Nucleotide diversity of the TNF gene region in an African village

(2001) 2, 343 348 2001 Nature Publishing Group All rights reserved 1466-4879/01 $15.00 www.nature.com/gene Nucleotide diversity of the TNF gene region in an African village A Richardson 1, F Sisay-Joof 2, H Ackerman 1, S Usen 2, P Katundu 3, T Taylor 3, M Molyneux 3, M Pinder 2 and D Kwiatkowski 1 1 Wellcome Trust Centre for Human Genetics, Oxford University, UK; 2 MRC Laboratories, Fajara, The Gambia; 3 Wellcome Trust Research Laboratories and Malawi Project, College of Medicine, University of Malawi, Blantyre, Malawi The wide variety of disease associations reported at the TNF locus raises the question of how much variation exists within a single population. To address this question, we sequenced the entire TNF gene in 72 chromosomes from healthy residents of a village in The Gambia, West Africa. We found 12 polymorphisms in 4393 nucleotides, of which five have not been previously described, giving an estimated nucleotide diversity ( ) of 5.6 10 4. A significantly higher frequency of polymorphisms was found in the promoter region than in the coding region (8/1256 vs 0/882 nucleotides, P = 0.02). All polymorphisms with the exception of one rare allele were found to be present in Malawi, which is both geographically and genetically distant from The Gambia. Genotyping of 424 Gambian and 121 Malawian adults showed a significant frequency difference between the two populations for eight of the 12 polymorphisms, but the average fixation index across the variable sites was relatively low (F ST = 0.007). We conclude that, at the TNF locus, the nucleotide diversity found within a single African village is similar to the global value for human autosomal genes sampled across different continents. (2001) 2, 343 348. Keywords: The Gambia; TNF gene; Malawi Introduction TNF, the gene encoding tumour necrosis factor (TNF), resides in the central part (class III region) of the major histocompatibility complex (MHC) surrounded by a large number of other immunological genes. 1 Many disease associations have been reported with HLA alleles that are in linkage disequilibrium with TNF and in recent years there have been growing reports of disease linkage and association with microsatellite polymorphisms and single nucleotide polymorphisms close to the TNF gene itself. 2 11 A precise description of the nucleotide diversity of this region is therefore fundamental to investigations of genetic susceptibility to infectious and inflammatory disease. As well as serving as a genetic marker for the surrounding gene region, TNF polymorphisms may be of intrinsic functional relevance. TNF is a key mediator of the inflammatory response and is critical for host defence against a wide variety of pathogenic microbes. The dual role of TNF, acting as an agent of both innate immunity and inflammatory pathology, poses a considerable challenge for gene regulation. When exposed to a serious infection, the survival of the host depends on an adequate but not excessive level of TNF production. This optimal response is likely to vary for different infectious diseases, resulting in conflicting evolutionary pressures that might be expected to generate phenotypic diversity Correspondence: Prof D Kwiatkowski, Wellcome Trust Centre for Human Genetics, Roosevelt Drive, Oxford OX3 7BN,UK E-mail dominic.kwiatkowski paediatrics.ox.ac.uk This work was funded by the Medical Research Council. Received 3 May 2001; revised and accepted 11 July 2001 in TNF regulation. There is strong evidence that individuals vary in their level of TNF responsiveness to standard stimuli such as bacterial lipopolysaccharide, and that this has a significant heritable component. 12 Many papers have been published on the relationship between TNF polymorphisms and levels of TNF production, based on various human investigations and experimental models of gene regulation, but there is currently little consensus concerning whether any specific polymorphism is truly functional and, if so, how it operates (reviewed in Knight and Kwiatkowski 8 ). These observations raise the question of exactly how much nucleotide diversity exists at the TNF locus within a single population. Considering the large body of literature on disease associations with human TNF polymorphisms, there is a surprising lack of data to address this topic. Although eight polymorphisms are known to exist in the 5 flanking region, we are not aware of any systematic analysis of the level of nucleotide diversity throughout the coding, intronic and flanking regions. African populations are typically more diverse than Caucasian or Asian populations, 13 and the Gambian TNF locusisofparticularinterest because of the various TNF and MHC associations that have been observed with susceptibility to malaria and chlamydial infection in this population. 2,5 7,14 We therefore sequenced the entire TNF gene in 36 healthy residents of a single West African village. The total sequence under investigation was 4393 nt, extending from 1389 nt 5 of the transcription start site to +3004 nt 3 of the polyadenylation site.

