C&EN’s Year in Chemistry 2021

In 2021, plastics recycling got a whole new vibe. At least that’s the way Bert Weckhuysen, an inorganic chemist at Utrecht University, puts it. The scientific community has known for years that recycling plastics is an important but difficult problem that has progressed slowly toward solutions.

In the past year, however, “there is really a surge—almost everyone is jumping on things,” Weckhuysen says. Chemists are studying many aspects of the plastics recycling problem. They are researching ways to break down plastics with catalysts, studying mechanical recycling, and rethinking how plastics are made in the first place. Plastics have entered a stage of circular thinking, Weckhuysen says, in which scientists design the polymers with a focus on what will happen to the materials after their initial use.

One way researchers designed plastics with reuse in mind was to make plastics out of polymers that can be chemically broken down into monomers, which can then be made into the polymers again. For example, this year, Cornell University chemist Geoffrey W. Coates and coworkers synthesized a polymer that chemists can both separate into monomers and recover from a mixture of plastics (Science 2021, DOI: 10.1126/science.abh0626).

On its own, the polyacetal polymer is stable at temperatures up to 325 °C. But after the researchers added a strong acid catalyst, they could break the polymer down into its monomer constituents above 73 °C. The monomers are liquid at this temperature, which made it easy for the team to separate them from other plastics that were still solids under these conditions. Brooks A. Abel, who is one of the study’s authors and is now at the University of California, Berkeley, told C&EN earlier this year that the researchers picked through recycling bins, chopped up the plastics they found, and dumped the mix into the flask with their polymer to test how easy it was to separate out the broken-down monomers. These samples contained labels, glues, “and probably a little bit of Gatorade,” Abel said. Even so, the team recovered as much as 98% of the monomer.

In this same vein, Junpeng Wang and colleagues at the University of Akron developed a polymer that scientists can break into monomers with the help of ring strain (Nat. Chem. 2021, DOI: 10.1038/s41557-021-00748-5). The group created a polymer made of cyclooctene monomers in which a four-membered cyclobutane ring is fused to the eight-membered ring. With a ruthenium catalyst, the cyclooctene zips itself into a polymer below room temperature. The material can fall apart into monomers again when the chemists heat it and the catalyst to 50 °C. Without the catalyst, the polymer is stable up to 370 °C. The team can change functional groups on the polymer backbone to tailor its properties. Because of this variability, the researchers think that the polymers could find multiple uses, including as plastics or rubbers.

Developing new types of plastics and polymers that have built-in recycling features is an important approach to the plastics recycling problem, Utrecht’s Weckhuysen says. One major question is whether to make plastics biodegradable by design, he says, and this comes down to whether the polymers can still be durable. “Everything is made to be stable, to be durable, which is good,” Weckhuysen says. But how do you make something that’s biodegradable but also stable?

Credit: Nature

Encapsulated enzymes distributed throughout polycaprolactone can accelerate the breakdown of the material in the presence of heat and humidity. Photos show the material before and after composting for 3 days.

Ting Xu, a chemical engineer at the University of California, Berkeley, and colleagues have figured out how to make a plastic that’s both. The group encased plastic-chomping enzymes in a protective coating and incorporated the resulting nanoparticles into polycaprolactone and polylactic acid plastics as they were being made (Nature 2021, DOI: 10.1038/s41586-021-03408-3). When the researchers exposed these plastics to humidity and either heat or ultraviolet light, the protective coating broke down, releasing the enzymes. The team could break down as much as 98% of the polymers in 30 h, depending on the polymer and the temperature. The depolymerization process leaves behind lactic acid, which people can pour down the drain or add to their garden, Xu told C&EN earlier this year.

New plastics recycling technologies and strategies are needed because our society will keep using plastics, Weckhuysen says. “People can complain about plastic, but plastic has brought us a lot,” he says. We’ve seen the importance of plastics during the COVID-19 pandemic, he says. Hospitals need to use plastics to protect staff and patients against infection, such as in plastic face shields or medical waste bags. Our need for plastics will continue in the future, which Weckhuysen says will require chemists to rethink classical plastics.

Nitrogen-deletion chemistry can transform pyrrolidines into cyclobutanes.

When chemists construct drug candidates, there’s an aspect of the design process that’s kind of like grappling in the dark. Would the molecule fit its target better if it had a five-membered ring instead of a six-membered ring? What if there were a nitrogen instead of a carbon at a particular spot? Such subtle tweaks are easy to make on paper but often require chemists to build the redesigned molecule from scratch. Molecular editing—reactions which selectively insert, delete, or exchange an atom in a complex molecule—aims to change that.

