Modulation of cholinergic activity through lynx prototoxins: Implications for cognition and anxiety regulation.

The functional importance of the cholinergic system

The cholinergic system helps to transduce experiences through the modulation of functional circuits in the brain. It is a neurotransmitter system implicated in many complex cognitive processes; including attention, fear and anxiety, memory storage and cognitive flexibility. Cholinergic neurons fire in response to salient stimuli which contributes to heightened attentional states. The release of the excitatory neurotransmitter, acetylcholine, functions to amplify signal to noise ratios (Disney et al., 2007). For example, behaviorally driven cholinergic transients have been shown to aid in cue-detection (Sarter et al., 2014), maintaining information in the presence of distractors (Suzuki et al., 1997) and thus influencing attentional performance. In addition, cholinergic activity has been detected in response to unexpected reward (Hangya et al., 2015), and sustained attention (Gill et al., 2000; Himmelheber et al., 2000).

The cholinergic system aids in the encoding of experiences into long-term memories (Letzkus et al., 2011). The cholinergic signalling plays a role in the sleep-wake cycle and memory consolidation during sleep. Activation of cholinergic neurons is associated with sleep-wake cycles, and during wakefulness acetylcholine facilitates thalamocortical signaling by directly exciting thalamocortical relay neurons while reducing activity in the reticular nucleus of the thalamus (Yamakawa et al., 2016). In addition, blockade of cholinergic signalling during REM sleep impaires off-line consolidation motor skills (Rasch et al., 2009), and low levels of acetylcholine during slow wave sleep mediate critical declarative memory consolidation processes (Gais and Born, 2004). Due to the complex nature of the attentional and memory-driven functions mediated by acetylcholine, several regulation mechanisms of the cholinergic system exist.

Structural elements of the cholinergic system

Cholinergic neuronal cell bodies are relatively localized, with cholinergic neuronal cell bodies resident in the basal forebrain and brain-stem. Their terminals, however, radiate broadly throughout the central nervous system (Zoli et al., 1999; Dani, 2001; Zaborszky, 2002; Dani and Bertrand, 2007; Guo et al., 2015). The release of acetylcholine is not confined within the synaptic cleft and thus extra-synaptic transmission to cholinergic receptors is possible. The widely radiating axonal projections of these discretely clustered cholinergic neurons might suggest a broad action for cholinergic signaling were it not for several layers of regulatory mechanisms. The connectivity of cholinergic neurons to their targets subserves functionally and spatially selective signaling (Zaborszky, 2002; Ballinger et al., 2016, Li et al., 2018). Previously, cholinergic signaling was thought to be carried largely through volume transmission with a slow diffuse release and spatially broad changes in ACh concentration over time and space (Runfeldt et al., 2014; Sarter et al., 2009). Emerging evidence from studies utilizing new technological advances such as optogenetics, however, suggests a role for temporally precise ACh release with acute cholinergic signaling (Ballinger et al., 2016). For instance, recent studies provide evidence for precise functional connectivity from cholinergic projections neurons to specific targets that results in a spatially selective transmission (Chandler and Waterhouse, 2012; Bloem et al., 2014; Unal et al., 2015; Zouridakis et al., 2019, Li et al., 2018). This precise functional connectivity may subserve distinct functioning to specific stimuli and represent targeted networks rather than the classical model of broad cholinergic signaling. Cholinergic projection neurons from the basal forebrain to the basolateral amygdala (BLA) can produce varying responses depending on the BLA cell type, which may result in state-dependent behaviors such as learning in fear conditioning only during specific patterns of unconditioned and conditioned stimuli presentations (Unal et al., 2015). In another study basal forebrain projections to the medial prefrontal cortex were demonstrated to display a frontocaudal organization with correlations between the rostral/caudal position in the basal forebrain with distribution in mPFC (Bloem et al., 2014). The role of cholinergic functional connectivity may facilitate control over precise stimuli or represent coordinated regulation between several brain regions for complex behaviors that is different from the regulation ensued by volume transmission. Acetylcholine signals through two families of neurotransmitter receptor classes, G-protein-coupled muscarinic receptors (mAChRs) and ion channel-containing nicotinic receptors (nAChRs), which bind muscarine and nicotine, respectively. Nicotinic receptors are present on both pre- and post-synaptic neuronal subdomains. Central cholinergic neurotransmission can therefore alter neuronal excitability by changing the presynaptic release of neurotransmitters, depolarizing neurons on which they are expressed, inducing secondary messenger cascades, and/or coordinating the firing of groups of neurons (Rice and Cragg, 2004; Kawai et al., 2007; Kutlu and Gould, 2015).

Both decreases and increases in cholinergic signaling can have deleterious or suboptimal effects (Picciotto, 2003; Dani and Bertrand, 2007; Picciotto et al., 2012). Optimal operation of the cholinergic system is dependent on several regulatory mechanisms that fine-tune the activity of the cholinergic system (Miwa et al., 2012). Factors include the number and activity of cholinergic neurons, the level of acetylcholine release, the presence of acetylcholinesterase, the state of calcium stores, or receptor composition. Acetylcholinesterase is highly efficient enzyme, breaking down the neurotransmitter acetylcholine, and shortening the duration of acetylcholine signal. Genetic abnormalities within the choline transporter is associated with attention deficit hyperactivity disorder (ADHD) and result in higher than normal levels of ACh synthesis (English et al., 2009), and aberrant cholinergic signaling is associated with schizophrenia (Higley and Picciotto, 2014).

