Cap analysis of gene expression (CAGE)

In brief

CAGE (cap analysis gene expression) is highly sensitive and accurate quantitative transcriptome analysis which utilizes a cap structure on 5’ end of RNA. This feature enables high-throughput identification of transcription start sites and ability to infer promoters and transcription factor binding motifs. This project established the availability of CAGE through development of an optimized, cost -efficient in-house protocol, benchmarked against a commercially available ‘gold standard’ kit. This protocol produces comparable results to the commercial kit while introducing savings in reagent costs and time.

Highlights

CAGE can be used to simultaneously locate transcription start sites and quantify the abundance of RNA transcripts.
This developed "In-house" CAGE protocol delivers high quality results at a ~1/3 of the reagent cost and 25% less hands-on time when compared to the commercial kit.
To our knowledge this is the first successful implementation of CAGE in Australia.

Introduction

CAGE (cap analysis gene expression) is highly sensitive and accurate quantitative transcriptome analysis, which utilizes a cap structure on 5’ end of RNA. This feature enables high-throughput identification of transcription start sites and the ability to infer promoters and transcription factor binding motifs, in addition to standard RNA transcript quantification obtained through conventional RNA-seq.

This development describes the generation of CAGE libraries via the cap-trapping method using an optimized protocol designed and validated in-house. This in-house protocol is based on the commercially-available kit from DNAFORM, but is shorter and includes a final PCR step to generate higher concentration libraries. Resulting libraries are full-length (un-fragmented), Illumina-compatible libraries containing both polyA and non-polyA transcripts, that have been minimally amplified by PCR. The entire process requires 5ug RNA/sample with RIN >7.

Brief summary of method:

RNA is reverse-transcribed to generate an RNA/DNA hybrid. The methylguanylate cap on the 5’ end of capped transcripts is oxidised and biotinylated, and the duplex digested with RNAse I to release and remove the biotinylated 3’ ends of RNA molecules. The biotinylated capped ends are captured with streptavidin beads, and 5’ end cDNAs are released from the beads with NaOH. 5’ and 3’ linkers are added through two rounds of linker ligation, followed by USER/SAP treatment to manage unreacted adaptors. Samples are pooled before low-cycle PCR to generate completed library ready for sequencing.

Quality control checks are performed after cap-trapping and RNA digestion, after linker ligation prior to pooling, and after PCR prior to sequencing.

Table 1. Comparison of costs for CAGE protocols
	Commercial Kit	"In-house" protocol
Cost/sample (reagents only)	$458.83	$153.58
Cost/sample (reagents, comsumables plus labour when 8 libraries are prepared at once)	$867.67	$464.92
Hands-on wet-lab time (when 8 libraries are prepared at once)	44.5 hr over 8 days	32.5 hr over 5 days
Setup cost	$26,239.03	$18,789.47

Procedure

Note : Please refer 'UQ_GIH_inhouseCAGE.pdf' in the side for full details of the procedure, including 'critical points' and ' Occupational Health and Safety'.

In-house CAGE workflow

Figure 1. Workflow of in- house CAGE protocol.

Day1

Step 1 - 1st strand cDNA synthesis and clean up (3 hours)

Step 2 – Oxidation and clean up (2 hours and 30 min)

Step 3 – Biotinylation (16 hours, overnight)

Day2

Step 4 - RNase I digestion and clean up (2 hours and 30 min)

Step 5 - Preparation of Streptavidin beads (1 hour and 30 min)

Step 6 - Cap-trapping, Releasing cDNA and clean up (3 hours)

Step 7 - cDNA Q.C (1 hour 30 min)

Step 8 - Single Strand Linker Ligation (16 hours, overnight)

Day3

Step 9 – 2x AMPure purification (2 hours and 30 min)

Step 10 - 3’ Linker Ligation (16 hours, overnight)

Day4

Step 11 – AMPure purification (1 hour 30 min)

Step 12 – SAP/USER treatment and clean up (3 hours and 30 min)

Step 13 – qPCR analysis for pooling (2hours and 30 min)

Day5

Step 14 – PCR (3 hours)

Step 15 – Exonuclease I treatment and clean up (2 hours and 30 min)

Step 16 - final Q.C (3 hours)

Anticipated results

The following checks can be used to assess the quality of the final library:

rRNA – ACTB ΔCT at ssDNA QC check 1 – expected to be <4
total quantity of ssDNA at ssDNA QC check 2 – expected to be 15-30ng
fragment length of final library at Library QC check 1 – expected to be >800bp
final library concentration at Library QC check 2 – expected to be 10-20 nM for 4-plex library

The following checks can be used to assess the quality of the final data:

Mapping rate – proportion of sequencing reads mapping to the genome. >95% is good, >90% is acceptable.
Read distribution – proportion of sequencing reads mapping to different regions of the genome. >80% mapping to the promoter region is good, and indicates good enrichment for 5’ end of RNA molecules.
Number of identified transcription start sites/sample – This metric is particularly sample-dependent, so should be compared with previous data obtained with the same or similar cell type.

Worked example

Basic QC data from the pilot project are as follows:

<ssDNAQC1> Checking concentration by Oligreen assay:

	ng/ml	X100 for dilution factor	pg/ul	Total ssDNA (pg) in 50 ul	In ng
ESC1041	4.575	457.5	457.5	22875	22.88
ESC1067	4.233	423.3	423.3	21165	21.17
ESC1271	4.238	423.8	423.8	21190	21.19
ESC1369	3.459	345.9	345.9	17295	17.23

<ssDNAQC2> Check rRNA contamination by qPCR:

	Mean Ct of rRNA	Mean Ct of ACTB	Delta Ct (ACTB-rRNA)
ESC1041	12.885	15.467	2.582
ESC1067	12.432	14.925	2.493
ESC1271	12.367	14.933	2.566
ESC1369	12.565	15.15	2.585

<LibraryQC1> Check the size distribution of the final product

1) QC after PCR amplification (5 cycles) followed by Exonuclease I treatment and MinElute clean up

2) QC after extra AMpure clean up to remove adaptor dimer

Pooled average fragment length = 895bp

<LibraryQC2> Check the DNA concentration using KAPA library quantification kit

The concentration of final library is 14.34 nM in 20 ul (4-plex).

Comparison of inhouse vs commercial CAGE

Figure 2. Genomic distribution comparison of sequencing reads. Graph shows the proportion of reads mapping to different genomic locations for the same ESC case and control samples used as input for both workflows. Data produced using commercial protocol (a); data produced using in-house protocol (b). A high proportion of reads mapping to promoters is indicative of a successful experiment.

Comparison of detected CTSS

Figure 3. Comparison of detected CTSS (CAGE transcription start sites). The graph shows the number of TSS detected for the same ESC case and control samples used as input for both workflows. Data produced using commercial protocol are shown in red; data produced using in-house protocol are shown in blue.

Referenced procedures

References

Hazuki Takahashi, Timo Lassmann, Mitsuyoshi Murata, Piero Carninci. 2012. “5′ end–centered expression profiling using cap-analysis gene expression and next-generation sequencing”. Nature Protocols, March. https://experiments.springernature.com/articles/10.1038/nprot.2012.005

Mitsuyoshi Murata, Hiromi Nishiyori-Sueki, Miki Kojima-Ishiyama, Piero Carninci, Yoshihide Hayashizaki, Masayoshi Itoh. 2014. “Detecting expressed genes using CAGE”. Springer protocols. https://experiments.springernature.com/articles/10.1007/978-1-4939-0805-9_7