1

GREGORY B. BULKLEY

Reactive Oxygen Species

The Evolution of a Scientific Concept

Free Radicals and

ver the past few decades, free radicalshighly reactive and thereby destructive molecules have come to be appreciated increasingly for their importance to human health and disease. Many common and life-threatening human diseases, including atherosclerosis, cancer, and aging, have free radical reactions as an underlying mechanism of injury. Over this period of time, our conceptual understanding of the interaction of such reactive oxygen species (ros) with living organisms has undergone a remarkable evolution that is perhaps instructive of how human knowledge of a scientific concept develops and matures. As this understanding has evolved, it has provided unprecedented opportunities for improving the quality and even length of human life.

Free Radicals

Free Radicals Molecules are made up of one or more atomic nuclei surrounded by orbiting electrons. The electrons are arranged in orbitals, depending upon their respective distances from these nuclei. In most molecules, the electrons within each orbital are paired with another electron thats spinning in the opposite direction. (The concept of spin, distinct from the concept of electrical charge, is less of a literal description of the electrons actual behavior than an analogy that helps us visualize this unseen world in familiar terms.) The paired electrons keep the molecule relatively stable (at a lower energy state) and thereby less reactive. When one or more electrons, especially within the outer orbital, is/are unpaired with respect to spin, the molecule becomes relatively unstable (at a higher energy state) and consequently more reactive with other molecules. A free radical is a molecule with one or more unpaired electrons in its outer orbital. Many of these molecular species are oxygen (and sometimes nitrogen) centered. Indeed, the molecular oxygen we breathe is a free radical. These highly unstable molecules tend to react rapidly with adjacent molecules, donating, abstracting, or even sharing their outer orbital electron(s). This reaction not only changes the adjacent, target molecule, sometimes in profound ways, but often passes the unpaired electron along to the target, generating a second free radical or other ros, which can then go on to react with a new target. In fact, much of the high reactivity of ros is due to their generation of such molecular chain reactions, effectively amplifying their effects many fold. The changes wrought on adjacent molecular targets can vary in magnitude, but because many of the components of the living cell are particularly susceptible to free radical injury, the molecular chain reactions can have substantial effects on the structure and function of living tissue. As a consequence, natural

ROS and Human Health

selection has driven the evolution of a number of intracellular defense mechanisms to neutralize or control the potentially destructive reactivity of ros. These include molecules that react preferentially with ros without passing that reactivity along. Some of these are simple molecules like vitamins E and C, while some are enzymes like superoxide dismutase (sod) and catalase, which catalyze such electron-quenching reactions. Often these compounds are referred to collectively as free radical scavengers. ROS and Human Health Because our bodies are continuously exposed to free radicals and other ros, from both external sources (sunlight, other forms of radiation, pollution) and generated endogenously, rosmediated tissue injury is a final common pathway for a number of disease processes. Radiation injury represents an important cause of rosmediated disease. Extreme examples include the physicalchemical reactions within the center of the sun and at the center of a thermonuclear blast. With respect to more commonly encountered levels of radiation, depending upon the situation, about two-thirds of the sustained injury is mediated not by the radiation itself, but by the ros generated secondarily. This applies not only to the acutely toxic forms of radiation injury, but the long-term, mutagenic (and hence carcinogenic) effects as well. Thus, the victims of the thermonuclear explosion in Hiroshima, both those who died instantly from the blast at ground zero and those in the suburbs who developed leukemia and lymphomas years later, were victims of free radical injury. An important clinical application of this principle is encountered regularly in the treatment of cancer by radiation therapy. Large tumors often outgrow their blood supplies and tumor cells die within the center, despite being well-oxygenated at the periphery. Between these two regions is an area of tumor

