and Immunity (2006) 7, 615–624 & 2006 Nature Publishing Group All rights reserved 1466-4879/06 $30.00 www.nature.com/gene

ORIGINAL ARTICLE polymorphisms in Toll-like receptors, interleukin-10, and interleukin-10 alpha and lymphoma risk

A Nieters, L Beckmann, E Deeg and N Becker Division of Clinical Epidemiology, German Cancer Research Center, Heidelberg, Germany

Interactions between environment and immune system play an essential role in the aetiology of immunopathologies, including lymphomas. Toll-like receptors (TLR) belong to a group of pattern recognition receptors, with importance for innate immune response and inflammatory processes. Interleukin-10 (IL-10) is a key regulatory cytokine and has been implicated in lymphomagenesis. Functional polymorphisms in these inflammation-associated genes may affect the susceptibility towards lymphoma. To test this hypothesis, we have genotyped DNA of 710 lymphoma cases and 710 controls within the context of a population-based epidemiological study for 11 functionally important single-nucleotide polymorphisms in TLR1, À2, À4, À5, À9, IL10 and IL10 receptor (IL10RA). The IL10RA Ser138Gly variant was underrepresented among lymphoma cases (odds ratio (OR) ¼ 0.81, 95 per cent confidence interval (95% CI) ¼ 0.65–1.02), mainly owing to an inverse association with Hodgkin’s lymphoma (HL). The TLR2 À16933T4A variant was associated with a 2.8-fold increased risk of follicular lymphoma (95% CI ¼ 1.43–5.59) and a decreased risk of chronic lymphocytic leukaemia (OR ¼ 0.61, 95% CI ¼ 0.38–0.95). Furthermore, the TLR4 Asp299Gly variant was positively associated with the risk of mucosa-associated lymphoid tissue lymphoma (OR ¼ 2.76, 95% CI ¼ 1.12–6.81) and HL (OR ¼ 1.80, 95% CI ¼ 0.99–3.26). In conclusion, this study suggests an effect of polymorphisms in factors of the innate immune response in the aetiology of some lymphoma subtypes. Genes and Immunity (2006) 7, 615–624. doi:10.1038/sj.gene.6364337; published online 14 September 2006

Keywords: TLR; IL10; IL10RA; lymphoma; epidemiology; SNP

Introduction cells associated with innate immunity. In humans, 10 different TLRs have been identified.9,10 They are im- The incidence of lymphomas has substantially increased portant pattern recognition receptors of microorganisms. over the past decades.1 The aetiology of lymphoma and For several TLRs, the binding of different ligands has the underlying causes for the observed increase are been demonstrated, allowing the immune system to largely unknown; however, genetic and environmental identify conserved structures as danger signals and factors involved in immune regulation and inflammation initiate an adequate immune response. Some TLR gene are suspected candidates.2 This is underlined by a large variants with potential functional relevance have been body of evidence such as (1) the increased risk of associated with various phenotypes including suscept- lymphoid malignancies among individuals with inher- ibility to infectious and inflammatory diseases as well as ited or acquired immunodeficiencies,3 (2) the association cancer.11–23 First evidence links a polymorphism in TLR4 of some lymphoma subtypes with infectious agents,4 (3) (Arg299Gly) with the susceptibility to gastric mucosa- the correlation with some indicators of the hygiene associated lymphoid tissue (MALT) lymphoma.16 hypothesis5–7 as well as (4) the repeatedly observed Signalling through TLRs is involved not only in the inverse association between allergies and and primary induction of inflammation, however, also anti- lymphoma risk.8,6 inflammatory mediators necessary for the resolution of The family of Toll-like receptors (TLRs) are important inflammatory processes are activated.24 Interleukin-10 factors mediating the interaction between infectious (IL10) is multifunctional cytokine crucial for the regulation agents and host immune response and bridging innate of immune homeostasis25 and is elevated in lymphomas.26 and acquired immunity. TLRs are expressed mainly Some studies have suggested a role of genetic variants on antigen-presenting cells, including dendritic cells, in IL10 in lymphoma aetiology and prognosis.27–30 / and B cells, as well as other We hypothesized that these factors at the interphase between environment and immune system and impor- tant players in inflammatory processes are implicated in Correspondence: Dr A Nieters, Division of Clinical Epidemiology, lymphomagenesis. Specifically, within our population- German Cancer Research Center, Im Neuenheimer Feld 280, based case–control study on lymphoid neoplasms invol- D-69120 Heidelberg, Germany. 5,8 E-mail: [email protected] ving 710 case–control pairs we have explored the Received 24 May 2006; revised 5 July 2006; accepted 24 July 2006; contribution of putative functional polymorphisms in published online 14 September 2006 TLR1, TLR2, TLR4, TLR5, TLR9, IL10 and interleukin-10 Variants in TLRs, IL10, IL10RA and lymphoma risk A Nieters et al 616 receptor alpha (IL10RA) to inter-individual differences in the TLR2 À16933T4A polymorphism showed a strong the susceptibility to lymphoma and lymphoma subtypes increased risk of FL among individuals who carried the (Table 2). A-allele (odds ratios (OR) ¼ 2.82, 95 per cent confidence interval (95% CI) ¼ 1.43–5.59). OR for the AT- and AA- genotypes are 2.71 (95% CI ¼ 1.33–5.50) and 2.95 (95% Results CI ¼ 1.38–6.30), respectively. The A-allele conferred a decreased risk estimate for CLL (OR ¼ 0.61, 95% Detailed demographic characteristics of the study popu- CI ¼ 0.38–0.95). The respective OR for the AT- and AA- 5,8 lation have previously been described. Therefore, only genotypes were 0.64 (95% CI ¼ 0.39–1.05) and 0.56 (95% a brief description of the age and sex distribution and the CI ¼ 0.31–1.01). availability of blood samples is summarized in Table 1. The TLR4 299G-allele was more prevalent in cases of MALT lymphoma (13.0%), T-cell non-Hodgkin’s lym- Association of gene variants with lymphoma and specific phoma (NHL) (9.1%) and HL (9.0%) compared to lymphoma subtypes controls (5.3%). The respective OR are 2.76 (95% The genotype distribution for the TLR1 À7202A4G, CI ¼ 1.12–6.81), 1.90 (95% CI ¼ 0.75–4.78) and 1.80 (0.99– TLR2 Arg753Gln G4A, TLR2 À16933T4A, TLR4 3.26). The stop-codon variant in TLR5 and the IL10 Asp299Gly A4G, TLR5 Arg392Stop C4T, TLR9 polymorphisms did not modify the risk of lymphoma À1486T4C, TLR9 À1237T4C, IL10 À3575T4A, IL10 nor any of the subtypes. A strong protective effect for HL À1082A4G, IL10RA Ser138Gly A4G and IL10RA was seen for the IL10RA Ser138Gly G-allele (OR ¼ 0.51, Gly330Arg G4A polymorphisms in all lymphoma cases 95% CI ¼ 0.31–0.84). and by lymphoma subtypes (diffuse large B-cell lym- phoma (DLBCL), follicular lymphoma (FL), chronic Haplotype associations lymphocytic leukaemia (CLL), multiple myeloma The two common haplotypes in TLR2 (TLR2 (MM), MALT lymphoma, T-cell lymphoma and Hodg- À16933T4A-TLR2 Arg753Gln) were found significantly kin’s lymphoma (HL)) is presented in Table 2. Genotype associated with lymphoma subtypes FL and CLL. For FL, distribution among the controls was in Hardy–Weinberg haplotype T-G had haplotype-specific empirical P-value equilibrium for all investigated polymorphisms. of P ¼ 0.008, and A-G had haplotype-specific empirical We found a 20% decreased risk of all lymphoma P-value of P ¼ 0.006. The global empirical P-value was associated with the IL10RA Ser138Gly G-allele and TLR9 P ¼ 0.022. For CLL, haplotype-specific empirical P-value À1237 C-allele, which were of borderline statistical for T-G was 0.029 and for A-G 0.038, with empirical significance. None of the other variants revealed a global P-value of P ¼ 0.092. For TLR9 (TLR9 À1486T4 marked association with the risk of all lymphoma C-TLR9 À1237T4C), the haplotype T-C was found subtypes combined. In the subtype-specific analysis, associated with CLL with haplotype-specific empirical P-value of P ¼ 0.052, and C-T was found associated with T-NHL with haplotype-specific empirical P-value of Table 1 Characteristics of cases and control subjects P ¼ 0.003. However, the global empirical P-values were not statistically significant (CLL: P ¼ 0.199; T-NHL: Characteristic Men Women P ¼ 0.051). For IL10RA (IL10RA Ser138Gly-IL10RA Cases/controls Cases/controls Gly330Arg), we found an association with MH with the haplotype G-A and a haplotype-specific empirical P- N % N % value of P ¼ 0.040 (global empirical P-value P ¼ 0.203). No significant haplotype associations were found be- Cases/controls 390/390 100/100 320/320 100/100 tween the control group and the group of all lymphoma. Also, no haplotype association was found for IL10 Age at diagnosisa (years) between the control group and lymphoma subtypes. 18–24 23/25 5.9/6.4 16/16 5.0/5.0 25–34 33/33 8.5/8.5 26/27 8.1/8.4 35–44 50/48 12.8/12.3 35/34 10.9/10.6 Interaction between TLR2 À16933T4A and IL10RA 45–54 53/55 13.6/14.1 50/51 15.6/15.9 Ser138Gly polymorphisms 55–64 97/93 24.9/23.9 69/64 21.6/20.0 In multivariate analyses of FL cases and controls, we 65–74 101/100 25.9/25.6 87/91 27.2/28.4 found indications for statistical interaction between the 4 ¼ 75 33/36 8.5/9.2 37/37 11.6/11.6 TLR2 À16933T4A and the IL10RA Ser138Gly poly- morphisms (P-value for interaction ¼ 0.057). The TLR2 Educational level Low 176/151 45.1/38.7 204/172 63.8/53.8 À16933T4A A-allele increased the OR for FL only in Medium or high 210/238 53.9/61.0 115/148 35.9/46.3 individuals who were AA homozygotes for the IL10RA Unknown 4/1 1.0/0.3 1/0 0.3/0.0 Ser138Gly polymorphism (P ¼ 0.0014) (Table 3). In carriers of the G-allele, the TLR2 À16933T4A is not Availability of blood samples associated with the risk for FL (P ¼ 0.81). Yes 380/367 97.4/94.1 302/297 94.4/92.8 The combined occurrence of the TLR2 À16933T4AA- After therapy 139/1 35.6/0.3 117/1 36.6/0.3 Before therapy 238/366 61.0/93.8 183/296 57.2/92.5 allele and the IL10RA Ser138Gly AA wild-type genotype Unknown 3/0 0.8/0.0 2/0 0.6/0.0 was associated with a 24% increased risk for all lymphoma combined (OR ¼ 1.24, 95% CI ¼ 1.00–1.54), No 10/23 2.6/5.9 18/23 5.6/7.2 largely owing to the strongly elevated risk for FL (OR ¼ 2.41, 95% CI ¼ 1.49–3.88) and to a lesser extent aAge of control subjects: at the time of diagnosis of the matched owing to the association with HL (OR ¼ 1.59, 95% case. CI ¼ 1.02–2.47). None of the other subtypes was affected

