Understanding Autophagy to Enhance Clinical Discovery: The 2016 Dr. Paul Janssen Award Symposium

Resources

Anding AL, Baehrecke EH. Vps15 is required for stress induced and developmentally triggered autophagy and salivary gland protein secretion in Drosophila. Cell Death Differ. 2015 Mar;22(3):457-464.

Castermans D, Somers I, Kriel J, Louwet W, Wera S, Versele M, Janssens V, Thevelein JM. Glucose-induced posttranslational activation of protein phosphatases PP2A and PP1 in yeast. Cell Res. 2012 Jun;22(6):1058–1077.

Hait WN, Versele M, Yang J-M. Surviving metabolic stress: of mice (squirrels) and men. Cancer Discov. 2014 Jun;4(6):646–649.

Howden EJ, Sarma S, Lawley J, Samels M, Palmer D, Everding B, Livingston S, Levine B. Dose matters: effect of two years of intensive supervised endurance training on aerobic capacity: 2446 June 3, 9:30 AM – 9:45 AM. Med Sci Sports Exerc. 2016 May;48(5 Suppl 1):671.

Kaushik S, Cuervo AM. AMPK-dependent phosphorylation of lipid droplet protein PLIN2 triggers its degradation by CMA. Autophagy. 2016;12(2):432–438.

Lee MY, Sumpter R, Zou Z, Sirasanagandla S, Wei Y, Mishra P, Rosewich H, Crane DI, Levine B. Peroxisomal protein PEX13 functions in selective autophagy. EMBO Rep. 2016 Nov 8.

Lin L, Baehrecke EH. Autophagy, cell death, and cancer. Mol Cell Oncol. 2015 Sep;2(3):e985913.

Madrigal-Matute J, Cuervo AM. Regulation of liver metabolism by autophagy. Gastroenterology. 2016 Feb;150(2):328–339.

Moore CEJ, Wang X, Xie J, Pickford J, Barron J, Regufe da Mota S, Versele M, Proud CG. Elongation factor 2 kinase promotes cell survival by inhibiting protein synthesis without inducing autophagy. Cell Signal. 2016 Apr;28(4):284–293.

Sumpter R, Levine B. Novel functions of Fanconi anemia proteins in selective autophagy and inflammation. Oncotarget. 2016 Jul 30.

Suzuki SW, Yamamoto H, Oikawa Y, Kondo-Kakuta C, Kimura Y, Hirano H, Ohsumi Y. Atg13 HORMA domain recruits Atg9 vesicles during autophagosome formation. Proc Natl Acad Sci USA. 2015 Mar 17;112(11):3350–3355.

Tasset I, Cuervo AM. Role of chaperone-mediated autophagy in metabolism. FEBS J. 2016 Jul;283(13):2403–2413.

Tracy K, Velentzas PD, Baehrecke EH. Ral GTPase and the exocyst regulate autophagy in a tissue-specific manner. EMBO Rep. 2016 Jan;17(1):110–121.

Yamamoto H, Fujioka Y, Suzuki SW, Noshiro D, Suzuki H, Kondo-Kakuta C, Kimura Y, Hirano H, Ando T, Noda NN, Ohsumi Y. The intrinsically disordered protein Atg13 mediates supramolecular assembly of autophagy initiation complexes. Dev Cell. 2016 Jul 11;38(1):86–99.

Yamamoto H, Shima T, Yamaguchi M, Mochizuki Y, Hoshida H, Kakuta S, Kondo-Kakuta C, Noda NN, Inagaki F, Itoh T, Akada R, Ohsumi Y. The thermotolerant yeast Kluyveromyces marxianus is a useful organism for structural and biochemical studies of autophagy. J Biol Chem. 2015 Dec 4;290(49):29506–29518.