ROS and Human Health

that is poorly oxygenated, yet remains viable. Radiation therapy of such tumors is particularly effective at the periphery, where an abundant concentration of oxygen is available to form tumorcidal ros. The poorly oxygenated center is injured to a significantly smaller degree. While the dead cells in the center dont survive anyway, the poorly oxygenated, yet viable, cells between these two areas can survive a safe dose of radiation therapy, and thereby seed a later local recurrence of the tumor. This is a major reason why many large tumors are treated by a combination of radiation therapy (to kill the tumor at its advancing edges) and surgical removal of the bulk of the tumor, including these particularly dangerous remaining cells. Cancer and other malignancies all entail unconstrained cell growth and proliferation based upon changes in the cells genetic information. In most cases, for example, one or more genes that normally constrain cell growth and replication is/are mutated, or otherwise inactivated. These genetic deficiencies correspond directly with deletions and sequence changes in the genetic code, resident in the cells dna. A frequently seen final common cause of such dna damage is free radical injury. Of the myriad injuries sustained by our dna on a daily basis, most are repaired by normal dna repair mechanisms within the cell, while some result in cell death. Since such injuries are sporadic and distributed somewhat randomly across the genome, most lethal dna injuries are clinically inconsequential, resulting in the loss of a few cells among millions. However, when a single cell sustains an injury that impairs growth regulation, it can proliferate disproportionately and grow rapidly to dominate the cell population by positive natural selection. The result is a tumor, frequently a malignant one, where the constraint of growth and proliferation is particularly deficient. Therefore, free radical injury to the genetic material is a major final common pathway for carcinogenesis.

ROS and Human Health

ROS can be generated within the cell not only by external sources of radiation, but also within the body as a by product of normal metabolic processes. An important source of endogenous free radicals is the metabolism of some drugs, pollutants, and other chemicals and toxins, collectively termed xenobiotics. While some of these are directly toxic, many others generate massive free radical fluxes via the very metabolic processes that the body uses to detoxify them. One example is the metabolism of the herbicide paraquat. At one time, drug enforcement authorities used this herbicide to kill marijuana plants. Growers realized they could harvest the sprayed crop before it wilted, and still sell the paraquat-laced product. Many who smoked this product subsequently died of a fulminant lung injury. Fortunately, this approach has been abandoned as a particularly inhumane way to solve the drug problem. While the paraquat story is a particularly striking example of a metabolic mechanism of free radical toxicity, many commonly encountered xenobiotics, including cigarette smoke, air pollutants, and even alcohol are toxic, and often carcinogenic to a large degree by virtue of the free radicals generated by their catabolism within our bodies. Moreover, there is accumulating evidence that a diet rich in fruits and vegetables, which are high in natural antioxidants, and low in saturated fat (a particularly vulnerable target for damage by ros), reduces the risk of atherosclerosis and cancer. Atherosclerosis is a complex process that leads to heart attack, stroke, and limb loss by the plugging of the arteries with atherosclerotic plaque. This plaque is a form of oxidized fat. When free radicals react with lipids, the consequence is lipid peroxidation, the same process by which butter turns rancid when exposed to the oxygen in the air. While a number of factors influence the development and severity of atherosclerosis, a major factor is the ros-mediated peroxidation of our low density

ROS and Human Health

lipoproteins (ldls, or bad cholesterol). The dietary approach to the prevention of heart disease and stroke is based partially on adding dietary antioxidants to limit ldl oxidation, as well as decreasing the intake of fat itself. These approaches already have made significant inroads into the mortality from heart disease, but offer the potential for safe pharmacological prevention in the future that is not as dependent upon willpower as are diet and exercise. Degenerative neurological diseases affect millions of Americans. A number of these diseases, including amyotrophic lateral sclerosis (als, or Lou Gehrigs disease), Parkinsons disease, and Alzheimers disease, appear to have ros toxicity as a central component of their underlying mechanism of nerve cell destruction. Unfortunately, there is little evidence that simply eating more dietary or even pharmocologic antioxidants will prevent or arrest the neural degeneration; not surprisingly the mechanism is too complex to lend itself to such a simplistic remedy. Nevertheless, improving our understanding of these complex injury mechanisms offers a real potential for improved clinical outcomes in the near future. Ischemia/reperfusion injury is a particularly fascinating example of ros-mediated disease. When an organ is deprived of its blood supply (ischemia) it is injured, not just by the temporary loss of oxygen, but also by the ros that are generated by reaction with the oxygen that is reintroduced at reperfusion, when the blood supply is restored. In some clinical situations, we can prevent this injury by giving antioxidants, sometimes even after the period of ischemia, but just prior to reperfusion. For example, the preservation of kidneys, livers, and other organs in solutions that contain antioxidants, as well as other agents, is now routine prior to their transplantation. Another example is the use of drugs that block the function of free radical generating enzymes prior to stopping the heart for cardiac surgery. These drugs help

