Sampling Terrestrial Environments for Bacterial Polyketides

Abstract

Bacterial polyketides are highly biologically active molecules that are frequently used as drugs, particularly as antibiotics and anticancer agents, thus the discovery of new polyketides is of major interest. Since the 1980s discovery of polyketides has slowed dramatically due in large part to the repeated rediscovery of known compounds. While recent scientific and technical advances have improved our ability to discover new polyketides, one key area has been under addressed, namely the distribution of polyketide-producing bacteria in the environment. Identifying environments where producing bacteria are abundant and diverse should improve our ability to discover (bioprospect) new polyketides. This review summarizes for the bioprospector the state-of-the-field in terrestrial microbial ecology. It provides insight into the scientific and technical challenges limiting the application of microbial ecology discoveries for bioprospecting and summarizes key developments in the field that will enable more effective bioprospecting. The major recent efforts by researchers to sample new environments for polyketide discovery is also reviewed and key emerging environments such as insect associated bacteria, desert soils, disease suppressive soils, and caves are highlighted. Finally strategies for taking and characterizing terrestrial samples to help maximize discovery efforts are proposed and the inclusion of non-actinomycetal bacteria in any terrestrial discovery strategy is recommended.

Keywords: bioprospecting; microbial ecology; polyketides.

Figure 1

Microbial ecology approaches used to examine both cultivatable and uncultivatable fraction of the microbial community structure. This includes molecular methods like shotgun metagenome sequencing, community fingerprinting of polymerase chain reaction (PCR) products, sequencing of PCR products, and phospholipid fatty acid analysis.

Figure 2

The overlap between Sanger sequencing results of three 16S clone libraries at 99% identity from a beach sand using two DNA extraction methods and two primer pairs. Clone library 1 Extraction method 1 PCR primer pair 1, Clone library II Extraction method 2 PCR primer pair 1, Clone library III II Extraction method 1 PCR primer pair 2. Reprinted by permission from Macmillan Publishers Ltd: ISME J. [45], copyright (2009).

Figure 3

Bacterial 16S 454 pyrosequencing results from 88 non-agricultural soils from across North and South America. Reproduced with permission from Lauber et al., Appl. Environ. Microbiol., published by the American Society for Microbiology, 2009 [71].

Figure 4

Forward Terminal Restriction Fragment Length Polymorphism (T-RFLP) community fingerprints of the Actinobacteria from Uncultivated soil (blue) Cultivated soils (green), animal associated sediments (orange) and street dust (red). Reproduced from Hill et al., Microb. Ecol., published by Springer International Publishing AG., 2011 [46].

Figure 5

Overlap of sequences amplified from three soils from the desert sites in Arizona (green), Utah (blue) and California (red). Sequence are from non-ribosomal polypeptide (NRP) Adenylation (AD) domains, type I polyketide synthase PKS ketosynthase domains (KS), type II PKS alpha ketosynthase domains (KSα) and 16S ribosomal sub units. Adapted with permission from Reddy et al., Appl. Environ. Microbiol., published by American Society for Microbiology, 2012 [120].

Figure 6

(A) Leaf cutter ant colonies in Texas. Photo courtesy of Texas A&M AgriLife Extension/Josh Blanek; (B) Leaf cutting ant nest in Costa Rica. In this case the nest was exposed when a rain barrel was moved, normally they are found at greater depth. Photograph courtesy of Herster Barres, Reforest the Tropics.

Figure 7

Polyketide and non-ribosomal peptide natural products isolated from bacteria associated with social ants.

Figure 8

Microtermolides, produced by termite-associated microbes, and the originally proposed and revised structure of the related NRP vinylamycin.

Figure 9

The known antibiotic geldanamycin and the related analogs discovered from a termite associate Streptomyces.

Figure 10

Polyketides discovered from beetle associated bacteria and known related compounds.

Figure 11

Beewolf with honey bee prey. Courtesy of Simon Jenkins http://www.simon-jenkins.photography.

Figure 12

Bacterial secondary metabolites associated with non-social wasp species.

Figure 13

Tripartilactam and pederin from non-social beetles and related natural products from non-insect associated microbes.

Figure 14

Type II polyketides found in metagenomic libraries from desert soils and their ring systems. Reproduced with permission from Feng et al., Proc. Natl. Acad. Sci. USA., published by United States National Academy of Sciences, 2011 [187].

Figure 15

Pentangular polyphenols discovered from desert soil derived metagenomic libraries.

Figure 16

Natural products from desert isolates.

Figure 17

Natural products from disease suppressive soils.

Figure 18

Cervimycins, isolated from cave derived Actinomycetes, are related to known tetracycline antibiotics 74 and 75.

Figure 18

Cervimycins, isolated from cave derived Actinomycetes, are related to known tetracycline antibiotics 74 and 75.

Figure 19

Type II polyketides isolated from an Actinomycetes isolate from Hardin’s cave.

Figure 20

(A) An example of Moonmilk (above) in Goatherds Chasm in Switzerland. Photograph provided by Olivier Gallois of the Groupe Spéléologique Archéologique Mandeure; (B) Yellow Microbial mats from the volcanic cave Gruta de Terra Mole in the Azores. Photo courtesy of Pedro Cardoso. Reproduced from Riquelme et al., Front. Microbiol., 2015 [248].

Figure 21

Polyketides discovered from cultured Actinomycete strains isolated from highly acidic mine drainage and highly basic mine tailings.

Figure 22

Polyketides identified from Actinopolyspora erythraea.

Figure 23

Average percent composition at the phylum-level classification of bacterial communities at the soil surface or at 1.5, 4.5, 7.5, or 18 m above the surface. Reproduced from Weber and Werth, Front. Microbiol., 2015 [260].

Figure 24

An example of actinobacterial 16S fingerprinting. Forward and reverse Terminal Restriction Fragment Length Polymorphism (T-RFLP) from a Sanger sequencer for a Russian taiga forest soil (Rusforest-Yenisei). PCR products of 16S actinobacterial specific primers were labelled with the dyes hexachloro-6-carboxyfluorescein (blue, forward) and carboxyfluorescein (green, reverse). Red peaks are the ROX 1000 size standards. Size is shown on the X axis in bp, fluorescence on the Y axis. The size range 81–677 bp was used for clustering analysis of forward T-RFLP patterns from a range of samples (shown in Figure 4).