Speakers

Yoshinori Ohsumi, PhD

Tokyo Institute of Technology
website | publications

Eric H. Baehrecke, PhD

University of Massachusetts Medical School
website | publications

Eric H. Baehrecke obtained his PhD from the University of Wisconsin–Madison, and was a Howard Hughes Medical Institute Fellow of the Life Sciences Research Foundation at the University of Utah during his postdoctoral studies. He was on the faculty of the University of Maryland from 1995–2007, and is currently a professor in the Department of Molecular, Cell and Cancer Biology at the University of Massachusetts Medical School. His team studies the regulation and function of autophagy in cell health and cell death.

Ana Maria Cuervo, MD, PhD

Albert Einstein College of Medicine
website | publications

Ana Maria Cuervo is the Robert and Renee Belfer Chair for the Study of Neurodegenerative Diseases, professor in the Departments of Developmental and Molecular Biology and of Medicine at the Albert Einstein College of Medicine, and co-director of the Einstein Institute for Aging Studies. She obtained her MD degree and a PhD in biochemistry and molecular biology from the University of Valencia (Spain) and received postdoctoral training at Tufts University, Boston. At the Albert Einstein College of Medicine, Cuervo's group is interested in understanding how altered proteins can be eliminated from cells through the lysosomal system (autophagy) and how malfunction of autophagy in aging is linked to age-related disorders, such as neurodegenerative and metabolic diseases.

Cuervo has received the P. Benson Award in Cell Biology, the Keith Porter Fellow in Cell Biology, and the Saul Korey Prize in Translational in Medicine Science, among others. She is currently co-editor-in-chief of Aging Cell and associate editor of Autophagy.

Beth Levine, MD

University of Texas Southwestern Medical Center, Howard Hughes Medical Institute (HHMI)
website | publications

Beth Levine received an AB from Brown University, an MD from Cornell University Medical College, and completed her postdoctoral training in infectious diseases/viral pathogenesis at the Johns Hopkins University School of Medicine. In 1993, she joined the faculty at Columbia University College of Physicians & Surgeons where she became associate professor of medicine. In 2004, she became the Jay P. Sanford Professor and chief of the Division of Infectious Diseases at UT Southwestern Medical Center. In 2011, she became the director of the Center for Autophagy Research at UT Southwestern and the Charles Cameron Sprague Distinguished Chair in biomedical science. Since 2008, she has been a Howard Hughes Medical Institute Investigator.

Levine has received numerous awards, including the 2008 Edith and Peter O'Donnell Award in Medicine from The Academy of Medicine, Engineering and Science of Texas, election to the National Academy of Sciences in 2013, and the 2014 American Society of Clinical Investigation Stanley J. Korsmeyer award.

Matthias Versele, PhD

Janssen Research & Development
website | publications

Matthias Versele is a biology team leader in the oncology discovery group in Beerse (Belgium). He currently leads drug discovery teams focusing on hematological malignancies, primarily B-cell lymphomas and multiple myeloma; he is involved in identifying new drug targets for the hematologic disease area stronghold; and he is the preclinical lead in support of the clinical development of Ibrutinib (Imbruvica). His main research interests are kinase signaling and stress response mechanisms relevant to tumor biology (unfolded protein response, autophagy).

Versele joined Janssen in 2006, after his post-doctoral training at UC Berkeley, working on the molecular mechanisms through which septin filament formation is coordinated with cell cycle progression and cell morphology. Versele completed his PhD in biochemistry at the KU Leuven (Belgium) in 2000, studying nutrient sensing and signaling mechanisms controlling stress resistance and cell growth.

Introduction

Thirty years ago, textbook illustrations of single-celled eukaryotes such as yeasts included large, empty holes called vacuoles. Even the name suggested nothing interesting to be found inside. Researchers understood that all cells continuously broke down and recycled proteins, nucleic acids, and other structures through a process dubbed autophagy, and that the vacuole was involved in that process in yeast. Yet they had no clear sense of how the process unfolded.

The last three decades have seen dramatic progress in the field, with profound implications for possibly treating many chronic diseases.

In 2016, Yoshinori Ohsumi won the Nobel Prize for his research on autophagy. Ohsumi began studying the process almost 30 years ago. Using yeast as a model system, Ohsumi and his colleagues identified the family of genes responsible for autophagy, and characterized many of the underlying mechanisms driving it. Because the autophagy machinery is highly conserved, those findings were directly relevant to subsequent work in animals and humans.