ROS and Human Health

prevent reperfusion injury when the heart is restarted and flow is restored. This reperfusion injury mechanism also has been found to play an important role in patients suffering from multiple organ failure after trauma, massive surgery, or shock. Multiple organ failure is now the leading cause of death in intensive care units, and extensive efforts are under way to understand better how ros contribute to this syndrome. Aging is a remarkably complex process that has managed to remain relatively opaque to scientific understanding. However,we now understand it to be a process per se, i.e., a series of controlled mechanisms, and not just the passive accumulation of wear and tear over the years. Put simply, our bodies age in the ways that are far more complex and more regulated than the processes by which our automobiles wear out. But if aging is a series of processes, its logical to conclude that it is potentially controllable, or at least modifiable. One of the most important of these processes is comprised of an accumulation of the molecular injuries that are mediated by free radicals and other ros. For example, since structural lipids are the primary component of our cell membranes, the integrity of which defines cell viability, aging is partially a matter of our going rancid as our lipids are progressively oxidized. While this is an oversimplification of this complex process, it reflects the optimism of some investigators of the aging process. Recent studies indicate that the therapeutic manipulation of ros metabolism can actually extend the total life span of mice to a significant degree. This was the first time that life span has been successfully altered experimentally by treatment. When one considers that the demographic, and consequent social, economic, and ecological impacts of even a 10 percent increase in human life span, a likely eventuality within the next decade or two, would far exceed that of a 100 percent cure for cancer (which is far less likely), the importance of this potential becomes evident.

Normal Biological Function

Normal Biological Function Given the large impact that ros can have on biological systems, it is not surprising that natural selection has found ways to use this reactivity to the organisms advantage. Accordingly, a number of biochemical mechanisms have evolved to do just this. For example, every high school science student learns that we use oxygen to burn our food to produce the energy we need to function. Effectively, the energy stored within the complex carbon molecules within our food is used to reduce oxygen to water and, in the process, to produce atp a molecular form of high energy storage which is used by the cell to run a whole host of functions. This process takes place within the mitochondria, intracellular organelles whose function it is to make this energy by a process termed oxidative phosphorylation. In the absence of oxygen, our cells are forced to make energy by anaerobic fermentation, a mechanism only one-third as efficient as aerobic oxidative phosphorylation. Interestingly, while students learn (correctly) that aerobic metabolism is normal, and the anaerobic pathway is a special situation, evolutionarily it is the other way around. For most of the period that life has existed on Earth (about two billion years) there were only negligible levels of oxygen in the atmosphere, and virtually every organism made energy by anaerobic fermentation. When oxygen first began to appear in the environment, a waste product of plant photosynthesis, it proved to be highly toxic and was literally of no earthly use. Indeed, evolutionary biologists tell us that the antioxidant defense mechanisms that we see in cells today evolved very early, in parallel with the appearance of oxygen in the environment. Many believe that the mitochondrial mechanism of oxidative phosphorylation evolved later, first as a highly efficient mechanism to scavenge these highly toxic oxygen molecules. (If oxygen were a new drug, it would be far too toxic to be approved by the US Food and Drug Administration.)

