Journal of Hepatology 47 (2007) 203211 www.elsevier.

com/locate/jhep

A Toll-like receptor 7 single nucleotide polymorphism protects from advanced inammation and brosis in male patients with chronic HCV-infectionq
Eckart Schott1,, Heiko Witt1,, Konrad Neumann2, Stefan Taube3, Djin-Ye Oh4, Eckart Schreier3, Sandra Vierich1, Gero Puhl5, Alexandra Bergk1, Juliane Halangk1, Viola Weich1, Bertram Wiedenmann1, Thomas Berg1,*
1 Department of Hepatology and Gastroenterology, CVK, Charite Universitatsmedizin Berlin, Germany Department of Medical Biometry and Clinical Epidemiology, CCM, Charite Universitatsmedizin, Berlin, Germany 3 Robert Koch-Institute, Berlin, Germany 4 Institute for Microbiology and Hygiene, CCM, Charite Universitatsmedizin, Berlin, Germany 5 Department of General, Visceral, and Transplantation Surgery, CVK, Charite Universitatsmedizin, Berlin, Germany 2

See Editorial, pages 165167

Background/Aims: HCV-infection leads to development of liver brosis, causing morbidity and mortality. Multiple factors inuence the progression of brosis, including genetic factors. Since HCV is an RNA virus, a role for TLR7 in the immune response against HCV is likely. No systematic analysis of TLR7 single nucleotide polymorphisms (SNPs) has been published. Methods: We sequenced TLR7 in 52 women and investigated SNPs with an allele frequency >5% in 807 patients with chronic HCV-infection by melting curve analysis. We analyzed the eect of TLR7 SNPs on grade of inammation and stage of brosis as determined by liver biopsy. Results: We detected ve TLR7 SNPs, three of which showed a frequency >5%. One variant, c.1-120T > G, was more common in patients with no or little inammation than in patients with grades 24 (10.7% vs. 6.1%; P = 0.034). The variant was also enriched in patients with no or little brosis compared to those with higher stages (12.6% vs. 6.6%; P = 0.005). The dierence was fully attributable to male patients. Conclusions: This is the rst analysis of TLR7 SNPs in patients with chronic HCV-infection. Our data suggest that the c.1-120G TLR7 allele oers protection from the development of inammation and brosis in male patients with chronic HCV-infection. 2007 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: HCV; Fibrosis; Toll-like receptor; Genetic polymorphism; Innate immunity

Received 4 January 2007; received in revised form 5 March 2007; accepted 20 March 2007; available online 12 April 2007 q The authors declare that they do not have anything to disclose regarding conict of interest with respect to this manuscript. * Corresponding author. Tel.: +49 30 450 553071; fax: +49 30 450 553903. E-mail address: thomas.berg@charite.de (T. Berg). These authors contributed equally to this work. Abbreviations: HCV, hepatitis C virus; TLR, Toll-like receptor; SNP, single nucleotide polymorphism; HBV, hepatitis B virus; UTR, untranslated region; pDC, plasmacytoid dendritic cell.

1. Introduction HCV-infection leads to chronic liver inammation in the majority of patients. A substantial proportion of patients develop brosis or cirrhosis, causing HCVrelated morbidity and mortality. Multiple factors inuence the progression of brosis, including gender, age at infection, and alcohol consumption [1,2]. In addition,

0168-8278/$32.00 2007 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2007.03.021

204

E. Schott et al. / Journal of Hepatology 47 (2007) 203211 mation and brosis were classied according to the semiquantitative scoring system described by Scheuer [19]. The degree of inammation was graded on a scale of 04 (0: absent; 1: minimal; 2: mild; 3: moderate; 4: severe). The degree of brosis was staged on a scale of 04 (0: absent; 1: mild without septa; 2: moderate with few septa; 3: numerous septa without cirrhosis; 4: cirrhosis). Liver biopsies were performed before treatment for HCV-infection. Patient characteristics are listed in Table 1.

genetic factors inuence progression of brosis. Association of genetic polymorphisms with progression of brosis was demonstrated among others for the hemochromatosis [3], transforming growth factor-b1 [4], complement factor-5 [5], angiotensinogen [6], monocyte chemotactic protein-1 [7], and microsomal epoxide hydrolase genes [8]. Fibrosis is the result of defective repair of liver damage resulting from inammation caused by eector cells of the immune system. The cytokine interferon-a is a key mediator at the interface between innate and adaptive immunity. It is mainly produced by plasmacytoid dendritic cells (pDCs) after the engagement of Toll-like receptors (TLRs) [9]. TLR7 is a promising candidate for an immune mediator in HCVinfection since it is expressed on pDCs, binds single stranded RNA, and its activation stimulates secretion of interferon-a [1012]. Two lines of evidence support a role for TLR7 in HCV-infection: rst, a clinical study demonstrated an antiviral eect of the TLR7 agonist, isatoribine [13], which is mediated by immune cells [14]. Second, activation of TLR7 on hepatocytes exerts an antiviral eect independent of interferon [15]. Binding of the TLR7 ligand SM360320 to hepatocytes leads to expression of antiviral genes and suppression of HCV-replication. Thus, two compartments are involved in the TLR7dependent antiviral response in HCV-infection: immune cells such as pDCs [16,17], which produce interferon-a and hepatocytes, which up-regulate antiviral proteins. The TLR7 gene is located on the X-chromosome [18], spanning three exons. Single nucleotide polymorphisms (SNPs) have been described for most TLRs, but no analysis of TLR7 SNPs has been published. We determined the prevalence of TLR7 SNPs in a cohort of patients with chronic HCV-infection and analyzed the eect of these SNPs on histology as determined by liver biopsy.