Beth Levine of the University of Texas Southwestern Medical Center drove that point home by describing her discovery of Beclin-1, a mammalian protein that is a molecular and functional homolog of the yeast autophagy regulator Atg-6. Beclin-1 integrates a collection of cellular signals to stimulate autophagy in response to stress. That response is critical for longevity and immunity in animals and humans, and defects in the Beclin-1 pathway accompany numerous diseases. Many of the benefits of regular exercise appear to result from stimulating autophagy. Levine's team is now developing drugs that target Beclin-1 to produce the same benefits for patients unable to exercise.

Eric Baehrecke of the University of Massachusetts Medical School studies how mitochondria get targeted for degradation and replacement. Baehrecke and his colleagues have found extensive autophagy of mitochondria, or mitophagy, during the development of the fruit fly intestine. Using that model, the researchers identified a specific protein that appears to link the process of mitochondrial fission to autophagy, ensuring the selective degradation of defective mitochondria. Mitophagy is also a prominent feature of neurodegenerative diseases such as Alzheimer's and Parkinson's.

Ana Maria Cuervo discussed the chaperone system. Chaperone proteins bind other proteins and target them for destruction, often by interacting with the autophagy system. Using an extensive collection of mouse models, Cuervo has shown that a protein called LAMP-2A controls chaperone-mediated autophagy. LAMP-2A levels normally decline with age, but a high-fat diet accelerates that decline, and models of neurodegenerative diseases such as Alzheimer's, Huntington's, and Parkinson's all feature LAMP-2A defects as well. Cuervo and her colleagues are now searching for drugs that can regulate chaperone-mediated autophagy, in the hope of treating many of these conditions.

Matthias Versele brought the discussion neatly back to the beginning. Versele initially sought to target protein elongation, a process tumor cells routinely inhibit in order to survive nutrient shortages—boosting it should hasten those cells' deaths. However, his team found that the protein elongation factor EF2 interacts with autophagy mediator Beclin-1. That led the researchers to identify compounds that can inhibit one or both of the two pathways, to attack tumors through either protein elongation or autophagy, at different points in their growth.

The meeting concluded with a wide-ranging panel discussion. All of the speakers highlighted the seminal importance of Ohsumi's work in founding the field and driving many of its most important early discoveries. On the subject of his own contributions, Ohsumi simply pointed to the huge range of mysteries that still remain: "I believe we are still at an early stage of autophagy study, there are so many fundamental questions to be answered."

Highlights

Studies in yeast revealed the mechanisms of autophagy, which recycles proteins and organelles.
Highly conserved homologs of yeast autophagy proteins serve similar functions in higher organisms.
Autophagy defects appear to affect longevity and drive numerous diseases in humans.

You eat what you are

Every day, an average human body manufactures billions of new cells, recycling a similar number of old ones in the process. At the molecular scale, we make about 240 grams of new proteins while consuming only 70 grams in food, relying on recycling to make up the difference. This vigorous turnover isn't unique to humans, or even to animals. Leaves turn brown in autumn as plants devour their own photosynthetic machinery, storing the amino acids in roots or seeds. Yeast cells destroy old proteins and synthesize new ones to adapt to environmental changes. "Life is just an equilibrium state of synthesis and degradation," said Janssen Award recipient Yoshinori Ohsumi, whose presentation anchored the symposium.

Researchers began to understand protein turnover in the mid-20th century. The discovery of a targeted protein degradation pathway—in which ubiquitin binds proteins and directs them to a degradation complex called the proteasome—accelerated that work. However, the proteasome is a selective system for removing damaged or unneeded proteins; the bulk of protein turnover happens through the more general autophagy pathway in the cellular lysosome. The lack of good assays for tracking autophagy made it much harder to study.

