Pathologists are morphology pros. For them, the delicate swoops and swirls and dabs of color in stained tissue slices are open books, keys to differentiating diseased tissue from benign and revealing clues to guide treatment decisions. The key word, of course, is “stained.”
Tissue slices are largely featureless under a light microscope, so pathologists (and researchers) must stain the cells to see them.
In many cases, hematoxylin and eosin (H&E) staining, which reveals basic tissue architecture, suffices. But sometimes more molecular information is required. In those cases, pathologists have two basic options.
In immunohistochemistry (IHC), a sample is stained with an antibody to a specific diagnostic protein, such as cytokeratin-18, revealing that molecule’s distribution across the tissue. Fluorescence in situ hybridization (FISH) uses nucleic acid hybridization to probe for specific sequences, for instance to reveal chromosomal abnormalities or pathogens.
Neither technique is new, of course. But that doesn’t mean researchers and tool developers are content. Here we take a look at what’s new in the world of in situ hybridization.
DNA FISH
FISH relies on the hybridization of a labeled probe designed to light up specific sequences, such as telomeres or specific genes. For instance, researchers can use probes that recognize the (normally separate) genes BCR and ABL to look for cells in which those two genes have been fused, as occurs in some cases of chronic myelogenous leukemia [1]. Patients bearing this particular mutation are responsive to the drug imatinib (Gleevec), so testing is crucial in deciding on a course of treatment.
The problem with in situ hybridization in general is one of signal strength. With relatively few hybridization events per cell, signal amplification typically is required. One approach is simply to heavily label the probe sequence with multiple fluorophores. Alternatively, researchers can use a detection modality that boosts the signal of each bound probe molecule. Thermo Fisher Scientific’s Tyramide Signal Amplification (TSA™) process, for instance, relies on a horseradish peroxidase (HRP) substrate called tyramide, which is coupled to a fluorophore (e.g., Alexa Fluor® 488). Incubation of tyramide with HRP-conjugated secondary antibodies generates hundreds of fluorescent molecules per antibody site, producing a sharply amplified signal, says Kamram Jamil, product manager for cellular analysis products at Thermo Fisher Scientific.
A soon to be released, updated version of the reagent, SuperBoost, will include three improvements, Jamil says: a simplified workflow; stronger signal amplification with the inclusion of up to three HRP molecules per antibody; and an HRP “stop” solution, which will enable researchers to multiplex reactions and create sharper images. “We have cellular data that suggest that this technology is sensitive enough that even a few tags on a probe can be amplified,” he says.
A CRISPR FISH
Though simple in theory, the FISH protocol is both laborious and punishing to a sample, saysTimothée Lionnet, a project scientist at the Howard Hughes Medical Institute Janelia Research Campus in Ashburn, Va. To make DNA available for hybridization, the sample must be denatured at high temperature. The hybridization step itself typically runs for hours or even overnight.
In a recent issue of PNAS, Lionnet described an alternative strategy that overcomes these difficulties [2]. A clever variant on Cas9-based genome editing, CASFISH relies on an enzymatically inactive form of the Cas9 nuclease that is conjugated to the Promega HaloTag® protein. HaloTag is an enzyme that covalently couples chloroalkane substrates to the protein of interest, thereby providing a mechanism for labeling the protein with useful molecular handles, including fluorescent dyes.
CASFISH replaces the traditional hybridization step with guide-RNA-directed binding of the Cas9 protein to such genomic locations as telomeres, centromeres and specific gene loci. The process takes just 15 minutes, Lionnet says, and it is theoretically compatible with live cells. “So it’s a much more gentle protocol and also much faster,” he says. In addition, thanks to a broad palette of HaloTag-compatible fluorophores, the protocol is easy to multiplex, Lionnet notes [3]. The CASFISH vector is available through Addgene.
RNA ISH
Of course, in situ hybridization doesn’t need to be confined to DNA analyses. Though not yet widely used in the clinic, RNA-based in situ hybridization (RNA ISH) offers several advantages over IHC and DNA FISH for certain clinical problems, says Guy Afseth, director of oncology product marketing at Affymetrix. Advantages include the ability to detect noncoding transcripts, specific splice variants, mRNA expression of secreted proteins and biomarkers for which commercial antibodies are either unavailable or suboptimal.
Albumin, for instance, is a useful biomarker for distinguishing primary liver tumors from metastatic growths in the liver, says Afseth: “If your tumor makes albumin, it is liver-derived; if not, it isn’t.” But albumin is difficult to evaluate by IHC, he notes, as it is a secreted protein. Thus, his company has developed assay-specific reagents to quantify albumin expression at the RNA level.
Another application is detection of human papillomavirus (HPV) in certain head and neck cancers. High-risk HPV-positive tumors have a better prognosis than those without the virus, says Rob Monroe, chief medical officer at Advanced Cell Diagnostics (ACD), which commercializes the so-called RNAscope RNA ISH assay format, “so identifying those oropharyngeal carcinomas that are HPV-related helps … in informing patients about prognosis and potential treatment options.”
Current HPV testing in head and neck cancer relies on IHC detection of a surrogate biomarker that is expressed as a result of HPV infection, p16, or DNA FISH analysis to assess chromosomal integration and subtyping of the HPV genome, Monroe says. But the former approach is nonspecific, while the latter is relatively insensitive. “The advantage of the [ACD] RNAscope assay is it is highly sensitive relative to DNA ISH, and it’s highly specific relative to p16. So it gives you the best of both.”
ACD’s RNAscope and Affymetrix ViewRNA® assays rely on so-called branched DNA. Binding of two or more adjacent probes creates a scaffold upon which a signal-amplification process can occur—up to 96,000-fold in the case of ViewRNA, Afseth says. This strategy works well even on fragmented RNA from formalin-fixed paraffin-embedded (FFPE) samples, Afseth says, as only a few binding events are required for visualization.
LGC Biosearch Technologies’ Stellaris probes, on the other hand, require no amplification step. Instead, a minimum of 25 unique tiled oligonucleotide probes (out of a maximum 48 probes per gene) must bind to a transcript to produce a measurable signal, says product manager Jessica Kaplunov. As a result, the signal is highly specific and quantitative. “If you see a bright spot, you can infer there is more than one copy compared to other dimmer spots. With amplification, you lose that, because amplification may be uneven,” she says.
Both Affymetrix and ACD say they are working with Leica Biosystems to automate their RNA ISH protocols on the company’s BOND-III clinical laboratory staining system. “This is a big step forward for clinical diagnostics labs,” Monroe says.
References
[1] Bishop, R, “Applications of fluorescence in situ hybridization (FISH) in detecting genetic aberrations of medical significance,” Bioscience Horizons, 3:85-95, 2010.
[2] Deng, W, et al., “CASFISH: CRISPR/Cas9-mediated in situ labeling of genomic loci in fixed cells,” PNAS, 112:11870-5, 2015. [PMID: 26324940]
[3] Grimm, JB, et al., “A general method to improve fluorophores for live-cell and single-molecule microscopy,” Nature Methods, 12:244-50, 2015. [PMID: 25599551]