Answer: mGRASP, or Mammalian GFP Reconstruction Across Synaptic Partners, is an experimental strategy for tracing and visually identifying synaptic connections.
One of the major challenges in neuroscience today is trying to build the human connectome: the complete map of which neurons are connected with which other neurons. Such techniques as diffusion tensor imaging (DTI), rabies virus tracing, or CLARITY are all pushing the field forward towards helping us understand how our neurons are wired together.
An experimental strategy that has been developed to help researchers trace neurons is called mGRASP, or mammalian GFP reconstitution across synaptic partners. The basic premise of mGRASP relies on using genetic tools to create small spots of GFP fluorescence. The technique uses artificial protein constructs which are essentially a normal GFP split into two units. One of these is expressed in a presynaptic transmembrane protein and colocalizes with dendritic proteins (pre-GRASP), and the second unit is expressed in the postsynaptic site and colocalizes with axonal proteins (post-GRASP). When the two proteins come together at a synapse, the GFP molecule is reconstructed, and the site where the two protein constructs meet therefore emits fluorescence. MGRASP is considered revolutionary since the technique has been performed in drosophila and C. elegans, but was not demonstrated in a mammalian model until 2011, when the technique was published by the Magee lab (mGRASP enables mapping mammalian synaptic connectivity with light microscopy).
Each synapse, the non cellular space in between two neurons that act as the site of neuron-to-neuron communication, is about 20-30 nm across. The mGRASP construct itself is a transmembrane protein which protrudes into the extracellular space, and is able to bridge the distance, allowing the pre-GRASP to interact with the post-GRASP protein. This creates fluorescence that can be visualized and quantified. Each spot of GFP represents a single synapse.
MGRASP is an ideal technique whenever a reliable genetic line can be identified, since the mGRASP constructs can be packaged into cre-dependent viral constructs. For example, it is possible to use a cre line to express one of the mGRASP pairs in one select cell population, and a non-cre line to express the other mGRASP partner more broadly. The technique can be applied either for dissecting out microcircuits within a brain region, or for tracing connectivity at a long range.
An alternative strategy to identifying synapses is to use a fluorescence based, viral-mediated transfection approach. By having the virus only function under the control of pre-synaptically or post-synaptically expressed proteins, you can theoretically limit expression of a fluorescence signal to the synaptic sites by searching for overlapping signals. However, there may be difficulty interpreting these fluorescent dots, since it is often difficult to resolve which z-plane the fluorescence signal is located on. MGRASP is a good experimental strategy that allows someone to be more confident in their assessment of whether an overlapping fluorescence signal is a genuine synapse.
One downside to mGRASP is that background noise cannot always be eliminated, as in many microscopy-based imaging techniques. Autofluorescence cannot always be eliminated, which may lead to false positives, inflating the count of synaptic contacts.