Ohsumi realized that the budding yeast Saccharomyces cerevisiae was an ideal model organism for deciphering the underlying mechanisms of autophagy. Yeast cells have two major physiological states: vegetative growth and sporulation. When food is abundant, the cells grow vegetatively, but a scarcity of nitrogen drives them to initiate sporulation. Because sporulation occurs during nitrogen starvation, when the cells can't manufacture fresh amino acids, it must involve extensive protein recycling.

Capitalizing on the power of yeast genetics, Ohsumi and his colleagues looked at mutant cells with deficient proteinases. When starved, these cells developed polka dot patterns of membrane-bound bodies inside their vacuoles. "Just a portion of cytoplasm is taken up by these single membrane bound structures, and we named these structures autophagic bodies," said Ohsumi. The autophagic bodies contain proteins, molecular assemblies, and even whole organelles that have been targeted for recycling. In wild-type cells the proteinases break down the autophagic bodies' contents too quickly to see, but in the mutants they accumulate and become visible.

Next, the investigators screened for additional mutations that would prevent the autophagic bodies from forming in the proteinase mutant background. This revealed fourteen genes, now labeled Atg-1 through Atg-14, involved in autophagy. All of the genes were novel; no other labs had found homologies or functions for anything like them, probably because the mutants only reveal their defects during nitrogen starvation.

Ohsumi's team ultimately found that the Atg genes undergo a lengthy cascade of interactions, leading to the formation of a multiprotein complex that activates autophagy. Homologs of the yeast Atg genes have since been found throughout eukaryotes, with their sequences highly conserved even in mammals. "This means the autophagic machinery [is a fundamental] development in the evolution of the eukaryotic cell," said Ohsumi.

On the heels of these initial discoveries, research on autophagy has blossomed. Scientists have since found that the same fundamental mechanism helps control longevity, immunity, development, and tumor suppression. Defects in autophagy lie at the roots of numerous human diseases.

Meanwhile, Ohsumi's laboratory has continued to study the details of autophagy at the cellular level. That work has identified the Atg-13 protein as a critical regulator of the process. Under starvation conditions, cells dephosphorylate Atg-13, causing it to bind additional Atg proteins to form a large pre-autophagosomal structure, or PAS. The team is now studying how the PAS drives the formation of autophagic bodies, and exploring the range of conditions that can induce autophagy.

While autophagy has been easiest to study in yeast growing under starvation conditions, Ohsumi emphasized that it could be induced by a variety of stimuli in different organisms, leading to subtly different outcomes. "There are many distinct modes of autophagy, we shouldn't say it's only one [thing]," said Ohsumi.

Highlights

The mammalian Beclin-1 protein is molecularly and functionally homologous to the yeast Atg-6 autophagy regulator.

Drugs targeting Beclin-1 could help treat a wide range of disorders.

Gut development in fruit flies provides a good model for studying selective autophagy.

Survival of the fit

Beth Levine began her scientific career as a virologist. But in the course of her work trying to understand how the anti-apoptotic protein Bcl-2 protected mice against a lethal central nervous system infection, screening for molecules that interact with Bcl-2 revealed a novel protein Levine named Beclin-1. Beclin-1 shares about 25% identity at the amino acid level with the yeast Atg-6 protein, which Ohsumi had previously identified as an autophagy regulator.

After contacting Ohsumi about the finding, Levine quickly received a sample of yeast lacking the Atg-6 gene. She found that expressing human Beclin-1 in these cells rescued their ability to carry out autophagy, confirming that human Beclin-1 doesn't just look like an autophagy regulator, it is one. Human cancer cells with low levels of Beclin-1 expression also fail to boost their autophagy rates like wild-type cells do when cultured in starvation media. Adding more Beclin-1 restores the mutant cells' autophagy responses, "demonstrating that Beclin-1 was essential for autophagy in mammalian cells," said Levine. Levine and her colleagues have since found that Beclin-1 integrates multiple cellular signals to boost autophagy in response to stress.

