2022

The world’s major food crops—rice, wheat, corn—must be planted anew for every harvest. That’s a lot of work for farmers and can contribute to environmental problems such as soil erosion. Perennial grains that survive and produce year after year could ease the burden, but breeding plants that are long-lived and productive enough has been a challenge. This year, researchers in China showed perennial rice can meet those benchmarks and save farmers many weeks of backbreaking labor.

Called Perennial Rice 23 (PR23), the variety was created years ago by crossing a commercial variety of Asian rice with a perennial wild rice that grows in Africa. Improving its yield and quality took more than 2 decades. Finally, in 2018, researchers at Yunnan University and other institutions released PR23 to farmers in China, enlisting them in a large-scale experiment to find out how many times the rice can be harvested and measure the yield and other benefits.

PR23 yielded just as much grain as regular, seasonally planted rice, the team reported last month in Nature Sustainability. In the first year, planting and cultivation cost about the same. But in the second year, farmers could eliminate a major task: transplanting young rice seedlings into a paddy, grueling work often done by women and children. Skipping this step, thanks to the perennial rice, reduced the amount of work per hectare by as much as 77 person-days each season, and helped lower farmers’ costs by half. Soil nutrients also increased in the fields containing perennial rice. By the fifth year, however, yields dropped so much the perennial rice needed to be replanted.

More and more farmers are cultivating PR23, thanks to technical assistance from Yunnan University and government promotion. More than 15,000 hectares were planted in southern China last year, a fourfold increase from 2020. PR23 and similar varieties are being tested in Africa as well. Perennial rice could also reduce soil erosion in the terraced uplands of Southeast Asia. But plant breeders still need to develop a strain adapted to that environment’s drier and less fertile soil. Researchers also worry about long-term impacts. One concern is that weeds and pathogens will accumulate in the unplowed fields, requiring more herbicide than conventional rice does. Another question is whether the rice emits more nitrous oxide—a potent greenhouse gas. But as cultivation spreads, the costs and benefits of perennial rice should come into focus.

Artificial intelligence (AI) is making inroads in areas once considered uniquely human, including artistic expression and scientific discovery. The machines’ encroachment was slow at first, but this year it turned into a landgrab.

The most visually stunning evidence—inescapable on social media—came from so-called text-to-image models. They use machine learning to analyze pairings of text and images online, finding patterns that allow them to create new images based on new text. Last year, the research lab OpenAI presented a software system called DALL-E that when asked for “an armchair in the shape of an avocado” could spit out several charming examples. This spring, OpenAI released a large upgrade, DALL-E 2. It implemented a machine-learning technique called diffusion, in which images emerge from “noise,” guided by context or text descriptions. The method can efficiently generate realistic and alluring pictures. Several diffusion models became available for public use this year, and an artist using one won a fine art competition, stirring both curiosity and acrimony. At the same time, Meta, Google, and others released diffusion models that can conjure videos.

Machine learning is also showing off its creativity in science, math, and programming. Science’s 2021 Breakthrough of the Year honored AI tools that predict the 3D structure of proteins from the sequence of their amino acid building blocks. Expanding on that work, researchers have now used AI to design entirely novel proteins that could be used in vaccines, building materials, or nanomachines. One technique, called “hallucination,” starts with random sequences and mutates them toward sequences that other AI tools are confident will fold up into stable proteins.

Meanwhile, DeepMind announced a tool called AlphaTensor, an algorithm that designs more efficient algorithms for multiplying blocks of numbers called matrices—an operation useful for computer graphics, physics simulations, and machine learning itself. It found shortcuts that human mathematicians had overlooked for decades. The company also presented AlphaCode, a system that writes programs to solve numerical problems (such as calculating how many binary strings of a given length don’t have consecutive zeroes). It uses a model trained on previous programs and their descriptions to produce many candidate programs, then picks the best prospects. Pitted against human programmers, AlphaCode places in the middle of the pack.

Aside from philosophical debates about whether these feats of silicon count as real creativity, they raise practical and ethical dilemmas. Some observers worry the artificial coders and artists will violate copyright, perpetuate stereotypes, spread misinformation, or eliminate jobs. But there’s no doubt humans will harness these tools to extend our own creativity, much as we did in the past with looms, cameras, and other once-unsettling inventions.