Genes and Immunity Table 2 Odds ratios for all lymphoma combined and lymphoma subtypes associated with polymorphisms within TLR1, À2, À4, À5, À9, IL10 and IL10RA among case subjects and control subjects from a lymphoma case-control study in Germany

Controls n (%) All lymphoma n (%) DLBCL n (%) FL n (%) CLL n (%) MM n (%) HL n (%) T-NHL n (%) MALT n (%)

TLR1 À7202A4G AA 382 (58) 383 (57) 90 (60) 48 (55) 56 (55) 39 (53) 58 (53) 18 (55) 16 (59) AG 237 (36) 245 (36) 52 (34) 32 (37) 38 (37) 28 (38) 44 (40) 14 (42) 8 (30) GG 44 (7) 47 (7) 9 (6) 7 (8) 8 (8) 6 (8) 7 (6) 1 (3) 3 (11) NA 6 3 1 0 0 0 2 0 0 OR (95% CI) 1 1.04 0.92 1.06 1.23 1.29 1.09 1.10 0.99 AG/GG vs AA (0.84–1.29) (0.64–1.32) (0.67–1.67) (0.80–1.80) (0.79–2.10) (0.70–1.69) (0.54–2.23) (0.45–2.17)

TLR2 À16933T4A TT 178 (27) 166 (25) 33 (22) 10 (11) 36 (36) 18 (25) 26 (24) 11 (33) 4 (15) AT 313 (47) 348 (52) 85 (57) 49 (56) 43 (43) 38 (53) 59 (54) 14 (42) 17 (63) AA 173 (26) 154 (23) 30 (20) 28 (32) 21 (21) 16 (22) 25 (23) 8 (24) 6 (22) NA 5 10 4 0 2 1 1 0 0 OR (95% CI) 1 1.19 1.26 2.82 0.61 1.06 1.25 1.64 2.07 AT/AA vs TT (0.85–1.66) (0.82–1.92) (1.43–5.59)y (0.38–0.95)& (0.60–1.87) (0.75–2.07) (0.55–4.89) (0.70–6.10)

TLR2 Arg753Gln G4A GG 611 (92) 628 (93) 140 (93) 82 (94) 93 (93) 69 (95) 104 (94) 29 (88) 27 (100) GA 52 (8) 47 (7) 11 (7) 5 (6) 7 (7) 4 (5) 7 (6) 4 (12) 0 AA 1 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 0 NA 5 3 1 0 2 0 0 0 0

OR (95% CI) 1 1.11 0.90 0.67 0.89 0.68 0.79 0.74 — risk lymphoma and Nieters IL10RA A IL10, TLRs, in Variants GA/AA vs GG (0.87–1.42) (0.46–1.78) (0.26–1.74) (0.39–2.04) (0.24–1.95) (0.33–1.86) (0.35–1.55)

TLR4 Asp299Gly A4G al et AA 596 (89) 590 (87) 137 (91) 74 (85) 94 (94) 65 (89) 91 (82) 27 (82) 20 (74) AG 71 (11) 84 (12) 14 (9) 13 (15) 6 (6) 8 (11) 20 (18) 6 (18) 6 (22) GG 1 (0) 1 (0) 0 0 0 0 0 0 1 (4) NA 1 3 1 0 2 0 0 0 0 OR (95% CI) 1 1.19 0.87 1.51 0.49 1.0 1.80 1.90 2.76$ AG/GG vs AA (0.85–1.66) (0.47–1.59) (0.79–2.87) (0.21–1.17) (0.46–2.17) (0.99–3.26) (0.75–4.78) (1.12–6.81)

