As society reopens, scientists aim to close in on COVID
Illustration by Nate Kitch

Forefront

As society reopens, scientists aim to close in on COVID

Illustration by Nate Kitch

For many, news of the first “breakthrough” COVID cases was alarming. But for scientists, it was expected—and presented an opportunity.

“We always knew there would be a certain number of people who develop infections even after being fully vaccinated,” says Robert B. Darnell. “What we didn’t know was what those cases would look like.” How severe would they be? Would some SARS-CoV-2 variants prove more adept at breaking through the vaccines’ protection than others? How would these cases impact the course of the pandemic?

As the global health crisis enters its second summer, the nexus of COVID research is shifting. We’ve come to understand the basics of how the virus infects host cells and replicates, and we’ve learned enough about the body’s immune response to create several good vaccines. But the world’s long-term relationship with this coronavirus, and other viruses like it, is still an open question.

One clue to how things will progress comes from surveillance within the Rockefeller community. Since January, mandatory weekly COVID testing of all on-site Rockefeller personnel has been conducted in-house by Darnell’s lab, using a saliva-based PCR test he and his colleagues developed (for more on the test, see Building a better COVID test). In addition to keeping the community safe, this program has produced a wealth of information, and it was among the first to document and explore what breakthrough cases look like at the clinical and genetic level.

Teresa Rozza and Salina Parveen
Teresa Rozza and Salina Parveen prepare saliva samples for COVID-19 testing. Photo by Matthew Septimus

The results suggest reasons for both confidence and caution. The vaccines are holding up well to known variants, such as those originating in Brazil and the United Kingdom, and they prevent severe disease. But even a highly successful vaccination program doesn’t mean the end of COVID.

“Based on what we’ve seen, routine testing of any individual with flu-like symptoms, or those who have had contact with a positive case, will remain an important tool to prevent the spread of this disease for some time,” says Darnell, who is Robert and Harriet Heilbrunn Professor.

Meanwhile, pursuing treatments for COVID remains as important as ever. Monoclonal antibodies have shown exceptional promise over the past year, and one version developed at Rockefeller—a combination of two antibodies originally isolated from COVID patients who successfully fought off the infection early in the pandemic—entered clinical trials this January.

Similar antibody-based drugs have been used experimentally in thousands of COVID patients, and these drugs help stop the infection in its early stages before it progresses to severe disease. The cocktail developed by Michel C. Nussenzweig, the Zanvil A. Cohn and Ralph M. Steinman Professor, and his collaborators including virologists Paul Bieniasz and Theodora Hatziioannou, recently licensed to Bristol Myers Squibb, is designed to help minimize the risk of the virus mutating and developing resistance to the therapy.

Scientists are also pursuing new antiviral drugs that, similarly to broad-spectrum antibiotics, might be effective against multiple pathogens. A group led by Nobel Prize–winning virologist Charles M. Rice, the Maurice R. and Corinne P. Greenberg Professor in Virology, mapped a network of more than a hundred human proteins that SARS-CoV-2 hijacks as it takes over a cell’s replication machinery. One of them, a little-known protein called TMEM41B, stands out for its use by four different coronaviruses as well as by viruses that cause Zika, yellow fever, and other diseases. The team is investigating ways to disrupt TMEM41B’s ability to support an infection.

Covid Illo

Data

There are many thousands of SARS-CoV-2 variants, and over 2,600 distinct lineages have been discovered so far. Four are considered “variants of concern” by the CDC.

Other researchers are studying how the virus impacts lung cells specifically. Because SARS-CoV-2 first enters the body via the lungs, its interaction with cells in the airways and alveoli is what allows it to establish a foothold in the body. A team led by Ali H. Brivanlou, Robert and Harriet Heilbrunn Professor, has used stem cell technology to produce tissue that mimics lung buds, the embryonic precursor to mature lungs (see Synthetic micro lungs). Beyond providing a realistic model to investigate the mechanisms of viral infection, the method can quickly produce vast amounts of lung tissue for drug-screening purposes.

As the pandemic evolves, so do our questions. What does the immune response to SARS-CoV-2 look like months or a year after infection? How does vaccination impact people who have already been infected? How well do our antibodies adapt to deal with the emerging variants of the virus? Bieniasz and Hatziioannou are studying the shifting relationship between our antibodies and the virus. Working with Nussenzweig, their team has found that in those who recover from COVID, the immune system retains a memory of the coronavirus, building a long-lasting defense in which antibodies are continually refined and improved.

