Expression of Cytokines (TNF-α,IL-1α,and IL-2) in Chronic Hepatitis C: Comparative Hybridocytochemical and Immunocytochemical Study in Children and Adult Patients

Abstract

Hepatitis C virus (HCV) is one of the principal causes of hepatitis, which in more than 80% of cases leads to chronic lesions in the liver and involvement of extrahepatic organs. It remains unknown why the infection so frequently turns chronic, independently of patient age. Using immunocytochemistry (IHC) and in situ hybridization (ISH) (both linked to the ImmunoMax technique) we examined cell sources of TNF-α, IL-1α, and IL-2 in control and HCV-infected children and adults. We demonstrated augmented expression of all the cytokines in HCV-infected patients compared to controls. No differences were detected in amounts of studied transcripts or cytokine proteins between biopsies taken from HCV-infected children and adults. Expression of TNF-α was localized mainly in liver sinusoidal cells (macrophages, endothelial cells). A high proportion of hepatocytes demonstrated expression of TNF-α, IL-1α, and IL-2. In both groups of patients, higher amounts of cytokine proteins than studied transcripts were demonstrated. The augmented expression of TNF-α, IL-1α, and IL-2 in liver with a similar proportion of involved cells (mainly hepatocytes) in children and in adults points to participation of the cytokines in the pathogenesis of chronic hepatitis C. The expression is insufficient to terminate the infection and may be linked with the comparably frequent chronic transformation of HCV infection noted in children and adults.

Keywords

chronic hepatitis C TNF-α IL-1α IL-2 children adult patients grading staging immunocytochemistry in situ hybridization ImmunoMax technique

An increase in serum levels of pro-inflammatory cytokines has been noted in liver diseases, including chronic hepatitis C (Tilg et al. 1992; Larrea et al. 1996). The most important pro-inflammatory cytokines, TNF-α and IL-1(α and β) are expressed in the liver mostly by resident macrophages (Kupffer cells) and T-cells (Beutler and Cerami 1989; Winwood and Arthur 1993; Hoffmann et al. 1994). Apart from Kupffer cells, macrophages, newly recruited from circulating monocytes, may become an appreciable cytokine source (Tsukamoto 1999). Production of pro-inflammatory cytokines was demonstrated also in endothelial cells of liver sinusoids, in activated hepatic stellate cells (HSCs), and in biliary epithelial cells (Hoffmann et al. 1994; Tsukamoto 1999). Hepatocytes were also shown to be capable of augmented production in response to liver injury, including viral injury (Gonzalez-Amaro et al. 1994; Oyanagi et al. 1999; Kasprzak et al. 2002). The role of the hepatocyte-derived cytokines remains unknown. Studies on intrahepatic expression of inflammatory cytokine genes in human chronic hepatitis C virus (HCV) infection showed that HCV liver injury was associated with an increase in hepatic expression of the Th1 cytokines IL-2 and IFN-7 (Napoli et al. 1996; Bertoletti et al. 1997; McGuinness et al. 2000). The augmented hepatic expression of the cytokines correlated with grading and staging (Napoli et al. 1996). Untreated patients with chronic hepatitis C showed increased TNF-α mRNA levels in the liver and mononuclear cells compared to healthy controls (Larrea et al. 1996; McGuinness et al. 2000). Other studies on intrahepatic expression of IL-1β and TNF-α in chronic HCV infection showed that immunoreactivity of both cytokines was predominantly localized in sinusoidal cells, with IL-1β immunoreactivity and mRNA levels being lower in chronic HCV than in control samples. In contrast, no significant differences in TNF-α mRNA levels were observed between chronic hepatitis C samples, primary biliary cirrhosis (PBC), or control samples (Dumoulin et al. 1999).

Because of the lack of comparative studies on cytokine expression in children and in adults with chronic hepatitis C and relatively few literature data on cell sources of the proteins, we have decided to continue our earlier studies (Kasprzak et al. 2002) on the model of chronic hepatitis C infection.

The aim of the study was to investigate the intrahepatic localization of mRNAs for pro-inflammatory cytokines produced mainly by macrophages, TNF-α and IL-1α and their respective proteins and the T-helper type 1 (Th1)-associated cytokine IL-2 in liver specimens from children and adults chronically infected with HCV by use of a non-isotopic variety of ISH, ICC alone, and both combined with ImmunoMax technique to determine the cellular source of these cytokines. We planned to compare expression of the mentioned cytokines between children and adults with chronic hepatitis C on the one hand and control individuals on the other. Attempts were also made to correlate the cytokine expression with (a) inflammation activity (grading) and/or advancement of fibrosis (staging), and (b) serum alanine transaminase (ALT) levels in groups of children and adult patients.

