About Chemical Innovation - Subscription Information
November 2000
Vol. 30, No. 11, 3 – 4.
Leading the Way

Table of Contents

Mining for biological activity

Doreen Gillespie is a postdoctoral researcher at the University of Wisconsin, Madison (Dept. of Plant Pathology, 1630 Linden Dr., Madison, WI 53706; 608-262-9815; deg@plantpath.wisc.edu). photo of Doreen GillespieShe received her B.A. degree in biology at Bryn Mawr and Haverford Colleges and her Ph.D. in genetics from the University of Washington, Seattle. Before joining the metagenomics research team in Madison, she was a postdoctoral fellow at the Seattle Biomedical Research Institute.

The microorganisms found in soils are a gold mine for drug discovery research. In fact, microorganisms from soil produce most of the antibiotics used in medicine today. Traditional studies of bacteria used to identify medicinal agents have been limited by their dependence on bacterial culturing. Recent evidence shows that <1% of soil microbes are recovered by standard culturing methods. The research group that I joined in 1998, led by Robert Goodman and Joe Handelsman at the University of Wisconsin, Madison, predicts that uncultured soil microorganisms contain a myriad of uncharacterized, medicinally useful products. The challenge is to access these products in spite of the challenges of culturing the microorganisms. Our approach to overcoming this challenge is to clone the DNA isolated directly from soil into libraries, thus bypassing the culture barrier. The genomes of these bacteria represent a portion of the collective genetic information found in soil, a collection that we term the metagenome.

The screening process

My research focuses on screening libraries for biological activities. We currently have ~30,000 individually arrayed clones, which are being subjected to various activity screens. The clone inserts are of sufficient size to contain multigene biosynthetic pathways capable of producing small, biologically active molecules (Figure 1). Screening existing libraries for phylogenetic markers by using the polymerase chain reaction has revealed that the cloned sequences are derived from diverse bacterial divisions that are historically difficult to culture. Thus, our phylogenetic data indicate that we have the raw material in our libraries for producing and detecting novel gene products that previously could not have been found by the traditional culturing methodology.

Before my involvement in this project, the laboratory identified clones with antibacterial, lipase, amylase, and nuclease activities. We have since identified 33 hemolytic clones, one of which produces two compounds that inhibit the growth of Gram-positive and Gram-negative bacteria, which include Salmonella, Bacillus, Staphylococcus, Streptomyces, and Streptococcus species. I used transposon mutagenesis to identify a single open reading frame responsible for increased production of the compound.

I have targeted my research efforts toward the development of a screening technique that will address three major challenges in the design of bacteria-based libraries: access to intracellular molecules; rate of screening throughput; and ease of positive clone recovery. Our heterologous expression system uses Escherichia coli to recognize and express foreign genetic information (Figure 2). This system will not lead to the expression of genes from all clones. In addition, small molecules that would be secreted in the source microorganism may not be efficiently secreted from E. coli. Because of the costs of labor and materials, screening our current and future libraries depends on the design of high-throughput screening that will minimize input and maximize yield. One means of minimizing input is to screen large pools of clones, but subsequent identification and recovery must then be rapid and relatively simple.

The optimal design

I entered this project with a molecular biology background, having worked most recently with Plasmodium falciparum, the causative agent of malaria. An awareness of strat egies pursued in antiparasitic agent research led me to propose screens optimized to detect inhibitors of characterized pathways or enzymes. With this in mind, I have developed an approach that will allow us to screen for specific inhibitory activities in large pools of library clones and easily recover individual library clones that produce potential inhibitors. The concept is to assemble, in the library cells, the cloned metagenomic DNA and genes that encode the molecular needed to detect a particular inhibitory activity. For intracellular inhibitor screening, I will introduce into pools of library cells a “detection plasmid” encoding one or more elements necessary to detect library clones to produce inhibitory molecules.

Optimal design of the detection elements is critical to the success of this screening approach and will permit rapid recovery of positive clones from large pools of library cells. For example, detecting inhibitors of a characterized protease is possible by introducing a plasmid that encodes the protease and a substrate containing the protease cleavage site into library cells. The substrate is designed to provide either positive selection of library clones containing protease inhibitors or an intracellular alteration in a rapidly assessed phenotype, such as fluorescence or luminescence, upon inhibition of the substrate cleavage.

Specifically, inhibitors of the peptide cleavage activity of sortase, a protein required for virulence in Staphylococcus aureus, will be detected by expressing in library host cells the genes for sortase and a protein, TetA, that confers tetracycline resistance on the cell. The TetA for this screen is modified to contain an internal cleavage site recognized by sortase. If active sortase is in the cell, the modified TetA is cleaved, rendering it incapable of providing resistance to tetracycline and resulting in clone death. In the presence of an inhibitory molecule, however, the modified TetA remains intact and provides tetracycline resistance, enabling rapid recovery on tetracycline selection of library clones containing potential inhibitors of sortase cleavage activity. Similarly, a substrate requiring cleavage for inhibition of fluorescence activation may be substituted for the TetA substrate, permitting screening and clone recovery by fluorescence-activated cell sorting.

The promise of screening

This strategy will give us the power to screen thousands of clones a day, and it can easily be adapted to other target activities, including inhibitors of other proteases, molecules that disrupt or stabilize protein dimerization, enzymes involved in metabolic pathways, and proteins or small molecules required for regulation of selected heterologous gene expression. With the potential represented in the soil metagenome libraries, we anticipate exciting new natural product discoveries from these and other ongoing screens in our group.


I would like to acknowledge the other contributing members of my research team whose work is mentioned in this article: Robert M. Goodman, Jo Handelsman, Alan D. Bettermann, and Mark R. Liles at the University of Wisconsin, Madison; Jon Clardy and Sean F. Brady of Cornell University; and Michelle R. Rondon at Ohio State University.

Return to Top || Table of Contents