Transcriptome annotation using tandem SAGE tags

Abstract

Analysis of several million expressed gene signatures (tags) revealed an increasing number of different sequences, largely exceeding that of annotated genes in mammalian genomes. Serial analysis of gene expression (SAGE) can reveal new Poly(A) RNAs transcribed from previously unrecognized chromosomal regions. However, conventional SAGE tags are too short to identify unambiguously unique sites in large genomes. Here, we design a novel strategy with tags anchored on two different restrictions sites of cDNAs. New transcripts are then tentatively defined by the two SAGE tags in tandem and by the spanning sequence read on the genome between these tagged sites. Having developed a new algorithm to locate these tag-delimited genomic sequences (TDGS), we first validated its capacity to recognize known genes and its ability to reveal new transcripts with two SAGE libraries built in parallel from a single RNA sample. Our algorithm proves fast enough to experiment this strategy at a large scale. We then collected and processed the complete sets of human SAGE tags to predict yet unknown transcripts. A cross-validation with tiling arrays data shows that 47% of these TDGS overlap transcriptional active regions. Our method provides a new and complementary approach for complex transcriptome annotation.

Figure 1.

(A) Schematics of search for tag-delimited genomic sequences (TDGS, double arrows). Upper part: procedure for assembling 5′G- 3′C pairs. Starting from the previously identified C-tag (n − 1), the program searches the next site on which an experimental C-tag (n) can be positioned (star 1). The genome sequence is then scanned for G-tags (x − 1, x) and stops (star 2) when the shortest G–C pair is found. In search of the next pair (stars 3 and 4), the G-tag (z) potential tag sequence is skipped because it does not match any G-tag in the experimental dataset. Lower part (B,C and D): illustration of the various causes of success and failure in assembling TDGS. Numerical values in B and C are taken from the study of 489 well-annotated sequences identified in the macrophage SAGE libraries. Cases schematized in D are detailed (d) in Figures 3 and 4.

Figure 2.

Number of distinct C-tags (left y-axis, black square) in five consecutive classes of occurrence, i.e. abundance, (x-axis) from 1 to >2048 counts summed up over UniSAGE, and percentage of tags matching RefSeq annotated sequences (right y-axis, gray diamonds).

Figure 3.

Characterization and annotation of validated TDGS. Alignments of the TDGS# 20 and # 212 to the UCSC human genome browser. For RT–PCR validation, Macrophage poly(A)+ RNA were extracted from MDM and the cDNA were synthesized using mRNA and oligo-dT primer. (A) TDGS# 20 corresponds to an example of Class 2 transcript localized near the coding region of CDH23. For PCR, a primer pair was respectively designed in the 3′-end of CDH23 and in the TDGS # 20. The existence of this new variant transcript was confirmed in macrophage by sequencing. (B) TDGS # 212 is an example of class 3 transcript. Experiments without reverse transcription (A, C) and with DNAse treatment (C, D) were performed to detect DNA contamination. For transcript validation, a first PCR was realized with primers pairs designed on TDGS # 212 and a second one with primers respectively in the 3′-end of EST EB10260 and TDGS #212. The 3′-end of the transcript was validated by 3′RACE (3′RACER kit, Invitrogen, France). The sequenced PCR products validated the existence of a transcript in inverted orientation relatively to the Cathepsin B gene (CTSB, NM_0019082).

Figure 4.

Characterization and annotation of validated TDGS. Alignments of the TDGS# 5, 6, 54 to the UCSC human genome browser. For RT–PCR validation, Macrophage poly(A)⁺ RNA were extracted from MDM and the cDNA were synthesized using mRNA and oligo-dT primer. (A and B) TDGS # 5 (376 bp) maps on chomosome 1, corresponding to the sequence of new full-length cDNA registered in GenBank as CR601947. TDGS # 6, shares with TDGS # 5 the same C-tag (dotted boxes) but its sequence is longer (821 bp) because its G-tag is located in 3′ of the TDGS # 5 one. (A and C) In addition, TDGS # 54, reveals another site, in a region of chromosome 14. The same conditions as described in Figure 4 were used to PCR validation. RT–PCR analysis followed by sequence checking confirm the existence of this new transcript. The sequence of chromosome 14 matched a sequence registered in the Affymetrix dataset harbored at UCSC (Affy Txn Phase2)

Figure 5.

Length of the TDGS assembled from the whole UniSAGE data. Frequency of TDGS length (in base pairs) on well-known annotated TDGS (R1 TDGS) (A) and on unkown TDGS (unmatched TDGS) (B). Each point represents a TDGS (in gray)(∼99 and 92.5% TDGS are shown, respectively), and the curve (in black) is a power regression curve.