TLR5 Arg392StopC4T CC 582 (88) 603 (89) 132 (89) 78 (90) 92 (90) 64 (88) 101 (91) 30 (94) 24 (89) CT 79 (12) 67 (10) 20 (13) 8 (9) 8 (8) 8 (11) 9 (8) 2 (6) 3 (11) TT 4 (1) 6 (1) 0 1 (1) 2 (2) 1 (1) 1 (1) 0 0 NA 4 2 0 0 0 0 0 1 0 OR (95% CI) 1 0.85 1.07 0.82 0.71 0.97 0.62 0.46 0.87 TC/CC vs TT (0.61–1.19) (0.63–1.82) (0.40–1.71) (0.35–1.44) (0.46–2.04) (0.30–1.30) (0.11–1.97) (0.25–2.95)

TLR9 À1486T4C TT 230 (34) 227 (34) 57 (38) 21 (24) 28 (28) 27 (37) 42 (38) 7 (21) 10 (37) TC 331 (50) 332 (49) 65 (43) 49 (56) 58 (57) 36 (49) 53 (48) 14 (42) 14 (52) CC 106 (16) 117 (17) 29 (19) 17 (20) 15 (15) 10 (14) 16 (14) 12 (36) 3 (11) ee n Immunity and Genes NA 2 2 1 0 1 0 0 0 0 OR (95% CI) 1 1.04 0.85 1.61 1.33 0.88 0.91 2.00 0.88 TC/CC vs TT (0.83–1.31) (0.59–1.23) (0.96–2.71) (0.83–2.13) (0.53–1.46) (0.58–1.42) (0.85–4.69) (0.40–1.97) 617 ee n Immunity and Genes 618

Table 2 Continued

Controls n (%) All lymphoma n (%) DLBCL n (%) FL n (%) CLL n (%) MM n (%) HL n (%) T-NHL n (%) MALT n (%)

TLR9 À1237T4C TT 475 (71) 507 (75) 111 (73) 70 (80) 80 (78) 52 (71) 85 (77) 28 (85) 16 (59) TC 181 (27) 156 (23) 36 (24) 15 (17) 22 (22) 21 (29) 24 (22) 5 (15) 9 (33) CC 11 (2) 15 (2) 5 (3) 2 (2) 0 (0) 0 (0) 2 (2) 0 2 (7) NA 2 — 0 0 0 0 0 0 0

OR (95% CI) 1 0.83 0.94 0.62 0.66 1.01 0.82 0.44 1.72 risk lymphoma and IL10RA IL10, TLRs, in Variants TC/CC vs TT (0.66–1.06) (0.63–1.39) (0.36–1.09) (0.39–1.09) (0.59–1.73) (0.50–1.35) (0.17–1.16) (0.78–3.80)

IL10 À108214G AA 208 (32) 211 (32) 39 (27) 32 (37) 33 (33) 18 (25) 38 (35) 12 (39) 8 (30) AG 302 (46) 298 (45) 70 (48) 35 (40) 44 (44) 35 (49) 53 (49) 10 (32) 12 (44) GG 150 (23) 155 (23) 38 (26) 20 (23) 23 (23) 19 (26) 17 (16) 9 (29) 7 (26) NA 9 14 5 0 2 1 3 2 0 OR (95% CI) 1 0.99 1.25 0.79 0.88 1.26 0.97 0.79 0.97 AG/GG vs AA (0.78–1.25) (0.83–1.87) (0.49–1.26) (0.56–1.39) (0.71–2.21) (0.61–1.53) (0.38–1.67) (0.42–2.28)

IL10 À3575T4A Nieters A TT 262 (40) 268 (40) 56 (38) 37 (43) 39 (39) 20 (28) 45 (41) 16 (49) 11 (41) TA 306 (46) 302 (45) 69 (47) 33 (38) 50 (50) 39 (54) 56 (50) 10 (30) 11 (41) AA 93 (14) 100 (15) 23 (16) 17 (20) 11 (11) 13 (18) 10 (9) 7 (21) 5 (18) al et NA 8 8 4 0 2 1 0 0 0 OR (95% CI) 1 0.99 1.05 0.87 1.02 1.62 1.07 0.73 0.88 TA/AA vs TT (0.79–1.23) (0.73–1.53) (0.55–1.37) (0.66–1.58) (0.94–2.79) (0.69–1.66) (0.36–1.48) (0.40–1.93)

IL10RA Ser138Gly A4G AA 434 (65) 470 (70) 96 (64) 64 (74) 73 (73) 47 (64) 84 (76) 22 (67) 17 (63) AG 217 (33) 187 (28) 50 (33) 23 (26) 26 (26) 25 (34) 22 (20) 9 (27) 8 (30) GG 16 (2) 18 (3) 5 (3) 0 (0) 1 (1) 1 (1) 5 (5) 2 (6) 2 (7) NA 2 3 1 0 2 0 0 0 0 OR (95% CI) 1 0.81 1.11 0.67 0.74 1.09 0.51 0.89 1.10 AG/GG vs AA (0.65–1.02) (0.77–1.61) (0.40–1.11) (0.46–1.19) (0.66–1.81) (0.31–0.84)# (0.42–1.88) (0.49–2.46)

IL10RA Gly330Arg G4A GG 290 (44) 310 (46) 68 (46) 39 (45) 47 (47) 30 (41) 55 (50) 15 (45) 11 (41) GA 310 (47) 296 (44) 66 (44) 38 (44) 43 (43) 38 (52) 45 (41) 13 (39) 12 (44) AA 66 (10) 65 (10) 15 (10) 10 (11) 10 (10) 5 (7) 10 (9) 5 (15) 4 (15) NA 3 7 3 0 2 0 1 0 0 OR (95% CI) 1 0.90 0.93 0.94 0.89 1.11 0.77 0.93 1.09 GA/AA vs GG (0.72–1.12) (0.65–1.33) (0.60–1.48) (0.58–1.37) (0.68–1.81) (0.50–1.19) (0.46–1.88) (0.50–2.41)

