GFP is no longer just a label. It can now be used to activate a virtually limitless set of functional tools in the cell types that it marks.

The transformation of GFP from reporter to reagent was driven by the desire of Connie Cepko at Harvard University and her colleagues to understand the complex mammalian retina. In the thicket of cell populations in the central nervous system, “you're always looking for ways to functionally assay an individual cell type,” says Cepko.

As a young graduate student in the Cepko lab, Jonathan Tang identified transgenic mice in which interesting subtypes of retinal interneurons were labeled with GFP. The victory soon resembled a dead end as he questioned whether GFP had uses beyond imaging. Digging in the GFP literature, Tang discovered nanobodies.

Nanobodies deliver functional capabilities to GFP-expressing cells. Image courtesy of J. Tang and T. Cherry.

Nanobodies are derived from a single-chain protein found in camelid species and have excellent binding specificity in a much smaller, more stable and less aggregation-prone form than those of typical antibody reagents. “Obviously camelids—these are alpacas, llamas, camels—aren't your garden-variety lab creature,” says Cepko, but a number of nanobodies that bind GFP have been developed to brighten or dim fluorescence or to target reporter fusion proteins for degradation. Tang's insight was to use GFP as a scaffold that brings pairs of GFP-binding nanobody fusion proteins together to reconstitute a functional biological complex, enabling a host of new activities in the cell.

He attached one nanobody to a DNA-binding domain and another to a transcriptional activation domain and then screened a large number of nanobody pairs for the ability to simultaneously bind GFP and transcribe a luciferase reporter in human cells. The best pairs gave several-hundred-fold higher expression than background.

The researchers delivered the GFP-dependent transcription system into mouse retinal tissue by electroporating a mix of three plasmids: two bearing the nanobodies and one carrying a responsive promoter fused to a red fluorescent protein or Cre recombinase. They observed sensitive red fluorescence or Cre readouts in GFP-expressing cells, even when GFP was below levels that could be visualized directly. Little to no target gene expression was detected in cells lacking GFP, with occasional leakiness attributed to the common promoters used.

A plethora of mouse lines exist with genetic elements flanked by recombination sites, so this method should allow a variety of genes to be turned on or off by Cre in GFP-positive cells. The researchers also showed GFP-dependent expression of a light-controlled ion channel in sensory neurons, from which they could record electrical activity. “Pairing GFP specificity with optogenetics power was really exciting,” says Cepko.

Cepko notes that electroporation works robustly with multiple plasmids, though uptake is not uniform across a tissue. The lab is also developing adeno-associated virus delivery vectors and hopes to generate a series of transgenic mice that will express nanobody reagents ubiquitously.

The system is very flexible. The researchers successfully tested the Gal4, LexA and chemically inducible rTetR DNA-binding domains and also tuned the degree of transcriptional activity using variants of VP16 or an alternative activation domain. They proved portability by using the system in zebrafish to direct several types of GFP-expressing embryonic cells to express a red fluorescent protein.

GFP has long been a powerful imaging tool for biologists. As a scaffold it expands the repertoire of countless existing GFP transgenic lines into the functional realm.