The discovery of a giant bacterium with complex innards shook biology this year. Microbes are supposed to be microscopic, but this one, tentatively dubbed Thiomargarita magnifica, can be 5000 times bigger than many bacterial cells—as long as a pushpin. The single, threadlike cells were first spotted on the surfaces of dying leaves in a mangrove swamp in the French Antilles.

Bacteria need to be tiny, researchers thought, because they lack the internal transport systems found in other cells and depend on diffusion to move nutrients and wastes. Diffusing molecules can’t travel very far, limiting how big a bacterium can be—or so the thinking went. Textbooks also say bacteria typically lack internal compartments, but T. magnifica has several, researchers reported in February.

One is a water-filled sac that may have allowed the microbe to become a macrobe. It pushes all the cell’s proteins and other components against the outer cell envelope, putting them in range of oxygen, sulfur, and other essential molecules diffusing in and out of the cell. By adding customized amino acid building blocks to the bacteria and tracing their incorporation into proteins, the researchers demonstrated that protein production takes place near the periphery of the cell. A few other bacteria, including one no bigger than a poppy seed found off the Namibian coast in 1999, have similar structures.

Other features appear to be unique to T. magnifica. The DNA of virtually all other known bacteria floats freely in their cells, but T. magnifica packages its huge 12-million-base genome into membranous sacs the researchers call pepins, along with the molecular machinery for making proteins. And whereas most bacteria produce the energy molecule ATP in their cell envelope, T. magnifica has a whole network of internal membranes that also make ATP, enabling it to produce enough fuel for such a large cell.

These structures shake up the traditional division of life into eukaryotes and prokaryotes. Eukaryotes include plants, animals, and other organisms with complex cells that segregate their components into membrane-lined compartments called organelles. Prokaryotes include bacteria and other single-celled organisms that lack organelles and have sometimes been characterized as simply “bags of proteins.” T. magnifica seems to represent something in between—perhaps mirroring transitional forms that evolved billions of years ago.

Large scale clinical trials of two vaccines against respiratory syncytial virus (RSV) have finally proved they can safely protect the two groups hardest hit by this common infection: infants and the elderly. Both vaccines prevented severe disease in people
over age 60 without causing serious side effects. One also protected infants for 6 months when given to their mothers late in pregnancy, so they could pass the antibodies to their fetuses.

RSV usually only causes mild, coldlike symptoms, but in babies the virus can inflame small airways in the lungs, and in the elderly, it can worsen existing lung and heart conditions. RSV vaccine development was derailed for decades after a clinical trial of an experimental candidate more than 50 years ago killed two children and hospitalized 80% of those who received it. Scientists subsequently figured out the key reason: Made from a chemically inactivated version of the entire virus, the vaccine only elicited relatively weak antibodies, which not only failed to stop the virus, but, through little-understood mechanisms, helped RSV damage airways.

The new vaccines avoid this problem by relying on a key advance made by Barney Graham and co-workers at the National Institute of Allergy and Infectious Diseases in 2013. A viral surface protein used in the vaccines changes its shape after it docks onto a cellular receptor and the virus fuses with the cell, establishing an infection. Led by Graham, who is now at the Morehouse School of Medicine, the team figured out how to lock the protein into its prefusion state. As a result, vaccination triggers far higher levels of potent antibodies.

The good news from this year’s trials, run by GSK and Pfizer, vindicated that strategy. More results will come soon: Janssen Pharmaceuticals and Bavarian Nordic have efficacy trials underway of their own RSV vaccines for older adults. Both vaccines performed well in the earlier phases of development.

Developers remain skittish because of past disappointments: GSK in February stopped its maternal RSV vaccine after unspecified “safety signals” surfaced in clinical trials. But none of the other studies has reported red flags to date, and several of the candidate vaccines could receive approval from regulators around the world next year.

Drawing on a vast trove of military medical records, researchers this year showed a common herpes virus is an essential player in multiple sclerosis (MS), a disease in which the immune system attacks neurons. The findings may lead to new ways to treat or prevent the mysterious disorder, which causes mild symptoms—including blurred vision, fatigue, and numbness—in some of its 2.8 million sufferers around the world, but gradually leaves others unable to speak or walk.

A leading suspect in MS has long been Epstein-Barr virus, which infects most people in childhood, then lies dormant in certain white blood cells. Transmitted mainly through saliva, the virus can lead to infectious mononucleosis, or “kissing disease,” in newly infected teens and young adults. Nearly all people with MS have antibodies to Epstein-Barr virus, but so do 95% of healthy adults, making it difficult to nail down the virus as a cause.

