A genetically encoded neurotransmitter sensor, engineered from a tick lipocalin, detects serotonin in the living brain.
It is possible that the 5-HT in our brains plays an essential part in keeping us sane. —Sir John Gaddum, 19541.
The serotonin (5-hydroxytryptamine, or 5-HT) system is the most commonly targeted neuronal pathway for treating mental disorders. The system innervates almost the entire brain and is involved in nearly every aspect of human behavior. Adding to the complexity, there are 15 different types of 5-HT receptor2. Despite decades of research, 5-HT remains one of the most mysterious neurotransmitters.
Genetically encoded neurotransmitter sensors have emerged as powerful tools for measuring neurotransmitters in the living brain. These sensors light up fluorescently when they capture a specific neurotransmitter, reporting on the levels of transmitters in real time and with single-cell resolution in the brains of living animals3. In this issue, Zhang et al. introduce a genetically encoded neurotransmitter sensor that can be used to monitor 5-HT release, allowing neuroscientists to study the precise mechanisms of 5-HT signaling4.
A perfect sensor should measure the transmitter within the physiological range without disturbing endogenous processes. This requirement poses a paradox for scientists developing neurotransmitter sensors. Substantial engineering and optimization is necessary to achieve ample brightness, sensitivity to physiological concentrations, high selectivity over other transmitters, and fast binding kinetics while minimizing interactions with host systems.
Genetically encoded neurotransmitter sensors are generally assembled from two components: a transmitter sensing domain and a reporter domain (for example, fluorescent protein). Previously described sensing domains have used G-protein-coupled receptors (GPCRs) or bacterial periplasmic binding proteins (PBPs)3. Recently, two 5-HT sensors have been developed: a GPCR-based sensor named GRAB5-HT (ref. 5) and a PBP-based one named iSeroSnFR (ref. 6) (Fig. 1). The Achilles’ heel of GPCR-based sensors is their slow dissociation kinetics. Meanwhile, PBP-based sensors require extensive effort to improve their sensitivity.
GRAB5-HT is based on serotonin receptor HTR2C. GRAB5-HT1.0 was shown to detect serotonin in cultured cells, mouse brain slices, and living fly and mouse brains5 (Table 1). GRAB5-HT1.0 exhibits up to a threefold fluorescence increase upon 5-HT binding and excellent sensitivity, but slow dissociation kinetics.
iSeroSnFR was designed with the help of machine learning algorithms by modifying a previously described PBP-based acetylcholine sensor, iAChSnFR6. It was applied in cultured cells and neurons and the brains of living mice. As a result of its wide detection range and fast kinetics (Table 1), iSeroSnFR has advantages in reporting 5-HT from populations of cells and screening of drugs targeting the 5-HT system.
To create ideal sensors for different applications, sensor engineering calls for new ideas based on more diverse sensing domain scaffolds. The green fluorescent genetically encoded serotonin sensor (G-GESS) represents a novel design using an untapped class of proteins for the sensor component: the serotonin-binding lipocalin from the soft tick Argas monolakensis. The design was inspired by the ability of blood feeders to release soluble proteins into their hosts at the feeding site, some of which bind naturally to 5-HT.
Zhang et al. inserted a serotonin-binding lipocalin into a circular permutated green fluorescent protein (cpGFP)4. Guided by available structures of lipocalins, the researchers employed directed evolution to optimize insertion linkers and random mutagenesis to create improved sensors. Eventually, G-GESS emerged as the winner, with a fluorescence increase of up to fourfold when bound to 5-HT and a dynamic range compatible with physiological concentrations of 5-HT.
In cultured neurons, G-GESS has high sensitivity, comparable to that of GRAB5-HT, and exhibits superior dissociation kinetics. Like iSeroSnFR, G-GESS can be modified to detect serotonin signaling both inside and outside the cell. The researchers also successfully applied G-GESS to image 5-HT changes in the brains of live mice in response to drugs and environmental stimuli. Moreover, the success of the first lipocalin-based 5-HT sensor suggests the lipocalin family as a promising scaffold for generating more biosensors.
Viewed as “being involved in everything, but responsible for nothing”7, 5-HT has posed enormous challenges for the study of its function and signaling mechanism. It has been argued that the way forward is to divide and conquer the different 5-HT receptors7. But challenges remain. Six out of seven general families of 5-HT receptors are GPCRs, and their function and activation may depend on their subcellular locations and the properties of downstream cells. Furthermore, the connectivity to downstream cells can be ambiguous. Classical fast transmitters like glutamate and GABA often act in a point-to-point manner, faithfully passing information from a neuron’s terminals to its postsynaptic partners. However, neuromodulators like 5-HT may not act on defined recipient cells after being released, but rather diffuse through the extracellular space in a process called volume release. In fact, it is believed that 90% of 5-HT axonal projections utilize this kind of strategy2. Last but not least, some 5-HT neurons can co-release other transmitters when activated2. For these reasons, the relationships among the activity of 5-HT neurons, the release of 5-HT and its influences on the downstream neurons are difficult if not impossible to directly characterize. Burning questions like when, where and how 5-HT is released during normal daily life and in mental illness conditions will not be answered unless 5-HT can be ‘seen’ in real time.
Thanks to the newly developed serotonin sensors, 5-HT in both cultured cells and living animals can be monitored. Further improvements, including increased signal strength (brightness), more robust membrane or other subcellular targeting, and faster dissociation kinetics, are expected to facilitate comprehensive studies at the subcellular level in physiological conditions. Different color variants of sensors are also in high demand to facilitate multiplexed imaging in conjunction with other probes. The development of far-red chemogenetic sensors8 opens news doors for this purpose.
We expect innovative tools like G-GESS to help lift the mysterious veil covering 5-HT. Hopefully, in the near future, we can obtain satisfying answers to many questions around 5-HT. For instance, what triggers 5-HT release? How are the axonal terminals organized to achieve the volume release of 5-HT? How far away can 5-HT diffuse to activate downstream sites? And how do disruptions to these steps contribute to the onset and progression of mental disorders?
Gaddum, J. H. in Hypertension: Humoral and Neurogenic Factors (eds Wolstenholme, G.E.W. & Cameron, M.P.) 75–77 (Ciba Foundation Symposium, 1954).
Muller, C. P. & Cunningham, K. A. Handbook of the Behavioral Neurobiology of Serotonin (Academic Press, 2020).
Sabatini, B. L. & Tian, L. Neuron 108, 17–32 (2020).
Zhang, S., Li, X., Zhao, S., Drobizhev, M. & Ai, H.-w. Nat. Methods https://doi.org/10.1038/s41592-021-01078-7 (2021).
Wan, J. et al. A genetically encoded GRAB sensor for measuring serotonin dynamics in vivo. Preprint at bioRxiv https://doi.org/10.1101/2020.02.24.962282 (2020).
Unger, E. K. et al. Cell 183, 1986–2002.e1926 (2020).
Jacobs, B. L. & Fornal, C. A. in Psychopharmacology: The Fourth Generation of Progress (eds Bloom, F. E. & Kupfer, D. J.) 461–469 (Raven, 1995).
Deo, C. et al. Bright and tunable far-red chemogenetic indicators. Preprint at bioRxiv https://doi.org/10.1101/2020.01.08.898783 (2020).
The authors declare no competing interests.