PERSPECTIVES

36. Casjens, S. Prophages and bacterial genomics: what 45. Blaisdell, B. E., Campbell, A. M. & Karlin, S. Similarities is, however, that any application has to be
have we learned so far? Mol. Microbiol. 49, 277–300 and dissimilarities of phage genomes. Proc. Natl Acad.
(2003). Sci. USA 93, 5854–5859 (1996). designed with enzymatic constraints in mind,
37. Lander, E. S. & Waterman, M. S. Genomic mapping by leading to suboptimal process and reaction
fingerprinting random clones: a mathematical analysis. Acknowledgements
Genomics 2, 231–239 (1988). Support from the National Science Foundation and Moore conditions. Instead of designing a process to
38. Angly, F. et al. PHACCS, an online tool for estimating Foundation is gratefully acknowledged. fit a mediocre enzyme, it is conceivable that
the structure and diversity of uncultured viral
communities using metagenomic information. BMC Competing interests statement the uncultivated microbial diversity, together
Bioinformatics 6, 41 (2005). The authors declare no competing financial interests. with in vitro evolution technologies, might
39. Bak, P. How Nature Works: The Science of Self-
Organized Criticality (Springer–Verlag, New York, be used to find a suitable natural enzyme(s)
1996). Online links that can serve as a backbone to produce a
40. Wommack, K. E. & Colwell, R. R. Virioplankton: viruses
in aquatic ecosystems. Microbiol. Mol. Biol. Rev. 64, DATABASES designer enzyme that optimally fits proc-
69–114. (2000). The following terms in this article are linked online to: ess requirements that are solely dictated by
41. Paul, J., Jeffrey, W. & DeFlaun, M. Dynamics of Entrez: http://www.ncbi.nlm.nih.gov/Entrez
extracellular DNA in the marine environment. Appl. ΦX147 | Streptococcus pyogenes strain MGAS315 | Xylella
substrate and product properties7.
Environ. Microbiol. 53, 170–179 (1987). fastidiosa strain Temecula1
42. Paul, J., Jiang, S. & Rose, J. Concentration of viruses
and dissolved DNA from aquatic environments by FURTHER INFORMATION
Novelty. For industries that produce bulk
vortex flow filtration. Appl. Environ. Microbiol. 57, Forest Rohwer’s laboratory: http://phage.sdsu.edu commodities such as high-performance
2197–2204 (1991). BLAST: http://www.ncbi.nlm.nih.gov/BLAST
43. Simon, M. & Azam, F. Protein content and protein FASTA: http://www.ebi.ac.uk/fasta33
detergents, a single enzyme backbone with
synthesis rates of planktonic marine bacteria. Mar. GenBank: http://www.ncbi.nlm.nih.gov/Genbank/index.html superior functionality that has an entirely new
Ecol. Prog. Ser. 51, 201–213 (1989). PHACCS: http://phage.sdsu.edu/phaccs
44. Rohwer, F. Construction and analyses of linker- Phrap computer program: http://www.phrap.com
sequence would be useful to avoid infringing
amplified shotgun libraries (LASLs) [online], Phred computer program: http://www.phrap.com/phred competitors’ intellectual property rights. This
<http://www.sci.sdsu.edu/PHAGE/LASL/index.htm> Sequencher: http://www.genecodes.com/sequencher
(2002). Access to this interactive links box is free online.
problem is illustrated by the fact that substitu-
tions at nearly every position in the mature
275 amino acid BPN (bacillus protease
Novo type, from Bacillus amyloliquefaciens)
subtilisin have been claimed in patents8.
O U T LO O K
Maximum diversity. The pharmaceutical
and supporting fine-chemicals industries
Metagenomics and industrial often seek entire sets of multiple diverse
biocatalysts to build in-house toolboxes for
applications biotransformations9. These toolboxes need to
be rapidly accessible to meet the strict time-
lines of a biosynthetic-feasability evaluation
Patrick Lorenz and Jürgen Eck in competition with traditional synthetic
chemistry.

Abstract | Different industries have different enzymes, were identified in a single effort Elusive metabolites. Many pharmacologically
motivations to probe the enormous resource that sampled marine prokaryotic plankton active secondary metabolites are produced by
that is uncultivated microbial diversity. retrieved from the Sargasso Sea. bacteria that live in complex CONSORTIA (see
Currently, there is a global political drive to Glossary) or by bacteria that inhabit niches
promote white (industrial) biotechnology An industrial perspective that are difficult to reconstitute in vitro10. So,
as a central feature of the sustainable In this perspective, the discussion is limited although there are reports on how to circum-
economic future of modern industrialized to prokaryotes, as their genomes are most vent this general problem of microbial culti-
societies. This requires the development easily targeted by the functional screening vation either by mimicking natural habitats11
of novel enzymes, processes, products tools available in metagenomics and because or by allowing for interspecies communica-
and applications. Metagenomics promises it is assumed, based on published literature, tion after single cell micro-encapsulation12,
to provide new molecules with diverse that the largest biodiversity occurs in the the cloning and heterologous expression of
functions, but ultimately, expression systems bacterial lineages 4–6. Different industries biosynthetic genes that encode secondary
are required for any new enzymes and are interested in exploiting the resource of metabolites (usually present as gene clusters)
bioactive molecules to become an economic uncultivated microorganisms that has been is the most straightforward and reproduc-
success. This review highlights industrial identified through large-scale environmental ible method of accessing their biosynthetic
efforts and achievements in metagenomics. genomics for several reasons detailed below. potential.

