Establishing conditions for miniprimer PCR. (A) Minimum primer length. Beginning with a 20-nt primer, a series of shortened primers having a common 5′ end was designed. These primers were used in combination with a 19-nt reverse primer (B) and a 10-nt primer (C) to demonstrate that 10-nt primers can be used for PCR. (B) Amplification using the forward primer series. When electrophoresed on an agarose gel, PCR with S-Tbr produces detectable amplification using primers as short as 10 nt (F10/R19). (C) Miniprimer PCR. Agarose gel electrophoresis of PCR products with S-Tbr and a range of annealing temperatures demonstrates robust amplification for a pair of 10-nt primers (F10/R10). In panels B and C, agarose gels are stained with ethidium bromide and arrows mark the expected size of PCR products.

Miniprimers designed and tested. (A) Sequences of miniprimers tested. (B) Performance of miniprimer pairs. Forward and reverse miniprimers were used to PCR amplify 16S rRNA genes from bacterial (Escherichia coli [Ec]) or archaeal (Halobacterium salinarum[Hs]) genomic DNA. PCR was determined successful if a band of the correct size was detected with minimum background. Success or failure with each template is indicated with + or −, respectively, in the appropriate column; a mark of +/− indicates that the product was in low abundance or contained a moderate amount of background but may be clonable. In several cases, amplicons were sequenced to confirm them to be the expected 16S rRNA sequences; confirmation is indicated by a + in the column labeled “16S.” nd, not determined.

Comparison of putative miniprimer and long-primer amplicons in the GenBank environmental sequence database. Pairs of long primers and miniprimers were used to identify putative 16S rRNA gene amplicons in the GenBank environmental sequence database. Horizontal lines separate aligned, matched pairs of long primers (gray shading) and miniprimers (no shading) that target the same 16S rRNA gene region. Sequences delimited by forward and reverse primer sequences separated by an appropriate distance were identified. The numbers of total sequences found (Sequences), unique sequences present within the total (Unique), and 16S rRNA gene sequences present within the unique sequences (16S) are indicated in the appropriate columns. 16S rRNA gene sequences were identified by a BLAST score ratio of >0.80 (see Materials and Methods). Degenerate positions are indicated using R for A or G, M for A or C, and W for A or T; na, not applicable.

Distribution of microbial mat library sequences in major bacterial divisions identified. The 10 taxonomic divisions with the highest representation in the libraries are shown as their relative proportions within the miniprimer (M) and two types of long-primer (P, 27F-P/1492R-10; H, 27F-HT/1492R-HT) libraries. The aggregate distribution of all libraries is also shown (all). Phylogenetic assignments were performed in Arb; the “other divisions” category comprises 31 divisions whose representation is less than ∼3 to 5% in each library. Proper Proteobacteria class names with standing have been shortened using Greek letters.

Distribution of microbial mat best database matches. The fractions of sequences whose best database matches fall into the designated distance intervals are shown for the long-primer (P, 27F-P/1492R-10; H, 27F-HT/1492R-HT) and miniprimer (M) libraries. Distances of 0.05, 0.10, and 0.20 approximately define the taxonomic levels of genus, order, and division, respectively (32). Plotted values and errors are the averages and standard deviations of six maximum likelihood calculations using six different taxonomies (15).

Microbial mat sequences with low-scoring database matches. Miniprimer (red), long-primer (green and blue), and reference (black) sequences were assembled into a phylogenetic tree by use of maximum likelihood (35). The best-scoring tree of 20 independent calculations is shown; branches supported by bootstrap values of greater than 50 (100 iterations) are labeled. Some bootstrap values near closely spaced terminal clusters have been removed for clarity; none of the removed values were less than 50 and most were greater than 80. Uncultured isolates are identified by their accession numbers and by the environment from which the sequences were cloned; cultured isolates are in italics. Taxonomic groups are indicated for clusters of three or more sequences that could be classified. Seven archaeal 16S rRNA gene sequences were used to root the tree.

Mismatches of 27F-P and 27F-HT primers to microbial mat miniprimer library sequences. (A) Portions of long primers 27F-P and 27F-HT that bind past the 3′ end of miniprimer 27F-10. (B) Bases represented by degenerate codes in 27F-HT. (C and D) The relative (C) and absolute (D) base frequencies observed within the 598 sequences of the miniprimer sequence libraries at each position within the 3′ halves of the 27F-P and 27F-HT binding sites. Note that at some positions the quantities of bases shown in panel D are not large enough to be resolved by the representation shown in panel C.