Research Area: Genetics

Countering mitochondrial stress

Scientists discover a pathway that monitors a protein import into mitochondria and elicits a cellular response when the process goes awry.

Raleigh McElvery | Department of Biology
April 13, 2018

If there’s one fact that most people retain from elementary biology, it’s that mitochondria are the powerhouses of the cell. As such, they break down molecules and manufacture new ones to generate the fuel necessary for life. But mitochondria rely on a stream of proteins to sustain this energy production. Nearly all their proteins are manufactured in the surrounding gel-like cytoplasm, and must be imported into the mitochondria to keep the powerhouse running.

A duo of MIT biologists has revealed what happens when a traffic jam of proteins at the surface of the mitochondria prevents proper import. They describe how the mitochondria communicate with the rest of the cell to signal a problem, and how the cell responds to protect the mitochondria. This newly-discovered molecular pathway, called mitoCPR, detects import mishaps and preserves mitochondrial function in the midst of such stress.

“This is the first mechanism identified that surveils mitochondrial protein import, and helps mitochondria when they can’t get the proteins they need,” says Angelika Amon, the Kathleen and Curtis Marble Professor of Cancer Research in the MIT Department of Biology, who is also a member of the Koch Institute for Integrative Cancer Research at MIT, a Howard Hughes Medical Institute Investigator, and senior author of the study. “Responses to mitochondrial stress have been established before, but this one specifically targets the surface of the mitochondria, clearing out the misfolded proteins that are stuck in the pores.”

Hilla Weidberg, a postdoc in Amon’s lab, is the lead author of the study, which appears in Science on April 13.

Fueling the powerhouse

Mitochondria likely began as independent entities long ago, before being engulfed by host cells. They eventually gave up control and moved most of their important genes to a different organelle, the nucleus, where the rest of the cell’s genetic blueprint is stored. The protein products from these genes are ultimately made in the cytoplasm outside the nucleus, and then guided to the mitochondria. These “precursor” proteins contain a special molecular zip code that guides them through the channels at the surface of the mitochondria to their respective homes.

The proteins must be unfolded and delicately threaded through the narrow channels in order to enter the mitochondria. This creates a precarious situation; if the demand is too high, or the proteins are folded when they shouldn’t be, a bottleneck forms that none shall pass. This can simply occur when the mitochondria expand to make more of themselves, or in diseases like deafness-dystonia syndrome and Huntington’s.

“The machinery that we’ve identified seems to evict proteins that are sitting on the surface of the mitochondria and sends them for degradation,” Amon says. “Another possibility is that this mitoCPR pathway might actually unfold these proteins, and in doing so give them a second chance to be pushed through the membrane.”

Two other pathways were recently identified in yeast that also respond to accumulated mitochondrial proteins. However, both simply clear protein refuse from the cytoplasm around the mitochondria, rather than removing the proteins collecting on the mitochondria themselves.

“We knew about various responses to mitochondrial stress, but no one had described a response to protein import defects that specifically protected the mitochondria, and that’s exactly what mitoCPR does,” Weidberg says. “We wanted to know how the cell reacts to these problems, so we set out to overload the import machinery, causing many proteins to rush into the mitochondrion at the same time and clog the pores, triggering a cellular response.”

“What makes our cells absolutely dependent on mitochondria is one of those million-dollar questions in cell biology,” says Vlad Denic, professor of molecular and cellular biology at Harvard University. “This study reveals an interesting flip-side to that question: When you make mitochondrial life artificially tough, are they programmed to say ‘help us’ so the host cell comes to their rescue? The possible ramifications of such work in terms of human development and disease could be very impressive.”

A pathway to understanding

Roughly two decades ago, researchers began to notice that the genes required to defend cells against drugs and other foreign substances — together, called the multidrug resistance (MDR) response — were also expressed in yeast mitochondrial mutants for some unknown reason. This suggested that the protein in charge of binding to the DNA and initiating the MDR response must have a dual purpose, sometimes triggering a second, separate pathway as well. But precisely how this second pathway related to mitochondria remained a mystery.

“Twenty years ago, scientists recognized mitoCPR as some kind of mechanism against mitochondrial dysfunction,” Weidberg says. “Today we’ve finally characterized it, given it a name, and identified its precise function: to help mitochondrial protein import.”

As the import process slows, Amon and Weidberg determined that the protein that initiates mitoCPR — the transcription factor Pdr3 — binds to DNA within the nucleus, inducing the expression of a gene known as CIS1. The resultant Cis1 protein binds to the channel at the surface of the mitochondrion, and recruits yet another protein, the AAA+ adenosine triphosphatase Msp1, to help clear unimported proteins from the mitochondrial surface and mediate their degradation. Although the MDR response pathway differs from that of mitoCPR, both rely on Pdr3 activation. In fact, mitoCPR requires it.

“Whether the two pathways interact with one another is a very interesting question,” Amon says. “The mitochondria make a lot of biosynthetic molecules, and blocking that function by messing with protein import could lead to the accumulation of intermediate metabolites. These can be toxic to the cell, so you could imagine that activating the MDR response might pump out harmful intermediates.”

The question of what activates Pdr3 to initiate mitoCPR is still unclear, but Weidberg has some ideas related to signals stemming from the build-up of toxic metabolite intermediates. It’s also yet to be determined whether an analogous pathway exists in more complex organisms, although there is some evidence that the mitochondria do communicate with the nucleus in other eukaryotes besides yeast.

“This was just such a classic study,” Amon says. “There were no sophisticated high-throughput methodologies, just traditional, simple molecular biology and cell biology assays with a few microscopes. It’s almost like something you’d see out of the 1980s. But that just goes to show — to this day — that’s how many discoveries are made.”

The research was funded by the National Institutes of Health and by the Koch Institute Support (core) Grant from the National Cancer Institute. Amon is also an investigator of the Howard Hughes Medical Institute and the Glenn Foundation for Biomedical Research. Weidberg was supported by the Jane Coffin Childs Memorial Fund, the European Molecular Biology Organization Long-Term Fellowship, and the Israel National Postdoctoral Program for Advancing Women in Science.

Study suggests method for boosting growth of blood vessels and muscle

Activating proteins linked to longevity may help to increase endurance and combat frailty in the elderly.

Anne Trafton | MIT News Office
March 22, 2018

As we get older, our endurance declines, in part because our blood vessels lose some of their capacity to deliver oxygen and nutrients to muscle tissue. An MIT-led research team has now found that it can reverse this age-related endurance loss in mice by treating them with a compound that promotes new blood vessel growth.