Materials and Methods

Patient Selection

Twenty-one white patients with chronic hepatitis C and with active replication of the virus, confirmed serologically [anti-HCV-positive in all children and all adults; HCV RNA positive in all children, including the quantitative tests, and in all adults, in whom in eight cases quantitative tests and in the remaining four cases only qualitative tests were performed using an ELISA third-generation assay (Abbott; Highland Park, IL)] for anti-HCV antibody, and qualitative assays of HCV RNA using the AMPLICOR HCV test, version 2.0 (Roche; Mannheim, Germany). The test was standardized against the WHO International Standard for HCV RNA. Quantitation of HCV RNA was conducted using the Digene Sharp Signal System (Gaithersburg, MD). The mean age of children (five boys, four girls) at time of liver biopsy was 14 ± 1 year (range 7-16 years) and the mean age of adult patients (four men, eight women) was 48 ± 5 years (range 22-74 years). All the tests were routinely performed for diagnostic reasons and before the anti-viral therapy. In addition, all patients' liver biopsies were positive for two HCV proteins (C and NS3) (n = 19) or at least one HCV protein (NS3) (n = 2) using ICC (data not shown). The biopsies originated from children of the Outpatient Clinic of Liver Disease, Specialized Unit for Health Care over Mother and Child in Poznań, Poland and from adult patients of the Chair of Gastroenterology and Human Nutrition, University of Medical Sciences in Poznań and Ward of Infectious Diseases, Józef Strus Hospital in Poznań, in whom the biopsies were performed in 1999-2002. All patients were seronegative for both HBsAg and HBeAg and for IgM antibody against cytomegalovirus (CMV) or against Epstein-Barr virus (EBV) by standard enzyme immunoassays (ELISA). Other cases of liver damage (e.g. α₁-antitrypsin deficiency, Wilson's disease, alcohol dependence) were ruled out. All the patients had exhibited elevated serum ALT activity (normal level below 40 IU/liter) for at least 6 months. The principal criterion for selection of patients for studies on cytokine expression involved active virus replication, persisting for over 6 months, and histopathological lesions in liver biopsies in the form of chronic hepatitis C of various grades and staging induced by the infection. It should be mentioned that for ethical reasons no liver biopsies in healthy humans are permitted in Poland. Therefore, the control group consisted of liver biopsies from serologically HCV-, HBV-, HCMV-, EBV-, and HIV-negative patients with nonspecific changes in the liver (n = 4, children) or chronic hepatitis of unknown etiology (n = 2, adults) (mean age of the control group was 24 ± 8 years, range 9-53 years). In all patients from the control group, normal ALT activity was demonstrated on the day of liver biopsy (below 40 IU/liter). Paraffin sections of reactive lymph nodes from three patients without liver disease were used as a positive control.

Tissue Sample Processing

In total, 27 liver biopsy specimens were obtained during routine diagnostic liver biopsies after receiving informed consent from the patients. Tissue specimens were fixed in 10% formalin and embedded in paraffin for light microscopy. Histopathological lesions were evaluated after the classical H+E staining as well as the silver technique and the trichromate technique according to Masson and periodic acid-Schiff with diastase pretreatment. At least 10 sections were prepared from each biopsy. Each biopsy specimen was evaluated based on a simple numerical scoring system for the grade of portal/periportal necroinflammation (G1) (0-4), the grade of lobular necroinflammation (G2) (0-4), and for the stage of fibrosis (S) (0-4), as proposed by Scheuer et al. (2002). The histological lesions considered to be characteristic of chronic hepatitis were also evaluated. Histological evaluation was performed independently by two experienced histopathologists.

In Situ Hybridization

For the studies, 5-μm-thick sections were cut and mounted on SuperFrost/Plus microscope slides from the total of 27 liver biopsies. Two digoxygenin-labeled oligoprobes specific for human cytokine mRNA were used, including (a) human TNF-α probe cocktail (R&D Systems; Minneapolis, MN), (b) human IL-1α probe cocktail (R&D Systems), detected with sheep anti-digoxygenin monoclonal antibodies (MAbs) (Fab fragments), labeled with horseradish peroxidase (Roche). In the hybridocytochemical studies, the classical ones and those amplified by the ImmunoMax technique (Komminoth and Werner 1997), sequential sections of the tissue material were applied, which was pre-tested by ICC techniques. The ISH protocol of R&D Systems was employed with our own modification and ImmunoMax amplification of the signal. Subsequent stages involved paraffin removal from the section using xylene, ethanols, and washing in diethyl pyrocarbonate (DEPC) (Sigma Chemical; St Louis, MO)-treated distilled water, each lasting 1-5 min, and washing in 0.05 M Tris (AccuGene; Landover, MD) for 5 min at room temperature (RT). Endogenous peroxidase was blocked by a 15-min incubation of the section in 1% H₂O₂ at RT. Then the sections were digested with proteinase K (DAKO; Carpinteria, CA) 100-200 μg/ml for 10-15 min at 37C. Next, the sections were washed in 1 × PBS for 10 min at RT and fixed again in 0.4% PFA at 4C for 15 min. They were washed in DEPC-distilled water for 1 min each. The prehybridization solution (Sigma) was applied under a coverslip and the preparation was incubated for around 1 hr at 37C. Various concentrations of the oligoprobe were tested, ranging from 50 to 500 ng/ml, dissolved in the prehybridization solution plus 30% formamide. The hybridization itself took place overnight (16-18 hr) at around 38C under a coverslip in a humid chamber. After removal of the coverslips, the sections were washed in decreasing concentrations of SSC (Promega; Madison, WI) plus 30% formamide (Sigma), and then in 0.1 M Tris-HCl, pH 7.6, and 0.1% Triton X-100 in TBS for 15 min at RT. Sheep anti-digoxygenin MAb-horse radish peroxidase (HRP) (Roche) complex was applied for around 45 min at RT at a dilution of 1:30. After a wash in PBS/Tween-20, biotinylated tyramine (Perkin-Elmer Life Sciences; Boston, MA) was applied at 1:50 dilution for around 3 min at RT. The streptavidin-peroxidase complex was added for 30 min at RT. The hybrids were detected using peroxidase substrate, 0.05% DAB (DAKO) in 0.05 M Tris-HCl, pH 7.6, supplemented with 0.001% H₂O₂ for 5 min at RT. Some preparations were counterstained with hematoxylin and mounted with a glycerol-based mounting medium. As controls for ISH, lymph nodes from three healthy donors served as initial positive controls. Negative controls included (a) hybridization without addition of the molecular probe and (b) incubations of slides in an RNase A solution (R&D Systems), 20 mg/ml in 2 × SSC/10 mM MgCl₂, pH 8, for 30 min at RT before hybridization. All the experiments were performed at least twice.

