Previous Article | Next Article 
The Journal of Neuroscience, March 15, 2003, 23(6):2323
Absence of Whisker-Related Pattern Formation in Mice with NMDA
Receptors Lacking Coincidence Detection Properties and Calcium
Signaling
York
Rudhard1, 2, *,
Matthias
Kneussel1, 2, *,
Mohammed A.
Nassar1, 2,
Georg F.
Rast1, 2,
Alexander
J.
Annala1, 2,
Philip E.
Chen1, 2,
Cezar M.
Tigaret1, 2,
Isabel
Dean1, 4,
Juergen
Roes5,
Alasdair J.
Gibb2,
Stephen P.
Hunt3, and
Ralf
Schoepfer1, 2
1 Wellcome Laboratory for Molecular Pharmacology,
2 Department of Pharmacology, 3 Department of
Anatomy, 4 Wellcome Trust Neuroscience PhD program, and
5 Department of Medicine, University College London, London
WC1E 6BT, United Kingdom
 |
ABSTRACT |
Precise refinement of synaptic connectivity is the result of
activity-dependent mechanisms in which coincidence-dependent calcium
signaling by NMDA receptors (NMDARs) under control of the
voltage-dependent Mg2+ block might play a special role.
In the developing rodent trigeminal system, the pattern of synaptic
connections between whisker-specific inputs and their target cells in
the brainstem is refined to form functionally and morphologically
distinct units (barrelettes). To test the role of NMDA receptor
signaling in this process, we introduced the N598R mutation into the
native NR1 gene. This leads to the expression of functional NMDARs that
are Mg2+ insensitive and Ca2+ impermeable.
Newborn mice expressing exclusively NR1 N598R-containing NMDARs do not
show any whisker-related patterning in the brainstem, whereas the
topographic projection of trigeminal afferents and gross brain
morphology appear normal. Furthermore, the NR1 N598R mutation does not
affect expression levels of NMDAR subunits and other important
neurotransmitter receptors.
Our results show that coincidence detection by, and/or
Ca2+ permeability of, NMDARs is necessary for the
development of somatotopic maps in the brainstem and suggest that
highly specific signaling underlies synaptic refinement.
Key words:
barrelette; somatosensory; whisker; trigeminal
pathway; pattern formation; topographic map; NMDA receptor; NMDAR; coincidence detection; point mutation; knock-in; homologous
recombination; cytochrome oxidase; DiI labeling; Mg2+ block; Ca2+-dependent
signaling; brainstem; Cre recombinase; loxP; Tenascin; boundary
| |
Introduction |
The precise pattern of
neuronal connectivity results from both activity-dependent and
activity-independent mechanisms (Goodman and Shatz, 1993
). Current
models implicate coincident activity of independent inputs in the
refinement of synaptic connections within developing topographic maps.
The rodent trigeminal pathway is an excellent system to study the
molecular mechanisms underlying activity-dependent map formation (Molnar and Hannan, 2000; Erzurumlu and Kind, 2001). Whisker-related sensory inputs are topographically mapped at multiple relay stations from the trigeminal input to the primary sensory cortex. At each level,
presynaptic afferents and postsynaptic target cells form discrete
cytoarchitectural and functional units that replicate the whisker
pattern on the muzzle in a one-to-one relationship (Woolsey and Van der
Loos, 1970; Van der Loos, 1976; Ma and Woolsey, 1984; Woolsey,
1990).
Whisker-related patterns develop along a peripheral to central temporal
gradient (Woolsey, 1990). At birth, patterning can be detected by
staining for cytochrome oxidase (CO) at the first central synapse of
the pathway, located in the brainstem trigeminal complex (BSTC) (Ma,
1993; Li et al., 1994; Jhaveri et al., 1998). Pattern formation
then progresses to thalamic barreloids, followed by cortical barrels.
Coincidence detection-dependent calcium signaling by NMDA receptors
(NMDARs) may be crucial in activity-dependent synaptic refinement. The
role of NMDAR-mediated activity in the development of whisker-related
patterns has been studied pharmacologically and genetically. Local
application of the NMDAR antagonist APV disrupted the topographic
refinement of thalamocortical connectivity (Schlaggar et al., 1993; Fox
et al., 1996). Deletion of the NMDAR NR1 or NR2B gene prevents the
formation of barrelettes in the brainstem (Li et al., 1994; Kutsuwada
et al., 1996; Iwasato et al., 1997), and cortex-restricted
deletion of the NR1 gene impairs whisker-related patterning in the
barrel cortex (Iwasato et al., 2000). Thus, NMDAR-mediated activity is
essential for the refinement of whisker-related synaptic patterns.
However, the relevance of coincidence detection by the NMDAR has not
been addressed so far.
Special biophysical properties enable the NMDAR to act as coincidence
detector (Bourne and Nicoll, 1993). Removal of the channel-blocking Mg2+ ion by postsynaptic depolarization in
synchrony with presynaptically released glutamate is necessary to open
the NMDAR channel. The Ca2+ signal
generated by such a coincidence event may trigger signaling cascades,
ultimately leading to changes in synaptic connectivity (Goodman and
Shatz, 1993).
Both Mg2+ block and
Ca2+ permeability are abolished by the N
(Asp) to R (Arg) point mutation at the Q/R/N site in the channel-lining region of membrane domain M2 (Burnashev et al., 1992; Sakurada et al.,
1993). NMDARs are heterooligomers formed from developmentally and
regionally regulated NR2 subunits (A to D) and the mandatory NR1
subunit (for review, see Cull-Candy et al., 2001). Thus, the introduction of the N598R mutation in the NR1 subunit (NR1 N598R) impairs coincidence detection and Ca2+
signaling of all NMDARs.