2.2. Detection of TLR7 SNPs

We isolated genomic DNA from EDTAblood using spin columns (Qiagen) and analyzed TLR7 exons by bi-directional DNAsequencing. DNA was selected from 52 consecutive females from our DNA-bank regardless of diagnosis and ethnicity. DNA-fragments were amplied by PCR, generating a fragment (primers 1F/ 2R) covering the 5 0 -UTR, exons 1 and 2, and complete intron 1. Two fragments were generated covering exon 3 (primers 31F/34R, 35F/38R). Primers were designed based on published sequence (AC005859) as depicted in Table 2 and were synthesized by TIB MOLBIOL. We performed PCR (0.75 U AmpliTaq Gold (Applied Biosystems), 400 lM dNTPs, 1.5 MgCl2, 0.1 lM of each primer) in a volume of 25 ll. The reaction mix was denatured (95 C for 12 min) followed by 40 cycles of 95 C for 20 s, 58 C for 40 s, 72 C for 90 s, and a nal extension step (72 C for 2 min) in a Biometra thermocycler. We digested PCR products (shrimp alkaline phosphatase (USB), exonuclease I (USB)) and performed cycle sequencing using BigDye terminator mix (Applied Biosystems). Sequencing was performed using 10 primer pairs (Table 2) that covered the whole sequence and generated fragments of at least 400 bases. Reaction products were puried with Sephadex G-50 (Amersham) and loaded onto an ABI 3100 uorescence sequencer (Applied Biosystems). Table 1 Patient characteristics Sex Male Female Age Mean (range) Ethnicity German Turkish Other Caucasian Unknown HCV genotype 1 2 3 4 Unknown Grade of inammation 0 1 2 3 4 Stage of brosis 0 1 2 3 4 427 (52.9%) 380 (47.1%) 47 (1878) 487 (60.3%) 42 (5.2%) 49 (6.1) 229 (28.4%) 455 (56.4%) 25 (3.1%) 74 (9.2%) 8 (1.0) 245 (30.4%) 25 (3.1%) 282 (43.3%) 289 (44.4%) 52 (8.0%) 3 (0.5%) 85 (10.5%) 233 (28.9%) 183 (22.7%) 82 (10.2%) 218 (27.0%)

2. Methods 2.1. Study subjects

The study was approved by the Ethics Committee and all patients gave informed consent. Chronic HCV-infection was conrmed by persistence of anti-HCV antibodies and HCV-RNA over 6 months. All patients were negative for HBs-antigen and HIV-antibodies. We performed HCV-genotyping by reverse hybridization assay (Inno Lipa HCV II, Innogenetics). Patients from the sequencing cohort were female with a median age of 54 years (range 3275). Forty-ve patients were German, three were Turkish, and four were of other Caucasian ethnicity. Patients suered from chronic HCV-infection (n = 35), primary biliary cirrhosis (n = 6), and alcoholic or non-alcoholic fatty liver disease (n = 3 each). One patient each suered from autoimmune hepatitis, chronic HBV-infection, drug-induced liver disease, arteriovenous malformation, and cryptogenic liver disease, respectively. The test cohort consisted of 807 patients with chronic HCVinfection, data regarding histological staging and grading were available in 801 and 651 patients, respectively. Hepatic inam-

E. Schott et al. / Journal of Hepatology 47 (2007) 203211 Table 2 Oligonucleotide sequences used for PCR amplication of TLR7 Exon 1 1F 1R Exon 2 2F 2R Exon 3 31F 31R 32F 32R 33F 33R 34F 34R 35F 35R 36F 36R 37F 37R 38F 38R 5 0 -AAGCCCATATCTGGGGTGGC-3 0 5 0 -CTCAAAGTTAGGTAGACTGGAGAC-3 0 5 0 -GTCTCCAGTCTACCTAACTTTGAG-3 0 5 0 -TCAATATTGCTCTTGCCAGCTGG-3 0 5 0 -AATGTTGCAAAAGAGAGGCAGC-3 0 5 0 -TCTAGTAGCTGGTTTCCATCCAGG-3 0 5 0 -CATGTGCATCAAGAGGCTGCAG-3 0 5 0 -TTTACAGGGATCTGTAGGGGAG-3 0 5 0 -CCATTTCCTTGTGCGCGCCGTG-3 0 5 0 -CTTCACTTGAATCTCCTGAAGGTG-3 0 5 0 -AAATTGCTAACCTCAGCATG-3 0 5 0 -GCCGGTTGTTGGAGAAGTCC-3 0 5 0 -CTCAAATGCCTGAATCTGTCAGG-3 0 5 0 -GACACTGGAGTTTCTTCCAACTG-3 0 5 0 -CCTAAGTTTCTTGCCTTCTGGAG-3 0 5 0 -CAAGTCACATCTGTGGCCAGG-3 0 5 0 -CTGTGCACCTGTGATGCTGTG-3 0 5 0 -TTGTCTGTCATCACAAACACTGTC-3 0 5 0 -GGAAAGGGACTGGTTACCAGG-3 0 5 0 -ACAGCCAGGCCTCTCCTTGG-3 0

205

Apart from haplotype reconstruction, all statistics were calculated using SPSS 12.01 software.

