Peroxisome proliferator carcinogenesis remains a controversial subject among the scientific community, in spite of the fact that it has matured from infancy to adolescence. "In the last 10 years, we've obtained a significant amount of information regarding peroxisome proliferators and human risk assessment," says James Popp, international toxicology director at Sanofi-Winthrop. "A decade ago, the discussion was based largely on hypotheses and limited data. Today, our discussions focus on a much broader body of scientific knowledge."
James Popp--Animal data is useless if you can't extrapolate the data to humans.
Source: Sanofi-Winthrop
|
Peroxisomes are single-membrane cytoplasmic organelles that contain a fine granular matrix. Present in all mammalian cells except red blood cells, peroxisomes carry out peroxidative functions through oxidase enzymes by metabolizing long-chain fatty acids. The number of peroxisomes, their size, and enzymatic profile vary by tissue location. When exposed to certain chemicals called "peroxisome proliferators," these organelles readily multiply. To date, more than 70 chemicals, ranging from hypolipidemic agents to phthalate ester plasticizers to solvents and herbicides to dietary factors and hormones, have been cited in the scientific literature as peroxisome proliferators.
While these compounds seem to produce similar qualitative changes in rodent liver cells, their quantitative response is dose dependent and species specific. Although the liver is the prime target of peroxisome proliferators, morphological and/or biochemical changes occur in several tissues including the testis, thyroid, kidney, intestine, adrenal glands, brown adipose tissue, and heart. Also, in addition to liver tumors, a subset of peroxisome proliferators increase testicular and pancreatic tumors in rats. Rodent hepatic cells, though, are the most responsive to peroxisome proliferators and the most extensively studied tissue.
Extensive experimental evidence proves there is more than a casual link between peroxisome proliferator-elicited hepatomegaly (enlarged liver) and subsequent liver tumors in rodents. This link, though, has not been established in humans. Peroxisome proliferators are among the most extensively studied nongenotoxic carcinogens. Still, proposed peroxisome proliferator mechanisms of action, including oxidative stress, induction of hepatic cell proliferation, and the role of the peroxisome proliferator-activated receptor (PPAR), need to be sorted out. Until these mechanisms of rodent cancer induction are fully elucidated, it will be difficult to assess the health risk to humans.
Peroxisomal Pathways
Hepatic peroxisomes are usually round, measuring between 0.5 and 1.0 microns. Rodent hepatocytes can contain enormous numbers of peroxisomes. Liver peroxisomal enzymes in rodents are induced within a few hours to a few days after exposure to peroxisome proliferators. The activity of catalase, the marker enzyme for these organelles, and urate oxidase, a key enzyme in catabolism of nitrogenous compounds in rodents, increases nearly 2-fold; the fatty acid beta-oxidation enzyme system can increase 20- to 30-fold. Differential regulation of genes encoding peroxisomal enzymes, many known to be controlled by PPAR, account for this difference in activity. Hepatomegaly and hepatic hypertrophy (enlarged liver cells) occur within a few days, usually reaching a steady-state within two weeks, and persist as long as the chemical treatment continues.
Peroxisome proliferators are nongenotoxic carcinogens, which, unlike genotoxic carcinogens, do not directly damage DNA. Rather, they alter gene expression via biological effects. In the liver, for example, nongenotoxic carcinogens may cause hepatomegaly through their action on endoplasmic reticulum, by increasing the number of peroxisomes, and/or by increasing the number of mitochondria. Many nongenotoxic carcinogens are now being identified. Many of them have no detectable effects in short-term tests for genotoxicity, yet they cause malignant tumors in rodents in long-term experiments.
Oxidative stress. Northwestern University Medical School's Janaradan Reddy and Narendra Lalwani first proposed the "active oxygen" hypothesis as a molecular mechanism to link peroxisome proliferators to hepatocarcinogenicity. This hypothesis suggests that the cells are over-challenged by peroxisome proliferators in the following way. Peroxisome proliferators stimulate increases in peroxisomal beta-oxidation. The beta-oxidation system may become 15- to 25-fold more active, while catalase production, though stimulated, is minimally affected (2- to 3-fold). In the peroxisomal membrane, long-chain fatty acids are oxidized, ultimately yielding hydrogen peroxide. Within the peroxisome, hydrogen peroxide is normally detoxified by catalases, but if hydrogen peroxide were to escape the peroxisome, toxicity may result. Oxygen radicals from hydrogen peroxide in vitro can induce several responses including lipid peroxidation, membrane damage, and accumulation of lipofuscin. Oxidative damage to DNA could ultimately initiate tumorigenesis. To date, however, only lipofuscin accumulation has been consistently shown to support the oxidative stress hypothesis.
