Abbreviations used



amyloid precursor protein


ionotropic glutamate receptor


kainate receptor


neuropilin and tolloid-like protein 1




post-synaptic density-95


transmembrane AMPA receptor regulatory protein

N-methyl-D-aspartate (NMDA) receptors (NMDARs) are glutamate-gated ion channels (iGluRs) comprised of two obligatory GluN1 and two GluN2(A-D) pore-forming subunits (Paoletti et al. 2013). NMDARs are crucial for neuronal communication and plasticity including long-term potentiation and long-term depression, which are likely to explain their importance for learning and memory (Paoletti et al. 2013). Furthermore, NMDAR dysfunctions are involved in wide range of neurological and psychiatric disorders and there is major interest in developing new drugs that target these receptors (Collingridge et al. 2013; Paoletti et al. 2013). However, recombinant and native iGluRs often differ in their pharmacological and biophysical properties, which can hinder the drug discovery process. This mismatch suggests that heterologously expressed receptors lack modulatory components that can influence key characteristics (Jackson and Nicoll 2011; Copits and Swanson 2012). The discovery of transmembrane alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptor (AMPAR) regulatory proteins (TARPs) and other transmembrane auxiliary subunits has solved many of these discrepancies for AMPAR subtypes of iGluRs (Jackson and Nicoll 2011). Also, TARPs have been shown to modify the pharmacology of AMPAR agonists and antagonists. For example, the antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) acts as a partial agonist when AMPARs are associated with TARPs (Jackson and Nicoll 2011).

Neuropilin and tolloid-like 1 and 2 (Neto1 and Neto2) were identified as auxiliary subunits of kainate-type iGluRs (KARs) that are responsible for the characteristic slow kinetics and high agonist affinity of native KARs in the central nervous system (Zhang et al. 2009; Copits et al. 2011; Straub et al. 2011; Tang et al. 2011; Copits and Swanson 2012). Although Neto1 is mainly expressed in the hippocampus, particularly where high affinity KARs are known to be enriched (in the CA3 stratum lucidum), Neto2 is expressed in cerebellar granule cells and cortical neurones. Neto2 interacts with GluK2 subunits of KARs to increase peak amplitude and open probability and to slow the decay time course of both GluK2 recombinant receptors and KAR-mediated miniature excitatory postsynaptic currents (mEPSCs) in cerebellar granule cells (Zhang et al. 2009). Neto1 confers high agonist affinity and slow kinetic properties on GluK2 containing KARs in the CA3 region of the hippocampus (Straub et al. 2011). Neto1 seems to have an opposite action on GluK1 subunit-containing KARs (Copits et al. 2011) therefore caution should be taken in generalizing these effects. In recombinant expression systems, Neto2 had no impact on GluK2 surface expression (Zhang et al. 2009). However, Neto2 but not Neto1 promotes surface localisation and delivery to synapses of GluK1-containing KARs in cultured hippocampal neurones (Copits et al. 2011).

The AMPAR and KAR auxiliary subunits described above are defined by four key criteria: (i) they are not an integral component of the channel pore, (ii) they display a direct and stable interaction with pore-forming subunits, (iii) they affect multiple aspects of receptor function, pharmacology and subcellular trafficking or targeting, (iv) their co-assembly is required for proper neuronal functionality of the native receptors in vivo (Jackson and Nicoll 2011; Copits and Swanson 2012; Yan and Tomita 2012). Therefore, auxiliary subunits fundamentally differ from other iGluR interacting proteins that are involved in transient and often dynamic interactions and influence singular aspects of receptor function (e.g. biogenesis, trafficking or synaptic localisation) (Copits and Swanson 2012).

While it is widely recognized that like other iGluRs, NMDARs are also part of a complex multi-protein assembly (Husi et al. 2000; Collingridge et al. 2013), their regulation by transmembrane auxiliary subunits is not clear. It has been reported that in addition to KARs, Neto1 also interacts with GluN2A and GluN2B subunits of NMDARs and is a candidate NMDAR auxiliary subunit (Ng et al. 2009). Neto1 was enriched in the post-synaptic density where it was shown to co-distribute with the scaffold protein, post-synaptic density-95 (PSD-95), and GluN1 NMDAR subunit (Ng et al. 2009). Neto1 was critical for the delivery and/or stability of GluN2A subunit-containing NMDARs within the post-synaptic density (Ng et al. 2009). Furthermore, Neto1 (−/−) mice were found to have deficits in long-term potentiation at Schaffer collateral-CA1 synapses and NMDAR-dependent spatial learning and memory (Ng et al. 2009). Surprisingly, a more recent study found no biochemical or functional interactions between Neto1 and NMDARs (Straub et al. 2011), which raised doubts about the role of Neto1 as NMDAR auxiliary subunit.

