miRCURY LNA™ Arrays for microRNA expression profiling

At a glance

435 miRPlus™ microRNAs - access to unique human microRNAs
Sensitivity - microRNA expression profiling possible from 30 ng total RNA
Specificity - efficient discrimination between closely related microRNA family members
Reproducibility - high reproducibility with 99% correlation between arrays
Dynamic range - greater than 4 orders of magnitude
Diversity - most comprehensive probe set available
Open platform - protocols available for Tecan and MAUI hybridization stations, and for manual hybridization

Product coverage
We currently offer two miRCURY LNA™ microRNA Arrays:

V. 11.0 - hsa, mmu & rno array: An array specific for human, mouse and rat microRNAs that contains more than 1700 capture probes, covering all microRNAs annotated in miRBase 11.0, as well as all viral microRNAs, related to these species. The coverage to the latest miRBase v.12.0 is 98%, 97% and 100% for human, mouse and rat, respectively. (See Table 1)
In addition, this array contains capture probes for 435 new miRPlus™ human microRNAs. They are proprietary microRNAs not included in miRBase. The significance of these sequences is underscored by the fact that 30 % of the currently annotated human microRNAs, were available in our miRPlus™ range of products prior to publication in miRBase.
The 9.2 all species array contains more than 2000 capture probes, making it possible to profile miRBase 9.2 microRNAs from any organism – vertebrate, invertebrate, plant and virus – and to cross profile between species.

miRPlus™ - proprietary human microRNAs
In addition to the microRNAs annotated in miRBase, our miRCURY LNA™ microRNA Arrays contain miRPlus™ capture probes. These probes target proprietary microRNAs that have been identified by Exiqon using cloning and sequencing of human normal and diseased tissue.

Before being added to our miRCURY LNA™ microRNA Arrays, the sequences are subjected to strict quality control to ensure that they are truly different from annotated microRNAs and that they are found in several clones. Once they have been added to the arrays, the miRPlus™ sequences give scientists unique information about microRNAs not available elsewhere (Figure 1).

The first miRPlus™ capture probes appeared on the miRCURY LNA™ microRNA Array in the middle of 2006 and over the years, 274 miRPlus™ probes have been featured on various versions of the arrays. Out of these sequences, 236 are annotated in miRBase v. 12, which means that almost 30% of the currently annotated human microRNAs were available on our arrays at some point before they were included in miRBase.

This clearly demonstrates the high quality of our miRPlus™ microRNAs and represents yet another benefit for scientists using our miRCURY LNA™ microRNA Arrays.

435 new human miRPlus™ capture probes On version 11 of our human, rat and mouse array, 435 new miRPlus™ capture probes have been introduced. This means that the miRCURY LNA™ microRNA Array can now detect close to 1300 human microRNAs and an additional 80 human viral microRNAs (Table 1). This is by far the highest coverage of human microRNAs available on the market.

847 Mature human microRNAs
80 Mature human viral microRNAs
435 Mature miRPlus™ human microRNAs

The miRPlus™ sequences are normally submitted to miRBase within 6-12 months; however, researches can get access to the sequences by signing a non-disclosure agreement.

Tm-normalized LNA™ capture probes
Detection of short target sequences, such as microRNAs, is inherently difficult and is complicated by the large variation in base composition of the microRNAs, which range in GC content from 25 to 90% (for human microRNAs). In order to be comparable, highly discriminative, and sensitive, microarray capture probes should optimally have similar melting temperatures (i.e. be Tm-normalized).

However, using pure DNA probes, Tm -normalization compromises probe design because the maximum hybridization temperature will be dictated by the lowest full-length microRNA capture probe duplex Tm on the array, i.e. the Tm of the most AT-rich duplex. In order to adjust the Tm of capture probes targeting the more GC-rich microRNAs (high Tm), significant truncations generating capture probes as short as 8-9 nucleotides, are required.