Loss-of-function studies in several multicellular organisms have linked Beclin-1 to longevity, innate immunity, tumor suppression, protection from neurodegenerative diseases, and mediating the beneficial effects of exercise. Beclin-1 loss commonly accompanies cancer as well as Alzheimer's and Huntington's diseases in humans. Harkening back to Levine's original studies, several viral virulence proteins also target Beclin-1. "Thousands of studies by many people throughout the world have really provided very strong evidence that autophagy is involved in many different diseases," said Levine.

Given the fundamental importance of autophagy in general and Beclin-1 in particular, researchers are understandably keen to develop drugs that can target this pathway. To do that, Levine has focused on the role of Beclin-1 in caloric restriction and exercise, two interventions long known to provide manifold benefits to patients. When transgenic mice expressing an autophagy reporter protein run on a treadmill, they display increased autophagy in cells from skeletal muscle, liver, pancreas, adipose tissue, and brain.

Mutant mice defective in their ability to induce autophagy show no improvement in glucose sensitivity, cholesterol, or other physiological measures when they exercise, implying that autophagy mediates the beneficial effects of exercise. "These findings suggest that exercise-mediated autophagy may be important in protection against diabetes and cancer progression, and we speculate it may also be important in protection against other diseases," said Levine.

Unfortunately, many of the people in greatest need of the benefits of exercise-mediated autophagy are either unwilling or unable to exercise. To address that, Levine and her colleagues have been developing a Beclin-1 targeting peptide drug. The scientists found an exposed region of the Beclin-1 protein critical for its autophagy function. They then determined that a short amino acid sequence from that region, joined to a sequence from the HIV Tat protein, is a potent inducer of autophagy.

The fusion protein penetrates cells well, and shows promise against diseases ranging from HIV to fatty liver, at least in preclinical experiments. Though peptides can be used directly as drugs, they generally have to be delivered by injection and tend to be relatively unstable. The researchers are now trying to build a fully synthetic chemical compound that will provide the same benefits, but which can be more easily delivered to patients.

Turning the gut

While stressors such as exercise and calorie restriction may generally increase autophagy in multiple cell types, the effects of autophagy vary dramatically depending on its cellular context. For example, autophagy can either enhance or suppress growth, and can promote cell survival or cell death. In cancer, the picture is even more nuanced. "Even within a single epithelial growth model, depending on the oncogene, you get different impacts of autophagy on growth," said Eric Baehrecke, adding that "understanding how different cells are activating this process and how it may be regulated ... may be the future of targeted therapies."

To study these diverse effects, Baehrecke and his colleagues have sought animal models where autophagy happens naturally during development. The fruit fly Drosophila melanogaster provides an especially good example. Between the third instar larval stage and the prepupa, the fly's intestine shrinks dramatically in just eight hours. Mutations in autophagy genes inhibit this shrinkage, indicating that the gut shrinkage relies on autophagy. A fluorescently tagged protein provides a convenient visual assay for the process, and looking at various mutants in this system reveals that autophagy actually shrinks the individual cells to reduce the intestine's size.

As the intestinal cells shrink, they're not just digesting random parts of themselves. Baehrecke found that the cells are specifically removing mitochondria. This phenomenon, called mitophagy, has also been associated with Parkinson's disease, and indeed researchers found that homolog of the human Parkinson's-associated protein Parkin is required for mitophagy.

The most common receptor for cellular cargo to use in entering the autophagy pathway is a protein called p62, but that receptor isn't required for mitophagy in the fly gut system. Some other receptor had to be responsible. To find it, the investigators searched databases for fly proteins with ubiquitin-binding domains, as ubiquitin often tags damaged structures for degradation in the cell, identifying 133 such domains. They then screened the associated genetic loci and identified three required for mitophagy in their system.

"The big shocker to us came when we analyzed the morphology of these mitochondria" in mutant flies, Baehrecke said. Besides failing to be cleared from the cell, the mitochondria stood out for being enormous. While wondering what caused these "mighty chondria" to form, the researchers began investigating the process of mitochondrial fission and fusion. Mitochondria can fuse with each other to grow larger or to dilute the effects of damaged structures, or can jettison damaged sections as smaller mitochondria. Mutating a protein that regulates fission appears to influence mitophagy.

