 16S rRNA gene marker  intra-gene variability  primer selection  size & information content Primer selection, information content, alignment and length

16s rRNA gene marker Conserved 2º structure Natural gene amplificationGenealogy reconstruction Ludwig and Schleifer, 1994 FEMS Rev 15:

Intra-gene variability  secondary structure shows differences in the conservation of homologous sites  highly conserved zones give information on deep-genealogies (higher resolution for distantly related)  hypervariable zones give information on recent events (higher resolution for close relatives) Anderson et al., 2008 PLoS ONE, 3: e2836 Stahl and Amann, 1991 John Wiley and Sons

Primer selection  universality  Universal primers target highly conserved regions  Universality depends on the known dataset  Different phyla may have differences in the “universal” regions (e.g. EUB 338)  Primers used for rRNA cloning may give biased results  Metagenomics without amplification steps may reveal hidden diversity EUB338 I Most Bacteria GCTGCCTCCCGTAGGA GT EUB338 II Planctomycetales GCAGCCACCCGTAGGT GT EUB338 III Verrucomicrobiales GCTGCCACCCGTAGGT GT Daims et al System Appl Microbiol 22,

Primer selection  size of the amplicon GM Valt 8 GM5 GM5-clamp F R 518 GM R F 945 Bac1055F R 1529 S 1505  ideally the almost complete gene (~ 1520 nucleotides) should be sequenced  many amplifications skip sequencing the helix 50 (~ 1490 nucleotides)  many clone libraries are based on just partial amplicons (~ 900 nucleotides)  Pairs GM3 (8) – GM4 (1492) most widely used

16S rRNA sequencing has grown exponentially in parallel to the development of sequencing techniques Yarza et al., Nature Revs : Tamames & Rosselló-Móra 2012 TIM 20: rRNA cataloguing radioactive Sanger sequencing non- radioactive Sanger sequencing reverse transcription sequencing NSG The database is exponentially increasing 99% environmental sequences 1% cultured organisms 3.8 x10 6 sequences 700,000 / year (last three) Sources of sequences and quality  rRNA Cataloguing (up to late 80’s), bad quality  reverse transcription sequencing (up to late 90’s), bad quality  Sanger methods (radioactive, biotin-labelled, terminal-dye… still in use)  cloning DNA, good quality  direct amplification, good quality  DGGE/TGGE, short sequences, bad quality  NSG, short sequences  454 technology (now up to 800nuc, mean of 500nuc), moderate quality  illumina (now 2x 250nuc), too short

16S rRNA sequencing has grown exponentially in parallel to the development of sequencing techniques Quast et al., 2013, Nuc Acid Res. 41: D590-D596 SILVA release 119 (July 2014) rate of rejection of about 30% of the existing sequences short sequences are generally worse than long stretches

We divided the 16S rRNA gene into 6 regions of 250 nucleotides -Calculated taxa recovery in each stretch -Compare with that of the full sequence Regions V1 & V2 Regions V3 & V4 Regions V5 & V6 Categoryminimum Species98.7% Genus94.5% Family86.5% Order82.0% Class78.5% Phylum75.0% Yarza et al., Nature Revs :

-77% of the 16S rRNA gene sequences < 900pb -The 5‘region (V1-V2) overestimates species -The remaining regions tend to underestimate all taxa -Increases in length tend to mirror that of the full sequence Yarza et al., Nature Revs :

Size & information content  complete sequences give complete information  partial sequences lose phylogenetic signal  short sequences lose resolution 1500 nuc 900 nuc 300 nuc

Primer selection & size of amplicons  selection of primers is important for representative results  the length of the amplified/sequenced gene  adequate phylogenetic signal  short sequences may lose resolution