3. Results 3.1. Detection of TLR7 polymorphisms By sequencing all 3 TLR7 exons in 52 women, we detected 5 SNPs (Table 3), three of which were present in the SNP database (http://www.ncbi.nlm.nih.gov/ SNP/snp_ref.cgi?rs=179008). Three SNPs showed an allele frequency above 5%, namely c.1-120T > G, c.32A > T, and c.2403G > A. The variant c.32A > T results in an amino acid change from glutamine to leucine at codon 11 (Q11L), while c.1-120T > G is located in intron 1 (IVS1-22), and c.2403G > A aects the third base of codon 801, representing a synonymous alteration. 3.2. Prevalence of TLR7 genotypes and haplotypes in a cohort of patients with chronic HCV-infection We analyzed the three frequent SNPs in a cohort of 807 patients with chronic HCV-infection. The distribution of genotypes in patients with HCV-infection is depicted in Table 4. Distributions of genotypes in women were in accordance with the HardyWeinberg equilibrium and frequencies of hemizygous males approximated half of the combined frequencies of homozygous and heterozygous women. Genotype distributions were similar among patients with dierent ethnicities. Haplotypes in females were reconstructed using the PHASE software. Five haplotypes covered more than 98% of individuals, namely c.1-120T c.32A c.2403G (64.5%), c.1-120T c.32A c.2403A (6.0%), c.1-120T c.32T c.2403G (21.5%), c.1-120T c.32T c.2403A (1.9%), and c.1-120G c.32A c.2403G (4.5%). Haplotype frequencies were similar in males: c.1-120T c.32A c.2403G (65.3 %), c.1-120T c.32A c.2403A (5.0%), c.1-120Tc.32Tc.2403G (20.3%), c.1-120T c.32T c.2403A (3.1%), c.1-120G c.32A c.2403G (5.4%). 3.3. TLR7 polymorphisms are associated with grade of inammation and stage of brosis The distribution of c.1-120T > G genotypes was signicantly dierent when patients were stratied according to grade of inammation as determined by liver biopsy. The variant allele was present in 20.0% of patients with no inammation, in 9.9% of patients with grade 1, in 5.9% with grade 2, in 5.8% with grade 3, and in 33.3% of patients with grade 4 (P = 0.031, Fig. 1a). Of note, only three patients showed grade 4 inammation on liver biopsy who were excluded from analysis in Fig. 1, since the number is insucient for statistical analysis. When patients were grouped as having either no or little (grades 0/1) vs. advanced (grades 24)

We analyzed SNPs c.1-120T > G, c.32A > T, and c.2403G > A in a cohort of patients with chronic HCV-infection. Genotyping was carried out by melting curve analysis with uorescence resonance energy transfer (FRET) probes in a LightCycler (Roche Diagnostics) after PCR amplication as described above with the exception that 48 cycles were performed. FRET probes were complementary to variant alleles. For detection of c.1-120T > G, we used probes 5 0 -TCTTTGAAATGTAAACTTGGATGTCTTCTCT-FL as sensor and 5 0 -LC705-TCTCTTAGTTGATGCTATTGGGCCCATCTC AAG-ph as anchor. For identication of c.32A > T, the sensor probe was 5 0 -AAAAAGGATAAGAATTAGTCTCTTCAGTG-FL and the anchor probe was 5 0 -LC640-CCACATTGGAAACACCTG AAAGTATAAGAAAGAAC-ph, and for c.2403G > A, we used 5 0 AGGAATAGTCACCTCTGTATGGTTAACCC-FL as sensor and 5 0 -LC 705-CCAGACAAACCACACAGCATCACAGGTGC-ph as anchor probes. FRET probes were designed and synthesized by TIB MOLBIOL.

2.3. Detection of IL-6

PBMCs from healthy donors characterized for TLR7 variants were isolated by Ficoll gradient centrifugation. Cells were cultured in 96well plates at 3 105 cells/well and stimulated with imiquimod (Invivogen, 15 lM) or lipopolysaccharide (LPS Re 595, Sigma, 50 ng/ml) for the time indicated. IL-6 was measured by ELISA, matched antibodies MQ2-13A5 and MQ2-39C3 were from Pharmingen, IL-6 standard from R&D Systems.

2.4. Haplotype reconstruction and statistical analysis

We performed haplotype reconstruction using the PHASE v2.02 software (www.stat.washington.edu/stephens/phase.html) [20]. Permutation tests with 10,000 permutations were used to detect dierences in the distribution of haplotypes in two groups. We performed statistical analysis by contingency tables using v2 statistics and Fishers two-sided exact test where applicable. Logistic regression analysis was performed to determine the inuence of age, sex, and c.1-120T > G on the degree of brosis. A variable was included in the regression model for age, sex, heterozygosity, or homo-/hemizygosity.

206

E. Schott et al. / Journal of Hepatology 47 (2007) 203211

Table 3 TLR7 single nucleotide polymorphisms detected by direct DNA-sequencing Variation c.1-346T > A c.1-120T > G c.32A > T c.1035G > A c.2403G > A NA, not applicable. Accession number NA NA rs 179008 rs 5743780 rs 864058 Location IVS1+110 IVS1-22 Codon 11 (exon 3) Codon 345 (exon 3) Codon 801 (exon 3) Amino acid exchange NA NA Q11L No No Frequency 1/104 6/104 33/104 2/104 8/104