Molecular editing is having a moment. In 2021, chemists published several high-profile papers that feature molecular editing. Chemists in Mark D. Levin’s lab at the University of Chicago developed a reaction that could snip the nitrogen out of a secondary amine, allowing chemists to make, for example, cyclobutanes from pyrrolidines (Nature 2021, DOI: 10.1038/s41586-021-03448-9). Levin’s lab also reported a reaction that could add carbons to certain nitrogen-containing rings, converting indoles to quinolines (J. Am. Chem. Soc. 2021, DOI: 10.1021/jacs.1c06287). Richmond Sarpong’s lab at the University of California, Berkeley, discovered a light-powered rearrangement that turns saturated heterocycles with six atoms into five-carbon rings with pendant amines, alcohols, or thiols (Science 2021, DOI: 10.1126/science.abi7183).

Blue light rearranges tetrahydropyrans into cyclopentyl alcohols.

Transformations like these tantalize medicinal chemists with the prospect of quickly and cleanly making dramatic changes to the complex molecules in their companies’ libraries. In fact, Sarpong and Levin say they were inspired to pursue these reactions after reading calls from pharmaceutical chemists to develop new reactions, like molecular editing, that could alter existing compounds to explore new chemical space.

Molecular editing “evokes something that I think everybody agrees on, which is that synthesis should be easier,” Levin says. “It should be as easy as editing a structure on your computer.”

Even with these remarkable new reactions, the phrase molecular editing causes some chemists to bristle. “If you take it on its face, it’s a synonym for a chemical reaction,” Levin says. “At a certain level, every reaction is a molecular edit because you are making a change to the molecule.”

“There are many in our community who will roll their eyes,” Sarpong adds. “They don’t want to add to the lexicon of terms that already exists in chemistry.”

Indoles react with a chlorodiazirine to form quinolines.

But Levin and Sarpong say molecular editing is distinct from well-established chemistry and from reactions that change substituents on the periphery of a molecule. Molecular editing, they say, transforms a molecule’s core structure—typically a ring of some kind—using just one or two reactions that aren’t intuitive. Levin says he sometimes describes molecular editing in terms of remodeling a house: you make a substantial structural change rather than tearing the whole house down and rebuilding.

“The aspect of it that truly, in my opinion, resonates with people is the precision that you can achieve,” Sarpong says. He points to Levin’s nitrogen-deletion paper as a prime example. “Add your pixie dust, and in a single step, you can get this magical transmogrification.”

Levin and Sarpong say that molecular editing is still in its infancy as a field. They liken it to where the field of C–H functionalization was 30 years ago. In that field, chemists coax reactivity out of specific C–H bonds on a molecule. C–H functionalization encouraged chemists to think about changing parts of a molecule that used to be considered inert, Levin says. Molecular editing takes that a step further, he says. It asks chemists to look at the whole molecule as an opportunity.

“There’s a lot that we can’t do yet,” Levin says. But he’s hopeful that someday soon, swapping a nitrogen atom for a carbon in benzene to make it pyridine, for example, will be as easy to do in a flask as it is to do with a pencil and an eraser.

Credit: Ian Haydon/UW Medicine Institute for Protein Design

Individual proteins (shown in various colors) fit together into complexes. Emerging modeling techniques have been used to predict these structures.

Structural and computational biologists made tremendous advances this past year using artificial intelligence techniques, like neural networks, to predict the way proteins fold. Methods that can take a protein sequence and predict how it will fold are critical to helping scientists better understand biological systems, including guiding drug developers to design molecules to hit specific targets. But more importantly, experts say this boom in protein structure prediction is the beginning of something, not the end.

The challenge of predicting protein shape and geometry is in taking the primary structure of proteins—the long list of individual amino acids—and spitting out a reliable 3D structure. There are 20 natural amino acids, each with a different side chain and therefore unique chemistry. These units interact with one another and the surrounding environment, causing protein strands to shimmy into various functional shapes, including small peptides that act as signaling molecules and large, complicated biomachines that run the basic functions of life.

Traditionally, solving these puzzles of structural biology involved experimental techniques such as X-ray crystallography and cryo-electron microscopy. But for decades, researchers have tried to build computer programs that could predict structures using just the basic information held in the linear sequence of amino acids.