Using homologous recombination in embryonic stem (ES) cells, we
now generated a mouse model with NR1 N598R mutant NMDARs. We found that
whisker-related somatosensory pattern formation (barrelettes) is
impaired in these mice. Our results show that the presence of
functional NMDARs is not sufficient for this type of pattern formation
when their Mg2+ block and
Ca2+ permeability are impaired.
| |
Materials and Methods |
Animal procedures. Mice were kept at the Biological
Services Unit of the University College London, and procedures on mice were performed according to the Animal Scientific Procedures Act of
1986 and under license of the Home Office.
Generation and analysis of mutant alleles and mice. The NR1
N598R targeting vector was constructed from 129 SVJ mouse genomic DNA
(Stratagene, La Jolla, CA).
A four-primer PCR created a fragment from exons 14 to 17, harboring the N598R mutation, silent mutations at codon 599, a
diagnostic AvaII site overlapping codon 598/599, and a
silent SalI site in exon 14 (oligos always in 5'-3'
notation, GluN-PCR25s (CCTTTCAGTCGACACTGTGGCTGCTGGTGGGGC), GluN-PCR26a
(AGCCACAGTGTCGACTGAAAGGGCTGCATGAATG), GluN-PCR6a
(CAGGACGCCCCAGGAAAACCACAT), and GluN-Mut2s (GTGGTTTTCCTGGGGCGTCCTGCTGCGGTCCGGC).
The floxed herpes simplex virus promoter (HSV)-neo
selection cassette from pL2-neor was
inserted into the BamHI site in intron 18. The HSV-tk
selection cassette from pIC19R/MC1-TK (Mansour et al., 1988) was added
5' to the NR1 targeting sequences. The final targeting vector had a 6.2 kb long 5' arm and a 0.9 kb short 3' arm flanking the 1.3 kb
floxed neo cassette and was linearized 3' of the NR1 sequences.
Transfected embryonic day 14 (E14) ES cells (Handyside et al.,
1989) that were G418 resistant were screened for homologous recombination by PCR [primers Neo-Seq-1 (GGTGTTGGTCGTTTGTTCGGATCT) and
mNR1-PCR-26a (TGGTGTCAGAGGTGCTTGGATGAT)]. The presence of mutations at
codons 598/599 were confirmed by AvaII digests and sequencing of PCR-products [primers mNR1-Seq-29s
(TCCCTTTGGCCGATTTAAGGTGAA) and mNR1-Seq-30a
(GAAAGCTGGGAATGCAGCCATCCA)]. PCR-positive clones were verified by
Southern blot analysis of SpeI/XbaI-digested ES
cell DNA (probe: SacII/KpnI cDNA
fragment binding to exon 8 and exon 9) and of
SalI/BglII digests (probe B: the genomic
ApaI/SmaI fragment, introns 20 and 21). Probe B
was also used on blots of EcoRV-digested mouse tail DNA as
shown in Figure 1.
In addition, a neo and a HSV-tk probe were used to exclude additional
random integration of the targeting vector. Correctly targeted ES cell
clones were injected into C57BL/6J blastocysts, and chimeric animals
were backcrossed with C57BL/6J mice to yield NR1+/Rneo animals. Mating I is
NR1+/Rneo × NR1+/Rneo.
Mice with the NR1 null allele (NR1+/
)
were produced likewise (details to be described elsewhere) by insertion
of an 8 kb large DNA cassette containing the HSV-neo gene into exon 1 upstream of the NR1 coding sequence. Probe A
(Bsu36I/XhoI DNA fragment, 5' of exon 1) was used
on blots of EcoRV-digested mouse tail DNA for genotyping.
Generation of mice with five different NR1 genotypes.
NR1+/Rneo,
NR1+/, and "Deleter" mice carrying
an X-chromosomally linked transgene for the recombinase Cre from
bacteriophage (Schwenk et al., 1995) were each backcrossed with
C57BL/6J to at least F6 generation.
Genotyping was done by PCR on mouse tail DNA, after the assay had been
verified by genomic Southern blots. The Cre transgene yielded a 486 bp
product [primers Cre-Seq1-s (AGATGTTCGCGATTATCTTCTA) and Cre-Seq2-a
(AGCTACACCAGAGACGG)]. NR1+/ mice were
intercrossed (mating III) to yield
NR1/ mice [primers mNR1-Seq38-s
(ACCAGTCGCACAGTCCAGGCAGCT) together with GluN-Seq3a
(GGCGTTGAGCTGTATCTTCC) and TA-Seq2-a (CTAGCTTCTGGGCGAGTTTACGGGT); wild-type product, 404 bp; null allele product, 323 bp].
Global activation of the NR1 N598R allele by mating
NR1+/Rneo mice with Deleter mice is
dominant lethal. Thus, to generate mice exclusively expressing the NR1
N598R mutant allele, female Deleter Cre+/+
mice were first bred with NR1+/ mice to
obtain
NR1+//Cre+/
mice. Backcrossing of
NR1+//Cre+/
with Cre+/+ mice yielded
NR1+//Cre+/+
mice. Female
NR1+//Cre+/+
mice were mated overnight with male
NR1+/Rneo mice (mating II) to obtain
NR1R/ mice along with
NR1R/+,
NR1+/, and
NR1+/+ mice.