Immunohistochemistry

The following specific MAbs were used: (a) mouse anti-human TNF-α antibody (R&D Systems), (b) goat anti-human IL-1α antibody (R&D Systems), (c) mouse anti-IL-2 antibody (R&D Systems), (d) mouse anti-human macrophage antibody CD68 (DAKO), and (e) anti-human Von Willebrand Factor antibody (DAKO Epos; Glostrup, Denmark). The first stages of the ICC reaction, up to blocking of endogenous peroxidase, followed the procedure applied for hybridocytochemistry. Then the sections were treated with primary MAb at 1:20 dilution (anti-IL-1α) and 1:100 dilution (anti-TNF-α and anti-IL-2) or ready-for-use dilutions (anti-CD68 and anti-human Von Willebrand Factor) overnight at 4C, then with the secondary biotinylated link of anti-mouse and anti-rabbit IgG and with the streptavidin-biotin-peroxidase complex (ABC) (DAKO). The studies followed the classical ABC technique (Hsu et al. 1981), alone or associated with the ImmunoMax technique (Komminoth and Werner 1997). In our studies we did not use antigen retrieval but amplification of the final signal using biotinylated tyramine. In the technique, termed the ImmunoMax technique, the key reaction involved a 3-min incubation with biotinylated tyramine (Perkin-Elmer) at 1:50 dilution at RT. This was followed by another application of the streptavidin complex. The color reaction was evoked with the HRP substrate 0.05% 3,3-diaminobenzidine tetrahydrochloride (DAB; DAKO) in 0.05 M Tris-HCl buffer, pH 7.6, supplemented with 0.001% H₂O₂. In tissue sections, endogenous peroxidase was blocked with 1% H₂O₂. Control reactions employed control sera of the respective species in 0.05 M Tris-HCl, pH 7.6, supplemented with 0.1% BSA and 15 mM sodium azide (negative control) (DAKO). Most of the histological preparations were subjected to double ICC reactions, including a macrophage marker (CD68) or a marker for endothelial cells (Von Willebrand Factor) and one of the cytokines. The terminal reaction product was visualized with DAB (brownish-black signal) or DAB supplemented with nickel (blue-gray colour). Some preparations were counterstained with hematoxylin and mounted with a glycerol-based mounting medium.

Quantification of ISH and ICC Results and Statistical Analysis

The contents of cytokines and of respective mRNAs in liver biopsies demonstrated using hybridocytochemistry and the ABC method, both combined with the ImmunoMax technique, were evaluated by the semiquantitative technique, relating the score of 0 to 4 points to the fraction of stained cells (score 0, 0% cells; 1, less than 5% cells; 2, 5-20% cells; 3, 20-40%; and 4, more than 40% positive cells). It should be mentioned that semiquantitative evaluation was conducted exclusively in preparations obtained using the ImmunoMax technique (both in ICC and in ISH), which yielded an intense color reaction. The preparations were examined under a light microscope at ×400 magnification. In each section of the liver biopsy, five fields in hepatic lobules and the periportal area and at least five different portal spaces in each specimen were examined. The calculations were performed on the mean of three consecutive sections of liver biopsy material from HCV-infected patients and resulted in mean numbers separate for the lobular/periportal area and the portal area. Cellular localization of the studied proteins and respective transcripts was evaluated (cell nucleus, cytoplasm, membranes).

StatXact Software (Cytel Software; Cambridge, MA) was used for statistical analysis of small data sets. To determine statistical significance of variations in percentage of cells with transcripts and proteins studied using the ImmunoMax technique, we first calculated the mean values of immunostaining or hybridocytochemical signal scores for both liver biopsy groups (adults and children) as well as control biopsy specimens, separately for the lobular/periportal area and the portal spaces. The values were then compared using the Mann-Whitney U-test for nonparametric independent data and the Wilcoxon test for nonparametric dependent data. Correlations between data rows were determined with Spearman's rank correlation index. In determining the correlations, means of the total expression of cytokine mRNA or of the respective protein were taken into account (i.e., expression in the lobules, periportal and portal spaces) and related to the sum of grading (G1+G2), staging (S), and ALT, respectively, in the groups of children and adult patients.

Results

Histology

All 21 liver biopsy samples of anti-HCV- and HCV RNA-positive patients were diagnosed by histological criteria as consistent with chronic hepatitis C. Both the grade of inflammation and the stage of fibrosis were significantly lower in children than in adults [grade of inflammation (G1 + G2) 1.9 ± 0.3 vs 3.7 ± 0.4, respectively; p<0.004)] and stage of fibrosis (1.1 ± 0.3 vs 2.5 ± 0.3, respectively; p<0.008) (Table 1). For histological patterns, large groups of lymphoid cells in the dilated portal spaces prevailed in adult patients (11/12). Infiltrates, consisting of individual lymphoid cells or groups, as well as traits indicative of fibroblast proliferation, were also observed in hepatic lobules. In both adults and children, many biopsies demonstrated enlarged hepatocytes, sometimes with evident fatty degeneration involving the entire cell. Moreover, pathology of hepatocyte nuclei was often observed (variable size and irregular shape, altered chromatin arrangement, many nucleoli, lesions in the nuclear envelope, frequent binary cell nuclei, variable staining). In control liver biopsies, nonspecific changes in the liver (n=4) and hepatitis (n=2) were observed, with individual cells of inflammatory infiltrate in portal spaces (final grading 1.2 ± 0.2) but with no traits of liver fibrosis (n=4) or with fibrosis restricted to portal spaces (n=2). Mean staging for all patients was 0.3 ± 0.2.