Nicotinic acetylcholine receptors (nAChRs) are nonselective ligand-gated cation channels that exist as pentamers composed of many variations of 15 possible subunits (Changeux et al., 1998; Picciotto et al., 2001). Nicotinic receptors typically exist as heteromeric combinations of α (2–10) and β (2–4) subunits (most commonly α4β2) or as α homopentamers (α7, α9, etc.) (Picciotto, 2003; Albuquerque et al., 2009) and are dispersed on the surface of neurons, including presynaptic terminals, cell bodies, and axons (Hill et al., 1993; McGehee et al., 1995; Wonnacott, 1997; Nashmi and Lester, 2006; Hurst et al., 2013). The most abundant subtypes in the brain are α4β2 and α7 nAChR subtypes. Receptor composition gives rise to specificity of the cholinergic response, as each combination displays distinct biophysical and pharmacological properties, such as agonist affinity and desensitization kinetics (Brown and Wonnacott, 2014). Differences in stoichiometry among heteromeric subunits results in differential response profiles and sensitivity to ligands, for instance the low sensitivity (LS) stoichiometry of α4β2 consists of three α and two β subunits (α4)3(β2)2, whereas the high sensitivity (HS) stoichiometry consists of two α and three β subunits, (α4)2(β2)3 (Marks et al., 1999; Gotti et al., 2007; Govind et al., 2012). The net sum of cholinergic activity and the effects on behavior will depend upon the integration of these multiple factors.

Lynx prototoxins, protein modulators of nicotinic receptors

Nicotinic receptor function is regulated by prototoxin protein modulators of the Ly6/uPAR superfamily. ly6/uPAR superfamily members adopt a three-loop β-rich fold structure stabilized by cysteine bonds (Lyukmanova et al., 2011; Tsetlin, 2015; Vasilyeva et al., 2017; Miwa et al., 2019) that is also observed for elapid snake venom α-neurotoxins that bind to nAChRs and other receptors with high affinity. Snake toxins evolved via functional mimicry of pathways operating in the prey, such as endogenous prototoxins (Fig. 1). The identification of Ly6/uPAR family members in the mammalian brain (Kuhar et al., 1993) and their homology to the cysteine-rich signature of α-neurotoxins, was suggestive that this family of prototoxins might bind to and regulate similar molecular targets (Miwa et al., 1999).

Fig. 1.

Schematic of lynx1 interaction with nAChRs.

A model of lynx1 (orange) is depicted interacting with A. two a4 nAChR subunits (from Nissen et al., 2018) and B. two a7 nAChR subunits (Lyukmanova et al., 2011). From Hoffman et al., 2019.)

Members of the uPAR superfamily include CD59, lymphocyte antigen genes, ly6A-H, transforming growth factor β receptor ectodomains, and uPAR. In total, at least 2583 sequences within seven subfamilies have been identified (Kessler et al., 2017). Further, the human genome contains at least 45 genes encoding the three-fingered domain (Galat et al., 2008). Three-fingered proteins exert an influence over a wide-array of physiological processes, including cell proliferation, differentiation, inflammation, and neuromodulation. Within this large superfamily, prototoxin members with significant expression in the brain include lynx1 (Miwa et al., 1999), lynx2/lypd1 (Dessaud et al., 2006; Wu et al., 2015), lypd6 (Darvas et al., 2009; Zhang et al., 2009), lypd6B (Demars and Morishita, 2014), PSCA (Jensen et al., 2015), and Ly6H (Horie et al., 1998). These are considered to be peripheral membrane proteins, attached via a GPI-anchor embedded into the neuronal membrane. The focus of this review will explore the in vivo role of prototoxins, members of this family expressed in the brain.

Biophysical mechanism of lynx1 action

The first characterized and most well-studied brain-expressed member of the prototoxin family is lynx1. Lynx1 exerts its modulatory effect on the cholinergic system by direct interactions with nAChRs. In pull-down experiments from rat cortical extracts, lynx1 has been shown to interact with all nAChR subunits tested (α3–7, β2, and β4) (Thomsen et al., 2016). lynx1 has been shown to functionally modulate α4β2 (Ibanez-Tallon et al., 2002; Nichols et al., 2014), α3β4, α5α3β4 (George et al., 2017), α6 (Parker et al., 2017), and α7 (Lyukmanova et al., 2011; 2013) nicotinic acetylcholine receptor subtypes. The influence of this interaction on heteromeric nAChR function can be multifactorial, influencing agonist affinity, desensitization kinetics, receptor number at the cell surface, and single-channel kinetics (Ibanez-Tallon et al., 2002; George et al., 2017; Nichols et al., 2014) and dependent on nAChR subtype (Parker et al., 2017), and isoform of lynx1, whether GPI-anchored or soluble (Lyukmanova et al., 2013; Thomsen et al., 2014).

Oocytes co-expressing α4β2 nicotinic receptors and lynx1 have a faster rate of desensitization to agonist, acetylcholine, and the agonist sensitivity is reduced, as assessed by a right-ward shift in the EC50 to acetylcholine (Ibanez-Tallon et al., 2002). These effects could be due to changes in receptor stoichiometry or the gating functions of lynx1. Removal of the GPI anchor by PI-PLC treatment did not alter the ACh dose-response properties of α4β2 nAChRs (Nichols et al., 2014), suggesting that influence on nAChR stoichiometry through receptor assembly could be the predominant effect of lynx1. A shift in the single channel species from faster inactivating, larger amplitude currents openings (Ibanez-Tallon et al., 2002), commonly thought to be correlated with LS nAChR stoichiometry, supports the idea the lynx1 can influence stoichiometry during nAChR assembly.

