Skip to main content Skip to secondary navigation
Main content start

Should I stay, or should I go: measuring brain-wide synaptic protein lifetime at the single synapse level

Rapid and reliable signaling between neurons is essential for communication in the brain. Chapters 2 and 3 of Principles of Neurobiology detail how information is received by and subsequently transmitted from one neuron to another. In these chapters, we learned how electrical signals, arising from changes in ion concentration across a neuron’s membrane, propagate down an axon (Sections 2.10–2.12) to trigger neurotransmitter release from its axon terminals (Section 3.1). Neurotransmitter is released at synapses which are comprised of presynaptic active zones, containing all of the machinery necessary for transmitter release, juxtaposed by a postsynaptic specialization housing neurotransmitter receptors and scaffolding proteins.

Though synapses are the fundamental units of neuronal communication, the proteins comprising them have finite lifetimes and are continuously being synthesized and degraded in response to both physiological and pathological cues. This process is called protein turnover and it allows for the removal and replacement of old or defective synaptic proteins with newly synthesized ones. Synaptic protein turnover is critical to maintain synaptic function. For instance, synaptic proteins can be damaged or become worn out and slow turnover of these proteins would likely impair synapse integrity. On the other hand, learning (discussed in Chapter 11) is thought to induce specific modifications on synaptic proteins that aid in maintaining the memory of the learned event. Therefore, high levels protein turnover at these synapses may erase learning-induced modifications and adversely impact memory retention. Thus, all synapses do not experience turnover at the same rates and differences in protein turnover may underlie the functions of distinct synaptic connections.

In a recent study, Bulovaite et al., measured synaptic protein lifetime at single synapse resolution across the entire mouse brain. Specifically, the authors genetically coupled endogenous PSD-95, the postsynaptic scaffolding protein enriched in glutamatergic synapses (Figure 3–27), to a protein called HaloTag. Systemic injection of the HaloTag ligand, a small molecule that penetrates through tissues and cells and irreversibly binds the PSD-95-HaloTag, enables visualization and date-stamping of the PSD-95 protein within excitatory synapses in the brain (Figure 1A). By varying the ligand injection time or the when the PSD-95 is visualized, Bulovaite et al., assayed PSD-95 dynamics in mice of different ages and in an autism and schizophrenia mouse model. 

The authors evaluated protein lifetime on three scales. At the brain-wide level, they measured 110 brain regions and observed that excitatory synapses have a wide range of PSD-95 lifetimes extending from hours to several months and synapses with short PSD-95 lifetimes were enriched in younger animals. They defined synapses with PSD-95 that lasted hours to days as having a short lifetime and PSD-95 that persisted for weeks or months as having long lifetimes (Figure 1B). Cortical structures contained synapses with the longest protein lifetimes, whereas subcortical regions displayed much shorter protein lifetimes (Figure 1C). The authors next measured PSD-95 within the same region and found differences in PSD-95 lifetime within the same region. The layers of the cortex, for example, displayed decreasing PSD-95 lifetimes from the superficial layer to the deeper layers. Additionally, within the hippocampus, a structure important for memory formation and retrieval, CA1 synapses had longer PSD-95 lifetimes than those in the CA2 or CA3 sub-regions. Differences in cell type, connectivity, or activity within these sub-regions may explain the variability in PSD-95 lifetime; however, future experiments are needed to test this. Lastly, the authors measured PSD-95 lifetime in the dendrites of CA1 pyramidal neurons and found that PSD-95 lifetimes varied at distinct synapses within dendrites of the same cell. Together, these data indicate a high level of region-, cell-type-, and subcellular compartment-specific control of PSD-95 lifetime.

Finally, the authors evaluated how PSD-95 lifetime is affected by mutations that lead to neurological and neurodevelopmental disorders. Dlg2 encodes a postsynaptic protein that can cooperate with PSD-95 to cluster neurotransmitter receptors and ion channels and mutations in this gene can cause autism and schizophrenia in humans. Bulovaite et al., found that PSD-95 lifetime increased in all brain regions in Dlg2 mutants compared to controls (Figure 1C). The cortex and hippocampus, two regions that show dysfunction in individuals with autism or schizophrenia, were the brain areas most affected by loss of Dlg2. These data indicate that Dlg2 plays a role in postsynaptic protein lifetime and may suggest that PSD-95 turnover is defective in autism and/or schizophrenia.

Bulovaite et al., provide a comprehensive look at PSD-95 dynamics in both physiological and pathological states. They pave the way for future studies evaluating the necessity for distinct PSD-95 lifetimes at the brain-wide or single-cell level, as well as those testing the cellular factors regulating postsynaptic protein duration. Yet, it is still unclear whether the changes in PSD-95 reflect protein turnover or whether they result from the wholesale removal of a synaptic connection. It is likely that both processes can contribute to PSD-95 dynamics, and future experiments parsing the two explanations may provide insight into synapse/synaptic protein regulation. Nevertheless, this tool has broad applications for understanding how PSD-95 dynamics contribute to brain function and how they may go awry in disease or disordered brain states.

Figure 1. (A) Experimental workflow using PSD-95-HaloTag mice to measure PSD-95 lifetime in the mouse brain. (B) Schematic of PSD-95 labeling in postsynaptic regions with short and long PSD-95 lifetimes. (C) Overview of experimentally determined PSD-95 lifetimes in distinct ages and disease models. Adapted from Bulovaite et al., 2022.



Bulovaite, B. et al. (2022) A brain atlas of synapse protein lifetime across the mouse lifespan. Neuron.