Nucleotide diversity of the TNF gene region 344 Figure 1 Schematic of the TNF locus, taken from Genbank accession no. M16441. Black bars represent translated sequence; grey bars represent the 5 and 3 untranslated regions of the primary transcript. Results Figure 1 depicts the polymorphisms identified by sequencing 72 Gambian chromosomes, and Table 1 describes the primer pairs used to genotype them. All locations refer to the transcription start site of the TNF gene as identified on Genbank accession no. M16441: 15 thus we refer to nucleotide 4095 on that sequence as +1, and nucleotide 4094 as 1. All polymorphisms found by sequencing were heterozygous and were confirmed by sequencing in the reverse direction. The polymorphisms at 1031, 863, 857, 376, 308, 244 and 238 nt have been previously described. In this sample we did not detect the polymorphisms at 575 163, +70 and +488 nt reported by other investigators in different populations. The polymorphisms at 1073, +467, +851, +943 and +1304 nt are novel. The 12 polymorphisms detected by sequencing represent an overall frequency of 1 per 338 nucleotides. Eight were located in the promoter region, defined here as the 1256 nt sequence extending from the TNF transcriptional start site to the polyadenylation site of the LTa gene (1 per 157 nucleotides). Four polymorphisms were found in the three introns which comprise 1094 bases (1 per 273 nucleotides). No polymorphisms were found with four exons covering 882 bases or in the 3 UTR which covers 789 bases. The difference between the frequency of promoter polymorphisms and the frequency of coding polymorphisms (8/1256 vs 0/882) is statistically significant (P = 0.02 by two-tailed Fisher s exact test). For comparison of nucleotide diversity with other gene regions, we used the normalised number of variant sites as employed by Cargill et al: 13 = K/ n 1 i=1 i 1 L where K is the observed number of variant sites (12), L is the total sequence length (4393 nt) and n is the number of chromosomes sequenced (72). The respective values are: 5.64 10 4 for the whole of the region; 0 for the coding region; 7.05 10 4 for the non-coding region overall; 7.54 10 4 for the introns; and 13.14 10 4 for the 5 flanking region. To obtain a more accurate estimate of allele frequencies in this population, a sequence specific oligonucleotide PCR method was used to determine genotypes for an independent sample of 424 unrelated healthy Gambian adults. The results are shown in Table 2. All alleles detected by sequencing were found by genotyping in this separate group of individuals, with the exception of the TNF+943 polymorphism which was detected by genotyp- Table 1 Consensus primer sequences were located as follows: for SNPs 1077 to 863, consensus primer at 637 nt; for SNP 857, consensus primer at 1129; for SNPs 376 to 244, consensus primer at 2; for SNP 238, consensus primer at 637; for SNPs +467 and +488, consensus primer at +820 con; for SNPs +846 and +938, consensus primer +1274; for SNP +1299, consensus primer at +1016 con sense. All reactions included the positive control primers 63 and 64 which amplify a conserved 796 bp segment of exon 3 of the HLA-DRB1 gene Allele 1 primer Allele 2 primer Conserved primer TNF 1073 GGA CTC ACC AGG TGA GGC C GGA CTC ACC AGG TGA GGC T CCG GGA ATT CAC AGA CCC C TNF 1031 CAA AGG AGA AGC TGA GAA GAT CAA AGG AGA AGC TGA GAA GAC CCG GGA ATT CAC AGA CCC C TNF 863 CGA GTA TGG GGA CCC CCC GAG TAT GGG GAC CCC CA CCG GGA ATT CAC AGA CCC C TNF 857 TCT ACA TGG CCC TGT CTT CG TCT ACA TGG CCC TGT CTT CA AAG GAT AAG GGC TCA GAG AG TNF 376 CCT GCA TCC TGT CTG GAA G TCC TGC ATC CTG TCT GGA AA GGC TGG GTG TGC CAA CAA C TNF 308 ATA GGT TTT GAG GGG CAT GG TAG GTT TTG AGG GGC ATG A GGC TGG GTG TGC CAA CAA C TNF 244 CCA GAA GAC CCC CCT CG CCA GAA GAC CCC CCT CA GGC TGG GTG TGC CAA CAA C TNF 238 CCC CAT CCT CCC TGC TCC CCC CAT CCT CCC TGC TCT GGG GTC TGT GAA TTC CCG G TNF+467 GTG CGC TGA TAG GGA GGG GTG CGC TGA TAG GGA GGA CTC TTT CCC TGA GTG TCT TC TNF+851 TGC TGG AAG GTG AAT ACA CG ATG CTG GAA GGT GAA TAC ACA AAG ACA CAT CCT CAG AGC TC TNF+943 CTT TAA GGG TGA CTC CCT CG CTT TAA GGG TGA CTC CCT CA AAG ACA CAT CCT CAG AGC TC TNF+1304 CCA TCA GCC GGG CTT CAA T CCA TCA GCC GGG CTT CAA C GTA AGT GTC TCC AAA CCT CTT Control TGC CAA GTG GAG CAC CCA A GCA TCT TGC TCT GTG CAG AT 63/64

