease.

Quantification of Gene Expression by Real-Time Polymerase Chain Reaction.

To investigate whether the ability of monocytes to migrate through matrigel was related to the expression of metalloproteinase enzymes, a series of preliminary experiments was performed to identify which was the most expressed matrix metalloproteinase (MMP)-encoding genes among MMP1, MMP9, MMP19, and MMP26,28 MMPs characteristically expressed by monocytes. MMP19 was the most frequently expressed MMP gene in monocytes from six AILD patients and three healthy subjects and was therefore selected for further testing. Total RNA was extracted from purified monocytes (1 × 10⁵ to 5 × 10⁵) using TRizol reagent (Invitrogen Life Technologies, Paisley, UK), and messenger RNA (mRNA) was reverse transcribed as previously described.7 MMP19 gene transcripts were quantified by real-time polymerase chain reaction (PCR) using gene-specific probes and TaqMan Master Mix (Applied Biosystems, Warrington, UK). Polymerase chain reaction amplification conditions were as previously described.22 Samples were run in triplicate using a real-time polymerase chain reaction thermocycler (ABI-Prism 7000 Sequence Detection System; Applied Biosystems, Foster City, CA), and results were analyzed by matched software. Relative gene expression was determined by normalizing to glyceraldehyde-3-phosphate-dehydrogenase gene expression in each set of samples according to the manufacturer's instructions.

Intracellular TNF-α and IL-10 Staining.

The ability of monocytes to produce proinflammatory and anti-inflammatory cytokines, namely, TNF-α and IL-10, was assessed before and after LPS stimulation, in the presence or absence of cT-regs, tT-regs, or CD4+CD25+CD127+ T cells. T-regs were added to autologous monocytes (3 × 10⁵/well) at a ratio of 1/8. The two cell populations were cultured in Roswell Park Memorial Institute-1640 medium supplemented with 2 mM L-glutamine, 25 mM 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid, 100 U/mL benzyl penicillin, 0.1 mg/mL streptomycin, 2.5 μg/mL amphotericin B, and 5% heat-inactivated fetal bovine serum in the absence or presence of 100 ng/mL LPS (Sigma-Aldrich Ltd., Ayrshire, UK). After overnight incubation, cells were exposed to Brefeldin A to block secretion of cytokines and induce their intracellular accumulation. After 5 hours at 37°C and 5% CO₂, cells were washed and stained with fluorescein isothiocyanate–conjugated anti-CD14 monoclonal antibodies, permeabilized, and fixed with Cytofix/Cytoperm, counterstained with PE-conjugated anti–TNF-α (BD Biosciences) or anti–IL-10 (BD Biosciences) monoclonal antibodies, washed, and analyzed by flow cytometry. Changes between the percentage of TNF-α–producing and IL-10–producing monocytes before and after T-reg addition were calculated in each individual as fold increase/decrease, expressed as the ratio between the levels of cytokine production after T-reg addition over baseline levels. Percentages of TNF-α–producing and IL-10–producing monocytes are determined after gating the monocyte subpopulation in the forward/side scatter plot.

Statistical Analysis.

The normality of variables distribution was assessed by the Kolmogorov-Smirnov goodness-of-fit test and, once the hypothesis of normality was accepted (P > 0.05), comparison was performed by paired and unpaired Student t tests as appropriate. If the values were not normally distributed, analysis was performed by Wilcoxon rank-sum test or Mann-Whitney test as appropriate. Correlation analysis was determined by Pearson's or Spearman's correlation coefficient. A value of P < 0.05 was considered significant. Results are expressed as mean ± standard error of the mean.

Monocyte Frequency and Phenotype.

The frequency of monocytes (CD14+ cells) in ID (15.7 ± 2.1) and AD (19.6 ± 2.5) patients was higher than in healthy subjects (10.2 ± 1.6; P = 0.044 for both). The frequency of monocytes in ID patients was lower than in AD patients (P = 0.017).

The frequency of HLA-DR–positive monocytes in ID (66.9 ± 3.8) and AD (64.1 ± 4.4) patients was lower than in healthy subjects (94.1 ± 4.5; P < 0.001 for both), whereas the frequency of CD86-positive, CD54-positive, and CD40-positive monocytes did not differ amongst the three groups, exceeding 90% for CD86 and CD54 and 70% for CD40 in all subjects.

In AILD, there was a positive correlation between the frequency of monocytes and AST (R = 0.35; P = 0.03) and IgG (R = 0.75; P < 0.001) levels and a negative correlation between the frequency of HLA-DR–positive monocytes and AST (R = −0.51; P = 0.001) and IgG (R = −0.53; P = 0.002) levels.

Immunohistochemistry.

CD68+, CD3+, and CD79+ cells were detected in the portal tracts of the liver biopsy specimens from all five AIH patients (two ID and three AD). Figure 1 shows CD68+ monocyte staining in one ID (Fig. 1A) and one AD (Fig. 1B) patient. Portal infiltration by T and B cells (Fig. 1C, D) is also shown for the AD patient.