Abbreviations: CI, confidence interval; CLL, chronic lymphocytic leukaemia; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; HL, Hodgkin’s lymphoma; MALT, mucosa- associated lymphoid tissue; MM, multiple myeloma; NA, not analysed; OR, odds ratio; T-NHL, T-cell non-Hodgkin’s lymphoma. y P-value ¼ 0.003. & P-value ¼ 0.03. $ P-value ¼ 0.03. # P-value ¼ 0.008. All OR adjusted for age (continuous) and sex. Variants in TLRs, IL10, IL10RA and lymphoma risk A Nieters et al 619 Table 3 Interaction between TLR2 À16933T4A and IL10RA À16933T4A polymorphism was inversely associated Ser138Gly polymorphisms in follicular lymphoma with the risk of FL, the TLR4 Arg299Gly variant with the risk of MALT lymphoma (Table 2). Controls FL OR (95% CI) To our knowledge, this is the first study that explored the association of a number of functionally important At IL10RA Ser138Gly GG/GA variants in TLR genes and IL10RA and the risk of TLR2 À16933 T4ATT 55 5 1 lymphoma and different subtypes. TLR2 À16933 T4A AT/AA 177 18 1.14 (0.40–3.25)$ IL10 mediates its biological effects by signalling At IL10RA Ser138Gly AA through the IL10 receptor, a heterotetramer consisting TLR2 À16933 T4A TT 123 5 1 of two dimeric receptor chains, IL10Ra and IL10Rb.25 TLR2 À16933 T4A AT/AA 309 59 4.62 (1.81–11.81)& Recently, two functionally important non-synonomous NA 5 0 polymorphisms have been identified in IL10RA.31 Homo- zygosity of the glycine variant of the IL10RA Gly330Arg Abbreviations: FL, follicular lymphoma; NA, not analysed. polymorphism was associated with reduced sensitivity $P-value ¼ 0.81. of IL10 and consequently reduced inhibition of tumour &P-value ¼ 0.001. necrosis factor-a (TNF-a) production.31 This variant was All OR adjusted for age (continuous) and sex. recently correlated with hepatic cirrhosis in infection.32 For the second variant (Ser138Gly), structure analyses suggested a potential impact on the by this combination (data not shown). The TLR2- conformation of the IL10–IL10RA complex.31 In our IL10RA-genotype combination association with FL risk study, this polymorphism was associated with a reduced was not modified by adjustment for other variables that risk of lymphoma, mainly driven by the 50% risk previously had been correlated with decreased (‘re- reduction for HL. It can be speculated that more effective ported herpes at other sites’) or increased (‘urticaria’)8 cellular immunity due to suboptimal IL10 signalling is a risk of FL (OR ¼ 2.2, 95% CI ¼ 1.36–3.59) (data not favourable factor for the pathobiology of HL. For this shown). group of lymphoid neoplasms, an infectious aetiology is suggested. The association of Epstein–Barr virus (EBV) Association with selected other diseases infection with a proportion of cases is well established In previous analyses, we had identified several variables and also for the non-EBV-associated cases an infectious 33 from questionnaire information that were positively or aetiology is discussed. An imbalanced T-helper (Th)1/ negatively associated with lymphoma risk.5,8 For some of Th2 response in the direction of Th1 may also underlie these, such as allergies and childhood viral infections, the positive association of these variants with reported biological rationale is given for an association with the non-B hepatitis and the inverse association with reported here explored gene variants in TLRs and members of the hay fever. IL10 family. Therefore, we explored – in the control The À16933A-allele in TLR2 did show a strong group – the distribution of variants in TLRs, IL10 and association with the risk of FL, and was underrepre- IL10RA in cases of other reported diseases and the sented among individuals with CLL. TLR2 recognizes respective controls (individuals that did not report that bacterial-derived molecules (lipoproteins, peptidogly- disease). Significant associations with reported allergies cans and glycolipids) and viral-derived envelope pro- were found for the TLR5 stop codon variant and the teins and lipopeptides and cooperates with TLR1 and 34 IL10RA variants (hay fever) (Table 4). The TLR4 299G- TLR6. Following the activation of the innate immune allele was highly overrepresented among cases of response, TLR2 appears to be important for the resolu- reported food allergy (OR ¼ 3.08, 95% CI ¼ 1.64–5.81). tion of inflammation and the generation of a homeostatic 24 The same variant conferred a decreased risk of aphthae. balance of the immune system after invasion. Genetic Several polymorphisms were associated with a de- variants in TLR2 may contribute to a shift of this balance creased (TLR1-7202A4G, IL10-1082A4G) or increased and consequently predispose towards immunopatho- (IL10RA variants) risk for reported non-B hepatitis. logies. An overly active anti-inflammatory response, for Reported childhood infections (measles and mumps) instance, may lead to immunosuppression and persis- were associated with variants in TLR9, reported scarlet tence of microorganisms in the host. This eventually fever with TLR1 À7202A4G (Table 4). results in chronic antigen stimulation, a known risk factor for lymphomagenesis.35,36 Polymorphisms in TLR2 have been associated with susceptibility to bacterial infection in mice and in humans.37,19,22 The À16933T4A Discussion variant has not been analysed functionally; however, it Factors at the interphase between environment and appears to be of clinical relevance. The T-allele has been immune system such as TLRs and IL10 are discussed associated with a decreased risk of asthma, atopic as risk factors for lymphoid neoplasms and other sensitization and current hay fever symptoms among immune response-related diseases. In this study, none children of farmers,13,38 indicating an impact of this of the IL10 variants was associated with lymphoma risk variant on immune responsiveness. Host immune nor with the risk of any of the subtypes. However, the response appears to be of particular importance for the IL10RA Ser138Gly variant showed a weak association clinical course of FL.39,40 The strong association of FL risk with the risk of all lymphoma combined, mainly owing with the TLR2 À16933T4A variant and some indication to the strong protective effect for HL risk. Furthermore, for a role of the TLR9 À1486T4C suggests that genes of we found an association of some TLR gene variants with innate immunity may also be aetiologically relevant. For specific lymphoma subtypes. Most strongly, the TLR2 the TLR2 variant, a 2.8-fold increased risk of FL

Genes and Immunity ee n Immunity and Genes 620

Table 4 OR for different reported diseases associated with polymorphisms within TLR1, À2, À4, À5, À9, IL10 and IL10RA among case subjects and control subjects from a lymphoma case–control study in Germany

Outcome Gene and SNP No. of frequent No. of No. of rare OR (95% CI) P-value genotype (%) heterozygote (%) genotype (%)

case/control case/control case/control risk lymphoma and IL10RA IL10, TLRs, in Variants

Reported allergy TLR5 Arg392Stop CC CT TT CT /TT vs CC 267 (90.8)/299 (84.5) 25 (8.5)/53 (15.0) 2 (0.7)/2 (0.6) 0.54 (0.32–0.90) 0.017

Reported hay fever IL10RA Ser138Gly AA AG GG AG/GG vs AA 69 (72.6)/363 (63.8) 24 (25.3)/192 (33.7) 2 (2.1)/14 (2.5) 0.60 (0.37–0.99) 0.043 IL10RA Gly330Arg GG GA AA GA/AA vs GG 51 (53.7)/237 (41.7) 39 (41.1)/270 (47.5) 5 (5.3)/61 (10.7) 0.61 (0.39–0.95) 0.028

Reported food allergy TLR4 Asp299Gly AA AG GG AG/GG vs AA 55 (77.5)/537 (90.7) 16 (22.5)/54 (9.1) 0 (0.00)/1 (0.2) 3.08 (1.64–5.81) 0.0005 Nieters A

Reported aphthae TLR4 Asp299Gly AA AG GG AG/GG vs AA tal et 88 (95.7)/505 (88.3) 3 (3.3)/67 (11.7) 1 (1.1)/0 (0.00) 0.35 (0.12–0.99) 0.047

Reported non-B hepatitis TLR1 À7202 A4G AA AG GG AG/GG vs AA 46 (74.2)/329 (55.7) 15 (24.2)/219 (37.1) 1 (1.6)/43 (7.3) 0.42 (0.23–0.77) 0.005 IL10RA Ser138Gly AA AG GG AG/GG vs AA 35 (55.6)/394 (66.3) 25 (39.7)/187 (31.5) 3 (4.8)/13 (2.2) 1.73 (1.00–2.97) 0.048 IL10RA Gly330Arg GG GA AA GA/AA vs GG 19 (30.2)/267 (45.0) 39 (61.9)/265 (44.7) 5 (7.9)/61 (10.3) 1.96 (1.1–3.48) 0.022 IL10 À1082 A4G AA AG GG AG/GG vs AA 25 (41.0)/179 (30.4) 24 (39.3)/272 (46.2) 12 (19.7)/138 (23.4) 0.54 (0.31–0.94) 0.031

