Marcia S. Osburne was Director of Molecular Biology, Trudy H. Grossman was Staff Scientist, Paul R. August was Senior Research Scientist, and Ian A. MacNeil was Staff Scientist at ARIAD Pharmaceuticals, Inc., Cambridge, Mass., at the time this article was written. Marcia S. Osburne, Paul R. August, and Ian A. MacNeil are currently at Aventis Cambridge Genomics Center, Cambridge, Mass., and Trudy H. Grossman is currently at Vertex Pharmaceuticals, Inc., Cambridge, Mass.

Links to Other ASM Pages:

• Table of Contents
• ASM News Issues

Some microbiologists are probing the rich diversity of their backyards instead of going far afield to find useful natural products

Marcia S. Osburne, Trudy H. Grossman, Paul R. August, and Ian A. MacNeil

For many decades, medicinal and agricultural chemists maintained a keen interest in the broad spectrum of natural products produced by microbiota. Soil, in particular, is an intensively exploited ecological niche from which many useful natural products are derived, including clinically important antibiotics such as tetracycline, erythromycin, vancomycin, b -lactams, cephalosporins, and rifampicin. Actinomycetes in particular are the source for the majority of commercially exploited natural products, exhibiting diverse chemical structures. Unicellular bacteria, on the other hand, more commonly, but not exclusively, produce peptide antibiotics.

The intense competition among soil microbes for limited nutrients is thought to be a major reason why some of these organisms produce antimicrobial "secondary metabolites"—molecules that, although not required directly for growth, serve useful purposes for the microbes that make them, giving them an edge over those that do not. However, soil microbes make other natural products besides antibacterials, including anticoccidial agents, antifungal drugs, herbicidal agents, anticancer drugs, insecticidal and nematocidal agents, immunomodulating compounds, and enzyme inhibitors. Soil microorganisms also produce a variety of lipopeptides, lipoproteins, glycolipids, and lipopolysaccharides with surface-acting properties, as well as industrially important enzymes including cellulases, amylases, proteases, and lipases (used extensively in textile applications). Other microbial enzymes are important for research in universities and throughout the biotechnology industry. For example, restriction enzymes and enzymes used in the polymerase chain reaction are used routinely in recombinant DNA technology.

In the 1980s, many large pharmaceutical companies conducted natural products drug discovery programs. Typical procedures included isolating microorganisms from soil samples, growing those microorganisms in large flasks at various temperatures in a variety of selective or nonselective media, harvesting and extracting cultures with organic solvents, and testing the extracts in a spectrum of targeted screens for activity against infectious disease agents and for other medical or veterinary applications.

Already by that time the rediscovery rate for natural products was quite high; novel molecules were discovered only rarely. In an effort to find new molecules, employees who happened to be traveling to distant locations were sent off with soil collection bags, with the idea that unusual microbes producing untapped natural products might populate those remote geographical areas. Company-sponsored collection trips to exotic locations became routine events. Despite the ancillary attraction of these long-distance excursions, however, some experts openly speculated that there might be ample microbial diversity in a cubic centimeter of the soil from our own backyards—if only we could assess that diversity and then recover the organisms that embody it.

Microbial Diversity in Soil Environments

By the early 1990s, scientists realized that only a small fraction of microorganisms in any soil sample can be cultured by standard techniques, grossly underrepresenting the actual diversity there. The methods and media available in a standard microbiology laboratory simply cannot reproduce the conditions necessary to recover the vast majority of bacteria in a particular soil environment. The number of species currently cultivatable from soil is thought to represent 1\% or less of the total population. More important, the uncultured species represent spectacular diversity.

These statistics emerge from PCR-based studies in which DNA in the soil is amplified without culturing the microbes that encode it. Analysis of highly conserved prokaryotic 16S rRNA gene sequences is part of a standard method for assessing microbial diversity. In one study of PCR-amplified 16S rRNA-encoding gene sequences, Jo Handelsman, Robert Goodman, and their collaborators at the University of Wisconsin, Madison, discovered a high percentage of novel phyla of Archaea in locally obtained soil samples. Handelsman and Goodman refer to the collective genomes of soil microflora as constituting a "metagenome."

Archaea are the recently described prokaryotic third domain of life, thought previously to occur predominantly in "extreme" environments. Other studies of soil samples show spectacular diversity in the bacterial domain as well. Thus the large number of unculturable and largely unknown prokaryotic species in the soil provides a promising source of untapped genetic diversity, with the potential of yielding useful natural products. These remarkable observations are encouraging others, including our group, to analyze systematically soil metagenomes from various locations in search of novel and potentially useful natural products.

Accessing the Soil Metagenome

In 1998, a small group of scientists at ARIAD Pharmaceuticals, in collaboration with Handelsman and Goodman, began a program to access the soil metagenome and discover new drugs. To make this theoretical resource a reality, we planned to generate natural products libraries of the soil metagenome, screen those libraries, and identify and analyze "hits" on selected target activities as a proof-of-concept for the program.