Metagenomics1 has the potential to sub- The ideal biocatalyst. For any industrial Industrial enzyme applications
stantially impact industrial production. The application, enzymes need to function Enzymes are used in a wide range of applica-
dimensions of the enormous biological and sufficiently well according to several appli- tions and industries13. They are required in
molecular diversity, as shown by Torsvik2, cation-specific performance parameters only minute quantities to synthesize kilo-
Venter3 and their co-workers, are truly aston- (FIG. 1). With the exception of yeasts and grams of stereochemically challenging chiral
ishing. A pristine soil sample might contain filamentous fungi, access to novel enzymes SYNTHONS that are used as building blocks to
in the order of 104 different bacterial species. and biocatalysts has largely been limited by produce highly active pharmaceuticals14, and
More than one million novel open reading the comparatively small number of cultiva- at a kiloton/year scale as active ingredients
frames, many of which encode putative ble bacteria. A corollary of this limitation for bulk products such as high-performance

510 | JUNE 2005 | VOLUME 3 www.nature.com/reviews/micro

© 2005 Nature Publishing Group
 FO C U S O N M E TA G E N OM I C S

Activity Stability Besides the involvement of the food,

Turnover frequency (kcat) Temperature stability feed, detergent and politically heavily-
pH stability promoted biofuel industries, it is the globally
Specific activity (kat/kg, U/mg)
operating chemical and pharmaceutical
Temperature profile Ingredient/byproduct stability industries that are riding this ‘third wave
of biotechnology’18 (a term illustrating the
chronology of developments in which red
pH profile Solvent stability and green biotechnology come first and sec-
ond, respectively). As industries face increas-
6 5 4 3 2 1 1 2 3 4 5 6 ing low-cost competition, particularly from
East Asia, and political pressure to reduce
Efficiency Specificity
Space-time yield
their environmental impact and resource
Substrate range
consumption to improve sustainability, it
is felt that there is a strong need for smart
Product inhibition Substrate specificity (Km, kcat/Km) and innovative technologies, processes and
products to remain competitive.
Byproduct/ingredient inhibition Substrate regioselectivity and The McKinsey consultancy predicts that,
enantioselectivity by 2010, biotechnology could be applied in
Producibility/expression yield Substrate conversion (%), yield the production of between 10% and 20% of
Figure 1 | Multi-parameter footprint analysis. This figure illustrates the ideal biocatalyst concept. Each all chemicals sold (amounting to a value
enzyme candidate from the metagenome is ranked, from low (rating of 1) to high (rating of 6) using a of $160 billion) and that up to 60% of all fine
specific set of criteria, to produce a multi-parameter fingerprint (shown in yellow). Criteria include in vitro chemicals (medium-volume products used
enzyme activity, efficiency, specificity and stability. This decision matrix reveals the strengths and weaknesses as intermediates in the manufacturing of
of every candidate enzyme, so that the most promising candidate enzymes from diverse enzyme libraries products such as pharmaceuticals, flavours,
can be selected for further process development by re-screening, protein engineering or directed
fragrances, agro-chemicals and detergents)
evolution methods. kat, catalytic reaction rate; kcat, catalytic constant; Km, Michaelis constant; U, unit.
might be produced using biotechnology17.
Even for the traditional mainstay of the
chemical industries, the polymer market
laundry detergents8. Their versatility allows processes as diverse as the production of acryl- (typical bulk-volume products), McKinsey
their use in many applications, including amide (Mitsubishi Rayon, Japan) and the use predicts that biotechnology might account
processes to degrade natural polymers such of enzymes in oil-well completion (British for up to 10% of the output.
as starch, cellulose and proteins, as well as Petroleum Exploration, UK). Currently, the The European Chemical Industry Council
for the REGIOSELECTIVE or ENANTIOSELECTIVE movement towards implementing sustain- and EuropaBio have responded to a call from
synthesis of asymmetric chemicals. able technologies and processes is gaining the European Commission, which asked for
Global industrial enzyme sales were esti- momentum, particularly in Europe. suggestions for ten central technological plat-
mated at $2.3 billion in 2003 REF. 15. The ‘Industrial’ or WHITE BIOTECHNOLOGY is forms, by proposing a ‘European Technology
main profits were divided among detergents currently a buzzword in the European bio- Platform for Sustainable Chemistry’. White
($789 million), food applications ($634 mil- business community. The term was coined biotechnology will be one of three pillars of
lion), agriculture/feed ($376 million), textile in 2003 by the European Association for the platform, together with materials sci-
processing ($237 million), and pulp/paper, Bioindustries (EuropaBio), based on a case ence, and reaction and process design. As a
leather and other applications including study report, and it denotes all industrially corollary, this triggered the nomination of
enzymes for the production of fine and bulk harnessed bio-based processes that are not sustainable chemistry as one of the thematic
chemicals ($222 million). In the face of soar- covered by the RED BIOTECHNOLOGY (medical) priorities in the European Union’s seventh
ing energy costs, dwindling fossil resources, or GREEN BIOTECHNOLOGY (plant) labels17. White framework research programme, with the
environmental pollution and a globalized biotechnology has its roots in ancient human aim of strengthening the scientific and tech-
economy, the large-scale use of biotechnol- history and its products are increasingly part nological bases of European industry and
ogy instead of, or to complement, traditional of everyday life, from vitamins, medicines, encouraging its international competitive-
industrial production processes, particularly biofuel and bioplastics to enzymes in deter- ness. Clearly, this joint European initiative
in the chemical sector, is viewed as both an gents or dairy and bakery products. It is the does not come without ‘stimulating’ prec-
opportunity and a necessity. In the future, belief of industrial promoters, analysts and edent from forward-looking international
novel biotechnological applications should policy makers that white biotechnology has peers in the United States or Japan, and
boost the market for industrial enzymes. the potential to impact industrial produc- leading European stakeholders (among them
tion processes on a global scale. The main DSM19, Degussa, BASF, Henkel, Novozymes
White biotechnology long-term applications of white biotechnol- and Genencor) have urged policy makers to
The notion that environmentally sound, ogy will be focused on replacing fossil fuels act to ensure a favourable competitive posi-
commercially viable biotechnological proc- with renewable resources (biomass conver- tion. The promises of industrial biotechnol-
esses can take their respected place in a sion), replacing conventional processes with ogy cannot be realized without coordinated
global industrial environment has been bioprocesses (including metabolic engineer- and well-funded research and development
acknowledged for some time16. Companies ing) and creating new high-value bioprod- (R&D) efforts, a supportive and guid-
from Europe, Canada, Japan, South Africa ucts, including nutraceuticals, performance ing political framework and early public
and the USA reported on their experiences in chemicals and bioactives. involvement to pre-empt concerns over new