The study found that the compound, which re-activates longevity-linked proteins called sirtuins, promotes the growth of blood vessels and muscle, boosting the endurance of elderly mice by up to 80 percent.

If the findings translate to humans, this restoration of muscle mass could help to combat some of the effects of age-related frailty, which often lead to osteoporosis and other debilitating conditions.

“We’ll have to see if this plays out in people, but you may actually be able to rescue muscle mass in an aging population by this kind of intervention,” says Leonard Guarente, the Novartis Professor of Biology at MIT and one of the senior authors of the study. “There’s a lot of crosstalk between muscle and bone, so losing muscle mass ultimately can lead to loss of bone, osteoporosis, and frailty, which is a major problem in aging.”

The first author of the paper, which appears in Cell on March 22, is Abhirup Das, a former postdoc in Guarente’s lab who is now at the University of New South Wales in Australia. Other senior authors of the paper are David Sinclair, a professor at Harvard Medical School and the University of New South Wales, and Zolt Arany, a professor at the University of Pennsylvania.

Race against time

In the early 1990s, Guarente discovered that sirtuins, a class of proteins found in nearly all animals, protect against the effects of aging in yeast. Since then, similar effects have been seen in many other organisms.

In their latest study, Guarente and his colleagues decided to explore the role of sirtuins in endothelial cells, which line the inside of blood vessels. To do that, they deleted the gene for SIRT1, which encodes the major mammalian sirtuin, in endothelial cells of mice. They found that at 6 months of age, these mice had reduced capillary density and could run only half as far as normal 6-month-old mice.

The researchers then decided to see what would happen if they boosted sirtuin levels in normal mice as they aged. They treated the mice with a compound called NMN, which is a precursor to NAD, a coenzyme that activates SIRT1. NAD levels normally drop as animals age, which is believed to be caused by a combination of reduced NAD production and faster NAD degradation.

After 18-month-old mice were treated with NMN for two months, their capillary density was restored to levels typically seen in young mice, and they experienced a 56 to 80 percent improvement in endurance. Beneficial effects were also seen in mice up to 32 months of age (comparable to humans in their 80s).

“In normal aging, the number of blood vessels goes down, so you lose the capacity to deliver nutrients and oxygen to tissues like muscle, and that contributes to decline,” Guarente says. “The effect of the precursors that boost NAD is to counteract the decline that occurs with normal aging, to reactivate SIRT1, and to restore function in endothelial cells to give rise to more blood vessels.”

These effects were enhanced when the researchers treated the mice with both NMN and hydrogen sulfide, another sirtuin activator.

Vittorio Sartorelli, a principal investigator at the National Institute of Allergy and Infectious Diseases who was not involved in the research, described the experiments as “elegant and compelling.” He added that “it will be of interest and of clinical relevance to evaluate the effect of NMN and hydrogen sulfide on the vascularization of other organs such as the heart and brain, which are often damaged by acutely or chronically reduced blood flow.”

Benefits of exercise

The researchers also found that SIRT1 activity in endothelial cells is critical for the beneficial effects of exercise in young mice. In mice, exercise generally stimulates growth of new blood vessels and boosts muscle mass. However, when the researchers knocked out SIRT1 in endothelial cells of 10-month-old mice, then put them on a four-week treadmill running program, they found that the exercise did not produce the same gains seen in normal 10-month-old mice on the same training plan.

If validated in humans, the findings would suggest that boosting sirtuin levels may help older people retain their muscle mass with exercise, Guarente says. Studies in humans have shown that age-related muscle loss can be partially staved off with exercise, especially weight training.

“What this paper would suggest is that you may actually be able to rescue muscle mass in an aging population by this kind of intervention with an NAD precursor,” Guarente says.

In 2014, Guarente started a company called Elysium Health, which sells a dietary supplement containing a different precursor of NAD, known as NR, as well as a compound called pterostilbene, which is an activator of SIRT1.

The research was funded by the Glenn Foundation for Medical Research, the Sinclair Gift Fund, a gift from Edward Schulak, and the National Institutes of Health.

A blueprint for regeneration

Whitehead Institute researchers uncover framework for how stem cells determine where to form replacement structures.

Lisa Girard | Whitehead Institute
March 15, 2018

Researchers at Whitehead Institute have uncovered a framework for regeneration that may explain and predict how stem cells in adult, regenerating tissue determine where to form replacement structures.

In a paper that appeared online March 15 in the journal Science, the researchers describe a model for planarian (flatworm) eye regeneration that is governed by three principles acting in concert, which inform how progenitor cells behave in regeneration. The model invokes positional cues that create a scalable map; self-organization that attracts progenitors to existing structures; and progenitor cells that originate in a diffuse spatial zone, rather than a precise location, allowing flexibility in their path. These principles appear to dictate how progenitor cells decide where to go during regeneration to recreate form and function, and they bring us closer to a systems-level understanding of the process.

From previous work, the researchers knew that stem cells are likely reading out instructions from neighboring tissues to guide their path, and it became clear that the process faces some serious challenges in regeneration. “We realized that positional information has to move; it needs to change during regeneration in order to specify the new missing parts to be regenerated. This revised information can then guide progenitor cells that are choosing to make new structures to differentiate into the correct anatomy at the correct locations,” says the paper’s senior author Peter Reddien, a Whitehead Institute researcher, an MIT professor of biology, and a Howard Hughes Medical Institute (HHMI) investigator. “There is a puzzle that emerges, however. Since positional information shifts after injury during regeneration, there is a mismatch between the positional information pattern and the remaining anatomy pattern. Realizing this mismatch exists was a trigger for our study. We wanted to understand how stem cells making particular tissues decide where to go and differentiate. Is it based on anatomy, or is it based on positional information? And when those two things are not aligned, how do they decide?”

Reddien and his lab have spent over a decade unraveling the mysteries of regeneration using a small flatworm, called the planarian. If a planarian’s head is amputated, or its side is removed, each piece will regenerate an entire animal. In order to understand how progenitors decide where to go in the noisy environment of animal regeneration, the researchers used the planarian eye, a visible organ that is small enough to be removed without serious injury and has the added advantage of having defined progenitor cell molecular markers.

