Dealing with stress


  • Neil Kaplowitz

    Corresponding author
    1. Division of Gastroenterology, Keck School of Medicine, University of Southern California, Los Angeles, CA
    • Division of Gastroenterology, University of Southern California, Keck School of Medicine, 2011 Zonal Avenue, HMR101, Los Angeles, CA 90033
    Search for more papers by this author
    • fax: 323-442-3243

  • Potential conflict of interest: Dr. Kaplowitz consults for GlaxoSmithKline, Johnson & Johnson, Bristol-Myers Squibb, Roche, TEVA, Merck, Novartis, Hepregen Biotech, Daiichi-Sankyo, ISIS, and Enanta.

I am honored to contribute to this outstanding Master's Perspective series and wish to thank the Editors for inviting me. The colleagues who have previously contributed to this section of HEPATOLOGY have all been role models, and their stories are inspirational for me even at this late stage of my career. I will tell you my story and hope that younger investigators and physician–scientists will gain some insights into the approach to a career in academic medicine from my experiences.


AASLD, American Association for the Study of Liver Diseases; AP-1, activator protein 1; APAP, acetaminophen; ATF, activating transcription factor; Bcl2, B cell lymphoma 2; Bcl-XL, B cell lymphoma extra large; BSP, bromosulfopthalein; cFLIP, cellular FLICE-inhibitory protein; CHOP, CCAAT/enhancer-binding protein homologous protein; CREB, cyclic adenosine monophosphate response element-binding; CRTC, cyclic adenosine monophosphate response element-binding regulated transcription coactivator; CURE, Center for Ulcer Research and Education; DILI, drug-induced liver injury; ER, endoplasmic reticulum; ETC, electron transport chain; GSH, glutathione; GST, glutathione S-transferase; HLA, human leukocyte antigen; HSF, heat shock factor; IRE-1, inositol-requiring gene 1; JNK, c-Jun-N-terminal kinase; MAP, mitogen-activated protein; MAP2K, mitogen-activated protein kinase kinase; MAP3K, mitogen-activated protein kinase kinase kinase; MHC, major histocompatibility complex; MOMP, mitochondrial outer membrane permeability; MPT, mitochondrial permeability transition; NAPQI, N-acetyl-p-benzoquinone imine; NF-κB, nuclear factor kappa B; Nrf-2, nuclear erythroid 2 p45-related factor 2; PERK, protein kinase RNA-like endoplasmic reticulum kinase; PGC-1α, peroxisome proliferator-activated receptor γ coactivator 1α; P-JNK, phosphorylated c-Jun N-terminal kinase; ROS, reactive oxygen species; Sab, synthetic antigen binder; sXBP1, spliced X-box binding protein 1; TNF, tumor necrosis factor; UCLA, University of California Los Angeles; UPR, unfolded protein response; USC, University of Southern California.

Early Years

I was born in the Bronx in 1943. My father emigrated from Belarus in 1922, and my mother was born in New York, the child of German, Hungarian immigrants. They had no opportunity to receive higher education but very strongly encouraged me to do well in school and to read as much as possible. A turning point for me was the opportunity to attend Stuyvesant High School. This is a public school, and admission is determined competitively by entrance examination; the highest-scoring students were assigned to Stuyvesant or Bronx High School of Science, depending on the location of their residence. My class had approximately 700 students, every single one of whom went on to college; many of the students were brilliant, and the teachers were top notch and delighted to have such a motivated group. In high school, despite the emphasis on math and science, I developed an interest in art, particularly sculpting and drawing, and felt the joy of self-expression by using my imagination. After graduating, I attended NYU–University College in the Bronx, now an extinct campus. I comajored in French literature and Far Eastern history. My father wanted me to be an architect, but my mother convinced me to apply to medical school in my junior year just to see if I could get in. I reluctantly agreed, although torn by the possibility of spending my senior year abroad, but was swayed by the altruistic goal of helping to alleviate human suffering from disease. To my surprise, having only partially fulfilled the prerequisite science courses, I was accepted to NYU School of Medicine and was able to receive my bachelor's degree at the end of my first year of medical school. My acceptance was contingent upon taking organic chemistry in summer school. This course had a profound impact on me. I found the chemical reactions and syntheses, which are built logically on a set of facts, to be a thing of beauty.

Medical school began with the basic sciences. This was challenging for me as a liberal arts major with minimal science background. However, two individuals made major impressions on me. First, there was Professor Severoa Ochoa, a Nobel Laureate and Chair of Biochemistry. He gave almost half of the lectures in our biochemistry course. These were presented as experiments rather than facts and pathways. He would identify a question or hypothesis and tell us how he or others designed experiments to test the hypothesis and so on and so on. Even our examinations were structured on how to design an experiment that would prove or disprove a biochemical fact. The course was truly inspirational and gave me insight into the creativity involved in the approach to medical science. I learned that the creative process in science and art are very similar and that a career in scientific discovery might fulfill my need to find self-expression.