NATURE REVIEWS | MICROBIOLOGY VOLUME 3 | JUNE 2005 | 511

© 2005 Nature Publishing Group
PERSPECTIVES

So, there is ample demand for novel

Application requirements
and process constraints enzymes and biocatalysts, and metagenomics
is currently thought to be one of the most
likely technologies to provide the candidate
molecules required 26,27 . The diversity of
Production of potential substrates for enzymatic trans-
metagenomic library formations in the fine-chemical industry and
the short time that is usually available to pro-
pose viable synthetic routes (particularly for
the pharmaceutical industry) make it useful
Sequence-based- Activity-based-
Primary screening screening screening to produce pre-characterized enzyme librar-
ies using generic substrates, before screen-
ing for a specific enzyme that is required for
biotransformation of a particular substrate of
Enzyme library —
precharacterized interest (FIG. 2).
candidate set
In vitro evolution Secondary Screening for industrial enzymes
and enzyme screening
engineering Application a Application b After Torsvik’s seminal report in 1980 on the
extraction and digestion of genomic DNA
Enzyme
candidates from bacterial samples prepared from soil28
and the proof of the concept for generating
Scale-up application gene libraries directly from environmental
DNA29, it was only in the next decade that
Pilot-scale the first such metagenome libraries were
process reported. These were constructed from non-
Scale up
(PCR-)amplified genomic DNA fragments,
isolated directly from marine plankton30
and from enriched consortia sampled from
cellulose digesters31. The research on the
enriched consortia was particularly impor-
Enzymatic application tant, in that it describes the first successful
biocatalytic process
expression screening of metagenomic DNA.
Once this was achieved, the usefulness of the
Figure 2 | Industrial enzymes — from the metagenome to applications and processes. This figure metagenome approach for biotechnology
illustrates industrial rational bioprospecting of the metagenome. A library of cloned DNA is produced and was clear. It was already evident that most
primary screening, based on application requirements such as the conversion of an indicator substrate, environmental bacteria were recalcitrant
produces enzyme libraries (different clones, which encode different enzymes, are indicated by different
to cultivation32, but metagenomics meant
colours) that serve as platforms for subsequent development. Secondary screening of the enzyme library
can identify process-specific properties such as substrate specificity, activity or stability. Primary and that the biotechnological potential of these
secondary screening both involve several stages based on different criteria, therefore the screening stages uncultivated bacteria might be accessible by
are multi-layered. Enzyme engineering and in vitro gene evolution can form a part of the development directly cloning DNA sequences retrieved
process. Subsets of cloned enzymes (red clones represent one subset, blue clones represent another from the environment. To our knowledge,
subset) are then used in scale-up application or process testing to identify suitable enzyme candidates. Diversa (San Diego, California, USA) and
Finally, process improvement by enzyme optimization and process engineering are carried out. Economic
TerraGen Discovery (Vancouver, Canada,
feasibility must be proven in a pilot-scale process environment as a prerequisite for scale-up to production
or final application scale. Examples of the use of enzyme libraries in an industrial context include
taken over in 2000 by Cubist Pharmaceuticals,
nitrilases36 (Diversa), alcohol dehydrogenases9 (Schering Plough) and glycosylhydrolases39 (BRAIN AG). Cambridge, Massachusetts, USA) were the
first companies to file patents on specific
aspects of metagenome technology in 1996.
Diversa claimed methodologies to isolate
technologies and generate the demand that increasingly popular with the chemical and normalize environmental DNA prior to
industry requires for economic success20. industries and are viewed as indispensable cloning and screening and described a result-
Whereas the food, feed and detergent for the modern organic chemist 21 . With ing fosmid library created from marine pico-
industries typically concentrate on a many successfully implemented processes plankton33. TerraGen’s claims were focused
limited number of enzyme reactions and operating worldwide22,23 and the number of on the retrieval of environmental DNA
substrates, the chemical and pharmaceu- industrialized biotransformations having sequences from soil to express industrially
tical industries deal with thousands of doubled every decade since 1960 REF. 24, it relevant xylanases34. Since then, several small-
chemically and structurally diverse mol- is estimated that biocatalysis might provide to medium-sized enterprises (SMEs) have
ecules, and production of each of these a superior synthetic solution over classical used uncultivated environmental microbial
requires individual enzymatic solutions. chemistry in 10% of processes25. In fact, the resources to retrieve relevant enzymes and
Consequently, owing to a wealth of poten- unavailability of an appropriate biocatalyst biocatalysts. Most industrial metagenomic
tially useful biocatalysts, biotechnological is thought to be a limiting factor for any discoveries reported so far have been made
applications of microbial resources are biotransformation process22. by SMEs or academic groups collaborating