The morning after the mating was referred to as E0. All four genotypes
could be distinguished by multiplex PCR using primers 1 (mNR1-Seq103-s,
GTCCATACTCAAGTGAGTCTGCCC), 2 (mNR1-Seq10-a, CAGGGGCATTGCTGCGGGAGTC), 3 (Neo-Seq4-s, GCTGCATACGCTTGATCCGGCTACC), and 4 (Neo-Seq3-a,
GAAGGCGATAGAAGGCGATGCGC), generating 508 and 615 bp PCR products for
the NR1+ and
NR1R alleles, respectively (the increase
in size in NR1R is attributable to
the remaining loxP site and adjacent polylinker sequence),
and 414 bp for the neo cassette, indicating the knock-out allele.
mRNA quantification. Reverse transcription (RT)-PCR
(Ready-to-Go RT-PCR; Amersham Biosciences, Arlington
Heights, IL) (oligo-dT primed first strand) was performed on total RNA
from postnatal day 0 (P0) whole brain from
NR1R/+ mice, using the primers
mNR1-PCR101s (CAGGGTACCTCCCTTTGGCCGATTTAAGGTGAA) and
mNR1-PCR104a (TGAGAATTCCAGGGGCATTGCTGCGGGAGTC). The resulting amplicons of 665 bp covering exons 14-19 of the NR1 allele were subcloned into M13.
Phage plaques harboring NR1+ or
NR1R cDNA inserts, respectively, were
quantified by differential oligonucleotide hybridization (42°C in
50% formamide; followed by final stringency wash in 0.3× SSPE
[1× SSPE (in mM): 150 NaCl, 10 NaH2PO4, and 1 EDTA, pH 7.4] at 60°C) with
oligonucleotides mNR1-Is3-a, (CCCCAATGCCAGAGTTGAGCAGGACGCCCCAG) and
mNR1-Is4-a (CCCCAATGCCgGAccgcAGCAGGACGCCCCAG) for wild-type and mutant
alleles, respectively (Higuchi et al., 1993). The relative abundance of
NR1+ and NR1R
transcripts was determined as the ratio of the number of plaques hybridizing with one probe to the total number of NR1
transcript-containing plaques. A number of
NR1+- and
NR1R- containing plaques were verified by
DNA sequencing.
Northern blot analysis.
Poly(A+) RNA was prepared from whole brain
of newborn mice, and Northern blot analysis was performed as described
previously (Specht and Schoepfer, 2001) using a 318 bp cDNA fragment
binding to NR1 exons 15-17. Reprobing with a 0.9 kb actin cDNA
fragment served as control for equal loading.
Western blot analysis. Membrane protein fraction was
prepared from frozen P0 brain of genotyped mice as described previously (Forrest et al., 1994), except that 100 mg of tissue were homogenized in 2 ml volume, and protein degradation was inhibited using Complete protease inhibitor cocktail (Roche Products, Hertforshire,
UK). Protein concentrations were determined by the Bio-Rad
(Hercules, CA) DC protein assay.
For SDS-PAGE, samples containing 50 µg of protein were supplemented
with equal volumes of protein sample buffer [62.5 mM
Tris-HCl, pH 6.8, 2% SDS, 20 mM DTT, 10% glycerol, and
0.003% (w/v) Pyronin-Y], samples were incubated at 50°C for 15 min,
and urea was added to samples to a final concentration of 4 M. SDS-PAGE on 6% gels was followed by electrotransfer
onto polyvinylidene difluoride membrane (Hybond-P; Amersham
Biosciences). Immunodetection was performed as described
previously (Specht and Schoepfer, 2001), with primary antibodies used
at the appropriate dilution: rabbit (rb) anti-NR1, 1:500
(G8913; Sigma, St. Louis, MO); rb
anti-NR2A, 1:750 (AB1555P; Chemicon, Temecula, CA); rb
anti-NR2B, 1:1000 (AB1557P; Chemicon); rb anti-NR2C, 1:100
[serum K21450, raised against amino acids (aa) 434-447; M. Herkert,
University of Erlangen, Erlangen, Germany]; rb anti-NR2D, 1:1000
(serum K23, aa 1046-1062; M. Herkert); rb anti-ionotropic glutamate
receptor subunit 1 (GluR1), 1:1000 (AB1504;
Chemicon), rb anti-GluR2/3, 1:5000 (AB1506;
Chemicon); rb anti-GluR4, 1:100 (AB1508;
Chemicon); mouse anti-metabotropic GluR1
(mGluR1),
1:500 (number 556331; BD Phar-Mingen); rb
anti-GABAAR-
3, 1:5000
(Brandon et al., 2000); and guinea pig
anti-GABAAR-g2, 1:200
(Brandon et al., 2001).
Specific immunoreactivity was detected using enhanced chemiluminescence
(ECL Plus; Amersham Biosciences).
Tissue culture and electrophysiology. Hippocampal
organotypic slice cultures were prepared from embryos at stage
E18.5-E19.5 as described previously (Stoppini et al., 1991). Briefly,
embryos were decapitated, and the brain was removed from the skull and placed into ice-cold dissecting medium (MEM 22370-027, including 100U/ml penicillin and 100 µg/ml streptomycin;
Invitrogen, Gaithersburg, MD). Hippocampi were
dissected out, and tissue slices were taken at 350 µm using a
conventional tissue chopper (McIlwain). Slices were placed on Millicell
(Millipore, Bedford, MA) 35 mm tissue culture inserts with
membranes (0.4 µm pore diameter) supplied with prewarmed culture
medium [two parts MEM and one part HBSS (24020-091;
Invitrogen), one part heat-inactivated horse serum (26050-070; Invitrogen), 100 U/ml penicillin, and 100 µg/ml streptomycin]. Cultures were then held in a moistened
atmosphere containing 5% CO2 at 37°C for up to
3 weeks.
