Vesl/Homer proteins are a family of scaffolding molecules, encoded by three genes (Homer-1, Homer-2, and Homer-3) that are abundantly expressed in a variety of tissues, including the brain, retina, cardiac muscle, skeletal muscle, smooth muscle, liver, kidneys, spleen, testis, thymus, placenta, and intestine. Their primary function is to cluster proteins and modulate their activity. Despite their system-wide expression, Homer proteins are best characterized in the brain and the central nervous system, where their primary role is to cluster and modulate the function of synaptic proteins.
Molecular Determinants of Vesl/Homer Proteins
In mammals, Vesl/Homer proteins are encoded by three genes, Homer-1, Homer-2, and Homer-3, that give rise to at least 22 splice variants. Homer-1a and Ania-3 are short isoforms encoded by the Homer-1 gene (Sgambato-Faure et al. 2006; Duncan et al. 2005). Both splice variants were originally identified after induction of their expression levels following excitatory synaptic activity, including induced convulsive seizures, long-term potentiation, and ischemic injury. Interestingly, compared to other immediate early genes (IEGs) that typically encode transcription factors, Homer-1a and Ania-3 are synaptic proteins that can directly modulate protein–protein interactions (Sgambato-Faure et al. 2006; Duncan et al. 2005).
Other typically long Homer isoforms are constitutively expressed and likely fulfill different functions compared with the short IEG isoforms of Homer-1 (Duncan et al. 2005).
Homer proteins contain various conserved protein–protein interaction sites. Isoforms both short and long possess an Ena/vasodilator-stimulated phosphoprotein homology-1 (EVH1) domain within the first 110 N-terminal amino acids. The EVH1 domain is evolutionarily conserved and mediates the binding of Homer to its interaction partner. Through the EVH1 domain, Homer binds receptors and channel proteins such as the metabotropic glutamate receptor (mGluR), the inositol triphosphate (IP3R), the ryanodine receptors (RyR), and transient receptor potential canonical channels (TRPCs). Other targets of Homer include GTPases, anchoring proteins such as shank, as well as transcription factors, including nuclear factor of activated T cells (NFAT). Furthermore, the Homer EVH1 domain shares the RX5GLGF domain found in PDZ proteins (PSD/Disks/ZO-1 [epithelial tight junction protein]), which is thought to ensure correct targeting of ion channels.
All long Homer isoforms also share a C-terminal coiled-coil (CC) domain containing two leucine zipper motifs, which mediate the multimerization between Homer monomers. It is thought that long Homer isoforms form tetrameric structures with their CC domains aligned in parallel, exposing four EVH1 domains for ligand binding (Hayashi et al. 2006).
Expression and Localization of Homer Isoforms
Homer proteins are present in many different tissues, both excitable and non-excitable. In the brain, Homer isoforms are expressed abundantly and at similar levels in the cortex, hippocampus, and cerebellum. All isoforms are particularly highly expressed in cerebellar Purkinje cells. Similarly, all Homer proteins have been detected in the skeletal and smooth muscle as well as the heart. In non-excitable tissues, Homer protein expression has been reported for the kidney, intestine, lung, spleen, liver, thymus, ovary, and testis. The expression of Homer isoforms has been reviewed in detail by Duncan et al. (2005).
Localization studies have shown that all Homer isoforms co-localize with their binding partner Group I mGluRs in the postsynaptic density of excitatory synapses in the brain (as reviewed in Duncan et al. 2005). Specifically, Homer proteins are predominantly localized to the soma and apical dendrites. For example, in the spinal dorsal horn of the mouse spinal cord, Homer and AMPA receptor GluR2 subunits were colocalized at excitatory synapses, while no colocalization of Homer with the inhibitory synapse-associated clustering protein gephyrin was found (Gutierrez-Mecinas et al. 2016). Subcellular fractionation studies have detected all Homer isoforms in the crude nuclear pellet, synaptosomal pellet, microsomal pellet, and PSD fraction. The additional presence of Homer-2a/Homer-2b in the soluble and synaptic vesicle fractions suggest a role for Homer-2 in receptor trafficking.