The researchers devised a simple experiment to resolve the question of how the progenitors decide where to go: amputate the animal’s head, then after three days remove one of the eyes from the head piece. What they found was that progenitor cells would nucleate a new eye in a position anterior (closer to the tip) to the remaining eye, rather than in the “correct” position specified by the anatomy, symmetrical with the current eye. However, if the same experiment is done, except with one of the eyes removed from the head piece earlier — the same day as the amputation, rather than three days later — there is a different result: The new eye is nucleated at a position symmetrical to the remaining eye, at the “correct” position according to the anatomy, suggesting that when the choices are conflicting, anatomical self-organizing dynamics win.

These simple rules guide the system to successful regeneration and also yield surprising outcomes when the system is pushed to its limits, producing alternative stable anatomical states with three-, four-, and five-eyed animals. Resecting both the side of an animal and amputating the head can place progenitors far enough from existing, self-organizing attractors that they miss them, allowing them to nucleate a new attractor — a third eye — in the head fragments. “If the migrating progenitors are too close to the attractor that is the existing eye, it will suck them in, just like a black hole; if they are far enough, they can escape,” says Kutay Deniz Atabay, first author on the paper and a graduate student in the Reddien lab.

When the researchers did the same surgery on a three-eyed animal, the progenitors miss the attraction of existing eyes and form a five-eyed animal. In each of these cases, all of the eyes are functionally integrated into the brain and exhibit the same light-avoidance response of the planarian eye. The researchers also observed that these alternative anatomical states endure, revealing rules for maintenance of existing and alternative anatomical structures. “This recipe helps provide an explanation for a fundamental problem of animal regeneration, which is how migratory progenitors make choices that lead the system to regenerate and maintain organs,” Atabay says.

“These studies contribute to a proposed framework for regeneration in which competing forces, self-organization, and extrinsic cues, are the guideposts impacting the choice of progenitor targeting in regeneration; and those two forces together determine the outcome,” Reddien says.

This work was supported by the National Institutes of Health (NIH) and the HHMI. Additional funding was provided by the MIT Presidential Fellowship Program and a National Defense Science and Engineering Graduate Fellowship.

Study: Fragile X syndrome neurons can be restored

Whitehead Institute researchers are using a modified CRISPR/Cas9-guided activation strategy to investigate the most frequent cause of intellectual disability in males.

Nicole Giese Rura | Whitehead Institute
February 15, 2018

Fragile X syndrome is the most frequent cause of intellectual disability in males, affecting one out of every 3,600 boys born. The syndrome can also cause autistic traits, such as social and communication deficits, as well as attention problems and hyperactivity. Currently, there is no cure for this disorder.

Fragile X syndrome is caused by mutations in the FMR1 gene on the X chromosome, which prevent the gene’s expression. This absence of the FMR1-encoded protein during brain development has been shown to cause the overexcitability in neurons associated with the syndrome. Now, for the first time, researchers at Whitehead Institute have restored activity to the fragile X syndrome gene in affected neurons using a modified CRISPR/Cas9 system they developed that removes the methylation — the molecular tags that keep the mutant gene shut off — suggesting that this method may prove to be a useful paradigm for targeting diseases caused by abnormal methylation.

Research by the lab of Whitehead Institute for Biomedical Research Founding Member Rudolf Jaenisch, which is described online this week in the journal Cell, is the first direct evidence that removing the methylation from a specific segment within the FMR1 locus can reactivate the gene and rescue fragile X syndrome neurons.

The FMR1 gene sequence includes a series of three nucleotide (CGG) repeats, and the length of these repeats determines whether or not a person will develop fragile X syndrome: A normal version of the gene contains anywhere from 5 to 55 CGG repeats, versions with 56 to 200 repeats are considered to be at a higher risk of generating some of the syndrome’s symptoms, and those versions with more than 200 repeats will produce fragile X syndrome.

Until now, the mechanism linking the excessive repeats in FMR1 to fragile X syndrome was not well-understood. But Shawn Liu, a postdoc in Jaenisch’s lab and first author of the Cell study, and others thought that the methylation blanketing those nucleotide repeats might play an important role in shutting down the gene’s expression.

In order to test this hypothesis, Liu removed the methylation tags from the FMR1 repeats using a CRISPR/Cas9-based technique he recently developed with Hao Wu, a postdoc in the Jaenisch lab. This technique can either add or delete methylation tags from specific stretches of DNA. Removal of the tags revived the FMR1 gene’s expression to the level of the normal gene.

“These results are quite surprising — this work produced almost a full restoration of wild type expression levels of the FMR1 gene,” says Jaenisch, whose primary affiliation is with Whitehead Institute, where his laboratory is located and his research is conducted. He is also a professor of biology at MIT. “Often when scientists test therapeutic interventions, they only achieve partial restoration, so these results are substantial,” he says.

The reactivated FMR1 gene rescues neurons derived from fragile X syndrome induced pluripotent stem (iPS) cells, reversing the abnormal electrical activity associated with the syndrome. When rescued neurons were engrafted into the brains of mice, the FMR1 gene remained active in the neurons for at least three months, suggesting that the corrected methylation may be sustainable in the animal.

“We showed that this disorder is reversible at the neuron level,” says Liu. “When we removed methylation of CGG repeats in the neurons derived from fragile X syndrome iPS cells, we achieved full activation of FMR1.”

The CRISPR/Cas-9-based technique may also prove useful for other diseases caused by abnormal methylation including facioscapulohumeral muscular dystrophy and imprinting diseases.

“This work validates the approach of targeting the methylation on genes, and it will be a paradigm for scientists to follow this approach for other diseases,” says Jaenisch.

This work was supported by the National Institutes of Health, the Damon Runyon Cancer Foundation, the Rett Syndrome Research Trust, the Brain and Behavior Research Foundation, and the Helen Hay Whitney Foundation. Jaenisch is co-founder of Fate Therapeutics, Fulcrum Therapeutics, and Omega Therapeutics.

Reading and writing DNA

Department of Biology kicks off IAP seminar series with a lecture by synthetic-biology visionary George Church.

Raleigh McElvery | Department of Biology
January 31, 2018

Thanks to the invention of genome sequencing technology more than three decades ago, we can now read the genetic blueprint of virtually any organism. After the ability to read came the ability to edit — adding, subtracting, and eventually altering DNA wherever we saw fit. And yet, for George Church, a professor at Harvard Medical School, associate member of the Broad Institute, and founding core faculty and lead for synthetic biology at the Wyss Institute — who co-pioneered direct genome sequencing in 1984 — the ultimate goal is not just to read and edit, but also to write.

