Molecular fingerprinting of cyanobacteria from river biofilms as a water quality monitoring tool

Abstract

Benthic cyanobacterial communities from Guadarrama River (Spain) biofilms were examined using temperature gradient gel electrophoresis (TGGE), comparing the results with microscopic analyses of field-fixed samples and the genetic characterization of cultured isolates from the river. Changes in the structure and composition of cyanobacterial communities and their possible association with eutrophication in the river downstream were studied by examining complex TGGE patterns, band extraction, and subsequent sequencing of 16S rRNA gene fragments. Band profiles differed among sampling sites depending on differences in water quality. The results showed that TGGE band richness decreased in a downstream direction, and there was a clear clustering of phylotypes on the basis of their origins from different locations according to their ecological requirements. Multivariate analyses (cluster analysis and canonical correspondence analysis) corroborated these differences. Results were consistent with those obtained from microscopic observations of field-fixed samples. According to the phylogenetic analysis, morphotypes observed in natural samples were the most common phylotypes in the TGGE sequences. These phylotypes were closely related to Chamaesiphon, Aphanocapsa, Pleurocapsa, Cyanobium, Pseudanabaena, Phormidium, and Leptolyngbya. Differences in the populations in response to environmental variables, principally nutrient concentrations (dissolved inorganic nitrogen and soluble reactive phosphorus), were found. Some phylotypes were associated with low nutrient concentrations and high levels of dissolved oxygen, while other phylotypes were associated with eutrophic-hypertrophic conditions. These results support the view that once a community has been characterized and its genetic fingerprint obtained, this technique could be used for the purpose of monitoring rivers.

Fig 1

Physicochemical parameters measured at each sampling point in the Guadarrama River during 1 year. DIN, dissolved inorganic nitrogen; SRP, soluble reactive phosphorus. Error bars indicate standard deviation (n = 3). Sampling sites: 1, white bars; 2, gray bars; 3, black bars.

Fig 2

Comparative analysis of TGGE genetic profiles of PCR-amplified fragments of 16S rRNA genes obtained from epilithic biofilms collected from the Guadarrama River. In the gels, three sampling points (1, 2, and 3) with three replicates (stones a, b, and c) during the different climatological seasons are represented. TGGE gels obtained with primer set CYA359F and CYA781R(b) (A) and using primers CYA359F and CYA781R(a) (B) are shown. A mixture of PCR products derived from five cyanobacterial strains isolated from the epilithon was applied as a reference marker (M) (from top to bottom: Aphanocapsa muscicola UAM 385, Chamaesiphon investiens UAM 386, A. rivularis UAM 390, Leptolyngbya boryana UAM 391, and Cyanobium sp. UAM 406).

Fig 3

UPGMA dendrogram obtained from the cluster analysis performed among sampling sites over 1 year, based on the TGGE bands. Seasons: A, autumn; W, winter; Sp, spring; S, summer. Sampling sites: 1, white circles, 2, gray circles; 3, black circles. Replicates: stones a, b, and c. Dark vertical lines correspond to clustered sites at 50% similarity.

Fig 4

Relationship between TGGE band richness and (A) dissolved inorganic nitrogen (DIN), (B) soluble reactive phosphorus (SRP), and (C) conductivity.

Fig 5

Neighbor-joining trees representing (A) unicellular cyanobacteria and (B) filamentous cyanobacteria and plastids based on the analysis of the 16S rRNA gene, showing the TGGE band sequences obtained in the present study (in bold). Numbers near nodes indicate bootstrap values greater than or equal to 60% in neighbor-joining (NJ), maximum-likelihood (ML), and maximum-parsimony (MP) analyses. TGGE bands were named according to the climatological seasons (A, autumn; W, winter; Sp, spring; S, summer), sampling sites (1 [white circles], 2 [gray circles], and 3 [black circles]), and replicates (stones a, b, and c).

Fig 5

Neighbor-joining trees representing (A) unicellular cyanobacteria and (B) filamentous cyanobacteria and plastids based on the analysis of the 16S rRNA gene, showing the TGGE band sequences obtained in the present study (in bold). Numbers near nodes indicate bootstrap values greater than or equal to 60% in neighbor-joining (NJ), maximum-likelihood (ML), and maximum-parsimony (MP) analyses. TGGE bands were named according to the climatological seasons (A, autumn; W, winter; Sp, spring; S, summer), sampling sites (1 [white circles], 2 [gray circles], and 3 [black circles]), and replicates (stones a, b, and c).

Fig 6

Diagram of the TGGE banding profiles after computer-assisted characterization and phylogenetic analyses of excised bands.

Fig 7

Canonical correspondence analysis biplot based on the TGGE bands, with respect to environmental variables. DIN, dissolved inorganic nitrogen; SRP, soluble reactive phosphorus; DO, dissolved oxygen. Climatological seasons: A, autumn; W, winter; Sp, spring; S, summer. Sampling sites: 1, white circles; 2, gray circles; 3, black circles. Bands are indicated with black triangles.