T-reg Phenotype.

In both AILD and healthy subjects, cells with high expression of CD25 (CD25^high), CD45RO (CD45RO^high), and CD62L (CD62L^high) and low expression of CD154 (CD154^low) were also positive for FOXP3, whose expression was lower in patients than in healthy subjects, in accordance with our previous findings.7 The frequency of CD127+ cells within T-regs and the density of CD127 molecules on individual cells (expressed as mean fluorescence intensity) were higher in AILD (38.4% ± 6.9 and 28.1 ± 7.7) than in healthy subjects (19.3% ± 5% and 16.2 ± 3.2; P = 0.03 for both). Figure 2 shows the expression of conventional T-reg markers in one representative ID AIH-1 patient and in one healthy subject.

Chemotaxis Assay.

Monocyte migration (%MM) through the transwell membrane was assessed in 22 AILD patients (13 ID, 9 AD) and in 12 healthy subjects (Table 2). In the absence of chemoattractant, the %MM in patients was higher than in healthy subjects, in both ID (P < 0.0005) and AD (P < 0.001). In the presence of chemoattractant, the %MM remained unchanged in ID patients, it decreased in AD patients (1.2-fold, P = 0.07), and it increased in healthy subjects (three-fold, P < 0.0005). In the presence of chemoattractant and after addition of cT-regs, the %MM increased in both ID (1.4-fold, P = 0.05) and AD (1.5-fold, P = 0.08) patients, whereas it decreased in healthy subjects (1.4-fold, P = 0.017). Addition of cT-regs to the lower chamber, separated from the monocytes by the membrane, had no effect on the %MM in either the presence or absence of chemoattractant. Addition of CD4+CD25− T cells instead of cT-regs either directly to the monocytes or into the lower chamber had no effect on the %MM in the presence or absence of chemoattractant. When tT-regs were added to monocytes, %MM in the presence of chemoattractant decreased in both AILD patients (1.2-fold in both ID and AD) and healthy subjects (1.8-fold; P = 0.03), whereas it increased after addition of CD4+CD25+CD127+ T cells in patients (1.3-fold in ID [P = 0.05] and 1.2-fold in AD) and decreased in healthy subjects (1.4-fold [P = 0.06]), though less markedly than in the presence of tT-regs (Table 2).

Chemotaxis Assay in the Presence of Matrigel Matrix.

Monocyte migration through the matrigel-coated transwell membrane was tested in 21 AILD patients (11 ID and 10 AD) and in 11 healthy subjects (Table 3). In the absence of chemoattractant, the %MM did not differ significantly between the three groups. In its presence, the %MM increased in ID patients (1.4-fold, P < 0.001) and in healthy subjects (2.7-fold, P = 0.002), whereas it decreased in AD patients (1.3-fold, P = 0.09). In the presence of chemoattractant and after cT-reg addition, the %MM increased marginally in ID (1.1-fold) and in AD (1.3-fold) patients, whereas it decreased in healthy subjects (1.7-fold, P = 0.006). Addition of cT-regs or CD4+CD25− T cells to the lower chamber, separated from the monocytes by the membrane, and of CD4+CD25− directly to monocytes had no effect on the %MM in the presence or absence of chemoattractant. Addition of tT-regs decreased the %MM in both patients (1.2-fold in ID and 1.3-fold in AD) and healthy subjects (1.6-fold, P = 0.09), whereas addition of CD4+CD25+CD127+ T cells increased the %MM in patients (2.3-fold in ID [P < 0.001]; 1.7-fold in AD), leaving it unchanged in healthy subjects (Table 3).

Quantification of MMP19 Gene Expression.

The MMP19 gene, a surrogate marker of monocyte ability to cross barriers, was quantified in monocytes purified from 19 patients with AILD (9 ID and 10 AD) and in 13 healthy subjects. MMP19 transcripts were detectable in both AILD and healthy subjects, the relative level of MMP19 mRNA in ID being higher than in AD patients (4.9 ± 1.2 versus 2.6 ± 1.1 [P = 0.054]) and similar to that observed in healthy subjects (5.2 ± 1.9). Whereas no correlation between the expression of MMP19 and the ability of monocytes to migrate was noted in AILD, a positive correlation was observed in healthy subjects (R = 0.8, P = 0.016).

TNF-α and IL-10 Intracellular Staining.

The ability of monocytes to produce proinflammatory and anti-inflammatory cytokines, namely, TNF-α and IL-10, and the effect of T-regs on their production were assessed in 23 AILD patients (12 ID and 11 AD) and in 12 healthy subjects. Before LPS stimulation, the percentage of TNF-α–producing and IL-10–producing monocytes was negligible and was not affected by cT-reg addition. After LPS stimulation (Table 4), the frequency of TNF-α–producing monocytes in AD was higher than in ID (P = 0.0001) patients and in healthy subjects (P = 0.017). After cT-reg addition, the frequency of TNF-α–producing monocytes increased in both AILD patients (3.4-fold [P < 0.001] in ID and 1.5-fold [P = 0.01] in AD) and healthy subjects (2.4-fold [P = 0.016]), remaining higher in AD compared with ID (P = 0.022) patients and healthy subjects (P = 0.055).