What’s more, their work suggests that vaccination further boosts the neutralizing power of antibodies: Individuals who receive vaccines after having recovered from COVID should enjoy high levels of protection, even against emerging variants they haven’t yet encountered. Vaccinated individuals who haven’t been exposed to the virus, however, retain some vulnerability to the variants, their work shows.

“It’s a complex situation,” Bieniasz says. “And it suggests that vaccines may need occasional updates in the future to keep up with the mutating virus.”

A Nobel like no other
In lieu of holding a press conference, Rice recorded TV interviews. Photo by Santa Maria Pecoraro Di Vittorio

Forefront

A Nobel like no other

In lieu of holding a press conference, Rice recorded TV interviews. Photo by Santa Maria Pecoraro Di Vittorio

To receive his 2020 Nobel Prize, Charles M. Rice had to travel no farther than midtown Manhattan. For the first time in decades, the event involved no trip to Stockholm, no lavish banquet, and certainly no handshake with the king of Sweden—only a quiet socially distanced ceremony at the Swedish Consulate.

Still, as is usually the case with Nobel laureates, Rice spent many long hours on camera; working double time as a spokesperson for basic science while conducting intensive investigations into COVID-19.

Rice, the Maurice R. and Corinne P. Greenberg Professor in Virology, shares his Nobel Prize in Physiology or Medicine with two other scientists for discoveries that led to the identification and characterization of the virus responsible for hepatitis C. After proving the pathogen’s role in causing the disease, his continued research enabled the creation of new drugs, a combination of which was ultimately shown to cure it.

This fly likes its fruit fresh
Illustration by Erick M. Ramos

Forefront

This fly likes its fruit fresh

Illustration by Erick M. Ramos

Compared to other pests, the fruit fly is relatively docile. It is tiny, it is quiet, it doesn’t bite—nor is it out to destroy anything of great value. In general, members of the Drosophila genus are attracted to rotting produce, food that nobody wants anyway.

The spotted wing drosophila, known to scientists as Drosophila suzukii, is an exception that has developed a taste for ripe summer fruits. It feasts in orchards and fields while fruit is still early in the ripening stage, damaging crops and leaving behind microscopic larvae. That box of fresh, plump cherries from the market? It looked good to D. suzukii, too.

“The preference for ripe fruit is a novel behavioral trait that is causing significant agricultural losses,” says Li Zhao, assistant professor and head of the Laboratory of Evolutionary Genetics and Genomics. “Understanding how it emerged may lead to new ways of controlling the damage it causes.”

308.9 million

Data

Cost of damage caused by D. suzukii to California strawberry crops in 2008, the first year the pest was observed in the state.

D. suzukii arrived in North America about 12 years ago, seemingly out of nowhere. Long confined to East and Southeast Asia, it had voyaged all the way to California, where farmers were at first dumbfounded trying to figure out what was ruining their crops. It was around this time that Zhao, then a postdoc at the University of California, Davis, got involved in a USDA-funded project to put together an early draft of the fly’s genome using next-generation sequencing technology.

Zhao’s lab is devoted to the study of how novel genes develop. Now she is getting to the bottom of what exactly has caused D. suzukii’s preferences to change as it evolved. “Suzukii’s strange behavior is a perfect case study for us,” Zhao says.

The obvious place to look is in genes that are important for sensory perception, such as those coding for smell receptors. Previously, scientists hypothesized that D. suzukii, too, parted ways with its evolutionary kin after a mutation. Perhaps its perception of a meal’s sugar or alcohol content, which varies as fruits ripen and then rot, changed.

But when Zhao and her colleagues compared D. suzukii to its closest relatives, they found something quite different. It turns out that this fly picks fresh fruit over rotten not as a matter of taste or smell but based on the firmness of the fruit. When offered servings of a gelatinous milk shake containing varying amounts of alcohol, sugar, acetic acid, and agar, D. suzukii consistently chose the firmest, regardless of its chemistry. And a detailed genetic analysis of 200 individual flies revealed that some of the most rapidly evolving genes in D. suzukii are those coding for mechanosensory receptors.

