Metabolic engineering is the science (and art) of engineering living cells, usually bacteria or single-celled yeasts, to produce valuable molecules that can then be extracted, purified, and sold. Engineered microbes are already used to make familiar compounds like ethanol and acetone, and also more exotic molecules like 3‐hydroxypropionic acid and squalene.¹ The list of chemicals produced in research labs at small scales is much longer.² The market for microbial products is currently estimated to be around $200 billion per year.³

But forget, for just a moment, about engineering a microbe to produce a new compound. Consider, instead, a (hypothetical) future in which it is possible to isolate and sell all the molecules that E. coli already produces naturally. Imagine if there were a technology that made it simple — and inexpensive — to pulverize a bacterium, smear it out, collect each molecule in its own tube,⁴ and sell these vials. How much would E. coli be worth?

The answer, according to my own estimates, is that the raw metabolites and macromolecules isolated from one liter of E. coli cells would be worth more than $600,000. This is more of a thought experiment than a serious economic analysis, but it suggests that we are collectively undervaluing the power and “technological” sophistication of even the smallest and simplest organisms.

To be clear, “fractionate E. coli and sell off its parts” is not a serious business proposal. Extracting even a single molecule from a microbial broth at high purity can be difficult, finicky, and expensive. And extracting all of the molecules, individually and at a high degree of purity, far exceeds today’s technical expertise. Nevertheless, quantifying the “parts price” of a microbe is a valuable way to understand the power and complexity of the “cellular factories” we use to make our medicines, fuels, and foods.

{{signup}}

The Method

Bacteria are neither expensive nor difficult to grow. A 1-liter culture of E. coli bacteria can be made in a single night by mixing 25 g of Luria broth base ($2-6) with 1 liter of water, preferably deionized (salt-free). This liquid broth is put in an autoclave, or high-temperature pressure-cooker, and then transferred to a shake flask. A small amount of E. coli is dropped into the sterile liquid, and then the whole thing is placed in a shaking incubator at 37° C.

After 8 hours or so, the culture is “spun down” in a centrifuge to yield around one gram (about half a thumb’s worth) of waxy sludge. That’s the bacteria! The equipment required for all this is inexpensive and readily available. The whole procedure requires perhaps an hour of labor ($10-30).

The actual mass of bacteria yielded from a saturated overnight culture varies. Biologists typically gauge culture growth using “optical density” (OD), which measures how much light of a defined wavelength (typically 600 nm) is scattered when passing through the media. E. coli cells in LB media start to slow their growth somewhere between 0.5 and 1 OD. For the sake of this exercise, I’ve assumed a harvest density of exactly 1 OD, which, for a liter of media, is about one trillion E. coli cells.⁵ One trillion cells is roughly equivalent to a 1 g pellet, representing 300 mg of dry weight.⁶

With that baseline in hand, the rest of the calculation just involved translating biology into numbers. I used many different sources to make my estimates, most notably the book Cell Biology by the Numbers, particularly its sections on the macromolecular composition of a cell and the concentration of metabolites in a cell. All of the sources and calculations used to write this essay are available for download.

Metabolites

First, I looked at E. coli’s metabolites and small molecules — nucleotides, amino acids, simple sugars (maltose, glucose, hexose … ), their precursors and derivatives, and all the other small organic molecules contributing to a bacterium’s core metabolic loops. I didn’t include elemental ions or metals, since these are not made by the bacteria themselves.⁷

The difficulty with cataloguing metabolites is that there are a lot of them. In an ideal world, one could randomly sample metabolites, perhaps by selecting a few particularly abundant or pricey ones, and extrapolate from those. In practice, however, we can use the data presented in this 2009 study, which measures the concentrations of about one hundred common metabolites in E. coli.⁸ Whereas most metabolomics studies only measure relative concentrations of their targets, this particular paper used carefully applied quantitative standards to arrive at absolute concentrations.

Obtaining prices for individual metabolites was tricky. Most chemical compounds are sold on the market at wildly different purities, each with a different value.⁹ Many chemicals can be bought in different forms or counterbalanced with different ions.¹⁰ Prices vary wildly between vendors — sometimes by orders of magnitude — because of bulk discounts, small market sizes, and volatile supply.

Ultimately, I relied on Sigma-Aldrich prices,¹¹ taking the cheapest per-gram option I could find. I ignored purity and excluded odd or high-value counterions — sodium, potassium, free acids, and magnesium stayed in; lead, gold, ammonia, or carbon-based salts were out. Readers could certainly challenge my low-purity assumption; after all, we’re already relying on near-magical levels of purification for this thought experiment, so why not go all the way and assume we can extract them with near-perfect purity? But still, I wanted to get a reasonably stable, worst-case baseline.