ISH

Hybridocytochemical studies were conducted to detect mRNAs for two cytokines, TNF-α and IL-1α. The TNF-α transcript was demonstrated in 4/9 children and in 8/11 adult patients. mRNA for IL-1α was detected in 7/9 children and in 5/7 studied adults (Tables 2 and 3). In both groups of patients, lobular localization of both transcripts and of their respective proteins generally prevailed over expression within portal spaces (Table 4). Transcripts for both cytokines were located first of all in cells of liver sinusoids (macrophages, endothelial cells) and in individual cells of inflammatory infiltrates (lymphoid cells, macrophages) in individual patients (Figures 1A and 1B). Although in a lower percentage of cells, transcripts for both cytokines were also observed in hepatocytes. The reaction product was detected mainly in the cell cytoplasm. Compared to the detected amounts of cytokines, a much smaller percentage of cells with transcripts in both groups of patients was detected (Table 4). No significant differences were detected in the percentage of cells with transcripts for the two cytokines between children and adult patients and between expression of TNF-α mRNA and IL-1α mRNA in individual groups of patients (Tables 4 and 5). Control liver biopsies contained no cytokine transcripts. Comparative analysis demonstrated significantly more abundant TNF-α mRNA of lobular localization in adults with HCV compared to the control (p<0.05) and no differences compared to children. Compared to control, IL-1α mRNA in the lobuli was more abundant in both children and adults (p<0.05 in each case). No significant differences were detected between expression of the transcripts for the two cytokines in portal spaces between HCV-infected patients and control specimens (Table 4).

Table 1

Epidemiological, histological, and biochemical findings in chronic HCV-infected patients a

	Children	Adults
	(n = 9)	(n = 12)	p value
Age (mean ± SE, yr)	14 ± 1	48 ± 5	<0.004
Gender (male/female)	5/4	4/8
Grade of inflammation (mean ± SE)	1.9 ± 0.3	3.7 ± 0.4	<0.004
Stage of fibrosis (mean ± SE)	1.1 ± 0.3	2.5 ± 0.3	<0.008
ALT (mean ± SE, IU/liter)	110 ± 21	106 ± 23	NS

^aNS, statistically nonsignificant.

Table 2

Histological findings and semiquantitative appraisal of TNF-α, IL-1α, and respective mRNAs and IL-2 in liver biopsies of children with chronic hepatitis C a

	Score of histological diagnosis b			Cytokines
Patient/age (yr)/sex	G1	G2	S	TNF-α mRNA/Protein c		IL-1α mRNA/Protein		IL-2 Protein
1. PW; 14,M	1	0	0	2/0	3/0	2/0	4/0	4/0
2. AM; 7,F	1	2	3	1/0	2/0	0/0	0/0	4/0
3. SK; 15,F	1	0	2	2/0	2/0	2/0	3/0	4/0
4. ND; 15,F	1	0	0	2/0	1/0	2/0	3/0	3/0
5. PW; 15,M	1	2	0	0/0	2/0	1/0	3/0	4/0
6. KD; 13,M	1	1	1	0/0	0/0	1/0	1/0	Nt
7. JM; 15,F	1	0	1	0/0	4/0	0/0	1/0	4/1
8. PG; 16,M	2	1	1	0/0	3/0	1/2	1/0	4/1
9. PK; 14,M	1	1	1	0/0	3/0	2/0	2/0	4/0

^aG1, grading in portal spaces; G2, grading in lobules; S, staging.

^bScoring system: see Materials and Methods. 0, negative signal; 1, less than 5% positive cells; 2, 5–20% positive cells; 3, 20–40% positive cells; 4, more than 40% positive cells; nt, not tested.

^cResults are given in two numbers, the first denoting expression in cells of hepatic lobules (Kupffer cells, endothelium, hepatocytes) and the other expression in cells of inflammatory infiltrates in portal spaces (lymphocytes, macrophages, endothelium).

Table 3

Histological findings and semiquantitative appraisal of TNF-α, IL-1α, their respective mRNAs, and of IL-2 in liver biopsies of adults with chronic hepatitis type C

	Score of histological diagnosis a			Cytokines b , c
Patient/Age (yr)/Sex	G1	G2	S	TNF-α mRNA/Protein		IL-1α mRNA/Protein		IL-2 Protein
1. AG; 41,F	2	2	2	0/1	2/0	0/0	3/1	3/1
2. WS; 22,M	1	2	2	2/0	2/0	2/1	3/0	2/0
3. MS; 39,M	1	2	2	1/0	2/0	2/0	2/0	3/0
4. JP; 55,F	3	2	4	1/0	4/0	2/0	4/1	4/0
5. JS; 51,F	2	3	4	0/0	3/0	nt	3/0	4/0
6. JK; 54,F	2	1	3	nt	3/2	nt	3/0	3/2
7. MB; 29,M	1	2	3	0/0	2/0	nt	4/0	Nt
8. IP; 35,F	1	1	1	1/0	1/0	1/0	1/0	Nt
9. BS; 56,F	2	2	2	0/0	0/0	nt	4/2	4/0
10. ER; 74,M	3	3	4	2/1	3/0	nt	nt	Nt
11. KF; 72,F	2	2	2	0/0	3/1	0/0	4/0	4/0
12. RR, 43,F	1	1	1	1/0	1/0	0/0	4/0	4/1

^aG1, grading in portal spaces; G2, grading in lobules; S, staging.

^bScoring system: see Materials and Methods. Score: 0, negative signal; 1, less than 5% positive cells; 2, 5–20% positive cells; 3, 20–40% positive cells; 4, more than 40% positive cells; nt, not tested.

^cResults are given in two numbers, of which the first denotes expression in lobules (Kupffer cells, endothelium, hepatocytes) and the other expression in inflammatory infiltrates in portal spaces (lymphocytes, macrophages, endothelium).