The second individual was Norman Javitt. Norm gave a few lectures in our clinical pathology course in year 2, which focused on the clinical chemistry of the liver, particularly pyrrole (bilirubin, porphyrins) and sterol (cholesterol/bile acid) metabolism. This was pure organic chemistry—a beautiful, logical series of chemical reactions modifying organic substances to change their physical chemical properties and elimination. Research in hepatology 45 years ago was mainly focused at the interface of organic chemistry and biology. I charged up to him at the end of the lecture and asked if I could do research with him. That was the beginning of a relationship in medical school and later in fellowship as I followed him to Cornell University and Rockefeller University, where Attallah Kappas was my comentor. This led to a number of my early publications on porphyrin and bile acid metabolism.1-4 Norm was a great mentor for my personality. He is rigorous, very critical, and a brilliant individual. I have always been a very curious, fiercely independent person who prefers to figure things out on his own. We were perfect together. He would say to me “Why don't you figure out why coproporphyrin isomer distribution in bile is different than urine or are there bile acid binding proteins in liver?” Then I would go off and develop an experimental approach and return with the data, which he would critique, and I would return to the laboratory and so on. These were wonderful years for me, because I was given the opportunity to develop my skills and approach experiments under the tutelage of a mentor who asked the tough questions. At the same time, I learned to integrate existing knowledge and use my imagination to develop my own ideas. The completion of a manuscript for publication which described the hypothesis, experimental results, and their interpretation gave me the same sense of fulfillment as painting a picture or sculpting a statue.

During the third year of my fellowship at Cornell, Ian Percy-Robb, a biochemist from Edinburgh, UK, came to Norm's laboratory at Cornell for sabbatical. We hit it off and worked closely together for a year. During that time, he helped me with my research project to discover bile acid binding proteins in the liver.

This brings me to the story of the fraction collector. In the early 1970s when doing column chromatography, we would manually collect timed fractions eluting from the column, which was very tedious. However, the neighboring laboratory had just purchased an automated fraction collector. Percy-Robb had taught me to perform gel filtration column chromatography to size-fractionate liver cytosol proteins. The mission I accepted from Norm was to find bile acid binders, so I decided to start by identifying the known bromosulfophthalein (BSP) organic anion binding proteins, called Y (ligandin) and Z (fatty acid binder) as a frame of reference. I added BSP to rat liver cytosol and loaded the column in the cold room, turned on the automatic fraction collector, and went home. Had I done this manually, I would have stopped after the last of the protein eluted, leaving the remainder of the excess unbound BSP to drain into a waste receptacle. When I returned the next day, there were hundreds of tubes. I added a drop of NaOH to each tube. BSP turns visible purple, and one can measure its absorbance at 580 nm wavelength. After doing this on hundreds of samples, I graphed (manually of course) the data expecting to see 2 peaks of bound BSP (Y+Z) followed by a peak of unbound BSP after the protein had eluted from the column. But to my surprise, there were two peaks of unbound BSP. This was very strange, but I remembered that my mentor had previously shown that BSP is conjugated with glutathione (GSH), and this is mediated by enzymes in the liver cytosol known as GSH S-transferases (GSTs). I checked the two unbound peaks by thin-layer chromatography, and indeed one contained BSP and the other contained BSP-GSH. I then assayed GST enzyme activity and found that the Y peak contained the enzyme.5

The story eventually turned out to be a bit more complicated as some of the family of GSTs also exhibit a nonsubstrate organic anion binding site for ligands such as bilirubin. Nevertheless, this was an example of fortuitous discovery which I owe to the fraction collector and curiosity. This event propelled me into the fields of biochemistry and physiology of GSTs6-8 and GSH metabolism. The study of these vital components of antioxidant defense eventually lead me into the fields of toxicology, drug toxicity, and oxidative stress. In other words, after some exploration of porphyrin metabolism, I was preparing for a career in bile acid metabolism, following in my mentor's footsteps, when chance turned me in a completely unforeseen direction.

After finishing fellowship training at Rockefeller University and Cornell University with Attallah Kappas and Norm Javitt, and 1 year on the faculty at Cornell, I spent 2 years (1973-1975) in research at the Oakland Naval Hospital under the Berry Plan. This was a great opportunity to set up my own lab and continue my research on the biochemistry of GSTs.6-8 There, I met John Kuhlenkamp, who has been my technician and laboratory manager to this day. Also, being close to University of California San Francisco gave me the opportunity to develop a friendship with Rudi Schmid, a charismatic figure and a giant in our field. His kindness and interest in me were very important and provided a certain level of mentorship for much of my career.