After at least 7 d in culture, slices were placed in a recording
chamber perfused with artificial CSF (aCSF) [in mM:
126 NaCl, 2.5 KCl, 1 CaCl2, 1.2 NaH2PO4, 26 NaHCO3, and 20 glucose, pH 7.4 (saturated with
95% O2-5% CO2)]
containing 200 nM TTX, 10 µM bicuculline methiodide, 1 µM strychnine, and 5 µM DNQX.
Whole-cell patch recordings were obtained with glass pipettes filled
with recording solution [in mM: 130 CsCl, 2.5 NaCl, 5 MgCl2, 10 HEPES, and 10 EGTA, pH 7.2 with CsOH
(4-6 M
resistance)] using an Axopatch 2B or 200A amplifier. Bath
application of aCSF containing either [20 µM NMDA] or
[20 µM NMDA plus 500 or 100 µM
Mg2+] or, as a control, [20
µM NMDA, 500 or 100 µM
Mg2+, and 20 µM APV], plus
10 µM glycine at all times, was accomplished using an
ALA Scientific Instruments (Westbury, NY) and NPI
Electronics (Tamm, Germany) Multivalve system. Cells were held at 
60
mV (series resistance was compensated at 95%), and I-V
curves before, during, and after application of the respective
combination of drugs were established by a ramp protocol covering the
range from 100 to 40 mV. Data acquisition and control of the
experimental protocol was done using the National
Instruments (Austin, TX) PCI-MIO-16XE-10 board and an INT-20
interface and CellWorks software (NPI Electronics). For data
evaluation, IGOR (WaveMetrics, Lake Oswego, OR) software was used.
Histology. Pups were injected intraperitoneally with an
overdose of sodium pentobarbitone, and death was confirmed by absence of paw-withdrawal reflex. Next, mice were transcardially perfused with
cold 0.1 M phosphate buffer (PB), pH 7.4, followed by 4% paraformaldehyde in 1× PBS. P0 pups were postfixed
overnight in toto.
For Nissl staining, fixed brains were dissected out and equilibrated in
30% sucrose-0.1 M PB, pH 7.4. Transverse-coronal
sections were cut with a cryostat (10-15 µm) or a freezing microtome
(50 µm), respectively, and mounted on gelatinized slides. Mounted sections were dried overnight, soaked in tap water for 10 min, stained
in 0.05% thionin solution (BDH Laboratory
Supplies, Poole, UK), washed in tap water, differentiated in
95% ethanol, dehydrated in an ethanol series (70, 90, and 100%),
cleared in Histoclear (National Diagnostics, Atlanta, GA),
and coverslipped in DPX (BDH Laboratory Supplies).
CO staining was performed as described previously (Wong-Riley, 1979).
Pups were prepared as described above and sucrose equilibrated in
toto. Serial 50 µm transverse brainstem sections were cut from whole heads using a freezing microtome and collected in 0.1 M PB, pH 7.4. Sections were transferred into CO
staining solution [0.1 M PB, pH 7.4, 4%
sucrose, 0.4% cytochrome c (C-7752; Sigma), and 0.5% diaminobenzidine] and incubated for 2-4 hr in a 37°C incubator. Reactions were stopped by transferring sections into 0.1 M PB, pH 7.4. Sections were rinsed in 0.01 M PB, pH 7.4, mounted on gelatinized slides, and
processed as above without Nissl staining. Each CO staining experiment
included at least one wild-type positive control from the same litter.
For Tenascin-C (TN-C) immunostaining, affinity-purified IgG fraction of
a polyclonal rabbit antiserum raised against complete Tenascin-C
(Faissner and Kruse, 1990) was used at 1:15000 and visualized with
biotinylated goat anti-rabbit antibody at 1:600 (BA-1000; Vector
Laboratories, Burlingame, CA), followed by the Vectastain Elite
ABC kit (1:600).
DiI labeling. For axonal labeling of primary trigeminal
afferents (Erzurumlu and Jhaveri, 1992), small crystals of DiI
(Molecular Probes, Eugene, OR) were applied to single
whiskers under visual control, i.e., to follicle B1 on one side and B2
on the opposite side of the face of fixed pups. Pups were kept in
fixative at 37°C for at least 3 months for the dye to diffuse. Brains
were carefully dissected out and embedded in 40°C equilibrated 2%
low-melting-point agarose (Flowgen). Vibratome-cut 100 µm transverse
brainstem sections were mounted onto polysine microscrope slides
(BDH Laboratory Supplies) and coverslipped with 40°C 0.6%
low-melting-point agarose in 1× PBS. Fluorescent staining was studied
using a Zeiss (Oberkochen, Germany) Axiophot
Photomicroscope equipped with a rhodamine filter set, followed by
confocal microscopy using a Leica (Nussloch, Germany)
upright microscope fitted with a Leica TCS SP scan head. DiI-labeled axonal projections in the trigeminal nucleus were visualized by confocal laser scanning microscopy (568 nm excitation) using a 25×, 0.75 numerical aperture oil immersion objective, 2×
electronic zoom, 4× scan accumulation, and 0.2 µm distance between
optical sections. Z-projections (standard deviation method) of
image stacks were produced using ImageJ software
(http://rsb.info.nih. gov/ij/). The location (l) of
a DiI patch in the subnucleus caudalis (nVc) is
l = a/b, where a is
the horizontal distance from center of patch to lateral margin of
section, and b is the horizontal distance from midline of
section to lateral margin of section.
| |
Results |
Targeted mutation of the NR1 subunit
Through homologous recombination in ES cells, we altered the DNA
sequence coding for amino acid 598 of the mature NR1 subunit. The
wild-type codon for the 598 asparagine residue in exon 15 was replaced
by a codon for arginine. At the same time, a floxed (flanked
by loxP sites in the same orientation) neo selection cassette was introduced into intron 18, yielding the
NR1Rneo allele (Fig.