inammation, the variant G allele was present in 10.7% vs. 6.1%, respectively (P = 0.034). The dierence was fully attributable to males (9.8% vs. 3.8%, P = 0.044, Table 5), while the dierence in females was statistically insignicant (11.7% vs. 8.7%, P = 0.457). No association of c.32A > T and c.2403G > A variations with grade of inammation was noted. The distribution of c.1-120T > G genotypes also differed signicantly when patients were stratied according to stage of brosis. The variant allele was present in 18.8% of patients with no brosis, in 10.3% of patients with stage 1, in 3.8% with stage 2, in 7.3 with stage 3, and in 8.7% of patients with stage 4 (P = 0.002, Fig. 1b). Patients with no or little brosis (stages 0/1) were more likely to carry the c.1-120G allele than those with higher stages of brosis. The variant G allele was present in 12.6% and 6.6%, respectively (P = 0.005). Again, the dierence was only signicant in men (11.7% vs. 3.6%, P = 0.003, Table 5) but not in women (13.3% vs. 10.8%, P = 0.524). No association was observed for the c.32A > T and c.2403G > A alterations. Next, we analyzed the distribution of haplotypes and their inuence on inammation and brosis (Table 5). The distribution of haplotypes was dierent between patients with no or little (stages 0/1) and

higher degrees (stages 24) of brosis, mainly due to the overrepresentation of the c.1-120Gc.32Ac.2403G haplotype in the former group. In turn, the c.1120Tc.32Ac.2403G haplotype was more frequent in patients with higher degrees of brosis. The dierence was solely attributable to men and no dierence was observed in women. The same trend was seen regarding the distribution of haplotypes according to grade of inammation, however, the dierence did not reach statistical signicance. The duration of infection is a determinant of the development of brosis. Since the date of infection is rarely known, the age of a patient may serve as a surrogate marker for the duration of infection. We investigated whether the protective eect of c.1-120G is dependent on the age of the patients. Age was a strong predictor of brosis (data not shown). The eect of the c.1-120G allele remained signicant in male patients below and above 50 years of age (Table 6). The c.1120T > G alteration was an independent predictor of brosis by multivariate analysis including age and sex as additional variables, arguing against an age-related bias as the explanation for the observed dierences. In the stepwise logistic regression analysis, age above 50 years was predictive of higher degrees of brosis, while female sex and the presence of the c.1-120T > G alter-

Table 4 Distribution of TLR7 genotypes according to gender and ethnicity in patients with chronic HCV-infection TLR7 genotype c.1-120T > G TT TG GG c.32A > T AA AT TT c.2403G > A GG GA AA
a b

All patientsa n = 807 (%) 91.1 5.5 3.5 66.9 18.1 15.0 87.4 7.4 5.2

Male patients n = 427 (%) 93.7 6.3 76.3 23.7 91.1 8.9

Female patients n = 380 (%) 88.2 11.6 0.3 56.3 38.4 5.3 83.2 15.8 1.1

Germana n = 487 (%) 93.0 4.7 2.3 66.9 19.1 14.0 88.9 6.8 4.3

Turkisha n = 42 (%) 85.7 7.1 7.1 73.8 11.9 14.3 92.9 4.8 2.4

Othera Caucasian n = 49 (%) 91.8 4.1 4.1 63.3 14.3 22.4 89.8 8.2 2.0

Unknowna n = 229 (%) 87.8 7.0 5.2 66.4 17.9 15.7 82.5 9.2 8.5

P valueb

0.218

0.647

0.167

Hemizygous men were grouped as if they were homozygous for clarity. Analysis for dierences between patients of dierent ethnicities was carried out using the v2 test.

E. Schott et al. / Journal of Hepatology 47 (2007) 203211

207

Fig. 1. The percentage of patients who carry at least one variant c.1120G allele is depicted for the group as a whole as well as for male and female patients separately according to grade of inammation (a) or stage of brosis (b). Grade of inammation and stage of brosis were analyzed by liver biopsy. The number of individuals that are contained in each group is indicated below the respective bar. Signicance levels (v2 test over all grades/stages) are depicted.

ation were predictive of lower degrees of brosis. HCV genotypes were not predictive of the degree of brosis (data not shown). 3.4. PBMCs from individuals harboring the c.1-120 variation show increased IL-6 production in response to TLR7 ligation To test the functional relevance of the c.1-120T > G alteration, we incubated PBMCs from healthy donors with TLR7 ligand imiquimod or TLR4 ligand LPS. After stimulation of TLR7 with imiquimod, we observed increased release of IL-6 from PBMCs from individuals carrying the variant allele compared to PBMCs from wildtype individuals, while no dierence was observed after TLR4 ligation (Fig. 2).

4. Discussion This is the rst report on TLR7 SNPs in patients with chronic HCV-infection. SNPs of other TLRs inuence

the ecacy of immune responses in settings of bacterial, viral, and parasite infections. TLR2 binds lipoproteins and polymorphisms have been implicated in Borrelia burgdorferi infection [21] as well as in tuberculosis and leprosy [22]. TLR4 binds lipopolysaccharide and polymorphisms inuence the course of disease in meningococcal infection [23], candidiasis [24], gram-negative bacteremia [2527], respiratory syncytial virus infection [28], and others. Polymorphisms of the gene of agellin-binding TLR5 are associated with susceptibility to Legionnaires disease [29]. TLR9 binds bacterial and viral CpG-DNA and polymorphisms have been implicated in autoimmune and allergic disorders [3034]. Although TLR9 stimulation shows antiviral eects in HCV-infection [35], binding to TLR9 has only been demonstrated for DNA viruses [36] and binding of HCV to TLR9 is unlikely. No TLR7 polymorphisms have been described to date that inuence the course of a human disease. Given the binding of single stranded RNA to TLR7, an implication in HCV-infection is likely. We sequenced 104 alleles of women with chronic liver diseases of viral and non-viral origin to detect SNPs relevant in our study population and to identify SNPs that are prevalent enough to be studied in a larger cohort. Women were chosen for sequencing since they provide twice the number of X-chromosomal alleles as men. The frequencies of hemizygous males approximated half of the combined frequencies of homozygous and heterozygous females, demonstrating equal distribution of mutant alleles in men and women. We identied ve SNPs, three of them with an allele frequency above 5%, which was the cut-o value for further analysis. c.1-120T > G is located within a putative branching site in intron 1, implicating that it might alter splicing patterns. The branch point is functionally important in splicing, located 1050 nucleotides upstream from the splice donor site, and characterized by the consensus sequence 5 0 -YTRAY-3 0 (R = purine, Y = pyrimidine) [37]. The c.1-120T > G transversion aects the highly conserved T at the second position of this consensus. The complete functional consequence of the c.1-120T > G variation remains to be determined but our pilot experiments suggest altered patterns of cytokine secretion after TLR7 ligation in individuals harboring the variation. We chose IL-6 as the readout since IL-6 is a well-characterized eector cytokine released after TLR7 ligation [11], is essential for liver regeneration [38,39], and has potential antibrotic eects [40]. The c.32A > T transversion results in an amino acid exchange (Q11L), whereas the c.2403G > A transition represents a synonymous alteration. In HCV-infection, an ecient T-cell response is necessary to clear the virus. The quality of antigen presentation likely determines the number of epitopes recognized by T-cells, which is a predictor of the outcome of the immune response against HCV [4143]. If

