Earlier this year, experts and law enforcement agencies worried about the possibility of cheap, home-brew heroin when two teams of scientists reported that they had created genetically engineered yeast strains that could almost make morphine from sugar. With morphine-making yeast, a would-be Walter White could brew heroin in an Albuquerque basement without relying on opioids extracted from poppies grown halfway across the world. The new genetically engineered yeast couldn’t quite go all the way from sugar to morphine, but it was clearly only a matter of time before someone made a strain that could.
Three months later, a third team of scientists, led by synthetic biologist Christina Smolke at Stanford University, finished the job. In a paper out last month, the scientists describe two genetically engineered yeast strains that metabolize sugar into opioids. While the risk of home-brewed heroin posed by these yeast strains is small—the opioid yield is very low, and, instead of morphine, the yeast make opioids that are harder to process into heroin—the implications of this work are large. These genetically engineered microbes are part of a major technical advance that could change how we make, distribute, price, and regulate drugs.
With morphine-making yeast, a would-be Walter White could brew heroin in an Albuquerque basement without relying on opioids extracted from poppies grown halfway across the world.
Though the new yeast strains are an important development, the use of genetically engineered microbes to make drugs is not new. Back in 1982, Eli Lilly began selling recombinant human insulin, the first drug produced by genetically modified bacteria. Before then, diabetics relied on animal insulin obtained from pork or beef pancreases, which were supplied in limited quantities by the meat industry. Making insulin from genetically engineered microbes has been a major success: It's now a $10 billion industry that manufactures substantially improved forms of the drug and in quantities sufficient for meeting growing demand.
But the genetic engineering methods used to manufacture recombinant human insulin are limited to proteins, such as insulin, antibodies, and growth factors. These drugs are directly encoded by single genes and that makes the genetic engineering straightforward. To make bacteria produce human insulin, for example, you simply transfer the human gene for insulin into bacteria. In practice, additional optimization tweaks are typically needed to get bacteria to yield large quantities of the drug in a usable form, but the fundamental process is easy.
Most drugs are not like insulin, however—they are not proteins directly encoded by genes. Instead, many drugs are small, often complex molecules that are made by a series of distinct chemical reactions, meaning that they can’t be produced by simply transferring a single gene into bacteria. Currently, the chemical reactions necessary for yielding these drugs occur either in industrial-scale chemical reactors, or they are carried out by the metabolic machinery of naturally occurring organisms, primarily plants. In other words, to make most of our drugs, we extract them from plants, synthesize them in chemical factories, or do a little of both. While successful in many cases, making drugs this way has some major disadvantages. Supplies of plant-based raw materials are limited by slow annual growing seasons and are vulnerable to shifts in climate that impact harvest yields. And chemical syntheses are limited to what's economically and chemically feasible on an industrial scale. Genetically engineered, drug-secreting microbes don't suffer from these disadvantages: They can easily be grown in very large quantities, year-round, using cheap and readily available raw materials. By supplementing or even replacing traditional manufacturing methods with genetically engineered microbes, we could ensure a more reliable and possibly cheaper supply of drugs.
The challenge is that the genetic engineering required to make complex, small- molecule drugs is much more extensive than the methods used to produce a protein drug like insulin. To create yeast or bacteria that make these complex molecules, it's necessary to extensively re-engineer the organism’s metabolic machinery, essentially turning it into a miniature chemical factory capable of carrying out a long sequence of specifically designed chemical reactions. This requires modifying many genes in a single strain, which has become technologically feasible only recently. In 2006, researchers at the University of California–Berkeley—led by synthetic biologist Jay Keasling—reported that they had created a genetically engineered yeast strain that produced artemisinic acid, which is normally extracted from the sweet wormwood plant to make anti-malarial drugs. To make a yeast strain that carried out the proper chemical reactions, Keasling and his team transferred the genes of four wormwood metabolic enzymes into yeast, and also modified several other native yeast genes. The pharmaceutical company Sanofi is now using genetically engineered yeast to manufacture artemisinic acid on an industrial scale.
Engineering opioid-producing yeast was an even greater challenge. Smolke's Stanford team engineered 23 different genes into a single yeast strain. They took genes from a variety of sources: Genes from rat, bacteria, and three different species of poppy were transferred into yeast to carry out all of the chemical steps necessary to metabolize sugar into opioids. The result is a major technical achievement, but the job isn't finished—the opioid yield of this yeast strain is too small to work on an industrial scale. To be commercially viable, the scientists estimate that "over a 100,000-fold improvement ... would be required for yeast-based production of opioids to be a feasible alternative to poppy farming for commercial production." However, such a scale-up is less of a challenge than it sounds: With a successfully engineered strain now in hand, the researchers expect major improvements in yield over the next several years.
The low yield of the current yeast strains means that, for the time being, home-brew heroin isn't a threat. The researchers point out that "a single dose of hydrocodone, as used in Vicodin® (5 mg), would require thousands of liters of fermentation broth, which no home brewer would reasonably pursue." However, more efficient yeast strains will certainly come along, and, as Kate Wheeling reported for Pacific Standard in May, biotech policy experts are arguing that governments and drug enforcement agencies need to start thinking about how to deal with the challenge these strains will present.
Home-brew heroin grabs headlines, but DIY illegal drugs are only one of the potential consequences of a new generation of pharmaceutical-producing microbes. One major consequence is that, while developing these microbes is still hard (the Stanford researchers spent more than 10 years developing their strains), distributing and culturing them is easy. Growing large batches of bacteria and yeast is cheap, fast, and insulated from the large fluctuations in costs associated with bad growing seasons for poppies and wormwood. Additionally, these microbial chemical factories can be copied by putting a few drops of cell culture on a confetti-sized piece of filter paper. The ease of copying is clearly a challenge to intellectual property protection, but it also presents an opportunity to develop and distribute open-source strains—designed in academic or government labs—that could be used to manufacture inexpensive generics in countries unable to afford brand-name drugs. Yeast-produced opioids from open-source strains could fill a major need: One recent study estimated that "5.5 billion people ... live in countries with low to nonexistent access" to effective pain medication. Though home-brewed illegal drugs will some day be a genuine risk, this new generation of drug-producing microbes also offers opportunities for making important drugs affordable and available worldwide.
Inside the Lab explores the promise and hype of genetics research and advancements in medicine.