"I feel like this makes a lot of sense, you would want this process of mitophagy to be very tightly associated with the recognition of the damaged portion of the mitochondria," said Baehrecke. Understanding this mechanism could also be important for developing new treatments for neurodegenerative diseases such as Parkinson's, where mitophagy appears to help drive pathogenesis.

Highlights

The cellular autophagy system interacts extensively with chaperone proteins.

The LAMP-2A protein drives chaperone-mediated autophagy in mammals, and declines in LAMP-2A correlate with many chronic diseases.

Beclin-1 links autophagy to protein elongation, a process cancer cells often inhibit.
Drugs targeting the elongation-autophagy interface could help kill tumors resistant to current chemotherapies.

Never too old for a chaperone

Autophagy degrades proteins, so it's no surprise that the autophagy system interacts extensively with the chaperone system, a major cellular mechanism for controlling protein production and activity. Exposing specific peptide structures on a protein can cause a particular chaperone to bind it, directing the marked protein for degradation by autophagy.

In outlining this system, Ana Maria Cuervo presented a straightforward diagram of the process. "This is much more simple than the [current model] of macroautophagy, but mostly because we have not been able to use yeast as a model," she explained, in regards to why there currently isn't as sophisticated an understanding of mammalian chaperone-mediated autophagy as of macroautophagy. The chief protein responsible for chaperone-mediated autophagy in mammals, LAMP-2A, is not conserved in yeast, so researchers studying this process must use more cumbersome cell culture systems and whole organisms, which has slowed progress.

So far, Cuervo's team has developed a series of mammalian cell culture and in vitro models for studying chaperone-mediated autophagy, as well as transgenic mice carrying fluorescent markers that allow them to watch the process in different tissues. These experiments have revealed numerous situations where chaperone-mediated autophagy fails or declines. Aging, neurodegenerative and kidney diseases, and metabolic and immune disorders all feature decreases in this type of autophagy, but it increases in models of cancer and lupus.

In mice, LAMP-2A levels decrease with age, and environmental factors can speed the decline. "You don't have to become old to get these decreased levels of LAMP-2A, you just have to go to McDonald's [quite] often," said Cuervo. Letting animals age, or putting them on a high-fat diet, destabilizes LAMP-2A in many of their tissues and also disrupts the protein's ability to multimerize at the lysosomal membrane, a crucial step in chaperone-mediated autophagy. Cuervo's colleagues have found similar defects in LAMP-2A levels and activity in models of Parkinson's, Alzheimer's, and Huntington's disease. In all of these systems, LAMP-2A proteins appear to reach the surface of the lysosome, but then get stuck, either because they fail to multimerize or fail to disperse later in the autophagic process.

The researchers have found that chaperone-mediated autophagy is critical for both protein quality control and amino acid recycling in many different cell types. "These functions really translate to the whole organism," said Cuervo. Her laboratory has developed numerous transgenic mouse strains with defects in chaperone-mediated autophagy in different organs. Eliminating the process in the brain reproduces some features of human neurodegenerative diseases such as Alzheimer's. Disrupting chaperone-mediated autophagy in the liver destroys glycogen production and causes that organ to accumulate huge deposits of fat. Looking more closely, the researchers found that the proteins responsible for glycolysis remained active in these animals as they were not degraded, and lipolysis declined because perilipin proteins were no longer being removed. That combination led to a massive accumulation of lipids.

"Unfortunately, we are all going to get old and our [chaperone-mediated autophagy] is going to go down, so we have to do something before that happens," said Cuervo. Her team is now looking for drugs that can regulate the process. "Hopefully [we'll] be able to develop treatments for some of these diseases through manipulation of the different forms of autophagy," said Cuervo.

Shortening a long story

Like Levine, Matthias Versele didn't set out to work on autophagy; he wanted to develop new cancer drugs that would target protein elongation. Tumors commonly outgrow their local nutrient supplies, forcing cancer cells to adapt to starvation conditions in order to survive. The largest user of ATP energy in a typical cell is protein elongation, so starving cells must inhibit that process. "Really the only mechanism to do that is to control the activity of elongation factor 2," said Versele.