208

E. Schott et al. / Journal of Hepatology 47 (2007) 203211

Table 5 Distribution of TLR7 genotypes and haplotypes according to grade of inammation, stage of brosis, and gender Grade of inammation Male patients 01 Genotype c.1-120T > G TT TG GG G presentb c.32A > T AA AT TT T presentb c.2403G > A GG GA AA A presentb Haplotypef c.1-120T c.32A c.2403G c.1-120T c.32A c.2403A c.1-120T c.32T c.2403G c.1-120T c.32T c.2403A c.1-120G c.32A c.2403G n = 153 90.2 9.8 9.8 79.1 20.9 20.9 88.9 11.1 11.1 n = 150 63.3 8.0 18.7 2.0 8.0
a

Stage of brosis Female patients 01 24 n = 161 91.3 8.1 0.6 8.7 57.1 37.9 5.0 42.9 83.2 16.1 0.6 16.8 n = 160 66.8 6.1 22.0 1.9 3.2 P Male patientsa 01 n = 145 88.3 0.354c 0.457d 11.7 11.7 76.6 0.937 0.733 23.4 23.4 87.6 0.558 1.000 12.4 12.4 n = 142 59.2 7.7 19.7 3.5 9.9 24 n = 279 86.4 3.6 3.6 76.3 23.7 23.7 93.2 6.8 6.8 n = 278 69.4 3.6 20.9 2.9 3.2 P Female patients 01 n = 173 86.7 13.3 0.0 13.3 57.2 38.2 4.6 42.8 83.2 16.2 0.6 16.8 n = 170 68.0 5.1 21.2 2.0 3.7 24 n = 204 89.2 10.3 0.5 10.8 56.4 37.7 5.9 43.6 83.3 15.7 1.0 16.7 n = 201 66.3 6.1 22.3 2.2 3.1 P

24 n = 183 96.2 3.8 3.8 73.8 26.2 26.2 91.3 8.7 8.7 n = 182 67.0 3.3 21.4 4.9 3.3 P

n = 154 88.3 11.7 0.0 11.7 59.1 36.4 4.5 40.9 83.1 14.9 1.9 16.9 n = 151 69.7 5.8 19.8 2.2 2.4

0.044d

0.003

0.441 0.524

0.304

1.000

0.863 0.917

0.469

0.069

0.902 1.000

0.057c

0.918e

0.014

0.894

All values except P values are in percent. a For genotype analysis, hemizygous men were grouped as if they were homozygous for clarity. b Combines all patients who carry at least one variant allele. c Analysis was carried out using the v2 test. d Analysis was carried out using Fishers exact test. e It was assumed that both X-chromosomes contribute equally to the individuals genotype. Analysis was carried out using the PHASE software. f Haplotype analysis was restricted to the ve frequent haplotypes that covered more than 98% of patients.

viral clearance is not achieved, the immune response determines the degree of damage implied on the liver in chronic infection since the virus itself is only weakly cytopathic. Liver damage is rather caused by the, albeit inefcient, immune response against infected hepatocytes. We found an association of c.1-120T > G with histological parameters as determined by liver biopsy. The c.1-120G allele was associated with a more favorable histological diagnosis, both regarding grade of inammation and stage of brosis. Although the mechanism by which this variant causes variability in the response to TLR7 activation is incompletely understood, a role for an endogenous antiviral immune mediator such as TLR7 for progression of inammation seems plausible. Altered responsiveness to viral stimuli on hepatocytes or immune cells that inltrate the liver could reduce the amount of damage caused by an antiviral immune response. In consequence, progression of brosis would also be slowed down. Therefore, TLR7 activation in the course of HCV-infection is a double-edged sword: activation likely is warranted in the setting of acute infection to clear the virus but is responsible for liver scarring in the setting of chronic disease.

