Salk Group 'Humanizes' The Mouse
SXR gene work advances research front on drug-drug interactions
By A.J.S. Rayl
Researchers under the direction of Ronald M. Evans at the Salk Institute for Biological Studies in La Jolla, Calif., have created a "humanized" mouse capable of detecting potentially toxic substances in the body. In the process, they have uncovered what they believe to be the primary source of the xenobiotic response within a specific gene they call SXR, steroid and xenobiotic receptor. SXR "may represent the critical biochemical mechanism of human xenoprotection," the authors wrote in a report.1 And the transgenic mouse--known as the SXR mouse--may well profoundly affect the course of drug development.
"We have unequivocally demonstrated the basis for the xenobiotic response residing within a specific molecule, which is a nuclear receptor," explains Evans, director of the Gene Expression Laboratory at Salk and senior author of study. The work " ... makes the study of the xenobiotic response now amenable to molecular analysis. [Because it is] so straightforward, we were able to create the humanized mouse."
These experiments are undeniably intriguing and important to the pharmaceutical industry and to the public at large. Although the rodent model has been shown to be unreliable in numerous studies, the pharmaceutical industry has had to rely primarily on mice and rat models to test for xenobiotic reactions in development of drugs because they were pretty much the only game around.
"Now we are providing a very viable animal model, an ideal system [we can use to] study human drugs and also use as a tool for research in general," says Wen Xie, a doctoral researcher in Evans' lab and the study's lead author.
Beyond its obvious role in drug development, the SXR mouse has the revolutionary capacity to address one of the most irksome problems in the pharmaceutical industry: drug-drug interactions. In fact, the pharmaceutical industry is already embracing the new mouse. "I could immediately see how this could be applied at ... early stages of drug discovery for in vitro analysis [and at late-stage discovery to] choose ... compounds to take forward into preclinical development, then into man," says Paul Spence, vice president of biotechnology for Pharmacia.
Marco M. Gottardis, associate director of oncology at Bristol-Myers Squibb, calls the SXR mouse "the beginning of a new era" in preclinical drug profiling.
A Smoking Gun and Red Herrings
Evans' team first isolated SXR in 1992 while Bruce Blumberg, now an assistant professor at University of California, Irvine, was searching for the human homologue for a frog gene called BXR. Blumberg had cloned BXR in the early 1990s at the University of California, Los Angeles, when he was searching for new developmental signaling molecules. His desire to discover the endogenous hormone for the BXR receptor took him to Salk in 1992. There, he discovered that the frog gene had a number of different complex components.
"That really put us on the road," says Evans. "The frog receptor responded to an unusual set of compounds º: anesthetics. It was very odd for a receptor of this class to be responsive to a drug, a prescription drug that was used in surgery. That's when we began to think, 'This, in fact, may be a receptor for external activators, a type of sensor for foreign compounds.'"
Not long after that, Blumberg uncovered SXR. "When we had human cDNAs, about 1995-'96, we started screening for ligands and were amazed that many, many things activated this receptor," Blumberg remembers. "Since the first things we saw activate it were steroids, we thought, 'Great, this is some new kind of steroid receptor.' We eventually realized this thing really is activated by lots of steroids." Evans adds, "We knew, I think, relatively early in the process that we were onto something. We had the smoking gun sense about it."
Although they had considered the possible existence of a xenobiotic receptor, other labs at that time were concluding that the xenobiotic sensor or receptor was not related to the gene they were studying, that there was no unifying mechanism. "We were discouraged by those original descriptions," admits Evans. "But we persevered."
About the same time, "Steve Kliewer at Glaxo Wellcome discovered a mouse gene that responded to an unusual class of steroidal ligands," recalls Evans. "But there are many different types of steroids." The actual moment of xenobiotic realization arrived unpretentiously one afternoon. Blumberg recalls, "It probably came when we were sitting in [Evans'] office one day ... saying, 'Why is this stupid receptor activated by so many different things?' One of us said, 'OK, let's take a different tactic. Let's just assume this means something. If it's supposed to be activated by all these different things, then maybe it's a sensor. We know it's expressed in liver, so maybe it's involved in breaking [things] down.'" At that point, they were, for all intents and purposes, there.
Evans and Blumberg expanded the concept into the Steroid Sensor Hypothesis, which marked the shift in their thinking.2 The Steroid Sensor Hypothesis basically states that "since there are all these bioactive compounds in the body, if you want to use them to do things, you need to get rid of them as well as to produce them," says Blumberg. "You can't just signal all the time; you have to take away the signal. And since the synthesis of steroids is regulated, why not the breakdown? Perhaps there is a receptor or a small number of receptors [that] are involved in sensing the levels of these compounds and then breaking them down."
A "key component" of their discovery was the idea that one or perhaps a few receptors might be a target for many different types of molecules. "The nature of the xenobiotic response itself requires the body to be able to respond to literally thousands, if not an infinite number, of compounds in the external world," Evans points out. "It seems unlikely that the genome could encode receptors dedicated to thousands of unknown compounds that the body has never seen, so it must have to evolve a way to have a small number of receptors that can register the presence of a large number of compounds. That was the hypothesis that we went in with--looking for one or a few genes that might explain this unusual type of response the body has to toxins and to external and orally active compounds."