Immunocytochemistry

TNF-α in liver biopsies from patients with chronic type C hepatitis was detected in 8/9 children and 11/12 adults examined (Tables 2 and 3). Expression of TNF-α was noted mostly in hepatocytes and only in individual cells of liver sinusoids (macrophages, endothelial cells) and in few inflammatory cells in portal spaces (lymphocytes, macrophages) (Figures 2A and 2B). Expression of IL-1α was present in liver biopsies in 8/9 children and in all adult patients examined (n = 11). In children, the proportion of IL-1α-containing cells did not differ from the frequency of cells with expression of TNF-α. More IL-1α was detected in lobules in adults compared to children (p<0.04) but the total content of IL-1α did not differ between children and adult patients (Table 4). In adult patients, the product of the ICC reaction for IL-1α was observed in a significantly greater percentage of cells than cells containing TNF-α (p<0.03) (Table 5). In both groups of patients, signal for IL-1α was observed in the cell population similar to that containing TNF-α. IL-2 was detected in all children studied (n = 8) and in all adults studied (n = 9). The product of the ICC reaction for IL-2 was also observed mostly in hepatocytes and only in individual sinusoidal cells or inflammatory cells. It involved a significantly greater percentage of cells compared to TNF-α-carrying cells in children and adults (p<0.02 and p<0.03, respectively) and, also, a significantly greater number of cells than that of IL-1α-positive cells in children (p<0.03) (Table 5). The signal was strong and evenly distributed in cytoplasm of hepatocytes or in one of the cell poles (Figure 2C). Application of the ABC technique to localize macrophages in liver biopsies studied (anti-CD68) demonstrated a typical distribution of the cells in liver sinusoids and, in low numbers, in portal spaces (Figure 3). This was accompanied by differences in the total macrophage numbers among individual biopsies. The source of all cytokines studied and of the two transcripts was identified by double staining with the antibody to CD68 or anti-Von Willebrand Factor (Figure 4). In lymph nodes (n = 3), expression of TNF-α, IL-1α, and IL-2 was intense even with classical hybridoimmunocytochemical techniques. The reaction was present in many cells of the CD68 phenotype, in lymphocytes, and in endothelial cells. In the control liver biopsies (n = 6), only scattered individual cells with all cytokines were detected among hepatocytes and sinusoidal cells (Figure 5). Significantly higher amounts of cytokines of lobular localization were demonstrated in children and in adults with chronic hepatitis C compared to control biopsies (p<0.05) (Table 4). The studies confirmed a higher sensitivity of the ImmunoMax technique compared to classical ICC and hybridocytochemical techniques, because the percentages of transcript-positive as well as all cytokine-positive cells detected with the ImmunoMax technique were about twofold higher than those detected using the classical version of ICC technique (data not shown).

Table 4

Comparative hybridocytochemical and immunocytochemical scoring of TNF-α, IL-1α, and IL-2 in children and adults with chronic hepatitis C and in control biopsy specimens (± SE)

	Children (n = 9)			Adults (n = 12)			Control (n = 6)
Cytokines	L		PS	L		PS	L		PS
mRNA TNF-α	0.78 ± 0.32		0.00 ± 0.00	0.73 ± 0.24 b	0.18 ± 0.12	0.00 ± 0.00		0.00 ± 0.00
		NS			NS			NS
TNF-α	2.22 ± 0.40 b , c	0.00 ± 0.00	2.17 ± 0.32 b , d	0.25 ± 0.18	0.50 ± 0.22		0.17 ± 0.17
		p<0.02			p<0.004			NS
mRNA IL-1α	1.22 ± 0.28 b		0.22 ± 0.22	1.00 ± 0.38 b		0.14 ± 0.14	0.00 ± 0.00		0.00 ± 0.00
		p<0.05			NS			NS
IL-1α	2.00 ± 0.44 b , c		0.00 ± 0.00	3.18 ±0.30 a , d		0.36 ± 0.20	0.50 ± 0.22		0.00 ± 0.00
		p<0.02			p<0.004			NS
IL-2	3.88 ± 0.13 b		0.25 ± 0.16	3.44 ± 0.24 b		0.44 ± 0.24	0.83 ± 0.17		0.17 ± 0.17
		p<0.02			p<0.008			NS

^a p<0.04 vs the respective group of patients.

^b p<0.05 vs HCV-infected children and adults and controls.

^c p<0.05 between protein and mRNA of respective cytokine in each group.

^d p<0.02 between protein and mRNA of respective cytokine in each group.

NS, statistically not significant; L, cytokine expression in hepatic lobules; PS, cytokine expression in portal spaces.

Figure 1

Fragments of hepatic lobules from patients with chronic HCV infection in hybridocytochemistry combined with the ImmunoMax technique. (A) IL-1α mRNA in sinusoidal cells (mainly Kupffer cells and endothelial cells). Bar = 20 μm. (B) TNF-α mRNA in cells of portal spaces (macrophages, lymphocytes, endothelial cells). Bar = 40 μm.

Figure 2

Fragments of hepatic lobules from patients with chronic HCV infection in ICC combined with the ImmunoMax technique. (A) localization of TNF-α in hepatocytes from cirrhotic liver. (B) TNF-α expression in individual sinusoidal macrophages. (C) IL-2 expression in a group of hepatocytes. Bars = 20 μm.

Figure 3

Fragment of hepatic lobule from adult patient with chronic HCV infection, immunostained with anti-CD68 antibodies. ABC method. Bar = 40 μm.