To firm up the link, epidemiologists trawled 20 years of medical records for more than 10 million U.S. military recruits and analyzed some of their stored blood samples. Of 801 soldiers who developed MS, all but one had previously tested positive for Epstein-Barr virus. And among soldiers who were initially negative, a subsequent infection raised MS risk 32-fold, the team reported in January in Science. That exceeds the increase in lung cancer risk caused by smoking.

Other researchers identified a possible mechanism, reporting in Nature just days later that the hibernating virus may awaken and cause nerve damage through so-called molecular mimicry. One of Epstein-Barr’s proteins resembles a protein made in the brain and spinal cord, which apparently tricks the immune system into attacking the sheathing around nerve cells that’s essential for conducting electrical signals. About 20% to 25% of MS patients sampled had antibodies in their blood that bind both proteins.

These discoveries are spurring efforts to develop drugs to treat MS by targeting the virus. And if one of the Epstein-Barr vaccines now in clinical trials proves effective and is given to children worldwide, someday MS could even go the way of polio and be virtually wiped out.

For decades, U.S. scientists have led the world in documenting the risks of climate change, and U.S. diplomats have cast global warming as a dire threat in international fora. Those warnings rang hollow, though, because unlike many wealthy countries, the world’s second largest producer of greenhouse gases had never passed a law to substantially reduce those emissions. This summer, attempts to pass such a bill appeared doomed to fail yet again.

Then, in a legislative instant, it all changed when a key senator suddenly dropped his opposition. The climate provisions of the so-called Inflation Reduction Act (IRA) amount to the biggest step the United States has ever taken to slow global warming. The legislation provides $369 billion over 10 years to support electricity from renewable sources and nuclear power, while also spurring a wholesale move to electric vehicles and research into ways to reduce industrial emissions. Several independent research groups have calculated it should put the United States on track to cut its greenhouse gas emissions by 40% from 2005 levels by decade’s end.

Yet the IRA alone is not enough for the United States to meet its commitment under the 2016 Paris agreement to cut emissions by 50% by 2030. For that to happen, analysts say, individual states will have to increase their clean energy generation. The Environmental Protection Agency will also have to issue, and enforce, long-expected greenhouse gas regulations for electric utilities—and future presidents and the courts will need to sustain them. Some climate activists have also criticized the IRA for its incentives for capturing carbon from smokestacks and its provisions to allow continued oil and gas drilling in the Gulf of Mexico. Those measures won the support needed to pass the bill, but critics see them as perpetuating a fossil fuel industry in no need of lifelines.

Meanwhile, a world that has already warmed 1.2°C since preindustrial times has little time left before the global average surpasses 1.5°C, the threshold of “dangerous climate warming” set in international negotiations. Emissions are expected to rise again this year globally, rather than falling as needed, and many climate scientists believe the world is certain to overshoot 1.5°C. Although the IRA is a step in the right direction, activists say, humanity needs much more action—and soon.

Ever since the Black Death killed one-third to one-half of the people living in Europe 700 years ago, researchers have wondered how the deadly plague left its mark on survivors. Such a devastating pandemic must have acted as a potent selective force, favoring people with particularly effective immune defenses. But detecting its legacy has been impossible in living people because our immune genes change frequently in response to new pathogens.

This year, researchers harnessed tools for studying ancient DNA to look at differences in immune genes in the very people who lived and died during the plague—and identified a dramatic effect. The team analyzed ancient DNA from the bones of more than 500 people buried before, during, and after the Black Death in London and Denmark. In October, they reported in Nature that survivors were much more likely to carry gene variants that boosted their immune response to Yersinia pestis, the flea-borne bacterium that causes the plague.

An astonishing 245 gene variants rose or fell in frequency after the Black Death in London; four of them also changed in ancient DNA from people in Denmark. One gene in particular stood out: ERAP2. It encodes a protein called endoplasmic reticulum aminopeptidase 2, which has been shown to help immune cells recognize and fight threatening viruses.