Recently, evidence for epigenetic control of the Homer-1a promoter region has been described. Using an established cued fear conditioning paradigm, differential and distinct epigenetic control of the transcriptional regulation of Homer-1a expression in the mouse amygdala and hippocampus was identified (Mahan et al. 2012).
Binding Partners of Homer Proteins
Homer proteins interact with a variety of synaptic proteins. Particularly, the interaction of Homer proteins with calcium channels and neurotransmitter receptors both at the plasma membrane and at the endoplasmic reticulum (ER) has recently attracted interest due to the possible involvement of these interactions in age-related diseases of the nervous system.
IP3Rs are intracellular ER channels of which activation results in calcium release from the ER lumen to the cytosol. Homer-1 proteins directly interact with the C-terminus of the IP3R and decrease calcium flux through the channel (Duncan et al. 2005). Interaction with Homer-1c potentiates IP3R activation (Tanaka et al. 2006), resulting in a decrease in dendritic branching (Duncan et al. 2005; Tanaka et al. 2006). The short isoform of Homer-1a has been shown to antagonize this direct interaction, resulting in reduced IP3R activity and a concomitant increase in dendritic branching (Duncan et al. 2005; Tanaka et al. 2006).
Similarly to IP3Rs, ryanodine receptors (RyRs) are large endoplasmic channels that extrude calcium from the ER lumen into the cytosol. Sequence analysis of RyRs has shown a number of EVH1 binding domains, and a direct interaction between Homer and RyR has been experimentally verified (reviewed in Pouliquin and Dulhunty 2009). Binding of Homer proteins to the RyR provides for a physical tethering of the RyR to the cytoskeleton and for RyR cross talk with plasma membrane proteins (Duncan et al. 2005; Pouliquin and Dulhunty 2009). However, the direct effects of Homer proteins on calcium release through the RyR are subject to controversy and speculation (Duncan et al. 2005; Pouliquin and Dulhunty 2009; Westhoff et al. 2003; Hwang et al. 2003).
Group 1 metabotropic glutamate receptors (mGluRs) are plasma membrane neurotransmitter receptors, which when activated initiate a pathway resulting in calcium release through the IP3R. Homer proteins have been found to traffic Group 1 mGluRs to the plasma membrane of neurons (Duncan et al. 2005). mGluR binding of long Homer isoforms (specifically Homer-1b and Homer-1c) creates a ready pool of mGluRs on the ER membrane. Upregulation of Homer-1a displaces the long Homer isoforms and facilitates the trafficking of mGluR to the plasma membrane (Duncan et al. 2005; Kammermeier 2008). Competition between long and short Homer isoforms hence serves as a cellular mechanism of regulating neuronal excitability in response to stress and/or injury. Recent data furthermore indicates that Homer proteins facilitate the cross talk between mGluRs and N-methyl-d-aspartic acid NMDA receptors (Bertaso et al. 2010), corroborating the modulatory role of Homer in excitatory glutamate signaling.
Transient receptor potential canonical (TRPC) channels are plasma membrane, nonspecific cation channels. Interestingly, TRPC channels have the requisite EVH1 sites at both the N- and C-termini to facilitate the cooperative binding of Homer proteins. Binding of the long Homer isoform inhibits the TRPC channel, where the short isoform has the opposite effect (Duncan et al. 2005). An upregulation of Homer-1a in response to depletion of ER calcium stores thus provides a mechanism for activation TRPC channels and subsequent store refilling (Duncan et al. 2005; Rychkov and Barritt 2007). Binding of Homer to polycystin-1 has implications for regulation of intracellular calcium signaling in a number of cell types and organs and has been reported for the central nervous system (Stokely et al. 2006).
As discussed previously, Homer proteins interact and link plasma membrane receptors/channels to intracellular ER channels at neuronal synapses. In addition, Homer proteins also interact with a variety of other proteins that facilitate and/or modulate synaptic function.