Reported measles TLR9 À1486T4C TT TC CC TC/CC vs TT 177 (37.0)/20 (22.7) 221 (46.2)/55 (62.5) 80 (16.7)/13 (14.8) 0.48 (0.28–0.83) 0.009

Reported mumps TLR9 À1486T4C TT TC CC TC/CC vs TT 140 (41.2)/54 (24.8) 152 (44.7)/125 (57.3) 48 (14.1)/39 (17.9) 0.47 (0.32–0.69) o0.0001

Reported scarlet fever TLR1 À7202 A4G AA AG GG AG/GG vs AA 62 (49.2)/255 (60.6) 59 (46.8)/134 (31.8) 5 (4.0)/32 (7.6) 1.58 (1.06–2.37) 0.025

Abbreviations: CI, confidence interval; OR, odds ratio; SNP, single-nucleotide polymorphisms. Only associations with Po0.05 are depicted. All OR adjusted for age (continuous) and sex. Variants in TLRs, IL10, IL10RA and lymphoma risk A Nieters et al 621 (P ¼ 0.003) and a 40% decreased risk of CLL (P ¼ 0.03) participant for ‘physician-diagnosed food allergies’. It was presented. At first sight it appears implausible that cannot be excluded that misclassification has occurred. the same variant shows these opposite associations with CpG DNA from and bacteria is recognized lymphoma subtypes. However, it is increasingly recog- by TLR9 expressed in endosomes. In this study, the nized that lymphoma subtypes appear to differ in association of variants in TLR9 with lymphoma (FL, aetiology and molecular pathogenesis.41,42 The observed CLL) is weak. However, we observed a strong associa- associations could be explained if one assumes that TLR2 tion of TLR9 variants with childhood viral infections. modulates the effect of an environmental factor which is Individuals with the TLR9 À1486C-allele were less likely a risk factor of FL and slightly protective for CLL. As to have reported measles or mumps infection. Both subsequently discussed, we cannot exclude the possibi- investigated TLR9 single-nucleotide polymorphisms lity that one or both results are chance findings. (SNPs) have no known functional effects, but the C- We found evidence for a statistical interaction on the allele of the À1237T4C variant was associated with an multiplicative scale between the TLR2 À16933T4A and increased risk of asthma18 and Crohn’s disease.51 IL10RA Ser138Gly variants. TLR2 induces IL10 produc- Evidence for a link between measles virus and TLR9 tion and is linked to IL10-mediated immune suppres- was recently provided.52 Mice data point towards a sion.43–45 Furthermore, TLR2-dependent signals were critical role of TLR9 in viral sensing.53,54 TLR1À7202 recently shown to mediate the generation of down- A4G, previously associated with prostate cancer23 and modulating T-regulatory cells, IL10-secreting immune TLR5 Arg392Stop, correlated with impaired flagellin- cells with important function in balancing immune mediated signalling15 did not modulate the risk of response.46,44 lymphoma subtypes. The latter variant was associated We hypothesize that under the circumstance of with a reduced risk of reported allergies, which ‘enhanced’ IL10 sensitivity owing to the IL10RA AA- seems plausible considering the Th2 eliciting effect of genotype, the production of TNF-a and further pro- flagellin.55 inflammatory cytokines is decreased, with negative The observed associations with subgroups of lympho- consequences on cellular immunity. Impaired cellular ma are unlikely owing to population stratification. The immunity may confer an increased risk of FL owing study is population-based, 99% of the participants are to the weakened anti-viral and antitumour immune Caucasians and 485% are of German origin. It cannot be response. excluded that the observed associations reflect an The TLR4 Asp299Gly polymorphism was associated association with unknown causal variants in tight with an elevated risk for MALT lymphoma, HL (border- linkage disequilibrium with the investigated polymorph- line nonsignificant) and T-NHL (nonsignificant). This isms. The statistical power is 485% to identify risk variant attenuates receptor signalling12,20 and patients estimates of 1.5 and larger for the combined group of carrying this polymorphism had an increased suscept- lymphomas. For subtype-specific analysis, however, the ibility to Gram-negative infections.17,47 Furthermore, this study is underpowered and can be considered explora- variant has been associated with several other infectious tory in nature. Additionally, owing to multiple testing of and inflammatory conditions.14,17,22 TLR4 activates sev- multiple SNPs in various strata and for the analysis of eral intracellular signalling pathways including the interactions between SNPs in IL10RA and TLR2,we nuclear factor-kB pathway, resulting in the release of cannot exclude that some of the associations may be due inflammatory mediators, anti-microbial peptides and to chance. Therefore, replication in independent popula- other molecules that link innate and adaptive immunity. tions is needed to confirm or refute our findings. Disruption of TLR4 signalling may have negative Indicators of the so called ‘hygiene hypothesis’ such as consequences on the interplay of innate and adaptive frequent contact to animals at young age, family size and immune responses necessary to eradicate pathogens like childhood diseases appear to be correlated also with the Helicobacter pylori. Chronic inflammation by H. pylori is risk of lymphoma among adults.5–7 It may very well be implicated in the aetiology of MALT lymphoma48 and its that TLRs and other factors of the innate immune system eradication with antibiotics alone resulted in regression as sensors of the environment are crucial factors in of gastric MALT lymphoma.48,49 Evidence from this mediating these effects. The more in depth investigation study and a previous report16 suggests that TLR4 of genetic variability in these genes and downstream Asp299Gly variant may be one of the host factors signalling molecules may also help to bring into the affecting the risk of lymphoma among H. pylori-infected inverse association between allergies and lymphoma that individuals. The variant was underrepresented among have repeatedly been reported and that are unexpected cases of gastric MALT lymphoma in a previous study16 on the basis of the hygiene hypothesis. and is positively associated with MALT lymphoma in our study population. Although both study populations have a similar ethnic background, an unmeasured Materials and methods ‘causal’ mutation might be in linkage disequilibrium with different variants of TLR4 Asp299Gly. The TLR4 Study population Asp299Gly variant is also associated with a threefold A population-based case–control study was carried out increased risk of reported food allergy. Interestingly, in between 1999 and 2002 in six regions of Germany among support of our finding, evidence for a greatly heightened 18- to 80-year-old adults, 1:1 matched for gender, age susceptibility to allergic response to food antigens has ( þ /À1 year of birth) and study region (Table 1). Details been presented in strains of mice lacking a functional on study design have been published elsewhere.5,8 The TLR4.50 However, for the interpretation of our study cases were recruited from hospitals and office-based result it should be emphasized that ‘food allergies’ were physicians involved in diagnosis and treatment of not verified by a clinician, instead we asked each lymphoma in the study regions, interviewed by trained