In the early 1990s, computational biologist John Moult of the University of Maryland and several other scientists set up an international competition called the Critical Assessment of Protein Structure Prediction (CASP) to advance the field. Every 2 years, groups of scientists around the world competed in analyzing data sets and protein sequences and predicting structures. Over time, the algorithms got better and better. But then a team from the AI company DeepMind changed the game.

Credit: DeepMind

This model shows the viral protein ORF8 from SARS-CoV-2 used in the most recent Critical Assessment of Protein Structure Prediction competition. The prediction (blue) closely matches the experimentally determined structure (green).

At the end of November 2020, at CASP, the company demonstrated an AI tool called the AlphaFold algorithm, which could predict protein structures that matched what traditional structural biology techniques could obtain, the gold standard for evaluating such prediction algorithms.

“You could feel something extraordinary has happened,” Moult says, referring to AlphaFold’s success at CASP. But what’s more, “AlphaFold provided the proof of concept” for others in the field, he says. Rather than feel defeated, other computational chemists in the field seem to have been energized by the success of AlphaFold. In the last year, the publication rate of new papers on protein structure prediction has surprised even veterans of protein structure prediction. As Moult says, the announcement last year “was the beginning of something, not the end of something.”

For example, Sameer Velankar of the European Molecular Biology Laboratory (EMBL) and colleagues have worked with DeepMind scientists to use the AlphaFold code to predict huge numbers of unknown protein structures and create databases to hold them all (Nature 2021, DOI: 10.1038/s41586-021-03828-1). Meanwhile, David Baker’s team at the University of Washington has improved its prediction algorithm and provided the code to researchers in the community to use (Science 2021, DOI: 10.1126/science.abj8754).

Credit: Ian Haydon/UW Medicine Institute for Protein Design

This image shows the structure of human interleukin-12 bound to its receptor, as predicted by the software RoseTTAFold.

More recently, groups have managed to make progress in predicting protein-protein interactions and the structures of protein complexes, something that researchers thought would take several more years to solve. Baker’s team studied such complexes in the yeast proteome (Science 2021, DOI: 10.1126/science.abm4805), and a collaboration between Pedro Beltrao at the EMBL’s European Bioinformatics Institute and Arne Elofsson at Stockholm University worked with data relating to the human proteome (bioRxiv 2021, DOI: 10.1101/2021.11.08.467664). Throughout 2021, a friendly arms race played out on preprint servers, where research groups uploaded their findings as quickly as possible.

Credit: DeepMind

AlphaFold predicted this structure for mediator of RNA polymerase II transcription subunit 23.

These recent advances should provide scientists with more tools to study the molecular details of biological systems. “Most biological functions at the molecular level rely on recognition and specificity between molecules,” says Janet Thornton, director emeritus of the European Bioinformatics Institute. Researchers need detailed knowledge of the various biological structures involved to understand these interactions—for example, when designing new therapeutics. “DNA sequencing for almost any organism is now routine, and being able to convert a protein sequence to a structure opens many new possibilities,” she says.

Alphabet, the parent company of DeepMind, obviously sees the possibilities too, having recently launched a drug discovery company called Isomorphic Labs that plans to build on AlphaFold’s success. But Moult believes that there are other avenues of research that AI could aid. These include predicting how molecules bind proteins, designing and repurposing drugs, and interpreting spectroscopy data. Looking further ahead, he expects AI and machine learning “will open up biology we don’t even understand.” Right now, he says, everything has changed but the way we think.

Credit: Tristan Croll/Coronavirus Structural Task Force

The Coronavirus Structural Task Force found SARS-CoV-2 protein models with errors (left) and corrected them (right).

The scientific response to the COVID-19 pandemic was swift and forceful. Researchers from every discipline around the globe turned their attention to fighting the virus and saving lives, and the results are a credit to science. Safe and effective vaccines were available in less than a year, and new treatments continue to emerge. But there have been missteps and false starts as well. Mistakes and dead ends are nothing new to scientists, although the urgency of the pandemic and the public’s hunger for hopeful news may have contributed to the frequency of missteps and their visibility in mass media.