Given these assumptions, the economic value of metabolites isolated from one liter of E. coli is a paltry $30-40.

‍

Surprisingly, the total parts price of metabolites is dominated by a handful of expensive molecules. Just three molecules (6-phosphogluconate, phosphoribosyl pyrophosphate, and succinyl-CoA) account for about half of the total economic value. All of the highest-value molecules are also fairly high in abundance and quite valuable on a per-molecule basis. At the end of the day, though, bacterial metabolites simply aren’t worth that much.

Bulk Macromolecules

E. coli’s bulk macromolecules are, at least on paper, far more valuable than its small metabolites. These include cell membrane lipids, the carbohydrates coating the cell’s exterior, and glycogen molecules, which are used for long-term energy storage; anything that is a more complex arrangement of molecules. These molecules come in a wide range of compositions, and they aren’t generally sold sourced from E. coli, so I’ve used prices for comparable macromolecules from the other species noted below. I’ve also excluded one major cell membrane component, phosphatidylserine, because I was unable to find any suitable price listings.

‍

Proteins

Cell membranes and cell walls are neither the most abundant nor the most complex cellular macromolecules. That honor goes to proteins, the molecular machines that do much of the work of maintaining and growing a cell.

Proteins are hundreds of times larger than simple metabolites; compared to cell wall and membrane macromolecules, they are also made from a greater diversity of subunits and constructed far more specifically. Together, the proteins in an E. coli also weigh more than all other types of molecules, including lipids, metabolites, and carbohydrates, combined. Therefore, from the beginning of this thought experiment, I suspected most of the economic value of E. coli would come from proteins.

Unfortunately, E. coli has around 4,300 distinct protein-coding genes, each encoding a unique protein. As with metabolites, that’s too long a parts pricelist to cover comprehensively here. Therefore, I decided to focus on commonly-used proteins, abundant proteins, and some random selections.

In the first scenario, I thought most of the dollar value of the E. coli proteome might come from the proteins that scientists already use in large quantities. After all, all things being equal, high demand drives higher prices. I picked recA,¹² alkaline phosphatase,¹³ and the exonuclease V complex¹⁴ as representative highly-used proteins, based on a combination of personal familiarity and my ability to find prices and in vivo concentrations for each. Secondly, I looked at a few of E. coli’s most abundant proteins. Specifically, I used the two main ribosomal proteins, as well as the five most abundant proteins by copy number.¹⁵ And finally, I randomly sampled proteins from the E. coli genome, selecting ten proteins for which I could find both copy number and price information.

The outer membrane protein Lpp — which is the most abundant protein in E. coli and lends the cell membrane its structural rigidity — is an interesting outlier, worth almost $50,000 per liter on its own. The per-gram cost of the other proteins, however, doesn’t vary all that much. Most proteins cost between $1 million and $3 million per gram.¹⁶ This lets us extrapolate a total parts price for the entire E. coli proteome without having to price out all 4,300 individual proteins; assuming an average price of $3 million/g, a liter of E. coli contains almost exactly half a million dollars worth of assorted proteins.

‍

Nucleic Acids

The last major class of E. coli parts to price out are the nucleic acids: DNA and RNA.

Large DNA with a relatively arbitrary but fixed sequence isn’t nearly as valuable as, say, large DNA synthesized to order with a desired sequence, but it does have its uses. Bulk bacteriophage DNA is commonly used as a sequencing control and, a bit less frequently, as a backbone for DNA origami structures. Using New England Biolabs’ price for M13 phage DNA as a guide, I estimate that a liter of E. coli genomes is worth about $21,500. This number assumes that an E. coligenome is just as useful per gram as a phage genome, even though it is about a thousand times larger. It’s unclear whether that should make it more useful¹⁷ or less useful.

RNA is simpler to price out, even though it comes in three major forms with vastly different shapes and functions (messenger RNA transcripts, transfer RNAs, and ribosomal RNAs). Thermo Fisher Sciences already sells total E. coli RNA at a current price of $423 for 200 micrograms, which implies a parts price of about $127,000 per liter. This is not as much as the entire proteome’s economic value, but it is more than twice the value of the six most abundant proteins in E. coli. This bulk price is also a lower bound figure; it’s entirely possible that the RNA pool could sell for even more as individual, purified RNAs.