What if you could engineer a cell resistant to all viruses, even the ones it hadn’t yet encountered? What if you could grow your own liver in a pig to replace the faulty one you were born with? What if you could grow an entire brain in a dish? In his lecture on Jan. 24 — which opened the Department of Biology’s Independent Activities Period (IAP) seminar series, Biology at Transformative Frontiers — Church promised all this and more.

“We began by dividing the Biology IAP events into two tracks: one related to careers in academia and another equivalent track for industry,” says Jing-Ke Weng, assistant professor and IAP faculty coordinator for the department. “But then it became clear that George Church, Patrick Brown, and other speakers we hoped to invite blurred the boundaries between those two tracks. The Biology at Transformative Frontiers seminar series became about the interface of these trajectories, and how transferring technologies from lab bench to market is altering society as we know it.”

The seminar series is a staple in the Department of Biology’s IAP program, but during the past several years it has been oriented more toward quantitative biology. Weng recalls these talks as being relegated to the academic sphere, and wanted to show students that the lines between academia, industry, and scientific communication are actually quite porous.

“We chose George Church to kick off the series because he’s been in synthetic biology for a long time, and continues to have a successful academic career even while starting so many companies,” says Weng.

Church’s genomic sequencing methods inspired the Human Genome Project in 1984 and resulted in the first commercial genome sequence (the bacterium Helicobacter pylori) 10 years later. He also serves as the director of the Personal Genome Project, the “Wikipedia” of open-access human genomic data. Beyond these ventures, he’s known for his work on barcoding, DNA assembly from chips, genome editing, and stem cell engineering.

He’s also the same George Church who converted the book he co-authored with Ed Regis, “Regenesis: How Synthetic Biology Will Reinvent Nature and Ourselves,” into a four-letter code based on the four DNA nucleotides (A, T, C, and G), subsisted on nutrient broth from a lab vendor for an entire year, and dreams of eventually resurrecting woolly mammoths. He’s being featured in an upcoming Netflix Original documentary, so when he arrived at the Stata Center to give his lecture last week he was trailed by a camera crew.

According to Church, the transformative technologies that initially allowed us to read and edit DNA have grown exponentially in recent years with the invention of molecular multiplexing and CRISPR-Cas9 (think Moore’s Law but even more exaggerated). But there’s always room for improvement.

“There’s been a little obsession with CRISPR-Cas9s and other CRISPRs,” said Church. “Everybody is saying how great it is, but it’s important to say what’s wrong with it as well, because that tells us where we’re going next and how to improve on it.”

He outlined several of his own collaborations, including those aimed at devising more precise methods of genome editing, one resulting in 321 changes to the Escherichia coli genome — the largest change in any genome yet — rendering the bacterium resistant to all viruses, even those it had not yet come into contact with. The next step? Making similarly widespread changes in plants, animals, and eventually perhaps even human tissue. In fact, Church and his team have set their sights on combatting the global transplantation crisis with humanlike organs grown in animals.

“Since the dawn of transplantation as a medical practice, we’ve had to use either identical twins or rare matches that are very compatible immunologically, because we couldn’t engineer the donor or the recipient,” said Church.

Since it’s clearly unethical to engineer human donors, Church reasoned, why not engineer animals with compatible organs instead? Pigs, to be exact, since most of their organs are comparable in size and function to our own.

“This is an old dream; I didn’t originate it,” said Church. “It started about 20 years ago, and the pioneers of this field worked on it for a while, but dropped it largely because the number of changes to the genome were daunting, and there was a concern that the viruses all pigs make — retroviruses — would be released and infect the immunocompromised organ recipient.”

Church and his team successfully disrupted 62 of these retroviruses in pig cells back in 2015, and in 2017 they used these cells to generate living, healthy pigs. Today, the pigs are thriving and rearing piglets of their own. Church is also considering the prospect of growing augmented organs in pigs for human transplantation, perhaps designing pathogen-, cancer-, and age-resistant organs suitable for cryopreservation.

“Hopefully we’ll be doing nonhuman primate trials within a couple of years, and then almost immediately after that human trials,” he said.

Another possibility, rather than cultivating organs in animals for transplant, is to generate them in a dish. A subset of Church’s team is working on growing from scratch what is arguably the most complicated organ of all, the brain.

This requires differentiating multiple types of cells in the same dish so they can interact with each other to form the complex systems of communication characteristic of the human brain.

Early attempts at fashioning brain organoids often lacked capillaries to distribute oxygen and nutrients (roughly one capillary for each of the 86 billion neurons in the human brain). However, thanks to their new human transcription factor library, Church and colleagues have begun to generate the cell types necessary to create such capillaries, plus the scaffolding needed to promote the three-dimensional organization of these and additional brain structures. Church and his team have not only successfully integrated the structures with one another, but have also created an algorithm that spits out the list of molecular ingredients required to generate each cell type.

Church noted these de novo organoids are extremely useful in determining which genetic variants are responsible for certain diseases. For instance, you could sequence a patient’s genome and then create an entire organoid with the mutation in question to test whether it was the root cause of the condition.

“I’m still stunned by the breadth of projects and approaches that he’s running simultaneously,” says Emma Kowal, a second-year graduate student, member of Weng’s planning committee, and a former researcher in Church’s lab. “The seminar series is called Biology at Transformative Frontiers, and George is very much a visionary, so we thought it would be a great way to start things off.”

The four-part series also features Melissa Moore, chief scientific officer of the Moderna Therapeutics mRNA Research Platform, Jay Bradner, president of the Novartis Institutes for BioMedical Research, and Patrick Brown, CEO and founder of Impossible Foods.

Decoding the genetics of roundworm mating

Sixth year graduate student Zoë Hilbert investigates how C. elegans react to changes in their environment — and how these changes affect physiology, gene expression, and behavior

Raleigh McElvery
January 30, 2018

Sixth year graduate student Zoë Hilbert is sure of many things. After performing her first dissection in third grade, she was sure she liked science. Before she started college, she was sure she wanted to major in a biology-related discipline. And as she finished her final year at Columbia University, she was sure she would leave the East Coast immediately upon graduation. What she did not anticipate, however, was falling in love with the Cambridge biotechnology hub, applying to MIT for graduate school, and switching fields from biochemistry to genetics.