Experts on science policy and science communication don’t fault scientists for mistakes in the lab; those are normal at any time and especially in the early stages of research. “We knew science would go completely wrong in the process,” says Dietram Scheufele, a professor of science communication at the University of Wisconsin–Madison. “But we basically didn’t acknowledge that at all.” Scheufele and other experts say that researchers and public health officials could have done better when talking to people about ongoing research on COVID-19. Instead, researchers and government officials presented new findings and recommendations—like whether to wear a mask or who should get a vaccine booster—with more certainty than science could really justify. And although people continue to trust science and scientists in general, some groups have been skeptical of research findings and public health recommendations, even embracing treatments with no demonstrated benefit.

Science communication experts say that one possible solution is being more transparent about the mechanisms that make the scientific process trustworthy, such as independent peer review and journal article retractions, and being forthright about when the science might be wrong.

This year provided several examples of scientists encountering, and sometimes correcting, COVID-19 research that wasn’t quite up to snuff or didn’t have the expected impact. The year also provided scientists with ways to continue advancing their research fields.

Structural issues

Credit: Lorenzo Casalino, Amaro Lab (UC San Diego)

Computational scientists have made detailed models of the SARS-CoV-2 spike protein since the early months of the COVID-19 pandemic.

One of science’s first toeholds in the long climb out of this pandemic was the genomic sequence of SARS-CoV-2, the virus that causes COVID-19. It was deciphered in early January 2020 (Nature 2020, DOI: 10.1038/s41586-020-2008-3). Shortly thereafter, researchers began solving the structures of the virus’s proteins, which enable the pathogen to enter our cells and replicate. Those structures, and the detailed computer models of the virus and its proteins that followed, helped scientists and doctors develop vaccines and therapeutics.

But protein structures are only as helpful as they are accurate. That’s why members of the Coronavirus Structural Task Force spent time this year checking new SARS-CoV-2 protein structures uploaded to the Protein Data Bank (PDB), one of the world’s largest open structure databases. They looked for out-of-place amino acid side chains, unfavorable steric interactions, and other errors. Inaccurate structures can impede drug designers’ ability to predict molecules that might target those proteins. When the task force found a mistake, it shared a corrected structure on its website and contacted the authors of the original structure, who can—and in most cases did—upload the correct protein structure to the PDB.

Because automated structural biology tools and protein sequence data are becoming more available, researchers with little experience or specialist knowledge can produce and share new protein structures in just weeks, before other researchers can check the models’ accuracy. This problem was exacerbated last year in the rush to address the COVID-19 pandemic. To solve this issue, the task force has called on the structural biology community to change the way it shares data and documents and addresses structural errors (Nat. Struct. Mol. Biol. 2021, DOI: 10.1038/s41594-021-00593-7).

Repurposed drugs faded

Drug developers leaped into the fight against COVID-19 from the pandemic’s earliest days. Some of the therapeutics they researched, like monoclonal antibody treatments and antivirals, including remdesivir, proved helpful. But many other molecules failed to deliver on their initial promise.

Some repurposed drugs, such as hydroxychloroquine, showed early promise against SARS-CoV-2, but most proved to be ineffective treatments.

The most prominent example is hydroxychloroquine. Remembered now as the drug that former US president Donald J. Trump promoted as a cure for COVID-19, hydroxychloroquine was one of several already-approved drugs that showed promising antiviral activity in very early cell studies. Finding new uses for treatments already on the market is known as drug repurposing, and it can be a faster route to new therapies than starting from scratch. But further investigation of hydroxychloroquine and almost two dozen other repurposed drugs found that these compounds were harming infected cells but not stopping the virus (Science 2021, DOI: 10.1126/science.abi4708).

Computational chemists also tried finding COVID-19 therapeutics, without much success. One review of the efforts found that hundreds of published computational studies lacked rigor, failing in particular to back up computational results with lab studies (Chem. Soc. Rev. 2021, DOI: 10.1039/d0cs01065k). At one point, the preprint server bioRxiv stopped accepting computational papers on possible COVID-19 therapeutics if they lacked experimental verification. And although the artificial intelligence company Insilico Medicine announced in February 2020 that it had predicted two potential COVID-19 therapeutics, neither has panned out so far.

Water wasted?

Credit: Shutterstock

Public health officials can test wastewater to spot COVID-19 outbreaks, but the technique has limits.