The Middle Years

In 1975, after completion of my military service, I took a position at University of California Los Angeles (UCLA) based in the Center for Ulcer Research and Education (CURE) at the Wadsworth Veterans Administration (VA) Hospital. John Kuhlenkamp accompanied me, and we set up a small laboratory with seed support and then spent the next 15 years building our research program. I was the only hepatology investigator at UCLA during that period, which suited my desire for independence. Nevertheless, I benefited greatly from the intellectual environment and outstanding scientists at CURE, such as Mort Grossman, Charles Code, Jon Isenberg, John Walsh, and George Sachs. The gastroenterology fellows during that era were amazing: Tachi Yamada, Ian Taylor, Andrew Soll, Steph Targan, Emeran Mayer, Fergus Shanahan, Andrew Stolz, Bob Gish, Laurie DeLeve, Stuart Sherman, and Shelly Lu, to name a few who have gone on to illustrious careers. I benefited tremendously from guidance and support of the CURE colleagues. Clearly, a young investigator's chances for success are greatly enhanced by working in an outstanding environment with role models, cutting edge science, and access to highly motivated trainees; I was fortunate that some of these talented gastroenterology fellows chose to work with me. Laurie DeLeve, Shelly Lu, and Andrew Stolz established their own successful independent research programs after joining me at the University of Southern California (USC) in Los Angeles. During this period, Murad Ookhtens, a Ph.D. in biomedical engineering, joined my group, which led to a lasting professional relationship and close friendship. Also, a series of postdoctoral fellows from Japan came to Los Angeles to work with me, and this connection with Japan continues.

Together with some of my early postdoctoral fellows, Yuichi Sugiyama, Motonobu Sugimoto, Hajime Takikawa, and Andrew Stolz, we did return to explore and discover bile acid binders and their role in bile acid metabolism and transport.9-14 However, I was really captivated with GSH metabolism and its role in detoxification, and this occupied much of my focus from the early 1980s (Fig. 1). The following individuals were among the postdoctoral fellows who worked with me in the GSH field over the years: Tachi Yamada, David Eberle, Tak Yee Aw, Murad Ookhtens, Jose C. Fernandez-Checa, Laurie DeLeve, and Shelly Lu. These individuals were a remarkably talented group of trainees who took up the challenges with intelligence, dedication, and resourcefulness as we discovered that GSH is transported into bile and blood,15-23 that compartmentation of GSH in mitochondria is dysregulated in alcohol-fed animals,24-28 that mitochondrial GSH is vital for defense against physiological and pathological oxidative stress, that the synthesis and transport of hepatic GSH is regulated by various hormonal factors,29-31 and that sinusoidal endothelial GSH is vital for defense.32, 33 Nearly all of these postdoctoral fellows have gone on to very successful careers in academic medicine and/or science. They were all wonderful to work with. Being a mentor is a two-way street. The fun of scientific discovery is greatly enhanced by sharing it with others. Mentorship is more than simply telling trainees what to do; it is sharing one's curiosity and the joy of exploring uncharted territory. I try to teach two other important things. First, the novel or original observation one makes is a small, albeit crucial, part of the scientific process; much of the work has to do with considering alternative explanations for the findings and thus the design of as many experiments as possible to try to prove oneself wrong. These are the controls. The second is that a “failed” experiment, one in which the results are completely opposite to what is expected, represents a unique opportunity to gain a much deeper understanding of what you are studying. Remember the tale of the fraction collector.

Figure 1.

Falk meeting on GSH in Basel, Switzerland, in 1992.

Transition to Stress Biology

I will digress here to provide a little conceptual context about stress as we understand it today. Disease-causing agents promote subcellular organelle dysfunction (e.g., cytoplasm, endoplasmic reticulum [ER], mitochondria, or nucleus) to varying degrees, which set in motion the activation of transcription factors and signal transduction pathways that are designed to facilitate adaptation to the stress. However, if the cell is unable to sufficiently adapt, the cell enters a fail-safe mode of self-destruction (Fig. 2). For any type of injury scenario, there are opposing forces battling to sustain or overcome the balance. This occurs at the interorgan, intraorgan, intercellular, cellular, and subcellular level. This concept of balance is now well recognized in the adaptive and innate immune systems, regenerative responses, and fibrogenesis. My particular research focus developed along the lines of organelle stress responses, and I will describe how and why this came about.

Figure 2.

