compounds in

The marine environment is estimated to be one of the most significant sources of biological activity in the world. In the last few decades an increase in the research intensity conducted on marine microorganisms has been observed, which confirms the great potential of these organisms in the field of bioactive compounds' production. In order to efficiently use the natural resources of the marine environment, metagenomics can be applied. This powerful technique allows for efficient screening of microbial biodiversity for bioactive compounds. The primary aim of this review is to present some aspects of the construction of metagenomic libraries, and strategies of screening for novel bioactives in the marine surrounding. This paper also illustrates several examples of the application of metagenomic methods in the discovery of novel enzymes and drugs in various marine environments.


BACKGROUND
Oceans and seas that cover more than 70% of the Earth's surface are now one of the most mysterious environments in the world (Zhao, 2011).They were formerly considered to be uninhabited by any life forms (Zobell, 1946).However, recent studies have proven that oceans constitute a rich source of living microorganisms, with cell counts of 10 6 -10 9 cells per milliliter (Rheinheimer, 1992;Fenical & Jensen, 2006).Extreme conditions like low temperatures and high pressures, which are present in the marine habitat, support the production of bioactive compounds (bioactives).The ability of marine microorganisms to adapt to the difficult life conditions has resulted in the development of unique biochemical features which may find their application in virtually all industries (Andersen et al., 1974).Simultaneously, adjustment to the challenges of the deep sea has led to greater adaptability in both signaling pathways involved in the detection of nutrients and greater metabolic capabilities in utilizing them (Kennedy et al., 2010).Furthermore, the marine surfaces are proven to be a significant source of bioactives due to the existing competition between microorganisms associated on the surfaces of eukaryotes (Penesyan et al., 2010).
The last ten years have exposed a new tendency in the exploration of the microorganisms' biodiversity, which is based on culture-independent techniques and the examination of the complex microbial population for basic and applied research (Konig et al., 2006).In the study of microbial communities, this approach was developed in order to supplement or replace culture-based technologies, especially since nowadays it is commonly known that only 1-5% of the consortia of living microorganisms can be cultivated under laboratory conditions (Handelsman, 2004).A crucial method of exploring the DNA in uncultivated microorganisms is metagenomics.The microbial DNA isolated from the sample obtained from natural habitats is called the metagenome (Mitra et al., 2010).This sample of DNA theoretically represents all microbes contained in the environmental sample, whose number can vary from several hundred to several thousand of species in a single metagenome (Raes et al., 2007;Gurgui & Piel, 2010;Thomas et al., 2012).

DNA isolation and selection of the vector system
The first step in planning the construction of a metagenomic library is the selection of a vector system.This choice is related to the quality of the extracted environmental DNA, which can be described by means of four parameters: yield, purity, fragment size and representativeness (Morgan et al., 2009;Ekkers et al., 2012).In practice, it is very difficult to achieve satisfactory values of all these factors due to the negative effect that one may inflict on the other.For instance, high yield of DNA is often related to an increased contamination and low average fragment size of the DNA sample.Consequently, small average DNA fragment size limits the analysis of larger operons, in the case of which capital DNA fragments are required (Williamson et al., 2005).
The isolation of DNA from a specific environment, which is destined to be used to construct a metagenomic library, can be performed by using one of the following strategies: direct or indirect extraction of the DNA from the environmental sample.The former approach is based on the isolation of the DNA directly from the sample, while the latter requires a prior selection of the cells of interest (Foote et al., 2012).The indirect method of environmental DNA isolation was proposed by Faegri et al. (1977) and applies four steps: dispersion of the en-vironmental sample, cell separation, cell lysis and DNA isolation.However, Roh et al. (2006) have proven that direct isolation methods allow for receiving a hundred fold higher amount of DNA on average than the indirect ones.
Occasionally standard techniques of DNA isolation are not sufficiently effective in obtaining the satisfactory yield of DNA for further analysis.In these cases in order to enhance the DNA weight, pre-amplification is required, by means of the multiple displacement amplification (MDA) technique (Woyke et al., 2010).This technique permits whole genome amplification of small amounts of DNA.Spits et al. (2006) developed the threehour procedure of MDA-based-amplification of DNA from a single cell with the use of Phi 29 polymerase to generate 1-2 g of DNA.A possible disadvantage of MDA is the risk of artifact inherence, such as chimeric rearrangements (Neufeld et al., 2008, Gilbert et al., 2010).
The size of the obtained DNA fragment is crucial in the selection of the vector system used in the construction of the metagenomic library.Small-size metagenomic libraries require the DNA sample of about 10 kb and exploit the high-copy-number plasmids with leaky-promoters such as pUC19 (Handelsman, 2004;Godiska, 2009).DNA fragments between 20 to 40 kb require to be inserted in fosmids or cosmids (Hallam et al., 2004;Kim et al., 2006).Bacterial artificial chromosomes (BACs) are commonly used to construct metagenomic libraries with the 100-200 kb DNA fragments (Rondon et al., 2000;Uchiyama & Miyazaki, 2009).

