What is the main focus of the research project you are describing?

We developed a double fluorescence assay on paraffin samples combining LNA-based miRNA in situ hybridization (ISH) and immunohistochemistry (IHC) and demonstrated its application using miR-21 as an example (5).

miR-21 is highly expressed in many tumors, and we now know that miR-21 primarily originates from tumor associated fibroblasts, e.g., in colon and breast cancers (4, 8). One of the most widely accepted targets of miR-21 is PDCD4, and we therefore combined miR-21 ISH with PDCD4 IHC in a breast cancer cohort (5).

How can combined ISH/IHC contribute to our understanding of miRNA targeting?

The differential expression of PDCD4 (in cancer cells) and miR-21 (in stromal cells) cannot alone form the basis for arguing that miR-21 is dynamically targeting PDCD4. Interestingly, we identified miR-21 in a subset of normal mammary glands and in some high-grade breast tumors – including pre-invasive DCIS lesions (ductal carcinoma in situ, considered the earliest form of breast cancer). We observed that miR-21 can indeed be expressed in normal epithelial and in cancer cells (6).

From our studies on co-localization of miRNA and target protein, we find that it is essential to show that differential expression is occurring in the same cell population, and therefore using a cytokeratin antibody, we show the epithelial origin together with the presence or absence of PDCD4 and/or miR-21.

What was your previous experience with miRNA ISH?

We have worked with LNA probes for miRNA ISH over the last 5 years at Bioneer and have been involved in many exciting research projects with customers worldwide.

We perform LNA ISH on an automated platform that allows identical processing of hundreds of samples, both frozen and paraffin sections, using a one-day protocol. Automation helps to increase reproducibility and enables us to obtain relative expression estimates using image analysis (1, 2, 4).

How challenging was it to combine miRNA ISH with IHC?

A limiting parameter in combined ISH/IHC assays is that LNA ISH requires proteolytic pretreatment, and many commercial primary antibodies have not been evaluated using such pretreatment. Therefore, identification of appropriate antibodies requires some literature searches and scrutinizing the antibody vendor descriptions. We identified a PDCD4 antibody suitable for IHC that was compatible with proteolytic pretreatment and we saw staining typically of epithelial cells, including epithelial cancer cells.

Examination of double or triple fluorescence stained slides can be quite time-consuming, but I recently worked with scanned digital whole slides of fluorescence-stained sections, which strongly facilitates this process.

Can you give any details about your experimental set-up?

The automated ISH protocols that we have established are quite robust, but we have not found them advantageous for fluorescence detection. Therefore all our LNA-based miRNA ISH using fluorescence is performed manually (5).

The fluorescence miRNA ISH assay is just as sensitive as the chromogenic assay, and allows combination with IHC, which is a major advantage. We have combined LNA ISH with one or two antibodies and DAPI counterstain in various combinations.

How do you feel about your results?

Examining the miR-21 and PDCD4 stained slides is a very exciting process. The more I have looked into the (co-)localization patterns, the more exciting it appears, but also the more complex, with different patterns identified in different tissue structures, e.g., in DCIS lesions.

What do you find to be the main benefits of the miRCURY LNA miRNA Detection Probes and One Day ISH protocol for your miRNA ISH Services at Bioneer?

In all our internal and customer-related studies, we use the miRCURY LNA miRNA Detection Probes for miRNA ISH analyses. In my accumulated experience with many different probes and working with tissue from different organs and embedding matrices, I find that the LNA probes show consistent performance even between different probe batches. The high signal-to-noise that we see in many samples is essential for quantification of the ISH staining by image analysis.

What would be your advice to colleagues about getting started with miRNA ISH analysis?

I always recommend starting out with the miRCURY LNA miRNA ISH Optimization Kit, because it provides the essential probes and reagents, together with a detailed protocol, to implement the technology successfully.