“Our ability to control this invasive species could rely on a better understanding of mechanosensation—such as the processes by which flies are able to detect how much force is required to manipulate an object,” says Zhao. “It’s a new direction to explore.”

The Prozac problem
Laboratory of Molecular and Cellular Neuroscience

Forefront

The Prozac problem

Laboratory of Molecular and Cellular Neuroscience

Prozac doesn’t always work—and when it does, it takes too long to kick in.

“The rate of suicides drops after nine days of treatment, and people start to feel better only after two to three weeks,” says Revathy Chottekalapanda, a senior research associate in the laboratory of the late Paul Greengard.

122.4 million

Data

Number of SSRI prescriptions filled annually in the United States.

Why selective serotonin reuptake inhibitors (SSRIs) like Prozac take so long to start working—and why they fail some people entirely—is a mystery that dates back over 40 years, to when the drugs were first introduced. Chottekalapanda and her colleagues have a new theory, centered on a single gene that, in mice, ramps up exactly on day nine of Prozac treatment. This molecular switch triggers a cascade of gene-expression changes that transform the animals’ behavior, reducing symptoms of depression and anxiety.

“For the first time, we were able to put a number of molecular actors together at the crime scene in a time- and sequence-specific manner,” Chottekalapanda says of the findings, which were published in Molecular Psychiatry.

Wide-open spaces
Photo by Matthew Septimus

On Campus

Wide-open spaces

Photo by Matthew Septimus

At midday, the six-story atrium at the heart of the university’s Collaborative Research Center, which houses about a third of Rockefeller’s laboratories, is normally bustling with activity as scientists congregate to catch up, share data, and refine ideas. But places designed to spread knowledge are also good places to spread viruses. During the winter, when COVID policies required strict social distancing, in-person meetings were suspended and most of the furniture was removed. What remained was a calming quiet, and plenty of elbow room for those needing to catch up on reading or e-mail.

Quality genomes for all
Ancient and endangered, the kakapo is one of 25 species with a brand-new reference genome. Photo by Chris Cunningham

Forefront

Quality genomes for all

Ancient and endangered, the kakapo is one of 25 species with a brand-new reference genome. Photo by Chris Cunningham

There was no reliable genomic sequence for scientists to consult when studying the flightless kakapo of New Zealand. Nothing on the adorable vaquita porpoise or the blunt-snouted clingfish either. No error-free genetic database for bats or platypuses, Canada lynxes, or Goode’s thornscrub tortoises.

When it comes to vertebrates—other than humans, of course, and popular lab animals such as mice and zebra fish—scientists are often stumbling in the dark. Reference genomes of tens of thousands of species either don’t exist or are unusable, rife with errors and duplications.

Number 7,262

Data

Average number of manual “edits” required to properly assemble a high-quality vertebrate genome sequence.

“It is unconscionable to be working with some of these genomes,” says Rockefeller’s Erich D. Jarvis.

From the collective groan of frustrated scientists, the Vertebrate Genomes Project was born. Its goal is to build a library of more than 70,000 error-free reference genomes representing every vertebrate species alive today. Projected to take at least 12 years, the endeavor recently reached an early milestone with the release of its first 25 premium genomes. Reported in a series of papers in Nature, this work provides a proof of concept for a new method that merges several sequencing tools into one lean pipeline.

Reference genomes of tens of thousands of species either don’t exist or are unusable, rife with errors and duplications.

“We call it the kitchen sink approach, combining tools from several DNA sequencing companies to make one high-quality genome,” says Jarvis, who chairs the project. Reference genomes that once took years to generate are now rolling out in weeks or months, and scientists at several institutions are already using the new approach in their research. “It often pays off to do some hard work on the front end so that we can get high-quality data on the back end,” says Jarvis, whose Laboratory of Neurogenetics of Language studies vocal learning in songbirds, hummingbirds, and other species.

But plenty of work still lies ahead. “The next step is to sequence all 1,000 vertebrate genera, and then all 10,000 vertebrate families, and eventually every single vertebrate species.”

The perils of long telomeres
Human telomeres (green) at the ends of chromosomes (blue).

Forefront

The perils of long telomeres

Human telomeres (green) at the ends of chromosomes (blue).