Selection of the host(s)
A standard host in most studies based on metagenomic techniques is Escherichia coli.However, it is predicted that the use of E. coli as the host in metagenomic library construction allows to express only 40% of genes contained in the eDNA of an average sample (Craig et al., 2010;Parachin & Gorwa-Grauslund, 2011;McMahon et al., 2012).Due to this limitation, an alternative host strain from Bacillus, Pseudomonas or Streptomyces genera can be applied (Lorenz & Eck, 2005;Aavick et al., 2009).There are also few archaeal genera (Methanococcus, Pyrococcus, Sulfolobus, Thermococcus), which have been successfully employed in designing of the stable host-vector expression system (Angelov & Liebl, 2010).
Another strategy is the application of broad-hostrange vectors, which is related to the concept that a significant number of genes presented in a metagenomic library cannot be expressed in a single host.Thus, this approach may have a positive effect on the detection frequency of genes that confer novel functions (Simon et al., 1983;Wexler et al., 2005).For instance, the application of bacterial genera such as Pseudomonas, Rhizobium or Streptomyces, which have over 15 RNA polymerase σ factors (Escherichia coli, which is the most commonly used host in the metagenomic library construction has only seven), may be crucial in the expression of genes that require specialized σ factors (Gabor et al., 2004b).

SCREENING STRATEGY
When the metagenomic library is constructed, a screening strategy should be selected.There are two fundamentally different approaches -activity-based screening (alternatively named functional screening) and sequence-based screening -that may be applied to detect novel bioactives in metagenomic libraries (Fig. 1).Selection of a screening method is connected with vari-ous factors such as knowledge about the sequence similarity of the target gene to known genes, type of desired gene product activity, or availability of rapid methods of identifying the transformants of interest (Knietsch et al., 2003;Yun & Ryu, 2005).