The Dynamins form part of a superfamily of proteins that participate in membrane trafficking events following localization to cytoplasmic and membrane compartments (Reems et al. 2008). Homer-1 and Homer-2 proteins bind to Dynamin III through their EVH1 domain, whereas the physical link between Dynamin III and Homer has been reported to be involved with positioning the endocytic zone near the PSD (Duncan et al. 2005; Lu et al. 2007). As Dynamin III also interacts with mGluR5, Homer proteins have been suggested to play a role in recruiting Dynamin III to mGluR5 at the PSD (Duncan et al. 2005).
Shank is a scaffolding protein to which Homer can bind through the EVH1 domain and C-terminal leucine zipper motifs (Duncan et al. 2005). Crystallographic analysis of the postsynaptic density revealed that Homer together with Shank binds to form a mesh-like matrix structure, where the Homer–Shank complex may provide structural support to neuronal dendritic spines and provide an assembly stage for other PSD proteins (Hayashi et al. 2009). Overexpression of Homer-1c in hippocampal neurons resulted in the synaptic localization of Shank being reduced and actin being increased (Duncan et al. 2005). Additionally, depolymerization of actin reduced synaptic localization of both Homer-1c and Shank, suggesting Homer-1c to be involved in the accumulation of synaptic F-actin (Duncan et al. 2005).
Cupidin, or Homer-2, interacts with actin cytoskeletal regulators, Cdc42 and Drebrin, in dendritic spines (Shiraishi-Yamaguchi et al. 2009). The interaction between Cupidin and activated Cdc42 has been suggested to possess a possible role in the formation of mushroom-type spines in hippocampal neurons (Shiraishi-Yamaguchi et al. 2009). Drebrin is a dendritic spine F-actin binding protein and interacts with Cupidin via the N-terminal EVH1 domain (Shiraishi-Yamaguchi et al. 2009). These interactions suggest that Cupidin may play an important role in spine morphology by scaffolding multiple dendritic spine actin regulators (Shiraishi-Yamaguchi et al. 2009).
Huntington interacting protein (Hip1) and Hip1 protein interactor (Hippi) form complexes and activate caspcase-8 which leads to mammalian cell death during Huntington’s disease (Sakamoto et al. 2007). Homer-1c has been shown to protect striatal neurons from Hippi–Hip1-induced cell death. A mutant form of Homer-1c lacking the C-terminal region failed to protect the neurons from cell death or bind to Hippi, suggesting that a direct interaction between the Homer-1c C-terminal region and Hippi is required to induce protection and concluding that a Homer-1c/Hippi complex may be an important regulator for neuronal death during Huntington’s disease (Sakamoto et al. 2007).
Soluble N-ethyl-maleimide-sensitive attachment protein receptor (SNARE) syntaxin 13 can bind with Homer-1c, which may participate in endosomal trafficking as co-expression in COS-7 cells results in the co-localization of Homer-1c and syntaxin 13 in intracellular vesicular structures (Duncan et al. 2005).
Role of Homer Proteins in Disease Pathophysiology and Drug Development
Given the central role of Homer proteins in synaptic clustering and plasticity, it is not surprising that Homer proteins have been implicated in a variety of disease pathophysiologies. Following ischemic injury to the retina, Homer-1c was identified as an early marker for subtle changes prior to more severe neurodegeneration (Kaja et al. 2003). Similarly, lower levels of Homer-1a protein correlated with a decline in cognitive function in aged mice as assessed with behavioral paradigms (Kaja et al. 2013). Additionally, transcranial magnetic stimulation, which enhances hippocampal long-term potentiation and reduces cortical excitability, led to the Homer-1a-dependent recovery of suppressed large conductance calcium-activated potassium channel activity in a transgenic Alzheimer’s disease mouse model (Wang et al. 2015).
Haas and colleagues determined that cellular prion protein associates with a complex of intracellular proteins including Homer-1b/Homer-1c via the transmembrane protein mGluR5 in AD models (Haas et al. 2016), which emphasizes the critical role of Homer in controlling neuronal excitability and synaptic plasticity, but also its potential as a drug target.
