CAN bacteria be logical? Yes they can, say researchers who have made
genetically engineered bugs that do the same job as the components of a
microchip.
These smart bugs will crunch on chemical inputs rather than digital bits.
They could one day be sent into waste-water plants to hunt out toxic chemicals.
Or they could tell doctors what proteins are present in body fluids.
Just like a silicon-based AND gate that requires two electrical inputs, the
bacterial logic gate will only produce an output when both its chemical inputs
are present. It is the combination of such simple functions that allows
computers to perform highly complex tasks, says Michael Simpson, a physicist at
the Oak Ridge National Laboratory in Tennessee.
Advertisement
Nature already combines simple functions to good effect, says Simpson. 鈥淲hen
you look at genetic circuits you see things that look like binary logic and
feedback patterns.鈥� The challenge is to manipulate them to your advantage.
So Simpson, together with Gary Sayler and James Fleming, modified
Pseudomonas putidacells to produce both AND and OR gates
(see Graphic).
They are now trying to produce more complex gates. The trick, he says, is to
choose the genes to suit the kind of inputs and outputs you want.
In the case of the AND gate, for example, they used chemical 鈥渋nducers鈥� as
inputs. One causes a gene to make a protein that the second input inducer must
have to express the output enzyme. You need both inducers to produce an output,
which could, for example, be a bioluminescent enzyme.
PCs with living logic circuits are still a long way off, warns Tim Gardner, a
bioengineer and president of Cellicon Biotechnologies in Boston. For one thing,
the basic components are not as good as silicon. 鈥淭hey are slower, noisier and
more difficult to manipulate,鈥� he says.
But Simpson is optimistic. 鈥淲e鈥檙e trying to take advantage of what cells do
well鈥攑rocessing information on the molecular scale,鈥� he says. With the
right sequence of inputs and outputs, you could have a number of gates inside a
single cell, with the output of one gate becoming an input of another. In
theory, a single cell could produce massively parallel complex functions.
Because the chemical inputs don鈥檛 alter the normal control metabolism, the
cells can still reproduce and survive. But in a competitive environment with
other unmodified cells they would be at a disadvantage, says Simpson, as they
would be expending energy on computations instead of survival.