Association of tumour necrosis factor-alpha -308 G/A polymorphism with primary open-angle glaucoma


  • Conflict/competing interest: None.

  • Funding sources: This project was supported by grants from TUBITAK (The Scientific and Technological Research Council of Turkey) – SBAG-HD-152 (106S226).

Associate Professor Banu Bozkurt, Selcuk University, Meram Medical Faculty, Department of Ophthalmology, Konya 42080, Turkey. Email:


Background:  Tumour necrosis factor-alpha (TNF-α) is an important proinflammatory cytokine driving axonal degeneration and retinal ganglion cell apoptosis in glaucoma. The aim of the study was to evaluate the association of TNF-α -308 G/A and -238 G/A polymorphisms with primary open-angle glaucoma (POAG).

Design:  A prospective, case–control study, university hospital setting.

Participants:  Eighty-six POAG patients and 193 healthy unrelated controls.

Methods:  TNF-α polymorphisms were screened by using direct gene sequencing.

Main Outcome Measures:  Frequency of TNF-α -308 G/A and TNF-α -238 G/A promoter polymorphisms in glaucoma and healthy subjects.

Results:  The frequencies of TNF-α -308 GA genotype and ‘A’ allele were higher in patients with POAG (22.1% and 12.2%, respectively) in comparison with the control group (10.9% and 6%, respectively) (P = 0.046 and 0.02, respectively), with odds ratios of 2.45 (P = 0.01, 95% CI = 1.23–4.87) and 2.19 (P = 0.013, 95% CI = 1.18–4.08), respectively. Genotype distribution of the TNF-α -238 variants did not yield a statistically significant difference between the two groups (P = 0.87).

Conclusion:  TNF-α -308 G/A polymorphism seems to be associated with POAG in Turkish population. However, population-based studies with large number of subjects and long-term follow-up are needed to verify the association of TNF-α -308 G/A polymorphism with glaucoma susceptibility.


Glaucoma is classified as a neurodegenerative optic neuropathy and the most prevalent type in Caucasian populations is primary open-angle glaucoma (POAG), in which elevated intraocular pressure (IOP) is the major risk factor.1–3 The pathophysiological mechanisms by which elevated IOP causes the characteristic optic neuropathy may include mechanical compression of the cribriform plates of the lamina cribrosa on the axons of retinal ganglion cell (RGC), pressure-induced ischaemia, local cellular response mechanisms and apoptosis.4–8 In cases of elevated IOP or ischaemia, some of the damage is caused immediately by the insult itself and some is secondarily caused by changing the neuronal environment, which in turn increases the vulnerability of spared neurons even if the primary causative factor is no longer present.9 Immune system is the key modulator in deciding whether the RGCs will survive or die and neuroprotection of RGCs has been emphasized an important goal for managing glaucoma.

Tumour necrosis factor-alpha (TNF-α, OMIM 191160) is a cytokine that plays a pivotal role in inflammation, immunity, haematopoiesis and apoptosis.10,11 It is a potent activator of neurotoxic substances such as nitric oxide (NO) and excitotoxins and binding of TNF-α to its receptor leads to the activation of the caspase system and apoptosis. Post-mortem histopathological studies, experimental glaucoma models and cell culture data have suggested that TNF-α plays an important role in glaucomatous degeneration and progression.12–17 Histopathological sections of the glaucomatous optic nerve head (ONH) and retina revealed an increased expression of mRNA of TNF-α and TNF-α receptor-1, parallel with the progression of ONH damage.12,13,15 As RGCs selectively express TNF-α receptor-1, they are the targets for TNF-α-induced neurodegeneration. Furthermore, a cell culture study showed that glial cells exposed to elevated hydrostatic pressure or stimulated ischaemia secreted increased amounts of TNF-α, subsequently leading to apoptotic death of co-cultured RGCs.14 The importance of TNF-α in glaucoma is further strengthened by a study, which showed that anti-TNF-α neutralizing antibody or deleting the genes encoding TNF-α or its receptor blocked the deleterious effects of elevated IOP.17

The TNF-α gene is located within the highly polymorphic major histocompatibility complex class III region on chromosome 6p21.3. There are many microsatellites and polymorphic areas in the promoter region of TNF-α gene.18–21 TNF-α -308 G/A and -238 G/A single nucleotide polymorphisms are the most common genetic variations seen in the Caucasian population. TNF-α -308 G/A (rs1800629) is characterized by a G to A substitution at position -308 and TNF-α -238 G/A (rs361525) is characterized by a G to A substitution at position -238. TNF-α -308 A allele has been reported to increase transcriptional activation in some experimental systems.20,21 Lipopolysaccharide-stimulated whole blood cell cultures as well as peripheral blood mononuclear cells stimulated with anti-CD3 and anti-CD28 monoclonal antibodies from subjects carrying the TNF-α -308 GA genotype showed a significant increase in TNF-α production compared with individuals carrying the TNF-α -308 GG genotype.22,23 The -238 G/A polymorphism is not known to be of functional significance, but is sited close to a putative repressor site.24