It didn't take long for them to realize that the gene they were studying in xenobiotics was basically the same homologue Kliewer was studying in a set of steroid activators. "That led us to deduce that this was probably in fact a xenobiotic receptor, because the xenobiotic response includes responses to active steroid drugs such as tamoxifen, androgens, estrogen, [and] glucocorticoids, as well as to xenobiotic compounds from the environment," explains Evans.
During the ensuing years, Evans' group put the receptor through thousands of tests and discovered that SXR:
* Resides primarily in the liver and intestine--not surprising, because those two organs are responsible for degrading and eliminating foreign substances and toxins from the body.
* Serves as a biological lookout that detects potentially harmful substances.
* Triggers the cytochrome P450s, a family of enzymes that functions as the body's metabolic garbage disposal, grinding up the potentially harmful substances and flushing away the residue.
"The P450 enzyme that primarily recognizes and degrades the drugs and xenobiotics that we ingest is called CYP3A--this pathway exists in rabbit, mouse, and in most vertebrates--and in humans it is regulated primarily by SXR," elaborates Blumberg. "A very interesting feature of this receptor and what makes it so interesting for the pharmaceutical industry is that º [it] behaves very differently [from one species to the next]. In many cases, the metabolism in mice and humans is the same. But in some cases, it's not. So the human receptor has a very different pharmacology than the mouse receptor [has].
º A good example would be ... rifampicin, an antituberculosis drug, which is a very strong activator of the human system but doesn't touch mouse."
Once that xenobiotic response is turned on, it's working on all foreign invaders. In one drug-drug interaction, it has already been demonstrated that women who take rifampicin and birth control pills may have so-called "miracle babies," as Evans dubs them, because women are on the best method of contraception available.
Though the first set of published experiments showed that SXR could respond to many different kinds of compounds, it really did not prove that this receptor was capable of sustaining the xenobiotic response in a living organism, admits Evans.
Step by Step
The goal of the most recent work was, Evans says, "to use targeted mutation by homologous recombination or gene knockout technology to mutate the rodent gene and to show that the xenobiotic response in the rodent was lost, which, in fact, was the case." The second step was to introduce the human gene into the rodent and show the rodent had acquired the human xenobiotic response.
In comparing the action of a popular anesthetic on a normal mouse with its action on a transgenic SXR mouse whose xenobiotic response is constantly turned on (making it resistant to the effect of most drugs), they found that while the normal mice slept for at least a half-hour following administration of the anesthetic, their transgenic counterparts remained wide awake. "This demonstrates how the activation of a genetic network, in this case the xenobiotic response network, confers resistance to drugs," Evans maintains. When that experiment succeeded, says Evans, "We popped the cork."
The plan now, according to Evans, is to make SXR mouse available across the board as soon as possible. "That's a good move, because this will be a very valuable tool that will benefit the public, allowing us to be even more accurate in choosing compounds [that] are going to be safe in the clinic," says Spence. Evans "wants to create a standard, and that really is the only way," suggests Gottardis. "I think [it will be an advantage for everybody to have the same mouse]. That will have the [Food and Drug Administration] sit up and take notice."
Evans is confident the FDA will not only sit up and take notice, but that it will take action. "It is my guess that during the next few years, a test through a humanized mouse will become essential in drug development," predicts Evans. "It might even be mandated by the FDA, because the rodent by itself is such an unreliable monitor of drug interaction and drug toxicity in humans. The need for a humanized rodent is tremendous. Most people don't understand the nature of the problem of drug-drug interaction, the complexity of the xenobiotic response itself, and the rather elegant solution to that problem." There are a few other receptors that respond to drugs and xenobiotics, and several more are expected to emerge from the human genome projects. Still, the Salk team feels that SXR is the key player. "The fact of the matter is that the P450 enzyme system, which SXR regulates, recognizes and degrades the vast majority of the compounds," says Blumberg. "We know a lot about these orphan nuclear receptors.3 We've been studying them for years. Considering that SXR sees so many compounds leads me to believe that it's still going to be the major player--even if we discover more pathways."
No matter how you look at it, Blumberg adds, "The SXR mouse is a giant leap, and [it] provides a really major key that can be exploited to everyone's benefit." Evans sums it up: "It has been a long haul to get to the top of the mountain. But when we got there, suddenly we realized it was a great view."S A.J.S. Rayl (ajsrayl@loop.com) is a freelance writer in Malibu, Calif.
References
1. W. Xie et al., "Humanized xenobiotic response in mice expressing nuclear receptor SXR," Nature, 406:435-9, July 27, 2000.
2. B. Blumberg et al., "SXR: a novel steroid and xenobiotic activated nuclear hormone receptor," Genes and Development, 12:3195-3205, 1998.
3. "Orphan nuclear receptors: new ligands and new possibilities," Genes and Development, 12:3149-55, 1998.
the-scientist.com |
