The Application of MicroRNAs in Cancer Diagnostics - RNA Interference from Biology to Therapeutics - page 263

high affinity DNA analogs (e.g., LNA) into the miRNA-specific primers to obtain
sufficient specificity in subsequent real-time PCR [ 112 ]. In most cases, either the
DNA intercalating dye SYBR Green I or TaqMan probes are used for real-time
detection of amplification products. SYBR Green I is not sequence specific, but
postrun dissociation curve analysis can be performed to validate that only one prod-
uct is amplified. In principle, TaqMan probes can provide an extra level of sequence
specificity beyond that provided by the RT-PCR primers, but due to the short length
of mature miRNAs, TaqMan probes will often have to overlap RT-PCR primer
sequences and, therefore, in reality may not add specificity to the reaction when
analyzing miRNAs. For global profiling of miRNA expression, RT-qPCR assays
can be multiplexed and parallelized [ 109, 113 ]. Several RT-qPCR arrays are com-
mercially available for analysis of tens to hundreds of miRNAs in 96- or 384-well
format [ 111 ], including, e.g., TaqMan low-density microRNA cards, which are 384-
well micro- fl uidic cards [ 114 ] .
All hybridization- and PCR-based methods suffer from the limitation that only
known miRNAs can be detected due to reliance on predesigned primer/probe
sequences. In contrast, NGS ( aka massively parallel sequencing or deep sequencing)
enables digital miRNA expression profiling as well as discovery of new miRNAs
and miRNA variants [ 115, 116 ]. There are a number of different commercial plat-
forms available for NGS, which differ somewhat with respect to sequencing princi-
ples, chemistries, and methods of detection, but all share the ability to sequence
thousands of DNA molecules in parallel, thus producing massive amounts of data. In
a typical small RNA sequencing experiment, oligonucleotide linkers are added to
small RNAs extracted from a biological sample and the resulting library is amplified
by PCR and sequenced (Fig. 13.3 , right panel). Typically, over 10 million small
RNA sequence reads are produced for each sample [ 117 ] . When sequence reads are
mapped back to a reference genome, the number of reads for a particular miRNA
species can be taken as a measure for its relative expression level, whereas the exact
sequences of individual reads can be used to identify new miRNAs and polymor-
phisms [ 118 ]. Drawbacks include the risk of sequencing errors and introduction of
bias during library preparation and amplification as well as high analysis costs and
scarcity of standardized computational tools for data analysis [ 119, 120 ] . Nevertheless,
NGS is likely to become the technology of choice in future miRNA and transcrip-
tome profiling studies as sequencing prices are constantly reduced, experimental
protocols optimized, and new computational analysis tools developed.

13.3 MicroRNAs as Cancer Biomarkers

13.3.1 Early miRNA Pro fi ling Studies of Cancer

One of the first expression profiling studies to highlight the potential of miRNAs as
cancer biomarkers was published in 2005 by Lu et al. [ 23 ]. The study used a bead-
based hybridization method to measure the expression of 217 miRNAs in 334
samples representing multiple human malignancies. The authors found that miRNA