GPI-anchored vs secreted versions of prototoxins

Clear gating effects of lynx1 applied acutely as an engineered water-soluble version (ws-lynx1), have also been demonstrated, which are differentiated from the native GPI-anchored native membrane protein. Ws-lynx1 has an inhibitory effect on agonist sensitivity and peak amplitude and can inhibit α7, α4β2, and α3β2, although the functional effects were more pronounced with α4β2 nAChRs. The inhibitory effect is concentration specific and apparently can act in a non-competitive manner (Lyukmanova et al., 2011; 2013). Evidence supports lynx1 function on both gating and receptor assembly of α4β2 nAChRs. It should be noted that differential effects have been reported for lynx1 when co-expressed with nAChRs - and thus GPI-anchored membrane proteins - (Ibanez-Tallon et al., 2002) as opposed to when delivered as a soluble form of protein (Lyukmanova et al., 2013; Miwa and Walz, 2012). Thus in vitro studies on ws-lynx1 should be considered in this context. It is clear from the actions of elapid snake toxins and secreted mammalian family members such as SLURPs, that significant gating effects are capable when prototoxins are bound to the nAChR and not otherwise attached to the membrane (Vasilyeva et al., 2017; Durek et al., 2017; Chimienti et al., 2003; Lyukmanova et al., 2016). The functional effects of SLURPs on nAChRs have been reported mostly outside the central nervous system (Adeyo et al., 2014; Chimienti et al., 2003).

Concatameric nAChR studies

To constrain the number of variables between multiple factors of stoichiometry, receptor number, and gating, concatemeric nAChRs have been used to fix nAChR receptor stoichiometry (George et al., 2017). In these studies, the five subunit cDNAs for α3β4 nAChRs were fused into a single polypeptide, inhibiting any potential heterogeneity in stoichiometry at the fifth position. In these studies, the effect of lynx1 on either the (α3)2(β4)3 or (α3)2(β4)3 nAChR stoichiometry can be carried out independently. The identity of the nAChR subunit in the fifth position is an important determinant of desensitization and agonist sensitivity (Wu and Lukas, 2011). In these studies, lynx1 reduced cell surface expression but not gating of (α3)2(β4)3, whereas it had marked single-channel effects of the (α3)3(β4)2 stoichiometry (decreased unitary conductance, altered burst proportions, and enhanced long closed dwell-times). The differential effect of lynx1 on stoichiometry lends support to the hypothesis, generated from data on α4β2 nAChRs (Nichols et al., 2014), that lynx1 could have preferential binding affinity for α:α over β:β interfaces. Reduced cell-surface expression and increased closed dwell times accounted for the reduction in (α3)2 (β4)2α5 function mediated by lynx1. More defined structure-function studies will be required before a clearer understanding can be determined.

Structural information on lynx1 nAChR complexes

The lynx1 protein demonstrates some topological features of snake venom toxins, such as the the three loop toxin fold(Lyukmanova et al., 2011). α-btx is one of the most widely used snake toxins for the study of nAChRs. It exhibits nearly irreversible affinity for nicotinic receptors, while in contrast lynx1 is able to compete reversibly with other nAChR ligands. Residues on loop II and III are important for the interaction although the effects can be more diverse than that seen with toxin binding and function. There is limited structural information for prototoxins, relative to α-neurotoxins (Tsetlin, 2015), due to the lack of crystallographic data for lynx1. nAChR subunit structural information is emerging as more subunits (Kouvatsos et al., 2014; Zouridakis et al., 2019), or AChBP complexes (Shahsavar et al., 2012; Kaczanowska et al., 2014) are being reported (Giastas et al., 2018). The NMR solution structure of lynx1 has been solved, indicating the β-sheet rich three finger structure reminiscent of the toxin fold of α-neurotoxins (Lyukmanova et al., 2013; Tsetlin, 2015). This structure indicates the lynx1 contains multiple β-sheets forming the first and second loop structures and disulfide pairing similar to α-neurotoxins. This solution structure also indicates a flexible, relatively disordered third loop, a feature which could make crystallization efforts more difficult. Mutagenesis studies have indicated residues which are also important for lynx1 binding or function (Lyukmanova et al., 2013), particularly in the key second loop region which has been mapped as important for binding within the related α-neurotoxins (Tsetlin, 2015). Computational models of lynx1 with nAChRs have indicated possible interactions within important parts of the nAChR structure, for instance the cys-loop and C-loops (Lyukmanova et al., 2013; Nissen et al., 2018; Dong et al., in press), which have been shown to be involved in the transduction of agonist binding to the receptor open state, and with ligand binding, respectively.

These computational models support interfacial binding of lynx1 to the nAChR, for instance with α7 or (Lyukmanova et al., 2011; 2013), or α4* (Nissen et al., 2018) nAChRs=. This helps to bolster vitro experimental evidence of interfacial binding in studies carried out in mammalian cells (Nichols et al., 2014) and oocytes (George et al., 2017).