Genes and Immunity Variants in TLRs, IL10, IL10RA and lymphoma risk A Nieters et al 622 interviewers and asked for a 20 ml blood sample. Eligible a role in lymphomagenesis.27 The IL10 À1082 A4G cases were asked for participation as soon as possible polymorphism has been associated with multiple in- after diagnosis. The median time from diagnosis to fectious and inflammatory diseases60 and affects suscept- interview was 27 days (mean: 38 days). Participation rate ibility and outcome of NHL.29,28 For the IL10RA was 87.4%. Diagnoses were collected in form of copies of Gly330Arg, IL10-mediated inhibition of TNF-a produc- the official pathology or physicians’ reports, which were tion in human monocytes was demonstrated.31 The for about 46% of the cases verified by reference IL10RA Ser138Gly variant was selected because structur- pathologists. Additionally, a 10% random sample of all al modelling indicated potential conformational changes included cases was re-evaluated within a reassessment with potential impact on IL10 signalling.31 system in the context of the European collaboration, Epilymph. All cases were classified according to the Genotyping World Health Organization (WHO) classification system Genotyping for TLR2 Arg753Gln was performed by by Jaffe´ et al.56 The study comprises 710 case–control polymerase chain reaction-restriction fragment length pairs (390 male, 320 female subjects) including 554 cases polymorphism (PCR-RFLP),61 genotyping for IL10 of B-NHL, 158 cases of DLBCL, 92 cases of FL, 104 cases À1082A4G by allele-specific PCR.62 Genotyping of the of B-CLL, 76 cases MM, 29 cases of MALT lymphoma, 35 other variants was performed using Pyrosequencing cases of T-NHL and 115 cases of HL. One case was technology.63 The respective primers and representative diagnosed as HL and NHL.8 The controls were drawn pyrograms are listed in Supplementary Tables 1 and 2. randomly from population registers of the study regions All PCRs were performed in an Eppendorf Mastercycler which have an almost 100% coverage owing to compul- (Eppendorf, Hamburg, Germany). PCR mixtures con- sory registration by law. Participation rate was 44.3% tained 1 Â Reddy Mix PCR buffer (ABgene, Epsom, UK), including all people who could not be contacted and 1 U of Thermoprime DNA polymerase (ABgene), deox- who refused participation owing to illness. The response ynucleoside triphosphates (Stratagene, La Jolla, USA rate was 53.3% excluding these groups. The study was each at 250 mM) and primers (each at 1 pmol) in a final approved by the local ethical committees and written volume of 10 ml. Cycling conditions were identical for all informed consent was obtained from each subject. investigated polymorphisms: 941C 2 min, 35 Â (941C40s, 571C40s,721C 40 s). Biological samples Blood was drawn in Vacutainer CPT tubes (Becton Quality control procedures. Pyrosequencing technology Dickinson, NJ, USA) for plasma and lymphocytes and provides the sequence of the investigated polymorph- Vacutainer SST tubes (Becton Dickinson) for serum. isms as well as the surrounding sequence. Therefore, it is Blood was taken before therapy for 62% of the cases. a very accurate method to detect SNPs. At least six per From the participants (710 cases, 710 controls), blood cent of the samples were selected at random for repeated specimens were obtained from 96.1% of the patients and genotyping with Pyrosequencing technology or PCR- 93.5% of the controls. Blood was processed by centrifu- RFLP, respectively. Furthermore, samples with ambig- gation according to the manufacturer’s recommendation uous results were repeated. All 96-well plates contained

(Becton Dickinson). Lymphocytes were washed with negative (H2O) controls, and a set of three positive phosphate-buffered saline, step-wise frozen in foetal calf controls of the respective genotypes (homozygote for serum 10% dimethyl sulphoxide solution and stored at allele 1, heterozygote, homozygote for allele 2). For the À801C. Genomic DNA from lymphocytes was extracted rare variants, only the first two genotypes were included. using the QiaAmp Blood according to the manufac- For some variants (TLR2 À16933T4A, TLR2 Arg753Gln, turer’s instructions (Qiagen, Hilden, Germany). IL10 À3575T4A, IL10 À1082A4G), DNA samples with the respective genotype from 102 ethnically diverse Selection of polymorphisms individuals were used as positive controls that have The TLR1 À7202 A4G promoter polymorphism alters been sequenced and genotyped as part of the the putative core binding site of the proto-oncogene SNP500Cancer project.64 PU.1. and was associated with prostate cancer.23 The TLR2 À16933T4A polymorphism was shown to have an Statistical analysis effect on asthma and atopy among farmer’s children.13 All lymphoma subtypes combined were analysed using The TLR2 Arg753Gln variant was chosen because it was matched conditional logistic regression taking the shown to affect the response to bacterial peptides in matching variables gender, age and study region into vitro,19 and to be associated with several immune account. Owing to sparse data, subentities were analysed pathologies.11,21 The TLR4 Asp299Gly variant has been by unconditional logistic regression involving all control related to hyporesponsiveness to inhaled endotoxins in subjects and using the matching variables for adjust- humans12,20 and with several infectious and inflamma- ment. Age at diagnosis of the case and age of the control tory conditions.14,17,9 The TLR5 Arg392Stop variant in at the time of diagnosis of the respective matched case the ligand-binding domain abolished flagellin-induced was taken. Relative risk was estimated by OR and signalling in CHO-K1 cells57 and was associated with associated 95% CI using the SAS procedure LOGISTIC the susceptibility to Legionnaire’s disease15 and systemic for unconditional logistic regression. Lupus erythematosus.58 The TLR9 À1237T4C variant In a logistic regression model including multiplicative was associated with asthma in a European-American terms, interaction between genes was explored. Evidence study population and in conjunction with the of statistical interaction on the multiplicative scale was À1486T4C polymorphism differentiates between the declared if the difference of the likelihood ratio of the most frequent haplotypes.18 The IL10 À3575 T4A was nested model including the main variables and an selected because it affects IL10 production59 and has interaction term and the model including only the main