Is E. coli a Money Printer?

At first glance, it seems there’s a great business opportunity in cannibalizing E. coli for parts. At $627,000/liter, a modest biology lab could easily grow millions of dollars’ worth of E. coli overnight, every night. Surely someone could make a killing that way?

Unfortunately not. Fractionating an E. coli cell — that is, blending up a cell and isolating its parts — is much more expensive than growing it! There are already scientific kits available on the market that can be used for extracting individual proteins or nucleic acids for tens of dollars per purification. But those kits also waste, or discard, all other cellular components. Getting all of the individual proteins (where most of the economic value sits) out of a bacterium at high efficiency and purity would require new technology.

RNAs are a better proposition, as there are relatively cheap kits for extracting total RNA from a variety of biological samples. But then you’d be directly competing with Thermo Fisher on a functionally identical product! Does the high price on Thermo Fisher’s total RNA mean that they’re making insane margins that a competitor could cut in on? Or does it mean that the extraction of bacterial RNA at scale is more expensive than it looks at a glance?

The second problem with selling E. coli parts is a lack of demand. The proteins expressed in a liter of cells may be worth $500,000 on paper, but unfortunately, there isn’t much of a market for most of them.

Big lab suppliers like Thermo Fisher or Sigma Aldrich don’t bother selling obscure proteins such as “sapF” to scientists because hardly anyone needs them. There’s no real market. To find prices for a large number of proteins, I had to look at smaller vendors like MyBioSource, which will sell you just about any protein in tiny amounts. But in those cases, you’re really paying for the service of custom purification, not for the protein itself. The price reflects the hassle of isolating it in small batches, not steady demand. If someone actually tried to mass-produce every E. coli protein this way, they’d run out of customers very quickly.

Put differently, that $627,000 per liter is only a notional value — what today’s customers might pay for one more liter of E. coli, given current catalog prices. The figure is high because, somewhere out there, a few biologists need small amounts of obscure proteins for unusual experiments. Once those needs are met, though, the extra liters wouldn’t fetch nearly as much. In the end, breaking down E. coli for parts would run into the same problem as asteroid mining: the market isn’t nearly as big as the sticker prices make it look.

Based on our parts price stories, it seems the value of E. coli-derived molecules has a lot less to do with producing individual molecules and more to do with linking those molecules together in intricate, precise ways. Monomers and high-energy metabolites are barely worth more than the broth that E. coli feed on. Crude, bulk polymers of sugars and lipids are much more valuable. But the real economic value of living machines seems to be in their ability to produce high-complexity, highly-specific protein (and to a lesser extent, RNA) polymers.

Transmuting Trash

As a synthetic biologist in graduate school, I used to grow a lot of E. coli. Sometimes I grew them as experimental subjects. More often, though, I used them as factories to pump out plasmids destined for use elsewhere.¹⁸ A typical day in the lab might start by spinning down a couple dozen tubes of 5 mL cultures of overnight E. coli growth, each with a different plasmid, yielding pale yellow, gummy smears or plugs of bacteria that I would break open and carefully filter for the plasmids they carried. At the end of that day, after transforming a new batch of bacteria with new sets of plasmids, the bacteria would be seeded into 5 mL tubes of fresh media, which would go into a shaking incubator to grow overnight for harvesting the next morning.

The work made it easy to develop a distaste for E. coli. I was always focused on extracting the precious plasmids they carried; the bacteria themselves were just trash, as easy to make as ice in a freezer and just as easy to toss down the drain (after proper sterilization, of course). Everyone in the lab knew that manufacturing a few thousand bases of DNA cost us about $10-20 in media, purification equipment, and labor. We never considered that every purification also involved throwing away thousands of dollars' worth of assorted proteins, RNAs, lipids, glycans, and metabolites.

If you listen to a metabolic engineer describe their work, you might hear them hail bacteria as quasi-magical self-replicating nanofactories capable of transmuting literal gunk into gold. But if you then visit that same engineer’s laboratory, you will see those very nanofactories treated as base sludge — used and then discarded. It’s hard for us, gifted as we are with eyes that can’t actually see a bacterium, to hold E. coli’s sophistication in our heads for very long. My hope is that, by breaking down the bacteria into its component parts — at least on paper — we can more easily come to appreciate it.

{{divider}}

Samuel Clamons is a bioinformatics scientist at Illumina, Inc. with a PhD in Bioengineering and training in applied mathematics and computer science. Outside of his day job, he writes science fiction and researches theoretical questions in biology at Asimov Press.