Such short capture probes present a significant problem because of their poor specificity due to the high frequency of 8-9-meric sequences in the transcriptome. Furthermore, probe sensitivity is reduced due to the short probe length. This means that Tm –normalization of pure DNA capture probes is not possible without compromising specificity and sensitivity of some of the probes.

This paradox is solved by incorporation of LNA™ in the capture probes of the miRCURY LNA™ microRNA Arrays. By adjusting the LNA™ nucleoside content as well as the length of the probes, the capture probes have been Tm -normalized to ensure that all microRNA targets hybridize to the array with equal affinity under high-stringency hybridization conditions. LNA™ capture probes have been designed according to empirically derived algorithms to maximize their affinity and specificity for their microRNA target. Figure 2 illustrates how the Tm of LNA™ capture probes is increased significantly and the Tm range is narrowed significantly compared to DNA probes. This makes miRCURY LNA™ microRNA Arrays superior to DNA-based arrays, especially when it comes to the detection of AT-rich microRNAs (Figure 3).

Sensitivity
The sensitivity of miRCURY LNA™ microRNA Arrays have been assessed by empirically testing 705 human microRNA capture probes using synthetic microRNAs. More than 90% of the LNA™ capture probes on the array have a detection limit of ≤10 amol, enabling microRNA profiling even with very small amounts of total RNA (Figure 4).

Sample input

The high sensitivity of miRCURY LNA™ microRNA Arrays means that reliable results can be obtained from as little as 30 ng of total RNA (Figure 5). However, without prior knowledge of the microRNA content of the sample, we recommend using 250 to 1000 ng total RNA.

Specificity
miRCURY LNA™ microRNA Arrays are highly specific for their microRNA targets. The combination of Tm -normalized LNA™ capture probes and hybridization conditions optimized for high stringency binding allows for accurate detection of microRNA expression and increases the specificity of the capture probes. The optimized LNA™ capture probe design provides superior distinction between closely related microRNAs and will, in most cases, be able to specifically distinguish between microRNAs that differ by only one nucleotide (Figure 6).

Dynamic range
miRCURY LNA™ microRNA Arrays offer superior dynamic range over more than 4 orders of magnitude, ensuring that microRNAs with high and low expression levels will be detected well within the linear detection range (Figure 7).

Improve data quality with spike-ins
All miRCURY LNA™ microRNA Array products contain 10 synthetic spike-in microRNAs that can be detected on the arrays by specifically designed capture probes. When the spike-in microRNAs are added to the labeling reactions before a dual-color array hybridization, the signals from the spike-in capture probes can be used:

as a control for the labeling reaction and hybridization
to calibrate/adjust scanner settings between channels
as a control for the data normalization procedure
to estimate the variance of replicated measurements within arrays
to assess technical variability between different parts of the array

Figure 8 illustrates the position of the 10 spike-in microRNAs in 1 ug total RNA.

Reproducibility
The miRCURY LNA™ Array features very high reproducibility through an optimized manufacturing processes that ensures high quality uniform spots (Figure 9). This results in very low CV values of the four replicate spots as well as excellent inter-slide correlation (Figure 10).

High cross-platform correlation
Data obtained with the miRCURY LNA™ microRNA Array demonstrate excellent cross-platform correlation with data obtained using other profiling technologies. An example of such correlation is demonstrated in Figure 11. A comparison was made of miRCURY LNA™ microRNA Array data and published profiling data of placenta, ovary and liver tissue using small RNA cloning (Landegraf et al., 2007) or real-time PCR (Liang et al., 2007). The RNA used for these studies were from different sources and were, therefore, subjected to significant differences with respect to variations between donors, tissue fractions and sample processing procedures. Moreover, the cloning data from some of the less abundant microRNAs constituted only 15 counts in total from the three different tissues, resulting in some “stochastic noise”. Nevertheless, the hierarchical cluster analysis showed excellent correlation between the three different techniques, resulting in tight tissue clustering independent of platform.