The danger of extrapolation: humans and rodents differ in response to PCBs.

Although awareness of the extreme toxicity of polychlorinated biphenyls (PCBs) led to their 1976 ban in the United States, these chemicals' virtually indestructible molecular nature and environmental ubiquity make them a continuing serious health threat. Much of what we know about the health effects of PCBs comes from rodent studies, and current thinking in toxicology and risk analysis is based on the assumption that results in animal models will extrapolate to humans. This month, Michelle M. Tabb of the University of California, Irvine, and colleagues report on an investigation comparing human biological response to PCB exposure to that of rodents [EHP 112:163-169]. Their findings indicate there are significant physiological differences between humans and rats that have important implications for risk assessment.

Highly stable and flame-resistant, PCBs have long been used in the manufacture of industrial products such as electrical insulation, machinery lubricants, plastics, wood products, paints, inks, and agricultural pesticides. Although PCBs are no longer made in the United States, their long-term, widespread use and persistence in the environment have made them virtually ubiquitous in soil, water, and air today. PCBs accumulate in animal liver and fat tissue. They disrupt normal hormonal functioning and cause brain and nervous system damage, cancer, and other health problems.

Scientists already knew that when PCBs bind to the pregnane X receptor (PXR) in rodents, the ligand-bound receptor elicits production of enzymes that metabolize PCBs and other contaminants. The current study was a comparative investigation to determine what happens when PCBs bind to the human counterpart to PXR--the steroid and xenobiotic receptor (SXR).

The research design incorporated multiple approaches to evaluate reactions from exposure of human and rat cells to at least 26 different PCBs in separate assays. Activation assays tested the ability of the compounds to activate mRNA transcription of the receptor in transfected cells. The authors used Northern blots to compare the amount of mRNA produced in exposed versus control cells. In another type of test, RNA isolated from exposed cells was converted to cDNA and amplified by QRT-PCR (quantitative real-time reverse transcriptase polymerase chain reaction) to quantify gene expression.

The scientists discovered that the most stable and abundant PCBs found in human tissues activated PXR and induced expression of target genes for metabolizing the toxicants, producing a protective physiological reaction in rodents. In contrast, these same PCBs antagonized SXR, blocking the expression of target genes and inhibiting the human body's ability to physiologically counteract harmful effects of exposure (several less stable PCBs did not antagonize SXR, however). They also found that PCBs were able to increase the amount of a key metabolic enzyme in rat liver cells but not in human liver cells. QRT-PCR analysis showed that treatment with antagonistic PCBs reduced the expression of genes encoding key metabolic enzymes in human cells, confirming the results of the activation assays and Northern blots.

The demonstration that many PCBs block SXR in humans rather than activate it--as previously thought, based on animal models--challenges the assumption that animal models extrapolate to humans. This discovery brings important new insight for development of sound, rational, science-based public policy to protect human health.