limits of metabolic engineering

=chemistry =biochemistry =metabolic engineering

 

 

Genetic engineering can be used to get microbes to convert sugar to a wide variety of complex chemicals with good efficiency. So, is that the future of industrial chemistry? Yes, but only to a limited extent.

 

 

limits of enzymes

 

Enzymes are catalysts, and operate under the same physical principles and fundamental limitations that other catalysts do. They're very good catalysts in some ways, but their advantages are also linked to some limitations.

Enzymes have very good selectivity. They can act on very dilute and very specific molecules and rarely act on others. That selectivity requires binding strongly to their substrates, which also requires that they bind strongly to their products for most reactions to happen, which means that most enzymes act very slowly when the concentration of their products is high. This strong binding to substrates also requires complex structures.

Enzymes can do reactions at low temperature, but this requires complex structures, which makes enzymes unstable at high temperatures. This means reactions can't usually be driven by heat differences.

To summarize, enzymes can work at low concentrations and low temperatures, but for the same reasons, they can usually only work at low concentrations and low temperatures.

 

 

 

limits of genetic engineering

 

If you take a microbe, and add in an enzymatic process that greatly reduces its growth and has no benefits, then that microbe will quickly evolve to downregulate those enzymes. It's possible to make high-value products unrelated to core metabolism, like monoclonal antibodies, but it's very expensive. If you want to get high concentrations and high yields of chemicals from fermentation, they need to be primary metabolic products.

For something to be a primary metabolic product, production of it needs to produce net ATP. This is a fairly restrictive criteria for anaerobic fermentation. If you start adding oxygen, then your microbes definitely have ways to make ATP without making your product, and it becomes more difficult to keep them from doing less of that. Also, you need to actually get the oxygen to cells, and that can be a big challenge. It's not so bad for fermentation making citric acid, where you can use small bubbles and mixers and only need a little oxygen, but if you want to, say, grow organs in a tank without a heart and circulatory system and lungs, getting oxygen to the cells is a huge problem.

 

 

 

input costs

 

Fermentation of sugar produces at most 53% of its mass in ethanol and ethanol has 65% the energy per mass of gasoline. So, if sugar is $300/ton, it takes at least $870 worth of sugar to make ethanol fuel equivalent to a ton of oil. That's not particularly cheap, and it doesn't account for the costs of fermentation or distillation.

Methanol is cheaper per joule than sugar, and it's theoretically possible to make some fermentation products from methanol, but the options are more limited than with sugar, and this approach isn't used yet industrially. There's currently no economic incentive to pursue it now, either, considering US corn subsidies and the development costs.

 

 

 

toxicity

 

Ethanol is, by far, the chemical most produced by industrial fermentation. Final concentration reaches ~11%. Distillation of ethanol afterwards is energy-intensive and expensive, though still less expensive than the sugar used.

Most chemicals are more toxic than ethanol and harder to separate. Propanol is substantially more toxic than ethanol to cells. Butanol is substantially more toxic than that. A lot of compounds that could be made by fermentation are toxic to microbes at <1% concentrations.

 

 

 

separation

 

 

distillation

 

Ethanol separation by distillation is currently the cheapest separation of a fermentation product.

Distillation is only feasible for small molecules with low boiling points. (For example, succinic acid is valuable and easy to make with fermentation, but because it can't be distilled out, it's too expensive to separate.) That's a very limited set of choices, and small molecules are also easy to make without using fermentation. The high selectivity of enzymes is more valuable for more complex products with more synthesis steps.

 

 

phase separation

 

Ethanol is toxic largely because, being partly polar and partly nonpolar, it disrupts cell membranes. What if you make some really hydrophobic compounds and try to get them to phase separate? For example, pinene, or some other terpene. (Let's set aside the ATP requirements of making the DMAPP used to make those.)

Most proteins need hydrophobic regions. If you have hydrophobic compounds at a concentration where they would spontaneously phase separate, they'll first go into proteins and disrupt them. So, you generally can't produce terpenes by fermentation to a concentration where they'll phase separate.

How, then, are organisms able to accumulate hydrophobic compounds? How does an oil palm fruit contain so much oil? Those compounds are generally stored in special organelles called lipid droplets. It's only by first accumulating oil in those then mechanically squeezing it out that pure oil can be produced. (Yes, I know, there are also soluble particles of fatty acids held by lipoproteins.)

(Speaking of oil palms, they require less land for a given amount of oil production than other plant oils, and there's nothing inherently more environmentally destructive about them, so I think the recent moral panic about palm oil as an ingredient is questionable.)

Could the same approach be used artifically? Yes, it's possible to put porous polymer beads in a fermentation tank, use those to absorb oils, and then squeeze out the oils. But, of course, that's expensive.

It's also possible to accumulate hydrophobic compounds in lipid droplets in the microbes used, but that's generally limited to <50% of the cell mass, while fermentation of sugar to ethanol produces several times the mass of cells in ethanol. Growing sugarcane to grow microbes to squeeze for oil is less efficient, and thus much more expensive, than growing plants to squeeze for oil.

 

 

extraction

 

A much cheaper approach than polymer beads is liquid-liquid extraction followed by distillation. For example, making isobutanol by fermentation, extraction with xylene, and then distillation. That's sometimes cheaper than direct distillation, but while product concentration in water is less important, it's still an issue: if it's low, (say, 0.1%) then mass transfer will be relatively slow.

Another issue with this approach is toxicity of the extractant; usually it just reduces growth rate, but this does limit your choices. You also don't want to be losing too much extractant in fermentation broth effluent.

It's still necessary to separate the product from the extractant, and usually this is done by distillation, which requires that both things have suitable boiling points.

 

 

precipitation

 

It's possible to precipitate products during fermentation. For example, tyrosine has a low enough solubility to do that. Unfortunately, this doesn't necessarily make separation easy. When tyrosine precipitates during fermentation, it forms tiny needles that stick to bubbles, and you get a messy foam.

Taking this approach, some other possible problems are solid coatings forming on valves and tiny crystals that stick to cells. Most ionic compounds from fermentation will have a large enough zeta potential (surface charge resulting from one type of ion dissolving more than the other) that particles won't flocculate.

Of course, it's possible to collect those solid products; it's just more expensive than distillation. This is also obviously limited to things that precipitate before they become toxic, which is a fairly restrictive criteria.

 

 

membranes

 

Cows ferment grass and get net energy from the fermentation products. How do cows handle the product separation problem?

Cow rumens get products through, basically, membranes. They have a lot of surface area, and they're self-cleaning; doing the same thing with artificial polymer membranes would be too expensive. The lining also actively transports things like lactate and butyrate, using monocarboxylate transporters; that's not practical to copy artificially.

To what extent are membranes feasible for separating fermentation products? I'm not actually sure how well people will be able to deal with fouling issues, but in the short term, I doubt that direct pressure driven membrane separation of fermentation products will be economically feasible.

 

 

 

conclusion

 

There are a lot of issues with producing industrial chemicals by fermentation. And yet, I still think it's an important topic, and will be much more widely used in the future than it is now. The point of this post is that it's a challenging problem, and that there are major limitations which mean "conventional" (non-fermentation) chemical processing will still be needed.