Phosphorylation of elongation factor 2 (EF2) by EF2 kinase stops protein elongation and saves the cell's energy, and many types of tumors have elevated levels of EF2 kinase that keep their protein elongation rates low. "The hypothesis of the project was that if we were able to inhibit EF2 kinase, we might be able to promote excessive protein elongation and excessive energy expenditure, and essentially drive the cancer cell into a metabolic crisis," said Versele.

The researchers screened a chemical library for compounds that inhibited EF2 kinase, revealing a few promising hits. Some of these chemicals promote excessive protein synthesis in starved cancer cells in culture, causing the cells to die. Adding a protein synthesis inhibitor reverses the drugs' effects, showing that they are indeed working by boosting protein production. However, in addition to EF2 kinase, the most promising compounds also bound p62, a cargo protein that normally degrades through autophagy. Treated cells show a gradual increase in p62 levels. "We took this as a first indication that these compounds not only were able to regulate protein elongation, but also inhibit autophagy," said Versele.

Next, the team looked for genes whose loss would strengthen the effect of the EF2 kinase inhibitors. That experiment pointed to a single gene: Beclin-1. As Levine had explained in the meeting's second talk, Beclin-1 is a critical component of the mammalian autophagy machinery, and Beclin-1 defects occur in many human cancers. Indeed, Versele found that cancer cells with higher levels of autophagy are more sensitive to compounds that inhibit EF2 kinase, suggesting that this treatment strategy might work especially well in tumors with Beclin-1 defects.

The link between EF2 kinase and autophagy isn't absolute, though. The extent to which an individual compound inhibited EF2 kinase doesn't correlate with the extent to which it inhibits cell growth; autophagy and protein elongation can be uncoupled. Using genome editing techniques, the researchers deleted the EF2 kinase gene in a cultured cell line, and found that the cells still accumulate p62 protein when treated with a powerful EF2 kinase inhibitor. "[This suggests] that the accumulation of p62 by this particular compound is not related to EF2 kinase, because that target is not present in these cells," said Versele.

Turning to biochemistry, the team coupled their EF2 kinase inhibitors to solid matrices and screened cell lysates to find proteins that bound the drugs. This revealed that different compounds bind with different affinities to EF2 kinase and Beclin-1. Some of the compounds are strong inhibitors of one protein or the other, while some inhibit both. The researchers are now pursuing two distinct strategies for different types of cancer cells: inhibiting EF2 kinase in established tumors and blocking autophagy in early tumor development.

A lot to digest

After the main presentations, the speakers returned to the stage for a wide-ranging panel discussion and questions from the audience. Versele began the session by asking Ohsumi what happens to proteins and organelles once they enter the lysosome or vacuole. Ohsumi replied that this phase of autophagy remains poorly understood. Next, the group discussed how the autophagy system interacts with the antigen presentation machinery. That area, too, holds many mysteries. "It is involved in different aspects of antigen presentation, but we don't have [a] granular understanding of the molecular mechanisms and biochemistry," said Levine.

Panelists also tackled the problem of measuring autophagy in humans, which will clearly be necessary as drugs targeting this process move toward the clinic. Cuervo suggested that at least initially, researchers will need to rely on surrogate measurements such as autophagy rates in isolated blood samples and skin fibroblasts.

In response to a question from an audience member, Baehrecke discussed emerging work suggesting that autophagy is important in wound healing. However, he emphasized that the data so far come primarily from artificial models, and mechanistic details are sparse. Another attendee asked how different autophagy pathways coordinate with each other within cells. "They are not in isolation, they really coordinate with each other," said Cuervo, adding that "when you block one of these pathways the others compensate."

In response to a different question, the panelists agreed that it will be important to study the specificity of autophagy inhibitors as they move into the clinic, as it may be necessary to increase or decrease it for different indications. However, a general promoter of autophagy could still prove useful. "I feel like there is a physiological range in which up-regulation of autophagy may be beneficial in protecting against certain diseases," said Levine.