Besides treating COVID-19, scientists also worked on ways to better identify infected people to try to stop the virus’s spread. Wastewater monitoring emerged as a relatively cheap way to catch COVID-19 outbreaks early, since viral particles appear in feces—and thus in the water we flush into municipal sewer systems—before infected people show symptoms. By the second half of 2021, researchers were beginning to see the limitations of that approach. Scientists can calculate case numbers from wastewater data, but not very precisely. And this kind of testing can’t say which individuals in a dormitory or office building—let alone a town or city—are infected. Researchers have also found it difficult to identify new variants of the virus in wastewater because they are rarely able to sequence entire viral genomes. Nonetheless, wastewater-monitoring scientists say that the lessons learned and equipment installed during the COVID-19 pandemic will make the technique a useful public health tool for the future.

Borophene’s hydrogenated cousin, borophane

Credit: Mark Hersam/Northwestern University

Borophane, the hydrogenated form of borophene, resists oxidation. B = teal; H = red.

Borophene, an atomically thin 2D form of boron, is chemically unstable. By adding dihydrogen to borophene in vacuum, researchers at Northwestern University converted borophene to the stabler borophane (Science 2021, DOI: 10.1126/science.abg1874). The hydrogenated form is stable in air for about a week, and the hydrogenation is reversible, so borophane could serve as a way to protect the unstable borophene.

Einsteinium forms a coordination complex

After 6 years of planning, researchers finally made a coordination complex with einsteinium, the heaviest element stable enough for conventional chemistry experiments. To make the eight-coordinate complex, the researchers treated ²⁵⁴Es with a hydroxypyridinone-based ligand (Nature 2021, DOI: 10.1038/s41586-020-03179-3). The complex will help chemists better understand trends across the actinide series of elements.

Molecule goes to infinity and beyond

Researchers at Nagoya University fused 12 benzene rings to make a molecule that loops like an infinity symbol (ChemRxiv 2021, DOI: 10.33774/chemrxiv-2021-pcwcc). In X-ray crystal structures, only 3.192 Å separated the upper and lower benzene rings where the molecule, called infinitene, crosses over itself. Through computational studies, the chemists confirmed that infinitene’s π electrons are delocalized within individual benzene rings, not across the entire molecule. The study, which appeared as a preprint, has not yet been peer-reviewed.

Sugar molecules can decorate RNA

Credit: Cell

RNA molecules decorated with sugars (colored shapes) stick out of cell surfaces. These RNA-glycan conjugates bind receptors involved in immune signaling.

Scientists have long known that sugar molecules decorate biomolecules such as proteins and lipids. This year researchers reported that sugars can also be found on RNA molecules (Cell 2021, DOI: 10.1016/j.cell.2021.04.023). The decorated RNA molecules stick out from the cell membrane, where they might act as signals for the immune system.

A crystalline thorium complex

Bonds between actinide elements are usually short lived. Researchers at the University of Manchester made a complex containing thorium-thorium bonds, and it was stable enough that they could determine its structure using X-ray crystallography. They combined a thorium precursor with a tetrasilyl cyclobutadiene dianion. The resulting complex contained a cluster of three thorium atoms with cyclooctatetraenyl ligands and bridging chlorines (Nature 2021, DOI: 10.1038/s41586-021-03888-3).

Catching chiral nitrogen compounds

Synthesizing chiral nitrogen compounds wasn’t very straightforward until chemists reported a single-pot method this year (Nature 2021, DOI: 10.1038/s41586-021-03735-5). They treated an amine with an alkyl bromide that replaces the electron lone pair on the amine. Combining the resulting mixture of enantiomers with 1,1′-bi-2-naphthol (BINOL) locked the molecules into one conformation. Subsequent removal of the BINOL leaves the desired enantiomer of the chiral amine.

POLL: OUR READERS HAVE VOTED INFINITENE AS THEIR FAVORITE MOLECULE OF 2021.

• 3.055 billion

  The number of base pairs in the complete human genome. Researchers filled in the last remaining holes in the human genome using new sequencing technologies. The gaps were largely highly repeating sections associated with regulatory and other functions, not protein-encoding regions (bioRxiv 2021, DOI: 10.1101/2021.05.26.445798).

  
  Credit: Michael Abbey/Science Source
  Human chromosomes

• 2.4 Å

  The distance between atoms at which a hydrogen bond transitions to a covalent bond, according to simulations of hydrogen fluoride reacting with a fluoride ion. The research predicted a blurrier line between hydrogen and covalent bonds than was previously thought (Science 2021, DOI: 10.1126/science.abe1951).

• 80-fold

  The relative improvement in energy storage by weight for mass-produced spools of lithium-ion fiber batteries compared with previous ones (Nature 2021, DOI: 10.1038/s41586-021-03772-0). These fiber batteries are designed to be integrated into textiles.