Cell stress in defense and injury. Adaptive responses are activated by organelle dysfunction, leading to specific transcriptional programs designed to mitigate the stress. If balance cannot be successfully achieved, the cell moves into a default mode, promoting injury. CHOP,; Mitoch, mitochondria; ROS, reactive oxygen species.

As I look back over the four decades of my academic career, starting with my first faculty position at Cornell in 1972, I can divide my career into two eras. The first 20-25 years was a period when scientists spent the bulk of their time developing reagents, purifying proteins, studying kinetics, and raising antibodies. The second period to the present, especially the last decade, has been the most exciting for physician–scientists. The study of disease pathogenesis, development of animal models, and the ability to gain deeper insights have been greatly facilitated by the availability of the reagents and tools at our fingertips, and therefore my laboratory has been able to focus most of its activity on the study of regulatory pathways in disease pathogenesis. This transformation that occurred in the 1990s brings me to the recent part of my career and the subject of stress, particularly in the context of hepatotoxicity from drugs and alcohol. The transition began as I moved my laboratory across town to USC in 1990 (Fig. 3). I have always preferred to ask questions for which the tools were available to generate meaningful answers, so the advent in the past 15-20 years of polymerase chain reaction, western blotting, microarrays, conditional knockouts, and so forth has enabled me to return to earlier studies in acetaminophen (APAP) toxicity and alcoholic liver disease with new vigor and begin to deepen my understanding of disease pathogenesis.

Figure 3.

Kaplowitz Lab just arriving at the University of Southern California in 1990. Standing from right to left, Neil Kaplowitz, Hilo Yoshida, Hideki Saijki, Shelly Lu, Jose Fernandez-Checa, Andrew Stolz, Carmen Garcia-Ruiz, John Kuhlenkamp, technician, Tony Koo, Laurie DeLeve, and Murad Ookhtens. Seated right to left, H. Zeng, Linda Hammond, Terry Maddatu.

During most of my academic career, I have maintained a strong interest in drug-induced liver injury (DILI), as I realized that my interest in GSH metabolism was very relevant to DILI. Looking back at the development of the field of drug hepatotoxicity, Hy Zimmerman certainly was responsible for putting the spotlight on DILI, especially organizing the clinical and scientific knowledge into a working classification and conceptual approach. But a scientific milestone was the work of Mitchell and colleagues34 in the early 1970s, which appeared as I was first exploring GSTs as detoxifying enzymes and organic anion binding proteins. Mitchell's group showed that APAP toxicity depended on the metabolism to an electrophilic product, N-acetyl-p-quinoneimine I (NAPQI), which was detoxified by GSH; once GSH defense was exhausted, NAPQI was free to bind to protein thiols and this was believed to somehow cause necrosis, presumably by inhibiting the function of key proteins. This work was very exciting to me, because it showed the importance of GSH in defense against drug toxicity and the value of boosting or maintaining its levels with N-acetylcysteine (NAC). In addition, as the role of oxidative stress in many diseases began to emerge, this gave me great reassurance that studying GSH metabolism, cellular defense and redox biology was directly relevant to the physiology and pathophysiology of the liver.

Approximately 10-12 years ago, my interest in GSH redox extended to gene regulation and cell death. I was interested in whether the GSH redox state affects susceptibility to tumor necrosis factor (TNF)-induced apoptosis. Indeed, we found that even modest GSH depletion sensitized hepatocytes to killing by TNF and this was due to impaired transcriptional activity of nuclear factor–κB (NF-κB). A number of antiapoptosis genes are regulated by NF-κB accounting for normal resistance of the liver to TNF-induced apoptosis; one aspect is that some of the up-regulated NF-κB–dependent genes dampen or inhibit c-Jun N-terminal kinase (JNK) so that its activation is only transient. When we depleted GSH, this feedback shut-off of JNK was blocked so TNF induced apoptosis could proceed, at least partly because JNK activation was sustained. In these experiments, one of the approaches to depleting GSH which we employed was to expose primary mouse hepatocytes (PMH) to APAP. Inhibition of JNK blocked the TNF-induced apoptosis. As a control, we used APAP alone (no TNF) which, as expected, caused dose-dependent necrosis. However, to our great surprise, the JNK inhibitor protected against APAP-induced necrosis. This was a minor aspect of the article we published in HEPATOLOGY35 and other publications from our laboratory which dealt with the role of oxidative stress and GSH redox status in sensitization to TNF-induced apoptosis36-38 which in itself was of potential importance. However, I could not let go of the unexpected protection against APAP by the JNK inhibitor and suggested that postdoctoral fellows Basuki Gunawan and Derick Han explore this further. Indeed, either a small molecule inhibitor of JNK or antisense silencing of JNK1 and 2 markedly protected against APAP toxicity in vivo.39 We then showed that glycogen synthase kinase 3β (GSK3β) was upstream and that translocation of these kinases to mitochondria was occurring during the early phase of toxicity.40

