There’s an inherent tension in convincing organisms to produce fuel for us. To grow and thrive, the organism has to direct its energy into a variety of chemicalsâ€”proteins, fats, DNA, and more. But for biofuels, we’re mostly interested in fats, which are long-chain hydrocarbons that already look a lot like our liquid fuels. Fat is easy to convert into biodiesel, for example.
So how do we convince an organism to do what we want, rather than what it needs? There have been two approaches to this so far. One is to take an organism that we understand well and engage in genetic engineering to direct its metabolism toward fuel production. The second approach is to search for organisms that naturally produce lots of the chemicals we’re interested in.
Now, researchers at the company Synthetic Genomics have taken what you might consider a hybrid approach. They’ve started with an algae that will produce oodles of fat, but only if you stop its growth by starving it of essential nutrients. And, by studying how this starvation response works, the scientists identified a key regulator and altered its activity. The engineered strain produces nearly as much fat as the normal strain, but it does so while continuing to grow.
The species in question is a single-celled algae called Nannochloropsis gaditana. It has two properties that make it great for biofuel production. One is that it’s part of a genus that is happy to grow in salt and brackish water, meaning that biofuel production doesn’t have to compete for fresh water. The second property is that it naturally produces a slew of fats (largely triacylglycerols). Starving Nannochloropsis for an essential nutrient (nitrogen) causes the algae to convert its spare energy to fat for storage, allowing it to ride out the adverse conditions. These lipids can end up accounting for 60 percent of the cells’ dry weight.
Unfortunately, starving the Nannochloropsis algae isn’t exactly conducive to continued growth. Rather than having a nice, continuously expanding culture that you can pull cells out of for fuel production, the entire population has to go through a boom-bust cycle. Researchers have tried for years to engineer a similar response that doesn’t require starvation, but their efforts have been slowed by the fact that there are no genetic tools for engineering Nannochloropsis, and we don’t know enough about the biology of its starvation process to really understand what to target.
The new work from Synthetic Genomics deals with both of these hurdles. To start with, the company’s researchers got the CRISPR-Cas gene-editing system working in Nannochloropsis. That allows them to target any gene they’d like for deletion, modification, or replacement.
But they also worked on understanding how the starvation process gets triggered. Changes in fat metabolism start to become apparent about five hours after all nitrogen sources are taken out of the culture. So, the team reasoned, changes in gene activity have to come before that. After three hours of starvation, the researchers looked for changes in the activity of a specific class of genes: those that bind to DNA and regulate nearby genes. These, they reasoned, could be controlling the starvation process.
They came up with a list of 20 genes. The researchers then targeted 18 of them individually for elimination using the CRISPR editing system.
One of these 18 genes, called ZnCys, turned out to be exactly what the researchers were hoping to find. Eliminating the gene caused the algae to build up three times more fat as the normal strain. Unfortunately, the edited version also acted like it was starving, with growth slowing to a crawl. As a result, the normal strain would outproduce the gene-edited version over the long run.
To get around this issue, the researchers started targeting sites near the part of the gene that encodes a protein. These nearby sequences often help control the amount of protein produced from a gene, so disrupting them could produce a version of the ZnCys that had lower activity than normal but wasn’t completely shut down. Their plan worked; the researchers ended up with three new strains, which saw ZnCys activity reduced by 20, 50, and 70 percent, giving them a nice range to test.
To an extent, all of the new strains worked. While total productivity of the three engineered lines was down compared to a normal strain, it was only down by anywhere from five to 15 percent. While there were definitely fewer cells, they incorporated large quantities of carbon, and they converted more than twice as much of it to lipids. This more than made up for the drop in cell number. Critically, the strains did fine in a continuous culture, meaning that you could siphon off 70 percent of the cells each day for biofuel production without shutting the whole culture down.
A closer examination of gene activity in the cells showed that the engineered versions had reduced activity of genes involved in importing and assimilating nitrogen. So even when nitrogen was present, the cells weren’t able to use as much of it, which nicely explains why they acted like they were semi-starving.
Ideally, I expect that Synthetic Genomics would prefer to generate a strain that produces a lot of lipids even when the strain is not nitrogen starved at all. As a result, the company probably viewed ZnCys as a bit of a disappointmentâ€”Synthetic Genomics would have probably preferred a gene that simply switched the metabolism into lipid production mode without messing with nitrogen.
Still, the study provides some indication of how the nitrogen response is regulated. One of the other 18 genes the researchers looked at (or the two they didn’t) may or may not be more directly involved in lipid production, but it didn’t show good performance in this screen because it had so many other effects. No doubt the team is continuing to dissect the pathways that get activated when nitrogen becomes limiting.
And, in the mean time, the researchers have a strain that can do continuous biofuel production at double the rate of the normal oneâ€”which is certainly better than what they started with.
from Ars Technica http://ift.tt/2tqkFRA