One rainy morning in May, a Romanian archivist named Tudor Arhire retrieved a brown envelope from a wooden filing cabinet, slid out a small, yellowed page and placed it carefully on the table. Arhire is the custodian of a government archive in Sibiu, Romania, a medieval city in the region of Transylvania. Inside the grand, historic building, elegant windows and parquet floors contrasted with yellowed lace curtains and battered upholstery; on a desk in the corner, a pile of books and parchments spanned hundreds of years. The document he produced was a letter, more than 500 years old. Despite the ancient creases and stains, its nine lines of flowing Latin script, translated long ago, were clearly legible. But nobody here was intending to read it. Instead, two visitors, a married couple named Gleb and Svetlana Zilberstein, waited eagerly with latex gloves and plastic tubes.

The letter is one of the archive’s most precious possessions. Dated August 4, 1475, it was written to the burghers of Sibiu by a man describing himself as “prince of the Transalpine regions.” He informed the townspeople that he would soon be taking up residence among them. He signed with a name sure to strike fear into their hearts: Vlad Dracula.

Dracula had previously ruled the neighboring region of Wallachia, and he was known for his cruelty, especially his practice of impaling enemies on stakes. Thus his nickname, Vlad the Impaler. Now he was preparing to gain the Wallachian throne once again. His letter to Sibiu’s residents is one of only a few sparse documents related to the notorious prince, who centuries later would inspire Bram Stoker’s fictional vampire, Count Dracula.

a group of four people examine an old manuscript
In Sibiu, Romania, archivists Tudor Arhire, left, and Marilena Mărginean, right, show a 17th-century manuscript to protein detectives Gleb and Svetlana Zilberstein. Andrei Pungovschi

Subscribe to Smithsonian magazine now for just $12

This article is a selection from the November/December issue of Smithsonian magazine

an orange ink portrait from a woodcut
Vlad Dracula, also known as Vlad the Impaler, from a 15th-century woodcut. Bridgeman Images

The Zilbersteins were interested not in the words on the page, however, but something else—physical remnants of the prince himself, including molecule fragments from his sweat, saliva and tears. Their work harnesses breathtaking advances in a field known as proteomics, which seeks to understand the interaction of proteins within living cells and organisms. Proteins have long been studied in the context of biology and medicine, but spectacularly sensitive analytical techniques now allow researchers to use protein traces to gather intimate information from materials that were once primarily the domain of historians and archaeologists, opening a new window onto the past. The project is part of a scientific revolution that is profoundly expanding the type of information that can be gleaned from historical texts and artifacts, from X-ray and CT scanning to carbon dating and genetic sequencing.

Already, DNA is used to identify individuals from biological remains and reveal large-scale relationships, from family trees to evolutionary timelines. But DNA remains constant throughout a person’s life, and it degrades badly over time. Which is why researchers are also interested in proteins, the molecules DNA encodes and that do most of the work inside our cells. If DNA keeps a static record of our ancestry, proteins, which metabolize our food, store and transport resources, and carry messages from one place to another, provide a running commentary on our health and habits. They leave evidence of our diets, our illnesses, the drugs we use, even our cause of death. And they are left behind on everything we touch.

Until recently, researchers hoping to detect traces of ancient proteins needed to destroy a small sample of the material in question in order to isolate enough molecules to get a “readable” signal. This isn’t usually a problem with biological remains such as bones or fossils, but few archivists are willing to damage an invaluable artifact like Dracula’s letter. But Gleb, an Israeli entrepreneur and inventor originally from Soviet Kazakhstan, has designed a material that can coax protein molecules from the surface of paper, parchment and paintings—even mummies and woolly mammoths—without damaging the objects themselves. Working with Pier Giorgio Righetti, an Italian chemist, he and Svetlana have used this method to explore a range of archives, arousing both excitement and consternation among historians, as the researchers report the unsuspected activities of iconic figures from Johannes Kepler to Joseph Stalin.

Alongside these unconventional pioneers, researchers around the world are experimenting with other minimally invasive methods to persuade protein molecules away from the objects they have clung to for decades, centuries or millennia. This means that curators like Arhire should brace themselves for a profound shift in identity. They are now keepers of not only texts and manuscripts but also biological stories reaching far beyond the written word.


Zilberstein and Righetti’s journey into the past began a decade ago, with a battered, crumbling Bible. Righetti, a chemist at Milan Polytechnic University, had spent most of his career developing methods to separate proteins with increasing precision.

Now 81, Righetti is tall, with blue eyes and a neat, white goatee. Energetic and seemingly unaffected by the heat when I met him in Milan one warm day this past spring, he led me on a blistering tour through the city’s streets and galleries, joyfully pointing out medieval courtyards—“magnificent cloisters!”—and a World War II submarine displayed on the grounds of the national science museum.

two people study historic manuscripts
Svetlana Zilberstein and Pier Giorgio Righetti take a sampling from Giacomo Casanova’s manuscript Histoire de ma vie in the French National Library in Paris. Courtesy of Gleb Zilberstein

Righetti grew up in grim poverty. His father fought in the Second World War, and the family became homeless after fleeing Allied forces. His earliest memory is of living in tents under the rafters of an old castle; families were separated by hanging sheets. After several years in a seminary, training to become a priest, at age 15 Righetti quit and rejoined his family in Milan; viewing science as his best chance for a more prosperous future, he studied organic chemistry, earning a PhD from the University of Pavia. He fell in love with proteins, and even now he speaks about them not as molecules but as characters in a human drama. “They build things, sweep the streets, put people in prison. Without them we couldn’t live.” Righetti became an expert in using electromagnetism to coax protein molecules through gel, a process known as electrophoresis, and he traveled extensively over the years to teach the techniques he pioneered to scientists all over the world.

In 2010, a group of Italian researchers asked Righetti, in the twilight of an already accomplished career, if he could retrieve anything from a few crumbs of a disintegrating Bible, written in Latin, that was taken to China in the 13th century and returned to Italy 400 years later. Hoping to learn more about the Bible, which some speculate may have belonged to Marco Polo, the researchers had painstakingly reassembled the book from more than 10,000 pieces and concluded that it was made in northern France in the 1230s. They wondered if further clues might be held in the leftover fragments.

Rather than use electrophoresis, Righetti opted to try a more powerful technique called mass spectrometry. Once used to analyze uranium isotopes during World War II, the technology works by converting molecules in a sample into a gas of charged ions, for example by bombarding them with a beam of electrons, then using electric and magnetic fields to separate the ions as they travel at high speed through a vacuum. The resulting “mass spectrum” (which shows the mass-to-charge ratios of the ions plotted against their intensities) acts like a molecular fingerprint.

Researchers were increasingly using the technology to study proteins, but applying it to ancient, degraded samples was an enormous challenge. Righetti’s first attempts to extract trace proteins from a tiny parchment fragment, using the digestive enzyme trypsin, failed to produce any results. Eventually, after first softening the sample in a microwave, he succeeded, and detected eight distinct proteins. The results showed that the parchment was made from calf tissue, rather than fetal lamb tissue, as was long assumed. The finding itself may have been of only niche interest, but Righetti was one of the first researchers to show that proteins really could yield information about historical artifacts. Still, the Bible, with its disastrous state of preservation, was a special case, and the technique’s potential for historical scholarship seemed limited; few other curators would be willing to let Righetti, or anybody else, destroy even a sliver of a treasured artifact in order to analyze it.

Then he got a call from Gleb Zilberstein, with whom he’d collaborated on other projects in the past. “I know how we could do it without needing to take a sample,” he said.

two hands in blue plastic gloves hold up a letter
Zilberstein examines Dracula’s letter after removing the strips of film, which he designed to capture proteins from a surface without damaging the artifact. Andrei Pungovschi

Like Righetti, Zilberstein saw science as a path to a better life. Now 53, he grew up in a Soviet industrial town on the West Siberian Plain. “Everything was gray,” he says. When he was 10, his uncle gave him a fossil collection, including ammonites and sharks’ teeth from the Kazakh desert, which once lay at the bottom of an ancient sea. His eyes light up as he talks about it, pointing his fingers toward his forehead as if zapping his brains with rays of energy. “To see these fossils, it makes you crazy!” he says. “Even if you live in a poor country, you can find bright things.”

He studied chemistry and physics at Novosibirsk State University in Siberia, where he met Svetlana, who was also a student, at a nightclub. Gleb’s degree was interrupted by two years of compulsory military service, but he avoided routine duties by inventing a new type of gas mask that captured volatile organic compounds using a filter made from dried blood.

When the Soviet Union collapsed, Gleb declined to apply for either Russian or Kazakh citizenship, and after college he and Svetlana emigrated to Israel, where they now live in Tel Aviv, near the ocean. Gleb has since founded a series of companies that commercialize his proprietary technologies, many of which capture molecules using their own minute electric charges.

One of Zilberstein’s products was designed to kill bacteria by pulling on their proteins to break apart their cells. After seeing Righetti’s journal article about the “Marco Polo” Bible, Zilberstein realized the technology might be adapted to historical research, by using the same method to tease molecules off an artifact’s surface while leaving the object intact. On the phone, he explained to Righetti that he’d already engineered several polymers that could attract proteins and other types of molecules according to their charge. Righetti was sold.

For their first experiment, Zilberstein traveled to a government archive in Moscow to study the original manuscript of The Master and Margarita, by the Russian novelist Mikhail Bulgakov, who died from kidney disease at age 48 in 1940. In the library, he covered the manuscript pages with ground-up polymer beads, then Righetti used mass spectrometry to analyze the molecules Zilberstein had captured and found liberal traces of morphine. In a subsequent paper, published in the Journal of Proteomics, they concluded that Bulgakov was self-medicating with the drug while he wrote.

Righetti describes the response as “a barrage of criticism.” One critic complained that the method risked contaminating fragile pages with the ground-up beads; another questioned their conclusion, pointing to the lack of any written record of Bulgakov taking morphine at this time (despite a period of addiction earlier in his life). If the drug was present on the pages, the critic said, it could have come from later readers, perhaps Russian secret service agents studying the subversive text, an objection that Righetti and Zilberstein agreed was “legitimate.”

an illustration of a bearded man in an red garb with a bow and arrows
Marco Polo, by the 18th-century watercolorist Giovanni Grevembroch. A Bible some speculate may have belonged to the famed explorer was among the first historical artifacts studied using advanced protein analysis. Bridgeman Images

So they tried again. They embedded the beads within a thin film of a plastic called ethylene vinyl acetate, or EVA, so that no debris would be left on the page when the film was lifted off. The film can be tailored for different analyses by embedding beads that attract different molecules. “We found saliva traces,” Righetti says, including “three markers of the kidney disease that took Bulgakov to the grave.”

For Righetti and Zilberstein, the kidney disease proteins they detected proved the molecules really did come from Bulgakov. Moreover, the analysis showed that their method could yield new discoveries without damaging an original document. It also identified human proteins left by individuals who’d handled an artifact in the past, opening up a whole new way of reading historical sources. A text might tell us what a person wrote—what they wanted us to know—but the physical residue of Bulgakov’s drug consumption showed how previously undiscovered chemical traces could reveal details about an author’s lifestyle and health and might even suggest his state of mind.

After that first dramatic study, Righetti and Zilberstein made a series of striking findings. They found tuberculosis proteins on the collar of the shirt that Anton Chekhov was wearing when he died, as well as, more surprisingly, a human protein Zilberstein says is produced when a stroke cuts off the blood supply to the brain, suggesting that this, and not the infection itself, was Chekhov’s immediate cause of death. They also identified TB proteins on an obscure letter George Orwell sent to a Russian editor—a particularly impressive finding, as the letter was typewritten. The traces were found on the corners of the page, where Orwell would have pulled it from his typewriter. He probably licked his fingers first, Righetti suggests.

Fascinated by the possibility of detecting past diseases, Righetti spent weeks sampling pages from medieval death records in Milan, written during a bubonic plague epidemic in 1630 that wiped out nearly half of the city’s population. The researchers reported finding more than 20 proteins from Yersinia pestis, the bacterium that causes plague, as well as traces of corn, carrots, chickpeas and rat droppings. For Righetti, an enthusiastic storyteller, the findings conjured overworked scribes in infested halls, eating as they recorded fatalities and leaving their books open at the end of each day. “At night the rats were scurrying, looking for food,” he says. “It was an incredible picture.”

These were vivid glimpses into the world of the dead, what Righetti colorfully calls “stroboscopic flashes of the netherworld.” “You put on Chekhov—flash! You put on Bulgakov—flash!” Whether you’re analyzing an object from 100 years ago or 1,000 years ago, the method “allows you to get something no one ever thought of.”


At the archive in Sibiu, Zilberstein leaned over the table, hands shaking slightly as he used tweezers to place several small, beige squares—the EVA films—onto Dracula’s 500-year-old letter, plus two others written by the prince. Heavyset and dressed in black, with a strong Russian accent and brown ringlets with bleached-blond tips, he created an impression somewhere between disco star and movie villain. He noted greasy spots on one of the letters, which he suggested could have come from the wax seal. Or maybe it’s blood, he said with a smile: “We will extract the spirit of Dracula!”

Svetlana—blond hair, freckles, dressed in neutral tones—hovered as he worked. Together with Righetti, the Zilbersteins are commercializing the EVA technology through a company called SpringStyle Tech Design, and plan to make it available to institutions such as museums, libraries and government archives. Svetlana whispered instructions as he worked, nudging him to adjust the placement of the squares. These would sit in place for an hour, weighed down by books, but after ten minutes Gleb peeked underneath to check for signs of damage to the letters. It’s important to be sure, Svetlana said.

She recalled how at the State Hermitage Museum in St. Petersburg they were given rare permission to study the Donna Nuda, a masterpiece from the school of Leonardo da Vinci (perhaps even made by the painter himself). When they lifted the films, they were horrified to see white patches on the surface of the priceless artwork. Fortunately, all that was affected was a recent coat of varnish, which responded more forcefully to the charged polymers than they anticipated. The implacable Russian curator wasn’t bothered; his assistants soon fixed the finish with cotton swabs dipped in alcohol. But at the time, “We were scared,” Svetlana laughed. “We expected they would send us to jail!” Arhire, watching with a grim face over his Dracula letters, did not appear to share the joke: “Here, they will just send me.”

The Donna Nuda study was the first that aimed to analyze the ingredients used for paint in Leonardo’s studio in such detail. It was a success: Through the varnish, the EVA films coaxed traces of linseed oil, conifer resin, rosemary oil and egg yolk (evidence of the egg-based painting medium tempera grassa). One expert described the findings as “the artistic equivalent of discovering the recipe for Coca-Cola.” The researchers suggested that rosemary oil, which Leonardo hadn’t been known to use, likely diluted the paint and delayed drying time, helping the artist blur features such as landscapes when he wanted to create a sense of depth.

a historic bible is open to the middle
The “Marco Polo” Bible. A 2012 study found the parchment was made from calf tissue by identifying bovine proteins, igniting a boom in biochemical analysis of historical artifacts. © 2022 Facsimile Finder – Giovanni Scorcioni – Republic of San Marino
two hands in blue plastic gloves hold up a smart phone
Zilberstein scans a document with a proprietary phone app; areas with the highest concentration of proteins and bacteria, which are the most promising targets for chemical analysis, are seen in red. Andrei Pungovschi

As we waited for the electrostatic forces to draw proteins from the letter written by Dracula, Arhire told me that, in his view, Dracula didn’t deserve his notorious reputation. For one thing, impalement was a normal punishment at the time, he said. Dracula spent his life fighting for control of his homeland against Hungarians to the north and Ottoman Turks to the south, which put him at the frontline of the clash between West and East, Christianity and Islam. For Romanians today, Arhire explained, Vlad isn’t a villain but a national hero.

That dynamic history is partly what interests Zilberstein. Transylvania is “a unique place,” he says, because it acted as a meeting point for soldiers, slaves and merchants from all over Europe and as far away as Mongolia and Persia. These migrants and travelers would have carried trade goods, cultural traditions and epidemics. Europe had also just come through a period of exceptionally cold climate, which Zilberstein suggests might have left its mark on people’s “proteomes,” the snapshot of proteins in a person’s body at a given time. (Accounting for all the modifications to basic proteins encoded by the 25,000 or so human genes, there may be a million or so human proteins.) When analyzing Dracula’s letters, Zilberstein and Righetti will look for proteins related to famine, stress and diseases such as syphilis and smallpox; dietary proteins from food or wine; and pests such as rats and flies. The hope is that this will provide a window into 15th-century life in this turbulent period, a few decades before Columbus’ first contact with the New World.

And of course there’s the tantalizing prospect of uncovering more about Dracula himself, who was killed by soldiers loyal to his Ottoman-backed rival in the winter of 1476-77, shortly after regaining the Wallachian throne. “We’ll see if he was sick,” Zilberstein said. “There are stories that Dracula cried tears of blood.” Such a condition, known as haemolacria, actually exists; if the tests turn up hemoglobin or other blood proteins on the letters, “that could be preliminary evidence of haemolacria.”

a letter with handwritten text shown from above
Another letter by Dracula, marred by red sealing wax. Experts searching for biochemical traces zero in on signatures, corners, even sweat or tear stains. Andrei Pungovschi

It’s a fascinating idea, but the ease with which Zilberstein and Righetti link dramatic narratives to the proteins they find makes some scholars uneasy. A few years ago, they investigated the notebooks of the 17th-century astronomer Johannes Kepler, which are held in St. Petersburg. They found no proteins but detected heavy metals including silver, gold, arsenic and lead, and concluded—to the horror of at least one Kepler expert, who insists the idea bears no relation to the historical record—that the astronomer was also a practicing alchemist.

In a subsequent study, they sampled Joseph Stalin’s personal copy of A. N. Tolstoy’s 1942 play Ivan Grozny, stored at the Russian state archives in Moscow, which the dictator read during the Second World War. Stalin’s scrawled, repetitive notes in the margins hint at an agitated mind. On those pages Righetti and Zilberstein detected lithium. “I said, he was bipolar like Winston Churchill!” Righetti recalls. He titled their paper about the study, which appeared in the journal Analytical and Bioanalytical Chemistry, “Stalin’s ‘Black Dog,’” even though at the time lithium was more commonly used to treat gout than manic depression.

Scholars have also called into question some of their earlier findings and interpretations. One leading researcher pointed out to me that the protein Zilberstein identified as suggesting that Chekhov had a stroke might more accurately be described as a general marker of inflammation, and is associated with a range of conditions, from gastric cancer to ALS. And this past July, in a wide-ranging overview of the field in the journal Chemical Reviews, three leading proteomics researchers cited the Bulgakov study, with its surprising discovery of kidney proteins, as an example of an intriguing but “improbable” result that required further analysis and confirmation. (In an email, the paper’s authors said that Zilberstein and Righetti did not account for how kidney proteins might have been transferred to a manuscript page. Zilberstein and Righetti counter that kidney proteins could have been secreted through Bulgakov’s sweat, saliva or urine and may have transferred to the manuscript because while he was ill Bulgakov wrote from his bed.) Still, Matthew Collins, a bioarchaeologist and world leader in ancient proteomics, and a co-author of the Chemical Reviews paper, says he remains “impressed” by the innovative design of the EVA technology and its ability to target different types of molecules.

An investigation the pair is now completing, into the death of the American author Jack London, highlights both the limitations and the singular opportunities of their technology. London died in 1916 in uncertain circumstances. The cause was recorded as kidney failure, but within days newspapers began to report rumors that he regularly used drugs such as opium, morphine and heroin, and may have committed suicide by taking an overdose. Working with Richard Rocco, a pharmacologist at Samuel Merritt University in California, Zilberstein and Righetti sampled several of London’s personal effects, including glass drug vials, a leather medicine case, and a pamphlet on rattlesnake bites. They found traces of opium and heroin on the vials, but only in combinations with other drugs that suggested they were part of over-the-counter remedies popular at the time for treating coughs and colds. “There were no remnants of heavy drugs,” says Righetti. “Our data exclude the hypothesis of suicide.”

Meanwhile, in a possible blood spot on a journal article about ingrown toenails, they found blood proteins modified by sugars, suggesting London had undiagnosed diabetes. Opioid use can both cause and exacerbate diabetes, and opioid painkillers are less effective in people with the disease, Zilberstein says. He suggests the three factors worked together: The pain from diabetes caused London to take increasing doses of over-the-counter opioids, which in turn damaged his kidneys and ultimately killed him.

scientific study of an ancient manuscript
Films placed on the document absorb protein fragments and other molecules. Andrei Pungovschi

Jay Williams, a Jack London scholar and biographer, welcomes the research, but suggests that when investigating the lives of authors scientists should work directly with humanities scholars to interpret their findings. He points out that scientific data alone can’t reveal a person’s motivations. Even if London’s only source for heroin was over-the-counter cough syrup, for example, this doesn’t mean London wasn’t addicted; it also doesn’t reveal whether an overdose might have been accidental or intentional. Moreover, Williams argues that biographers have already provided “a much fuller picture of London’s drug use” than the study admits. That research, he says, “should have been incorporated into this scientific examination to understand more exactly how London died.”

Kenneth Brandt, a literature scholar at Savannah College of Art and Design in Georgia and executive coordinator of the Jack London Society, agrees. Zilberstein and Righetti’s findings “offer valuable factual data that will enhance the biographical records of these writers,” he says. But he stresses that such data “will have to be carefully contextualized in relation to an informed understanding of the authors’ lives, which presents great collaborative opportunities for scientists and literary scholars.”

Righetti dismisses criticism that he and Zilberstein overreach in their interpretations. “Classical academics—they write papers so nobody can read them,” he says. “You have to make them fun, clear, understandable. If I can put in literature, history—why not?” After a career engaging only with chemists, his work is resonating with popular audiences around the world. If some people think his claims are too bold, “I don’t give a damn! I’m having the biggest fun of my life.”

Back in Sibiu, the hour is up, and Arhire can relax: The Dracula letters are unblemished. Gleb carefully seals the films in plastic trays and covers them in bubble wrap, and Svetlana hands out chocolates to celebrate. Afterward they walk out into Sibiu’s cobbled streets and order dessert wine at a nearby café, looking like any other tourists—except that tucked inside a black shoulder bag are hidden secrets from the reign of Vlad Dracula, ancient molecules that evoke distant worlds.


If Zilberstein and Righetti are self-styled outsiders, Matthew Collins might represent the academic establishment. Collins runs two labs, one in the archaeology department at Cambridge University in the United Kingdom and another at Denmark’s Natural History Museum in Copenhagen. Over the past two decades, Collins and his colleagues have used increasingly sophisticated mass spectrometry techniques to reach ever-further back in time, helping to transform fields such as archaeology and evolutionary biology. They revealed the first ancient proteome of a 43,000-year-old woolly mammoth, and used proteins found in 42,000-year-old bone fragments, from a cave in central France, to prove that they (and the delicate artifacts they were found with) belonged to Neanderthals, rather than modern humans, as some researchers claimed. They have even read partial protein sequences from a 3.8 million-year-old ostrich shell.

When I visited this summer, Collins’ Cambridge lab appeared like any other molecular biology workspace: clean, white countertops, with cupboards stocked with centrifuges, reagents and pipettes. But as we walked to a nearby storage room, we passed a window through which rows of human skulls were displayed, and in the room itself the shelves were full of boxes with labels like “lion cub,” “squirrel monkey” and “aardvark.” When I asked Collins what was in them, he shrugged and said, “tissues.”

Collins’ driving interest, ever since seeing Jaws as a teenager inspired a love of sharks, has been to understand the animal world, but the growing power of protein analysis is leading him to explore the human past too. He says medieval parchment, made from animal skins, is especially ripe for study, because it can provide insights into everything from farming practices to monastic life. Collins’ research group developed its own noninvasive sampling technique, which they came up with “by accident,” when his colleague Sarah Fiddyment began to study 13th-century pocket Bibles.

Collins had arranged with the relevant archivists that Fiddyment would take razor-thin slivers from the margins of certain pages to test for proteins. “But then,” he says, “she came to me ashen-faced.” The conservators had refused to let her touch their books. Desperate to save her project, Fiddyment spent two weeks with the archive staff, studying how they worked. She realized that they routinely removed dirt from ancient pages by rubbing them gently with common PVC erasers. Ingeniously, Fiddyment asked for the eraser crumbs, and she was able to extract proteins from them as well as if not better than from a piece of actual parchment. The proteins, she found, had been dragged from the page by electrostatic charges generated by rubbing, like when you generate static electricity by rubbing a balloon on your hair.

The beauty of it was that the conservators were already generating the samples. “They’re throwing away the eraser crumbs,” Collins says. “We thought, if they’re doing this routinely in conservation labs around the world and tipping the crumbs into the bin, can’t they tip them into tubes instead and send them to us?” Fiddyment’s project included more than 70 Bibles from archives all over Europe. Now, instead of having to knock on doors, curators are coming to them.

One project captured proteins from 500-year-old Iroquoian bone points from Canada simply by testing the plastic bags in which the tools had been kept. The analysis showed they were made not from readily available deer and beaver bones, as assumed, but from the bones of humans and bears. Oral traditions and clan emblems suggest some Iroquoians saw themselves as closely tied to bears, but finding physical evidence of how this bond played out in daily life has been challenging. The researchers concluded that the weapon-makers deliberately chose bear and human bones “to materially express their mutual entanglement, to symbolically transpose the hunting skills of bears into the hands of humans.” Without the proteins this dimension would have been completely missed.

The success of these varied approaches has inspired other researchers to test a range of minimally invasive methods, from digestive enzymes to abrasion. Earlier this year, Australian researchers used dermatological tape to pull proteins from the bone fragments of a 2,500-year-old Egyptian mummy. Collins says dermatological tape doesn’t work very well on parchment, but he’s getting excellent results with NanoTape, a commonly available adhesive strip inspired by the structure of gecko feet.

Simply capturing molecules isn’t enough, however. Interpreting the data is just as hard. For example, DNA can be precisely sequenced, but analyzing proteins by mass spectrometry is more like matching patterns. When proteins are ionized for mass spectrometry they tend to break into pieces. The jumble of sizes and shapes detected is checked against a database of known proteins; software then calculates which pattern of masses might result from known proteins and comes up with a list of potential matches. That means teasing out which proteins are present in a complex and degraded sample can be difficult. What’s more, most protein databases focus on diseases or economically important plants and animals, so matches can be biased toward those types of entries.

Then there’s the ever-present possibility of contamination—how do you tell ancient proteins on a sample from modern ones? In one study, researchers analyzing the proteins on a Neolithic pot from Germany reported the highly surprising discovery of Crimean-Congo hemorrhagic fever, a disease never found so far north at the time; a separate research group subsequently pointed out that the identification was based on a single virus protein that was being used as a research tool in the same lab, raising the possibility of contamination.

Collins is now looking for ways to identify distinctive damage patterns in ancient proteins that would allow researchers to separate out modern contamination and “read the authentic signal,” as researchers can already do with DNA.

In the zoological storage room in Cambridge, Collins’ colleague Matthew Teasdale, an archaeogeneticist, placed a small plastic box on the table and removed the lid. Inside was a set of neatly labeled plastic tubes about an inch tall. Each held at its tip a cluster of tiny white tendrils—the eraser crumbs collected by conservators as they worked. Barely visible, these flecks have been brushed away without a thought for years. Now they are “changing how you think about an archive,” Teasdale says. “Every archive is now a biological resource.”


A few years ago, John McNeill, then president of the American Historical Association, delivered a provocative address inspired in part by the work of Righetti and Zilberstein. “I will raise the question,” he said, “of whether or not we will reach a point that might be called ‘peak document,’” when the rate of historical information we can glean from rereading texts will be overtaken by data from scientific techniques.

McNeill pointed out that academic history is still often divided by national boundaries: Scholars specialize in “Imperial China,” for example, or “colonial Latin America.” In ten or 20 years’ time, he suggested, historians might specialize instead in interpretive approaches enabled by advanced tools, from the analysis of ancient proteins and DNA to paleoclimatology or the microscopic examination of teeth. Perhaps they will ask completely new kinds of questions. Zilberstein likes the idea of building a “molecular portrait of dictators,” looking for commonalities in habits, behavior and stress.

a plastic glove hand places a sample in a tray
Samples are ready for analysis by a technique called mass spectrometry. Andrei Pungovschi

Deb Donig, a literary scholar at California Polytechnic State University, has argued that proteomics could trigger a similar “radical reconfiguration” in literary analysis. Instead of focusing on national affiliations or character, such as Russian literature or Jewish literature, we might ask “how those who are undergoing stress write; how the elderly write; how people write in the context of famine.” Or, given the popularity of morphine among the authors studied by Righetti and Zilberstein, a researcher might investigate how modernist literature was informed by opioid drugs.

Another promise of ancient proteomics is access to people who are underrepresented in the written record. Conventional history suffers from a “text fetish,” McNeill argues, meaning that it tends to record wars, politics and economics, and generally emphasize record-keeping groups and societies. Proteins, on the other hand, can tell us of the activities, and therefore to some degree the beliefs and practices, of people whose concerns were never put down in ink.

Along these lines, Zilberstein and Righetti are planning to search for biochemical traces from 18th-century logbooks of ships carrying Black slaves from Africa to America. While the texts themselves were written by slave traders, the researchers hope proteins left behind will reveal something of the slaves’ unrecorded perspectives, such as their living conditions, including diet or diseases, on board the ships.

As usual, Zilberstein is already dreaming about the implications not just for history but for industry. Researchers already scour rainforests and oceans to look for microbes (and ultimately molecules) to aid drug development and other new technologies. Zilberstein believes libraries and museum collections could perform a similar role, as an “Aladdin’s cave” of rare biomolecules stored for hundreds or thousands of years, with potential uses from green technology to medicine.

When he and Righetti applied their EVA films to an Egyptian mummy, for example, they found traces of bacteria known to degrade plastics, which he thinks were selected for by the mummification process. He speculates that mummies might be a good place to search for previously undiscovered species with similar abilities: “It is a unique cosmos for organisms that like to eat oil.” Or, he says, traces of skin or blood left on clothing worn during epidemics might preserve biological responses, such as antibodies, that could inform the design of vaccines against coming pandemics.

In the meantime, ancient proteomics is already becoming part of the routine way certain historical documents, such as parchment manuscripts, are interpreted. So far Collins and his colleagues have used the eraser method to analyze more than 7,000 manuscripts from archives all over the world. And Zilberstein speaks about adding his EVA films to the rollers of high-throughput document scanners, to incorporate proteomics into digitization projects for government libraries, say, so that millions of pages can be swept for proteins and other molecules at the same time that they’re scanned. “We could do that now,” he says. Until that happens, however, he and Righetti will aim to rewrite history one document at a time.