Open Access

ARTICLE

Relation between the tumor necrosis factor-α (TNF-α) gene and protein expression, and clinical, biochemical, and genetic markers: age, body mass index and uric acid are independent predictors for an elevated TNF-α plasma level in a complex risk model

Susanne Schulz^1,2, Undraga Schagdarsurengin¹, Thomas Suss^1,2, Ursula Müller-Werdan², Karl Werdan², Christiane Gläser¹

1 Institute of Human Genetics and Medical Biology
2 Department of Medicine III, Martin-Luther-Universität Halle-Wittenberg

* Corresponding Author: Christiane Gläser,

European Cytokine Network 2004, 15(2), 105-111.

Accepted 06 April 2004;

Abstract

Background: Tumor necrosis factor-alpha (TNF-α) has been implicated in the pathogenesis of numerous complex diseases. The plasma level of this pro-inflammatory cytokine is associated with a variety of different risk factors, but little is known about the genetic background and the complex interactions. Methods: in this clinical study, correlations were studied between plasma levels of circulating TNF-α protein (ELISA), its mRNA expression in monocytes (RT-PCR) and genetic variants of TNF-α gene (SSCP), with several diseases, including obesity, atherosclerosis, diabetes mellitus, hypertension, as well as risk factors such as age, gender, inflammatory markers, the coagulation/fibrinolysis balance, and lipid metabolism. One hundred and ninety four clinically and biochemically well-characterized patients were enrolled. Results: At the transcriptional level, measured in monocytes, no association with any clinical or biochemical parameter investigated was found, including TNF-α protein level. Investigating the influence of genetic variants of the TNF-α gene on mRNA and protein levels, only one promoter polymorphism, namely c.-238G > A, was shown to be associated with transcriptional but not with translational expression. However, at the translational level, significant positive, but weak associations were determined for obesity (P = 0.037), age (P = 0.038), uric acid (P < 0.001), body mass index (P = 0.01), plasminogen (P = 0.013), and fibrinogen (P = 0.002) in bivariate regression analyses, whereas HDL-cholesterol (P = 0.005) was shown to be negatively correlated. However, investigating confounding effects in stepwise multivariate regression analysis, body mass index (P = 0.009), uric acid (P = 0.026) and age (P = 0.037) turned out to be significantly associated with plasma levels of circulating TNF-α (adjusted R² = 0.117; SE: 0.688).

---

Keywords

---