Figure 1

Our initial strategy entailed isolating large fragments (100–300 kb) of DNA directly from the soil or other microenvironments, without first cultivating microbes in our samples. This DNA was ligated into bacterial artificial chromosome (BAC) vectors, low-copy plasmids that can readily maintain large DNA inserts. These recombinant BAC vectors were then transformed into host microorganisms, such as E. coli, thereby generating a recombinant library (Fig. 1). The resulting clones can be screened for biological activity expressed directly in E. coli or in some other suitable surrogate expression host. Alternatively, the library can be probed for sequences of interest in a genomics-type application. This approach circumvents the need to culture microorganisms from environmental samples, and it also provides a relatively unbiased sampling of the genetic diversity of those environments.

At the start of our collaboration, two prototype libraries had already been constructed by Michelle Rondon in Handelsman’s laboratory: a BAC library of B. cereus DNA containing large (>100 kb) inserts (n), and a soil metagenome library containing approximately 4,000 clones with average insert size of about 30 kb. Because the known genes for natural-product biosynthetic pathways in bacteria sometimes exist in rather large gene clusters, our next goal was to develop a library with larger metagenomic DNA insert fragments.

Cloning large fragments into BAC vectors proved to be a difficult challenge. Cleaning up soil DNA to the point where it could be ligated into BAC vectors is difficult to accomplish without shearing the DNA. However, we developed a process in which soil DNA can be gently extracted, sized, and ligated into a BAC vector, resulting in our first "large" metagenome library of 12,000 clones. We obtained the soil from which this library derives from a local New England site.

Screening "Hits" from the Soil Metagenome Library and Follow-up

It can be argued that E. coli, an enteric organism, may not be the ideal host in which to express random metagenome genes. Nevertheless, we reasoned that if the general techniques proved successful, we could eventually express those genes in a variety of other hosts, including Streptomyces and Bacillus spp., to enhance the likelihood of detecting any particular gene or gene cluster.

Figure 2

With those options in mind, we set up a series of agar plate screens to test our metagenome clones for biological activity, and included robotically conducted screens for antibacterial and antifungal activities as well as a series of enzyme activity and siderophore screens. To our delight, we detected at least four clones producing antibacterial hits and several clones producing enzyme activity hits in our large-insert library. One of the four antibacterial-producing clones encodes a family of small molecules related to indirubin, a pigmented molecule with antileukemic properties and an inhibitor of tyrosine kinases (Fig. 2).

In traditional natural products screening programs, extracts that are "hits" in a screen of interest require follow-up analysis, including efforts to isolate and identify active compounds. This process, typically involving analytical chemists, often requires fairly large amounts of extract, in turn requiring the producing microbe to be cultured—"refermented" —on a large scale. One major frustration of traditional natural product screening is known familiarly as the "referm" problem: soil-derived cultures that produce an activity of interest the first time they are grown often cannot be made to produce that activity again when they are refermented. This problem is estimated to occur more than 50% of the time in traditional natural products screening endeavors, and probably occurs because scientists can only make educated guesses as to the best way to grow a soil microbe to enhance production of the secondary metabolites of interest.

However, the metagenome approach offers some distinct advantages over the traditional approach with regard to this frustrating issue. Although the metagenome approach also requires analytical chemistry to isolate and identify compounds with interesting biological activity, the DNA encoding the active molecule(s) is already isolated on a BAC plasmid, and the genes in question can be identified by selectively knocking out their functions via insertion mutagenesis. The relevant genes can then be sequenced and analyzed, quickly providing valuable information as to the identity of the molecule(s) with interesting activity. Once isolated, the relevant genes can be engineered to enhance expression of the natural product. We have recently developed methods to increase expression from BAC plasmids and facilitate high-throughput sequencing of their metagenomic DNA inserts.

Phylogenetic Diversity of Soil DNA

The soil used to construct our metagenome library comes from a local site in the Boston area, and was collected during March when the ground was beginning to thaw. We were interested to know the extent of microbial diversity in this DNA, both in terms of the number and types of species. One method frequently used to estimate microbial diversity is based on a phylogenetic analysis of DNA sequences encoding small subunit ribosomal RNA (SSU rRNA).

Purified DNA used to construct the metagenome library served as the template for PCR amplification, using degenerate oligonucleotide primers designed to amplify bacterial rRNA genes. From this mixture, more than 100 distinct 16S rRNA genes were cloned and sequenced. Altogether, they describe a broad spectrum of sequences, some of which appear to fall readily within known bacterial families isolated from all over the world. Equally striking, however, is that the majority of those sequences seem to represent unidentified bacterial families, presumably from still-uncultivated and otherwise undescribed bacterial populations.

Figure 3

To depict the relationships among these putative species, we arrayed some 70 sequences in a phylogenetic tree (Fig. 3). One subtree in this array contains an exact match to a Cytophaga species that typically is found

in glacial ice. Thus, our findings bring up the intriguing possibility that in the temperate climate found in the Boston area, with temperatures ranging from more than 100ºF in summer to below 0ºF in winter, the spectrum of diversity may encompass some microbes that are thought to populate only extreme environments.