Mitochondria have emerged as a critical organelle in generating ROS, which then play a key role in activating signal transduction pathways. Mitochondria were known to be a critical target of NAPQI. In the toxicology field, APAP-induced necrosis had been assumed to directly result from a combination of covalent binding and GSH depletion, particularly in mitochondria, leading to reactive oxygen species (ROS) and reactive nitrogen species (RNS) which induced mitochondrial membrane permeability pore transition (MPT) leading to a bioenergetic catastrophe and necrotic cell death. We found that activation and participation of signal transduction pathways were required in order for mitochondrial function to collapse and these pathways were downstream of covalent binding and GSH depletion. Indeed, this could be viewed as a unique type of programmed cell necrosis. One fascinating aspect of this work was that these signaling molecules (e.g., JNK) translocated to the mitochondria, and the binding of activated JNK appeared to worsen mitochondrial function leading to MPT, but only if mitochondria were made vulnerable by prior APAP treatment.41

With this hint that activated JNK binds to mitochondria, I asked another postdoctoral fellow, Sanda Win, to address the target of JNK binding in mitochondria and to determine if this interaction was necessary for toxicity to occur. We identified Sab,42 a protein described a decade ago to bind Bruton's tyrosine kinase in lymphocytes and P-JNK in fibroblast cell lines.43 First, we identified Sab exclusively in the outer membrane of liver mitochondria and then showed that activated JNK induced by APAP coimmunoprecipitated with Sab. We then knocked-down the expression of hepatic Sab in vivo and found that this markedly inhibited both APAP-induced necrosis and TNF/galactosamine-induced apoptosis. Silencing Sab prevented sustained JNK activation in both models.42 Another laboratory has recently generated complementary information in cell lines treated with anisomycin (JNK activator), demonstrating that isolated mitochondria from these cells are susceptible to activated JNK induced inhibition of the electron transport chain (ETC) which was prevented by silencing Sab.43 Our current working model is that Sab serves as a platform for assembly and activation of mitogen-activated protein (MAP) kinases and that this interaction promotes sustained ROS by mediating an inhibitory effect on electron transport; the sustained ROS then promotes sustained MAP kinase activation in a self-amplifying cycle (Fig. 4). During the past 10-12 years, the work in my laboratory on the pathogenesis of hepatotoxicity has been conducted by a sequence of outstanding postdoctoral fellows starting with Guopeng Feng, Hide Nagai, Katsu Matsumaru, Basuki Gunawan, Naoko Hanawa, and Mie Shinohara, who each spent 2 years with me and Cheng Ji, Derick Han, Huan Lou, Zhang-Xu Liu, Tin Aung Than, and Sanda Win, who have remained in the research group after postdoctoral training, and the current postdoctoral fellows, Lily Dara and Behnam Saberi (Fig. 5). This talented, dedicated group of young investigators make it fun for me to continue to pursue the science of hepatotoxicity as we explore new and often unexpected directions. As I noted above, the joy of scientific exploration is an important part of an academician's career. However, sharing this with trainees and young colleagues and helping them move to independent careers is equally satisfying.

Figure 4.

Model of JNK/Sab in hepatotoxicity. NAPQI or TNF in the presence of inhibition of NF-κB (galactosamine or actinomycin D) inhibit electron transport (ETC), promoting release of ROS from mitochondria. This leads to activation of MAP kinase kinase kinase (MAP3K), e.g., ASK-1, which in turn activates MAP2K (e.g., MKK4) which activates JNK (thin black arrows). P-JNK then binds to Sab which leads to further impairment of ETC and enhanced ROS production sustaining a high level of JNK activation in a self-amplifying cycle (thick red arrows). In the case of APAP, profound GSH depletion renders the mitochondria more susceptible to ROS leading to mitochondrial permeability transition (MPT) pore opening which induces a bioenergetic catastrophe causing the necrotic cell death characteristic of APAP toxicity. In contrast, in the case of TNF/galactosamine or TNF/actinomycin the sustained JNK activation promotes a proapoptotic combination of events mediated by the Bcl-2 family: Bax and Bim translocation to mitochondria, inhibition of Bcl-XL, t-Bid formation and translocation as a consequence of proteosomal degradation of cFLIP (caspase-8 inhibitor), all of which favors mitochondrial outer membrane permeability (MOMP). Adapted from Win et al.42

Figure 5.