512 | JUNE 2005 | VOLUME 3 www.nature.com/reviews/micro

© 2005 Nature Publishing Group
 FO C U S O N M E TA G E N OM I C S

Table 1 | Activity-based screening for industrially relevant enzymes and biocatalysts from metagenomic libraries
Function Habitat Library Average Number Library Substrate Number Hit rate Ref.
type insert of clones size (Mb) of hits (hit per Mb)
size (kb) screened
Esterase/lipase Forest soil Plasmid 8 67,000 536 Tributyrin 98 1/5.5 *
Esterase/lipase Forest soil Fosmid 40 19,968 799 Tributyrin 47 1/17 *
Esterase/lipase Sandy Fosmid 30 29,884 903 Tributyrin 49 1/18 *
ecosystem
Esterase/lipase Sandy Fosmid 40 25,344 1,014 Tributyrin 29 1/35 *
ecosystem
Esterase/lipase Soil Plasmid 6 286,000 1,716 Tributyrin 3 1/572 63
Esterase/lipase Soil Plasmid 6 730,000 4,380 Triolein 1 1/4,380 63
Esterase/lipase Soil BAC 27 3,648 100 Bacto Lipid 2 1/50 64
Oxidation of Soil Plasmid 3 900,000 2,700 1,2-ethanediol; 1,2- 15 1/180 65
polyols propanediol; 2,3-
butanediol
Alcohol Soil/enrichment Plasmid 4 400,000 1,600 Glycerol/1,2- 10 1/160 66
oxidoreductase propanediol
Amidase Soil/enrichment Plasmid 5 193,000 965 D-phenylglycine-L- 7 1/138 67
leucine
Amylase Soil Plasmid 5 80,000 400 Starch 1 1/400 68
Amylase Soil BAC 27 3,648 100 Starch 8 1/12 64
Biotin production Soil/excrement Cosmid 35 50,000 1,750 Biotin-deficient 7 1/250 69
enrichment medium
Protease Soil Plasmid 10 100,000 1,000 Skimmed milk 1 1/1,000 70
Cellulase Sediment λ phage 6 310,000 1,860 Carboxymethyl- 3 1/620 71
enrichment cellulose
Chitinase Seawater λ phage 5 825,000 4,125 Methylumbelliferyl- 11 1/375 72
diacetylchitobioside
Dehydratase Soil/sediment Plasmid 4 560,000 2,240 Glycerol 2 1/1,120 38
enrichment
4-Hydroxybutyrate Soil Plasmid 6 930,000 5,580 4-Hydroxybutyrate 5 1/1,116 73
conversion
β-Lactamase Soil Plasmid 5 80,000 400 Ampicillin 4 1/100 68
The screening host for all studies was Escherichia coli. *Unpublished results, K. Liebeton et al., BRAIN AG. BAC, bacterial artificial chromosome.

with larger companies, mainly in the chemi- serve the fine-chemical and pharmaceutical company BRAIN AG (Zwingenberg,
cal industry. Different methods for clon- industries36. Besides target-gene-specific Germany) collaborated with the detergent
ing environmental DNA and screening molecular enrichment strategies (see below), manufacturer Henkel (Düsseldorf, Germany)
for enzymes or bioactive molecules have a technical focus is also placed on the sen- to screen for novel glycosyl hydrolases for use
been devised by different companies. Only sitivity and throughput of screening by in laundry applications39, and with chemicals
a few selected examples per company are using cell-sorting technology37. Diversa has company Degussa (Düsseldorf, Germany) to
mentioned in the following section. partnered with several companies for the dis- use sequence-based screening strategies
Diversa, the largest and most prominent covery and development of enzymes, among to clone novel nitrile hydratases40.
specialist biotech company for the com- them DSM, Syngenta and BASF. Genencor Because of the ease of genetic manipula-
mercialization of metagenome technologies, International (Palo Alto, California, USA) tion of E. coli, combined with the availability
has described several approaches to access teamed up with researchers from Göttingen of elaborate cloning and expression systems,
‘uncultivable’ microorganisms. Applying a University (Germany) to search for new most of the screens to date have used gene
classical growth-selection-based expression dehydratases of potential use in the produc- libraries constructed in this bacterium.
strategy, diverse environmental libraries tion of 1,3-propanediol from glucose, using However, the reported hit rates for the
were constructed in Escherichia coli using a metabolically engineered strain of E. coli38. required targets vary between enzyme classes
phage λ and were screened for nitrile use. Both sequence-based screens, targeting new TABLE 1. This might be due to difficulties
After growth in media containing nitriles as genes that were related to known reference in expressing specific genes in E. coli, their
the sole nitrogen source, more than 100 new genes, and growth selection protocols, to scarcity in the metagenome pool or a com-
and diverse nitrilase genes were recovered35. screen for expressed and functional enzyme bination of both. To overcome the potential
The resulting enzyme library is marketed to activity, were used. The German biotech limitations of expressing genes in E. coli,

NATURE REVIEWS | MICROBIOLOGY VOLUME 3 | JUNE 2005 | 513

© 2005 Nature Publishing Group
PERSPECTIVES

Table 2 | Activity-based screening for antimicrobial metabolites from metagenomic libraries

Habitat Library type Average insert Number of clones Library Target Screening Positive Hit rate Ref.
size (kb) screened size (Mb) organism host hits (hit per Mb)
Soil BAC 44.5 24,546 1,092 Several * 3 1/364 74
Soil BAC 63 12,000 756 B. subtilis, E. coli 4 1/189 56
S. aureus
Soil BAC 27 3,648 100 B. subtilis E. coli 1 1/100 64
Soil Fosmid 30 13,440 403 S. aureus E. coli 3 1/134 ‡
Soil Cosmid nd 5,000 nd B. subtilis E. coli 1 nd 55
Soil Cosmid 30–45§ nd Unknown B. subtilis E. coli 10 1/300–900 75
*Producer colonies identified by production of a pigment. Isolated pigments were shown to have antimicrobial activity. ‡Unpublished results, R. Schulze and G. Meurer,
BRAIN AG. §Estimate based on specifications of the pWEB cosmid cloning system (Epicentre Technologies). BAC; bacterial artificial chromosome; B. subtilis, Bacillus
subtilis; E. coli, Escherichia coli; nd, not disclosed; S. aureus, Staphylococcus aureus.