Cortical plasticity influenced by lynx prototoxins

The overall effect of lynx1 on the major nAChR subtypes primarily acts as a negative allosteric modulator of nAChRs. The role of prototoxins in vivo has been addressed through transgenesis and knockout technologies. lynx1 null mutant (lynx1KO) mice have been a useful research tool for the understanding of the role of lynx1 in plasticity and learning and memory behaviors. The function of lynx1 through nAChRs has been shown to mediate plasticity in the adult visual cortex and auditory cortex (Morishita et al., 2010; Takesian et al., 2018). lynx1 mRNA and protein levels increase in the primary visual cortex (V1) during the time of closure of the critical period in primary visual cortex. Removal of lynx1 in lynx1KO mice lead to extended ocular dominance (OD) plasticity in adulthood (Morishita et al., 2010; Sadahiro et al., 2016). The mechanism of lynx1 action on ocular dominance plasticity is correlated with a functional association of lynx1 with tissue plasminogen activator (tPA), a molecule implicated in spine turnover (Morishita et al., 2010; Bukhari et al., 2015). The turnover of spines in V1 layers 5 and L2/3 pyramidal neuronal dendrites are doubled in lynx1KO mice, and there is a higher loss rate in layer 5 (Sajo et al., 2016). These observations indicate a role for lynx1 in the structural remodeling and spine dynamics required for plasticity in the visual cortex.

Multiple plasticity periods altered by lynx1 regulation

Within the auditory cortex, a nearly 2-fold developmental increase of lynx1 expression has been observed in the primary auditory cortex (A1) between postnatal days P11 and P20. This is consistent with the closing of the critical period in the primary auditory cortex. This was accompanied by a decrease in nAChR sensitivity in 5-HT3AR positive cells, as compared to plasticity in the visual cortex. lynx1 has also been linked to a reduction in auditory plasticity via association with the α4-containing nAChR in 5-HT3AR positive cells (Takesian et al., 2018). Heightened nicotine sensitivity was observed in A1 neurons of lynx1KO mice which was blocked using the α4 nAChR specific antagonist, DHβE (Takesian et al., 2018). There is an emerging concept that multiple critical periods associated with various brain areas and that critical periods for sensory processing are shorter and end earlier than critical periods for higher complex functions (Morrone, 2010). The reduction of α4 nAChR signaling across development by the expression of lynx1 within specific A1 cells may explain how lynx1 serves as a cortical plasticity brake in that region. It will be interesting to see if control of multiple critical periods can be explained by different temporal regulation of lynx1 within different cortical regions.

Removal of the lynx1 brake on nAChRs augments associative learning

Lynx1KO mice demonstrate augmented cued fear conditioned learning (Miwa et al., 2006), but no differences in basal anxiety or contextual conditioning, suggesting a specific role of lynx1 in associative learning. This type of learning involves fear and anxiety centers, such as the medial prefrontal cortex, amygdala, somatosensory and auditory cortices. The amygdala, auditory cortex, and medial prefrontal cortex receive cholinergic input (Woolf, 1991; Hill et al., 1993; Séguéla et al., 1993; Whalen et al., 1994; Mesulam, 1995; Mark et al., 1996; Passetti et al., 2000; Parikh et al., 2007; Mineur et al., 2007; Mansvelder et al., 2009; Poorthuis et al., 2014). Acetylcholine is released rapidly after aversive stimuli and is important for fear learning. A disinhibitory network is activated to regulate fear conditioning. Pairing of foot shocks with a sound causes activation of basal forebrain afferents to layer 1 interneurons in A1, which in turn inhibit layer 2/3 parvalbumin-positive interneurons (Letzkus et al., 2011). Reduction in intracolumar inhibition results in enhancement of excitatory output by layer IV pyramidal neurons. Such changes within A1 (Takesian et al., 2018) might be sufficient to account for all of the associate leaning augmentations, though other regions have not been explicitly addressed. Within other cortical regions, the mPFC expresses both α7* and α4β2* nAChRs which mediate cholinergic signaling in all layers (Poorthuis et al., 2013; Arroyo et al., 2014; Bloem et al., 2014; Verhoog et al., 2016). This fear conditioning lynx1KO phenotype does not appear to be due to alterations in pain processing, as nociception is not augmented in lynx1KO animals in a standard hot-plate assay (Nissen et al., 2018). Removal of lynx1, rather, has an augmented antinociceptive effect when lynx1KO animals are injected with nicotine relative to wild-type mice. This is consistent with reports that nAChR activation has an antinociceptive effect (Freitas et al., 2013) and hypersensitivity of nAChRs due to lynx1 removal (Miwas et al., 2006).

Subcortical effects of lynx1

In vitro studies indicate a positive functional effect of lynx1 on α6* nAChR activity in nicotine-evoked flux assays in striatal synaptosomes. This is somewhat at odds with the inhibitory role of lynx1 on most other nAChR subtypes studied thus far. Consistent with the restricted expression of α6* nAChRs in dopaminergic neurons, lynx1 KO mice demonstrate reduced levels of nicotine-evoked dopamine release from striatal synaptosomes. On the other hand, dopaminergic neurons from lynx1KO mice do not demonstrate changes in dopaminergic firing, nicotine-elicited responses in dopaminergic neurons, or dopamine measurements in fast-scan voltammetry studies (Parker et al., 2017). The lack of clear dopaminergic phenotype could be due to compensation due to the multiplicity of nAChR subtypes expressed within dopaminergic neurons. The composition of nAChRs expressed in the VTA dopaminergic projections to the NAc include α3, α4, α5, α6, α7 and β2, β3, and β4 subunits in various combinations (Klink et al., 2001; Mansvelder et al., 2009). Most VTA GABAergic neurons express α4 and β2 subunits while most VTA pre-synaptic glutamatergic terminals express α7* nAChRs (McGehee et al., 1995; Mansvelder et al., 2009). Cholinergic regulation of dopamine has been widely studied in the context of nicotine addiction. Nicotine can stimulate dopamine release in the NAc (Picciotto et al., 1998). Similarly, stimulation of VTA nAChRs increases excitability of dopamine neurons (Corrigall et al., 2002).

Complementary expression patterns of prototoxins in the brain

Spatial control over nicotinic receptor responses can also be achieved because prototoxins exhibit mostly non-overlapping expression patterns in the CNS (Miwa et al., 2012). For instance, within the hippocampus, lynx1 is expressed in the CA3 subfield and select cells in the hilar region, whereas lynx2 is expressed in the CA1 subfield and dentate gyrus. In general, however, the expression patterns of prototoxins are widespread. For instance, lynx1 has highest levels in the hippocampus and cerebellum (Miwa et al., 1999), but is also found extensively throughout other brain regions, and can be found in the retina (Maneu et al., 2010), the lung (Fu et al., 2012), and the spinal cord (Meyer, 2014). In addition to differential spatial expression, lynx1 has an interesting temporal expression profile with expression beginning at around postnatal week 2 or 3 in a mouse model (Miwa et al., 1999; Thomsen et al., 2014) correlating with the close of the critical period in the visual system. Other prototoxins, such as lypd6, are expressed in more limiting patterns, for instance, to defined GABAergic subpopulations (Demars and Morishita, 2014). The cell type and region specific expression can influence an array of distinct functions, in addition to the nAChR subtype binding specificity, and gating function on those receptors.

Orchestrating cholinergic responsiveness through opposing lynx gene expression: lypd6 and lynx1

Prototoxins exist within a larger family of genes (Tsetlin, 2015). Some evidence suggests that lynx1 opposed some of the effects of another prototoxin family member, lypd6. One piece of evidence lies in responsiveness to somatosensory processing. Overexpression of lypd6 (e.g. Synapsin-driven lypd6 transgenic mice) produced a reduced sensitivity to painful stimuli. In these studies, antinociception was also assessed by the writhing test, measuring the response to acetic acid (Darvas et al., 2006). These animals showed decreased responsiveness in this test. Nicotine-evoked nAChR currents in trigeminal neurons displayed higher calcium fluxes in transgenic mice as compared to those of wild-type mice. While the studies differed in the types of assays, genetic removal of lynx1 in lynx1 KO mice or addition of lypd6 in transgenic mice seems to produce similar cellular effects (e.g. elevated calcium levels), suggesting that they act differently at the biophysical level on nAChRs (Parker et al., 2017). These differences appear to manifest in apposite at the behavioral level too with respect to analgesia (Nissen et al., 2018; Darvas et al., 2006).

The pattern of expression of prototoxin family members within distinct interneurons places them in a position to potentially sculpt the activity patterns of circuits in the visual cortex. The two prototoxins, lynx1 and lypd6, demonstrate complementary temporal and spatial expression patterns within inhibitory subpopulations in the visual cortex (Fig. 2). Lynx1 is found in parvalbumin positive interneurons, whereas lypd6 is found in somatostatin-positive interneurons localized to deep cortical layers, a cell type in which lynx1 was not detected (Demars and Morishita, 2014). Somatostatin-positive interneurons synapse robustly onto parvalbumin-positive interneurons. Considering that lypd6 can augment the calcium component of nicotine-evoked currents, the release of acetylcholine could increase the inhibitory drive more robustly in somatostatin-positive interneurons, as compared to parvalbumin ones due to the differential expression of these two prototoxins and lypd6 is highly expressed in other brain regions and spinal cord of mice and humans (Darvas et al., 2009; Zhang et al., 2009). Lypd6 has been shown to form complexes with α3, α4, α5, α6, α7, β2, and β4 nAChR subunits and competes with α-btx for binding α7 subunits (Arvaniti et al., 2016). Previous studies have also reported, however, that blockade of α7 with α-btx and methyllycaconitine does not affect the modulation of nicotine-induced currents by lypd6 (Darvas et al., 2009). The function of lypd6 has been demonstrated in several model systems. For example, in PC12 cells, a soluble version of lypd6 completely inhibits nicotine-induced phosphorylation of ERK, which is an important pathway activated during plasticity induction. Furthermore, lypd6 KO mice exhibit decreased baseline levels of anxiety-like behavior in two-independent behavioral assessments (i.e., elevated plus maze and marble burying tests) (Arvaniti et al., 2018). Lypd6, however, also contains a Nxl motif, which allows it to bind LRP5/6, a member of the Wnt signaling pathway (Zhao et al., 2018), and therefore some of the phenotypes may be mediated by the Wnt coreceptor, low density lipoprotein receptor-related protein 6 LRP5/6.

Fig. 2.

Selective expression of lynx1 and lypd6 prototoxins within circuits. From Demars and Morishita (2014).

Lynx1 and disease relevance

Alzheimer’s Disease (AD) pathology is associated with an increase in soluble β-amyloid (Aβ), a peptide cleaved from the amyloid precursor protein. Aβ been shown to gain entry through nAChRs and to elicit toxic effects (Thomsen et al., 2016; Inestrosa et al., 2013). There is evidence that lynx1 and Aβ1–42 compete for binding to nAChRs (Thomsen et al., 2016). In pull-down experiments from rat cortical extracts, water soluble lynx1 (ws-lynx1) pulled down all nAChR subunits tested (α3–7, β2, and β4), but the only subunits in which Aβ1–42 led to reduced lynx1/nAChR interactions were the α3, α4, α5, and α7 nAChR subunits. In contrast, the α6, β2, and β4 nAChR subunits were not sensitive to Aβ1–42 competition (Thomsen et al., 2016). Although the authors speculated that the interactions occurred at the cell membrane these interactions were insensitive to β subunits suggesting a significant association of lynx1 with individual α subunits or non-pentameric receptors. Interactions of a nAChR lacking a β subunit are likely to occur beneath the membrane surface, consistent with the reported interaction of lynx1 with nAChR dimers in the endoplasmic reticulum prior to receptor maturation (Nichols et al., 2014). Conversely, when Aβ1–42 was used to pull-down nAChR complexes, it also pulled down all nAChRs tested, and lynx1 could compete at α7 and β2 subunits. Such findings are in accordance with the results of previous studies, which reported that Aβ1–42 can bind α7, α4β2, and α4α5β2 receptors (Dougherty et al., 2003; Lamb et al., 2005; Wu et al., 2004). These results indicate that lynx1 and Aβ1–42 bind at similar sites on nAChRs. Aβ1–42 is thought to bind at the orthosteric binding site, whereas the lynx1 binding site on nAChR has yet to be mapped.

If the lynx1 and Aβ1–42 interactions are significant in vivo, lynx1 may exert protective effects against the pathological progression of AD (Thomsen et al., 2016; Thomsen and Mikkelsen, 2012). Supporting this, a small but significant (e.g. 10%) reduction in lynx1 messagerRNA is associated with Alzheimer’s pathology (Thomsen et al., 2016), and wslynx1 has been shown to block the inhibitory effect of Aβ1–42 on long-term potentiation (Bychkov et al., 2018). A protective role for lynx is supported by in vitro (Miwa et al., 2006) and in vivo (Miwa et al., 2006; Kobayashi et al., 2014) studies in lynx1KO mice. Thus, further studies are required to elucidate the role of Aβ1–42 in normal and disease states (Kroker et al., 2013).

Although the role of lynx1 on disease states has focused primarily on Alzheimer’s, recently (Artoni et al., 2019) utilized the enhanced cholinergic tone of lynx1 KO mice to investigate if alterations in cholinergic circuit alter arousal dynamics similar to those observed in mouse models of Autism Spectrum Disorder (ASM). These studies determined lynx1 KO mice exhibited a shifted distribution toward maximal pupil size similar to ASD model mice (Artoni et al., 2019). lynx1 KO mice were also used to screen for genes differentially regulated and linked to genes in patients with risk for neurodevelopmental disorders such as epilepsy and schizophrenia (Smith et al., 2018).

Lynx2 prototoxin and anxiety regulation

Another member of the lynx family, lynx2, is expressed within key regions of the anxiety response circuitry, namely the amygdala and medial prefrontal cortex. Characteristic of the three-looped structure of the ly6/uPAR super family (Dessaud et al., 2006), the lynx2 protein binds to and suppresses the activity of nAChRs within these regions (Tekinay et al., 2009; Wu et al., 2015). In vitro immunoprecipitation experiments have demonstrated that lynx2 forms stable complexes with α7, α4β2, and α4β4 nAChRs (Tekinay et al., 2009; Wu et al., 2015). Co-expression of lynx2 and α4β2 leads to faster desensitization kinetics in response to acetylcholine (Tekinay et al., 2009) and a shift in the EC50 for acetylcholine (Tekinay et al., 2009), nicotine, and epibatidine (Wu et al., 2015). The presence of lynx2 also decreases the expression of α4β2 at the cell surface, suggesting an additional potential mechanism for the decreased response to agonists (Wu et al., 2015). There is also evidence that the lynx2 protein can blunt nicotine-induced upregulation of α4β2 (Wu et al., 2015).

Key regions associated with anxiety, the amygdala and medial prefrontal cortex, not only express lynx2 but also receive cholinergic input (Woolf, 1991; Hill et al., 1993; Séguéla et al., 1993; Whalen et al., 1994; Mesulam, 1995; Mark et al., 1996; Passetti et al., 2000; Parikh et al., 2007; Mineur et al., 2007; Mansvelder et al., 2009). Nicotinic receptors have been implicated in the regulation of anxiety responses (Picciotto, 2003; Klein and Yakel, 2006; Gozzi et al., 2010; Mineur et al., 2016; Jiang et al., 2016; Wilson and Fadel, 2017). Anecdotal evidence comes from smokers wherein individuals have reported using nicotine to ameliorate anxiety symptoms (Moylan et al., 2012). nAChRs have been shown to regulate activity in anxiety/fear circuits and have been linked to fear and anxiety-related behaviors in animal studies, particularly the amygdala and substructures of the amygdala such as the basolateral amygdala (BLA)(Picciotto, 2003; Klein and Yakel, 2006; Gozzi et al., 2010; Mineur et al., 2016; Jiang et al., 2016; Wilson and Fadel, 2017).

Endogenous acetylcholine modulates excitability of the BLA, as well as stimulating cortical-BLA inputs. Activation of nAChRs with nicotine increases glutamatergic transmission in the BLA and post-synaptic glutamatergic currents from cortical inputs into the BLA, whereas blockade of nAChRs decreases activity in the BLA (Mineur et al., 2007; Jiang and Role, 2008; Jiang et al., 2016). Behavioral consequences of altered nAChR signaling is highly dependent upon the activity of the amygdala and its inputs. For example, α7 nAChRs can modulate the activity of both excitatory and inhibitory neurons, and the outcome depends upon the starting conditions (Jiang and Role, 2008; Pidoplichko et al., 2013). Both of the most common nAChR subtypes are involved in anxiety-like behavior as focal knockdown of α7* and β2* (*indicates containing that subunit) within the amygdala have anxiolytic-like effects (Mineur et al., 2016). Only β2* nAChRs were shown to be involved in social defeat (Mineur et al., 2016). In fear extinction, activation of cholinergic terminals in the BLA hinders the acquisition of extinction and maintains the fear memory by increasing the firing of BLA principal neurons (Jiang et al., 2016).

Behaviorally, Lynx2 null mutant mice (lynx2 KO) exhibit increased sensitivity to nicotine in pyramidal neurons in the medial prefrontal cortex as compared to wild-type controls (Tekinay et al., 2009). These data suggest that ligand sensitivity is altered in the presence of lynx2, and that lynx2 also acts to inhibit the activity of nAChRs. The functional consequences of lynx2 deletion include increased anxiety-like behaviors across several paradigms (e.g. light-dark box, thigmotaxis, and elevated plus maze), as well as reduced social interaction in a standard social interaction test (Tekinay et al., 2009). Thus, lynx2 may play an important role in limiting or regulating the function of its cognate receptors to respond adaptively in circuits mediating anxiety-like behavior, and deletion of lynx2 in mice can serve as a robust model of excessive anxiety.

Prototoxin family members

Several members of the ly6/uPAR/neurotoxin superfamily have been shown to bind to and differentially modulate the function of multiple nAChRs (Wu et al., 2015; Arvaniti et al., 2016; Puddifoot et al., 2015) (Table 1). One example, LYPD6 interacts with α3, α4, α5, α6, α7, β2, and β4 nAChR subunits and is expressed in somatostatin interneurons of the V1 layers 5 and 6 (Darvas et al., 2009). Other prototoxins, however, are known to show decided preferences specific nAChR partners. For example, LYPD6B has been shown to modulate α3β4-containing but not α7-containing nAChRs (Ochoa et al., 2016), Ly6g6e interacts with potentiated α4β2 nAChRs (Wu et al., 2015), and Ly6h interacts with α7 nAChRs (Puddifoot et al., 2015) Prostate stem cell antigen (PSCA), another prototoxin modulator of nAChR function, has a preferential modulatory effect on α7-containing but not α3β4-containing nAChRs (Hruska et al., 2009). Recent work demonstrates that prototoxins can also have significant preferences for interacting with particular subunit interfaces within nAChR complexes.

Table 1

Spatial Control Over the Central Nervous System. Overview of select Ly6 family members with known expression within the rodent central nervous system.

Spatial Control Over the Central Nervous System
Gene	Expression patterns within rodent brain	Behavioral Consequences	nAChR Interactions	References
Lynx1	Substantia nigra, dorsal raphe nucleus, hippocampus, cortex, cerebellum, Isocortex, olfactory areas, cortical subplate, straitum, pallidum, thalamus, midbrain, hypothalamus, pons, and medulla; temporally, expression in mice is upregulated around postnatal week.	The loss of lynx1 in a knockout mouse results in enhanced associative learning, enhanced antinociception to nicotine or epibatidine, and can enhance motor activity	Interacts with α7, α4β2, α6, and α3β2.Lynx1 presence shifts Ach EC50 to the right and results in increased agonist desensitization	Miwa et al., 1999; Ibanez-Tallon et al., 2002; Lyukmanova et al., 2013; Thomsen et al., 2014; Parker et al., 2017; Nissen et al., 2018; Allen Brain Map
Lynx2/lypd1	Basolateral amygdala, medial prefrontal cortex, Isocortex, olfactory areas, hippocampal formation, cortical subplate, straitum, pallidum, thalamus, hypothalamus, midbrain, pons, and medulla.	Loss of lynx2 leads to increased fear- and anxiety-like behavior in a knockout mouse	Interact with a7 and α4β2 nAChRs; enhances receptor desensitization and decreases ACh sensitivity	Dessaud et al., 2006; Tekinay et al., 2009; Wu et al., 2015; Allen Brain Map
Lypd6	Known expression in somatostatin interneurons of V1 layers 5 and 6	Lypd6 overexpression results in increased locomototion. hypoalgesia, and pre-pulse inhibition of the acoustic startle response while the loss of lypd6 in knockout mice decreases anxiety-like behavior. In humans, mutations are reported to result in developmental delay, hypotonia, and autistic features.	Interacts with α3, α4, α5, α6, α7, β2, and β4 nAChR subunits, Lypd6 presence enhances the Ca2+-component of nicotine-evoked currents	Darvas et al., 2009; Chandler and Waterhouse, 2012 (22085900); Zhang et al., 2009; Arvaniti et al., 2018; Allen Brain Map
PSCA	Isocortex, olfactory areas, hippocampal formation, cortical subplate, striatum, pallidum, thalamus, midbrain, hypothalamus, pons, medulla, ciliary ganglion, neural and choroid plexus	PSCA is up-regulated in the cortex of AD patients	Interacts with α4 nAChRs, decreases nicotine-induced kinase phosphorylation, interacts with α7-containing chick nAChRs	Hruska et al., 2009; Jensen et al., 2015; Ochoa et al., 2016; Allen Brain Map
Ly6g6e	Olfactory areas, hippocampus, cortical areas, and midbrain	No in vivo rodent behavioral studies; hypothesized to regulate nicotine addiction	Interacts with potentiates α4β2 nAChRs and slows desensitization but effects persist in absence of extracellular calcium	Wu et al., 2015; Allen Brain Map
Ly6h	Frontal cortex, Isocortex, olfactory areas, hippocampal formation, cortical subplate, straitum, pallidum, thalamus, hypothalamus, midbrain, pons, and medulla; a temporal pattern exists: expression is first observed around postnatal week 1–2 in mouse models	Ly6h enhances nicotine-induced potentiation of hippocampal pyramidal neurons; hypothesized role in nicotine addiction	Interacts with α7 nAChRs and causes a right-ward shift of nicotine or epibatidine-evoked responses, demonstrated to both influence and not interact with α4β2 nAChRs in different assays	Horie et al., 1998 (9799603); Artoni et al., 2019; Tekinay et al., 2009; Thomsen et al., 2016; Puddifoot et al., 2015; Wu et al., 2015; Allen Brain Map
Lypd6B	Visual cortex, Isocortex, olfactory areas, hippocampal formation, cortical subplate, pallidum, hypothalamus, midbrain, pons, and medulla	Mutations in humans are linked to intellectual disability and developmental delay; no in vivo rodent behavioral studies	Inhibits α3β4-, not α7-mediated nAChR currents; specifically for α3β4: Lypd6B presence reduces EC50 for ACh in nAChRs containing 3 (not 2) α3 subunits	Chandler and Waterhouse, 2012 (22085900); Demars and Morishita 2014 (25359633); Ochoa et al., 2016 and 2019; Allen Brain Map