1).
View larger version (40K):
[in this window]
[in a new window]
|
Figure 1.
Generation of the NR1 N598R alleles.
A, Partial organization of the murine NR1 locus.
Wild-type locus (NR1+), the targeted locus after
homologous recombination (NR1Rneo), and the targeted
locus after Cre-mediated excision of the neo cassette
(NR1R). Exons are depicted as boxes.
Coding regions are in black, and the 3'UTR are in
gray. Exon 15 harbors the N598R mutation after
targeting. loxP elements are shown as filled
triangles, and the neo selection marker is shown as an
open box. The 5' and 3' limits of the targeting
construct are indicated (× symbols) on the
NR1Rneo and NR1R alleles.
Relevant EcoRV (E) restriction
sites and DNA fragments, as well as localization of hybridizing probe
B, are shown. PCR primers 1-4 (see Materials and Methods) for routine
genotyping are depicted as horizontal arrows.
B, Southern blot of EcoRV-digested
genomic mouse tail DNA. Lanes 1-3 show offspring from
mating I (NR1+/Rneo × NR1+/Rneo), lanes 4-7 show offspring
from mating II
(NR1+//Cre+/+ × NR1+/Rneo), and lanes 8 and
9 show offspring of mating III
(NR1+/ × NR1+/).
Top, The membrane was hybridized with probe B to test
for the different mutated NR1 alleles: wild-type (+ symbols; 5 kb); targeted
(Rneo; 6 kb); and targeted after
Cre-mediated resolution of the neo cassette (R; 4.6 kb).
Bottom, The membrane was stripped and hybridized with
probe A to test for the null allele: wild-type (+ symbols; 6.9 kb) and null ( symbols;
9.6 kb) (see Materials and Methods).
|
|
Chimeric mice of wild-type blastocysts and ES cells carrying the
NR1Rneo allele produced heterozygous
offspring of the NR1Rneo/+ genotype. These
NR1Rneo/+ mice were viable and fertile and
were backcrossed with C57BL/6J inbred mice for additional experiments.
Homozygous animals of the NR1Rneo/Rneo
genotype died within 12 hr after birth. The presence of the neo
cassette in exon 18 was found to disrupt the expression of the
NR1Rneo allele (data not shown). Thus,
NR1Rneo/Rneo mice should have the same
phenotype as homozygous NR1 knock-out animals and were therefore not
studied further.
Generation of NR1 mutant genotypes
Cre-mediated excision of the floxed neo cassette in
intron 18 of the NR1Rneo allele converts
the inactive neo-containing allele into an active NR1 allele that
carries the N598R mutation (NR1R allele)
(Fig. 1). Mating of NR1Rneo/+ mice with
Deleter mice (Schwenk et al., 1995) results in mice of the
NR1R/+ genotype. These mice express comparable amounts of
NR1+ and NR1R mRNA, as shown
by M13 plaque assays (NR1+ plaques,
52.3 ± 1.3%; NR1R plaques,
47.7 ± 1.3%; n = 3; means ± SD).
Because of the lethal phenotype of NR1R/+
(see below), it is not possible to generate
NR1R/R mice. To obtain mice that express
only NR1R mRNA (and no wild-type NR1
mRNA), NR1Rneo/+ mice were mated with Cre
mice (Cre+/+,
NR1+/) that carried in addition an NR1
null allele, NR1. The breeding scheme
results in a total of four different NR1 genotypes
(NR1+/+,
NR1R/+,
NR1R/, and
NR1+/).
NR1/ mice were generated by mating
NR1+/ mice.
The N598R mutation is dominant lethal
NR1R/+ mice were found to die within
6 hr from birth. They were in respiratory distress and displayed motor
hyperactivity to the extent that they did not remain in a seated
position but rolled over when placed on their limbs.
NR1R/ mice were found to have an even
more severe phenotype than NR1R/+ and
NR1/ mice; they hardly moved or
breathed, became cyanotic shortly after birth, and died within 1 hr.
Thus, both NR1R/+ and
NR1R/ genotypes differ phenotypically
from NR1+/ and
NR1/ genotypes.
NR1+/ mice appeared normal, whereas
NR1/ mice passed from a phenotypically
normal to a morbid phase and died within 12-20 hr as a result of
respiratory failure (Forrest et al., 1994; Li et al., 1994; Poon et
al., 2000).
Mice with NR1R/+,
NR1R/,
NR1+/, and
NR1/ genotypes had normal body weights
at birth compared with NR1+/+ mice.
Newborn pups of all five genotypes weighed between 1.2 and 1.8 gm,
varying with litter size rather than with genotype.
To investigate whether NR1R/+ pups were at
a selective disadvantage, phenotypically wild-type littermates were
removed immediately after birth. Although the mother did not neglect
pups isolated in this way, no milk uptake was observed in
NR1R/+ pups, and they still died at P0.
Assuming a tetrameric stoichiometry with two copies of the NR1 subunit
per receptor complex (Behe et al., 1995), mice of
NR1R/+ genotype should express 25% pure
wild-type receptors, in addition to 25% pure mutant receptors and 50%
mixed receptors
(NR1+/NR1R).
Because a reduced level of wild-type NMDAR expression to ~5% is
compatible with life (Mohn et al., 1999), the reduction of wild-type
receptor levels to 25% cannot be the cause of death in
NR1R/+ mice. Thus, the
NR1R allele represents a gain-of-function mutation.