alternative cloning hosts such as Bacillus Prokaria41 (Reykjavik, Iceland). Sequence- a target compound to come to the market is
subtilis, Streptomyces spp., Pseudomonas spp. specific enrichments (‘biopanning’), carried decisive, and in these applications it might be
or eukaryotic expression systems could be out either on the metagenomic input DNA even more important for a company to have
used. However, this option is limited by prior to cloning or on the library after clon- a large collection of biochemically diverse
the low throughput that is associated with ing in plasmid vectors, have been reported catalysts, even if these molecules are not
alternative cloning hosts and serves mainly by Diversa42. expressed in large amounts.
as an option for large-insert (>30 kb) clon- In addition to new technologies to
ing, which is used in bioactive molecule amplify DNA from limited resources using Searching for bioactive molecules
discovery (see below and TABLE 2). A dif- random primers and strand-displacing Despite large increases in the R&D expendi-
ferent strategy, whereby the abundance of DNA polymerase from phage Φ29 REF. 43, ture of big and medium-sized Pharma, the
target genes in the metagenomic DNA pool a strategy promoted by Lucigen (Middleton, number of pharmacologically active ‘new
is increased through microbial enrichment Wisconsin, USA), it is clear that current chemical entities’ (NCEs) has not increased
using low concentrations of externally added mass-sequencing efforts in several laborato- proportionally46, and this is most notable in
nutrients prior to DNA extraction, has been ries will facilitate the in silico identification the paucity of new antibiotics. Big Pharma
pursued by the Icelandic biotech company of open reading frames that might encode currently find antibiotics economically
potentially useful enzymes44. Another inter- unattractive to develop for various reasons
esting approach for a focused, high through- when compared with drugs to treat long-
Glossary
put screening system, substrate-induced gene term chronic conditions like obesity and
CONSORTIUM expression screening (SIGEX), was recently high cholesterol47. This could have dramatic
Physical association between cells of two or more types
of microorganisms. Such an association might be
applied to isolate catabolic enzymes from a consequences for the global antibiotic
advantageous to at least one of the microorganisms. groundwater metagenome45. Environmental resistance problem. Natural compounds
DNA was cloned upstream of a plasmid- and their derivatives will continue to have
ENANTIOSELECTIVE borne promoter-less gfp (green fluorescent an important role in drug discovery, as they
Selection of a particular enantiomer, one of a pair of protein) reporter gene in E. coli. Substrate have done in the past48 — natural products
stereospecific isomers.
(benzoate or naphthalene)-induced expres- constituted 63% of all newly approved anti-
GREEN BIOTECHNOLOGY sion of GFP in recombinant cells was used infectives between 1983 and 1994 REF. 49.
Biotechnology applied to agriculture, for example, to identify and isolate (by fluorescence acti- Many natural products are bacterial and
production of transgenic crop plants with genetically vated cell sorting, FACS) single cells carry- fungal secondary metabolites, but as most
engineered improved pest resistance.
ing metagenomic inserts (catabolic operons) microorganisms cannot easily be cultivated,
RED BIOTECHNOLOGY with substrate-responsive promoters and it is probable that many potentially active
Biotechnology in health care that uses substances associated downstream enzymes that were compounds have never been characterized.
produced in the human body to fight disease — subsequently shown to convert the inducer Metagenomics might have an invaluable role
medicines, vaccines, diagnostics and gene therapy. substrates to products. in the discovery process of new bioactive
REGIOSELECTIVE
Once new genes are cloned and screened molecules50,51. For this reason, metagenomes
Selection of a change that occurs with a greater for activity, the main stumbling block is have attracted the attention of academia and
frequency at one site than at several other sites, usually the expression of pure protein in sufficient specialized pharmaceutical companies.
involving a structural or positional isomer. amounts at reasonable costs. A cheap and TerraGen Discovery52,53 first incorporated
SYNTHON
efficient enzyme, usually produced in efficient PCR-amplified microbial-soil-metagenome
A molecule used as a chemical building block to expression systems like bacilli or filamentous genes into incomplete polyketide synthase
synthesize complex compounds. fungi, is a key factor for success, particularly clusters present in a recipient Streptomyces
when the enzyme functions as part of the final strain to show the usefulness of functionally
WHITE BIOTECHNOLOGY
(bulk) product such as in detergents. In the cloning and expressing environmental genes
Industrial use of biotechnology, for example, to produce
fine chemicals, biofuel, bioplastics, enzymes for use in
fine-chemical industry, this might be similar as part of a host cell’s secondary metabolite
detergents, food and feed. The boundaries between for bulk product synthesis. Particularly in the biosynthetic machinery. In a second step, the
white, green and red biotechnology are blurred. pharmaceutical industry, the time taken for same researchers reported the production

