Structures of the serotonin transporter protein SERT in complex with two different antidepressants shed light on how these drugs act, and point to possible targets for future drug development. See Article p.334
On page 334 of this issue, Coleman et al.1 describe the first high-resolution structure of the serotonin transporter (SERT) protein. This protein recaptures serotonin molecules that have been released by the cell and so modulates the effects of serotonin on neighbouring neurons. Those not familiar with SERT might question why solving this structure is such a big deal. First, it is technically challenging to purify the large amounts of SERT required for structural determination. Second, visualizing the detailed molecular structure of such a protein could provide unprecedented opportunities to develop more-selective and -efficacious therapies for diseases such as depression.
The neurotransmitter molecules serotonin, dopamine and noradrenaline are released from particular neurons and cross the synaptic cleft to bind to receptors on neighbouring neurons, where they modulate the effects of rapid excitatory or inhibitory signals from other neurotransmitters. SERT, together with the transporters for dopamine (called DAT) and noradrenaline (NET), belongs to the neurotransmitter:sodium symporter (NSS) family, which itself is part of the second-largest class of mammalian membrane-spanning protein — that is, proteins that move ions, small molecules and nutrients into and out of cells. SERT, DAT and NET are selectively expressed on the surfaces of presynaptic neurons2 and, by transporting serotonin, dopamine and noradrenaline back into the cell, act to terminate the modulatory effects of these neurotransmitters.
Genetic ablation studies3 conducted almost 20 years ago showed that the three transporters play an essential part in maintaining everyday control of neuronal signalling. Their significance has been further emphasized by the discovery4 that rare genetic mutations in the genes that encode them are linked to diseases such as autism, attention deficit hyperactivity disorder (ADHD) and parkinsonism.
Drugs such as cocaine, amphetamine and ecstasy exert their psychostimulatory action by hijacking these transporter proteins. Moreover, medications that target SERT, NET and DAT are used to treat depression and ADHD, among other conditions. Antidepressants, including selective serotonin reuptake inhibitors (SSRIs) and serotonin–noradrenaline reuptake inhibitors (SNRIs), block the transporters, thereby preventing reuptake of the neurotransmitter and consequently increasing its availability and so its overall activity in the synapse. In theory, this is how SSRIs and SNRIs alleviate some of the symptoms of depression and other mood-related conditions.
In 2005, the group that carried out the current study determined the structure5 of LeuT, a bacterial relative of SERT. LeuT is involved in nutrient uptake in these unicellular organisms and functions similarly to SERT. In 2013, the same group also solved the structure of DAT in fruit flies6. This study suggested that the structural organization of these proteins is highly evolutionarily conserved, and defined several structural elements that indicated that the mechanism by which NSSs transport their substrates is also conserved. In providing molecular insights into the structure of human SERT, the current study confirms that nature, when it finds a way to do something, does it over and over again. But in the case of SERT, there is an interesting twist.
To solve structures at high resolution using X-ray crystallography, proteins must be purified in large quantities. Membrane-spanning proteins are notoriously difficult to purify, because they are unstable once removed from the hydrophobic-bilayer environment of the membrane. Previously, membrane proteins have been successfully stabilized by introducing mutations, by inducing the formation of a complex with easily crystallized proteins, or even by developing stabilizing antibodies to the native proteins7. Coleman et al. used an elaborate screen to painstakingly identify a few amino-acid residues that could be mutated to stabilize purified SERT without markedly affecting its functional properties. The resulting crystals revealed that the more than 600 amino-acid residues of SERT thread through the cell membrane 12 times to form an intricate 3D structure that is designed to mediate ion-coupled neurotransmitter transport across membranes.
The authors used X-ray crystallography to determine two structures of SERT bound to an SSRI — either paroxetine (Prozac) or escitalopram (Lexapro; the S-enantiomer of citalopram). As expected from the structure of fruit-fly DAT, they found that one molecule of paroxetine bound tightly in a central cavity in SERT that lies deep within the plasma membrane and is also thought to bind serotonin, thus competitively blocking serotonin transport. However, two molecules of escitalopram bound to SERT: one, like paroxetine, in the high-affinity central cavity, and the second more loosely in an outward-facing external vestibule (Fig. 1).
The existence of a second binding site (known as an allosteric site) away from SERT's primary binding site was proposed more than 30 years ago8. In 2012, modelling and mutational experiments9 suggested that the site was situated in the external vestibule, as Coleman et al. now confirm. The previous studies suggested that, when present at sufficiently high concentrations, escitalopram binding in the allosteric site could markedly retard the dissociation of drugs bound to the high-affinity site — thus prolonging the SERT-blocking activity of escitalopram. This 'positive allosteric modifier' phenomenon has been proposed to explain the superior clinical efficacy of escitalopram compared with other SSRIs10, but confirming the mechanism in animal studies has been difficult11. Coleman and colleagues' molecular documentation of the allosteric site is therefore important, because their structure might provide previously unappreciated opportunities for selective drug development.
The presence of allosteric sites on membrane proteins might be more common than anticipated. An example of the possible power of this finding comes from G-protein-coupled receptors (GPCRs) — membrane proteins that bind serotonin, dopamine, noradrenaline and many other signalling molecules on postsynaptic neurons. Successful determination of the structures of more than two dozen of these receptors has identified many potential allosteric binding sites that are being leveraged to help develop positive or negative modifiers of signalling to combat disease12. The same opportunity could now present itself for transporters. Visualization of the molecular dynamics of transporter function, coupled with structure-based molecular modelling of drug binding, offers unprecedented opportunities for developing improved treatments for disorders of the central nervous system.
About this article
Computational identification of the binding mechanism of a triple reuptake inhibitor amitifadine for the treatment of major depressive disorder
Physical Chemistry Chemical Physics (2018)