• 1/2

  Proportion of childhood asthma cases attributable to air pollution in the predominantly Hispanic and Black neighborhoods of West Oakland, California. By comparison, air pollution is linked to one in five asthma cases in the Bay Area overall (Environ. Health Perspect. 2021, DOI: 10.1289/EHP7679).

• 
  Credit: Shutterstock
  Air pollution is linked to childhood asthma.

• 78 nm

  The resolution of a microscopy approach that combines stimulated Raman scattering with expansion microscopy to produce superresolution imaging of biological samples without fluorescent labels. The researchers used the approach to image cells undergoing division, a mouse hippocampus, and zebrafish embryonic retinas (Nat. Commun. 2021, DOI: 10.1038/s41467-021-23951-x).

• 
  Credit: Theor. Chem. Acc.
  OgTs₄

• 586

  The number of electrons in oganesson tetratennesside, the stable pentatomic molecule that would be made if the periodic table’s heaviest two elements reacted with each other. Researchers used relativistic quantum calculations to predict that this tetrahedral molecule would be stable (Theor. Chem. Acc. 2021, DOI: 10.1007/s00214-021-02777-2).

• 15–30 cm

  The sea level rise the world will experience by midcentury regardless of cuts in greenhouse gas emissions, according to the Sixth Assessment Report from the Intergovernmental Panel on Climate Change.

  
  Credit: Patterson Clark/Politico Pro DataPoint
  Map of global temperature change

• 0.5 Gg

  Predicted net annual emissions of trichlorofluoromethane (CFC-11) from the oceans in the 2070s. The compound’s production was banned in 2010, but oceans have absorbed large amounts of CFC-11 over the years. Scientists now estimate that the oceans are a significant and growing source of the compound in the atmosphere (Proc. Natl. Acad. Sci. U.S.A. 2021, DOI: 10.1073/pnas.2021528118).

• 
  Credit: NASA Ozone Watch
  Ozone hole over Antarctica

• 800–1,300 V

  The voltage used to accelerate a plasma of I⁺ and I²⁺ ions in an iodine-based ion drive that was tested as a thruster in spacecraft. Other ion drives in space probes have used costly xenon as an ion source, but iodine is cheap and abundant. Researchers demonstrated the iodine-powered thruster on an orbiting satellite and found it performed as well as a xenon-based thruster (Nature 2021, DOI: 10.1038/s41586-021-04015-y).

  
  Credit: ThrustMe
  Iodine-based ion drive
• 
  Credit: Jason McLellan/University of Texas at Austin
  SARS-CoV-2 spike protein

• <10 kg

  The mass of all the SARS-CoV-2 in humans as of June 2021. Researchers used the typical viral load in various tissues of those with COVID-19 to estimate that the total mass of the virus at that time was between 100 g and 10 kg (Proc. Natl. Acad. Sci. U.S.A. 2021, DOI: 10.1073/pnas.2024815118).

Snakes spitting pain

Most of the venomous snakes in the Liverpool School of Tropical Medicine’s collection immobilize and kill prey by injecting venom through their fangs. But some snakes protect themselves from predators by spitting venom—potentially causing pain, inflammation, and blindness. This year, a team of researchers at the school’s Centre for Snakebite Research and Interventions uncovered how the venom packs such a punch: the molecular cocktail in spitting snakes’ venom contains high levels of proteins that increase the sensation of pain (Science 2021, DOI: 10.1126/science.abb9303). The team found that, while each snake’s venom had a unique concoction of toxins, all the spitting snakes had higher levels of phospholipase A₂ (PLA₂) enzymes in their venom than nonspitting snakes. In tests on mouse nerve cells, mixtures of venom components including PLA₂ triggered more cell activity than cocktails without it, a sign of a greater pain response.

Tasting the shape of proteins

Credit: Bryan Shaw

Using food-safe models, students can feel the shapes of proteins with their tongues, which are more sensitive than fingertips. A Skittles candy (right) is shown for scale.