“I’m incredibly grateful for the MIT first-year program, because dedicating the fall semester solely to taking classes gave me a background in subjects I didn’t take in college,” Hilbert says. “I’d never taken genetics before, and now here I am in Dennis Kim’s lab — a genetics lab.”

Hilbert was enthralled by evolution from an early age, in particular the idea that entire organisms and their proteins change over time in response to internal and external pressures. She recalls becoming “obsessed” with the small and seemingly unremarkable stickleback fish, after she learned that researchers could map the evolution of physical features like additional belly fins or extra armor to variations in specific genes.

“When it came time for the first years to write our National Science Foundation proposals, we had the opportunity to work with a faculty member,” she recalls, “and I chose Dennis because one of the project ideas he’d listed was in a similar vein to the stickleback research. Coming into it, I didn’t know anything about his work or even his model system, but I ended up joining the lab after second semester rotations.”

The Kim lab investigates how the roundworm Caenorhabditis elegans reacts to changes in their environment — and how these changes not only affect physiology and gene expression, but behavior as well.

Today, Hilbert is as enamored by C. elegans as she once was with stickleback fish. With minimal prodding, she’s happy to rattle off their numerous advantages: they’re transparent, so there’s no need to do dissections to look inside; they’re ideal for studying development and the nervous system, because scientists have already charted all the cells in the body and how the neurons communicate; and they’re low-maintenance and easy to keep in lab. The list goes on.

But most pertinent to Hilbert is the fact that — like most species of animals — the two sexes of C. elegans, males and hermaphrodites, often behave differently in similar situations due to differences in gene expression. Take mating, for example.

Hermaphrodites are capable of self-fertilization, and can produce up to several hundred identical progeny over the course of several days. Males are much less common and unable to reproduce on their own, but by mating with hermaphrodites they can introduce some genetic variety into their offspring. Because males must locate a hermaphrodite in order to pass on their genetic material, they’ve developed some specific behaviors to find their mate. And that’s where Hilbert’s work comes in. She makes males choose between the two things they need most: food and mate.

There comes a time in every adult male’s life when finding a mate takes precedence over continuously eating, as younger worms are wont to do. If he is placed in a plate of yummy bacteria by himself, he runs away — not because he’s full, but because he’d rather spend his time searching for a mate. However, if he is placed in a plate of food along with a tempting hermaphrodite, his urge to escape is suppressed and he remains long enough to mate.

That said, C. elegans mating is not always so cut and dry. Researchers understand that a male’s behavior is also food-dependent. If you place a starving male on the plate of food, he no longer prioritizes mating over feeding, and will remain in the food instead of seeking a mate. He is constantly evaluating his priorities, which are heavily influenced by the situation at hand and — as Hilbert discovered — when and where certain genes are expressed.

“We’ve spent a lot of time monitoring how the expression of daf-7 changes in different food and mating situations,” Hilbert says. “When you starve the male, you suppress the gene and as a result you also suppress the fleeing behavior.”Hilbert demonstrated several years ago that this male-specific behavior is controlled by a gene known as daf-7, which encodes a signaling molecule and is expressed in two specific neurons in the male.  (No expression is normally seen in the hermaphrodite.) Curiously, the same gene in the same two neurons is also turned on when any worm — male or hermaphrodite — comes across a pathogen, sending a “WARNING: consume at your own risk” signal, and prompting the worm to avoid the noxious bacteria.

Expression appears to be dependent not only on nutritional state (hungry or full), but also environment (food and/or mate) and sex (since males express daf-7 differently than hermaphrodites).

“All these factors and signals are converging on this one gene,” Hilbert says. “It’s really quite incredible.”

The neurons that express daf-7 are “sensory,” and traditionally viewed as funnels to higher neural centers where information is processed and behaviors are generated. However, Hilbert’s data suggest this information processing is happening right there, directly within these neurons via changes in gene expression without waiting for instructions from on high.

What Hilbert finds particularly intriguing is that the worms rely on just one molecular pathway to dictate behavior in two very different situations: mating and pathogen avoidance. Although the worm flees food in both situations, precisely why one gene is implicated in two distinct settings remains a mystery. Hilbert is still asking herself, For what benefit?

She intends to spend her final semester at MIT tying up loose ends and conducting follow-up experiments to extend the work from her recent paper in the January 2017 issue of eLife, on which she was first author. She’s screening for molecules that could impact whether or not daf-7 is expressed, honing in on chemicals and signaling molecules used by neurons to communicate with one another.

“I’d advise prospective grads to be willing and open to change your mind about what you want to do,” she says. “I was really into protein biochemistry when I first arrived at MIT, and was really surprised when I fell in love with a discipline that was completely different from my initial interests.”

As Hilbert applies to academic postdoctoral positions, she’s still set on fulfilling her longtime dream of heading out West. She’s sure she’d like to end up someplace like California, Washington, or Utah, but only time will tell.

Photo credit: Raleigh McElvery
Combatting chemotherapy resistance

Graduate student Faye-Marie Vassel investigates a protein that helps cells tolerate DNA damage, sharing her expertise with budding scientists to further STEM education

Raleigh McElvery
December 8, 2017

Combatting chemotherapy resistance

Person with long, dark hair and lab coat stares into microscope.

Graduate student Faye-Marie Vassel investigates a protein that helps cells tolerate DNA damage, sharing her expertise with budding scientists to further STEM education

Raleigh McElvery

 

Faye-Marie Vassel has a protein. Well, as a living entity, technically she has many, but just one she affectionately refers to as her own. “My protein, REV7.” And it makes sense — if you were hard at work characterizing a single protein for all six years of your graduate career, you’d be pretty attached, too. Plus, the stakes are high. REV7, which aids in DNA damage repair, could ultimately provide insight into ways to combat chemotherapy resistance.

Although Vassel’s mother trained as an OB/GYN in Russia before moving to the U.S., serving as what Vassel describes as a “quiet” scientific role model, Vassel spent her early childhood emulating her father, a social worker, and engrossed in the social sciences. She intended to one day work in science policy — until high school when she joined an after-school program at the American Museum of Natural History in New York City, and discovered an additional interest.

Here, Vassel took a series of molecular biology classes and met her first female research mentor, a postdoctoral fellow at Rockefeller University, who encouraged her to participate in another, more advanced science program funded by the National Science Foundation.