Genetic studies of TNF-α polymorphisms and glaucoma have produced conflicting results. Some studies showed that TNF-α -308 A allele is strongly associated with POAG and pseudoexfoliative glaucoma (PEXG),25–27 whereas others have failed to detect such an association in both glaucoma types.28–30 In the present study, we sought to examine whether TNF-α -308 G/A and TNF-α -238 G/A promoter polymorphisms were associated with increased risk of POAG development in native Turkish population.


A case–control study was performed to determine plausible association between TNF-α polymorphisms and POAG. The procedures used in this study conformed to the tenets of the Declaration of Helsinki and the Ethics Committee of Hacettepe University Faculty of Medicine approved the research. Informed consent was obtained from the participating individuals following an explanation of the nature and possible consequences of the study.

Native Turkish subjects with POAG underwent a complete eye examination including visual acuity testing, slit-lamp biomicroscopy, fundus examination with 90-diopter (D) lens, IOP measurement with Goldmann applanation tonometry at least for three times in different hours, central corneal thickness (CCT) measurement with ultrasonic pachymetry and gonioscopy. Visual field assessment was performed with a Humphrey Field Analyzer, using the 30-2/24-2 Fastpac or SITA strategy (Zeiss Humphrey Systems, Dublin, CA, USA). Glaucoma was assessed by the consensus of two glaucoma specialists. Patients were classified as having POAG, when they had glaucomatous ONH or retinal nerve fibre layer structural abnormalities and/or visual field damage consistent with retinal nerve fibre layer damage, visual field loss in the upper hemifield that is different compared with the lower hemifield (abnormal glaucoma hemifield test), not explained by any other disease and open-angle at gonioscopy.

A control group of age- and sex-matched, native Turkish individuals was chosen randomly from a sample of patients admitted to the ophthalmology clinic for refractive errors, senile cataracts, routine ophthalmic examinations or medical staff with no ocular problems. They had no morphological or functional damage indicative for primary or secondary open-angle or angle-closure glaucoma.

DNA isolation and genotyping

Genomic DNA was isolated from 200 µL of peripheral blood using a QIA-Amp blood mini kit (Qiagen, Hilden, Germany). PCR products were amplified using the ABI 2× PCR Master Mix, with 50 ng of genomic DNA and appropriate amplification primer pairs to a final concentration of 5 ng/µL. The forward primer was 5′-CCT CCC AGT TCT AGT TCT AT-3′ and reverse primer was 5′-TTC TGT CTC GGT TTC TTC T-3′. PCR reactions followed a standard PCR program: 1 cycle of 94°C for 5 min; 32 cycles of 94°C for 30 s, 60°C for 30 s and 72°C for 30 s; a final elongation step of 72°C for 5 min. The size of the PCR product was 284 base pair. For optimal results, the PCR products were purified from the PCR cocktail with QIA Quick PCR Purification Kit (Qiagen, Hilden, Germany). The purified PCR products were directly sequenced on both strands on a ABI 310 Capillary Electrophoresis (ABI PRISM 310 Genetic Analyzer, Perkin Elmer Applied Biosystems Division, Foster City, CA, USA) using the Big Dye Terminator Cycle Sequencing Kit version 3.1 (Perkin Elmer Applied Biosystems Division, Foster City, CA, USA) according to the manufacturer's protocol. One investigator (LM), who was masked to the diagnosis of the participants being studied, performed the molecular analyses and evaluated the sequence electropherograms (Fig. 1).

Figure 1.

Sequence electropherograms (reverse strand) obtained from tumour necrosis factor-α promoter region showing polymorphisms at nucleotide positions -308 and -238. Blue, black, green and red dyes correspond to C, G, A and T nucleotides, respectively. (a) The arrow on the right side indicating a blue (C) peak at position -308 (GG genotype), and the arrow on the left side again showing a blue (C) peak at position -238 (GG genotype). (b) The arrow on the right side at position -308 indicating a nucleotide change of C→N, with both blue (C) and red (T) peaks (GA genotype), and the arrow on the left side indicating a blue (C) peak at position -238 (GG genotype). (c) The arrow on the right side at position -308 indicating a nucleotide change of C→T with one red peak (T) (AA genotype) and the arrow on the left side indicating a blue (C) peak at position -238 (GG genotype).