Normal Biological Function

Having evolved primarily as a defense mechanism against rosmediated tissue injury, the process of oxidative phosphorylation appears to have been exploited later by natural selection because it was more efficient for making atp. We are now so dependent on this high-efficiency system that humans and most mammals can survive only for brief periods without a continuous supply of oxygen. Free radicals are also an important component of the bodys defense systems. One of the most important determinants of survival in nature is the ability to avoid being eaten, not only by large carnivores, but by microorganisms in the course of an infection. In fact, more people die each year the world over from infectious diseases than from any other cause. Complex organisms, including humans, have highly developed immune systems that counter this threat. Major components of that defense system are macrophages, cells that circulate throughout the body and engulf and kill bacteria and other microorganisms by phagocytosis. The leukocytes (white blood cells) in our blood are an important example, although about 80 percent of the killing of circulating bacteria is effected by Kupffer cells, specialized macrophages within the liver. An essential component of this process is the killing of the bacteria that have been engulfed within a vacuole, termed the phagosome, without damaging the surrounding cell. The major mechanism of this killing is the generation of ros within the phagosome by a highly specialized enzyme on its inner membrane surface. The active, microbicidal product of this system is hypochlorous acid (hocl), the active ingredient of Clorox bleach. It is truly remarkable that we disinfect our bloodstreams with the same chemical that we use to disinfect our bathrooms. Here, natural selection has driven our bodies to employ opportunistically the toxic, destructive properties of ros to our advantage as a defense against microbial invasion.

Normal Biological Function

10

Weve also learned that ros play a critical role in cell signaling. We now know that organisms use very low fluxes of ros to send control messages within cells (intracellular signaling) and between cells (intercellular signaling) in the course of normal body function. While such signaling has long been appreciated, it traditionally has been thought to be effected exclusively by cytokines and other large, complex signaling molecules. Only recently have we come to realize that small ros can function in the same way. A prominent example of this is the molecule nitric oxide, actually a free radical, the function of which was the basis for the Nobel Prize in Physiology for Medicine in 1998. This tiny molecule is a major controller of blood flow and blood pressure throughout the body and also carries signals between nerve cells. It is the basis for the action of Viagra, which works through the control of blood flow through the penis. Signaling by ros, however, operates somewhat differently than more conventionally understood signaling. Cytokines are large molecules with relatively large, convoluted surfaces, with variations in electrical charge across the surface. These effect signaling by docking with receptors, complimentarily shaped and charged molecular surfaces on the target cells. The analogy is a key fitting into a lock and opening a door. This paradigm provides the essential element for cell signaling: specificity. If every cytokine activated every receptor the result would be chaos, not control. This need for specificity presents a real problem for those who believe that ros act as signaling molecules. Ros are small, relatively simple molecules that are so reactive that they often react with the first adjacent molecule they encounter, often over distances measured in angstroms and within nanoseconds. Indeed, their reactivity is promiscuous, not specific. Consequently, signaling specificity cannot be based upon molecular specificity. We have hypothesized that

Normal Biological Function

11

the necessary specificity of ros-mediated signaling is effected by subcellular compartmentalization, providing spatial proximity between the source of ros generation and its target. This is in fact how the organism protects itself against oxidant injury by endogenously generated ros: The compartmentalization of phagocytic microbicide within the phagosome is one example. Another is provided by the mechanism of energy (atp) generation by oxidative phosphorylation described earlier. Not only is this mechanism compartmentalized within the mitochondria, it is effected by each of the component molecules (cytochromes) being arrayed on the mitochondrial membrane, directly adjacent to one another to form the respiratory chain. This allows the unpaired electrons to be passed directly from molecule to adjacent molecule as oxygen is reduced to water. Because these electrons are so confined, they do not escape into the cytoplasm to react nonspecifically with other molecules. If the specificity of molecular signaling by ros is based upon subcellular compartmentalization, it presents real challenges to the investigator. Traditional biochemical techniques require the homogenization of cells and the separation of pools of particular components by differential density centrifugation of these components. Obviously, such an approach would destroy the very architectural structure that we wish to study. It would be analogous to try to understand how a Toyota is constructed by homogenizing a Toyota assembly plant and analyzing the homogenate for its composition of steel, rubber, plastic, oil, etc. Our current research centering on this problem of rosmediated signaling involves imaging these oxidant fluxes at the subcellular level in intact, living cells in (near) real time using advanced imaging techniques, including confocal fluorescence microscopy. We are optimistic that this approach will facilitate future studies of this kind.

Historical Developments in the Role of ROS

12

Historical Developments in the Role of ROS The evolution of our understanding of the role of free radicals and other ros in biology and medicine is a fascinating story in its own right. Physicists had long appreciated that violent free radical reactions were taking place within the sun and within a thermonuclear blast. Obviously, such extreme conditions are incompatible with any form of life as we understand it. As we began to control and use radiation for less violent purposes, we soon came to understand that some of the more subtle effects of radiation upon the body, including radiation sickness and carcinogenesis, were also caused by free radicals. Nevertheless, this conceptual framework was focused upon exogenous, physical/chemical sources of ros generation, and emphasized the destructive nature of their interaction with living things. Indeed, this mind set was so strong that the idea that such violently reactive molecules could be generated by living organisms was considered ridiculous. In 1969, when Joe McCord and Irwin Fridovich published a paper indicating a common, nearly ubiquitous intracellular protein (sod) functioned primarily as a free radical scavenger, thereby suggesting that the routine defense against endogenously generated ros was a major function of the cell, they were ridiculed by many of their colleagues. Today, their discovery is recognized as the basis of a whole new field of biology and medicine. It took about a decade for the scientific community as a whole to appreciate the importance of McCord and Fridovichs contribution, and this was greatly facilitated in 1980, when Neil Granger and his colleagues first described free radical-mediated reperfusion injury in the cat small intestine. From here, the pendulum swung rapidly toward the other extreme: Ischemic disease, especially heart attack and stroke, were (and continue to be) major threats to our lives, especially within industrialized

Historical Developments in the Role of ROS

13

countries where the drug markets are lucrative. Unfortunately, in many clinical situations the first opportunity to treat comes at the end of a period of ischemia but prior to reperfusion (e.g., cardiopulmonary resuscitation, cpr). If tissue has already died during the ischemic period, nothing can bring it back to life. But if much of the injury that had been heretofore ascribed to the period of ischemia was actually sustained at reperfusion, after treatment could be initiated, some of the damage might be prevented. The therapeutic (and economic) implications were enormous: Injuries thought to be irreversible at the point of treatment might not yet have occurred! Not surprisingly, a feeding frenzy ensued, catapulting a small cadre of previously obscure (and often ridiculed) free radical biologists into sudden prominence. Some of these reactions were truly excessive, some humorous, and some unfortunate. On the humorous side, I recall sitting at a scientific meeting in Point Clear, Alabama, in 1986 hosted by Neil Granger. Granger hoped to bring together those who were now working in the new field of reperfusion injury that he had largely initiated with his seminal study in 1980, which proposed a specific biochemical pathway for ros generation under these circumstances. Those of us, like Granger, who were interested in the rather obscure problem of intestinal ischemia were overwhelmed by the enthusiasm and interest shown by scientists and clinicians whose focus was on the far more fashionable problems of ischemia of the heart and brain. And now many of those who had ridiculed the idea earlier were fighting among themselves to claim credit for the discovery. The high point of this unseemly display was two separate speakers who each showed slides of Grangers previously published hypothesis as their own. I was appalled, but Granger, our host, was merely bemused. Fortunately, I directly followed these two speakers on the program to present some of our own work. This gave me