“I initially had my doubts, but just having that support changed everything,” Vassel says. “That was my first time doing research of any kind, and I got a sense of the sheer diversity of potential research projects. That’s also when I heard there was something called biophysics.”

From that point on, Vassel was hooked. As an undergraduate at Stony Brook University, she initially declared a major in physics before switching to biochemistry. Later, when it came time to select a graduate school, she was split between MIT and the University of California, Berkeley. As she recalls, MIT’s graduate preview weekend made all the difference.

“I had the chance to stay with biology students and speak with professors,” she says. “The whole experience made the department seem personal, and demystified the graduate school process by making it more tangible.”

She proposed a joint position between two labs: Graham Walker’s lab, based in Building 68, and Michael Hemann’s lab situated in the Koch Institute for Integrative Cancer Research. Walker’s lab focuses on microbiology, DNA repair, and antibiotic resistance, while Hemann’s lab investigates chemotherapy resistance in hopes of improving cancer therapies. After stumbling upon one of their joint papers, Vassel decided she’d like to combine the two.

“It’s invaluable to have both perspectives,” she says. “Mike’s lab just celebrated its 10th anniversary, while Graham‘s just had its 35th. It’s been interesting seeing the different ways they approach their respective research questions, because they were trained in such different scientific eras.”

Although Vassel is currently the only student formally working in both labs, the collaboration between Walker and Hemann, aimed at combatting chemotherapy resistance, has been ongoing.

Frontline chemotherapies, including one anticancer agent called cisplatin, kill cancer cells by damaging their DNA and preventing them from synthesizing new genetic material. Just how sensitive cancer cells are to cisplatin — and therefore how effective the treatment is — depends on whether the cell can repair the damage and bypass DNA-damage induced cell death. In some cases, cells increase production of “translesion polymerases,” which are specialized DNA polymerases that can help cells tolerate certain kinds of DNA damage by synthesizing across from damaged DNA or DNA bound to a carcinogen.

Vassel’s protein, REV7, is a structural subunit of one key translesion polymerase, and its expression is deregulated in many different cancer cells. As Vassel suggests, if one aspect of these translesion polymerases — say, the REV7 subunit — could be altered to hinder repair, then perhaps cancer-ridden cells could regain drug sensitivity.

Thanks to recently-developed CRISPR-Cas9 gene editing techniques, Vassel has removed REV7 entirely from drug resistant lung cancer cells, and watched as cisplatin sensitivity was restored. She also conducted rescue experiments, adding REV7 back into cell lines lacking the protein to see whether those cells become resistant to the drug once again. Most recently, she has been working in murine models to see whether REV7 has similar effects in a living system.

If her hypothesis is correct, REV7 would be a powerful target for drug development. Treatments that inhibit REV7, she explains, could be used in tandem with frontline chemotherapies like cisplatin to prevent resistance.

Since her foray into biology at the American Museum of Natural History almost a decade ago, Vassel has maintained her passion for science outreach. During her time at MIT, she has served as a math tutor for middle schoolers in the Cambridge public school system. She also volunteered as a science and math mentor for high school students, as part of a dual athletic and academic program founded by MIT.

As Vassel wraps up her final year of graduate studies, she is torn between completing an academic postdoc and indulging her early interest in science education policy.

“Growing up in New York City, it was not lost on me that — despite the city’s wonderful diversity — people from historically underserved groups were still missing from many science-related positions,” Vassel says. “It got me thinking about the dire need for policymakers to improve curricula to make science more inclusive of all life experiences. There’s this idea that science is apolitical when it’s really not, and that mindset can have detrimental effects on equity and diversity in science.”

Photo credit: Raleigh McElvery
An eye for a mouth: How regenerating flatworms keep track of body parts

Graduate student Lauren Cote identifies genes directing regeneration

Justin Chen
November 16, 2017

An eye for a mouth: How regenerating flatworms keep track of body parts

Person with brown hair in pony tail sits in front of computer and microscope.

Graduate student Lauren Cote identifies genes directing regeneration

Justin Chen

 

Peering down through a microscope at a petri dish, Lauren Cote, a sixth-year graduate student, watches the tip of a worm’s tail. Alone in the petri dish, the brown globule of tissue is regenerating an entirely new digestive system, a brain, and a pair of eye spots. After just a few weeks, the animal — a quarter-inch-long ribbon of flesh capped by a triangular head — is complete again. Swimming through the dish, the worm’s grainy, mahogany body fades to a translucent gray-blue along the edges, stretching and contracting as if hinting at its malleability.

Many animals regenerate. Salamanders replace their tails while zebrafish regrow damaged heart muscle. Even humans can renew large parts of their livers. However, few creatures can regenerate like planarians, a class of flatworms found in fresh and salt water habitats around the world — and in the Reddien lab at the Whitehead Institute.

Because planarians are masters of regeneration, able to replace any body part and even create a new animal from small chunks of tissue, they have become a focus of intense study. By examining the flatworm species Schmidtea mediterranea, Cote and other members of the Reddien lab have uncovered the ways cells communicate after injury to coordinate regeneration. Their work provides insight into how the ability to regenerate evolved, and how the healing process works in a variety of animals, including humans.

– –

Although regeneration seems mysterious, researchers have simplified the feat into two steps. First, planarians create the raw material to make new body parts by stimulating a group of rapidly dividing cells, called stem cells, that are the source of all new tissue in the worm. Second, these new cells need instructions to know what kind of tissue to become. Cote’s goal is to demystify this second step by locating a grid of information, like latitude and longitude lines on a map, that helps planarians keep track of their body parts and sense what is missing.

Hands suctioning small, black dots from petri dish.
Few creatures can regenerate like planarians, a class of flatworms found in fresh and salt water habitats around the world.

“The animal could have lost just the tip of its head or entire left side of its body,” Cote says, “and somehow it regrows the precise anatomy needed to make a complete worm.”

Over the past few years, research in the Reddien lab has demonstrated that a network of muscle cells spread throughout the worm’s body guides regeneration. To accomplish this task, muscle cells rely on a group of genes called position control genes (PCGs) which, based on Cote’s model, are predicted to encode proteins involved in cell communication. Depending on what PCGs are activated or expressed, muscle cells would send out a unique combination of signaling molecules that determine which body parts, such as eyes, stomach, or tail, would form.

“We like to imagine that muscle cells function like satellites and beam down information,” Cote says. “This allows stem cells to know where they are and what new body part to become.”