Cell proliferation. Liver size increases dramatically in rodents after exposure to many peroxisome proliferators. In addition to hepatic hypertrophy, hepatocellular replication is responsible for much of the increase, through a self-limiting and transient process that occurs during the early stages of exposure. During the first week after exposure to most peroxisome proliferators, the cell replication rate may increase dramatically but soon returns to its baseline rate. Higher doses of some peroxisome proliferators, such as the experimental drug Wy-14,643, may be the exception, where cell turnover may persist.
Dan Marsman --Short-term exposure may not be a significant risk for humans.
Source: NIEHS |
According to Reddy and his colleague, Sambasiva Rao, this early cell proliferation caused by peroxisome proliferators is a mitogenic response and may be less carcinogenic than cytotoxic, regenerative-type cell division. They report that the average latency period between exposure and occurrence of hepatocellular carcinomas ranges from 50 to 120 weeks in rodents. Animals exposed to potent peroxisome proliferators such as ciprofibrate develop tumors as early as 50 weeks after exposure, while those exposed to weaker peroxisome proliferators such as phthalate esters start developing tumors by week 90. Recent studies have shown that removal of rodents from peroxisome proliferator exposure, after the appearance of benign adenomas but before hepatocellular carcinomas appear, results in complete tumor regression. "Tumor regression such as this is consistent with the important role played by tumor promotion and/or progression by peroxisome proliferators," says Daniel Marsman, a pathologist at the NIEHS.
The classic definition of chemically induced carcinogenesis includes stages of initiation, promotion, and progression. There is some question as to whether peroxisome proliferators induce or promote cancer. "An active area of peroxisome proliferator investigation is their role as cancer promoters," says Marsman. "Peroxisome proliferators are not DNA reactive and have not been shown to act as initiators. Instead they may promote lesions initiated by other chemicals or those occurring spontaneously. The tumors induced are the result of a multistage process. Continuous, long-term peroxisome proliferator exposure is often needed for carcinogenic activity, unlike the one-shot, cancer-inducing exposure of many directly mutagenic chemicals. Corroboration of tumor regression studies by other scientists might indicate that short-term exposure to peroxisome proliferators may not be a significant risk factor in humans. What we may find instead is that only high-dose repeated exposure to these agents would raise human risk."
Some scientists believe that the hyperplastic response and the associated tumor promotion activity may become a better predictor of nongenotoxic hepatocarcinogenesis than peroxisome proliferation because there are so many unanswered questions regarding the link between peroxisome proliferation and neoplasia.
Peroxisome proliferator-activated receptor. Other investigations into long-term effects of peroxisome proliferators suggest that factors other than oxidative injury or cell proliferation may influence the carcinogenicity of these compounds.
The PPAR, recently isolated, cloned, and sequenced, may be associated with the former proposed mechanisms, or may represent a separate mechanism of carcinogenesis. A member of the steroid hormone receptor superfamily, these PPARs contain both a DNA-binding domain and ligand-binding domain. Activated by structurally diverse peroxisome proliferators or fatty acids, the PPAR modulates gene expression. In in vitro experiments, the PPAR regulates transcription of genes responsible for biochemical changes (inducing peroxisomal enzymes responsible for beta-oxidation and certain microsomal enzymes) observed in peroxisome proliferator-treated cells.
According to John Ashby and colleagues at Zeneca Central Toxicology Laboratory, if a peroxisome proliferator bound directly to PPARs, then the compound's potency could be determined by its ability to activate these receptors. However, peroxisome proliferators do not appear to bind directly to PPARs, but rather indirectly activate the receptors by influencing other cellular components.
"A few years ago, there was a lot of excitement when PPAR was first identified," says Marsman. "The marked increases in enzyme activity and the rapid induction of immediate to early oncogenes and cell proliferation may fit with a receptor mode of action. So far, however, we know PPAR to be involved primarily in lipid metabolism, and it's still too early to predict its full impact."
