Icecore Findings with Respect to Contaminant Database Building Likelihood Criteria

For ice samples with mineral inclusions, DNA extraction was performed directly from sediment material., It was performed from concentrated melt water for all others.

To amplify bacterial and archaeal 16S ribosomal RNA genes, a seminested PCR with broad-range primers was implemented. The resulting clones were only 405-430 bases in length because it was impossible to amplify full-sized 16S rRNA genes using different DNA polymerases and primers. This may be indicative of very small amounts of DNA or its ancient status (Cooper and Poinar 2000) due to possible damage, for instance, by oxygen excess content in the lake (Lipenkov and Istomin 2001). Note that the DNA was extracted from highly concentrated melt water and we were nevertheless forced to use a very sensitive seminested PCR scheme to be able to generate signals. One-round PCR did not work while PCR inhibitor tests performed with E. coli genomic DNA added to original ice DNA samples were always negative. Preliminary control experiments performed by E. coli cell counting followed by PCR extinction showed that the detection limit of our PCR schemes was equivalent to 5-8 cells per ml of melt water.

A total of 121 clones from both accretion (101 clones) and glacial (20 clones) ice samples were analysed. The ice clones showed high redundancy resulting in the recognition of twenty 16S rDNA phylotypes, of which 16 were seen in accretion ice and three also in glacier ice. We will discuss below the question of accretion-ice findings only.

Most phylotypes from accretion ice had close Genbank homologues (>98%), i.e. were identified (Stackebrandt and Goebel 1994) or assigned to known sequences, but three were distinctly different from their nearest neighbours (<96% similarity).

Bacterial taxa identified by 16S ribosomal DNA sequences were indexed on the basis of specially developed criteria. The contaminant database was very useful but not sufficient by itself (Table 7.3). Since the biological contents of the Lake Vostok are hard to predict, a degree of confidence was introduced in the table allowing us to discard obvious cases with the highest scores as contaminants, while those with the lowest scores were referred to as likely contaminants.

Table 7.3. Indexing contaminant criteria for bacteria recorded in Vostok ice core. The keys are also used in Table 7.4

Key score"



Taxa recorded in the contaminant database


Taxa proved to be contaminants (e.g., Tanner et al., 1998, Cisar et al.,

Was this article helpful?

0 0

Post a comment