Activity-based screening
Functional screening of metagenomic libraries exploits the direct detection of activity of interest.The screen of a metagenomic library is not dependent on the sequence information or sequence similarity of known genes (Lorenzo, 2005).However, this method, being distinct from the sequence-based screening, allows for the discovery of new classes of genes that encode both known and new activities (Suenaga et al., 2007).This approach is strictly bounded by the limitation of eDNA expression in a selected host.All the investigated genes in a given DNA fragment are expected to be expressed independently of its size and structure (Mori et al., 2008).However, there are many factors, which influence an efficient expression: recognition of promoters given in the gene library, regulatory agents of the transcription system of the host, toxicity of gene products, codon usage differences (CUB), correct protein folding, presence of proper initiation factors, or the capacity of the host to secret the gene expression product (Lorenz et al., 2002;Craig et al., 2010).Moreover, activity-based screening methods are directed at the discovery of already known compounds and limited by available assays and detection sensitivity (Kowalchuk et al, 2007).
The main advantage of applying this strategy of investigation of novel bioactives is that this approach allows to identify the inherence of a gene product, which cannot be detected by means of the sequence-based method due to the absence of, or insufficient homology to known genes.Furthermore, there is an assurance that the gene product is correctly expressed, which hastens the process of protein characterization (Jones et al., 2007;Park et al., 2007).
The activity-based analysis of eDNA in a metagenomic library might be divided into three different strategies: (1) direct detection of gene products in individual clones; (2) heterologous complementation of host strains or mutants; (3) induced gene expression (Rolf, 2005;Ferrer et al., 2005).
The first approach to the investigation of enzymatic activities of positive clones commonly applies chromophore-containing derivatives of enzyme substrates, which allow for observing the enzymatic reaction.For example, Lämmle et al. (2007) have successfully employed the fluorescent methods in the detection of novel lipolytic enzymes and phosphatases.
The second type of metagenomic library screening is based on the use of the host, which requires heterologous complementation by target genes to grow under selective conditions (Carola & Rolf, 2011).This technique is designed for screens of specific bioactive compounds, provided the gene product is in an active form.As an example for this strategy, Riesenfeld et al. (2004) successfully applied this method in the discovery of nine clones expressing resistance to aminoglycoside antibiotics, and one expressing resistance to tetracycline.
A different approach is presented in substrate-induced gene expression (SIGEX) screening of gene libraries designed by Uchiyama and Miyazaki (2009).This method is based on the identification of positive clones with a reporter system, which "traps" genes encoding biodegradative pathways located in the "operon-trap" vector.The screen of eDNA clones is connected with the assumption that operons encoding these pathways may be induced by the substrate of the pathway.Reported gene is immediately included downstream of the cloning site for the DNA insert, hence the expression of the insert gene is coupled with that of the reporter gene.The system is addressed especially for screening the catabolic genes which induction is connected with metabolically relevant compounds.Positive clones are detected by a reported signal coming from the fluorescence-activated cell sorter (FACS) in combination with a fluorescent reporter protein, such as the green fluorescent protein (GFP), which production is induced by the substrate of interest (Handelsman, 2005).

Sequence-based screening
Two different strategies of screening of gene libraries are employed in the sequence-based approach: (1) PCRbased and (2) hybridization-based techniques (Weiland et al., 2010).Both methods use the acquirement about conserved DNA sequences present in Databases.Thus, knowledge about the conserved domains of genes is the main limitation of the sequence-based screens.Additionally, the detection of sequence of interest does not guarantee an efficient expression of the target gene.Consequently, this approach is successfully applied in studies based on biodiversity research of environmental samples and identification of exact strains using the bacterial gene encoding 16S rRNA.
However, there are some examples proving that the sequence-based screening of gene libraries may be used successfully in investigating new bioactive compounds.Park et al. (2008) created a format of microarray (metagenome microarray technique -MGA), which is arrayed with fosmid library clone DNA samples on a glass slide.They evaluated the MGA applying random prime labeled fluorescent probes prepared using the PCR products of the target gene.The observation was that it is possible to obtain specific hybridization signals only for the DNA fragments that contained the target gene.

METAGENOMICS AND MARINE HABITATS
Microbial communities inhabiting the marine environment have evolved during extended evolutionary processes of physiological adaptation under the influence of a variety of selection pressures.It is estimated that marine habitat is nowadays one of the richest sources of bioactive compounds due to the diversity of metabolically complex microorganisms (Schloss & Handelsman, 2003).During the last five decades natural products from the marine environment have become a fascinating target for scientists in the discovery of lead compounds applied in medicine and industry (Cooper, 2004;Jones et al. 2009;Mayer et al., 2011).Recent examples of marine enzymes that can be potentially applied in different industries, such as the food industry (amylase), chemical industry (lipase, protease, esterase) or agriculture (cellulase), are presented in Table 1.

CONCLUSIONS
Oceans and seas appear to be especially rich in bioactive compounds produced by various marine organisms.Since most of marine microbes are not cultivable, metagenomic approaches can be considered as very promising in discovering previously unknown applicable compounds of important biological activities.It is likely that further research will provide new fascinating discoveries in this field.

Figure 1 .
Figure 1.Scheme of a gene library construction and approaches of the discovery of bioactive compounds from the marine environment.