Statistical analysis

Frequencies of TNF-α -308 G/A and TNF-α -238 G/A allele and genotype frequencies between controls and POAG patients were compared by χ2-test or Fisher's exact test. The odds ratio (OR) and 95% confidence intervals (CI) were also calculated using logistic regression analysis. The statistical package of spss version 11.5 for Windows (SPSS, Inc, Chicago, IL, USA) was used in the analysis of the data and P < 0.05 was considered as statistically significant.


There were 86 POAG patients (35 men, 51 women) with a mean age of 65.1 ± 10.2 years and 193 healthy controls (74 men, 119 women) with a mean age of 66.2 ± 7.8 years. There were no differences between the two groups according to age or gender (P = 0.2 and 0.8, respectively).

In glaucoma group, mean IOP was 27.6 ± 3.9 mmHg within a range between 22 and 39 mmHg and cup/disc (C/D) ratio was 0.7 ± 0.2. The mean CCT value was 546 ± 25.6 µm within a range between 484 and 614 µm. Mean values of MD and PSD were −12.2 ± 8.1 dB (between −4.1 and −29.5 dB) and 5.8 ± 4 dB (between 1.3 and 13.8 dB).

The frequencies of genotypes and alleles for TNF-α -308 G/A and -238 G/A polymorphisms are provided in detail in Table 1. The observed genotypes did not deviate from those predicted by the Hardy–Weinberg equilibrium in either the cases or the controls (P > 0.05). TNF-α -308 GA genotype was higher in patients with POAG (22.1%) in comparison with the control group (10.9%) (P = 0.046), with OR of 2.45 (P = 0.01, 95% CI = 1.23–4.87). TNF-α -308 A allele was higher in POAG patients (12.2%) compared with the control group (6%) (P = 0.02) and increased the risk of POAG with an OR of 2.19 (P = 0.013, 95% CI = 1.18–4.08). Genotype distribution of the TNF-α -238 variants did not yield a statistically significant difference between the two groups (P = 0.87). Homozygote variant A/A genotype of -238 polymorphism was not observed in any group.

Table 1.  Comparison of TNF-α -308 and TNF-α -238 genotypes and alleles in POAG and controls
Genotype and allele frequencies Controls n = 193POAG n = 86P-value
  1. POAG, primary open-angle glaucoma; TNF, tumour necrosis factor.

TNF-α -308 G/A polymorphismGG171 (88.6%)66 (76.7%)P = 0.046
GA21 (10.9%)19 (22.1%)
AA1 (0.5%)1 (1.2%)
G allele363 (94%)151 (87.8%)P = 0.018
A allele23 (6%)21 (12.2%)
TNF-α -238 G/A polymorphismGG180 (93.3%)79 (91.9%)P = 0.87
GA13 (6.7%)7 (8.1%)
G allele373 (96.6%)165 (95.9%)P = 0.87
A allele13 (3.4%)7 (4.1%)


Glaucoma is defined as a neurodegenerative disease and several immunological mechanisms have been shown to be involved in the pathogenesis.16 There is increasing evidence that supports the substantial role of the TNF- α/TNFRI pathway in the destruction of the optic nerve in POAG.31 TNF-α activates the transcription factor NF-κB through binding to the high affinity TNF receptor-2, which in turn mediates the expression of a wide range of genes essential for neuronal survival.32 However, it can also serve as a neurodegenerative factor when it binds to the low affinity death receptor TNF-R1 and induces the mitochondria-mediated apoptotic pathway. There is a delicate balance between these two pathways and any shift in equilibrium might have deleterious effects. Genetically determined higher production of TNF-α in an individual carrying TNF-α -308 A allele might create a more inflammatory environment in the eye and lead to RGC death, glaucomatous degeneration and disease progression in stress conditions.

In our study, the distribution of genotypes and alleles in the TNF-α -308 gene promoter region differed significantly between the two groups, with the GA genotype and the ‘A’ allele being more common in POAG patients than in the control subjects. All of our subjects had IOP measurements greater than 21 mmHg; therefore, TNF-α -308 polymorphism seems to contribute to optic neuropathy as a high IOP related genetic factor. A similar substitution at the -238 position of the TNF-α gene promoter was also examined and no differences in genotype and allele frequencies could be found between POAG and control groups.

The first study that evaluated the association of TNF-α -308 polymorphism with POAG was done by Lin et al.25 In that study, allele ‘A’ of TNF-α -308 polymorphism was found to occur more frequently in Chinese patients with POAG (42.5%) in comparison with the control group (21.4%). AA genotype was found in 31.7% of POAG patients and 6.8% of healthy subjects, which was significantly different (P < 0.05) (OR = 6.26, 95% CI = 2.5–15.56). In the study of Khan et al.26 TNF-α polymorphism G-308A is found to be strongly associated with PEXG in the Pakistani population. The GG wild-type genotype was found at a frequency of 87% in the controls and 43% in the PEXG patients. The GA heterozygous genotype was present in 32% of the PEXG patients, whereas only 10% of the controls were heterozygous. Of the PEXG patients, 25% contained the AA variant genotype, whereas in the controls this genotype was present at only 2%. Recently, Razeghinejad et al.27 showed a marked association between the TNF-α -308 G/A polymorphism and the risk of glaucoma development either when all cases with OAG were compared with controls (P = 0.0003) or when they were categorized as POAG (P = 0.001; OR = 4.33, 95% CI = 1.67–11.34) or PEXG (P = 0.001; OR = 4.35, 95% CI = 1.6–11.88) in the Iranian population.