514 | JUNE 2005 | VOLUME 3 www.nature.com/reviews/micro

© 2005 Nature Publishing Group
 FO C U S O N M E TA G E N OM I C S

of novel bioactive metabolites (which were engaged in metagenomic research for the 12. Zengler, K. et al. Cultivating the uncultured. Proc. Natl
Acad. Sci. USA 99,15681–15686 (2002).
patented as terragines) after cloning metage- sake of anti-infectives discovery, other uses 13. Kirk, O., Borchert, T. V. & Fuglsang C. C. Industrial
nomic DNA fragments in a Streptomyces for bioactive molecules in cancer therapy enzyme applications. Curr. Opin. Biotechnol. 13, 345–
351 (2002).
lividans host using E. coli shuttle cosmid and other medical fields will probably lead 14. Patel, R. N. et al. Enzymic preparation of (3R-cis)-3-
vectors. This approach addressed two impor- companies that are currently engaged in red (acetyloxy)-4-phenyl-2-azetidinone: a taxol side-chain
synthon. Biotechnol. Appl. Biochem. 20, 23–33 (1994).
tant issues of using heterologous expression biotechnology to use metagenomics as its 15. Global Industry Analysts. Industrial enzymes — a global
systems to obtain useful amounts of phar- potential unfolds61. multi-client market research project. (GIA, San José,
California, USA, 2004).
macologically active, naturally occurring 16. The application of biotechnology to industrial
secondary metabolites. First, it is important Metagenomics — an industrial future sustainability. (Organisation for economic co-operation
and development (OECD), 2001).
to choose an appropriate expression host. As the excitement about genetic access to 17. Schepens, H. et al. White Biotechnology: gateway to a
S. lividans is an actinomycete, a family of the boundless realms of microbial diversity more sustainable future [online] <www.mckinsey.com/
clientservice/chemicals/pdf/BioVision_Booklet_final.pdf>
high GC-content Gram-positive bacteria that slowly gives way to the reality of tapping (2003).
devote large proportions of their genomes to into this diversity, the usual challenge of 18. Riese, J. Surfing the third wave: new value chain creation
opportunities in industrial biotechnology (McKinsey &
biosynthetic genes that synthesize consid- heterologous gene expression needs to be Company, Frankfurt, Germany, 2003).
erable amounts of secondary metabolites. addressed to turn metagenomic technolo- 19. DSM Position Document, Industrial (white)
biotechnology: an effective route to increase EU
Therefore, an actinomycete expression host gies into commercial successes, particularly innovation and sustainable growth [online] <http://www.
is more likely than E. coli to provide the nec- in applications for which bulk enzyme or b-basic.nl/documents/DSMIndustrial-white-
biotechnology.pdf> (2004).
essary genetic, biochemical and physiological product quantities have to be produced at 20. Herrera, S. Industrial biotechnology — a chance at
background to produce bioactive molecules competitive prices. Metagenomics, together redemption. Nature Biotechnol. 22, 671–675 (2004).
21. Schoemaker, H. E., Mink, D. & Wubbolts, M. G.
from genes that are derived from unchar- with in vitro evolution and high-throughput Dispelling the myths — biocatalysis in industrial
acterized bacteria which are also thought screening technologies, provides industry synthesis. Science 299, 1694–1697 (2003).
22. Schmid, A. et al. Industrial biocatalysis today and
to be members of this taxonomic lineage. with an unprecedented chance to bring bio- tomorrow. Nature 409, 258–268 (2001).
This, however, might change in the future, molecules into industrial application. That 23. Schmid, A., Hollmann, F., Park, J. B. & Buhler, B.
The use of enzymes in the chemical industry in Europe.
as successful attempts to metabolically engi- this is possible has been shown by Diversa. Curr. Opin. Biotechnol. 13, 359–366 (2002).
neer E. coli to improve its use as a cloning A full cycle from the discovery of novel 24. Straathof, A. J., Panke, S. & Schmid, A. The production
of fine chemicals by biotransformations. Curr. Opin.
and production host for complex secondary molecular scaffolds from multiple resources, Biotechnol. 13, 548–556 (2002).
metabolites (for example, 6-deoxyeryth- including metagenomes, to in vitro evolution 25. Roberts S. M. Biocatalysts in synthetic organic chemistry.
Tetrahedron 60, 499–500 (2004).
ronolide B, an antibiotic precursor) have technology to generate an improved biocata- 26. Lorenz, P., Liebeton, K., Niehaus, F. &, Eck, J. Screening
been reported54. Second, this study showed lyst for a specific application was realized to for novel enzymes for biocatalytic processes: accessing
the metagenome as a resource of novel functional
the importance of developing high-capacity produce α-amylases for applications in the sequence space. Curr. Opin. Biotechnol. 13, 572–577
vectors and suitable transformation proto- hot and acidic process of starch liquefac- (2002).