In this issue, Cousins et al. (2013), provide new insights into the role of Neto1 as NMDAR-associated protein. The investigations also included the amyloid precursor protein (APP), which also appears to meet some of the criteria for NMDAR auxiliary subunit (Cousins et al. 2009; Hoe et al. 2009). The authors employed carefully controlled co-immunoprecipitation for the analysis of recombinant and native NMDAR complexes. The results indicate that Neto1 co-immunoprecipitates with heteromeric NMDARs via GluN2A or GluN2B subunits (Cousins et al. 2013). The previously reported contradictory observations (Ng et al. 2009; Straub et al. 2011) can be explained by the differential solubility of unassembled GluN1 and heteromeric GluN1/GluN2 subunit assemblies by various detergents and differences in immunoprecipitation conditions (Wenthold et al. 1999). While Neto1 does not co-immunoprecipitate with APP directly, Neto1, APP, GluN1/GluN2A or GluN1/GluN2B are all part of the same protein complex (Cousins et al. 2013) (Fig. 1). Studies of GluN1/GluN2A chimeras and truncated GluN2A constructs revealed that the GluN2A intracellular C-terminal domain mediates Neto1 association (Cousins et al. 2013) (Fig. 1). This is in contrast with previous studies where the extracellular GluN2A domains were identified as Neto1 interaction sites (Ng et al. 2009). These new findings indicate that the interaction between Neto1 and GluN2 subunits is indirect (Fig. 1), and therefore that Neto1 is member of the NMDAR signalling complex rather than a genuine auxiliary subunit for this family of iGluRs. While Neto1 can reduce surface expression of recombinant NMDARs with no change in total subunit levels, this effect is not apparent in the presence of PSD-95 or APP, both of which can facilitate NMDAR surface expression (Cousins et al. 2009, 2013). Because the majority of endogenous NMDARs are normally associated with PSD-95 (Kornau et al. 1995), it is unlikely that the Neto1 enhanced internalisation and/or decreased receptor insertion into the plasma membrane have a major effect on NMDARs in vivo.

In summary, Cousins and colleagues demonstrated that Neto1 is part of the NMDAR/APP trafficking complex (Cousins et al. 2013). The results indicate that the interaction of Neto1 with NMDARs is indirect (Fig. 1). These findings suggest that Neto1 is not an auxiliary subunit of NMDARs. Because the NMDAR/Neto1 association is not mediated via the obvious candidate scaffolding protein synapse associated protein 102 (SAP102) (Cousins et al. 2013), future studies need to identify the linker(s) between Neto1 and GluN2A/GluN2B. The other auxiliary subunit candidate APP and the scaffolding protein PSD-95 both facilitate NMDAR surface expression, which was unaltered in the presence of Neto1 (Cousins et al. 2013). Therefore, the previously reported role of Neto1 in the normal synaptic maintenance of GluN2A (Ng et al. 2009) is probably because of the clustering and/or stabilisation of GluN2A-containing NMDARs at the post-synaptic density. The studies of Neto1 and APP interactions with NMDAR illustrate that it is often very difficult to distinguish auxiliary subunits from transient interactors. However, better understanding of the molecular organisation/interactions of native iGluR complexes is essential for the understanding of their functional diversity. Future studies should also address the role of auxiliary subunits in influencing iGluR pharmacology, as this may have a significant impact on drug development.


Figure 1. Schematic representation of possible interactions between N-methyl-D-aspartate receptors (NMDARs), neuropilin and tolloid-like protein 1 (Neto1) and amyloid precursor protein (APP). The diagram depicts possible interactions by which Neto1 and APP associate with the NMDAR complexes (based on Cousins et al. 2009; Jackson and Nicoll 2011; Copits and Swanson 2012; Cousins et al. 2013). See text for details.

Download figure to PowerPoint


  1. Top of page
  2. Acknowledgements
  3. Conflicts of interest
  4. References

EM's research is supported by the Biotechnology and Biological Sciences Research Council, UK (grant BB/J015938/1). I thank Dr. Andrew Doherty for his help with the preparation of Fig. 1.


  1. Top of page
  2. Acknowledgements
  3. Conflicts of interest
  4. References