Photosynth Res DOI 10.1007/s11120-014-0066-9 EMERGING TECHNIQUES Challenges of metagenomics and single-cell genomics approaches for exploring cyanobacterial diversity Michelle Davison • Eric Hall • Richard Zare • Devaki Bhaya Received: 27 June 2014 / Accepted: 10 December 2014 Ó Springer Science+Business Media Dordrecht 2014 Abstract Cyanobacteria have played a crucial role in the optimized, setting a stage for further investigations of history of early earth and continue to be instrumental in individual cyanobacterial cells isolated directly from nat- shaping our planet, yet applications of cutting edge tech- ural populations. nology have not yet been widely used to explore cyano- bacterial diversity. To provide adequate background, Keywords Multiple displacement amplification (MDA) Á we briefly review current sequencing technologies and Whole genome amplification (WGA) Á Single cell Á their innovative uses in genomics and metagenomics. Next, Microfluidics Á Cyanobacteria Á CRISPR we focus on current cell capture technologies and the challenges of using them with cyanobacteria. We illus- trate the utility in coupling breakthroughs in DNA ampli- Background fication with cell capture platforms, with an example of microfluidic isolation and subsequent targeted amplicon Cyanobacteria represent an ancient, diverse, and ecologi- sequencing from individual terrestrial thermophilic cya- cally important phylum. They are ubiquitous in both ter- nobacteria. Single cells of thermophilic, unicellular Syn- restrial and marine environments (Coelho et al. 2013; Shih echococcus sp. JA-2-3-B0a(2-13) (Syn OS-B0) were sorted et al. 2013) and are of significant interest from a variety of in a microfluidic device, lysed, and subjected to whole viewpoints. Evolutionarily, cyanobacteria played a crucial genome amplification by multiple displacement amplifi- role in oxygenation of the early earth atmosphere (Hoehler cation. We amplified regions from specific CRISPR spacer et al. 2001), and were the first pioneers in primary endo- arrays, which are known to be highly diverse, contain symbiosis, giving rise to modern day plastids (Larkum semi-palindromic repeats which form secondary structure, et al. 2007). Ecologically, cyanobacteria are primary pro- and can be difficult to amplify. Cell capture, lysis, and ducers, performing oxygenic photosynthesis in a wide genome amplification on a microfluidic device have been range of associations and symbiotic relationships (Paerl et al. 2000; Lesser et al. 2004; Bergman et al. 2007), and play critical roles in global nitrogen and carbon cycles (Karl et al. 2012; Bar-Zeev et al. 2013). Environmental M. Davison (&) Á D. Bhaya Department of Plant Biology, Carnegie Institution of Science, nutrient imbalances lead to toxic blooms of cyanobacteria 260 Panama Street, Stanford, CA 94305, USA and freshwater eutrophication (Oliver 2012; McMahon and e-mail: [email protected] Read 2013). More recently, cyanobacteria have been examined for their potential in waste water remediation E. Hall SRI International, 333 Ravenswood Ave, Menlo Park, (Martins et al. 2011; Olguı´n 2012), and as an efficient CA 94025, USA chassis in biotechnology (Ducat et al. 2011; Hess 2011; Berla 2013). Although the ecology, physiology, and R. Zare molecular biology of cyanobacteria have all been exten- Department of Chemistry, Stanford University, 333 Campus Drive Mudd Building, Room 121, Stanford, CA 94305-4401, sively studied for many decades (Gupta and Carr 1981; Fay USA 1992; Robinson et al. 1995), it is only in the last few years 123 Photosynth Res that new genome sequencing technologies and single-cell metagenomics revealed the existence of specialized capture technologies have made a major impact in our cyanobacterial populations containing genes for specific understanding of cyanobacterial diversity. Many of these metabolic functions (Bhaya et al. 2007; Kashtan et al. technologies have been optimized for the study of model 2014). organisms such as E. coli or yeast, so there are specific challenges in their use with cyanobacteria and microbial Sequencing cyanobacterial isolates populations. To fully appreciate these issues, we begin with a mini-review of the major strategies that are currently used The focus on culturing of axenic strains and genome to capture genomic diversity. sequencing is widespread (Laloui et al. 2002; Parkhill and In the second section, we describe in greater detail the Wren 2011). However, the ability to isolate a strain in the methods available for single-cell capture and subsequent laboratory can often skew the distribution of sequenced sequencing strategies. Because such methods have not been individuals. Within the phylum of Cyanobacteria, almost widely used with cyanobacteria, we describe the optimi- 40 % of the