Jack VandenHeuvel --Lack of peroxisome proliferation in humans doesn't negate the risk.
Source: Purdue University.
|
PPAR subtypes have been discovered as well in humans, and two have been cloned. Jack VandenHeuvel, assistant professor in Pharmacology and Toxicology at Purdue University, expresses this opinion, "Since rodent and human PPARs are activated by peroxisome proliferators, I believe that receptor-mediated responses are going to be similar between the two species. As to risk assessment, it's hard to say at this time which responses are relevant to carcinogenesis. Once we identify the genes responsible for the carcinogenic process and identify those under direct PPAR control, we'll be able to bridge that gap."
Rodents at Risk
William Kluwe --It is important to concentrate on factors known to cause human cancers.
Source: Roger Riley
|
"We know that compounds that cause peroxisome proliferation in rodent liver also cause cancer in rodent livers," says William Kluwe, director of toxicology at Pfizer Laboratories. "Whether or not this can be extrapolated to cancer in man is the heart of the issue." According to Kluwe, peroxisomes that can be induced in vitro in rodent hepatocytes cannot be replicated in vitro in human hepatocytes, leading some scientists to conclude that peroxisome proliferation is not a significant human response. However, studies show that the link between peroxisome proliferation and liver cancer is not a confirmed cause-and-effect relationship. Another entirely different mechanism that also happens to generate peroxisomes may be at work. This neither supports nor precludes the concept of using peroxisome proliferation as a surrogate predictor of cancer risk.
Reddy and colleagues originally described the hepatocarcinogenic effect of nafenopin in mice in 1976. Soon after, several reports appeared in the literature establishing peroxisome proliferators as complete liver carcinogens in rodents. Chronic feeding of peroxisome proliferators to rats causes the following liver lesions to appear in a sequential fashion: altered areas, hepatocellular adenomas, and finally hepatocellular carcinomas.
Source: Cliff Elcombe
According to Ashby, there is an 81% correlation between chemicals that induce peroxisome proliferation in laboratory animal livers in the short term and those that induce hepatocellular tumors in those animals over the long term. However, peroxisome-proliferating carcinogens vary widely in their potency and effectiveness. Ashby compared the potency and carcinogenic effectiveness of ciprofibrate, diethylhexyl phthalate (DEHP), tibric acid, methyl clofenapate, nafenopin, and Wy-14,643 in the F344 male rat. The highly effective hepatocarcinogen ciprofibrate, at a low dosage of 300 parts per million (ppm), induced a high incidence of tumors. At the other end of the spectrum, DEHP required a dose of 12,000 ppm to produce tumors. Tumor latency period, an additional potency indicator, was not reported for DEHP because of insufficient data. Results indicate that potent peroxisome proliferators (those inducing at least a sixfold increase in hepatic peroxisomes) are also highly effective liver carcinogens. The DEHP study also demonstrates a good relative association between carcinogenesis and peroxisome proliferation. F344 rats fed 12,000 ppm DEHP had a tumor incidence of 24% and a 5-fold increase in peroxisomes. At a dose of 20,000 ppm, the figures were 78% and 7-fold, respectively. "This suggests that perhaps only a small induction of peroxisomes is necessary to produce tumors in this strain [of rat]," Ashby and colleagues report.
Rodents are particularly susceptible to peroxisome proliferation, but evidence indicates that other species such as the guinea pig, marmoset, and humans are much less susceptible. Gordon Gibson of the University of Surrey's School of Biological Sciences writes in an article in Toxicology Letters, "Much controversy exists as to species differences in susceptibility to the hepatocarcinogenicity of peroxisome proliferators, with obvious implications for human risk assessment."
Predicting Human Hazard
"If we could understand the mechanistic process of peroxisome proliferators in rodents, we could look for similar processes in humans," says Ronald Melnick, a senior toxicologist at the NIEHS. "That would give us a stronger foundation for evaluating human risk."
Popp declared at the 1990 Toxicology Forum, "It doesn't do us any good to continue to study a particular phenomenon in rodents and be able to study it to the nth degree if we cannot make a transition of that data over to humans. I think in this particular case the use of human hepatocytes is extremely important in taking our rodent toxicology data and trying to understand what it means for man."