Table 5

Comparative immunocytochemical scoring of TNF-α, IL-1α, and IL-2 in hepatic lobules in children and adults with chronic hepatitis C a

	Cytokines in hepatic lobules
	TNF-α	IL-1α	IL-2
Children (n = 9)	2.22 ± 0.40	2.00 ± 0.44	3.88 ± 0.13 b , c
Adults (n = 12)	2.17 ± 0.32	3.18 ± 0.30 d	3.44 ± 0.24 b

^aData are expressed as means ± SE.

^bDifference between TNF-α and IL-2 (in children p<0.02; in adults p<0.03).

^cDifference between IL-1α and IL-2 (p<0.03).

^dDifference between TNF-α and IL-1α (p<0.03).

Correlations Between Expression of Cytokines and Grading/Staging and ALT

In none of the group of patients could correlation be noted between expression of cytokines on the one hand and grading and ALT activity on the other. In adults, a positive correlation was detected between expression of TNF-α protein and staging (Spearman correlation coefficient of 0.689; p<0.05).

Discussion

HCV continues to represent the main causative agent of the hepatitis which, independently of patient's age, leads to chronic transformation of the process in over 80% patients (NIH Consensus Statement 1997). The mechanisms remain unknown that result in such a high frequency of chronic transformation of infection with this type of virus, independent of patient age, and in such infrequent permanent cures after treatment of the disease with, e.g., interferon-alpha (IFN-α). In discussions on the pathogenesis of chronic hepatitis C, involvement of several cytokines has for a long time been taken into account (Bertoletti et al. 1997; Shapiro et al. 1998; Tsukamoto 1999; McCaughan et al. 2000). An increase in serum levels of pro-inflammatory cytokines has been noted (Tilg et al. 1992; Larrea et al. 1996) as well as upregulation of intrahepatic Th1-like cytokines (Napoli et al. 1996). HCV may be resistant to inhibition by cytokines, and cytokines may therefore have a more prominent role in liver damage than control of or termination of viral replication (Shapiro et al. 1998; Koziel 1999).

For the present studies, three cytokines were selected (TNF-α, IL-1α, and IL-2). TNF-α is required for normal proliferation of hepatocytes in liver regeneration and it exhibits anti-apoptotic activity (Simpson et al. 1997) but, on the other hand, it represents a well-recognized mediator of hepatocyte death (Zhu et al. 1998). IL-1 plays other roles in the liver, including induction of the acute-phase response and autoactivation of Kupffer cells (Winwood and Arthur 1993). IL-2 is produced mainly by Th1 cells and its role involves stimulation of proliferation and activation of most T-lymphocytes, NK-cells, and B-lymphocytes. IL-2 is also known to function in some non-lymphoid cells (reviewed in Smith 1984; Peters 1996). Moreover, normal and neoplastically transformed epithelial cells can express endogenous IL-2, which is processed and secreted in the same way as in hematopoietic cells (Reichert et al. 1998). The few authors who have examined intracellular expression of cytokines in chronic hepatitis C in vivo mention liver sinusoidal cells and cells of inflammatory infiltrates as the most common sources of the cytokines (Gonzalez-Amaro et al. 1994; Dumoulin et al. 1999; Oyanagi et al. 1999). Under certain conditions, other cells, such as endothelial cells, HSCs, bile duct epithelial cells, and hepatocytes are able to synthesize TNF-α and IL-1 (Gonzalez-Amaro et al. 1994; Tsukamoto 1999; Kasprzak et al. 2002). Opinions on the amount of cytokines detected in livers with chronic hepatitis C or on correlation of cytokine expression on the one hand and grading and/or staging and/or intrahepatic viral load on the other are divergent (Gonzalez-Amaro et al. 1994; Larrea et al. 1996; Napoli et al. 1996; Dumoulin et al. 1999; Kinnman et al. 2000). No comparative studies have been conducted on intrahepatic expression of cytokines in chronic hepatitis C in children or adults, particularly those employing direct visualization of transcripts and/or proteins at the cellular level. Compared to control biopsies, our studies have demonstrated more pronounced expression of IL-1α mRNA in both groups of HCV-infected patients and TNF-α mRNA in adult patients, as well as of all cytokines in children and in adults with chronic hepatitis C. Comparison of cytokine expression in children with that in adults has demonstrated more intralobular IL-1α in adults, but total expression (in lobules and in portal spaces) did not differ between the two groups of patients, similarly to expression of TNF-α and IL-2. Augmented expression of transcripts for the two cytokines has been observed mainly in cells of liver sinusoids (macrophages and endothelial cells), in occasional cells of inflammatory infiltrate, and in a few hepatocytes, which has been partially consistent with earlier reports (Gonzalez-Amaro et al. 1994; Larrea et al. 1996; McGuinness et al. 2000). A similar localization of transcripts for cytokines in chronic HCV infection has also been noted by Dumoulin et al. (1999), but in their hands IL-1β immunoreactivity and mRNA levels were less pronounced in chronic HCV than in control samples. In contrast, no significant differences in TNF-α mRNA levels were observed between chronic hepatitis C samples, primary biliary cirrhosis (PBC), or control samples (Dumoulin et al. 1999). Compared to our observations, the differences might be explained by the fact that the authors studied cytokine expression in a relatively large group of patients with hemophilia. Even though we have observed a higher percentage of cytokine-producing cells in the liver sinusoids compared to controls, the observation has been paralleled by detection of a markedly lower percentage of cells with TNF-α and IL-1α transcripts compared to the amounts of the respective cytokine proteins, and unsuspected localization of the latter has involved mainly hepatocytes rather than cells in liver sinusoids. The participation of hepatocytes in production of cytokines in viral hepatitis has been documented with increasing frequency (Gonzalez-Amaro et al. 1994; Oyanagi et al. 1999; Kasprzak et al. 2002). TNF-α biosynthesis is regulated at transcriptional and post-transcriptional levels, and it is possible that TNF-α is processed by a different pathway in hepatocytes than in macrophages or T-lymphocytes (Kriegler et al. 1988; Beutler and Cerami 1989; Santis et al. 1992). Our studies have demonstrated expression of additional cytokines, i.e., IL-1α and IL-2, in HCV-infected hepatocytes. A probable although still unclear hypothesis remains their participation in the pathogenesis of infection. According to some authors, in patients with chronic hepatitis C the elevated serum TNF-α reflects the intensity of inflammatory lesions (grading) and decreases after anti-viral treatment (Neuman et al. 2002). In both groups of our patients, no correlation was observed between grading and cytokine expression, despite higher grading in adults. This points to participation of additional cytokines or other factors in intensification of inflammation. A significant involvement of TNF-α has also been suggested in the process of liver fibrosis in chronic hepatitis C (Kinnman et al. 2000). Both TNF-α, and IL-1 are known to belong to the group of cytokines capable of stimulating proliferation of Ito cells (HSCs), cells of key importance in the process of liver fibrosis (Tsukamoto 1999). Only in our adult patients with chronic hepatitis C has positive correlation been observed between TNF-α and staging. However, participation of other pathogenic factors (in addition to HCV and elevated TNF-α) in production of higher staging in adults cannot be excluded. Our studies have demonstrated a particularly high expression of IL-2 in hepatic lobules, both in children and in adult patients with chronic hepatitis C, compared to TNF-α and, in the case of children, also to IL-1α. The augmented IL-2 expression in patients with chronic hepatitis C is believed to provide evidence for an intrahepatic response of Th1, but cell sources of the cytokine have not always been documented (Napoli et al. 1996; Dumoulin et al. 1997; McCaughan et al. 2000). In our study, expression of the IL-2 dominated in hepatocytes, which has been unsuspected and, until now, such a high expression of the cytokine in liver lobules (and in hepatocytes) has found no confirmation in the literature. The role of augmented IL-2 expression in hepatocytes of patients with chronic hepatitis C requires further studies, including those using techniques of molecular biology. The very low expression of cytokines in cells of inflammatory infiltrates in our patients (both in portal spaces and in the lobules) might reflect weak stimulation of mononuclear cells, low numbers of infiltrating activated macrophages (CD68), or disturbed production of pro-inflammatory cytokines by cells of the infiltrate. Studies performed in vitro using the protein NS3 to stimulate responses of T-cells from infected individuals have shown both an absence of IL-2 protein secretion by the lymphocytes and a failure to amplify IL-2 mRNA on stimulation (Eckels et al. 1999). Other authors have documented a correlation between increased IL-2 expression and increased IFN-7 expression or grading/staging in chronic hepatitis C infection (Napoli et al. 1996). Our studies have not confirmed this observation.