27. Schloss, P. D. & Handelsman, J. Biotechnological
cols to accommodate the large gene clusters tion62, showing the feasibility of the ‘ideal prospects from metagenomics. Curr. Opin. Biotechnol.
that are involved in secondary metabolite biocatalyst’ concept. 14, 303–310 (2003).
28. Torsvik V. L. & Goksoyr J. Determination of bacterial DNA
biosynthesis (TABLE 2). Patrick Lorenz and Jürgen Eck are at in soil. Soil Biol. Biochem. 10, 7–12 (1980).
Similar strategies were used by researchers BRAIN AG, Darmstaedterstrasse 34, 29. Pace, N. R., Stahl, D. A., Lane, D. J. & Olsen, G. J.
The analysis of natural microbial populations by ribosomal
from Aventis Pharmaceuticals55 (Cambridge, 64673 Zwingenberg, Germany. RNA sequences. Adv. Microbial Ecol. 9, 1–55 (1986).
Massachusetts), which addressed the clon- Correspondence to P. L. 30. Schmidt, T. M., DeLong, E. F. & Pace, N. R. Analysis of a
e-mail: [email protected] marine picoplankton community by 16S rRNA gene
ing-vector capacity and expression host cloning and sequencing. J. Bacteriol. 173, 4371–4378
issues by designing bacterial artificial chro- doi:10.1038/nrmicro1161 (1991).
31. Healy, F. G. et al. Direct isolation of functional genes
mosomes that could be used for cloning 1. Handelsman, J. Metagenomics: application of genomics encoding cellulases from the microbial consortia in a
metagenome DNA in E. coli56, Streptomyces to uncultured microorganisms. Microbiol. Mol. Biol. Rev. thermophilic, anaerobic digester maintained on
68, 669–685 (2004). lignocellulose. Appl. Microbiol. Biotechnol. 43, 667–674
and Pseudomonas strains57 (key members 2. Torsvik, V., Ovreas, L. & Thingstad, T. F. Prokaryotic (1995).
of this research team have since joined diversity-magnitude, dynamics, and controlling factors. 32. Staley, J. T. & Konopka, A. Measurement of in situ
Science 296, 1064–1066 (2002). activities of nonphotosynthetic microorganisms in aquatic
ActivBiotics of Lexington, Massachusetts, 3. Venter, J. C. et al. Environmental genome shotgun and terrestrial habitats. Annu. Rev. Microbiol. 39, 321–
USA). These endeavours were paralleled by sequencing of the Sargasso Sea. Science 304, 66–74 346 (1985).
(2004). 33. Short, J. M. & Mathur, E. J. Production and use of
efforts to extract large clonable DNA frag- 4. Curtis, T. P., Sloan, W. T. & Scannell J. W. Estimating normalized DNA libraries. US Patent 5,763,239
ments (mainly from soil). This work was sup- prokaryotic diversity and its limits. Proc. Natl Acad. Sci. (1998).
USA 99, 10494–10499 (2002). 34. Radomski, C. C. A., Seow, K. T., Warren, R. A. J. &
ported by European biotech company BRAIN 5. Ward, B. B. How many species of prokaryotes are Yap, W. H. Method for isolating xylanase gene sequences
AG, collaborating with researchers from the there? Proc. Natl Acad. Sci. USA 99, 10234–10236 from soil DNA, compositions useful in such method and
(2002). compositions obtained thereby. US Patent US 5,849,491
University of Darmstadt, Germany58, and 6. Schloss, P. D. & Handelsman, J. Status of the (1998).
BioSearch Italia (Gerenzano, Italy; recently microbial census. Microbiol. Mol. Biol. Rev. 68, 35. Robertson, D. E. et al. Exploring nitrilase sequence
686–691 (2004). space for enantioselective catalysis. Appl. Environ.
rebranded as Vicuron Pharmaceuticals), col- 7. Burton, S. G., Cowan, D. A. & Woodley J. M. The search Microbiol. 70, 2429–2436 (2004).
laborating with the University of Warwick, for the ideal biocatalyst. Nature Biotechnol. 20, 37–45 36. DeSantis, G. et al. An enzyme library approach to
(2002). biocatalysis: development of nitrilases for enantioselective
UK59. Through examination of the patent lit- 8. Maurer, K. H. Detergent proteases. Curr. Opin. Biotechnol. production of carboxylic acid derivatives. J. Am. Chem.
erature, it seems that Diversa is using similar 15, 330–334 (2004). Soc. 124, 9024–9025 (2002).
9. Homann, M. J. et al. Rapid identification of 37. Short, J. M. & Keller, M. Method of high-throughput
methods for the discovery of small molecules enantioselective ketone reductions using targeted screening of an environmental library. US Patent
using direct cloning60. microbial libraries. Tetrahedron 60, 789–797 (2004). 6,806,048 (2004).
10. Piel, J. Metabolites from symbiotic bacteria. Nat. Prod. 38. Knietsch, A., Bowien, S., Whited, G., Gottschalk, G. &
So, the obvious potential of the meta- Rep. 21, 519–538 (2004). Daniel, R. Identification and characterization of coenzyme
genomic approach for drug discovery has 11. Kaeberlein, T., Lewis, K. & Epstein, S. S. Isolating B12-dependent glycerol dehydratase- and diol
‘uncultivable’ microorganisms in pure culture in a dehydratase-encoding genes from metagenomic DNA
been realized by companies around the simulated natural environment. Science 296, 1127–1129 libraries derived from enrichment cultures. Appl. Environ.
world. Whereas in the recent past, SMEs (2002). Microbiol. 69, 3048–3060 (2003).