The NR1R allele is expressed at
wild-type levels
To investigate whether the NR1R
allele is expressed at wild-type level in our mutant mice, we analyzed
expression of NR1 mRNA and protein (Fig.
2).
View larger version (64K):
[in this window]
[in a new window]
|
Figure 2.
Analysis of NR1 expression in brains from newborn
mice. A, Northern blots. Three hundred nanograms of
poly(A+) RNA from different genotypes were probed
with a cDNA fragment covering NR1 exon 15-17 and reprobed for actin as
a control. Lanes 1-4 show NR1+/+,
NR1R/+, NR1R/, and
NR1+/ littermates from mating II, and lanes
5 and 6 show NR1/ and
NR1+/+ littermates from mating III.
B, Western blots. Immunoblot analysis of 50 µg of
membrane protein from littermates as in A with an
antibody binding to the C terminus of NR1. The membrane was reprobed
for tubulin. Even prolonged autoradiographic exposures did not detect
specific NR1 signals in NR1/ samples.
|
|
Northern blot analysis of poly(A+) RNA
revealed comparable levels of NR1 mRNA in
NR1R/ and
NR1+/ mice. This indicates that neither
the altered coding sequence nor the remaining loxP site in
intron 18 have any obvious effect on NR1 transcription and RNA
stability. Furthermore, no aberrant splice variants were detected (Fig.
2A).
Western blot analysis of membrane fractions showed comparable protein
levels for wild-type and mutant subunits. This indicates that the
presence of the N598R mutation does not interfere with the regulation
of NR1 protein levels in vivo (Fig.
2B).
Analysis of expression of glutamate and GABAA
receptor subunits
Because glutamatergic and GABAergic neurotransmission are
influencing each other during development of neural networks (Ben-Ari, 2001), we investigated whether the presence of the NR1 N598R mutation alters the protein levels of other relevant neurotransmitter receptors. To this end, we performed Western blot analysis on brain membrane fractions of all five genotypes probing for NMDAR NR2 subunits, major
AMPA receptor and GABAA receptor subunits, as
well as the NMDAR-associated G-protein-coupled mGluR1 receptor (Fig.
3) (Husi et al., 2000). For all 10 glutamate and GABAA receptor proteins that we
analyzed, no obvious differences were detected in mice expressing N598R
mutant NR1 subunits when compared with mice expressing wild-type
NMDARs. In particular, the presence of the
NR1R subunits does not lead to reduced
NR2B protein levels, contrasting with
NR1/ knock-out animals, which have
downregulated levels of NR2B (Fig. 3A) (Forrest et
al., 1994). This indicates that NR1R
subunits are comparable with NR1 wild-type subunits with respect to
their role in regulating NR2B expression. In summary, Figure 3
indicates that the presence of the mutant NR1 subunit does not have
obvious effects on expression of fast neurotransmitter receptors.
View larger version (57K):
[in this window]
[in a new window]
|
Figure 3.
Western blot analysis of expression of glutamate
and GABAA receptor subunits. Membrane protein fractions (50 µg) from whole brains of newborn and adult mice were probed with
antibodies to specific receptor subunits. Genotypes of newborn mice
(P0) are from matings II and III. Adult wild-type (+/+)
is shown as control (ND, not done). A,
NMDA receptor subunits. Protein levels of NR2B were found not to be
altered in mice expressing the N598R mutant NR1 subunit, whereas
absence of the NR1 subunit lead to decreased expression levels of NR2B.
NR2A and NR2C were both found to be low in newborn mice of all
genotypes compared with higher protein levels in adult mice, suggesting
that the NR1 N598R mutation does not affect the protein levels of NR2A
and NR2C. NR2D protein levels were at comparable levels in newborn mice
of all genotypes. B, AMPA receptor subunits.
C, Metabotropic glutamate receptor. D,
GABAA receptor subunits. Expression levels of these
subunits were at comparable levels in all NR1 genotypes, suggesting
that neither the absence of NR1 nor the N598R point mutation interfere
with their expression.
|
|
NR1R mutant NMDARs are functional but lack
coincidence detection properties
We next investigated whether the mutant
NR1R subunits form functional NMDARs. Bath
application of NMDA resulted in inward currents into neurons from
NR1R/+ and
NR1R/ animals (Table
1). These whole-cell currents were
similar in size to those recorded from wild-type tissue
(NR1+/+ and
NR1+/). APV blocked NMDA-induced
currents in all genotypes. No NMDA-induced currents were detected in
mice lacking NMDARs (NR1/) (Table
1).
NMDA-induced currents in wild-type neurons
(NR1+/+ and
NR1+/) showed the expected
I-V relationship and were sensitive to
Mg2+ ions in the expected
voltage-dependent manner (Fig. 4, Table 1). In contrast, NMDA-induced responses from neurons, which express only the mutant NR1R subunit, were
insensitive to Mg2+ ions, even at a higher
concentration (500 µM). Currents from neurons
that express both NR1R and
NR1+ subunits show an intermediate
sensitivity to block by Mg2+, consistent
with a heterogeneous receptor population containing pure mutant, mixed,
and wild-type receptors.
View larger version (9K):
[in this window]
[in a new window]
|
Figure 4.