After LPS stimulation, the frequency of IL-10–producing monocytes (Table 5) in AD was higher than in ID (P = 0.034) patients and similar to healthy subjects (P = NS). After cT-reg addition, the frequency of IL-10–producing monocytes increased in both AILD patients (twofold [P = 0.014] in ID; 2.1-fold [P = 0.002] in AD) and healthy subjects (2.6-fold [P = 0.006]). No changes in TNF-α–producing and IL-10–producing monocytes were observed when CD4+CD25− T cells were added instead of cT-regs. Before cT-reg addition, the TNF-α–producing/IL-10–producing monocyte ratio was higher in AILD (1 ± 0.1 in ID and 1.8 ± 0.2 in AD, P = 0.007) than in healthy subjects (0.77 ± 0.1; P = 0.07 and P = 0.0005, respectively); after cT-reg addition, it remained higher in AILD (1.05 ± 0.04 in ID and 1.14 ± 0.1 in AD) than in healthy subjects (0.7 ± 0.1; P = 0.0005 and P = 0.006, respectively).

Addition of tT-regs did not affect the frequency of TNF-α–producing monocytes (Table 4) before T-reg addition, whereas it increased that of IL-10–producing monocytes (1.8-fold in ID [P = 0.045]; 1.4 in AD [P = 0.13]; 1.8-fold in healthy subjects [P = 0.009]) (Table 5). In contrast, addition of CD4+CD25+CD127+ T cells increased the frequency of TNF-α–producing monocytes in both AILD (threefold in ID [P = 0.0008]; 2.4 in AD [P = 0.001]) and healthy subjects (2.9-fold [P < 0.001]) (Table 4) but had no effect on that of IL-10–producing monocytes (Table 5). After tT-reg addition, the TNF-α–producing/IL-10–producing monocyte ratio decreased from 1 ± 0.1 to 0.71 ± 0.2 in ID patients, from 1.8 ± 0.2 to 1.5 ± 0.3 in AD patients and from 0.77 ± 0.1 to 0.44 ± 0.03 (P = 0.02) in healthy subjects. In contrast, after CD4+CD25+CD127+ T cell addition, the TNF-α–producing/IL-10–producing monocyte ratio increased to 2.8 ± 0.1 (P < 0.001) in ID to 5.4 ± 1.2 (P < 0.001) in AD and to 1.8 ± 0.2 (P < 0.001) in healthy subjects.

Monocyte TLR4 Expression.

The expression of monocyte TLR4, whose engagement is key to initiation of adaptive immune responses, was assessed on purified monocytes from 27 patients with AILD (15 ID and 12 AD) and 13 healthy subjects. Irrespective of LPS stimulation, the percentage of TLR4-expressing monocytes was greater than 95% in both AILD patients and healthy subjects. Monocyte TLR4 molecule density before LPS stimulation and in the absence of cT-regs was higher in AILD than in healthy subjects (P = 0.048 in ID and P = 0.03 in AD) (Table 6). After cT-reg addition, the TLR4 density increased more in AILD patients (1.5-fold [P = 0.01] in ID; 2.1-fold [P = 0.01] in AD) than in healthy subjects (1.3-fold [P = 0.015]). No change in monocyte TLR4 expression was observed when CD4+CD25− T cells were added instead of cT-regs. Addition of tT-regs had a negligible effect on monocyte TLR4 expression in AILD and decreased it marginally in healthy subjects (1.2-fold [P = 0.15]), whereas addition of CD4+CD25+CD127+ T cells increased it in both AILD (1.5-fold in ID [P = 0.003]; 1.4-fold in AD [P = 0.001]) and in healthy subjects (1.3-fold [P = 0.04]) (Table 6).

After LPS stimulation and in the absence of cT-regs, monocyte TLR4 expression intensity increased in AILD, though significantly only in AD patients (1.2-fold [P = 0.02]), and also in healthy subjects (1.2-fold [P = 0.024]). After LPS stimulation and cT-reg addition, monocyte TLR4 expression increased in both AILD (P = 0.01 in ID and P = 0.004 in AD) and healthy subjects (P = 0.037), this increase being more evident in the former (1.35-fold in ID; 2.1-fold in AD) than in the latter (1.2-fold) (Table 7). Addition of tT-regs had negligible effect on monocyte TLR4 expression in both AILD and healthy subjects (1.2-fold [P = 0.14]). In contrast, addition of CD4+CD25+CD127+ T cells increased monocyte TLR4 expression in both AILD and controls, though more markedly in the former (1.6-fold in ID [P = 0.01]; twofold in AD [P = 0.0007]) than in the latter (1.2-fold [P = 0.04]).

Table 8 represents a summary of monocyte immune responses in AILD and in health. Results were similar for patients with AIH-1 or autoimmune sclerosing cholangitis.