Possibly the cold temperatures that prevailed when our soil sample was collected led inadvertently to our obtaining this representative of a "cold-loving" bacterial family. It will be interesting to see whether a summer soil sample taken from the same site includes microbial representatives from families commonly found in more tropical climates. More extensive analysis should answer the question of whether a cubic centimeter of Boston area backyard soil might actually encompass sufficient microbial diversity to keep a drug discovery program busy for a long time.

Tapping into Microbial Diversity: Complementary Approaches, Challenges

Following the discovery of this vast, previously unappreciated cache of terrestrial microbial diversity, many academic and industrial researchers are beginning to develop alternative approaches to exploit these resources. For example, investigators at TerraGen Discovery Inc., a Vancouver-based biotechnology company, are using surrogate host expression systems to examine environmental DNA as a way of identifying novel products.

Meanwhile, researchers at Diversa in San Diego, California, are tapping into microbial diversity by inserting the DNA of unculturable organisms into expression libraries, thereby identifying stable biocatalysts extracted from extreme environments in Iceland, Costa Rica, and Yellowstone National Park. Investigators at several other companies, including Khosan Biosciences, Inc. in Hayward, Calif., and Maxygen, Inc. in nearby Redwood City, are taking advantage of microbial diversity by recombining environmentally derived genes into gene clusters encoding natural product biosynthetic pathways in order to derive novel products.

We are finding that the world is unexpectedly full of unknown microorganisms, and they offer much to discover in terms of unanticipated microbial DNA sources encoding potentially useful natural products. However, before BAC libraries prove effective sources of natural products, several important technological hurdles need to be overcome. In particular, expressing large multigene clusters of metagenomic DNA in surrogate host strains remains a difficult challenge. Moreover, constructing "high-quality" metagenomic libraries, in which a high percentage of clones in a particular library contain "interesting" genes whose activities can be detected in high-throughput screening formats, represents another critical interim goal that must be met before this technique can be considered a truly practical endeavor.

These and other challenges will require serious efforts from many researchers. Despite the challenges, however, the opportunity to move forward in this very exciting new field, which will not only furnish novel natural products but teach us much about the microbial world, seems irresistible.

ACKNOWLEDGMENTS

We thank all of our colleagues at ARIAD Pharmaceuticals, Inc., and the University of Wisconsin and Cornell University for their contributions to this work.

SUGGESTED READING

Bintrim S. B., T. J. Donohue, J. Handelsman, G. P. Roberts, and R. M. Goodman. 1997. Molecular phylogeny of Archaea from soil. Proc. Nat. Acad. Sci. USA 94:277–282.

Hoesse, R., S. Leclerc, J. A. Endicott, M. E. M. Nobel, A. Lawrie, P. Tunnah, M. Leost, E. Damiens, D. Marie, D. Marko, E. Niederberger, W. Tang, G. Eiserbrand, and L. Meijer. 1999. Indirubin, the active constituent of a Chinese antileukaemic medicine, inhibits cyclin-dependent kinases. Nature Cell Biol. 1:60–67.

Pace, N. R. 1997. A molecular view of microbial diversity and the biosphere. Science 276:734–740.

Rondon, M. R., P. R. August, A. D. Betterman, S. F. Brady, T. H. Grossman, M. R. Liles, K. Loiacono, B. A. Lynch, I. A. MacNeil, C. Minor, C. L. Tiong, M. Gilman, M. S. Osburne, J. Clardy, J. Handelsman, and R. M. Goodman. 1999. Cloning a soil metagenome: bacterial artificial chromosome libraries to harvest microbial diversity in natural environments. Appl. Environ. Microbiol. 66:2541–2547.

Rondon, M. R., S. J. Raffel, R. M. Goodman, and J. Handelsman. 1999. Toward functional genomics in bacteria: analysis of gene expression in Escherichia coli from a bacterial artificial chromosome library of Bacillus cereus. Proc. Nat. Acad. Sci. USA 96:6451–6455.

Shizuya, H., B. Birren, U.-J. Kim, V. Mancino, T. Slepak, Y. Tachiiri, and M. Simon. 1992. Cloning and stable maintenance of 300-kilobase-pair fragments of human DNA in Escherichia coli using an F-factor-based vector. Proc. Natl. Acad. Sci. USA 89:8794–8797.

Torsvik, V., K. Salte, R. Sorheim, and J. Goksoyr. 1990. Comparison of phenotypic diversity and DNA heterogeneity in a population of soil bacteria. Appl. Environ. Micro. 56:782–787.

Vining, L. C. 1990. Functions of secondary metabolites. Annu. Rev. Microbiol. 44:395–427.

Ward, D. M., R. Weller, and M. M. Bateson. 1990. 16S rRNA sequences reveal numerous uncultured microorganisms in a natural community. Nature 345:63–65.

Whitman, W. B., D. C. Coleman, and W. J. Wiebe. 1998. Perspective. Prokaryotes: the unseen majority. Proc. Nat. Acad. Sci. 95:6578–6583.