Genes and Immunity Variants in TLRs, IL10, IL10RA and lymphoma risk A Nieters et al 623 variables was larger than 3.84, corresponding to an alpha 10 Beutler B. Inferences, questions and possibilities in Toll-like level of 0.05 for 1 d.f. receptor signalling. Nature 2004; 430: 257–263. We used the software package Haplo.score65 to test 11 Ahmad-Nejad P, Mrabet-Dahbi S, Breuer K, Klotz M, Werfel T, for association of indirectly deduced haplotypes defined Herz U et al. The Toll-like receptor 2 R753Q polymorphism by the SNPs in TLR2, TLR9, IL10 and IL10RA with defines a subgroup of patients with atopic dermatitis having phenotypes defined by disease status (lymphoma) and severe phenotype. J Allergy Clin Immunol 2004; 113: 565–567. lymphoma subtypes. Haplo.score assigns the probability 12 Arbour NC, Lorenz E, Schutte BC, Zabner J, Kline JN, Jones M for each haplotype pair consistent with the genotype of et al. TLR4 mutations are associated with endotoxin hypo- responsiveness in humans. Nat Genet 2000; 25: 187–191. an individual by EM algorithm, and then directly models 13 Eder W, Klimecki W, Yu L, von Mutius E, Riedler J, Braun- the individual’s phenotype as a function of each inferred Fahrlander C et al. Toll-like receptor 2 as a major gene for haplotype pair, weighted by their estimated probability, asthma in children of European farmers. J Allergy Clin to account for haplotype uncertainty. Haplo.score pro- Immunol 2004; 113: 482–488. vides haplotype-specific association tests as well as a 14 Franchimont D, Vermeire S, El Housni H, Pierik M, Van Steen global test. To account for sparse haplotype data, we K, Gustot T et al. Deficient host–bacteria interactions in performed simulations to derive empirical P-values. The inflammatory bowel disease? The Toll-like receptor (TLR)-4 number of simulations for empirical P-values was set as Asp299gly polymorphism is associated with Crohn’s disease 20 000. Haplotype association analysis was adjusted for and ulcerative colitis. Gut 2004; 53: 987–992. sex and age. 15 Hawn TR, Verbon A, Lettinga KD, Zhao LP, Li SS, Laws RJ et al. A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with suscept- ibility to Legionnaires’ disease. J Exp Med 2003; 198: 1563–1572. Acknowledgements 16 Hellmig S, Fischbach W, Goebeler-Kolve ME, Folsch UR, Hampe J, Schreiber S. Association study of a functional Toll- We thank the participants of the study and the many like receptor 4 polymorphism with susceptibility to gastric colleagues who supported the performance of the study. mucosa-associated lymphoid tissue lymphoma. Leuk Lym- A detailed list is presented in previous publications.5,8 phoma 2005; 46: 869–872. Furthermore, we thank Marlen Auer, Bettina Ehret and 17 Kiechl S, Lorenz E, Reindl M, Wiedermann CJ, Oberhollenzer F, Bonora E et al. Toll-like receptor 4 polymorphisms and Georg Zahn for excellent technical assistance. The study atherogenesis. N Engl J Med 2002; 347: 185–192. was funded by the Federal Office for Radiation Protec- 18 Lazarus R, Klimecki WT, Raby BA, Vercelli D, Palmer LJ, tion (StSch4261 and StSch4420). The European Commu- Kwiatkowski DJ et al. Single-nucleotide polymorphisms in the nity supported the set up of a common protocol for Toll-like receptor 9 gene (TLR9): frequencies, pairwise linkage assessment of occupational exposures (SOC 98 201307 disequilibrium, and haplotypes in three US ethnic groups and 05F02) and implementation of additional study areas exploratory case–control disease association studies. Genomics (QLK4-CT-2000-00422). Genotyping was supported by 2003; 81: 85–91. the German Jose´ Carreras Leukaemia foundation 19 Lorenz E, Mira JP, Cornish KL, Arbour NC, Schwartz DA. A (DJCLS_R04/08). This work was supported by a novel polymorphism in the Toll-like receptor 2 gene and its Deutsche Forschungsgemeinschaft Grant CH117/3-1 potential association with staphylococcal infection. Infect (LB). Immun 2000; 68: 6398–6401. 20 Michel O, LeVan TD, Stern D, Dentener M, Thorn J, Gnat D et al. Systemic responsiveness to lipopolysaccharide and polymorphisms in the Toll-like receptor 4 gene in human References beings. J Allergy Clin Immunol 2003; 112: 923–929. 21 Ogus AC, Yoldas B, Ozdemir T, Uguz A, Olcen S, Keser I et al. 1 Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world The Arg753GLn polymorphism of the human Toll-like receptor 2 cancer burden: Globocan 2000. Int J Cancer 2001; 94: 153–156. gene in tuberculosis disease. Eur Respir J 2004; 23: 219–223. 2 Baris D, Zahm SH. Epidemiology of lymphomas. Curr Opin 22 Schroder NW, Schumann RR. Single nucleotide polymorph- Oncol 2000; 12: 383–394. isms of Toll-like receptors and susceptibility to infectious 3 Vineis P, Crosignani P, Sacerdote C, Fontana A, Masala G, disease. Lancet Infect Dis 2005; 5: 156–164. Miligi L et al. Haematopoietic cancer and medical history: a 23 Sun J, Wiklund F, Zheng SL, Chang B, Balter K, Li L et al. multicentre case control study. J Epidemiol Community Health Sequence variants in Toll-like receptor gene cluster (TLR6- 2000; 54: 431–436. TLR1-TLR10) and prostate cancer risk. J Natl Cancer Inst 2005; 4 Muller AM, Ihorst G, Mertelsmann R, Engelhardt M. 97: 525–532. Epidemiology of non-Hodgkin’s lymphoma (NHL): trends, 24 Netea MG, Van der Meer JW, Kullberg BJ. Toll-like receptors geographic distribution, and etiology. Ann Hematol 2005; as an escape mechanism from the host defense. Trends 84: 1–12. Microbiol 2004; 12: 484–488. 5 Becker N, Deeg E, Nieters A. Population-based study of 25 Moore KW, de Waal MR, Coffman RL, O’Garra A. Interleukin- lymphoma in Germany: rationale, study design and first 10 and the interleukin-10 receptor. Annu Rev Immunol 2001; 19: results. Leuk Res 2004; 28: 713–724. 683–765. 6 Grulich AE, Vajdic CM, Kaldor JM, Hughes AM, Kricker A, 26 Khatri VP, Caligiuri MA. A review of the association between Fritschi L et al. Birth order, atopy, and risk of non-Hodgkin interleukin-10 and human B-cell malignancies. Cancer Immunol lymphoma. J Natl Cancer Inst 2005; 97: 587–594. Immunother 1998; 46: 239–244. 7 Vineis P, Miligi L, Crosignani P, Fontana A, Masala G, Nanni O 27 Rothman N, Skibola CF, Wang SS, Morgan G, Lan Q, Smith et al. Delayed infection, family size and malignant lympho- MT et al. Genetic variation in TNF and IL10 and risk of non- mas. J Epidemiol Community Health 2000; 54: 907–911. Hodgkin lymphoma: a report from the InterLymph Consor- 8 Becker N, Deeg E, Rudiger T, Nieters A. Medical history and tium. Lancet Oncol 2006; 7: 27–38. risk for lymphoma: results of a population-based case–control 28 Lech-Maranda E, Baseggio L, Bienvenu J, Charlot C, Berger F, study in Germany. Eur J Cancer 2005; 41: 133–142. Rigal D et al. Interleukin-10 gene promoter polymorphisms 9 Cook DN, Pisetsky DS, Schwartz DA. Toll-like receptors in the influence the clinical outcome of diffuse large B-cell lym- pathogenesis of human disease. Nat Immunol 2004; 5: 975–979. phoma. Blood 2004; 103: 3529–3534.