Historical Developments in the Role of ROS

14

the delicious opportunity to point out that I shared one thing in common with the two previous speakers: Their hypothesis had not been my idea originally. I then expressed regret that it appeared that I was alone in appreciating that fact. Far less humorous consequences included predictably premature clinical trials of sod for myocardial infarction. Fortunately, no one was harmed in these trials, but a great deal of money and effort was wasted. Over time, it has become clear that while the reperfusion component of post-ischemic injury is quantitatively quite important in some clinical situations, it is less so in many others, including heart attack and stroke. The real value of the work in this field has been, in my view, a far better understanding of basic physiology, upon which future advances in the treatment of disease will be based. Viagra is indeed a prototype for this developmental process. Even in the 1980s and early 1990s, our perspective of free radical biology emphasized their destructive capacity. I sometimes think we envisioned them as tiny blowtorches running around the body burning things, while our bodies chased after them with antioxidant fire extinguishers. The analogy clearly is appropriate for microbicide and reperfusion injury, but it is also too narrow to encompass the full range of the impact that ros have on biology and medicine. In the past few years the appreciation of the role of small fluxes of highly compartmentalized ros in controlling normal cell function has greatly expanded this horizon. What have we learned from this experience? Beyond the obvious examples of human weakness, some general observations appear appropriate. One is an appreciation of the strikingly different impact that an agent can have on biological systems, depending upon the dose. In massive quantities, ros can vaporize living things; in moderate quantities they can kill pathogens to our advantage; and in tiny quantities at precise locations they can

Historical Developments in the Role of ROS

15

act very precisely as controlling molecular switches. Similarly, it is perhaps human nature to classify agents in our environment as good or bad for us, but nature doesnt always make such distinctions. This mindset clearly has slowed our progress of understanding free radical biology in the past. Perhaps the most striking lesson we can take from this is the awesome, opportunistic power of natural selection to drive the evolution of complicated mechanisms such as the mitochondrial respiratory chain, phagocytic microbicide, and even cell signaling to exploit the violent reactivity of ros to the organisms survival advantage. This is truly mind-boggling. Finally, this narrative illustrates how biased and narrow-minded we scientists can be, just like other people. I frequently am amused by the lengths we must go to show the absence of a conflict of financial interest, or even the appearance of one, when peer-reviewing a paper for publication or a grant for funding. Most scientists are far more subject to bias driven by their preconceived ideas and their own pet ideas and hypotheses than by financial considerations. The study of free radicals and other ros in biology and medicine is endlessly fascinating and quite intellectually rewarding. It can carry one rapidly from the ridiculous to the sublime. At the ridiculous level are the imaginative advertisements for antioxidant supplements, for the diet, to rub on the skin, or even as an enema (!) to counteract the ravages of aging. They dont and cant work, at least not yet. At the other end of the spectrum is human sexuality, which owes a lot more to ros than Viagra. There is a species of yeast (Saccromyces pombe) that reproduces by asexual budding in its normal, anaerobic environment. This is efficient, but does not allow for the recombination of genes characteristic of sexual reproduction, which produces the genetic variations that are acted upon by natural selection to drive evolution. When these yeast cells are exposed to the normal levels of oxygen of an

Historical Developments in the Role of ROS

16

aerobic environment, they are severely stressed by the ros that are generated, and this triggers sexual reproduction. Helen Bernstein at the University of Arizona has suggested that this primitive organism represents the evolution of the sexual process itself by organisms in general, in response to the stress of the oxygen accumulating within our early atmosphere, allowing natural selection to drive the development of antioxidant defenses, as well as some of the other adaptations outlined here. If she is right, we could owe the invention of sex itself to free radicals and other ros, and certainly nothing could be more important than sex.
Originally published in COSMOS Journal, 2002. Gregory B. Bulkley, M.D. received his medical degree from Harvard Medical School. He retired as the Mark M. Ravitch Professor of Surgery from Johns Hopkins University School of Medicine. Dr. Bulkley is the author of over 300 books, scientific articles, book chapters and reviews.