Comparative in vitro studies in human, guinea pig, hamster, and rat hepatocytes are underway. Although enzyme induction has been the primary focus thus far, the studies reaffirm relatively little, if any, response of human hepatocytes to peroxisome proliferators, concedes Penelope Fenner-Crisp, acting deputy director at EPA's Office of Pesticide Programs.
In the early literature, a few studies examined human liver biopsies from people who had received hypolipidemic drugs in the past. One qualitative study of nine patients who had taken gemfibrozil demonstrated no increase in peroxisomes. A quantitative study of 16 patients who had taken clofibrate found a 50% increase in the number of their hepatocyte peroxisomes. Cliff Elcombe, of Zeneca's Central Toxicology Laboratory, reported to the Toxicology Forum in 1990 that "some of these observations are questionable because peroxisomes have a half-life of about 36 hours. Once the peroxisome-inducing agent is stopped, peroxisome levels rapidly return to baseline levels."
Elcombe claims that hepatocyte culture systems eliminate problems of absorption of these agents from the gastrointestinal tract. At the forum, he related results from a study investigating hepatocyte cell cultures from the rat, guinea pig, marmoset, and humans. Maintained for four days in the presence of methylcofenapate, one of the most potent proliferators known, and using palmitoyl CoA oxidation as a peroxisomal enzyme marker, these cultures yielded a dose-related increase in the enzyme (more than 1500%) in rats, whereas enzyme activity in guinea pig, marmoset, and human cells was negligible.
William Stott, a toxicologist at Dow Chemical, says that peroxisome proliferators do not represent a potential carcinogenic threat to humans. Basic genotoxicity tests of many chemicals that cause peroxisome proliferation demonstrate that these agents do not cause DNA damage. Instead, their tumorigenic effects on the liver appear to be a secondary phenomenon to changes the proliferators induce such as cell proliferation. According to Stott, peroxisome proliferators generate a short burst of cell proliferation in rodents, called a true mitogenic wave of proliferation. "You don't see that happening in higher mammalian cells," he said.
Stott continues, "Changes such as peroxisome proliferation and increased cell proliferation are necessary for tumors to form. Without these effects, you don't get tumors. Long-term epidemiology studies in monkeys indicate that on exposure to peroxisome proliferators, their livers are less sensitive to these compounds than seen in rodents."
Kluwe says it is important to concentrate on factors known to cause human cancers, such as direct damage to DNA. "Results from studies looking for DNA damage in rodents are contradictory," states Kluwe. "Most of them suggest that there is no evidence of DNA damage by chemicals that induce peroxisome proliferation, even in rodents. Nonetheless, we still don't understand why rats get liver tumors; if it is not because of DNA damage, then what causes the tumors?"
Long before we knew what a peroxisome was, we were investigating potential carcinogenicity in chemicals," Kluwe continues. "Peroxisomes are supposed to be a shortcut. Rather than spending three years and lots of money to see if a compound caused liver cancer, we hoped a two-week study on 20 animals would give us the answer. So far, there are no shortcuts. Besides, looking at a single parameter and trying to decide if a chemical is a true carcinogenic risk based on that information is naive."
The Puzzle Persists
The controversy over the risk of peroxisome proliferators to humans continues. According to George Lucier of the NIEHS, "the use of mechanistic data is an emerging issue in risk assessment that's becoming increasingly compelling. There is an industry-wide push to develop strategies to identify which animal carcinogens may not be carcinogens in humans. The peroxisome proliferators are one important aspect of that issue. Mechanistic data would help considerably in our evaluations of this issue. In the absence of evidence that humans are not responsive, it is prudent to assume that rodent carcinogens pose a health risk to humans."
Part of the problem in defining human risk assessment is obtaining in vivo data for humans. Existing clinical data are difficult to obtain because some of the data are proprietary and difficult to use because of the limited number of human subjects.
Stott reports that investigations on the toxic effects of peroxisome proliferators in higher mammals, including man, are underway on autopsy material and in vitro hepatocytes. "We test for peroxisome proliferators on a selected basis. Data from these investigations, though not one of our standard registration type studies, is useful during product development. If a compound we want to register induces peroxisome proliferation, it's a fair bet we're going to get liver tumors in mice or rats. This data is only one piece of information that goes in the judgment call of whether development on the product continues or we look for a compound that has a similar efficacy but does not induce peroxisome proliferation."
Pharmaceutical and chemical companies alike report that they try to avoid developing products that induce peroxisome proliferation because of regulatory issues. The U.S. regulatory agencies have not yet deemed peroxisome proliferators a class of chemicals harmless to man. Peroxisome proliferators do not enjoy the status of chemicals that bind to alpha-2u-globulin, a process which is known to cause kidney cancer in male rats but is considered irrelevant to humans.
Nonetheless, there is strong feeling on one side of the scientific community that studies on peroxisome proliferators suggest no hazard to humans. "It should be recognized that a large body of existing data demonstrates clearly that human [hepatocytes] are resistant to chemically induced peroxisome proliferation," says Elcombe. "This suggests that peroxisome proliferation data, and consequential tumor data generated in rats and mice with nongenotoxic peroxisome proliferators, is inappropriate information to use in human risk assessment. The latest research suggests that subtle species differences in response to PPARs may explain the observed species differences in response to peroxisome proliferators."
Ashby and colleagues write, a November 1994 article in Human and Experimental Toxicology, "Although the correlation of peroxisome proliferation with hepatocarcinogenesis is striking, the generic mechanism by which this class of chemicals induces liver tumors is still not understood. On the one hand, it could be argued that until the precise mechanism by which peroxisome proliferator-induced cancer is elucidated, it cannot be concluded without doubt that this class of agents does not produce cancer by a mechanism totally distinct from, and independent of, peroxisome proliferation. A parallel exists with other forms of nongenotoxic carcinogenesis, such as rodent thyroid carcinogenesis associated with elevated levels of thyroid-stimulating hormone. In this example, the mechanistic evidence is supportive of, but does not provide conclusive proof of, the absence of human hazard."
"On the other hand," Ashby continues, "it could be argued that as long as there are data to support the association between peroxisome proliferation and related phenomena in the liver and hepatocarcinogenesis then the peroxisome proliferation effect can, and should, be used to establish the hazard of this class of agents to animals and man."
Looking at the economic effect of this controversial issue, Stott says, "Certain products have not made it to market in the United States because this issue has not been resolved. And Americans are not enjoying the potential benefits of these products."
On the other end of the spectrum, in an article in Critical Reviews in Toxicology, Reddy and Lalwani conclude, "a peroxisome proliferator carcinogenic in experimental animals, as with all chemical carcinogens . . . should be grounds for considering such a compound as a carcinogen irrespective of the carcinogenic mechanism(s) involved."
VandenHeuvel supports this position: "Just because we don't see an increase in the number of peroxisomes in human cells following exposure to peroxisome proliferators, we can't claim this negates human risk assessment." As evidence VandenHeuvel states that in rodent testis, even though increases are not seen in peroxisomes, there is a significant increase in Leydig cell tumors with exposure to peroxisome proliferators. Says VandenHeuvel, "Until we determine which genes are involved in the carcinogenic process, and how they are impacted by peroxisome proliferators and PPARs, I don't feel comfortable about saying that there is no risk."
"Regulatory agencies need to reach a consensus on handling peroxisome proliferators in terms of human risk assessment," Popp asserts, "setting up guidelines on whether to deal with them as a group or address each chemical individually. Then we'll be ready to look at the data and determine how to use it in risk assessment."
At a recent meeting of the International Association of Research in Cancer (IARC) in Lyon, questions relevant to the mechanisms of peroxisome proliferation were pondered. Although many issues remain unresolved, a consensus report on how to handle data revolving around peroxisome proliferation is forthcoming. Hirohi Yamasaki, unit chief of multistage carcinogenesis at IARC, says, "I think we've reached some reasonable conclusions. When we evaluate chemicals that induce peroxisome proliferation, for example, we should evaluate them case-by-case rather than grouping them as peroxisome proliferators and saying that as a group they are safe or dangerous in rodents or humans."
Melnick intimates that the pieces of the peroxisome puzzle are far from coming together. "Currently, there is no evidence that positively demonstrates peroxisome proliferators cause liver tumors in humans. However, a negative doesn't prove that it doesn't do it; it only means that the effect might not yet have been observed."
Source: Cliff Elcombe
Marilyn Citron
Marilyn Citron is a freelance journalist in Birmingham, Mississippi.
[Table of Contents]
Last Update: April 8, 1998