– –

To systematically identify PCGs from the roughly 20,000 genes expressed in Schmidtea mediterranea, Cote worked in tandem with postdoctoral researcher Lucila Scimone in the Reddien lab to perform a two-part study. First they created maps of gene expression by examining individual muscle cells. After inventorying the genes each individual muscle cell expressed, they aggregated the data into a whole body map, showing gene activity across the entire worm. Some genes were expressed in all muscle cells, implying a general function such as controlling contraction and relaxation. In contrast, other genes were expressed in precise regions of the worm, like the head or midsection, suggesting that they could act as PCGs by defining the identity of each area.

In the second half of the study, Cote and Scimone used molecular techniques to disrupt the activity of potential PCGs. “We hypothesized that if a gene were needed to direct regeneration, the worm would still be able to renew itself without that gene’s activity,” Cote says, “but the animal would end up with an abnormal body.”

Indeed, Cote found that disrupting four genes in particular, encoding signaling molecules and receptor-like proteins, led to defective regeneration; worms either grew extra eyes on their head or grew extra feeding tubes sprouting out of their midsection like elongated suction cups.  Together these four genes, along with a few previously identified genes controlling head and tail regeneration, comprise a short but expanding list of PCGs controlling the location and identity of new tissues. As scientists begin to understand the molecular details of planarian regeneration, they will test whether similar genes are used by other animals and humans.

– –

Although a biologist now, Cote began her academic life focusing on mathematics. As an undergraduate math major at the University of Chicago, she studied branches of mathematics such as analysis, algebra, and algebraic topology, a discipline that describes the properties of multidimensional shapes. After a summer project, Cote realized that — while she enjoyed learning mathematics — she found the research far too abstract.

“I was having a mid-college crisis,” she recalls. “I wanted to study something more visual where you could actually see what is going on.” Following this urge, Cote began to work in a lab examining fly development during her junior year. “I remember watching sheets of cells on the outside of a fly embryo folding in on themselves and sliding under the surface away from view. It made me wonder how cells make decisions and choreograph their movements to build a body. That’s how I got interested in developmental biology.”

After graduating from the University of Chicago, Cote worked as a lab technician for two years. During this time, she realized that her background in math and ability to think logically was an asset. “Putting together a mathematical proof is similar to publishing a research paper,” she says. “In both cases you are piecing together smaller bits of evidence into a cohesive argument.”

A series of blobs with white, green, purple and yellow specs inside them.
Gene expression maps from the first half of Cote’s and Scimone’s study. The head of the worm faces the top of the screen while the tail of the worm faces the bottom of the screen. Each worm is marked by purple, yellow, and green dots indicating the expression of three different genes expressed in muscle cells. These colors show how genes are localized to different areas of the worm and could act as PCGs.  In the second half of the study, Cote and Scimone identified PCGs by using molecular techniques to disrupt gene activity and looking for worms that regenerated abnormal bodies.

Encouraged by her successful venture into biological research, Cote decided to pursue a PhD in biology. She learned about the Reddien lab while taking a genetics course during her first year at MIT. Like Cote, many members of this group have backgrounds in other areas of science — including computational biology, development, evolution, biochemistry, and immunology — which helps them examine planarian regeneration from many perspectives.

“They were beginning to put together a story linking muscle cells to regeneration that was really intriguing,” Cote says. “I also liked the challenge of working with planarians because they are a fairly new lab animal. We’re still developing a lot of research tools so there is room to be creative and ask fundamental questions.”

By following an initial strand of curiosity as an undergraduate and identifying PCGs as a graduate student, Cote has begun to decipher the molecular language of regeneration.  As scientists learn more about how planarians replace missing body parts, new areas of exploration open. One pressing question­ is how planarian regeneration compares to that of other animals. To pursue that mystery, Cote plans on studying another animal as a postdoctoral researcher and eventually starting her own laboratory.

“I still haven’t made up my mind, “she says, “but I’m considering a lot of possibilities such as crustaceans, sea squirts, zebrafish, and axolotls.” Regardless of her final choice, Cote will be investigating how cells — essentially fatty membranes encasing a slurry of water and proteins — manage to form complex and intricate structures. She will be pursuing the same questions that first captivated her as an undergraduate in Chicago. “How do cells make decisions? How do they know to become an eye or a stomach or a brain?” she asks. “There is a lot more that I want to understand.”

Photo credit: Raleigh McElvery
Pairing mismatch helps impaired fish RNA cleavage proceed swimmingly
December 21, 2017

Beyond tending to its multitudes of genetic, metabolic, and developmental processes, eukaryotic cells must additionally be vigilant against invasion by parasitic sequences such as viruses and transposons. RNA interference (RNAi) is a defense used by eukaryotic cells to protect themselves from such threats to their genomic harmony. Cellular RNAi components slice and destroy invading double-stranded RNA sequences and also help snip and process microRNAs, RNA sequences encoded by the genome that play key roles in gene regulation. An important process that occurs naturally in our cells, RNAi has also been harnessed by scientists as a tool to study gene function in common models such as worms, fruit flies, and mice. While many researchers have been using RNAi to tease apart gene function for over a decade, those using zebrafish, a powerful vertebrate model, have been forced to use other approaches because RNAi just did not seem to work well in these animals. Now, researchers at Whitehead Institute have uncovered how small changes in the fish Argonaute (Ago) protein, an RNA slicing protein, that happened in its lineage an estimated 300 million years ago greatly diminished the efficiency of RNAi in these animals, while another ancestral feature, in a critical pre-microRNA, was retained that enabled the microRNA to still be produced despite the fish’s impaired Ago protein.

In an article published December 21 in the journal Molecular Cell, graduate student Grace Chen, along with both Whitehead Member David Bartel, also a professor of biology at Massachusetts Institute of Technology (MIT) and investigator with the Howard Hughes Medical Institute, and Whitehead Member and MIT professor of biology Hazel Sive, describe their discovery of a roughly 300 million-year-old, two amino acid substitutions in the fish Ago protein. The substitution is present in the ancestor all teleost fish, the class of fish which includes not only zebrafish but also the vast majority of fish species spanning those populating the ocean, aquarium, and supermarket. These two changes reside in and near the protein’s catalytic site and greatly decrease the ability of the fish Ago to perform its RNA slicing function, offering an explanation for why RNAi has not been a useful tool in zebrafish.