The team found two variants of ERAP2 that differ by just one letter in the genetic code. One produces a full-size protein, the other a truncated version. People who inherited two copies of the variant encoding the full-size protein were twice as likely to have survived the plague as those who inherited two copies of the other variant. The researchers also cultured immune cells from 25 modern-day British people in the lab, and found that cells with the full-size, protective version of ERAP2 produced more immune system proteins called cytokines when exposed to Y. pestis

The fast spread of this protective gene variant in Europe during the century after the Black Death is the strongest example yet of natural selection on the human genome. The protective variant of ERAP2 is still found in 45% of British people today. Its persistence suggests it continued to be favored by natural selection until recently—probably because the plague was endemic in Europe and Asia until the early 19th century. But this protection may have a price: The same variant also confers a higher risk of developing autoimmune diseases, such as Crohn’s disease and rheumatoid arthritis.

For thousands—if not millions—of years, a little moon named Dimorphos made laps around a larger asteroid, millions of kilometers from Earth. On 26 September, NASA smacked into it with a spacecraft, forever altering its orbit—and demonstrating a strategy that might one day save humanity.

When the fridge-size Double Asteroid Redirection Test (DART) satellite barreled into the 160-meter-wide Dimorphos at 6 kilometers per second, scientists celebrated the first-ever mock trial of a planetary defense mission. NASA’s goal was to knock Dimorphos slightly closer to its partner, shortening its orbital period and demonstrating a strategy for thwarting real threats, should future Earth-bound asteroids be detected.

For years leading up to the big event, scientists ran computer simulations and blasted small-scale replicas of asteroids with projectiles to forecast how much momentum would be transferred. Predictions varied widely, depending in part on whether the target was a monolithic rock or a gravity-bound heap of rubble.

In the minutes leading up to the final collision, DART’s onboard cameras streamed images of Dimorphos (above), which the ever-closer view revealed to be an egg-shaped rubble pile. Two weeks after the impact, scientists compiled observations to confirm that the moon’s nearly 12-hour orbit had shortened by 32 minutes—a change more than 26 times larger than NASA had set as its goal. The collision was a one-off, but it gives scientists a crucial data point for the momentum models they would use to design any future asteroid-deflection missions.

So far, however, astronomers have only detected about 40% of the estimated 25,000 near-Earth asteroids large enough to decimate a large city and common enough to pose a threat. The Near-Earth Object Surveyor, a long-anticipated space-based infrared telescope, would help locate more of these bodies, but it has faced repeated funding cuts and delays from NASA. DART’s bull’s-eye has shown what’s possible, but a capable planetary defense system will also require more intelligence about the threat.

Until recently, DNA’s shelf life was pegged at about 1 million years. Genetic material much older than that presumedly would be too badly degraded to read. This year, scientists wound back the clock further than they once thought possible, extracting tiny DNA snippets at least 2 million years old from frozen soil in an Arctic desert.

The study, hailed as a tour de force, demonstrates the power of environmental DNA, or eDNA, to reconstruct lost worlds: in this case, a coastal forest unlike any in existence today that flourished during a warm climate episode at the tip of northern Greenland. DNA fragments in 41 organic-rich samples from a thick layer of sediment heaped at the mouth of a fjord revealed a lush forest of poplars, thujas, and other conifers; black geese and horseshoe crabs; and mammals such as reindeer, lemmings—and mastodons. No one had expected the range of this extinct relative of elephants to extend that far north.

What preserved the DNA over the ages was not only the natural icebox of permafrost, but also grains of quartz and clay, especially smectite, whose charged surfaces bound and protected the DNA. The team spent years honing techniques for prizing DNA snippets from the minerals and then decoding them with high-throughput sequencing.

The findings raise the prospect of extracting eDNA at other high Arctic sites, where fossils are scarce. But the further back in time paleogeneticists delve, the harder it will be to identify some species—especially those in dead-end lineages whose genomes likely bear scant resemblance to contemporary species.

Profiling eDNA from ancient sediments could pay off by revealing genetic adaptations that allowed plants and animals to thrive in the far north at temperatures warmer than today’s. More controversially, novel gene sequences might even be plucked out of ancient genomes and, using CRISPR gene editing, stitched into present-day life forms—to help crops germinate earlier, for example. The prospect of resurrecting ancient genes is bound to make many scientists uncomfortable—but proponents argue the looming climate crisis demands drastic interventions.

At first, zero COVID was a success. But with time, China’s strict lockdown policy strained its economy, frustrated its citizens, and arguably did more harm than good to public health. This month the government belatedly started relaxing restrictions without formally ending the zero-COVID strategy.