Current research group. From left to right: Maria Ybanez, Michelle MacViegh-Aloni, Behnam Saberi, Zhang-Xu Liu, Lily Dara, Derick Han, Neil Kaplowitz, Cheng Ji, Sanda Win, John Kuhlenkamp, Tin Aung Than, Murad Ookhtens (missing, Huan Lou).

My longstanding interest in the pathogenesis of liver injury extended over the years to alcohol as a causative agent. This began back in the mid-1980s, when Jose Fernandez-Checa joined me as a postdoctoral fellow. We explored GSH metabolism in alcohol-fed rats and found that there was an impairment in the uptake of GSH by mitochondria.24-28 After 7 years working together at UCLA and USC, he moved back to Spain and has had a very successful career. We have continued to collaborate over the years on aspects of the role of mitochondrial GSH in alcoholic liver disease.44-51 Also, Hide Tsukamoto has had a very large impact on our work. He is an outstanding colleague who established an Animal Model Core at USC which has allowed us to perform many studies in the intragastric alcohol-fed mouse model. Then, Cheng Ji joined me as a postdoctoral fellow more than a decade ago. Cheng is an outstanding molecular biologist, so we decided to fish for new clues for pathogenesis by examining gene expression profiles in intragastric alcohol and pair-fed mice. To our surprise, a set of genes reflecting a response to ER stress stood out.52 Very little was known about ER stress response at the time but so this was a new finding and seemed important. Although we focused on alcohol, many others joined this field studying the role of ER stress in the metabolic syndrome and NAFLD/NASH. Cheng Ji has now successfully achieved independence, but we continue to collaborate as we try to understand how alcohol induces ER stress, particularly with respect to the role of one carbon metabolism and homocysteine, and how ER stress contributes to liver disease with respect to steatosis, cell death, fibrosis, and liver cancer.53-57 The latter work has been greatly facilitated by collaborations with our colleague Amy Lee and the creation of the liver-specific GRP78 knockout mouse model which spontaneously develops steatohepatitis and is very susceptible to alcohol, overfeeding, and drug/toxin-induced liver disease.56 ER stress response is an example in our view of a protective, adaptive response (unfolded protein response, UPR), which if overwhelmed or impaired, progresses to a fail-safe lethal outcome. However, along the way it plays a key role in dysregulating lipid metabolism and promoting insulin resistance.

So you can see that for many years we have been attempting to understand the role of various types of stress responses in the protection or worsening of liver injury. Our early efforts focused on oxidative stress and GSH defense, then addressed stress responses in the cytoplasm (e.g., MAP kinases and other yet-to-be-published signaling pathways) and ER. Having extensively explored the ER stress response, I turned to another talented postdoctoral fellow, Tin Aung Than, approximately 5 years ago, and asked him to consider whether something analogous to the UPR in response to ER stress could exist in mitochondria in response to mitochondrial stress. Because an aspect of UPR is the activation of a transcriptional program which promotes the production of more ER as well as chaperones to compensate for demand, we decided to focus on mitochondrial biogenesis in response to mitochondrial specific stress induced by a toxin, rotenone, which inhibits complex I of the ETC. It is known that the induction of peroxisome proliferator-activated receptor γ coactivator-1α (PGC-1α) serves as a master activator/regulator of mitochondrial biogenesis. Thus, we set up a model in PMH and HepG2 cells and showed that rotenone at nonlethal doses induced mitochondrial biogenesis and this was blocked by PGC-1α silencing. Thus, the question is what is the retrograde signal from mitochondria which activates the expression of the nuclear PGC-1α gene. PGC-1α is regulated by cAMP-response element-binding protein (CREB), a cyclic AMP responsive transcription factor. cAMP and other signals are known to promote the dephosphorylation and nuclear translocation of CREB-regulated transcription coactivator 2 (CRTC2) and CRTC3, a pair of transcriptional coactivators which activate CREB-mediated transcription of PGC-1α. Most knowledge in this area concerns CRTC2.

We hypothesized that CRTC2 and CRTC3 might respond differently to a mitochondrial-specific stress compared to a metabolic stress such as fasting. Indeed, we found that a small fraction of CRTC3 is present in mitochondria, but not CRTC2. We compared the effects of silencing expression of CRTC2 versus CRTC3 in our model of PGC-1α–dependent mitochondrial biogenesis in response to rotenone. To our surprise, the response required CRTC3 whereas CRTC2 was dispensable.59 We continue to explore the mechanism for the selective role of CRTC3 in the mitochondrial stress response. The point is that unique organelle stress responses appear to exist which depend on compartmentation of transcription factors and signaling molecules which elicit adaptation and fail-safe mechanisms designed to deal with the location of stress. In the case of mitochondrial biology and the role of mitochondrial stress in adaptive and injurious responses, we and others are moving toward trying to understand not only the role of mitochondrial biogenesis, but the regulation and importance of removal of damaged mitochondria, which eliminates a major source of ROS and apoptosis-inducing proteins by autophagy and mitophagy, as well as the contribution of fission and fusion of mitochondria. The field is on the doorstep of major advances in understanding the contribution of the execution versus impairment of these processes in determining the susceptibility to and pathogenesis of liver disease.