Current-voltage relationship of NMDA-induced
currents. NMDA-induced macroscopic currents in N598R mutant cells lack
rectification compared with wild type. Current-voltage relationships
were determined after application of 20 µM NMDA using
whole-cell patch recordings from CA1 pyramidal-shaped cells in
organotypic hippocampal slice cultures. Representative
I-V curves of genotypes NR1+/+,
NR1R/+, and NR1R/ are shown in
the presence of 100 µM Mg2+
(NR1+/+) or 500 µM
Mg2+ (NR1R/+,
NR1R/) (black traces), without
added Mg2+ (light gray traces), and
in the presence of 20 µM APV (dark gray
traces). Currents were normalized to the current at 30 mV. A
decreased rectification in NR1R/+ and a complete
lack in NR1R/ are clearly visible, even at an
increased Mg2+ concentration of 500 µM. For a quantitative evaluation, see Table 1.
|
|
So far, we showed that NR1R/ animals
express functional NMDARs that lack the coincidence detection
capability conferred by a voltage-dependent
Mg2+ block. In contrast,
NR1/ mice do not express NMDARs at all.
Neuroanatomy in N598R mutant mice
To investigate whether impairment of coincidence detection
properties in NR1R/ mice causes
neuroanatomical deficits, we analyzed brain morphology of various brain
regions of newborn mice.
No differences in external brain morphology and weight between
NR1+/ and
NR1R/ mice were detected. The early
cyanosis and death observed in NR1R/
mice prompted us to investigate the brainstem containing the regulatory
centers for respiration, cardiac function, and swallowing. Cytological
(Nissl) and histochemical (CO) staining methods did not reveal gross
neuroanatomical differences between
NR1+/ and
NR1R/ mice in brainstem structures (Fig.
5E-H). In the
cerebellum, no differences were seen with either method. The developing
cerebellar granule cell layers were present in both genotypes (Fig.
5A,B, and data not shown).
Likewise, in the hippocampus, no significant differences were seen. The
layers of the hippocampal cortex were developed equally, and the
dentate gyrus was emerging in both genotypes (Fig.
5C,D).
View larger version (103K):
[in this window]
[in a new window]
|
Figure 5.
Neuroanatomy of NR1+/ and
NR1R/ littermates. Nissl-stained sections of
cerebellum (A, B), hippocampus
(C, D), and brainstem (E,
F), as well as CO-stained sections of brainstem
(G, H) of newborn
NR1+/ (A, C,
E, G) and NR1R/
(B, D, F,
H) littermates are shown. Both Nissl and CO
staining showed unchanged gross CNS structure in
NR1R/ mice compared with control
NR1+/ mice. However, whisker-related patterns were
present in NR1+/ mice (G)
and absent in NR1R/ mice
(H). Scale bar: A-D, 250 µm; E-H, 500 µm. Section thickness:
A-D, 15 µm; E-H, 50 µm.
amb, Nucleus ambiguus; CA, cornu ammonis;
dg, dentate gyrus; hyp, hypoglossal
nucleus; nVi, subnucleus interpolaris of trigeminal
nucleus; io, inferior olive; D, dorsal;
L, lateral.
|
|
However, within the BSTC, CO staining failed to reveal whisker-related
patterns in NR1R/ animals that is found
in wild-type mice (Fig. 5G,H) (see below).
Absence of whisker-related patterns in the brainstem of NR1
mutant mice
NMDAR-mediated coincidence detection and
Ca2+ signaling have been implicated in
activity-dependent refinement of neuronal connections in developing
somatotopic maps (Crair, 1999). Therefore, we analyzed the formation of
whisker-related patterns (barrelettes) in the BSTC in our mouse model
in detail. Mice with the genetic background we used were born between
E18.5 and E20.5 and displayed the full array of large mystacial whiskers.
We found that, in the BSTC of wild-type animals, histochemical (CO
staining) but not cytological (Nissl staining) barrelettes can be
discerned at birth (Fig. 5E,G). CO
staining reflects the activity of mitochondrial cytochrome oxidase
mitochondria, which is particularly high in dendritic arborizations and
somata (Wong-Riley and Welt, 1980). The degree of whisker-related
patterning was furthest advanced in the subnucleus interpolaris (nVi)
(detected in all samples; n = 14) and the least obvious
in the principal nucleus of V (nVp) (one of five animals) (Fig.
6A-C). Identical findings have been reported by Ma (1993). No significant difference was
seen between NR1+/+ animals and
NR1+/ animals.
View larger version (151K):
[in this window]
[in a new window]
|
Figure 6.
Cytochrome oxidase-stained brainstem
sections in newborn NR1 wild-type and mutant mice. A-L,
CO-stained transversal sections of the BSTC are shown at the
level of the nVc (A, D, G,
J), nVi (B, E,
H, K), and nVp (C,
F, I, L) of
NR1+/ (A-C),
NR1/ (D-F),
and NR1R/ (G-L) animals.
In animals expressing only the wild-type NR1 subunit, whisker-related
rows (a-e) segregating into individual barrelettes were
consistently found in nVi (B). Emerging
whisker-related patterns were usually present in nVc
(A), whereas patterning in nVp
(C) was found only in a small number of animals.
In contrast, whisker-related patterns were never found in animals
exclusively expressing the N598R mutant NR1 subunit
(G-L) or lacking the NR1 subunit
(D-F). Examples from two
NR1R/ animals are shown in G-I and
J-L. MV, Motor nucleus of V;
amb, nucleus ambiguus. Scale bar, 200 µm.
|
|
NR1R/ mice express functional NMDARs
that lack coincidence detection properties as shown above. However, in
newborn pups of this genotype, no whisker-related patterning was
noticeable (n = 12) (Fig. 6G-L). In fact,
throughout the BSTC, CO-stained brainstem sections of
NR1R/ mice looked similar to those of
NR1/ mice (n = 5)
(Fig. 6D-F), which are known to lack
whisker-related patterning (Li et al., 1994). It appears that
NR1R/ mice show a somewhat uneven,
granular, but whisker-unrelated, CO staining in the BSTC, in contrast
to its amorphous appearance in NR1/ mice.