Figure 4

Fragment of hepatic lobule from child with chronic HCV infection with double staining for IL-2 (blue-gray hepatocytes) and with anti-CD68 antibodies (brown macrophages). ABC method. Bar = 8 μm.

Figure 5

Fragment of hepatic lobule from control biopsy specimen, with scattered cells with TNF-α expression. ABC method. Bar = 20 μm.

For studies on cytokine localization in liver, this study has for the first time employed the ImmunoMax technique, which has increased sensitivity above the classical ICC and ISH tests. Compared to controls, augmented expression of TNF- α, IL-1α, and IL-2 has been demonstrated in patients with chronic hepatitis C. On the one hand this indicates the role of cytokines in the chronic inflammatory process and, probably, also in a mechanism of liver regeneration. On the other hand, it reflects inefficient anti-viral defense. The comparable level of expression of the cytokines in HCV-infected children and adults, confronted with higher grading and staging noted in adult patients, points to participation of additional cytokines or other factors that intensify degeneration of the organ in HCV-infected adults. On the basis of our results, it is difficult to clarify causes of the high expression of proteins of the three cytokines in hepatocytes and only low levels of transcripts of such a localization. The result suggests that, in addition to the classical ones, other pathways of cytokine expression operate in various cells, e.g., under the effects of active HCV replication.

In this study, cellular localization has for the first time been compared and expression of TNF-α, IL-1α, and IL-2 has been semiquantitatively appraised in children and adults with chronic hepatitis C. We have confirmed expression of TNF-α in hepatocytes and, in addition, we have demonstrated the potential for IL-1α and IL-2 expression in the cells. Chronic hepatitis C is accompanied by augmented expression of cytokines such as TNF-α, IL-1α, and IL-2, comparable in groups of children and of adult patients. Nevertheless, the augmented expression remains insufficient to terminate the infection and this may be linked to the comparatively high frequency of chronic transformation of HCV infection in the two groups of patients.

Footnotes

Acknowledgements

Supported by a grant from the University of Medical Sciences, Poznań, Poland (no. 501-2-13-01).

