Fluorescent in situ hybridization (FISH) is a classic method of research that is used to locate the positions of specific DNA sequences on chromosomes. In FISH, in situ hybridization is coupled with fluorescent labeling where probes are labeled with an F-labeled nucleotide and then targeted to the DNA or gene or interest. In this coupled technology, dsDNA can be particularly identified and targeted and then visualized through a fluorescence microscope or other applicable imaging system.

FISH technology has been used extensively as a macromolecule recognition tool to effectively identify and map chromosomal, and sub-chromosomal, abnormalities during the annotation phase of the Human Genome Project. Now, FISH is used in research more so for its highly sensitive and specific capabilities, which may be used as a diagnostic tool for the identification of genetic and chromosomal disorders, principally directed for clinical diagnosis.

The FISH procedure uses fluorescent labels that are directly complementary to the nucleotide sequence of interest. FISH technology may therefore be used not only to target specific DNA sequences, but mitochondrial RNA (mtRNA), long noncoding RNA (lncRNA), and microRNA (miRNA) in various tissue and cell types, as well.

There are many slight variations in how FISH may be performed, however each procedure follows the same general principles. Firstly, a probe is constructed. The probe is then fluorescently labeled directly by a process called Nick translation. Next, the labeled probe and target DNA/RNA are denatured separately and then sequentially annealed to one another. Locations of the fluorescent probe within the target DNA/RNA can then be identified through various means of analysis. Below is an example FISH protocol using DNA from HeLa cells, using probes constructed through PCR and direct fluorescent labeling methods, along with hybridization and secondary washing steps.

The power of FISH can be amplified by the use of multiple fluorescent colors in tandem. Multicolor FISH can be used to identify as many labeled features as there are differing fluorophores used in the hybridization process. Single colors, as well as various combinations of colors, can be simultaneously detected.

Epifluorescent microscopy is commonly used for this multicolor analysis, and similar to other fluorescence microscopes, the excitation light from the light source is passed through a filter before reaching the objective lens and dichroic mirror. This excitation filter allows a single band of a particular wavelength through, dependent on the peak fluorescence excitation wavelength of the fluorophore. The excitation filter can be in the form of filter cubes or a filter wheel with multipass dichroic and barrier filters. The produced signals must be isolated by an emission filter that ensures that only the wavelength emitted from the targeted fluorophore in the sample passes through.

Traditionally, data was collected through digital charged coupled devices (CCDs), high-speed color film, or intensified video cameras, though these methods had issues with color fidelity and there existed problems when processing superimposed images, with different colors, within a single specimen.

Recently, much development has been made to overcome these limitations with the creation of quantitative imaging analysis software suites. These analysis programs are capable of capturing images across multiple wavelengths and/or focal planes, with the ability to retrieve and store large volumes of multicolor image data sets. The software offer the ability to focus at multiple wavelengths, record the slide locations of cells, quantitatively measure probe counts and/or intensity, and determine cell morphology. It may also be able to analyze multicolor metaphase chromosomes, map genes, offer comparative genomic hybridization ratios, as well as determine karyotype generation.

Confocal laser scanning microscopy may also be used for visualization. This technique uses light emitted by a laser, at a specific wavelength to maximally excite the fluorophore, which it will then emit at a somewhat longer wavelength. The difference between these two wavelengths is known as the Stokes shift.

The light runs through a pinhole aperture before hitting the dichromatic mirror directed towards the specimen, where secondary fluorescence is then emitted. Instead of illuminating the entire sample at once, the laser light only illuminates a given spot at a specified depth. Because of this, out-of-focus signals from features located above and/or below the focal plane contribute to the high signal-to-noise ratio of images produced. As the detector pinhole required for focusing images also may reduce signal strength, confocal laser scanning microscopy typically requires higher fluorescence signal yield to be effective.

Dozens of variations of FISH have been created to improve the sensitivity, specificity, and image resolution for the associated target samples. Depending on the technique, chemical and physical information can be generated for a number of biological targets. As the target DNA remains intact, (different from other forms of molecular analysis) genetic information about probe positions in relation to chromosome bands or other probes is directly, and efficiently, obtained. FISH has aided in rare disease diagnostics, evaluation of prognosis, and even the chance of remission for diseases like cancer.

In many cases, it may be necessary to multiplex spectral techniques to increase the number of detectable targets. Many experimental steps require empirical optimization specifically for the samples and probes used, which may prove time consuming and costly. Long hybridization times are essential to ensure the appropriate saturation of signals throughout the specimen, so FISH is also an inherently lengthy process and requires experienced personnel, though advances in assay automation may alleviate these limitations in the future.

Though many variations of FISH exist, a major downfall of this technique is that it can only detect known genetic aberrations. FISH can be used only to confirm previously characterized imbalances and abnormalities, and cannot serve as a screening test for novel chromosomal rearrangements.