Gender is one of several factors that inuence progression of brosis in HCV-infected patients. Female sex is associated with a decreased risk to develop brosis [1,2], potentially explaining why a stronger eect of c.1120T > G was observed in men: a protective variation should exert its eect most profoundly in individuals at higher risk. The fact that TLR7 is located on the X-chromosome makes it prudent to analyze men and women separately. The presence of two copies of a X-chromosomal gene may inuence its biological eect as well as the consequence of a genetic variation: variant alleles are twice as frequent in women but rarely occur as a homozygous trait, whereas men who carry a TLR7 alteration are hemizygous. Thus, a recessive trait is more prominent in men than in women, while a dominant trait is more frequent in women. Also, it must be considered that skewing of X-chromosomal activation may aect the penetrance of a mutation [44]. Since homozygous women are extremely rare, it is obvious to nd the most reliable results in men. Gender eects are also apparent in clinical settings of inammation: men are more prone to infections [45] while women are more likely to suer from autoimmune disease [46], ndings mainly

E. Schott et al. / Journal of Hepatology 47 (2007) 203211 Table 6 Inuence of age on the eect of the c.1-120T > G alteration on stage of brosis in patients with chronic HCV-infection Presence of the variant G allele Stage 01 n/n Univariate analysis All patients Age 6 50 years Age > 50 years Male patients Age 6 50 years Age > 50 years
a

209

P value Stage 24

n/n

30/241 10/77 13/116 4/29

12.4 13.0 11.2 13.8 ORc

15/215 17/268 5/138 5/141

7.0 6.3 3.6 3.5

0.059 0.088 0.026 0.047

P value

Multivariate analysis All patients Age >50 Heterozygous variation Homozygous variationd Female sex Male patients Age >50 Hemizygous variation
a b

4.187 0.791 0.315 0.501

(3.0355.776) (0.4051.545) (0.1370.726) (0.3640.692)

<0.001 0.493 0.007 <0.001 <0.001 0.002

4.126 (2.5636.643) 0.270 (0.1150.636)

Analysis was carried out using Fishers exact test. Analysis was carried out by logistic regression, introducing a variable for age above 50 years, female sex, heterozygosity for the c.1-120T > G alteration, and homo-/hemizygosity for the c.1-120T > G alteration. c OR Odds ratio, numbers in brackets represent the 95% condence interval The OR represents the relative risk compared to a male individual of 50 years or younger with a wildtype c.1-120 locus. No adjustment for additional variables was performed. d For clarity, homozygous women and hemizygous men were combined.

attributed to the action of sex hormones [47,48]. Innate immune responses linked to TLR4 also dier between men and women, rendering men more susceptible to sepsis or endotoxin shock [49,50]. A recent study has observed gender dierences in interferon-a production by pDCs in response to TLR7 stimulation unrelated to hormonal eects, strengthening the notion that the regulation of innate immune responses is gender-dependent

Fig. 2. PBMCs from healthy donors characterized for TLR7 SNPs were incubated with TLR7 ligand imiquimod (a, 15 lM) or TLR4 ligand lipopolysaccharide (b, 50 ng/ml), for the time indicated. All donors had wildtype genotypes at positions c.32 and c.2403 and diered only by the c.1-120 genotype (lled symbols: variant G allele, open symbols: wildtype T allele). Supernatants were analyzed for IL-6 content by ELISA. Data are depicted as means SEM from individual patients. Each data point represents data from two separate measurements, each of them performed in duplicate.

[51]. It would be interesting to test whether these results hold true when study individuals are stratied for TLR7 SNPs, which may inuence the response to TLR7 stimulation in a gender-dependent fashion. Since the duration of HCV-infection is a major predictor of brosis, it is essential to control for this parameter when other factors are investigated that may inuence the development of brosis. Unfortunately, data regarding the duration of infection are scarce given the fact that in many patients the mode of transmission is unknown. Thus, the patients age is frequently used as a surrogate marker. We nd an association of the c.1120G > T variant with brosis in patients below and above 50 years of age by univariate and multivariate analysis, arguing against an age-related bias in our cohort as the explanation for dierent degrees of brosis between the groups. In summary, we present the rst comprehensive analysis of TLR7 SNPs in patients with chronic HCV-infection. Our data suggest that c.1-120T > G is protective regarding the development of inammation and brosis. We propose that drugs that modify innate immune responses mediated by TLR7 may not only be useful for the treatment of chronic HCV-infections but may also inuence the degree of inammation and brosis, making TLR7 a promising target for pharmacological approaches in chronic HCV-infection.