It was once thought that all cells were immortal, forever able to replicate and generate fresh copies of themselves. But a cell’s days are, in fact, numbered, predestined by the length of its telomeres. Located at the tips of each chromosome, these structures shorten as they absorb the wear and tear of cell division. Eventually, a cell’s telomeres wither away entirely, capping the number of times it can divide at about 50.

There may be good reasons why this threshold hovers consistently around 50 and not, say, 25 or 500. Scientists have for decades suspected that telomere shortening isn’t just an unwanted side effect of cellular aging but a carefully calibrated process that proactively curtails cell division to prevent cancer. And the telomere reserve we are born with is key, with each telomere being long enough to allow normal development yet short enough to run out before rapidly proliferating cells start amassing into tumors.

Studying four Dutch families with striking cancer histories, scientists in the lab of Titia de Lange, the Leon Hess Professor, recently provided a real-world example of this theory.

The six individuals in the study had each developed one or several cancers of different types, including breast, colorectal, thyroid, and skin cancer. The researchers found that because these patients had mutations in TIN2, a protein that keeps telomere length in check, they had also been born with extremely long telomeres. The work was published last December in eLife.

All of which suggests you can thank your normal-sized telomeres for every single cell in your body that hasn’t run amok. They may not seem to stand against the cruel passage of time, but they likely have prevented many cancers from occurring in your lifetime.

Predicting the unprecedented
Illustration by Gizem Vural

Forefront

Predicting the unprecedented

Illustration by Gizem Vural

Standard mathematics can predict how cancer cells will multiply, how crop yields will fluctuate, and how insects will swarm—in much the same way that statistics can determine the average human height. As long as measured quantities have finite averages and variances, figuring out the specifics is a simple matter of applying a formula known as Taylor’s law, which relates a population’s mean to its variance.

But what about extreme events with no finite limits—pandemics like COVID-19 or financial fluctuations like the GameStop short squeeze?

The data sets that describe extreme events are known as heavy-tailed distributions. While most aspects of our daily lives huddle around an average—a neat bell curve of mundane behavior, minor disease outbreaks, small blips in a stable market—extreme events are plotted at distant tails of the graph. When there’s no finite limit to how extreme an event can be, then there’s no limit to how far its tail can be flung or how “heavy” it can grow. One extreme event can stretch the entire graph into unpredictable territory. It follows that Taylor’s law loses its footing in a heavy-tailed world.

Rockefeller’s Joel E. Cohen disagrees. His recent work on heavy-tailed distributions, which he published with colleagues at Columbia University and Cornell University in Proceedings of the Royal Society A, describes how Taylor’s law can predict even extreme outliers. The study proposes a novel way of looking at heavy-tailed variables that yields surprisingly orderly connections between the mean and the variance of a system.

Cohen’s discovery does not mean that scientists can now simply plug their numbers into an equation and foresee the next market coup. But it does raise the prospect that mathematical modeling may one day help scientists anticipate and manage extreme occurrences, “from daily precipitation to microbial evolution, from cortical oscillations in the human brain to global pandemics,” says Cohen, who is the Abby Rockefeller Mauzé Professor. “Advances like these are the mathematical analogue of bioimaging—they make it possible to see what was previously invisible.”

An old drug solves new problems
Campbell examines structural data. Photo by Mario Morgado

Forefront

An old drug solves new problems

Campbell examines structural data. Photo by Mario Morgado

Sorangicin A has never lived up to its potential. It’s been three decades since the compound was discovered to have antibiotic properties, yet it languishes in obscurity—ignored by all but a few scientists.

Meanwhile, antibiotic resistance grows. Every year, about half a million people fall ill with tuberculosis that doesn’t respond to conventional antibiotics such as rifampicin. Resistant TB strains, experts warn, are a ticking bomb.

It might be time to give sorangicin a second look. A recent study found that sorangicin, first discovered in the 1980s, can kill even drug-resistant TB. “Sorangicin inhibits regular strains in very much the same way as rifampicin, by targeting the molecular machinery that transcribes DNA to RNA,” says Elizabeth Campbell, a research associate professor at Rockefeller. “But now we show that, through a different mechanism, it also traps those variants that escape rifampicin.” The work was published in Proceedings of the National Academy of Sciences.

Campbell and colleagues are particularly excited about sorangicin as a potential drug candidate because of its compatibility with other medications. Rifampicin, on the other hand, has been shown to reduce the efficacy of HIV medications by up to 90 percent.