To teach blind students or those with low vision about molecular structures, educators have traditionally used aids like physical models or tactile diagrams consisting of raised images on paper. For large molecules like proteins, though, these tools can be impractical or expensive. This year, Bryan Shaw and colleagues at Baylor University proposed a new kind of aid: bite-size and detailed models of complex structures that students can feel with their lips and tongue (Sci. Adv. 2021, DOI: 10.1126/sciadv.abh0691). Since the mouth is more sensitive to touch than fingers, Shaw’s group wanted to see if the oral sense of touch could help students perceive structures. Using biocompatible resin, the researchers 3D printed protein models 2–20 mm in diameter. The researchers also cast edible gelatin versions. They then tested these oral teaching aids alongside visual and manual ones with college and elementary students who were not visually impaired. The students identified and remembered proteins about equally well across the three types. The team hopes to collaborate with other researchers to further develop this idea into a workable and widely used learning tool.

Bendable ice

Credit: Science

Researchers have grown microfiber crystals of ice that are surprisingly elastic.

Ice has a reputation of being rigid, brittle, and prone to shatter. So researchers led by Xin Guo and Limin Tong at Zhejiang University were surprised to find that, when grown as microscopic, fiber-like crystals, ice is bendy and elastic (Science 2021, DOI: 10.1126/science.abh3754). Naturally formed ice typically contains air bubbles and cracks. For this study, the researchers wanted ice without imperfections. Using an electric field in a freezing chamber, the team grew nearly perfect hexagonal crystals ranging from 10 μm to less than 0.8 μm in diameter. While most hard materials can’t bend when made very thin, like notoriously fragile glass fiber-optic cables, these microfiber ice crystals can bend into a radius of curvature that almost reaches the theoretical limit. No applications currently exist for flexible ice, but the finding sheds new light on how a common material defies expectations when made extremely small and under perfect conditions.

Glass catalyst

Credit: Yangjie Li

Glass can act as a strong base to catalyze reactions and degrade various reagents.

This year, a team led by Yangjie Li of Purdue University found that glass beads can facilitate many base-catalyzed reactions, suggesting that glass could be used as a green catalyst in place of toxic or expensive chemicals (Chem. Sci. 2021, DOI: 10.1039/d1sc02708e). The team added glass microspheres to thousands of reaction conditions to see how the glass would react. Glass’s surface is covered with dissociable silanol groups, giving it a negative charge when it is in contact with solution. With this negative charge, glass can act like a strong base, producing nucleophilic solvent anions that can catalyze reactions. The team also found that glass can degrade biomolecules such as lipids, meaning glass vessels used for bioanalytical experiments might disrupt those studies.

Heavy-metal ant teeth

Credit: Shutterstock

Ant teeth are sharp and durable thanks to distributed zinc atoms.

Thinner than a human hair but strong enough to slice through tough leaves and flesh, the average ant tooth is much more than it seems. This year, a team of researchers led by Robert Schofield of the University of Oregon found that ants’ teeth get their strength from a uniformly distributed layer of zinc atoms along the edges. This layer of atoms allows ants to use at least 60% less force than if their teeth had a similar composition to that of humans’ teeth (Sci. Rep. 2021, DOI: 10.1038/s41598-021-91795-y). Ant teeth are so small that electron microscopy can’t analyze their makeup. Using atom probe tomography, though, researchers were able to image the tips of ant teeth. Although human teeth and crab claws are hard, they contain nodules of interspersed minerals, making them vulnerable to crack. Ant teeth, however, are made of a protein that binds zinc atoms along the edges. This makes the edges razor sharp and able to evenly distribute force. The researchers are now looking into the composition of scorpion stingers and spider fangs to see what molecular adaptations they have.

Restoring touch with static charge

Credit: ACS Nano

A triboelectric nanogenerator like this restored sense of touch in rats.

We feel the world around us with the help of nerves in our skin. We lack a reliable way to restore the sense of touch when those nerves are damaged. This year, a team of researchers led by Ben M. Maoz of Tel Aviv University engineered tiny devices known as triboelectric nanogenerators (TENGs) that can be implanted under the skin of animals to power up injured nerves (ACS Nano 2021, DOI: 10.1021/acsnano.0c10141). The nanogenerators harness static electricity to deliver their power. The team created a TENG about 25 mm² by layering cellulose and polydimethylsiloxane. Compressing the TENG generated friction between the layers of those two materials, driving an exchange of electrons and creating a small voltage of up to 2 V. A wire connected to thin layers of gold that sandwiched the two materials conveyed the resulting current to an electrode that wound around a nerve, stimulating it. When tested on animals with damaged nerves, the devices restored sensation.

Correction:
This story was updated on Dec. 22, 2021, to correct the byline. It was written by Emily Harwitz, not Ryan Cross.