NATURE REVIEWS | MICROBIOLOGY VOLUME 3 | JUNE 2005 | 515

© 2005 Nature Publishing Group
PERSPECTIVES

39. Breves, R., Maurer, K. H., Eck, J., Lorenz P. & Zinke H. 56. MacNeil, I. A. et al. Expression and isolation of genes from microbial consortia. Appl. Environ. Microbiol.
New glycosyl hydrolases. PCT patent application WO antimicrobial small molecules from soil DNA libraries. 67, 89–99 (2001).
03/054177 (2003). J. Mol. Microbiol. Biotechnol. 3, 301–308 (2001). 70. Gupta, R., Beg, Q. K. & Lorenz, P. Bacterial alkaline
40. Liebeton K. & Eck J. Identification and expression in 57. Martinez, A. et al. Genetically modified bacterial strains proteases: molecular approaches and industrial
E. coli of novel nitrile hydratases from the metagenome. and novel bacterial artificial chromosome shuttle applications. Appl. Microbiol. Biotechnol. 59, 15–32
Eng. Life Sci. 4, 554–562 (2004). vectors for constructing environmental libraries and (2002).
41. Hobel, C. F. V. et al. Use of low nutrient enrichment to detecting heterologous natural products in multiple 71. Rees, H. C., Grant, S., Jones, B., Grant, W. D. &
access novel amylase genes in silent diversity of expression hosts. Appl. Environ. Microbiol. 70, Heaphy, S. Detecting cellulase and esterase enzyme
thermophiles. World J. Microbiol. Biotechnol. 20, 2452–2463 (2004). activities encoded by novel genes present in environmental
801–809 (2004). 58. Quaiser, A. et al. First insight into the genome of an DNA libraries. Extremophiles 7, 415–421 (2003).
42. Gray, K. A., Richardson, T. H., Robertson, D. E., uncultivated crenarchaeote from soil. Environ. Microbiol. 72. Cottrell, M. T., Moore, J. A. & Kirchman, D. L. Chitinases
Swanson, P. E. & Subramanian, M. V. Soil-based gene 4, 603–611 (2002). from uncultured marine microorganisms. Appl. Environ.
discovery: a new technology to accelerate and broaden 59. Berry, A. E., Chiocchini, C., Selby, T., Sosio, M. & Microbiol. 65, 2553–2557 (1999).
biocatalytic applications. Adv. Appl. Microbiol. 52, 1–27 Wellington E. M. Isolation of high molecular weight DNA 73. Henne, A., Daniel, R., Schmitz, R. A. & Gottschalk, G.
(2003). from soil for cloning into BAC vectors. FEMS Microbiol. Construction of environmental DNA libraries in
43. Detter, J. C. et al. Isothermal strand-displacement Lett. 223, 15–20 (2003). Escherichia coli and screening for the presence of
amplification applications for high-throughput genomics. 60. Short, J. Screening for novel bioactivities. US Patent US genes conferring utilization of 4-hydroxybutyrate. Appl.
Genomics 80, 691–698 (2002). 6,057,103 (2000). Environ. Microbiol. 65, 3901–3907 (1999).
44. McDonald, A. E. & Vanlerberghe, G. C. Alternative 61. Piel, J. et al. Antitumor polyketide biosynthesis by an 74. Gillespie, D. E. et al. Isolation of antibiotics turbomycin
oxidase and plastoquinol terminal oxidase in marine uncultivated bacterial symbiont of the marine sponge A and B from a metagenomic library of soil microbial
prokaryotes of the Sargasso Sea. Gene 349, 15–24 Theonella swinhoei. Proc. Natl Acad. Sci. USA 101, DNA. Appl. Environ. Microbiol. 68, 4301–4306 (2002).
(2005). 16222–16227 (2004). 75. Brady, S. F., Chao, C. J. & Clardy, J. Long-chain
45. Uchiyama, T., Abe, T., Ikemura, T. & Watanabe, K. 62. Richardson, T. H. et al. A novel, high performance N-acyltyrosine synthases from environmental DNA.
Substrate-induced gene-expression screening of enzyme for starch liquefaction. Discovery and Appl. Environ. Microbiol. 70, 6865–6870 (2004).
environmental metagenome libraries for isolation of optimization of a low pH, thermostable α-amylase J. Biol.
catabolic genes. Nature Biotechnol. 23, 88–93 (2004). Chem. 277, 26501–26507 (2002). Acknowledgements
46. Service, R. F. Surviving the blockbuster syndrome. 63. Henne, A., Schmitz, R. A., Bömeke, M., Gottschalk, G. & We would like to thank our collegues at BRAIN AG and particu-
Science 303, 1796–1799 (2004). Daniel, R. Screening of environmental DNA libraries for larly C. Schleper of the University of Bergen, Norway, and the
47. Leeb M. A shot in the arm. Nature 431, 892–893 (2004). the presence of genes conferring lipolytic activity on Zentrum für Evolutionsforschung und Biotechnologie (ZEB),
48. Cragg, G. M. & Newman, D. J. Medicinals for the Escherichia coli. Appl. Environ. Microbiol. 66, 3113–3116 Darmstadt, Germany, for fruitful discussions and sharing ideas
millennia: the historical record. Ann. N. Y. Acad. Sci. 953, (2000). and resources.
3–25 (2001). 64. Rondon, M. R. et al. Cloning the soil metagenome: a
49. Newman, D. J., Cragg, G. M. & Snader, K. M. Natural strategy for accessing the genetic and functional diversity Competing interests statement
products as sources of new drugs over the period of uncultured microorganisms. Appl. Environ. Microbiol. The authors declare competing financial interests; see Web
1981–2002. J. Nat. Prod. 66, 1022–1037 (2003). 66, 2541–2547 (2000). version for details.
50. Osburne, M. S., Grossman, T. H., August, P. R. & 65. Knietsch, A., Waschkowitz, T., Bowien, S., Henne, A. &
MacNeil, I. A. Tapping into microbial diversity for natural Daniel, R. Metagenomes of complex microbial consortia
products drug discovery. ASM News 66, 411–417 derived from different soils as sources for novel genes Online links
(2000). conferring formation of carbonyls from short-chain
51. Daniel, R. The soil metagenome — a rich resource for the polyols on Escherichia coli. J. Mol. Microbiol. Biotechnol. FURTHER INFORMATION
discovery of novel natural products. Curr. Opin. 5, 46–56 (2003). ActivBiotics: www.activbiotics.com
Biotechnol. 15 199–204 (2004). 66. Knietsch, A., Waschkowitz, T., Bowien, S., Henne, A. & Aventis Pharmaceuticals: www.aventispharma-us.com
52. Seow, K. T. et al. A study of iterative type II polyketide Daniel, R. Construction and screening of metagenomic BRAIN AG: www.brain-biotech.de
synthases, using bacterial genes cloned from soil DNA: a libraries derived from enrichment cultures: generation of a Cubist Pharmaceuticals: www.cubist.com
means to access and use genes from uncultured gene bank for genes conferring alcohol oxidoreductase EuropaBio: www.europabio.org
microorganisms. J. Bacteriol. 179, 7360–7368 (1997). activity on Escherichia coli. Appl. Environ. Microbiol. 69, The European Chemical Industry Council:
53. Wang, G. Y. et al. Novel natural products from soil DNA 1408–1416 (2003). www.cefic.org
libraries in a streptomycete host. Org. Lett. 2, 2401–2404 67. Gabor, E. M., de Vries, E. J. & Janssen, D. B. The European Technology Platform for Sustainable
(2000). Construction, characterization, and use of small-insert Chemistry: www.suschem.org
54. Pfeifer, B. A., Admiraal, S. J., Gramajo, H., Cane, D. E. & gene banks of DNA isolated from soil and enrichment Degussa: www.degussa.com
Khosla, C. Biosynthesis of complex polyketides in a cultures for the recovery of novel amidases. Environ. Diversa: www.diversa.com
metabolically engineered strain of E. coli. Science 291, Microbiol. 6, 948–958 (2004). Genencor International: www.genencor.com/wt/home
1790–1792 (2001). 68. Gabor, E. M. Harvesting novel biocatalysts from the Henkel: www.henkel.com
55. Courtois, S. et al. Recombinant environmental libraries metagenome. Thesis, Univ. Groningen (2004). Lucigen: www.lucigen.com
provide access to microbial diversity for drug discovery 69. Entcheva, P., Liebl, W., Johann, A., Hartsch, T. & Prokaria: www.prokaria.com
from natural products. Appl. Environ. Microbiol. 69, Streit, W. R. Direct cloning from enrichment cultures, a Vicuron Pharmaceuticals: www.vicuron.com
49–55 (2003). reliable strategy for isolation of complete operons and Access to this interactive links box is free online.

516 | JUNE 2005 | VOLUME 3 www.nature.com/reviews/micro

© 2005 Nature Publishing Group