Nucleotide diversity of the TNF gene region Table 2 Polymorphisms detected by sequencing 36 Gambian adults throughout the region 1389 nt to +3004 nt in relation to the TNF transcriptional start site identified on Genbank accession no. M16441. The column entitled variants seen on sequencing lists the number of occurences of the rare allele on sequencing 72 Gambian chromosomes. The differentiation of Gambians and Malawians is quantified by Wright s fixation index (F ST ) and by the P value of the 2 test comparing allele frequencies in the two populations 345 Position Region Common/ Variants Gambia Malawi Gambia vs Malawi rare allele seen on sequencing n Homozygotes/ Allele n Homozygotes/ Allele F ST P heterozygotes frequency heterozygotes frequency for rare allele for rare allele 1073 5 FR C/T 1 421 0/34 4% 121 0/14 6% 0.001 NS 1031 5 FR T/C 5 416 8/99 14% 119 7/46 25% 0.016 0.0001 863 5 FR C/A 2 424 3/42 6% 121 0/33 14% 0.016 0.0001 857 5 FR C/T 2 424 2/45 4% 120 0/1 0% 0.011 0.001 376 5 FR G/A 1 424 0/21 3% 119 0/15 6% 0.008 0.01 308 5 FR G/A 11 422 9/30 17% 120 1/24 11% 0.007 0.05 244 5 FR G/A 1 420 0/25 2% 118 0/20 8% 0.013 0.001 238 5 FR G/A 1 422 3/62 8% 118 0/24 10% 0.001 NS +467 Intron 1 G/A 7 420 0/25 2% 116 0/6 3% 0.000 NS +851 Intron 1 A/G 3 417 9/81 12% 119 4/32 17% 0.004 0.05 +943 Intron 1 G/A 1 420 0/0 0% 120 0/0 0% NS +1304 Intron 3 A/G 4 417 5/83 11% 118 4/31 17% 0.005 0.05 ing only in the individual who was originally found to have this polymorphism by sequencing. All allele frequencies were consistent with Hardy Weinberg equilibrium. Allele frequencies for the same polymorphisms were determined in 121 unrelated healthy Malawian adults. All of the Gambian polymorphisms were also found in Malawi apart from the rare TNF+943A allele. However eight of the polymorphisms showed a significant difference in allele frequency between the two populations when analysed by chi-squared test. In particular, the 1031C allele and the 863A allele were approximately double the frequency in Malawi compared to that in The Gambia (25% vs 14% and 14% vs 6% respectively, P 0.0001 for both comparisons). To quantify the effect of population substructure we calculated Wright s fixation index for each of the variable sites, comparing the level of heterozygosity observed within each of the two subpopulations to that observed in the total data set. 16 This gave values of F ST ranging from zero to 0.016, with an average of 0.007 across all 12 polymorphisms. From these genotypic data it is not possible to determine exact values for linkage disequilibrium, but an impression of strongly linked alleles may be gained by simple pairwise comparison of individual loci. This is illustrated in Table 3, which gives the odds ratio of carrying the rare allele (in either homozygous or heterozygous state) at locus 1 if an individual carries the rare allele (in either homozygous or heterozygous state) at locus 2. In the table, the shaded cells highlight those pairwise comparisons which give P 0.01 by chi-squared test. A very similar pattern of association emerges in both Gambian and Malawian populations. The TNF 1031C, 863A, 376A, 238A, +851G, and +1304G alleles appear to be linked to each other in both populations. Additionally, TNF 244A appears to be linked to TNF 376A, 238A, +851G and +1304G in Malawi, and TNF 1073T appears linked to 857T in The Gambia. Table 3 Odds ratio of carrying the rare allele (in either homozygous or heterozygous state) at one locus if an individual carries the rare allele (in either homozygous or heterozygous state) at the other locus. An asterisk denotes too few observations to infer an odds ratio. Shaded cells highlight those pairwise comparisons which give P 0.01 by 2 test

