A guy named Austin was wandering the halls of MIT's Stata Center this afternoon with a plasma bag. Its contents are a little darker and a little grayer than you'd expect blood to be - maybe the color of well-peppered Bloody Mary mix. It's also a little thinner. "We're having problems with the expression level of the hemoglobin," Austin told me when I poked at the bag.

Austin Day is the brains behind the bactoblood project - bacterially produced hemoglobin - brought by the UC Berkeley team. I mentioned this project in a previous post, and I was excited to see the presentation. The concept is to engineer E coli to produce a red blood cell substitute, and let the bugs loose in the body to create a cheap and universal system for blood transfusions.

One hurdle, as one of the presenters explained dryly, is that "injecting E coli into the blood stream may cause a few problems," such as sepsis -- bacteria in the blood stream is just what the immune system is designed to fight. So they encased their E coli in a chassis, or bacterial coat, specially developed by a Berkeley bioengineer to evade detection by the immune system (he's using it to create tumor-killing bacteria).

They then built a controller out of two plasmids, which basically turns on protein expression, and then, to prevent cells from replicating in the blood stream, self-destructs. They say the stuff can also be freeze-dried (Austin had the vials of brown flake to prove it), giving it a perk that real blood doesn't have. Adam Arkin, the group's faculty advisor, summarized the project in three simple clauses: "Express a bunch of protein, eat your genome, kill yourself."

I asked Arkin, whether that project, or in his opinion, others at the contest, would carry on. He wasn't convinced they would -- he said that the group probably will finish putting all the biological parts together and seeing how the whole system works (they hadn't gotten to this stage by contest time), but beyond that was unlikely. "We know it won't be a blood substitute in round 1," he said, and developing the idea further would take serious time and money.

He pointed out some obvious problems with bactoblood's clinical applications. One issue is, how much bacteria would a person have to receive to get enough blood, and would that amount cause sepsis? Another, well publicized in the last couple weeks, is that blood that doesn't retain nitrous oxide does not do well -- and some solution to that problem would have to be engineered. "I think a lot of projects are about teasing an idea," Arkin said. "The first few steps to show you could probably do it."