In addition to the proteins discussed prior, several members of the prototoxins family are expressed and function peripherally to inhibit signalling of the nicotinic system (e.g., SLURP1, SLURP2, lypdg6e, PATE-M, PATE-B, etc.). Although each of them plays an important role in signalling regulation and are therefore are worth mentioning, their distribution and targets remain generally peripheral and ultimately fall outside of the scope of this review.

Therapeutic potential of nicotinic receptors

Considerable efforts have been made to develop agonists against nAChRs for a number of indications. Nicotinic receptors suffer from some challenges as a therapeutic target. They are part of a large gene family with very similar sequences, making them difficult to target selectively. Secondly, nicotinic receptors are wide-spread throughout the brain, so a drug could have multiple functional consequences (Mineur and Picciotto, 2008). Nicotinic receptors also desensitize quickly and down-regulate, making selective targeting a short-term proposition in some cases (Quik and Wonnacott, 2011), thus subject to tachyphylaxis. Lastly, neurons usually express more than one subtype generally, and the potential for receptor compensation is a complexity that requires more exploration. Therefore combinatorial targeting of multiple subtypes within a neuron, which might be an ideal strategy, is a challenge. Also the inverted U-shaped curve of nicotinic receptor based therapeutics means that efficacy and safety are a hurdle higher than other targets (Colquhoun and Patrick, 1997). The ability to successfully create new nicotinic therapeutics has largely been unsuccessful with several failed clinical trials pointing to a translational gap (Bertrand and Terry, 2018; Vieta et al., 2013).

Several features make lynx prototoxins intriguing as possible therapeutic targets for the cholinergic system, and could possibly circumvent some of the issues associated with direct targeting of nicotinic receptors. Lynx genes have a degree of regional selectivity, some more restricted than others (Dessaud et al., 2006; Miwa et al., 1999; Thomsen et al., 2014). The advantage of multiple lynx family members is the possibility of exerting better spatial or temporal control over the cholinergic system by selectively acting on a single lynx family member which is associated with a specific function or region. Thus, targeting the lynx-receptor interface, could increase the possible specificity of action over the receptor alone. This has the potential of lowering unwanted side-effects. Removal of lynx1 leads to reduced levels of desensitization (Ibanez-Tallon et al., 2002), suggesting that lynx-based drugs could be less sensitive to either desensitization or tachyphylaxis. The reported effects on the number, stoichiometry, and desensitization kinetics of nicotinic receptors at the neuronal cell surface due to lynx interactions (George et al., 2017; Wu et al., 2015), suggest changes in nicotinic receptor number could also be a beneficial outcome of lynx targeting, and consideration of possible intracellular effects of a lynx-based therapeutics is warranted (Lester et al., 2012).

Overview

Shaping the responsiveness of acetylcholine through modulation of nAChRs has significant influence on a number of complex brain functions adaptive to the organism. Different prototoxins have been shown to be selectively expressed in different cell types within circuits and have differential binding capacity on nAChR subtypes. At present, the majority of reports of prototoxin function have been inhibitory on α4β2, α3β4, α5α3β4 (e.g. lowering agonist response property, lowering receptor number, or accelerating desensitization of nAChRs, etc.). Exceptions to this are the reports of lypd6, as well as the positive effects of lynx1 for α6β2, and α6* nAChR function. As of yet, no in vivo role of a prototoxin has been reported on mAChRs, so the nicotinic selectivity of prototoxins could also influence the relative degree of muscarinic drive over the nicotinic one in response to acetylcholine release. The detailed understanding of the relative weighting of nicotinic function imparted through nAChR subtypes in complexes with specific prototoxins will require further in vitro and circuit-based investigations. Lynx prototoxin regulation of the cholinergic system holds promise as a therapeutic target for a wide range of disordered states by imparting both spatial and subtype specific regulation of a widespread neurotransmitter system.