346 Discussion Nucleotide diversity of the TNF gene region In this systematic survey of 72 chromosomes sampled in the Gambian village of Sukuta, we found an overall nucleotide diversity ( ) of 5.6 10 4 in the TNF gene region. This compares to the average value of 5.3 10 4 obtained by Cargill et al 13 among 106 genes distributed throughout the genome, and 8.5 10 4 obtained by Halushka et al 17 among 75 genes. In both of these studies, the various ethnic groups from around the world were sampled. Thus the level of nucleotide diversity at the TNF locus within a single African village is comparable to the global value for human autosomal genes across different continents. These data do not allow us to draw firm conclusions about the relative nucleotide diversity of TNF compared to other loci, or about the overall level of genetic diversity in Sukuta compared to other communities, and systematic surveys of multiple gene loci in other well-defined communities are needed to address these issues. However an emerging body of data suggests that many of the common variable sites in the human genome may be found within a single African community, unless it happens to derive from a small number of founders and is highly inbred, and our present findings would be highly consistent with this notion. In this Gambian study population, we found a significantly higher frequency of polymorphic sites in the TNF promoter region compared to the TNF coding region (8 per 1256 nt vs 0 per 873 nt, P = 0.01). The corresponding values of are 13.1 10 4 for the promoter region and 0 for the coding region, which is a wider difference than observed in most other gene regions: Cargill et al 13 report average values of 5.43 10 4 and 5.30 10 4 for noncoding and coding regions respectively. However approximately 10% of genes studied by these authors had a nucleotide diversity of 14 10 4 in the non-coding region, while over 5% had no observed nucleotide diversity in the coding region. Taken together, it would seem that the TNF promoter region may be somewhat more variable and the coding region rather less variable than average, but both are within the range of values observed elsewhere in the genome. All of the polymorphisms that we identified in the Gambian population were also found in Malawians, with the exception of a rare TNF+943A allele which was present in only one of the 576 individuals that we analysed. Malawi and The Gambia are geographically remote and their populations are known to be genetically different, 18 so it is not surprising that eight of the 12 TNF polymorphisms are at significantly different frequency in the two populations. However the level of population differentiation was relatively low when assessed by Wright s fixation index, with an F ST value of 0.007 averaged across the 12 variable sites, compared to values in the region of 0.07 for a variety of other genes compared across different populations worldwide (Goddard et al 19 and G McVean, personal communication). Further data from other African and non-african populations may shed light on the demographic history of the TNF locus, including the notion that specific TNF alleles may confer selective advantage or disadvantage in relation to malaria and other infectious diseases. A growing number of disease associations have been reported for polymorphisms in the TNF promoter region. These include associations between infectious diseases and the TNF 308A allele, the TNF 238A allele and the TNF 376A allele. 2 8 To give a few examples of associations with non-infectious diseases, asthma has been associated with the TNF 308A allele, 9 multiple sclerosis with the TNF 376A allele, 10 rheumatoid joint damage with TNF 238G homozygotes, and Crohn s disease with the TNF 1031C allele. 11,20 There has been some debate as to the functional significance of these associations: for example, as the TNF 308A allele has been reported to increase transcriptional activation in some experimental systems 21,22 but not others. 