However, some studies could not find an association between TNF-α -308 A allele and POAG and PEXG.28–30 Tekeli et al.29 unexpectedly showed a higher frequency of the common G allele (low producer allele of TNF-α) in Turkish subjects with PEXG. ‘A’ allele was found in 3.2% of PEXG group and 8.2% in control group (P = 0.023) and -308G/A variant was reported to be a possible protective factor against PEXG. This finding is completely in conflict with the findings of other studies, which indicated that the presence of the TNF-α -308 A allele (higher producer allele of TNF-α) is a susceptibility factor for both POAG and PEXG.25–27 In the study of Mossböck et al.30 no significant differences in either genotype distribution or allelic frequencies of the TNF-α -308 G/A and the TNF-α -238 G/A polymorphisms were found between patients with PEXG (frequency of G alleles; 0.86 and 0.94, respectively) and control subjects (frequency of G alleles; 0.86 and 0.96, respectively) in an Austrian population. In their previous study,28 genotype distribution and allelic frequencies of the TNF-α -308 G/A and TNF-α -238 G/A polymorphisms were also similar between POAG and control subjects.

TNF-α -308 G/A polymorphism is rare in the Japanese population in comparison with Caucasians. Funayama et al.33 found that the TNF-α -308 GA genotype frequency was 1% in Japanese patients with POAG, 2.8% in normal tension glaucoma patients and 2.8% in healthy controls, whereas A/A was not found in any of the patients or controls. In contrary to TNF-α -308 polymorphism, POAG patients carrying both TNF-α−857T and optineurin/412A were shown to have significantly worse visual field scores than patients with only TNF-α -857T polymorphism, and POAG patients carrying TNF-α−863A polymorphism had worse visual field scores if they also had optineurin 603A polymorphism. They concluded that there is a possible interaction between optineurin and TNF-α polymorphisms, which could increase the risk of glaucoma.

The main problem in identifying the gene variants associated with susceptibility to common diseases is that the observed results are not replicated in subsequent studies that used different populations and/or larger numbers of cases versus controls.34 Four factors seem to cause confusion in the field; (i) different approaches for genetic investigation; (ii) inappropriate technology and statistical method for each genetical approach; (iii) inconsistent standards for interpreting levels of statistical evidence; and (iv) non-standardized strategies for choosing and evaluating phenotypes. A major limitation of the current study is that the sample sizes of our patients and controls are limited to represent the whole population, which makes it difficult to assess more complicated models and fully explore possible confounders. For most common diseases, increasing the sample size in a study is a crucial step in achieving statistically significant genetic mapping results. Another reason for the differences between the studies is the genetic differences between studied populations, as the prevalence of genetic polymorphisms is different in various ethnic populations. The percentage of TNF-α -308 polymorphism in our control group was less than those reported for Austrian28,30 and British populations,35 which means that TNF-α -308 A allele might be underrepresented in the control group and might lead to a remarkable difference in the genotype and allele distributions when compared with the POAG group. It might also be explained by different genetic backgrounds of Turkish population compared with those of Caucasians used in the study of Mossböck et al.28,30 In this study, we used direct gene sequencing to detect the transition of ‘G’ to ‘A’ in TNF-α -308 and -238 gene promoter regions, whereas restriction fragment length polymorphism agarose gel electrophoresis was used in previous studies.

Our results showed an association between TNF-α -308 G/A polymorphism and POAG in Turkish patients, whereas TNF-α -238 polymorphism does not have any great influence on the overall susceptibility to POAG. As we do not know the functional effect of the polymorphism in these subjects, further investigation should be done to determine the effect of TNF-α -308 G/A polymorphism on TNF-α production. In case of high IOP, genetically determined higher release of TNF-α might probably amplify the destructive immune process in the RGC and lead to progression of the disease. In those subjects, inhibition of TNF-α might be helpful in preventing neuronal degeneration in glaucoma patients. However, population-based studies with large number of subjects and long-term follow-up are needed to verify the association of TNF-α -308 promoter polymorphism with glaucoma susceptibility and severity.


This project was supported by grants from TUBITAK (The Scientific and Technological Research Council of Turkey) -SBAG-HD-152 (106S226).