“If sorangicin can be developed into a medication, it might be especially helpful for people with comorbidities,” she says.

The weight you don’t want to lose
Brown fat deposits are found between the shoulders and along the spine. Photo by Andreas G. Wibmer and Heiko Schöder, Memorial Sloan Kettering

Forefront

The weight you don’t want to lose

Brown fat deposits are found between the shoulders and along the spine. Photo by Andreas G. Wibmer and Heiko Schöder, Memorial Sloan Kettering

Brown fat can be just as hard to acquire as white fat is to lose. Newborns and animals have a surplus of the stuff, which burns rather than hoards calories. Adult humans, not so much.

“The natural question that everybody has is, ‘What can I do to get more brown fat?’” says Paul Cohen, the Albert Resnick, M.D. Associate Professor and senior attending physician at The Rockefeller University Hospital. “We don’t have a good answer to that yet.”

Recently, however, Cohen’s team discovered that brown fat has many benefits beyond waistline control. Published in Nature Medicine, their study of 52,000 people suggests that 10 percent of adults have detectable amounts of brown fat, and that these individuals are less likely to suffer from type 2 diabetes, heart disease, and hypertension. Brown fat also appeared to mitigate the negative health effects of white fat in those who were obese.

“We are looking into the possibility that brown fat tissue does more than consume glucose,” Cohen says.

How cells feel their way
Alushin (right) in the lab. Photo by Frank Veronsky

Forefront

How cells feel their way

Alushin (right) in the lab. Photo by Frank Veronsky

Cells are tactile little things. Whether bumping up against their neighbors or clinging alone to the bottom of a scientist’s Petri dish, they are able to sense their physical environment as well as we can feel the push of the ground beneath our feet. Scientists have long known that mechanical signals flow from the outside environment into the cell and inform its movements, but only recently have they acquired the technology to study this phenomenon, known as mechanosensation, in detail.

In Gregory M. Alushin’s Laboratory of Structural Biophysics and Mechanobiology, scientists have taken a major step toward describing how mechanosensation plays out on a molecular level. It all comes down to actin, a protein involved in giving the cell its shape, and its biochemical ally α-catenin, a so-called adhesion protein found in the cell’s outer rim. Using specially designed laser tweezers, the researchers were able to stretch out single actin filaments, which are about 15,000 times thinner than a human hair, to demonstrate that actin transmits a signal to α-catenin when stretched. The α-catenin protein heeds the call, responding to actin’s transmission by either tightening or loosening its grip on the external environment.

“The idea that actin filaments could potentially be tiny stretchy tension sensors in the cell has been banging around in the literature for a while, but we’ve proved it here,” Alushin says.

Although α-catenin is known to be critical in brain development and is frequently mutated in cancer, scientists have had a hard time pinpointing its exact role. “We know that if you get rid of it, everything else in the cell breaks, but not much more,” Alushin says. “But by defining the force-detector in α-catenin, we will enable researchers to manipulate the protein with better precision.” The results were published in eLife in September.

How mice miss the exit
The Jackson Laboratory

Forefront

How mice miss the exit

The Jackson Laboratory

Any mouse worth its whiskers can navigate a maze.

But mice are curious creatures, and they prefer to explore new arms of a maze rather than run the same route again and again. So they memorize the paths taken previously, and then on future maze runs, instead of turning down a well-trodden path, they make a point to seek out new adventures.

In the laboratory of Priya Rajasethupathy, however, a few forgetful mice often fail to find the road not taken. They hesitate at forks in the maze, wanting to turn down a new path but struggling to remember where they’ve previously scurried. They get it right half the time—which means they’re guessing instead of relying on their short-term memory.

“It’s rare to find a single gene with a strong influence on a complex cognitive function, but it happened in this case.”

In recent work published in Cell, Rajasethupathy and colleagues discovered that variations in a single gene, which codes for a brain receptor called Gpr12, can explain to a great degree these differences in short-term memory among mice. They found that mice with excellent short-term memories have more than twice the Gpr12 receptors as forgetful mice and that, by boosting the expression of this one gene, scientists can help absentminded mice make the right turn 80 percent of the time.

“It’s rare to find a single gene with a strong influence on a complex cognitive function like short-term memory,” says Rajasethupathy, the Jonathan M. Nelson Family Assistant Professor. “But it happened in this case, and it led us to the unexpected mechanisms involved.”