In particular, PhD students may benefit greatly from learning this manual technology. I also refer to our publications in Methods in Molecular Biology (5, 6, MiMB) for tutorials, further applications and troubleshooting.

Of course, studies of larger sets of probes and samples would be advantageous to run on an automated platform as we do in Bioneer, where the experimental conditions can be systematically optimized in parallel with the positive and negative control probes.

In your opinion what is the most important factor for a successful miRNA ISH experiment?

I consider three factors that affect the performance of an LNA ISH experiment:

• sample quality
• accessibility of the miRNA
• miRNA expression level

Sample quality – both of paraffin and frozen tissues, as well as human and mouse tissues – is assessed using two key probes: the U6 snRNA probe together with a positive control probe, like for example miR-126 (3).

A negative control probe is important to check signal specificity. In some samples, certain tissue structures and cell compartments may show positive staining with the negative control scramble-miR probe. This may be caused by the detecting reagents and should be tested with a no-probe control.

Fortunately, most of the samples we receive at Bioneer provide optimal staining using these control probes, and I consider these initial observations as key, before applying other probes to the samples.

However, good staining with the positive control probe does not necessarily mean that it will be possible to detect any other miRNA, since this is limited by the other two factors: accessibility to the miRNA, and the miRNA expression level.

What are the future perspectives for this research?

At Bioneer, we perform LNA ISH on an automated platform. The automated basis for tissue staining allows subsequent image analysis for quantitation of miRNA expression after obtaining digital whole slides. I expect the image analysis platform will be used increasingly in the future, to provide quantitative expression estimates, as a tool to evaluate miRNA biomarkers, in the context of cellular origin and tissue architecture.

Automation of the combined fluorescence assays is under development in our lab, but often the main work with double fluorescence stained slides is in image acquisition and microscopy because the questions addressed in such samples are usually more qualitative than quantitative. Acquisition of digital whole slides from double or multiplex fluorescence assays is a great move forward to simplify fluorescence microscopy.

With respect to miR-21 and its targets, like PDCD4, much work needs to be done in various cancer types to better understand how miR-21 works in different cell populations, and thereby hopefully understand some of its cancer related functions as an onco-miR, as well as its utility as a prognostic, diagnostic and predictive biomarker.

References

1. Eriksen et al. Intratumoral Heterogeneity of MiRNA Expression in Rectal Cancer. PloS One 2016, 11(6):e0156919. PMID: 27258547.
2. Knudsen et al. MiRNA-17 is the most up-regulated member of the miR-17-92 cluster during early colon cancer evolution. Plos One 2015, 10(10):e0140503. PMID: 26465597.
3. Jørgensen et al. Robust one-day in situ hybridization protocol for detection of miRNAs in paraffin samples using LNA probes. Methods 2010, 52:375-381. PMID: 20621190.
4. Nielsen et al. High levels of miRNA-21 in the stroma of colorectal cancers predict short disease-free survival in stage II colon cancer patients. Clin Exp Mets 2011, 28:27-38. PMID: 21069438.
5. Nielsen and Holmstrøm. Combined MiRNA In Situ Hybridization and Immunohistochemical Detection of Protein Markers. Book chapter in "Target Identification and Validation in Drug Discovery" in the Methods in Molecular Biology series 2013, 986:353-65. PMID: 23436423.
6. Nielsen et al. miR-21 expression in cancer cells may not predict resistance to adjuvant trastuzumab in primary breast cancer. Frontiers in Oncology 2014, 4:207. PMID: 25177545.
7. Nielsen et al. Chromogen Detection of miRNA in Frozen Clinical Tissue Samples Using LNA Probe Technology. Book chapter in "In Situ Hybridization Protocols" in the Methods in Molecular Biology series 2014, 1211:77-84. PMID: 25218378.
8. Rask et al. High expression of miR-21 in tumor stroma correlates with increased cancer cell proliferation in human breast cancer. APMIS 2011, 119:663-673. PMID: 21917003.