A nuclear localization signal (or NLS) is a small part of a DNA sequence that signals for a protein to be imported to the nucleus of a cell. Usually coding for several lysine and arginine amino acid residues, this small portion of a gene is embedded within the cDNA of the protein that it is effecting. Basically, this signal "tags" a protein to be processed in a special manner.
Nuclear localization signals were discovered in 1984 by Kalderon et al. By point mutating the simian virus 40 sequence, they were able to turn a nuclear localized protein into a cytoplasmically targeted one. The researchers discovered the region that dictated this behavior by mutation of the surrounding sequence, and thus the NLS was discovered.
The NLS of a tagged protein is first recognized by the alpha subunit of the alpha beta importin complex in the cytoplasm. Importins are a superfamily of proteins that regulate translocation of proteins. The tagged cargo, now bound to the complex, then translocates across the nuclear membrane via the nuclear pore complex, or NPC. This relocation requires the presence of guanosine triphosphatase Ran, guanosine triphosphate, and physiological temperature (37 degrees C works fine).
After traversing the nuclear membrane, the complex then dissociates in the nucleus. This occurs when RanGTP binds with the importin's beta subunit. This binding causes a conformational change in the alpha subunit, which then releases the NLS tagged protein into the nucleus. This is a directional mechanism due to the dissociation of the complex after import, and its reassembly in the cytoplasm.
Some importins only transfer NLS tagged proteins during certain stages of cell cycles, and some have even been hypothesized to be tumor suppressor proteins.
The interest that exists in NLSs is not only for investigating endogenous purposes, but also for the engineering of other proteins to be imported into the nucleus. The signal is small (4-8 amino acid residues, or 12-24 base pairs), therefore it is easily imbedded in the sequences of such reporter genes as eGFP and beta-galactosidase. Expression of the protein in the nucleus means that it is due to your gene being expressed in a cell, which is of some concern with proteins targeted to the cytoplasm. These proteins that lack NLS targeting can be translocated across intracellular space and be expressed in other cells. This means that the protein that your gene is creating is giving false positives, in a mechanism known as pseudotransduction. The NLS targeting of a protein traps that protein within one cell, and does not allow for this phenomenon.
In 1998, the field of bioinformatics came up with a handy program that predicts NLSs in a query protein. It utilizes an Hidden Markov Model (HMM) to find NLSs in an amino acid sequence.
Of other interest to researchers, NLS targeted proteins are thought of as being more easily seen, especially in vivo. I personally find the nuclei harder to see in vitro, but maybe that is due to poor microscopy skills, and not the actual visibility.
ICR annual report, Vol. 5, 1998; pages 52-53
Nature, 311, 33-38 (1984)
General Lab Knowledge