currently sequenced cyanobacterial genomes zation of a protocol by which single cyanobacterial cells cluster within the marine Synechococcus/Prochlorococcus were sorted in a microfluidic (‘‘chip’’) device followed by subclade isolated from various oceanic locations and lysis and whole genome amplification by on-chip multiple depths. This has provided a rich source of genomic-based displacement amplification (MDA). We focused on experimental data which have emerged from a number of amplification from specific regions of the genome includ- laboratories (Scanlan et al. 2009; Flombaum et al. 2013; ing regions from specific CRISPR spacer arrays, which are Mackey et al. 2013; Thompson et al. 2013; Axmann et al. highly diverse and are part of the recently identified 2014; Kashtan et al. 2014). On the other hand, until adaptive immune response. By doing so, we provide a recently, there was a conspicuous lack of genome proof of concept study that suggests that this pipeline has sequences from the diverse morphologies that represent the the potential to be used for the study of natural diversity in deeply branched cyanobacterial lineages. To address this, a cyanobacterial populations. major collective effort was made to sequence the full There are currently three primary strategies for acquir- genomes of a greater range of cyanobacterial lineages to ing sequence data to capture cyanobacterial diversity: capture phylogenetic and phenotypic diversity (Shih et al. 2013). Fifty-four strains covering all five morphological (i) classical isolation methods, by which axenic subsections, in addition to a range of lifestyles and strains are first isolated from environmental sam- metabolisms, were sequenced using isolates primarily from ples, followed by DNA extraction and sequencing the Pasteur collection (Shih et al. 2013). Based on this to get complete genome sequences study, over 21,000 novel proteins, with no homology to (ii) metagenomics, a culture-independent means by known proteins, were discovered, as well as the longest which total DNA is directly extracted from collection of CRISPR spacer-repeat units, 650, ever char- environmental samples and sequenced acterized in cyanobacteria. By sampling across a wide (iii) single-cell methods, by which individual cells are phylogenetic distribution, the groundwork has been laid for isolated (either directly from the environment or cyanobacteria to emerge as a powerful comparative geno- from axenic populations) and DNA is extracted mic model system (Shih et al. 2013). for amplification and subsequent sequencing. Initial sequencing projects focused on acquiring the Sequencing metagenomes genomes of axenic microbes from available culture col- lections, with Haemophilus influenzae being the first bac- In an alternate approach, metagenomic surveys have terial genome ever to be sequenced (Fleischmann et al. uncovered cyanobacterial signatures across many habitats, 1995) and the unicellular cyanobacterium Synechocystis sp. including many biofilm assemblages such as arctic mats PCC6803 being the second (Kaneko et al. 1996). However, and sub-zero sediments (Varin et al. 2012; Lay et al. 2013), with the advent of next generation sequencing, the glacial streams (Wilhelm et al. 2013), thermophilic hot choice of sequencing strategies and platforms has begun to springs (Bhaya et al. 2007; Heidelberg et al. 2009; Klatt have a significant impact on the sorts of questions that et al. 2011 #, Ionescu et al. 2010; Mackenzie et al. 2013), can be addressed. As there are advantages and problems non-thermophilic mats (Germa´n Bonilla-Rosso 2012; Kirk associated with each method (Fig. 1), it is also recom- Harris et al. 2013; Lindemann 2013; Mobberley et al. mended to use these methods in combination to mediate 2013) coastal mats (Balskus et al. 2011; Burow et al. 2013), the shortcomings of each individual sequencing strategy. as well as free-living populations, such as those in oligo- For example, the use of fully sequenced genomes as trophic ocean waters (Shi et al. 2011; Malmstrom et al. ‘‘anchors’’ or ‘‘references’’ in conjunction with population 2013), the Red Sea (Thompson et al. 2013), soil crusts 123 Photosynth Res Mixed Population Isolation of Sample Dilution/FACs/ DNA Extraction co- Micromanipulation/ cultures/ax Microfluidics enic cultures Total DNA Enrichment Single cells Environment DNA Extraction DNA Extraction WGA Library Construction Library Construction DNA Sequencing DNA Sequencing Metagenome Genome(s) Genome Pros: Pros: Pros: •Community-wide survey of •Entire genomic context •Entire genomic content gene content and diversity preserved and SNPs for an individual •Each read represents a cell single individual Cons: •Consensus genome may Cons: Cons: “mask” diversity •Partial genome coverage •Loss of entire