210

E. Schott et al. / Journal of Hepatology 47 (2007) 203211 synergizes with CD40 ligand to induce high amounts of IL-12. Eur J Immunol 2001;31:30263037. Lee J, Chuang TH, Redecke V, She L, Pitha PM, Carson DA, et al. Molecular basis for the immunostimulatory activity of guanine nucleoside analogs: activation of Toll-like receptor 7. Proc Natl Acad Sci USA 2003;100:66466651. Horsmans Y, Berg T, Desager JP, Mueller T, Schott E, Fletcher SP, et al. Isatoribine, an agonist of TLR7, reduces plasma virus concentration in chronic hepatitis C infection. Hepatology 2005;42:724731. Lee J, Wu CC, Lee KJ, Chuang TH, Katakura K, Liu YT, et al. Activation of anti-hepatitis C virus responses via Toll-like receptor 7. Proc Natl Acad Sci USA 2006;103:18281833. Chaperot L, Blum A, Manches O, Lui G, Angel J, Molens JP, et al. Virus or TLR agonists induce TRAIL-mediated cytotoxic activity of plasmacytoid dendritic cells. J Immunol 2006;176: 248255. Gibson SJ, Lindh JM, Riter TR, Gleason RM, Rogers LM, Fuller AE, et al. Plasmacytoid dendritic cells produce cytokines and mature in response to the TLR7 agonists, imiquimod and resiquimod. Cell Immunol 2002;218:7486. Du X, Poltorak A, Wei Y, Beutler B. Three novel mammalian toll-like receptors: gene structure, expression, and evolution. Eur Cytokine Netw 2000;11:362371. Scheuer PJ. Classication of chronic viral hepatitis: a need for reassessment. J Hepatol 1991;13:372374. Stephens M, Donnelly P. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet 2003;73:11621169. Schroder NW, Diterich I, Zinke A, Eckert J, Draing C, von Baehr V, et al. Heterozygous Arg753Gln polymorphism of human TLR-2 impairs immune activation by Borrelia burgdorferi and protects from late stage Lyme disease. J Immunol 2005;175:25342540. Texereau J, Chiche JD, Taylor W, Choukroun G, Comba B, Mira JP. The importance of Toll-like receptor 2 polymorphisms in severe infections. Clin Infect Dis 2005;41:S408S415. Smirnova I, Mann N, Dols A, Derkx HH, Hibberd ML, Levin M, et al. Assay of locus-specic genetic load implicates rare Toll-like receptor 4 mutations in meningococcal susceptibility. Proc Natl Acad Sci USA 2003;100:60756080. Van der Graaf CA, Netea MG, Morre SA, Den Heijer M, Verweij PE, Van der Meer JW, et al. Toll-like receptor 4 Asp299Gly/ Thr399Ile polymorphisms are a risk factor for Candida bloodstream infection. Eur Cytokine Netw 2006;17:2934. Agnese DM, Calvano JE, Hahm SJ, Coyle SM, Corbett SA, Calvano SE, et al. Human toll-like receptor 4 mutations but not CD14 polymorphisms are associated with an increased risk of gram-negative infections. J Infect Dis 2002;186:15221525. Barber RC, Aragaki CC, Rivera-Chavez FA, Purdue GF, Hunt JL, Horton JW. TLR4 and TNF-alpha polymorphisms are associated with an increased risk for severe sepsis following burn injury. J Med Genet 2004;41:808813. Montes AH, Asensi V, Alvarez V, Valle E, Ocana MG, Meana A, et al. The Toll-like receptor 4 (Asp299Gly) polymorphism is a risk factor for Gram-negative and haematogenous osteomyelitis. Clin Exp Immunol 2006;143:404413. Tal G, Mandelberg A, Dalal I, Cesar K, Somekh E, Tal A, et al. Association between common Toll-like receptor 4 mutations and severe respiratory syncytial virus disease. J Infect Dis 2004;189:20572063. Hawn TR, Verbon A, Lettinga KD, Zhao LP, Li SS, Laws RJ, et al. A common dominant TLR5 stop codon polymorphism abolishes agellin signaling and is associated with susceptibility to Legionnaires disease. J Exp Med 2003;198:15631572. Lammers KM, Ouburg S, Morre SA, Crusius JB, Gionchett P, Rizzello F, et al. Combined carriership of TLR9-1237C and

Acknowledgements The authors thank Sascha Gille (TIB MOLBIOL, Berlin, Germany) for design of the FRET probes. This work was partially supported by an investigational grant from Anadys Pharmaceuticals, San Diego, CA, by the German Competence Network for Viral Hepatitis (Hep-Net) funded by the German Ministry of Education and Research (BMBF, 01 KI 0437, 01 KI 0402), by the EU-Vigilance network of excellence combating viral resistance (VIRGIL, LSHM-CT-2004-503359), by a Rahel-Hirsch award to DYO, and by the Sonnenfeld-Stiftung, Berlin, Germany (to HW). References
[1] Cholet F, Nousbaum JB, Richecoeur M, Oger E, Cauvin JM, Lagarde N, et al. Factors associated with liver steatosis and brosis in chronic hepatitis C patients. Gastroenterol Clin Biol 2004;28:272278. [2] Poynard T, Ratziu V, Charlotte F, Goodman Z, McHutchison J, Albrecht J. Rates and risk factors of liver brosis progression in patients with chronic hepatitis c. J Hepatol 2001;34:730739. [3] Bonkovsky HL, Troy N, McNeal K, Banner BF, Sharma A, Obando J, et al. Iron and HFE or TfR1 mutations as comorbid factors for development and progression of chronic hepatitis C. J Hepatol 2002;37:848854. [4] Wang H, Mengsteab S, Tag CG, Gao CF, Hellerbrand C, Lammert F, et al. Transforming growth factor-beta1 gene polymorphisms are associated with progression of liver brosis in Caucasians with chronic hepatitis C infection. World J Gastroenterol 2005;11:19291936. [5] Hillebrandt S, Wasmuth HE, Weiskirchen R, Hellerbrand C, Keppeler H, Werth A, et al. Complement factor 5 is a quantitative trait gene that modies liver brogenesis in mice and humans. Nat Genet 2005;37:835843. [6] Powell EE, Edwards-Smith CJ, Hay JL, Clouston AD, Crawford DH, Shorthouse C, et al. Host genetic factors inuence disease progression in chronic hepatitis C. Hepatology 2000;31:828833. [7] Muhlbauer M, Bosserho AK, Hartmann A, Thasler WE, Weiss TS, Herfarth H, et al. A novel MCP-1 gene polymorphism is associated with hepatic MCP-1 expression and severity of HCVrelated liver disease. Gastroenterology 2003;125:10851093. [8] Sonzogni L, Silvestri L, De Silvestri A, Gritti C, Foti L, Zavaglia C, et al. Polymorphisms of microsomal epoxide hydrolase gene and severity of HCV-related liver disease. Hepatology 2002;36:195201. [9] Cella M, Jarrossay D, Facchetti F, Alebardi O, Nakajima H, Lanzavecchia A, et al. Plasmacytoid monocytes migrate to inamed lymph nodes and produce large amounts of type I interferon. Nat Med 1999;5:919923. [10] Diebold SS, Kaisho T, Hemmi H, Akira S, Reis e Sousa C. Innate antiviral responses by means of TLR7-mediated recognition of single-stranded RNA. Science 2004;303:15291531. [11] Heil F, Hemmi H, Hochrein H, Ampenberger F, Kirschning C, Akira S, et al. Species-specic recognition of singlestranded RNA via toll-like receptor 7 and 8. Science 2004;303:15261529. [12] Krug A, Towarowski A, Britsch S, Rothenfusser S, Hornung V, Bals R, et al. Toll-like receptor expression reveals CpG DNA as a unique microbial stimulus for plasmacytoid dendritic cells which [13]