Cite: Clamons, Samuel. “The Price of E. coli.” Asimov Press (2025). https://doi.org/10.62211/82ue-71kj

Thanks to Eryney Marrogi and Ella Watkins-Dulaney for reading a draft of this. Lead image by Ella Watkins-Dulaney.

{{divider}}

Footnotes

1. This article lists 13 companies active in 2016 selling products of metabolic engineering. To my surprise, as of 2025, eight still have active websites, with most selling product; two have gone bankrupt or shut down, and the other three still appear in industry summaries and news articles but either don’t have easily-accessible websites or appear to have had their websites hijacked.
2. See this article for more information about the MCF2Chem database of research-scale biosynthetically created compounds.
3. Though most of this value currently comes from food additives and from microbe-derived foods, such as beer and cheese. For scale, though, the current yearly bioplastics market alone is estimated at $26 billion, bio-alcohols are $11 billion, and biologics (many of which, to be fair, are made in eukaryotes instead of bacteria) are $0.5 trillion.
4. Perhaps using some hyper-effective form of liquid chromatography?
5. In practice, LB cultures easily reach 2 or 3 OD, and can be pushed quite a bit higher, though at high densities, OD isn’t linear with respect to actual bacteria density (so a 3 OD culture has quite a bit less than 3x as many bacteria as a 1 OD one). Consider 1 OD a conservative estimate.
6. An E. coli cell is 70 percent water by mass.
7. By this logic, I really shouldn’t include glucose or amino acids, which are provided directly to our precocious little cell factories in the LB broth. It turns out they don’t contribute much to the overall parts price for metabolites anyway, so I’ve merely included them for completeness.
8. By my calculations, the metabolites listed in Bennet et al. add up to about 17 percent of total E. coli mass, which is about six times higher than the total metabolite pool estimate from Cell Biology by the Numbers. Clearly, there’s some discrepancy between different sources, so adjust your confidence accordingly; I nevertheless feel comfortable claiming that this list of 100-odd molecules represents “most” of the E. coli metabolome by mass and molecule count.
9. As an example, Sigma-Aldrich sells citrate (as a salt with either sodium or potassium) in purities of 97 percent, 98 percent, 99 percent, 99.5 percent, 99-105 percent, “meets USP testing specifications,” “suitable for cell culture,” “Pharmaceutical Secondary Standard; Certified Reference Material,” “Molecular Biology Grade,” “EMPLURA®,” and more.
10. Elaborating on the previous example, citrate comes as either a solution (with or without pH balancing) or as a crystalline salt counterbalanced with sodium, zinc, potassium, magnesium, ammonium-iron, or lead, sometimes at different hydrations.
11. Sigma-Aldrich is a well-respected, broadly-stocked one-stop-shop for most purchasable chemicals, and many other companies sell through their website portal.
12. A critical enzyme for fixing DNA strand breaks, used in a variety of assays and reactions requiring DNA binding or recombination.
13. An enzyme that removes phosphate groups from a variety of molecules, used in molecular cloning to functionally “block” the ends of DNA from sticking end-to-end.
14. A complex of three enzymes (recB, recC, and recD) that chews back single-stranded overhangs in DNA complexes of mismatched length. Also known as recBCD.
15. According to Wiśniewski & Rakus (2014), whose supplementary data I used as a primary source for protein concentrations and molecular weights throughout.
16. The exceptions tell interesting stories of their own. Alkaline phosphatase, the cheapest protein by far, competes on the market with shrimp-derived alkaline phosphatase, functionally similar but extremely cheap to produce for some reason. Ribosomes are presumably cheaper to purify because they are both abundant and heavy. RecA is one of a small number of E. coli proteins produced at scale by New England Biolabs, making it cheaper than average. On the other end of the price spectrum, the exonuclease V complex, also sold by New England Biolabs, is likely expensive (per gram) because it is a complex of multiple proteins, tuned for maximum efficiency and conveniently packaged with buffers and optimized protocols — in other words, when you buy it, you’re really paying for more than just the proteins.
17. An E. coli scaffold could be used to make DNA origami structures much larger than those made from, say, bacteriophage genomes.
18. A plasmid is a small circle of DNA that replicates in a bacterial host, separate from the main chromosome. Plasmids typically hold between one and a few genes, and are frequently used by biologists as an easy-to-manipulate chassis for inserting new functionality. If the genome is a hard drive, a plasmid is a tiny flash drive.