Fluorescence In Situ Hybridization (FISH) is a method which has evolved many exotic variants, often identified by prefixes, such as “mFISH,” “Q-FISH,” “MERFISH,” or “CO-FISH.” Standard FISH, however, simply involves fragments of fluorescently labeled DNA (FISH probes) annealing to sample DNA. But even standard FISH leaves plenty of room for variability. One such aspect in which there is notable divergence has to do with how FISH probes are manufactured, with important consequences for probe performance. We can focus on two of the available FISH probe production approaches: to either use a synthetic oligonucleotide as the starting molecule, or to start instead with fragments of a bacterial artificial chromosome (BAC). Both currently enjoy widespread use and we will discuss significant features of each.

Building BAC probes starts with building a BAC library. To do so, DNA comprising an organism’s whole genome is digested into fragments with a typical size of a few hundred kilobases. These fragments are ligated into plasmid cloning vectors and transformed into bacterial host cells, typically E. coli, where they will be copied again and again as the cells multiply. This array of cultures containing the fragmented genome is the BAC library. To then build a BAC FISH probe from this library, the specific BAC clone containing a fragment of interest must be cultured and its DNA must then be extracted. The extracted DNA is then further fragmented down to strands ranging in length from hundreds to a few thousand base pairs long. This stage represents one of the principal differences between BAC probes and synthetic oligonucleotide probes because once the DNA is fragmented, strands of a variety of sizes result, and performance is heavily influenced by the quality control steps taken to screen out undesirable fragments. Fragments that are very long or very short, as well as those representing sequences complementary to more than one genomic locus, will be more likely to bind loci other than the intended target of the probe set. In a FISH reaction, this can produce generalized background fluorescence and localized off-target signals, both of which can confound analysis. On the other hand, one benefit of creating DNA fragments with a wide span of lengths and sequence characteristics is that the set as a whole may tolerate a wider range of hybridization and post-hybridization wash conditions.

Oligonucleotide FISH probes might be said to be engineered, by contrast. The availability of the human genome sequence enabled researchers to start mining the now known sequences of genomic targets for unique segments which would meet specific parameters suitable for FISH. Using bioinformatics, oligonucleotide sequences can be digitally pre-selected based on sequence length, melting temperature, GC percentage and a variety of other preliminary attributes. Sequences that meet those initial criteria can then be further screened to disqualify those which form hairpins, bind to more than a pre-defined number of off-target sites, or have more than a specific percentage of sequence complementarity to any off-target site. Sequences that satisfy all the requirements are then chemically synthesized and fluorescently labeled. The result is a probe set with performance characteristics that can be as narrow or as wide as their designer chooses.

Many factors influence the choice of manufacturing method, such as availability of genomic sequences, of design expertise, cost, resources like pre-existing procedures which have already been validated, and many other variables. These are tools with great value, and FISH is a method with fast turn-around time. While innovative new methods and platforms ever spring within the field of biotechnology, FISH is nevertheless a technology with enough legacy behind it to enjoy widespread use for some time to come. This means demand for FISH probes will continue to exist well into the future as well.