Ask Assistant Professor of Chemical Engineering Camilla Kao to describe a bacterium, and she’ll compare it to a factory capable of producing antibiotics, immunosuppressants and anti-cancer drugs that no chemist can synthesize. Bacteria normally produce antibiotics to inhibit other bacterial strains competing for resources. Pharmaceutical companies exploit this property to manufacture drugs, but the process of cultivating bacteria is slow and costly. By hijacking molecular synthesis with the latest advances in biotechnology&emdash;and inducing bacteria to overproduce antibiotics&emdash;Kao aims to greatly speed up drug development.
“Companies have proprietary mutant [bacterial] strains that have been worked on for more than 20 years, but many of the mutations still are unknown,” Kao said. “There’s no knowledge you can gain from these overproducers to turn around and make others.”
That’s where Kao comes in. On March 13, she presented her latest progress at the 229th annual meeting of the American Chemical Society in San Diego, focusing on research that earned her the 2004 Jay Bailey Young Investigator Award, which recognizes the best paper in the journal Metabolic Engineering.
Funded by the National Science Foundation and the National Institutes of Health, Kao’s research uses a “reverse engineering” strategy to learn from and exploit drug-company mutants&emdash;in this case, the bacterial species that produces erythromycin, an antibiotic given to treat bronchitis, diphtheria, whooping cough, pneumonia and other conditions. Reverse engineering involves taking something apart to see how it works. For Kao, it means asking how one strain of Saccharopolyspora erythraea produces 10 times the erythromycin of normal strains.
The answer, it seems, lies not in the cluster of genes known to produce the antibiotic but in a mutation of another gene that co-regulates the timing of expression for the antibiotic-producing genes. The result? The mutant produces erythromycin for five days. The nonmutant produces it for only one. If the specific mutation can be located, Kao reasons, perhaps it can be introduced into other antibiotic-producing bacteria to increase their yields.
DNA microarrays
Without a tool known as the DNA microarray, or DNA chip, Kao’s finding may not have been possible. First developed a decade ago at Stanford in the lab of biochemistry Professor Patrick Brown, microarrays use both computer technology and DNA base pairing to track the expression of thousands of genes simultaneously. The technique has tremendous potential for the study of biological processes where numerous genes work in concert&emdash;an apt description of antibiotic synthesis.
A microarray consists of single-stranded DNA arranged on a wafer the size of a postage stamp. The genetic sequence of the DNA tethered to the chip is known. Then researchers add a sample with an unknown composition of fluorescently labeled DNA molecules. The base pairs in the sample’s single-stranded DNA, groups of which correspond to individual genes, stick to similar, or homologous, regions of the DNA tethered to the chip. The presence of the resulting double-stranded DNA reveals the base-pair sequence and the amount of that sequence in the added sample.
One way to measure gene expression is to measure levels of messenger RNA&emdash;that’s RNA that translates the genetic information of DNA into proteins. Accordingly, one can go backward and make DNA copies of all RNA transcripts in the mutant bacterial cultures of interest. With this clever trick, one can get DNA that can bind to the DNA already on the microarray, and the concentrations accurately reflect gene expression.
In Kao’s experiment, DNA copies of RNA from cultures of erythromycin-producing bacteria were infused with fluorescent dye and applied to the microarray.
“This experiment revealed that the RNA transcript abundance of the erythromycin-producing genes remained higher for longer times in the Saccharopolyspora erythraea mutant strain,” Kao said.
Searching for mutations
When researchers reveal how mutants overproduce antibiotics, the real work has only just begun&emdash;successful reverse engineering hinges on the ability to reproduce the mechanism of overproduction in new strains. Ideally, the exact sequence of base pairs causing the mutation could be found and then directly introduced into a new strain.
Drug companies use chemicals and radiation to damage DNA and induce mutations. But it’s not easy to detect these single base-pair mutations in the so-called industrial mutant strains. Kao’s group, for example, detected the presence of an unidentified regulator gene altering the timing of erythromycin production but could not pinpoint the specific mutation.
A similar study by Kao’s group identified two genes responsible for overproduction of the antibiotic tylosin. Instead of increasing the length of time that a bacterium produces the antibiotic&emdash;the case with erythromycin&emdash;these two genes allow the bacterium to produce it faster. Again, the specific mutations eluded Kao’s group.
Currently, the researchers are trying to identify the mutations by chopping overproducer DNA into small pieces, several genes in length, and inserting them into normal bacteria to look for changes in either expression of antibiotic gene clusters or final antibiotic yields.
DNA shuffling
Studying mutant strains that already exist represents only half of Kao’s work; she also creates her own mutants. Creating the mutants on her own offers two advantages. First, instead of making mutants with chemicals and radiation, her group uses transposons&emdash;small DNA segments that can replicate on their own and insert their copies into new positions anywhere in the genome. Mutations induced by transposons are significantly easier to identify.
A technique called DNA shuffling, recently developed by the Redwood City-based company Maxygen, provides Kao with a second advantage&emdash;speed. After one round of generating mutants with transposons, only a handful of the many mutants generated will likely increase antibiotic yields. Shuffling creates combinations of all mutants that originally generated increased yields&emdash;similar to genetic recombination&emdash;and produces sequences with multiple mutations more quickly.
Think of a genome as a deck of cards, and inducing a beneficial mutation as slipping in a joker. It’s easy to see which decks have jokers&emdash;or which strains produce more antibiotics. Imagine mixing and then separating 10 decks that all have two jokers each. Do it with enough replications and you’ll get a deck with 20 jokers&emdash;or a bacterium with 20 beneficial mutations.
“In one year we should be able to generate a high producer that’s equivalent to 20 years of work with the traditional method, and then we can study the identifiable mutations,” Kao said.
Whether she’s studying existing mutants or creating her own, Kao’s main interest lies in learning how to manipulate cells to do what she wants. The family of soil bacteria she focuses on, the actinomycetes, produces thousands of bioactive molecules, with many more discoveries sure to come.
“When new molecules are discovered to be made by species of this family, the research we do will hopefully be used to quickly make mutant strains that make a lot more of it,” she said. “No drug company wants to wait 20 years to release a new antibiotic.”
This story was written by science-writing intern Kenneth M. Dixon.