[14]

[15]

[16]

[17]

[18]

[19] [20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

E. Schott et al. / Journal of Hepatology 47 (2007) 203211 CD14-260T alleles enhances the risk of developing chronic relapsing pouchitis. World J Gastroenterol 2005;11:73237329. Lazarus R, Klimecki WT, Raby BA, Vercelli D, Palmer LJ, Kwiatkowski DJ, et al. Single-nucleotide polymorphisms in the Toll-like receptor 9 gene (TLR9): frequencies, pairwise linkage disequilibrium, and haplotypes in three U.S. ethnic groups and exploratory case-control disease association studies. Genomics 2003;81:8591. Kikuchi K, Lian ZX, Kimura Y, Selmi C, Yang GX, Gordon SC, et al. Genetic polymorphisms of toll-like receptor 9 inuence the immune response to CpG and contribute to hyper-IgM in primary biliary cirrhosis. J Autoimmun 2005;24:347352. Torok HP, Glas J, Tonenchi L, Bruennler G, Folwaczny M, Folwaczny C. Crohns disease is associated with a tolllike receptor-9 polymorphism. Gastroenterology 2004;127: 365366. Berghofer B, Frommer T, Konig IR, Ziegler A, Chakraborty T, Bein G, et al. Common human Toll-like receptor 9 polymorphisms and haplotypes: association with atopy and functional relevance. Clin Exp Allergy 2005;35:11471154. Jacobson IM, Ghalib R, Lawitz E, Freilich B, Gordon SC, Kwo P, et al. Early viral response and on treatment response to CPG 10101 (Actilon TM), in combination with pegylated interferon and/or ribavirin, in chronic HCV genotype 1 infected patients with prior relapse response (abstract). Hepatology 2006;44:224A. Akira S, Uematsu S, Takeuchi O. Pathogen recognition and innate immunity. Cell 2006;124:783801. Harris NL, Senapathy P. Distribution and consensus of branch point signals in eukaryotic genes: a computerized statistical analysis. Nucleic Acids Res 1990;18:30153019. Galun E, Zeira E, Pappo O, Peters M, Rose-John S. Liver regeneration induced by a designer human IL-6/sIL-6R fusion protein reverses severe hepatocellular injury. FASEB J 2000;14:19791987. Zimmers TA, Pierce RH, McKillop IH, Koniaris LG. Resolving the role of IL-6 in liver regeneration. Hepatology 2003;38:15901591, author reply 1591. Bansal MB, Kovalovich K, Gupta R, Li W, Agarwal A, Radbill B, et al. Interleukin-6 protects hepatocytes from CCl4-mediated

211

[31]

[41]

[42]

[32]

[43]

[33]

[44]

[34]

[45]

[46] [47]

[35]

[36] [37]

[48]

[49]

[38]

[39]

[50] [51]

[40]

necrosis and apoptosis in mice by reducing MMP-2 expression. J Hepatol 2005;42:548556. Cooper S, Erickson AL, Adams EJ, Kansopon J, Weiner AJ, Chien DY, et al. Analysis of a successful immune response against hepatitis C virus. Immunity 1999;10:439449. Day CL, Lauer GM, Robbins GK, McGovern B, Wurcel AG, Gandhi RT, et al. Broad specicity of virus-specic CD4+ Thelper-cell responses in resolved hepatitis C virus infection. J Virol 2002;76:1258412595. Lauer GM, Ouchi K, Chung RT, Nguyen TN, Day CL, Purkis DR, et al. Comprehensive analysis of CD8(+)-T-cell responses against hepatitis C virus reveals multiple unpredicted specicities. J Virol 2002;76:61046113. Sharp A, Robinson D, Jacobs P. Age- and tissue-specic variation of X chromosome inactivation ratios in normal women. Hum Genet 2000;107:343349. Oner PJ, Moore EE, Bi WL. Male gender is a risk factor for major infections after surgery. Arch Surg 1999;134:935938, discussion 938940. Kast RE. Predominance of autoimmune and rheumatic diseases in females. J Rheumatol 1977;4:288292. Molloy EJ, ONeill AJ, Grantham JJ, Sheridan-Pereira M, Fitzpatrick JM, Webb DW, et al. Sex-specic alterations in neutrophil apoptosis: the role of estradiol and progesterone. Blood 2003;102:26532659. Posma E, Moes H, Heineman MJ, Faas MM. The eect of testosterone on cytokine production in the specic and nonspecic immune response. Am J Reprod Immunol 2004;52:237243. Marriott I, Bost KL, Huet-Hudson YM. Sexual dimorphism in expression of receptors for bacterial lipopolysaccharides in murine macrophages: a possible mechanism for gender-based dierences in endotoxic shock susceptibility. J Reprod Immunol 2006;71:1227. Imahara SD, Jelacic S, Junker CE, OKeefe GE. The inuence of gender on human innate immunity. Surgery 2005;138:275282. Berghofer B, Frommer T, Haley G, Fink L, Bein G, Hackstein H. TLR7 ligands induce higher IFN-alpha production in females. J Immunol 2006;177:20882096.