In conclusion, my story illustrates one individual's approach to a career as a physician-scientist. Whatever degree of success I have had is not the point for me. If some measure of contribution to the movement of the field has resulted, that is wonderful. However, the process of exploration in order to satisfy my curiosity and creating pictures that interpret what has been observed is what I find fulfilling. Interpretation of experimental findings is just that, namely, a mixture of reality and fantasy which is always in need of refinement and more detail. Along the way, the privilege and good fortune of working with talented people greatly amplifies the fun, and one hopes it makes some impression on them in the way that my early experiences did for me. Although I now am in the ranks of senior citizens, both literally and figuratively, remaining in the thick of the process of investigation and training continues to energize and excite me. I greatly appreciate the support of the Department of Veterans Affairs, National Institute of Diabetes and Digestive and Kidney Disease (NIDDK), and National Institute on Alcohol Abuse and Alcoholism over the years in allowing me and my associates to conduct our research.

Idiosyncratic DILI

Throughout my career, I have maintained an active interest in the clinical aspects of DILI, following the literature, consulting for pharmaceutical companies where I have gained a broad appreciation for the impact and manifestations of DILI in drug development, writing reviews, editing a textbook (Drug-Induced Liver Disease) with Laurie DeLeve (third edition), and addressing questions of interest in the Spanish registry of DILI cases in collaboration with Raul Andrade and Maribel Lucena.60-62

During the past decade, as we developed an appreciation of the signaling pathways involved in APAP toxicity, the field of drug-induced liver disease has seen considerable progress and our concepts of idiosyncratic hepatotoxicity have transformed from metabolic idiosyncrasy championed by Zimmerman to the notion that most examples of idiosyncratic DILI are mediated by the adaptive immune system responding in genetically susceptible individuals to reactive drug metabolites or parent drugs leading to either drug specific hypersensitivity or spreading to autoimmunity. This view is predominantly based on genetic susceptibility studies which have revealed strong human leukocyte antigen (HLA) (major histocompatibility complex 1 [MHC1] and MHC2) associations. Nearly all the recent examples of idiosyncratic DILI have such associations: flucloxacillin, lumiracoxib, lapatanib, ticlodipine, amoxicillin-clavulanate, and ximelagatran.63-68

Having been a follower of Zimmerman's concepts, which were very important in giving us a framework to approach the field of DILI, I viewed idiosyncratic DILI occurring in the absence of systemic allergic features to be metabolically based. After all, there was usually a higher incidence of mild asymptomatic ALT elevations than overt disease (Temple's corollary), adaptation with continued use, negative rechallenge and no systemic allergic features. However, as the examples cited above have accumulated and the evidence for polymorphisms in toxification and detoxification genes as major determinants of susceptibility have not materialized, I along with most workers in the DILI field have been forced to reconsider and accept the strength of the evidence supporting the hypothesis that idiosyncratic DILI is usually determined by adaptive immune responses. Certainly, absolute proof is still lacking but it is impossible to deny the evidence. Lumiracoxib and ximelagatran are compelling examples. The mild toxicity seen more frequently, as well as the overt Hy's law cases, only occurred in those with a specific HLA association. Then, as Uetrecht suggests, adaptation with continued drug use may be due to the development of immune tolerance and progression to overt disease may be a failure to develop tolerance.69 This is all quite hypothetical at present but provides the framework for a field poised to make progress based on this conceptual shift. Clearly, the new view can be considered perhaps overly simplistic as we have no understanding of what accounts for different phenotypes (cholestasis versus hepatitis) and most importantly we still have no explanation for why rare liver injury occurs in the presence of common polymorphisms. So despite strong associations, most subjects with the HLA susceptibility do not develop liver injury. One can view the development of liver injury via an adaptive immune response in a small fraction of genetically susceptible individuals to be determined by the regulation of a self-amplifying injury process modulated by adaptive and injury-promoting imbalances (Fig. 6). Obviously, the situation is very complex. Nevertheless, these recent examples of delayed idiosyncratic DILI demonstrate that the absence of the HLA allele associated with susceptibility is very strikingly predictive of little or no risk of toxicity.

Figure 6.

