GluN2B and GluN2D NMDARs dominate synaptic responses in the adult spinal cord
OPEN
SUBJECT AREAS:
ION CHANNELS IN THE
NERVOUS SYSTEM
CELLULAR NEUROSCIENCE
Received
19 August 2013
Accepted
30 January 2014
Published
13 February 2014
Correspondence and
requests for materials
should be addressed to
M.W.S. (mike.salter@
utoronto.ca)
* Current address:
Department of
GluN2B and GluN2D NMDARs
dominate synaptic responses in the adult
spinal cord
Michael E. Hildebrand1,2*, Graham M. Pitcher1,2, Erika K. Harding1,2, Hongbin Li1,2, Simon Beggs1,2
& Michael W. Salter1,2
1
Program in Neurosciences & Mental Health, Hospital for Sick Children, Toronto, ON, Canada, 2Department of Physiology,
University of Toronto, Toronto, ON, Canada.
The composition of the postsynaptic ionotropic receptors that receive presynaptically released transmitter is
critical not only for transducing and integrating electrical signals but also for coordinating downstream
biochemical signaling pathways. At glutamatergic synapses in the adult CNS an overwhelming body of
evidence indicates that the NMDA receptor (NMDAR) component of synaptic responses is dominated by
NMDARs containing the GluN2A subunit, while NMDARs containing GluN2B, GluN2C, or GluN2D play
minor roles in synaptic transmission. Here, we discovered NMDAR-mediated synaptic responses with
characteristics not described elsewhere in the adult CNS. We found that GluN2A-containing receptors
contribute little to synaptic NMDAR responses while GluN2B dominates at synapses of lamina I neurons in
the adult spinal cord. In addition, we provide evidence for a GluN2D-mediated synaptic NMDAR
component in adult lamina I neurons. Strikingly, the charge transfer mediated by GluN2D far exceeds that
of GluN2A and is comparable to that of GluN2B. Lamina I forms a distinct output pathway from the spinal
pain processing network to the pain networks in the brain. The GluN2D-mediated synaptic responses we
have discovered in lamina I neurons provide the molecular underpinning for slow, prolonged and
feedforward amplification that is a fundamental characteristic of pain.
Neuroscience,
Carleton University,
Ottawa, ON, Canada.
N
-methyl-D-aspartate receptors (NMDARs) are a prominent subtype of ionotropic glutamate receptor1,
critical for physiological synaptic plasticity in the developing and mature CNS, and for aberrant plasticity
and neuronal death in pathological disorders2. The NMDAR is a multiprotein complex, the core of which
is a heterotetrameric assembly of two glycine-binding GluN1 subunits and two glutamate-binding GluN2 subunits. GluN1 is encoded by a single gene, GRIN1, whereas there are four GluN2-subunit-encoding genes, GRIN2A
–D. Because NMDARs comprised of differing GluN2 subunits have distinctive functional properties and are
hypothesized to have differing physiological and pathological roles1,3, a major question has been to determine the
specific GluN1/GluN2 subunit composition contributing to NMDAR-mediated transmission at a given synapse.
GluN2A has emerged as the principal subtype of GluN2 mediating synaptic NMDAR responses in the adult
brain4–8. By contrast, the contribution of NMDARs comprised of GluN2B to synaptic responses is less than that of
receptors comprised of GluN2A throughout the adult brain, and at many synapses GluN2B contributes much less
than does GluN2A. The GluN2B subunit is nevertheless well-expressed in the adult but primarily localized at
extrasynaptic sites9. Conversely, during embryonic and early postnatal development synaptic GluN2B is more
prominent than is GluN2A6,10,11. GluN2C shows sparse expression in the adult brain, except for the cerebellum
and olfactory bulb where GluN2C is highly expressed beginning in the second postnatal week10,12. An enduring
mystery has been that, in contrast to the other GluN2 subunits, synaptic NMDAR responses mediated by GluN2D
are lacking in the adult brain9.
Here, while investigating glutamatergic synaptic responses of neurons in lamina I of the adult spinal dorsal
horn we discovered NMDAR-mediated synaptic responses with characteristics not previously described elsewhere in the adult CNS. Lamina I neurons form a critical part of the nociceptive neuronal network in the dorsal
horn which processes inputs from primary sensory afferents and transmits the resultant nociceptive signals to the
brain13,14. The nociceptive network in the dorsal horn is a highly interconnected matrix comprising monosynaptic
inputs from primary afferents as well as polysynaptic and monosynaptic inputs from local circuit neurons15–17. In
order to properly characterize the responses from glutamatergic synapses directly on lamina I neurons it was
necessary to avoid activating the nociceptive network because such network activation generates prolonged,
SCIENTIFIC REPORTS | 4 : 4094 | DOI: 10.1038/srep04094
1
www.nature.com/scientificreports
Figure 1 | Whole-cell patch recordings from lamina I neurons in rat parasagittal spinal cord slices. (A) Top, representative image of a parasagittal slice
and micropipette during patch-clamp recording of a lamina I neuron. Through visual discrimination, the lamina I neuron was identified in the darker,
striated, superficial-most layer of the dorsal horn (LI) compared to the more ventral, lighter substantia gelatinosa (LII), as indicated. Scale bar 5 200 mm.
Bottom, a two-photon image was taken from a lamina I neuron that was filled with lucifer yellow (0.1%, green) during whole-cell recording. Separately, an
immunohistochemistry image was taken following slice fixation and staining for CGRP (red) and neuronal (NeuN, blue) labeling. The two-photon image
was correctly scaled and superimposed onto the epifluorescence image stack to illustrate the location and orientation of the lamina I neuron within the
superficial dorsal horn. Note neuronal morphology corresponding to a type Ib fusiform lamina I neuron62 as well as neuronal location within the
outermost CGRP-stained dorsal horn region, characteristic of lamina I. Scale bar 5 100 mm. (B) Raster plot of a representative continuous whole-cell
recording from a lamina I neuron held at 260 mV. Scale bar x axis 5 1 s, y axis 5 10 pA.
largely polysynaptic excitatory responses which contaminate the
direct NMDAR-mediated synaptic responses. To this end, in lamina
I neurons we studied miniature excitatory postsynaptic currents
(mEPSCs), representing the overall population of direct synaptic
responses, and unitary primary afferent-evoked EPSCs, representing
individual synapses. Surprisingly, we found that GluN2A contributes
little to synaptic NMDAR responses while GluN2B dominates at
lamina I synapses in adult spinal cord. Moreover, we discovered that
there is a GluN2D-mediated synaptic NMDAR component of neurons in spinal lamina I and that the charge transfer mediated by
GluN2D far exceeds that of GluN2A and is comparable to that of
GluN2B.
Results
AMPAR and NMDAR components of mEPSCs in lamina I
neurons. We made whole-cell recordings from visually identified
SCIENTIFIC REPORTS | 4 : 4094 | DOI: 10.1038/srep04 (...truncated)