In the retina of a preclinical model for glaucoma, levels of Homer-1c mRNA and protein were increased, and the increase was correlated with a decrease in visual performance, while Homer-1a levels were not affected, indicating that differential expression of Homer isoforms could be a target of drug development (Kaja et al. 2014). Similarly, enhancement of stress-induced upregulation of Homer-1a was identified as the mechanistic framework underlying electroconvulsive therapy for refractory and drug-resistant forms of depression (Kaastrup Müller et al. 2015), and knockdown of Homer-1b/c in vivo reduced chemically induced seizure severity through inhibition of mTor signaling (Cao et al. 2015).
Taken together, their role in controlling neuronal excitability makes Homer proteins a viable target for drug development.
Synaptic functions of Homer proteins are to a large extent regulated by the interplay of the short Homer-1a isoform and the various long Homer proteins. Homer-1a regulates long isoforms in a dominant-negative fashion by competing with the long isoforms for Homer binding sites on Homer ligands, demonstrating the potential involvement of the inducible short Homer isoforms in the regulation of intracellular signaling and trafficking processes controlled by ubiquitously expressed long Homer isoforms. Furthermore, the general mechanism of short and long Homer proteins suggests a general stimulus-dependent mechanism using the molecular determinants of Vesl/Homer proteins to regulate intracellular signaling pathways (Duncan et al. 2005).
Homer proteins are a family of scaffolding molecules, encoded by three genes: Homer-1, Homer-2, and Homer-3. Two short Homer-1 isoforms are known (Homer-1a and Ania-3) that act as IEGs and are upregulated in response to seizures, LTP, and ischemic insult; in contrast, all long Homer isoforms are constitutively expressed. The long Homer isoforms are abundantly expressed in the brain, muscle, and various non-excitable tissues. Homer proteins possess an EVH1 domain for ligand binding and interaction and a CC domain providing for tetramerization of multiple Homer proteins into scaffolding complexes. Homer proteins physically link plasma membrane proteins (such as Group I mGluRs) with intracellular Ca2+ channels (such as the IP3Rs and RyRs) and hence are critically modulating synaptic activity. The short IEG Homer-1 transcripts lack a CC domain and are thought to competitively disturb this physical interaction in a response to increased synaptic activity as experienced during LTP or following seizure activity or ischemic insult (Sgambato-Faure et al. 2006; Duncan et al. 2005).
Homer proteins are important modulators of synaptic activity. However, their role in disease and in non-excitable tissues requires further investigation and could critically contribute to protective mechanisms (Duncan et al. 2010).
This chapter is part of the Encyclopedia of Signaling Molecules, 2nd Edition and is based on a chapter of the same name previously published in the first edition of Encyclopedia of Signaling Molecules. This study was supported in part by grants EY014227 and EY022774 from NIH/NEI; RR022570 and RR027093 from NIH/NCRR and NIH/NIGMS; and AG010485, AG022550, and AG027956 from NIH/NIA, by the Felix and Carmen Sabates Missouri Endowed Chair in Vision Research (PK) and the Dr. John P. and Therese E. Mulcahy Endowed Professorship in Ophthalmology (SK). We thank Margaret, Richard, and Sara Koulen for generous support and encouragement.
- Kaastrup Müller H, Orlowski D, Reidies Bjarkam C, Wegener G, Elfving B. Potential roles for Homer 1 and Spinophilin in the preventive effect of electroconvulsive seizures on stress-induced CA3c dendritic retraction in the hippocampus. Eur Neuropsychopharmacol. 2015;25:1324–31. PMID: 25935093.PubMedCrossRefGoogle Scholar
- Wang F, Zhang Y, Wang L, Sun P, Luo X, Ishigaki Y, Sugai T, Yamamoto R, Kato N. Improvement of spatial learning by facilitating large-conductance calcium-activated potassium channel with transcranial magnetic stimulation in Alzheimer's disease model mice. Neuropharmacology. 2015;97:210–9. PMID: 26051398.PubMedCrossRefGoogle Scholar