Despite the zebrafish’s deficiencies in RNAi, it is still able to produce the microRNA miR-451, an important regulator of red blood cell maturation and the only microRNA processed by Ago (the rest are produced with another protein called Dicer). MicroRNAs are short stretches of RNA that can regulate gene expression by inhibiting translation of mRNA into a protein and directing the destruction of mRNA before it can be used to make more protein. Since Chen had discovered that zebrafish lack an efficient Ago protein, it was mysterious as to how are fish were able to produce Ago cleavage-dependent miR-451. The Ago protein must process miR-451 by slicing the sequence out of a longer strand of RNA that has folded up on itself, forming a hairpin structure. What they determined was that in the pre-miR-451 hairpin in zebrafish, at a critical position in the miRNA, they found a “G–G” pairing mismatch that actually appears to facilitate cleavage by the impaired zebrafish Ago. No mismatch, no efficient cleavage.

Exploring the effects of a seed sequence mismatch on Ago-catalyzed cleavage kinetics further, they then tested its ability to slice other bound transcripts. The researchers discovered that while, as might be expected, a G–G mismatch slows Ago binding, it significantly enhances both slicing efficiency as well as the release of the bound product, more than off-setting the slower binding reaction kinetics and suggesting that non- “Watson–Crick” base pairing creates an exceptionally favorable geometry for the cleavage and release parts of the reaction.

These findings offer interesting insights into how animals can survive and thrive without an efficient RNAi system and suggest how the Ago protein could be “repaired” in order to allow zebrafish researchers to use RNAi in their experiments. Restoring a function that a lineage hasn’t had for 300 million years might also fuel additional findings into how the teleost class has diverged over time.

Written by Lisa Girard
***
David Bartel’s primary affiliation is with Whitehead Institute for Biomedical Research, where his laboratory is located and all his research is conducted. He is also a professor of biology at Massachusetts Institute of Technology and investigator with the Howard Hughes Medical Institute.
***
Paper cited:
Chen GR, Sive H, and Bartel DP. A Seed Mismatch Enhances Argonaute2-Catalyzed Cleavage and Partially Rescues Severely Impaired Cleavage Found in Fish. Molecular Cell, Dec 21 2017 DOI: 10.1016/j.molcel.2017.11.032.
Harnessing nature’s riches
December 19, 2017

Cambridge, MA – Researchers at Whitehead Institute have reconstructed the full suite of biochemical steps required to make salidroside, a plant-derived compound widely used in traditional medicine to combat depression and fatigue and boost immunity and memory. Their new study, which appears online this week in the journal Molecular Plant, resolves some long-standing questions about how this compound is manufactured by a type of high-altitude plant, known commonly as golden root. This work not only paves a path toward large-scale synthetic efforts—thereby protecting plants already in danger of extinction—but also provides a model for dissecting the biochemical synthesis of a host of natural products, which represent a treasure trove for modern medical discoveries.

“By cracking open the natural synthesis of this compound, known as salidroside, we have helped eliminate a major bottleneck in the broader development of plant-derived natural products into pharmaceuticals,” says Jing-Ke Weng, the senior author of the paper, a Member of Whitehead Institute, and an assistant professor of biology at Massachusetts Institute of Technology. “We simply can’t rely on the native plants as the sole sources of these biologically important molecules.”

Golden root, also called Tibetan ginseng, typically grows in high-altitude, arctic environments, such as Tibet. It is well known in Eastern cultures for its medicinal properties and produces a variety of chemical substances, particularly salidroside, which have garnered interest in the biomedical research community for their potential therapeutic effects.

“People have tried to farm golden root, but the medicinal value is much lower because the plants make much less salidroside when cultivated outside of their normal habitat,” says Weng.

That means collecting enough salidroside to fuel scientific studies is largely impossible, without risking the viability of these plants and their surroundings. So Weng and his team, including first author Michael Torrens-Spence, set out to find a better way. “If we can figure out how plants make these high-value natural products, then we can devise sustainable engineering approaches to recreate such molecules—we won’t have to destroy nature in order to harness its riches,” says Torrens-Spence, a postdoctoral researcher in Weng’s laboratory.

Torrens-Spence and his colleagues used a systematic multi-omics approach to characterize various tissues from a three-month-old, greenhouse-grown golden root plant. By correlating the active genes with the abundance of key metabolites between various tissue types, the researchers created a massive biochemical catalog of the plant’s tissues.

The researchers then mined these data and matched the likely biochemical precursors of salidroside with the candidate genes (and their corresponding enzymes) responsible for those compounds’ synthesis. This approach allowed Weng and his team to create a kind of draft blueprint of how salidroside is made in nature.

To test the validity of this draft blueprint—and the molecular players from the golden root plant that comprise it—the scientists turned to two well-studied laboratory organisms: the baker’s yeast Saccharomyces cerevisiae and the tobacco plant Nicotiana benthamiana. Normally, these organisms do not make salidroside. But if the researchers’ model was correct, by inserting the candidate genes involved in salidroside synthesis Weng and his colleagues should be able to bestow that special property upon them.

That is precisely what the researchers did. Using three key enzymes they identified through their “-omics” approach, including 4HPAAS (4-hydroxyphenylacetaldehyde synthase), 4HPAR (4-hydroxyphenylacetaldehyde reductase), and T8GT (tyrosol:UDP-glucose 8-O-glucosyltransferase), they engineered yeast and tobacco plants with the capacity to make salidroside. Notably, this biochemical pathway for synthesizing salidroside involves three enzymes, rather than four, as had previously been proposed.

“This is an exciting proof-of-principle for how we can systematically unlock the biochemistry behind a range of intriguing plant-derived natural products,” says Weng. “With this capability, we can accelerate biomedical studies of these unique compounds as well as their potential therapeutic development.”

Written by Nicole Davis
* * *
Jing-Ke Weng’s primary affiliation is with Whitehead Institute for Biomedical Research, where his laboratory is located and all his research is conducted. He is also an assistant professor of biology at Massachusetts Institute of Technology.
* * *
Full citation:
“Complete pathway elucidation and heterologous reconstitution of Rhodiola salidroside biosynthesis”
Molecular Plant, online December 19, 2017. DOI: 10.1016/j.molp.2017.12.007
Michael P. Torrens-Spence (1), Tomáš Pluskal (1), Fu-Shuang Li (1), Valentina Carballo (1) and Jing-Ke Weng (1,2).
1. Whitehead Institute for Biomedical Research, Cambridge, Massachusetts, USA
2. Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA