The Dangers of Mirrored Life
Creating mirrored organisms using synthetic biology could seriously harm extant life, a 300-page report claims. While the risks from mirrored life are uncertain, it is best not to find out.
The creation of a “mirrored” organism could “trigger severe ecological disruptions,” according to a 300-page technical report released today. Its authors claim such organisms could quickly spread across the world, fatally infect humans, and “directly drive vulnerable plant and animal species to extinction.” The report accompanies an article in Science, also released today, entitled “Confronting Risks of Mirror Life.”
But what, exactly, is a mirrored organism? To answer that, let’s consider how extant life works.
Proteins, sugars, lipids, and nucleic acids — key molecules used by cellular life — are all “chiral,” a term derived from the Greek for “hand.” Just as our hands cannot be perfectly aligned on top of one another regardless of how they are rotated, despite being mirror images, the same is true of chiral molecules. The two mirror images of a chiral molecule are called its enantiomers; they possess the same atoms connected in the same order, but with mirrored three-dimensional arrangements. One enantiomer of a molecule might fit perfectly into a protein receptor, while the other may not fit at all. This is no mere quirk of biology, but a crucial and ubiquitous feature of it.
A mirrored organism would use right-handed molecules everywhere a naturally-occurring organism uses left-handed ones, and vice versa. It could thus elude the typical chiral interactions by which microorganisms hunt their prey. A mirrored bacterium would be totally resistant to bacteria-infecting viruses, called bacteriophages (or just “phages”). It would also be undetectable by key parts of the human immune system. Such mirrored invaders could theoretically spread across the Earth while evading the biological defenses that have evolved to check such threats.
Experts say that mirrored organisms are still decades away from being a reality. But out of caution, the new report signals alarm well in advance.
For the article in Science, more than 30 scientists from ten countries are calling on the broader community to confront the risks of mirrored life. “In the absence of compelling evidence for reassurance,” they write, “mirror[ed] organisms should not be created.” If a mirrored bacterium were ever released, or accidentally escaped confinement, they warn, it could infect and kill humans, spread through animal or plant populations, and wreak havoc upon ecosystems.
Among the report’s authors are 16 members of national academies, two Nobel Laureates, and several researchers who have previously been frontrunners in the quest to create mirrored life. This includes recipients of a three-million-dollar grant, awarded by the National Science Foundation in 2019, to “design, construct, and safely deploy synthetic mirror cells.”1
Now, it seems, these researchers have changed their minds. Biochemists Kate Adamala and Michael Kay once had aspirations of helping create mirrored life; in a turnaround, they argue it shouldn’t be created.2 George Church, John Glass, and Craig Venter — all of whom have spent decades researching synthetic life forms — also co-author this call for restrictions.
It’s important to note that the 300-page report, while presenting detailed evidence for the potential harms of mirrored life, uses painstakingly speculative language. The words “may” and “possible” appear hundreds of times. This reveals an uncomfortable scientific truth: nobody knows exactly how a mirrored organism would spread, how infectious or deadly it would be to natural life if released, or whether it could somehow be contained. Nobody even knows when such an organism could be created — it may not happen for several decades. Still, the Science article’s authors argue that now is the appropriate time to “consider and preempt risks before they are realized.”
Regardless of speculation, synthetic biologists are inching closer to creating “artificial life,” and the tools for constructing mirrored organisms are improving rapidly. While biosecurity experts wrestle with when to publicize risks, there are cases where the downsides from a perceived threat are large enough — despite unknowns — that it’s sensible to call awareness to them early. It gives humanity time to steer away.
This is one of them.
Molecular Hands
The Last Universal Common Ancestor, or LUCA, was a single-celled organism that gave rise to all life on present-day Earth — from towering trees and microscopic bacteria to birds, bees, and humans. Evolutionary evidence recently published in Nature Ecology & Evolution suggests that LUCA lived about 4.2 billion years ago, at a time in Earth’s history when asteroid collisions were common and atmospheric conditions were long thought inhospitable for life.
Despite having descended from simpler lifeforms, LUCA was surprisingly complex; it may have had hundreds of genes and derived its energy from hydrogen gas and carbon dioxide. Some scientists even think it had a virus defense system; “researchers say LUCA likely housed 19 CRISPR genes, which bacteria use to slice up viral threats,” reports Quanta Magazine.
This ancient cell also adopted preferences for one molecular chirality over the other, passing those preferences down to its descendants. Nearly all living cells on Earth rely on the right-handed form of glucose, for example — and LUCA likely did the same.3 Proteins are constructed from 20 amino acids, of which all but one take the left-handed form.4 As far as we know, right-handed proteins never occur naturally. DNA and RNA molecules are also built from exclusively right-handed nucleic acids.
LUCA likely adopted these chiralities entirely by chance.5 But as its ancient molecules began replicating, those first chiralities inexorably spread. New chiral molecules emerged interdependently, like nuts and bolts. Once the first bolts tightened clockwise, that clockwise pattern proliferated.
It’s difficult to imagine any evolutionary path that could “reverse” the tangled molecules of life into their mirrored forms, because incremental mutations which swap out a few chiral molecules for their enantiomers would impair or kill the cell. As Richard Dawkins said of biological wheels: “the engineering solution can be seen in plain view, yet [it is] unattainable in evolution because it lies [on] the other side of a deep valley.” Evolving mirrored life would be like converting a Boeing 747 into a submarine; mid-flight.
Or, for another analogy, consider a hypothetical country with no traffic laws regarding which side of the road to drive on. Arbitrarily, drivers on a stretch of road begin using a particular side. Because it’s easier to go with the overall flow of traffic, other drivers join this side, and it soon becomes the norm. Signs, lights, and cars themselves are eventually adapted for people driving this chosen side. Although left- or right-hand driving can emerge spontaneously, once it becomes normalized, it now can’t switch spontaneously — it can only change by top-down design.
Because of a similar lock-in, the only way for a mirrored organism to arise today would be through human synthesis. But if we were to rewind the tape of life, there is no reason LUCA could not have chosen the opposite chiralities. Mirrored molecules have the same physical and chemical properties.6 They melt at the same temperatures and have the same solubilities in water. A mirrored organism, interacting with achiral molecules, would behave just the same as its natural counterpart. There’s nothing more fantastical about mirrored life than counterclockwise-tightening screws.
But mirrored molecules do have distinct biological properties. Just as a handshake is an interaction between two right hands or two left hands (the left-right version is an awkward mess), the success of molecular interactions similarly depends upon which enantiomer is involved.
Two enantiomers can bind to distinct proteins, for example, thus causing serious and often unpredictable problems. A drug called thalidomide, initially made by a West German pharmaceutical company called Chemie Grünenthal, was manufactured and given to patients as a mixture containing both enantiomers. One enantiomer acts as a mild sedative (it was prescribed for everything from pneumonia to the common cold or flu), whereas the other causes severe birth defects.7
During the few years that thalidomide was available on the market — and before Australian doctor William McBride warned of its consequences in a letter published in The Lancet in 1961 — it affected more than 10,000 babies, killing about half of them within months after birth.
In short, life has been “homochiral” for more than four billion years. Across the tree of life, organisms strictly require exactly one of the two chiral forms of their molecular building blocks — amino acids, nucleotides of RNA and DNA. These building blocks assemble into larger structures, such as proteins and genomes, that inherit the same handedness. Living matter has never been assembled from mirrored versions of these building blocks, and indeed L-ribose and L-nucleotides are essentially absent from nature. But in the future, perhaps, humans could jump start this process, doing what evolution has not.
Making Mirrors
Scientists have known about the chirality of life since the mid-19th century. In 1848, a 26-year-old researcher at the Ecole Normale Supérieure in Paris, named Louis Pasteur, discovered that crystals made from the same molecules could exist in two mirrored forms. According to a 2021 article in the journal Chirality:
[Pasteur] was observing crystals of the double sodium-ammonium salt of tartaric acid under a microscope. To his surprise, he noticed that each crystal has a tiny facet on one of its edges oriented sometimes to the right and sometimes to the left. He could separate them by hand with a tweezer.
By 1860, Pasteur had performed additional experiments on chiral molecules. He then delivered a pair of lectures during which he noted the existence of organic molecules that, “when dissolved in water, possess the rotative property,” referring to the way that some chiral molecules “twist” polarized light one way or the other.8
In 1950, the idea of mirrored life took root in science fiction. Arthur C. Clarke — the British writer most famous for 2001: A Space Odyssey — published a little-known story called Technical Error, in which a power plant employee is transformed into his mirror image after an accidental short-circuit. Even the often-prescient Clarke couldn’t have predicted that mirrored life would become the subject of scientific work only a few decades later.
One of the first times that researchers made large mirrored molecules came in 1992, when a trio of biologists at the Scripps Research Institute in California used a series of chemical reactions to build the HIV-1 protease enzyme in both its right- and left-handed forms. This enzyme chops up short strands of proteins, known as peptides. The synthesized enzymes — perfect mirror images of one another — “showed reciprocal chiral specificity on peptide substrates,” the scientists wrote. “That is, each enzyme enantiomer cut only the corresponding substrate enantiomer . . . ” In other words, mirror-image proteins behave in identical ways when acting upon their chiral counterparts.
Two Purdue chemists — James Brewster and Michael Laskowski, Jr. — read the paper and fired off a response. Not only are the HIV-1 protease results obvious, the two men argued, but they further suggest that “enantio-life will be as viable as ‘normal’ life in vitro.” Brewster and Laskowski, Jr. continued their letter prophetically, writing:
Escaped enantio-life would have a built-in immunity to attack from ‘normal’ life . . . Would-be synthesizers of life-based on amino acids and nucleic acids need to consider these matters in detail before getting started. Such organic or biochemists should prepare for trouble not only with the public and politicians but with their peers as well.
Despite their suggestion, interest in creating mirrored organisms has increased over the past few decades, as the ranks of “would-be synthesizers” have grown in number. The creation of artificial life has long been a keystone of synthetic biology, for example. And as scientists get closer to achieving that goal, they could plausibly adapt the same methods to create mirrored life. It would be an undeniable achievement, too, because rather than drawing from nature — as biotechnologists often do — such synthesizers would have to create it from scratch.
Bacteria — (mostly) single-celled organisms lacking a nucleus — would be the easiest kind of life to reconstruct in a mirrored form.9 Since bacteria reproduce asexually, only a single cell is needed, along with enough food and the right conditions for growth, to produce a culture of some ten billion bacteria in a liter flask. The key problem, then, is to build a single mirrored cell capable of self-replicating.
There is no simple way to build such a cell.
Several decades from now, though, perhaps scientists could build a mirrored cell from the “bottom-up.” First, one could assemble mirrored versions of the protein-assembling machinery, including ribosomes. Then, after chemically synthesizing a mirrored genome by stitching together L-nucleotides, and uniting the mirrored ingredients inside a lipid membrane, scientists could theoretically coax the ribosomal machinery to begin “reading” the mirrored genome and transcribing its information to generate the other protein components of the cell. The grand goal is to “boot up” these ingredients into a fully self-replicating organism.
It's hard to predict exactly how and when breakthroughs arrive. But given current rates of progress, these steps may be solved within 30 years. It wouldn’t be shocking if it took half that time, either.
The first step could be achieved using solid-phase peptide synthesis, a chemical method used to make peptides by stitching together individual amino acids. Pioneered in the 1950s, it works like this: chemists begin by attaching an amino acid to a solid surface with a covalent bond. Then, an atom on that amino acid is stripped away, forcing it to chemically react with the carbonyl group of a second amino acid. The two molecules click together, forming a small peptide with two amino acids. This process is repeated to construct larger protein fragments.
Solid-phase synthesis works equally well for left- and right-handed amino acids, and chemists have used it to build peptides stretching 164 amino acids in length.10 Chemically-made proteins behave identically to proteins made inside living cells, even though the two are made in opposite directions. (Chemists construct proteins from the carbonyl side to the amino group, whereas cells do the opposite). In 2014, biochemists at the University of Utah used solid-phase synthesis to build small, mirrored peptides and ligated them together into a 312-amino acid protein that worked just as well as its non-mirrored, or natural, counterpart.
Mirrored DNA is similarly simple to stitch, and current methods can string together up to 200 nucleotides before becoming impractical.11 Individual strands of mirrored DNA can then, in principle, be ligated (using a mirrored form of a ligase enzyme) to build entire mirrored genes stretching several thousands of nucleotides in length. Several companies, including a German company called Biomers, sell mirrored DNA strands online — albeit at a premium.
Scientists are slowly putting these pieces together to build simple, “mirrored” biological systems. In 2021, scientists created a mirrored version of DNA polymerase, an enzyme responsible for copying a cell’s genome, and demonstrated that it could accurately replicate an entire “mirror-image gene.” In 2022, the same laboratory repeated the process for RNA polymerase, an enzyme that transcribes DNA into RNA as the first step in protein synthesis. The mirrored enzyme seamlessly transcribed 2,900 bases of mirrored DNA into mirrored RNA.
In short, some of the required technologies to make mirrored life exist today and will continue to get better (not driven by a directed pursuit of this goal, but rather by ambient progress in synthetic biology). The remaining challenges are immense, however, and solutions to such challenges are not foreordained.
The first major challenge is constructing a mirrored ribosome. This has never been done. Ribosomes are the large biomachines that all organisms on Earth use to build proteins based on genetic instructions transcribed from DNA. They are made up of dozens of individual proteins and RNA strands, each of which must coordinate to function correctly.
As a ribosome “reads” an RNA template to build proteins, it must interact seamlessly with dozens of other molecules floating around the cell. EF-Tu, for example, is a protein that helps the ribosome grab onto nearby amino acids so that it can paste them into a growing protein chain.12 Each component of the protein translation “machine” is like a tiny gear in a mechanical watch; if a single gear doesn’t mesh smoothly with the other gears, the hands stop moving. In living cells, ribosomes assemble from many different proteins and RNA molecules, in a complicated process sculpted and honed by evolution. Replicating this delicate assemblage in a test tube, using synthesized copies of mirrored molecules, is no easy task.
Then there’s the cost. For a 2019 paper, researchers synthesized an entire, natural-chirality bacterial genome, stretching about 4 million base pairs in length. The project required several million dollars just for the DNA; not including time, labor, and other reagents. A decade-long effort to chemically synthesize a baker’s yeast genome, which includes about 12 million bases of DNA, has already run into the tens of millions of dollars — and it’s still not finished. Mirror-image DNA is significantly more expensive to create, though costs would presumably fall if demand increases.
The most difficult challenge, though, will be putting everything together. Even if someone makes a mirrored genome and all the other parts required for protein synthesis,13 there’s no guarantee that they will “boot up” to create a viable organism. Making life from scratch is uncharted territory: this step alone could take a decade or more of troubleshooting.14
But scientific efforts continue to bear fruit.
A group from Stony Brook in New York made a poliovirus from a chemically synthesized genome back in 2002. In 2010, researchers at the J. Craig Venter Institute chemically synthesized a genome for Mycoplasma mycoides, a bacterial parasite, transplanting it into Mycoplasma cells from which their natural genome had been removed. Cells receiving the genome transplant went on living — a feat that JCVI scientists touted as the “first self-replicating, synthetic bacterial cell.”15 And finally, biochemists have shown that purified enzymes and nucleic acids — placed within a lipid bubble — could copy genes, transcribe them into RNA, and even assemble viable proteins. Their “artificial cell” functioned for a time, but was not self-sustaining.
All of these experiments used molecules with natural chiralities. But most of the difficulty inherent in creating mirrored life is contained in the challenge of making artificial life in the first place. Once artificial cells are “booted up” using natural-chirality molecules, the same methods could be applied to mirrored variants.
Infected World
The considerable uncertainty surrounding discussions on mirrored lifeforms — and what might happen should scientists actually create one — means we can only speculate about the details. However, there are some high-level reasons, analyzed across 300 pages of the technical report, to think mirrored life could do immense harm.
If a mirrored organism is created in a laboratory, there are some obvious ways to confine the cells and physically prevent their release into the wild: sealed containers, directed airflow systems, and personal protective equipment, like masks and gloves. Every credible scientific institution requires stringent safety protocols for microbiology and synthetic biology experiments deemed to pose potential hazards.
Alas, lab leaks are surprisingly common. According to reporting in The Guardian, “hundreds of safety breaches happen every year at labs experimenting with dangerous pathogens” — and the public rarely hears about them:
. . . when a safety breach occurred in 2019 at a University of Wisconsin-Madison lab experimenting with a dangerous and highly controversial lab-created H5N1 avian influenza virus, the university never told the public – or local and state public health officials. The university made the decision to end the quarantine of a potentially exposed lab worker without consulting Wisconsin public health officials.
In 2001, in an outbreak that had nothing to do with a lab leak, foot-and-mouth disease in the U.K. caused damages of around £8 billion, and over 6 million cows and sheep were slaughtered to halt its spread. Six years later, though, a lab researching the disease under the highest level of biosecurity leaked the same virus through a sewage pipe, spreading viral particles into the groundwater and causing another outbreak that killed about 2,000 animals.
Yet, published accounts of lab leaks are only those incidents that were detected because they caused an infection. It is hard to estimate the likelihood of preventing any leaked material whatsoever. Biohazard containment measures are improving, but perhaps not enough for comfort.16
Not many bacteria need to “leak” to cause an infection, either. The infectious dose of some pathogenic strains of Escherichia coli can be fewer than 100 and as low as ten individual cells; hundreds of times smaller than a grain of sand. (A culture of E. coli on a petri dish contains billions of cells). If bacteria do escape from the laboratory, then many of the cells will naturally be killed by predators, like bacteriophages and protists, present in the environment. But this may not be the case for mirrored bacteria, according to the report.
If a few mirrored bacteria find a favorable environment, they might not need a living host to persist or begin replicating, and would also face far fewer predators in the meantime. Potential hosts would also lack key immune defenses, which can normally be trusted to prevent infection.
Early versions of mirrored cells would likely be weak and fragile, unable to feed on many of the chiral molecules consumed by “normal” cells. For example, most glucose is unavailable to mirrored life because it almost exclusively exists in its “right-handed” form, although scientists could engineer mirrored cells to catabolize such molecules (perhaps to facilitate growth in the laboratory).17 Mirrored cells would still be able to feed on achiral carbon sources, such as glycerol, and harvest nitrogen from ammonia, nitrate, and achiral organic compounds, including glycine.
Even if mirrored cells grew slower than their natural counterparts, it wouldn’t negate the potential risk, for they would die even more slowly. Mirrored lifeforms would be immune to phages and undetectable to protists, which use chiral molecules to hunt. If a single mirrored cell grew in number at a meager rate of 10 percent per day, and did not have nutrient limitations, it could create more than 1031 descendants in just two years — outnumbering all the bacteria currently on Earth.
The report argues that if a small number of mirrored cells were to invade the body via the lungs, small wounds, or the gut lining, they could grow uncontrollably and potentially cause fatal consequences.
Under normal circumstances, the “innate” immune system is first to react to bacterial invaders, preventing stray cells from proliferating into a serious infection. Cells in the innate immune system, such as macrophages (a type of white blood cell), hunt down pathogens by grabbing onto microbe-associated molecular patterns, or MAMPs;18 a suite of molecules found in many different microbes, such as the sugars present in cell walls or the proteins used to build bacterial flagella. MAMPs are a bit like keys that open padlocks to immunity. But alas, almost all known bacterial MAMPs are chiral:19 the reflected keys don’t open the same padlocks, and so mirrored cells may not be recognized by the innate immune system.20
This is bad news because the innate and adaptive immune systems work together. The first-line defense cells, including macrophages, normally present pieces of the pathogens they’ve killed to naïve T cells, thus activating them. Those T cells then stimulate B cells to manufacture antibodies specific to a pathogen. These antibodies bind to the bacteria, marking them for destruction. Antibodies also act as a defensive memory; they hang around the body for decades and can be used to respond more swiftly to similar infections in the future.
But again, there is a problem. To trigger adaptive immunity, protease enzymes break down bacterial proteins into short “peptide” chunks, and those peptides are presented to T cells via “major histocompatibility complex” (MHC) molecules. But mirrored proteins would resist being broken down because those enzymes are sensitive to chirality.
If mirrored cells go undetected by the first-line defenders — which rely upon chiral interactions — then it’s unlikely the body would make antibodies that are protective against the invaders.21
There is experimental evidence supporting this claim. Mice lacking pattern recognition receptors — proteins that recognize and bind to MAMPs — quickly succumb to bacterial infections that healthy mice can clear. Mice are also far less likely to survive routine infections if only a single type of pattern recognition receptor becomes defective. When mice are injected with mirrored proteins, they also do not produce detectable antibody molecules. They do make antibodies, however, when injected with non-mirrored versions of those same proteins.
There is also a rare genetic disorder that afflicts somewhere between 100 and 200 people (mainly in North Africa), in which the MHC-II proteins used to present peptides to T cells are mutated and defective — a vulnerability mimicking what might be expected for infections with mirrored bacteria. People with such deficiencies live their entire lives with acute and chronic infections, despite receiving prophylactic antibiotics.22 Survival is unlikely beyond age 10. Similarly, about 40 percent of people with a deficiency in either IRAK-4 or MyD88 — key signaling molecules just downstream of the pattern recognition receptors — die before reaching adulthood, despite receiving vaccines, antibiotics, and advanced medical treatment.
Evading the immune system, a mirrored microbe could eventually find its way into the bloodstream. It wouldn’t need any specialized equipment or chiral gymnastics to do so. Evidence suggests that microbes enter the blood of healthy people all the time, regardless of whether or not they are pathogenic.
In one study, 40 percent of people had bacteria enter their bloodstream after flossing their teeth (and 24 percent after brushing); but the cells were quickly cleared by the immune system. For a 2023 study, researchers sequenced circulating DNA from blood samples taken from nearly 10,000 people and found evidence of bacteria in 11 percent of them.
A few stray bacteria in our bloodstream are normally no big deal. But if mirrored bacteria divide more swiftly than they are killed, it’s difficult to see how an unlucky host could survive. Mirrored cells not adapted to catabolize glucose would have plentiful access to glycerol and other achiral carbon sources in the bloodstream, though the availability of iron might be a slowing factor. It’s hard to know how the infection would play out exactly, but it wouldn’t end well. Sepsis is one possibility, where the body launches fruitless but increasingly violent immune campaigns against an invading pathogen, ultimately undoing itself through organ failure or fatally low blood pressure.
Even if a patient were rushed to a sterile facility, there is no preventing every last bacterial cell from spreading to the environment or to other people. More widespread infections — between other humans, but also animals — would follow.
Not all pathogens spread quickly. The rabies virus infects most mammals and typically spreads through saliva (delivered via animal bites). A study of fox rabies in Europe found that the disease spreads at a rate of roughly 50 kilometers per year. SARS-CoV-2, on the other hand, was carried by humans, and it spread to every inhabited continent within four months — thousands of kilometers per month.
It is likely that mirrored bacteria would travel at the speed of human movement. Almost all civilization is tightly connected by transport networks that exchange goods across the world in hours or days. Lacking natural predators, mirrored bacteria could hitch a ride on transported animals, plants, or products made from them, on surfaces, or perhaps via inadvertent hitchhiking insects — all before reinfecting host organisms.
If humans are vulnerable to infection by mirrored bacteria, then it’s likely that other multicellular lifeforms would be, too.23 Many mammals, birds, fish, and other vertebrates, as well as invertebrates like insects (and plant life), largely rely on “lock and key” models for recognizing pathogens. None of these organisms have locks adapted for mirrored keys. When fruit flies lacking “lock and key recognition” mechanisms are infected with benign E. coli cells, for example, the insects usually die.
Insects occupy the bottom of food chains, and the disappearance of some species would have knock-on effects for every animal that eats them. And because they break down waste and recycle soil, even those creatures who are not directly reliant on eating insects would have their feedstocks and habitats severely disrupted. American naturalist E.O. Wilson once said: “If mankind were to disappear, the world would regenerate back to the rich state of equilibrium that existed ten thousand years ago. [But] if insects were to vanish, the environment would collapse into chaos.”
In the 1990s, ecologists began to observe population declines in amphibian species, which they tied to Batrachochytrium dendrobatidis (Bd), a pathogenic fungus. Global trade introduced the fungus to species without natural immunity and, since then, it has driven at least 90 species extinct and collapsed populations of hundreds more. But Bd, like other natural pathogens, is specialized to infect a specific family of hosts. Species extinctions caused by invasive pathogens are circumscribed. With mirrored bacteria, that “firewall” would not exist.
Plants, too, may be susceptible to infections by mirrored bacteria. Plant immune defenses are not well studied, though, and most experiments focus only on a handful of model species, such as Arabidopsis thaliana (mouse-ear cress). Nobody knows which plant species would be hardest hit, and this section of the technical report is highly speculative. Whether (and which) staple crops would be infected remains an open question.
Then there is the problem of predation. Phages, protists, and other organisms control bacteria populations in the environment, but these predators also rely on chiral mechanisms. Thus, mirrored bacteria could persist a long time outside of hosts — such as in oceans, rivers, and soils — and grow where nutrients are rich.
As mirrored microbes spread, they would pick up new mutations. During every cell division, there is a roughly one-in-ten-billion chance for each nucleotide to change into another. And because there are trillions of bacteria inside every human, every possible point mutation, at each nucleotide in the genome, could occur among the bacteria inside the first infected person or other large animal. These mutations could help mirrored cells become better at hunting down nutrients or eating an expanded repertoire of foods.
As a result, would mirrored bacteria drive every animal species to extinction? Unlikely.
While virtually all animal immune systems share common mechanisms, there is still room for great diversity. For example, while most animals use pattern-recognition receptors to detect pathogens, they vary in the molecules they have adapted to recognize. By luck, some species may be “pre-adapted” to also recognize mirror bacteria. Physiological differences could matter, too; the almost freezing body temperatures of deep ocean and polar fish hamstring the growth of mesophilic bacteria — that is, microbes adapted to grow in moderate temperatures — and extreme cold would have the same effect on mirrored forms. It’s impossible to predict the full range of responses, but some animal species might be much less susceptible than others.
Those mirror-resistant species could repopulate their environments — but only in the aftermath of a major collapse in biodiversity, perhaps wiping out some ecological niches or kinds of animals within them, entirely. Following the Permian–Triassic extinction event, the Earth’s most severe extinction event, it took biodiversity tens of millions of years to recover to pre-extinction levels. And the resulting Earth ecosystem is not guaranteed to resemble life as we know it: if a rainforest burns down, all memory of its ecology can burn down with it.
If mirrored bacteria were created, we don’t know exactly what would happen next. It will depend on the kind of bacteria released, variation in species response, and so on. Despite these and open questions, leading scientists, most of whom were initially skeptical, agree that the containment of mirrored bacteria would require a level of vigilance heretofore unseen.
Creating Countermeasures
Vaccine development can be fast. The first SARS-CoV-2 genome sequence was published on January 10, 2020. Moderna had designed an mRNA vaccine three days later and manufactured their first batch by February 7th. Those shots were shipped and ready for Phase I clinical trials on February 24, a total of 45 days after the SARS-CoV-2 sequence was initially uploaded online.
It’s possible that we could move just as swiftly against mirrored cells. The report maps out at least three options.
The most obvious defense would not be vaccines, but rather antibiotics; small molecules that treat or prevent bacterial infections. Most antibiotics work by binding to, and shutting down, key parts of the bacterial lifecycle that are absent in human cells.
A few antibiotics are either “racemic” (containing both enantiomers) or achiral, and could therefore be manufactured immediately at the start of an outbreak to provide protection against mirrored bacteria. This includes the commonly used and inexpensive quinolones, such as ciprofloxacin.
However, most widely used antibiotics, including penicillin, work through chiral mechanisms and will not work against mirrored microbes. This has been confirmed experimentally for enantiomer versions of chloramphenicol, cephalosporins, penicillins, and tetracyclines, which all fail to kill natural microbes. It is possible, however, that these enantiomers could target and kill mirrored microbes. Mirrored antibiotics would also interact differently with chiral molecules within the human body; they could well be toxic.
In the event of an outbreak, it’d be easiest to scale manufacturing for antibiotics that are made using chemical synthesis — rather than those harvested from living organisms — because the chemical reactions could simply be repeated using enantiomeric forms of each molecule. Carbapenems, monobactams, oxazolidinones, and several other classes of antibiotics are all made in this way.
Even if antibiotics were rapidly produced and administered to most of the world’s population, they might still have limited efficacy. Prophylactic use of antibiotics often fails to prevent infection in immunodeficient patients against natural chirality bacteria, and, plausibly, everyone would be similarly immune deficient against mirrored bacteria. For these reasons, antibiotics alone are unlikely to eliminate dangers from a mirrored bacteria outbreak.
The second option is to create anti-mirrored vaccines. Like other vaccines, these would train the human immune system to recognize and destroy mirrored pathogens. There are already vaccines against bacteria, including those responsible for diphtheria, tetanus, meningitis, and typhoid. But all existing bacterial vaccines appear to elicit immunity only against chiral antigens, where immunity is very unlikely to extend to the mirrored version. So a new vaccine, specific to the mirrored bacteria, would have to be developed.
But just as human immune mechanisms would likely fail to process the mirrored protein antigens from mirrored bacteria, vaccines that act by presenting mirrored antigens might similarly fail. Most bacterial vaccines rely on chiral protein antigens, either directly or indirectly, and so most classes of vaccines would be unable to recognize their target antigen and be ineffectual as a result.
The best option is to make a “conjugate” vaccine by fusing a mirrored antigen (a piece of the external wall of the mirrored organism, for example) to a non-mirrored “carrier protein.” The most comparable vaccines today are made through a delicate and slow process, with standard development timelines spanning 5-10 years. COVID-19 showed that vaccine timelines can be accelerated during a crisis, but majority vaccine coverage was only reached globally after most people in the world had been infected at least once. It is also more difficult to make conjugate vaccines than it is to make the mRNA vaccines used for COVID-19, and so timelines could be even slower.
A third countermeasure against infections is to design a predator for mirrored bacteria.
By number, the most abundant entity in the biological world is the bacteriophage. For every grain of sand in the world, there are roughly one trillion phages. Phages are viruses that infect bacteria and sometimes archaea, binding to receptors on a bacterium’s surface, transferring their genetic material into the cell, and hijacking the cell’s machinery to assemble more copies of the virus. They typically cause a form of cell death called “lysis,” where the cell wall and membrane rupture, allowing phages to spread to new bacteria.
Mirrored phages could be released wherever mirrored bacteria are proliferating, and they would spread automatically as they reproduce. Since they don't target or replicate within human cells, phages are not pathogenic to humans. They could help control the rate of spread and the abundance of bacteria in a given environment.24
But this plan would not solve the problem. In the same way that many bacteria evolve to outwit phages, the mirrored bacteria might evolve through mutations to evade this virus, leading to an arms race of evasion and speciation. More importantly, there is no plausible scenario where the phages “win” by entirely eliminating all of the mirrored bacteria: mirrored phages depend on mirrored bacteria to survive and replicate — and an ongoing risk of infection would persist so long as some mirror bacteria were present. Other possible “mirrored predators” (like certain protists) would face similar limitations.
While countermeasures such as these could help slow and treat infections for humans and animals, it is entirely unclear how to reverse a mirrored bacteria outbreak, in the same way that humanity already expects cyclical pandemics by virtue of how bacteria and viruses evolve. Antibiotics and vaccines might work for humans and selected animals, but the same countermeasures wouldn’t prevent plant infections, halt environmental contamination, or curb ecological damage. Wild ecosystems, the report says, would likely be impossible to protect.
Once extant, mirrored bacteria cannot be put back into the petri dish.
The Arc of Progress
There have been a few moments in the history of biology where scientists peered down an avenue of research, didn’t like what they saw, and agreed to go no further. If we don’t want to sleepwalk down the path to creating mirrored life, then history is surprisingly encouraging.
In 1996, lan Wilmut and colleagues announced the birth of Dolly, the first mammal cloned from an adult somatic cell. Once the stuff of sci-fi — like the Bokanovsky Process from Aldous Huxley’s Brave New World — human cloning was now feasible. But a strong ethical norm quickly spread among practicing scientists and the public. Federal funding would not support it and, in some cases, national laws were passed.25 With remarkable unanimity, the world agreed not to go there.
Similar cases have played out in fields outside biology. In the 1970s, scientists began noticing previously stable levels of ozone above Antarctica had begun to decline dramatically. It was also known that “chlorofluorocarbons” (CFCs) and other related substances were disproportionately damaging to the ozone layer. After years of debate, an international pact limiting emissions of ozone-depleting substances was signed in 1987. By the year 2000, emissions had fallen to around 10 percent of their peak, with one estimate suggesting this single pact has prevented more than a million cases of skin cancer per year. In 2013, it became the only international treaty to have been ratified by every country on the planet.
Of course, this kind of protective restriction is not always justified. There are plenty of cases where ethically motivated review processes end up throwing sand in the gears of worthwhile science, or even ground it to a halt — a real and pervasive problem. But consider the international prohibition on experimenting with live smallpox. Research institutions did not accidentally accumulate too many administrative hurdles to acquiring smallpox samples. Rather, the world made a clear-eyed and collective decision about refusing to experiment with smallpox. The severity and stakes of such research persuaded experts that it should fall outside of any kind of case-by-case review based on costs and benefits. It was, rather, a unilateral decision: just don’t go there.
If the broader scientific community agrees with the co-authors of this Science report, then one could imagine serious conversations involving scientists, policymakers, and the general public, working out policies or even international agreements with broad buy-in — similar to the Asilomar conference of 1975, organized in response to developments in recombinant DNA technology.
Public funders could declare intentions not to fund mirrored life and even withdraw general funding from institutions that receive funding for mirrored life projects from other sources. Perhaps legislation could get passed, banning mirror-image nucleic acids exceeding (say) 1,000 bases and proteins exceeding a certain length — shorter than needed for the full genome and proteome.
Ultimately, the question we should be asking is who gets to decide how (and whether) to proceed with research on the path to mirrored life. The co-authors of the Science article say “not us”: a small group of scientists shouldn’t be trusted to design major governance initiatives, any more than technologists alone should be trusted to design regulations on tech companies.
Instead, the authors call for more scientists to evaluate the risks, and to begin serious and inclusive conversations with policymakers and the public around how the world should handle those risks. They make a good case that “creating mirror life” should be added to the list of experiments where, through painful experience or foresight, the world has agreed to just not go there.
If we abide by this and ban mirrored life, what might we lose? Might we come to regret it? The first of these questions drives at whether mirrored life is worth creating. And indeed, a full mirrored cell could be used to manufacture mirrored peptides, which could resist degradation by enzymes in the body and lead to longer-lasting medicines.26 But the price of admission is to introduce a potentially unstoppable threat that, if mishandled, could lead to major loss of life and the collapse of ecosystems.
So just as we forego some theoretically interesting findings on smallpox research in the interest of avoiding a smallpox outbreak, we would lose out on some potentially interesting science and pharmaceuticals. But to stake nature and human life as we know it for scientific headlines and improved drug manufacturing methods seems immoral and improvident. Carefully weighing costs and benefits would be like debating the entertainment value of Russian roulette.
The second question asks us to consider whether initiatives to effectively ban mirrored life might do harm by blocking other useful research as “bycatch.” To build mirrored life, we’ll need to assemble complex mirrored proteins and long RNA and DNA sequences. Relatively short chains of mirrored DNA (called aptamers) might be functionally useful, but it’s hard to see any other practical use for synthesizing polynucleotides of mirrored DNA more than a few thousand base pairs long. It’s a similar story with mirrored RNA and mirrored proteins. If we never made certain mirrored molecules upstream of the self-replicating mirrored cell, it’s hard to see how that would block unrelated technologies.
It’s tempting to frame concerns around mirrored life in terms of a collision between safety and progress in the field of synthetic biology. But if we can just block a few critical precursors to mirrored life — diverting research efforts away from one back to hundreds of other streams — we could seal off the dangers without damming the overall flow of new discoveries in the field.
Excessive rules-based “safetyism” in other fields has sometimes proven to be self-defeating: an overly conservative approval process for a life-saving drug means more lives lost before it’s approved; steep regulatory burdens to installing nuclear power capacity contributes to more deaths from pollution. In such areas, we can learn from accidents and then iterate our way toward safety. But when it comes to building mirrored life, we shouldn’t plan to learn from experience. An outbreak could be uniquely bad; both irreversible and crippling to human progress. So here the outcome is upended: safety-motivated guardrails could protect human and scientific progress writ large.
In a sense, we got lucky. To prevent mirrored life would not require blocking an entire field of science or hugely valuable technologies. But we often can’t predict the lay of the scientific landscape; which technologies are downstream of which discoveries. And that raises the question: what if we weren't so lucky? What if continued progress in an entire field required a discovery with potentially life-threatening applications? The scientific community faces an important choice here — whether and how to prevent mirrored life — but the choice could have turned out much harder.
There is a great deal of speculation surrounding mirrored life — whether it can be made at all, how long it will take, and how dangerous such organisms would really be. But speculation should not encourage delay or complacency. In the absence of total certainty, one does not always need to find out.
Niko McCarty is a founding editor of Asimov Press.
Fin Moorhouse is a researcher at Forethought, a new research group. He was supported by the Mirror Biology Dialogues Fund while researching this piece.
Cite: McCarty N. & Moorhouse F. “The Dangers of Mirrored Life.” Asimov Press. DOI: 10.62211/33re-48kj
Similar grants have been awarded by Chinese and European agencies. The National Natural Science Foundation of China, for example, supported work toward the creation of a mirror-image central dogma (National Natural Science Foundation of China grants 32050178 and 21925702), and the European Commission ERA-Net MirrorBio consortium, with similar goals, was supported by the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung grant 031A461).
The “long-term goal” of Kay’s lab at the University of Utah is to “synthesize a D-ribosome . . . as a key stepping stone towards ultimately creating a fully synthetic mirror-image organism”. Eric Kool, Professor of Chemistry at Stanford University, notes on his website that with $1 billion to spend over a decade, he would build a team to construct mirror life: “It would be fun, and likely useful as well, to build a living bacterial cell from the ground up, constructed from the mirror image of DNA/RNA/protein/carbohydrate. Actually I think this is likely to happen in the next ten or twenty years anyway without me – people are starting to make real progress in this direction. I’ll be excited to see it become real.”
Left-handed glucose (L-glucose) can be synthesized artificially, but only in small quantities and at large expense.
The only exception is glycine, which is “achiral.” There are also right-handed amino acids present in nature, but they are not built into proteins.
LUCA likely first adopted its chirality by historical accident. There are many proposals for the specific origin of homochirality (some even point beyond Earth) but the matter is still being debated.
One of the four fundamental forces, the weak force, violates chiral symmetry. But it is so weak in relevant contexts that it is difficult to see how it has any effect on biology. Its effects on molecular systems have never been observed experimentally.
Thalidomide is commonly used as an example of enantiomers having different effects. However, even if one were to eat pure enantiomers of a molecule, some will still racemize — or change into the other enantiomer — once inside the body. This is a chemical process; factors like pH and temperature in the body can spontaneously cause one enantiomer to convert into its mirror image. Different molecules racemize at different rates.
As a product of plants, sugar molecules are “right-handed” and will twist polarized light in the clockwise direction. This effect which can be observed with a polarizing filter.
Viruses would be simpler to create but relatively uninteresting. Because they rely on other organisms to reproduce, mirrored viruses would be incompatible with natural-chirality hosts. Absent other mirrored life, they would remain inert. Eukaryotes — the domain of organisms which include plants, animals, and fungi — pose more of a challenge because their cells are more complex.
The average bacterial protein contains 320 amino acids.
D-RNA oligos up to 70 nucleotides in length can also be made using solid-phase chemical synthesis.
tRNAs — which carry amino acids to the translating ribosome — are made from both RNA and an amino acid. Their “hybrid” chemistry may be difficult to create in a mirrored form.
As few as 36 proteins are required for protein synthesis in cells, in addition to tRNAs and chemical forms of energy, such as ATP molecules. Note that ribosomes here count as one protein, but they are actually made from 50 individual proteins and several strands of RNA.
Instead of building a cell from the bottom-up (or from individual molecules), the 300-page technical report claims that one could build a cell using more top-down methods, by replacing “normal” molecules with their “mirrored” versions one at a time. This would be incredibly difficult, however: in particular, it would be challenging to keep such cells alive for long enough to complete the swaps.
This claim is overstated. The authors transplanted one genome for another, without replacing or augmenting any of the other components. Many scientists do not consider this to be a truly synthetic life-form, which implies an organism that has been made entirely from abiotic molecules.
Mirror bacteria might also be controlled with biological containment measures. For example, scientists could intentionally hobble mirror bacteria by engineering them to depend on food molecules not present in nature; or they could even reengineer functionally essential proteins to require amino acids which are very scarce in nature.
Some existing bacteria are able to catabolize L-glucose, thus providing a proof of concept for the mirrored equivalent (a mirrored cell catabolizing D-glucose).
Sometimes “pathogen-associated molecular patterns” or “PAMPs.”
There are some exceptions, though such MAMPs may not be sufficient on their own to trigger sufficient immune responses.
Some elements of the innate immune system can also be triggered when receptors in human cells detect damage-associated molecular patterns (DAMPs). Because these involve patterns of damage in our own cells, mirrored bacteria activity could eventually trigger the same response. But DAMPs are unlikely to trigger an immune reaction until late in an infection and seem unlikely to help the immune system target and eliminate infecting mirrored bacteria.
Even if they could be broken down, and did bind properly to MHC molecules, it still remains unclear whether T cells would be activated because they need stimulation from innate cells, which would also be impaired during a mirrored bacterial infection.
Prophylactic antibiotics are given to prevent infections, rather than treat an existing one.
All vertebrates have innate and adaptive immune systems; not just humans. All of these systems would likely have similar defects, as they rely on “lock and key” models for recognizing pathogens.
Over time, however, a large portion of all dissolved, organic carbon would be the opposite chirality of normal biomolecules. This could eventually result in ecosystem disruption as carbon becomes inaccessible to non-mirror life.
Interestingly, there is still no comprehensive federal law banning human cloning in the United States.
The most practically useful applications of mirrored bacteria (such as in drug discovery) may require high-risk strains, however. Mirrored bacteria adapted to manufacture biomolecules at scale would need to grow robustly enough to compensate for the metabolic costs of biomolecule production, and may therefore be engineered to metabolize natural-chirality nutrients, such as D-glucose.
 | A guest post by
|
 | A guest post by
|
Excellent analysis of both the risks and potential benefits that may be best foregone. Though, I wonder if we could use synthetic and cell free methods to still synthesize mirror biomolecules. A mirrored ribosome doesn’t have to exist only in a mirrored cell. Could that allow us to access some of those mirrored peptide medicines and other potentially more robust mirror biomaterials, without the risk of uncontrolled spread?
Wow, awesome analysis! You bring in so much more context and it is all fascinating and mind boggling. One possibility I haven't heard a lot of discussion on is building hybrid cells that contain both mirrored versions of the central dogma. A sort of partitioned life I guess. The ethics seem better since youre not abandoning the old chiral forms. But they might be able to pump out mirror enzymes faster than solid phase synthesis.