Idiosyncratic DILI. Genetically determined adaptive immune response triggers a self-amplification cycle that promotes liver injury. Usually, sufficient exposure to drug metabolite and the appropriate MHC class molecules are prerequisites for the development of idiosyncratic DILI, but the actual occurrence and severity of hepatotoxicity are likely to be determined by a variety of downstream adaptive and injurious responses. A self-amplifying injury-propagating cycle can be envisioned that could be modulated by genetic and environmental factors.

Clinical Medicine and Administration

Finally, I would like to touch on a few ancillary but important aspects of this tale, namely the other sides of academic medicine: clinical practice/bedside teaching and administrative work at both local and national levels. We academicians are fortunate to face choices with respect to the degree to which we emphasize laboratory or clinical research, practice of medicine, teaching in various contexts, and leadership. The emphasis placed on each of these usually changes over the course of a career, and all can be extremely fulfilling. I believe that each of us should seek fulfillment according to our personal strengths, weaknesses, and interests, while continually attempting to find the right balance. I have always enjoyed conducting teaching rounds and dealing with complex diagnostic and therapeutic problems, particularly in patients with liver disease. This has been very important for me, as it keeps me grounded and in touch with the major issues in hepatology as they relate to my own laboratory research. Although I have conducted only a modest amount of patient-related research, over the years the clinical problems I have encountered during teaching rounds have catalyzed a number of clinical descriptions that attempted to organize and describe what we encountered during teaching rounds.70-73 I view these as recreational activities, and I continue to dabble in this way. As for administration and leadership, this is a personal choice each of us must make. My advice is to remember the Peter Principle—avoid ascending to your level of incompetence—and at the same time recognize that one cannot avoid the impact that these activities have on your time and freedom to conduct research. Many colleagues who I admire have been able to ascend the ladder of academic administrative leadership and continue to do science. I made the conscious decision to resist the temptation to become a Chairman or Dean, but to stop at Division Chief and to take on this job in a centrally run Department of Medicine, which I have been doing at USC since 1990. Because of the central organization of our Department of Medicine, I have been largely free of the business and financial side of being a Division Chief and have mainly dealt with the academic side, such as recruitment, fellowship training, and creating and sustaining a supportive and intellectually stimulating environment.

I continue to write papers and grant proposals and spend a great deal of my “free” time reading the medical and scientific literature. In today's world with the fast and furious rate of progress, this is mandatory. One needs to make time for reading, writing, and thinking. I benefit from learning how basic scientists in cardiology, nephrology, and neurosciences address similar issues as I do in the liver and firmly believe that there is an overwhelming commonality in our work.

One aspect of my career has linked science and administration and has been especially rewarding, namely the establishment of the Research Center for Liver Disease at USC, which has been funded by NIDDK since 1995. This has provided vital Core Facilities, attracted many scientists, stimulated collaboration, and provided pilot project funding. We have approximately 50 center members, who along with me, depend on the technologies, both routine and highly specialized, provided by the center to enable our research productivity and efficiency. I look forward to reapplying for the 2015-2020 cycle in a few years.

Finally, on a national level, I encourage participation in AASLD and the American Gastroenterological Association. Serving on the Governing Board and the Presidency of AASLD were professional highlights for me and forged friendships with exceptionally talented colleagues (Fig. 7). Working for the benefit of one's peers at the local and national level has been a way to give back what I have received from others and help create a better environment. AASLD's contributions to scientific exchange through large and small meetings, its journals, practice guidelines, lobbying, and fundraising on behalf of young investigators have an enormous impact. It has been a privilege for me to contribute to AASLD, along with so many of you. Ultimately, the benefits of AASLD's activities undoubtedly have and will continue to trickle down to our patients and their lives will be better. After all, that is why I became a physician. My career path was determined by personal preferences, but never losing sight of the hope that what I love doing might eventually help in some small way to alleviate human suffering.

Figure 7.

AASLD Governing Board, circa 1995. Standing (left to right): Jay Hoofnagle, Steve Mezey, Caroline Riely, Monty Bissell, John Gollan, Bill Balestreri, Steve Strasberg, and Joe Bloomer; seated (left to right): Will Lee, Gene Schiff, Tony Tavill, Nick LaRusso, Marion Peters, and Neil Kaplowitz.

In conclusion, I am indebted to all the wonderful mentors, colleagues, fellows, and students who have participated with me in my journey. It is a story of friendships, optimism, curiosity, and perseverance. Most of all, I am grateful to my loving wife Fattaneh, for her support and understanding through the years. This exercise has given me the rare opportunity to look back over my career and actually try to understand what determined its evolution. In retrospect, one might think it was planned out, but truthfully, it represents a long series of fortunate events.

Chance favors the prepared mind.