Genes and Immunity Variants in TLRs, IL10, IL10RA and lymphoma risk A Nieters et al 624 29 Cunningham LM, Chapman C, Dunstan R, Bell MC, Joske DJ. 48 Farinha P, Gascoyne RD. Helicobacter pylori and MALT Polymorphisms in the gene promoter are lymphoma. Gastroenterology 2005; 128: 1579–1605. associated with susceptibility to aggressive non-Hodgkin’s 49 Wotherspoon AC, Doglioni C, Diss TC, Pan L, Moschini A, de lymphoma. Leuk Lymphoma 2003; 44: 251–255. Boni M et al. Regression of primary low-grade B-cell gastric 30 Lan Q, Zheng T, Rothman N, Zhang Y, Wang SS, Shen M et al. lymphoma of mucosa-associated lymphoid tissue type after Cytokine polymorphisms in the Th1/Th2 pathway and eradication of Helicobacter pylori. Lancet 1993; 342: 575–577. susceptibility to non-Hodgkin lymphoma. Blood 2006; 107: 50 Bashir ME, Louie S, Shi HN, Nagler-Anderson C. Toll-like 4101–4108. receptor 4 signaling by intestinal microbes influences suscept- 31 Gasche C, Grundtner P, Zwirn P, Reinisch W, Shaw SH, ibility to food allergy. J Immunol 2004; 172: 6978–6987. Zdanov A et al. Novel variants of the IL-10 receptor 1 affect 51 Torok HP, Glas J, Tonenchi L, Bruennler G, Folwaczny M, inhibition of TNF-alpha production. J Immunol 2003; Folwaczny C. Crohn’s disease is associated with a Toll-like 170: 5578–5582. receptor-9 polymorphism. Gastroenterology 2004; 127: 365–366. 32 Hofer H, Neufeld JB, Oesterreicher C, Grundtner P, Wrba F, 52 Schlender J, Hornung V, Finke S, Gunthner-Biller M, Marozin Gangl A et al. Bi-allelic presence of the interleukin-10 receptor S, Brzozka K et al. Inhibition of Toll-like receptor 7- and 9- 1 G330R allele is associated with cirrhosis in chronic HCV-1 mediated alpha/beta interferon production in human plas- infection. Genes Immun 2005; 6: 242–247. macytoid dendritic cells by respiratory syncytial virus and 33 Jarrett RF. Risk factors for Hodgkin’s lymphoma by EBV status measles virus. J Virol 2005; 79: 5507–5515. and significance of detection of EBV genomes in serum of 53 Ito T, Wang YH, Liu YJ. Plasmacytoid pre- patients with EBV-associated Hodgkin’s lymphoma. Leuk cursors/type I interferon-producing cells sense viral infection Lymphoma 2003; 44 (Suppl 3): S27–S32. by Toll-like receptor (TLR) 7 and TLR9. Springer Semin 34 Wetzler LM. The role of Toll-like receptor 2 in microbial Immunopathol 2005; 26: 221–229. disease and immunity. Vaccine 2003; 21 (Suppl 2): S55–S60. 54 Tabeta K, Georgel P, Janssen E, Du X, Hoebe K, Crozat K et al. 35 Gaidano G, Dalla-Favera R. Biologic and molecular character- Toll-like receptors 9 and 3 as essential components of innate ization of non-Hodgkin’s lymphoma. Curr Opin Oncol 1993; 5: immune defense against mouse cytomegalovirus infection. 776–784. Proc Natl Acad Sci USA 2004; 101: 3516–3521. 36 Suarez F, Lortholary O, Hermine O, Lecuit M. Infection- 55 Didierlaurent A, Ferrero I, Otten LA, Dubois B, Reinhardt M, associated lymphomas derived from marginal zone B cells: a Carlsen H et al. Flagellin promotes myeloid differentiation model of antigen-driven lymphoproliferation. Blood 2006; 107: factor 88-dependent development of Th2-type response. 3034–3044. J Immunol 2004; 172: 6922–6930. 37 Echchannaoui H, Frei K, Schnell C, Leib SL, Zimmerli W, 56 Jaffe´ ES, Harris NL, Stein H, Vardiman JW. Pathology and Landmann R. Toll-like receptor 2-deficient mice are highly Genetics of Tumours of Haematopoietic and Lymphoid Tissues. susceptible to Streptococcus pneumoniae meningitis because of IARC Press, International Agency for Research on Cancer: reduced bacterial clearing and enhanced inflammation. J Infect Lyon, 2001. Dis 2002; 186: 798–806. 57 Merx S, Zimmer W, Neumaier M, Ahmad-Nejad P. Character- 38 Lauener RP, Birchler T, Adamski J, Braun-Fahrlander C, Bufe ization and functional investigation of single nucleotide A, Herz U et al. Expression of CD14 and Toll-like receptor 2 in polymorphisms (SNPs) in the human TLR5 gene. Hum Mutat farmers’ and non-farmers’ children. Lancet 2002; 360: 465–466. 2006; 27: 293. 39 Dave SS, Wright G, Tan B, Rosenwald A, Gascoyne RD, Chan 58 Hawn TR, Wu H, Grossman JM, Hahn BH, Tsao BP, Aderem WC et al. Prediction of survival in follicular lymphoma based A. A stop codon polymorphism of Toll-like receptor 5 is on molecular features of tumor-infiltrating immune cells. associated with resistance to systemic lupus erythematosus. N Engl J Med 2004; 351: 2159–2169. Proc Natl Acad Sci USA 2005; 102: 10593–10597. 40 Coffman RL, Reiner SL. Instruction, selection, or tampering 59 Gibson AW, Edberg JC, Wu J, Westendorp RG, Huizinga TW, with the odds? Science 1999; 284: 1283, 1285. Kimberly RP. Novel single nucleotide polymorphisms in the 41 Harris NL, Stein H, Coupland SE, Hummel M, Favera RD, distal IL-10 promoter affect IL-10 production and enhance the Pasqualucci L et al. New approaches to lymphoma diagnosis. risk of systemic lupus erythematosus. J Immunol 2001; 166: Hematology (Am Soc Hematol Educ Program) 2001; 1: 194–220. 3915–3922. 42 Grulich AE, Vajdic CM. The epidemiology of non-Hodgkin 60 Hurme M, Lahdenpohja N, Santtila S. Gene polymorphisms of lymphoma. Pathology 2005; 37: 409–419. interleukins 1 and 10 in infectious and autoimmune diseases. 43 Jang S, Uematsu S, Akira S, Salgame P. IL-6 and IL-10 Ann Med 1998; 30: 469–473. induction from dendritic cells in response to Mycobacterium 61 Schroder NW, Hermann C, Hamann L, Gobel UB, Hartung T, tuberculosis is predominantly dependent on TLR2-mediated Schumann RR. High frequency of polymorphism Arg753Gln recognition. J Immunol 2004; 173: 3392–3397. of the Toll-like receptor-2 gene detected by a novel allele- 44 Netea MG, Sutmuller R, Hermann C, Van der Graaf CA, Van specific PCR. J Mol Med 2003; 81: 368–372. der Meer JW, van Krieken JH et al. Toll-like receptor 2 62 Perrey C, Turner SJ, Pravica V, Howell WM, Hutchinson IV. suppresses immunity against Candida albicans through ARMS-PCR methodologies to determine IL-10, TNF-alpha, induction of IL-10 and regulatory T cells. J Immunol 2004; TNF-beta and TGF-beta 1 gene polymorphisms. Transpl 172: 3712–3718. Immunol 1999; 7: 127–128. 45 Schaub B, Campo M, He H, Perkins D, Gillman MW, Gold DR 63 Fakhrai-Rad H, Pourmand N, Ronaghi M. Pyrosequencing: an et al. Neonatal immune responses to TLR2 stimulation: accurate detection platform for single nucleotide polymorph- influence of maternal atopy on Foxp3 and IL-10 expression. isms. Hum Mutat 2002; 19: 479–485. Respir Res 2006; 7: 40. 64 Packer BR, Yeager M, Staats B, Welch R, Crenshaw A, Kiley M 46 Caramalho I, Lopes-Carvalho T, Ostler D, Zelenay S, Haury et al. SNP500Cancer: a public resource for sequence validation M, Demengeot J. Regulatory T cells selectively express Toll- and assay development for genetic variation in candidate like receptors and are activated by lipopolysaccharide. J Exp genes. Nucleic Acids Res 2004; 32: D528–D532. Med 2003; 197: 403–411. 65 Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. 47 Lorenz E, Mira JP, Frees KL, Schwartz DA. Relevance of Score tests for association between traits and haplotypes mutations in the TLR4 receptor in patients with Gram- when linkage phase is ambiguous. Am J Hum Genet 2002; 70: negative septic shock. Arch Intern Med 2002; 162: 1028–1032. 425–434.

Supplementary Information accompanies the paper on Genes and Immunity website (http://www.nature.com/gene)

Genes and Immunity