References

Bertoletti

D'Elios

Boni

De Carli

Zignego

Durazzo

Missale

. (1997) Different cytokine profile of intrahepatic T cells in chronic hepatitis B and hepatitis C virus infection. Gastroenterology 112:193–199

Beutler

Cerami

(1989) The biology of cachectin/TNF-α primary mediator of the host response. Annu Rev Immunol 7:625–655

Dumoulin

Bach

Leifeld

El-Bakri

Fischer

Sauer-bruch

Spengler

(1997) Semiquantitative analysis of intrahepatic cytokine mRNAs in chronic hepatitis C. J Infect Dis 175:681–685

Dumoulin

Leifeld

Honecker

Sauerbruch

Spengler

(1999) Intrahepatic expression of interleukin-1β and tumor necrosis factor-α in chronic hepatitis C. J Infect Dis 180:1704–1708

Eckels

Tabatabail

Bian

T-H

Wang

Muheisen

Rice

Yoshizawa

. (1999) In vitro human Th-cell responses to a recombinant hepatitis C virus antigen: Failure in IL-2 production despite proliferation. Hum Immunol 60:187–199

Gonzalez–Amaro

Garcia–Monzon

Garcia-Buey

Moreno–Otero

Alonso

Yague

Pivel

. (1994) Induction of tumor necrosis factor alpha production by human hepatocytes in chronic viral hepatitis. J Exp Med 179:841–848

Hoffmann

Grewe

Hans-Christoph

Estler

Schulze-Specing

Decker

(1994) Regulation of tumor necrosis factor-α-mRNA synthesis and distribution of tumor necrosis factor-α-mRNA synthesizing cells in rat liver during experimental endotoxemia. J Hepatol 20:122–128

Hsu

Raine

Fanger

(1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques. J Histochem Cytochem 29:577–580

Kasprzak

Zabel

Wysocki

Seidel

Surdyk–Zasada

Filipiak

(2002) Expression of mRNA for cytokines (TNF-α and IL-1α) in human cytomegalovirus (HCMV) and hepatitis B (HBV) infections. Folia Histochem Cytochem 40:63–68

10.

Kinnman

Andersson

Hultcrantz

(2000) In situ expression of transforming growth factor-beta 1–3, latent transforming growth factor-beta binding protein and tumor necrosis factor-alpha in liver tissue from patients with chronic hepatitis C. Scand J Gastroenterol 35:1294–1300

11.

Komminoth

Werner

(1997) Target and signal amplification: approaches to increase the sensitivity of in situ hybridization. Histochem Cell Biol 108:325–333

12.

Koziel

(1999) Cytokines in viral hepatitis. Semin Liver Dis 19:157–169

13.

Kriegler

Perez

DeFay

Albert

(1988) A novel form of TNF/cachectin is a cell surface cytotoxic transmembrane protein: ramifications for complex physiology of TNF. Cell 53:45–53

14.

Larrea

Garcia

Qian

Civeira

Prieto

(1996) Tumor necrosis factor alpha gene expression and the response to interferon in chronic hepatitis C. Hepatology 23:210–217

15.

Management of hepatitis

(1997) NIH Consensus Statement 15:1–41

16.

McCaughan

Gorrell

Bishop

Abbott

Shackel

McGuinness

Levy

. (2000) Molecular pathogenesis of liver disease: an approach to hepatic inflammation, cirrhosis and liver transplant tolerance. Immunol Rev 174:172–191

17.

McGuinness

Painter

Davies

McCaughan

(2000) Increases in intrahepatic CD68 positive cells, MAC387 positive cells, and proinflammatory cytokines (particularly interleukin 18) in chronic hepatitis C infection. Gut 46:260–269

18.

Napoli

Bishop

McGuinness

Painter

McCaughan

(1996) Progressive liver injury in chronic hepatitis C infection correlates with increased intrahepatic expression of Th1-associated cytokines. Hepatology 24:759–765

19.

Neuman

Benhamou

Bourliere

Ibrahim

Malkiewicz

Asselah

Martinot–Peignoux

. (2002) Serum tumour necrosis factor-alpha and transforming growth factor-beta levels in chronic hepatitis C patients are immunomodulated by therapy. Cytokine 17:108–117

20.

Oyanagi

Takahashi

Matsui

Takahashi

Boku

Takahashi

Furukawa

. (1999) Enhanced expression of interleukin-6 in chronic hepatitis C. Liver 19:464–472

21.

Peters

(1996) Action of cytokines on the immune response and viral interactions: an overview. Hepatology 23:911–916

22.

Reichert

Watkins

Stanson

Johnson

Whiteside

(1998) Endogenous IL-2 in cancer cells: a marker of cellular proliferation. J Histochem Cytochem 46:603–612

23.

Santis

Companero

Alonso

Sanchez–Madrid

(1992) Regulation of tumor necrosis factor (TNF)-alpha synthesis and TNF receptors expression in T lymphocyte through the CD2 activation pathway. Eur J Immunol 22:3155–3160

24.

Scheuer

Standish

Dhillon

(2002) Scoring of chronic hepatitis. Clin Liver Dis 6:335–347

25.

Shapiro

Gerstein

Elias

Zuckerman

Salman

Lahat

(1998) mRNA cytokine profile in peripheral blood cells from chronic hepatitis C virus (HCV)-infected patients: effects of interferon-alpha (IFN-alpha) treatment. Clin Exp Immunol 114:55–60

26.

Simpson

Lukacs

Colletti

Strieter

Kunkel

(1997) Cytokines and the liver. J Hepatol 27:1120–1132

27.

Smith

(1984) Interleukin 2. Annu Rev Immunol 2:319–333

28.

Tilg

Wilmer

Vogel

Herold

Nolchen

Judmaier

Huber

(1992) Serum levels of cytokine in chronic liver diseases. Gastroenterology 103:264–274

29.

Tsukamoto

(1999) Cytokine regulation of hepatic stellate cells in liver fibrosis. Alcoholism. Clin Exp Res 23:911–916

30.

Winwood

Arthur

MJP

(1993) Kupffer cells: their activation and role in animal models of liver injury and human liver disease. Semin Liver Dis 13:50–59

31.

Zhu

Khoshnan

Schneider

Matsumoto

Dennert

Ware

Lai

MMC

(1998) Hepatitis C virus core protein binds to the cytoplasmic domain of tumor necrosis factor (TNF) receptor 1 and enhances TNF-induced apoptosis. J Virol 72:3691–3697