One such revelation came when the scientists began exploring the Gpr12 receptor, which they thought would be restricted to the prefrontal cortex, the brain region classically linked to short-term memory. Instead, the receptors primarily function in the thalamus and help establish synchronized brain activity during memory tasks. “These findings reveal a crucial dialogue between brain regions during short-term memory use,” Rajasethupathy says.

Thirftiness is hardwired
Marine microbes minimize their use of nitrogen or carbon depending on which nutrient is available in their environment.

Forefront

Thirftiness is hardwired

Marine microbes minimize their use of nitrogen or carbon depending on which nutrient is available in their environment.

Recycling is good. Microbes that eat plastic? That could be even better.

But when David Zeevi and Liat Shenhav set out in search of such organisms, they ended up making a surprising discovery about how cells conserve their own resources.

The plan was to scan microbes living in diverse areas of the Earth’s oceans and identify which genes are essential to those that prosper in polluted areas rife with plastics. The telltale sign of such genes is high resistance to change: Being crucial to survival, the genes would not endure random mutations during evolution and would remain largely unchanged across a species.

But the analysis turned up an unexpectedly high number of genes that were stable in this way. Soon, the team had embarked on a new project. “The question became, Why are so many microbial genes intolerant of change?” says Zeevi.

As they sequenced more organisms, a pattern emerged: The most stable genes were often linked to the use of carbon or nitrogen, which microbes need to make proteins. In a sense, the bacteria were conserving scarce resources.

Zeevi and Shenhav, who are fellows in the Center for Studies in Physics and Biology, suggest there could be something in the structure of the genetic code itself that leads to this phenomenon. The genetic code, shared among all life forms, is composed of short segments of DNA called codons that specify the amino acids to be used in protein manufacturing—thereby affecting overall nutrient requirements.

Using computational modeling, the researchers simulated one million imaginary, randomized genetic codes and measured the overall nutrient cost of all possible mutations. It turns out that mutations to the randomized codes resulted in higher nutritional requirements than did mutations to the authentic one.

“The standard genetic code is set in a way that makes it less likely for mutations to cost the cell extra carbon and nitrogen,” Zeevi says. “This is the case not only in microbes in the ocean but in all life.”

Synthetic micro lungs
Laboratory of Synthetic Embryology

Snapshot

Synthetic micro lungs

Laboratory of Synthetic Embryology

What exactly happens when a SARS-CoV-2 virus, catching a ride on a breath, flies through and lands on the cells of our lungs? To tease out the complex biological events playing out inside infected human cells scientists need models, the more realistic the better.

Researchers from Ali H. Brivanlou’s Laboratory of Synthetic Embryology used stem cell technology to produce lung “buds” that self-organize into three-dimensional tissue akin to early lungs. The tiny synthetic lung (red) allows researchers to study how SARS-CoV-2 (blue) attacks the airways and alveoli, structures known to be damaged in COVID.

It’s a realistic model for testing whether novel COVID-19 drugs curtail infection, says Brivanlou, who is Robert and Harriet Heilbrunn Professor.

The lemon aid
Pichest

Science gadget

The lemon aid

The saliva covid test developed at Rockefeller is both highly accurate and very convenient—except when your mouth is dry. In that case, you need a lemon.

Studies have shown that sour foods increase saliva production more than sweet foods do, and that lemon in particular gets our juices flowing. In one experiment, undiluted lemon juice was shown to increase saliva production from 0.61 to 1.44 milliliters per minute. (The Rockefeller test requires just a tenth of a milliliter.)

Among the first members of the Rockefeller community to receive weekly COVID testing were the children who attend the university’s faculty and staff childcare facility. While newborns typically don’t have trouble generating drool, toddlers and preschoolers can find it difficult to spit.

“We started recommending lemon last summer when the testing program began and some parents were struggling to get their kids’ saliva,” says Samara Wright, the childcare center’s nurse. “It definitely helps create more drool and speeds up the collection process.”

With adults, just the thought of a lemon was often enough. “We found that imagining a lemon has nearly the same effect,” says Myles Marshall, a technician in the Darnell lab who worked on the testing protocols.

For more on COVID testing with saliva, see Building a better COVID test.