23,24 We have found that the TNF 376A allele acts to recruit the transcription factor OCT-1 to this part of the TNF promoter region 7 and this appears to be associated with a modest increase in basal gene expression in the human monocyte line MonoMac6. Allele-specific binding of OCT-1 has also been observed to the TNF- 857A allele. 25 We have recently noted that the TNF 863A allele acts to reduce the binding of NF- B p50/p50, with little effect on p65/p50 binding, to this part of the TNF promoter region. 26 The effect appears to be to increase inducible TNF expression in primary human monocytes: this is consistent with observations by some investigators 27 but not others. 28,29 It remains to be determined whether these different observations are true inconsistencies, or simply reflect the true functional heterogeneity of varied cell types under different conditions of stimulation. The uncertain state of the functional evidence means that, if the above disease associations are replicated in independent studies, investigators must seriously consider the possibility that they relate to functional polymorphisms outside the TNF promoter region. In the case of disease associations within Caucasian populations, the true effect might turn out to map to a distant part of the MHC because of the strong linkage disequilibrium that exists across this region in Caucasians. The main import of the present study is to show that, at least in Africans, it is unlikely that the disease associations with TNF promoter polymorphisms can be ascribed to functional variation elsewhere in the TNF gene, except perhaps to two new intronic polymorphisms that we have identified at +851 and +1304 nt, that have allele frequencies as high as 17% in the Malawian population. We are currently carrying out a family analysis of polymorphisms of TNF and adjacent genes, with the goal of determining the precise extent of linkage disequilibrium at the TNF locus in Africans, and of building a detailed haplotypic map that will allow disease associations at this locus to be analysed with greater confidence. Methods Subjects To screen for novel polymorphisms, blood samples were obtained from 36 healthy individuals aged 2 to 44 years, all residents of the peri-urban village of Sukuta in the Western division of The Gambia. A sample size of 36 individuals, ie 72 chromosomes, gives 95% probability of detecting an allele of 4% frequency. To estimate allele frequencies, blood samples were obtained from 424 unrelated healthy Gambian adults and 121 unrelated healthy Malawian adults. All subjects gave informed consent and the study was approved by the Gambia Govern-

ment/medical Research Council Joint Ethical Committee and the College of Medicine Research Committee of the University of Malawi. Sequencing The first round of PCR amplification used genomic DNA (100 ng), oligonucleotide primers (0.5 M), dntps (0.8 mm), MgCl 2 (2.5 or 3.75 mm) and AmpliTaq Gold (1.25 U; Perkin Elmer, CA, USA) in 20 l PCR buffer (Perkin Elmer). Primer pairs were: 1389 (5 AGG CTG ACC AAG AGA GAA AG 3 ) and 484 (5 TTG AGT CCT GAG GCC TGT GT 3 ); 952 (5 GTT ACA GGA GAC CTC TGG GG 3 ) and +125 (5 GGA AGA GAA CCT GCC TGG C 3 ); 88 (5 AGGAAGTTTTCCGCTGGTTG3 ) and+1565 (5 TCA GCT TGA GGG TTT GCT GG 3 ); +1398 (5 ATA CTC AGA ACG TCA TGG CC 3 ) and+3004 (5 GAG TTG GAA ATT CCC ATG CC 3 ). An MJ Tetrad was programmed as follows: 95 C for 12 minutes; followed by 35 cycles of 94 C for 30 s, 58 C for 45 s and 70 C for1min; finishing with 10 min at 72 C before holding at 15 C. The second round of PCR amplification used firstround PCR product diluted 1:10 in distilled water (1 l), primers (0.5 M), dntps (0.8 mm), MgCl 2 (3 mm) and AmpliTaq Gold (1.25 U; Perkin Elmer) in 30 l PCR buffer (Perkin Elmer). Primer pairs were: 1389 (as above but with the M13 21 sequence added at the 5 end) and 900 (5 M13REV GAC ATT CTC CTA CCC ATT GC 3 ); 952 ( 21M13 5 GTT ACA GGA GAC CTC TGG GG 3 ) and 484 (as above but with the M13REV sequence added at the 5 end); 352 ( 21M13 5 TCC CCA AAA GAA ATG GAG GC 3 ) and +125 (as above but with the M13REV sequence added at the 5 end); 562 ( 21M13 5 TTT CCT GAG GCC TCA AGC CT 3 ) and 188 (M13REV 5 GTT GGG GAC ACA CAA GCA TC 3 ); 88 ( 21M13 5 AGG AAG TTT TCC GCT GGT TG 3 ) and +522 (M13REV 5 TCT CTT GCG TCT CCA TTT CC 3 ); +279 ( 21M13 5 TCT TCT CCT TCC TGA TCG TG 3 ) and +846 (M13REV 5 TGT GTA TTC ACC TTC CAG GC 3 ); +773 ( 21M13 5 AAG AAG ATA GGG TGT CTG GC 3 ) and +1329 (M13REV 5 AAG TTC TGC CTA CCA TCA GC 3 ); +1239 ( 21M13 5 CAT GTT GTA GGT AAG AGC TC 3 ) and +1565 (M13REV 5 TCA GCT TGA GGG TTT GCT GG 3 ); +1398 ( 21M13 5 ATA CTC AGA ACG TCA TGG CC 3 ) and +1769 (M13REV 5 ACC TTG GTC TGG TAG GAG AC 3 ); +1661 ( 21M13 5 TGT ACC TCA TCT ACT CCC AG 3 ) and +2240 (M13REV 5 TCA GGG ATC AAA GCT GTA GG 3 ); +2137 ( 21M13 5 TGG GAT TCA GGA ATG TGT GG 3 ) and +2676 (M13REV 5 CAG TTG GTC ACC AAA TCA GC 3 ); +2482 ( 21M13 5 ATT TGG GAG ACC GGG GTA TC 3 ) and +3004 (M13REV 5 GAG TTG GAA ATT CCC ATG CC 3 ). An MJ Tetrad was programmed as follows: 95 C for 12 min; followed by 5 cycles of 94 C for 30 s, 57 C for 30 s and 72 C for 45 s; then 30 cycles of 94 C for 1 min, 72 C for 1 min; finishing with 10 min at 72 C before holding at 15 C. DNA sequencing was performed on an ABI377 automated sequencer using M13 Dye primer and M13 Big dye primer sequencing kits (Perkin Elmer) and analysed using Factura and Sequence Navigator. Genotyping The amplification refractory mutation system (ARMS) method was used to genotype the polymorphisms identified at the previous stage of analysis. Primer sequences are given in Table 1. PCR amplification was performed Nucleotide diversity of the TNF gene region in 15 l reaction volume containing approximately 15 ng genomic DNA plus: 0.4 mm total dntps; 16.6 mm ammonium sulphate; 1.9 mm magnesium chloride; 67.9 mm Tris-base (ph 8.9); 0.1% v/v Tween 20; 20 ng of each allele specific and consensus primer; 5 ng of each positive control primer; 0.25 units of Bioline Taq. An MJ Tetrad PCR was programmed as follows: 96 C for 1 min; five cycles of 96 C for 35 s, 70 C for 45 s, 72 C for 35 s; 21 cycles of 96 C for 25 s, 65 C for 50 s, 72 C for 40 s; six cycles of 96 C for 35 s, 55 C for 1 min, 72 C for 1.5 min and hold at 15 C. The accuracy of the ARMS PCR method was tested on sequenced individuals before extending it to DNA samples of unknown genotype. Acknowledgements We thank the inhabitants of Sukuta for making this study possible, and Simon Correa, Idi Sambou, Yaya Dibba and Isatou Drammeh for excellent technical help and field work. FS-J is a WHO Training Fellow. References 1 Complete sequence and gene map of a human major histocompatibility complex. The MHC sequencing consortium [In Process Citation]. Nature 1999; 401: 921 923. 2 McGuire W, Hill AV, Allsopp CE, Greenwood BM, Kwiatkowski D. Variation in the TNF-alpha promoter region associated with susceptibility to cerebral malaria. Nature 1994; 371: 508 510. 3 Cabrera M, Shaw MA, Sharples CJ et al. Polymorphism in tumor necrosis factor genes associated with mucocutaneous leishmaniasis. J Exp Med 1995; 182: 1259 1264. 4 Nadel S, Newport MJ, Booy R, Levin M. Variation in the tumor necrosis factor-alpha gene promoter region may be associated with death from meningococcal disease. J Infect Dis 1996; 174: 878 880. 5 Conway DJ, Holland MJ, Bailey RL et al. Scarring trachoma is associated with polymorphism in the tumor necrosis factor alpha (TNF-alpha) gene promoter and with elevated TNF-alpha levels in tear fluid. Infect Immun 1997; 65: 1003 1006. 6 McGuire W, Knight JC, Hill AV, Allsopp CE, Greenwood BM, Kwiatkowski D. Severe malarial anemia and cerebral malaria are associated with different tumor necrosis factor promoter alleles. J Infect Dis 1999; 179: 287 290. 7 Knight JC, Udalova I, Hill AV et al. A polymorphism that affects OCT-1 binding to the TNF promoter region is associated with severe malaria. Nat Genet 1999; 22: 145 150. 8 Knight JC, Kwiatkowski D. Inherited variability of tumor necrosis factor production and susceptibility to infectious disease. Proc Assoc Am Physicians 1999; 111: 290 298. 9 Moffatt MF, Cookson WO. Tumour necrosis factor haplotypes and asthma. Hum Mol Genet 1997; 6: 551 554. 10 Fernandez-Arquero M, Arroyo R, Rubio A et al. Primary association of a TNF gene polymorphism with susceptibility to multiple sclerosis. Neurology 1999; 53: 1361 1363. 11 Negoro K, Kinouchi Y, Hiwatashi N et al. Crohn s disease is associated with novel polymorphisms in the 5 -flanking region of the tumor necrosis factor gene. Gastroenterology 1999; 117: 1062 1068. 12 Westendorp RG, Langermans JA, Huizinga TW et al. Genetic influence on cytokine production and fatal meningococcal disease [published erratum appears in Lancet 1997; 349: 656]. Lancet 1997; 349: 170 173. 13 Cargill M, Altshuler D, Ireland J et al. Characterization of singlenucleotide polymorphisms in coding regions of human genes. Nat Genet 1999; 22: 231 238. 14 Hill AV, Allsopp CE, Kwiatkowski D et al. Common West African HLA antigens are associated with protection from severe malaria. Nature 1991; 352: 595 600. 347

348 Nucleotide diversity of the TNF gene region 15 Nedospasov SA, Shakhov AN, Turetskaya RL et al. Tandem arrangement of genes coding for tumor necrosis factor (TNFalpha) and lymphotoxin (TNF-beta) in the human genome. Cold Spring Harb Symp Quant Biol 1986; 51: 611 624. 16 Wright S. Systems of mating. Genetics 1921; 6: 111 178. 17 Halushka MK, Fan JB, Bentley K et al. Patterns of single-nucleotide polymorphisms in candidate genes for blood-pressure homeostasis. Nat Genet 1999; 22: 239 247. 18 Hill AV, Allsopp CE, Kwiatkowski D et al. Extensive genetic diversity in the HLA class II region of Africans, with a focally predominant allele, DRB1*1304. Proc Natl Acad Sci USA 1992; 89: 2277 2281. 19 Goddard KA, Hopkins PJ, Hall JM, Witte JS. Linkage disequilibrium and allele-frequency distributions for 114 single-nucleotide polymorphisms in five populations. Am J Hum Genet 2000; 66: 216 234. 20 Kawasaki A, Tsuchiya N, Hagiwara K, Takazoe M, Tokunaga K. Independent contribution of HLA-DRB1 and TNF alpha promoter polymorphisms to the susceptibility to Crohn s disease. Genes Immun 2000; 1: 351 357. 21 Wilson AG, Symons JA, McDowell TL, McDevitt HO, Duff GW. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci USA 1997; 94: 3195 3199. 22 Abraham LJ, Kroeger KM. Impact of the 308 TNF promoter polymorphism on the transcriptional regulation of the TNF gene: relevance to disease. J Leukoc Biol 1999; 66: 562 566. 23 Brinkman BM, Zuijdeest D, Kaijzel EL, Breedveld FC, Verweij CL. Relevance of the tumor necrosis factor alpha (TNF alpha) 308 promoter polymorphism in TNF alpha gene regulation. J Inflamm 1995; 46: 32 41. 24 Stuber F, Udalova IA, Book M et al. 308 tumor necrosis factor (TNF) polymorphism is not associated with survival in severe sepsis and is unrelated to lipopolysaccharide inducibility of the human TNF promoter. J Inflamm 1996; 46: 42 50. 25 Hohjoh H, Tokunaga K. Allele-specific binding of the ubiquitous transcription factor OCT-1 to the functional single nucleotide polymorphism (SNP) sites in the tumor necrosis factor-alpha gene (TNFA) promoter. Genes Immun 2001; 2: 105 109. 26 Udalova IA, Richardson A, Denys A et al. Functional consequences of a polymorphism affecting NF-kappaB p50-p50 binding to the TNF promoter region. Mol Cell Biol 2000; 20: 9113 9119. 27 Higuchi T, Seki N, Kamizono S et al. Polymorphism of the 5 flanking region of the human tumor necrosis factor (TNF)-alpha gene in Japanese. Tissue Antigens 1998; 51: 605 612. 28 Skoog T, van t Hooft FM, Kallin B et al. A common functional polymorphism (C A substitution at position 863) in the promoter region of the tumour necrosis factor-alpha (TNF-alpha) gene associated with reduced circulating levels of TNF-alpha. Hum Mol Genet 1999; 8: 1443 1449. 29 Uglialoro AM, Turbay D, Pesavento PA et al. Identification of three new single nucleotide polymorphisms in the human tumor necrosis factor-alpha gene promoter. Tissue Antigens 1998; 52: 359 367.