Development of a universal microarray based on the ligation detection reaction and 16S rrna gene polymorphism to target diversity of cyanobacteria

Abstract

The cyanobacteria are photosynthetic prokaryotes of significant ecological and biotechnological interest, since they strongly contribute to primary production and are a rich source of bioactive compounds. In eutrophic fresh and brackish waters, their mass occurrences (water blooms) are often toxic and constitute a high potential risk for human health. Therefore, rapid and reliable identification of cyanobacterial species in complex environmental samples is important. Here we describe the development and validation of a microarray for the identification of cyanobacteria in aquatic environments. Our approach is based on the use of a ligation detection reaction coupled to a universal array. Probes were designed for detecting 19 cyanobacterial groups including Anabaena/Aphanizomenon, Calothrix, Cylindrospermopsis, Cylindrospermum, Gloeothece, halotolerants, Leptolyngbya, Palau Lyngbya, Microcystis, Nodularia, Nostoc, Planktothrix, Antarctic Phormidium, Prochlorococcus, Spirulina, Synechococcus, Synechocystis, Trichodesmium, and Woronichinia. These groups were identified based on an alignment of over 300 cyanobacterial 16S rRNA sequences. For validation of the microarrays, 95 samples (24 axenic strains from culture collections, 27 isolated strains, and 44 cloned fragments recovered from environmental samples) were tested. The results demonstrated a high discriminative power and sensitivity to 1 fmol of the PCR-amplified 16S rRNA gene. Accurate identification of target strains was also achieved with unbalanced mixes of PCR amplicons from different cyanobacteria and an environmental sample. Our universal array method shows great potential for rapid and reliable identification of cyanobacteria. It can be easily adapted to future development and could thus be applied both in research and environmental monitoring.

FIG. 1.

Main features of LDR method coupled with a universal microarray. After hybridization of a discriminating probe and a common probe to the target sequence (16S rRNA gene), ligation occurs only if there is perfect complementarity between the two probes and the template (A). The reaction is thermally cycled, generating single-stranded DNA fragments bearing a 5′ Cy3 fluorescent moiety and a 3′ czip code sequence. The cycling allows more common probe (and the corresponding czip code) to ligate to the discriminating probe, given a fixed amount of PCR target. (B) The LDR product is hybridized to a universal microarray, where unique zip code sequences have been spotted.

FIG. 2.

Partial alignment of group-specific consensus sequences and an example of probe selection for Microcystis. The discriminating probe is indicated by light gray box, and the common probe is indicated by an unshaded box. The important base (A) at the 3′ end of the discriminating probe is underlined.

FIG. 3.

NJ tree based on 16S rRNA gene sequences showing the 19 cyanobacterial groups. The probes used in the microarray were designed according to these groups. The P. marinus group is embedded in the large Synechococcus group. The tree contained 338 cyanobacterial 16S rRNA gene sequences. This phylogenetic tree stability has been supported by bootstrap analysis. For bootstrap analysis, 500 resamplings were performed by using the NJ algorithm in ARB. The bootstrap tree is available at http://www.ulg.ac.be/cingprot/midichip/output/publications/Castiglioni_Tree.htm.

FIG. 4.

Deposition scheme and several examples of LDR-universal array results. On the figure, the slide with eight arrays (top left corner), the deposition scheme of an array (top right corner), and a table specifying the cyanobacterial groups and the corresponding zip codes (bottom right corner) are shown. The hybridization-positive control is indicated by light gray shading, and the UNICYANO probe is indicated by boldface type. Each cyanobacterial group has four replicate spots. Hybridization results of the amplified 16S rRNA gene from the strains are shown in the bottom left corner. (A) Aphanizomenon sp. strain 202; (B) Calothrix sp. strain PCC 7714; (C) M. aeruginosa strain PCC 9354; (D) Plankthotrix sp. strain 1LT27S08; (E) S. major strain PCC 6313; (F) Synechococcus sp. strain Hegewald 74-30.

FIG. 5.

Testing of LDR sensitivity. (A) Correlation between signal intensity and template concentration. The effect of template concentration on LDR was tested with PCR products of Planktothrix sp. strain 1LT27S08, Calothrix sp. strain PCC 7714, and M. aeruginosa strain PCC 9354, ranging from 0.5 to 100 fmol. (B) Signal-to-noise ratio plotted against template quantity. Each data point represents the ratio between the mean signal intensity of the target-specific zip codes and the mean signal intensity of nonspecific zip codes. PCR products from Planktothrix sp. strain 1LT27S08 (1 to 100 fmol) were used as a template. The signal-to-noise ratio increases with growing template concentrations.

FIG. 6.

Microarray analyses of complex cyanobacterial samples. (A) The hybridization result of the unbalanced LDR mix shows how spot intensity (left) and measured fluorescence intensity (right) correspond to the concentrations of targets. The LDR mix contained 100 fmol of the PCR product from both M. aeruginosa strain PCC 9354 and Aphanizomenon sp. strain 202 and 5 fmol of the PCR product of both S. major strain PCC 6313 and Calothrix sp. strain PCC 7714. (B) The environmental sample 0TU27 from Lake Tuusulanjärvi (Finland) was analyzed with the array. The hybridization pattern shows the presence of Microcystis, Anabaena/Aphanizomenon, and Woronichinia spp.