China’s lockdown of Wuhan, the pandemic’s epicenter, contained the virus for 76 days in early 2020 until the city’s outbreak burned out. Life in Wuhan returned to nearly normal. New Zealand, Australia, Singapore, and Taiwan all adopted zero-COVID policies modified to suit their own legal and cultural norms and used them to buy time until vaccines arrived. Those countries then relaxed controls and transitioned to living with the virus.

China, however, despite having its own (less effective) vaccines, made zero COVID a goal in itself. To sniff out infections that are mostly asymptomatic, masses of citizens got tested every other day. They faced quarantine if positive and their apartment blocks were locked down. As the highly transmissible Omicron variant drove infection numbers to levels not seen since the Wuhan outbreak, the lockdowns grew more frequent and economically damaging. China’s gross domestic product growth, a robust 8% in 2021, is projected to slump to 3% this year.

China’s people have lost patience. On 14 November, residents of Guangzhou defied a lockdown and poured into the streets, toppling the barriers intended to keep them home. After 10 people died in a high-rise apartment building fire in Urumqi on 24 November, deaths many blamed on an ongoing lockdown, pent-up frustration erupted in cities throughout China. Protesters demanded the end of zero COVID and the end of mass testing; some even called on President Xi Jinping to step down. Now, authorities are hurriedly rolling back restrictions, despite an ongoing Omicron surge.

Ending zero COVID carries risks of its own, as China is still ill-prepared to live with the virus. Just 66% of those over age 80 are fully vaccinated and only 40% have gotten boosters, leaving them vulnerable to the expected wave of infections. China missed its chance to plan and execute a more orderly transition from zero COVID.

Escalating tensions severely bent—but didn’t always break—superpower science collaborations this year.

Within weeks of Russia’s invasion of Ukraine on 24 February, most major research funders in Europe announced they were severing relations with the Russian government, although some existing projects would be allowed to finish. The European Space Agency (ESA) suspended work on ExoMars, a nearly finished Mars rover mission that depended on Russian help. (ESA now has a plan to fly the rover without Russia, but much delayed.) And CERN, Europe’s particle physics laboratory, said it would no longer welcome scientists from Russia and its ally Belarus after pacts with those nations expire in 2024. In June, the United States followed suit, saying it would “wind down” most science projects with Russia, including Arctic research.

Collaborations between China and Western nations also frayed. In the United States, concerns that China is stealing the fruits of federally funded research led Congress to place new limits on the ability of government-supported scientists to work with Chinese institutions. In Europe, similar fears prompted some universities to back away from projects with Chinese partners. In August, China suspended work with the United States on an array of issues, including climate research, to protest a visit to Taiwan by a high-ranking member of Congress.

Yet superpower leaders still seem open to at least some scientific cooperation. Despite the Ukraine war, Russia continues to contribute to the ITER fusion reactor under construction in France and launch crew and supplies to the International Space Station. And in November, after a lengthy meeting aimed at defusing tensions, U.S. and Chinese leaders announced they would resume their paused work on climate and other issues. Such cooperation, each side said, is in the best interests of the world as a whole.

The Ukraine war roiled global energy markets, driving up prices and scrambling plans for cutting greenhouse gas emissions.

In Europe, steep cuts of natural gas imports from Russia upended plans to use gas as a lower carbon substitute for coal while economies transition to renewable sources such as wind and solar. Now, Germany and Austria say they will delay closing some coal-fired power plants—and even reopen shuttered stations—in order to keep the lights on. That could increase their carbon emissions by up to 20% over the next 2 years, researchers say. Europe is also seeking to import more liquefied natural gas from the United States and the Middle East, which could mean more leaks of methane, a potent warming gas, from wells, pipelines, and storage tanks.

But the crisis could ultimately end up speeding the transition to cleaner energy. It has highlighted the “unsustainability of the current global energy system,” the International Energy Agency (IEA) said in October, and prompted many nations to “accelerate structural changes.” Those changes—including major investments in renewables by the United States, Europe, Japan, and South Korea—will increase global spending on clean energy by at least 50% over the next decade, analysts estimate. IEA forecasts the crisis-driven uptick in coal use will be temporary as alternatives take hold.

Climate advocates hope such rosy scenarios come true. But forecasting global energy markets, they note, has proved hazardous.