Blind scoring found whisker-related patterning in nVi in mice
expressing both NR1+ and
NR1R subunits
(NR1R/+, n = 8) in
approximately one-half of the animals, in accordance with Single et al.
(2000).
To confirm the results obtained by CO staining, a marker for
delineating barrelette boundaries was used, similar to Li et al.
(1994). Brainstem sections from wild-type
NR1+/ and mutant
NR1R/ animals were stained
immunohistochemically for the extracellular matrix protein TN-C
(Faissner and Steindler, 1995). A pattern complementary to the one seen
with CO staining is revealed in NR1 wild-type mice
(NR1+/) (Fig.
7A-C). No whisker-related
pattern was found in animals lacking NMDARs
(NR1/), confirming the usefulness of
this stain (Fig. 7D-F). Equally, no whisker-related
pattern is seen in NR1R/ animals,
confirming the results obtained by CO staining for
NR1R/ mice.
View larger version (153K):
[in this window]
[in a new window]
|
Figure 7.
Barrelette detection by immunohistochemistry for
Tenascin-C. Transversal section through the BSTC of newborn mice were
immunostained with an antibody against the extracellular matrix protein
Tenascin-C. The immunopositive pattern reflects the boundaries between
barrelettes in control NR1+/ animals
(A-C). The emerging whisker-related
pattern is most distinct in nVi (B) in which
pronounced horizontal and finer vertical boundaries can be seen.
Arrowheads denote distinct boundaries delineating
whisker-specific rows. No emergent pattern within the BSTC can be
discerned in NR1R/ mice
(G-I) or in NR1/
mice (D-F). The contours of the
individual BSTC nuclei remain visible, and the sizes of the individual
nuclei appear comparable in all three genotypes. Scale bar, 200 µm.
|
|
Together, these results indicate that NMDARs in
NR1R/ mice fail to mediate
the postsynaptic signaling that is required for maturation of the
whisker-related somatosensory system.
Normal spatial order and axonal arborization of trigeminal ganglion
cell processes in N598R mutant mice
To ascertain that the absence of barrelettes in
NR1R/ mice was not caused by
aberrant projection of primary afferents within the
BSTC, we used DiI labeling of trigeminal ganglion cells (Erzurumlu and
Jhaveri, 1992) on mice that exclusively expressed either wild-type NR1
(NR1+/+,
NR1+/) or N598R mutant NR1 subunit
(NR1R/).
For the analysis of projections of primary afferents, small amounts of
DiI crystals were applied to a single whisker on each side of the face.
As in wild-type mice, NR1R/ mutants
displayed the typical arrangement of large mystacial whiskers in five rows.
In series of transverse sections of all genotypes, the trigeminal tract
was labeled, and the arborizations of axon collaterals projecting into
the trigeminal nuclei revealed nVc, nVi, and nVp as areas separated by
poorly stained zones (Fig.
8A,B,E-H).
Analysis of the spatial order of single whisker-related DiI patches
within nVc revealed correct correlation with the labeled whisker (Fig. 8A,B). Furthermore, in both
genotypes, the axonal arborization (Fig. 8C,D)
appeared to be similar, suggesting that
NR1R mutant NMDARs do not disturb the
crude topographic mapping of primary trigeminal afferents in the
BSTC.
View larger version (59K):
[in this window]
[in a new window]
|
Figure 8.
Projection and arborization of single
whisker-related trigeminal afferents. Low-magnification, light
microscopy (A, B,
E-H) images of coronal sections of newborn
NR1+/+ (A, E,
G) and NR1R/ (B,
F, H) mice through the brainstem
at the level of the nVc (A, B), nVi
(E, F), and nVp (G,
H) after DiI application to a single whisker, B1
or B2. Patches of DiI-labeled axonal arborization were centered
around the expected topographic location in the trigeminal nuclei. The
location (l) of the single
whisker-related patch in nVc for the B1 whisker was 0.160 (NR1+/+, n = 2) and 0.161 (NR1R/, n = 2) and for the B2
whisker was 0.239 (NR1+/+ or
NR1+/, n = 8) and 0.225 (NR1R/, n = 6). Z-projections
of stacks containing a series of 66 high-magnification confocal images
(C, D, insets of
A and B, respectively) gave comparable
overall impressions of axonal arborizations in both genotypes. Scale
bar: A, B, E-H, 100 µm;
C, D, 20 µm. Arrows mark
the trigeminal tract (A, B).
Arrowheads indicate axon collaterals branching off the
trigeminal tract (C, D). White
line depicts section margin (G,
H). D, Dorsal; L,
lateral.
|
|
| |
Discussion |
The objective of this study was to evaluate the relevance of
coincidence detection and Ca2+ signaling
by NMDARs for establishing patterned neuronal connectivity. Previous
structure-function studies on recombinant NMDARs have suggested that
the introduction of the N598R mutation into the NR1 gene would result
in a suitable animal model for this purpose (Burnashev et al., 1992;
Katz, 1994).
Lethality of the NR1 N598R mutation
The lethal phenotype of the NR1 N598R mutation has
been described previously (Single et al., 2000). However, in contrast
to the previous work, we found that NR1R/+
pups generated in our breeding population are unable to survive for
more than a few hours after birth. The reason for the somewhat different phenotype between the two animal populations that carry the
same mutation is unknown, although subtle differences in genetic background might be relevant.
The mutant NMDARs are functional but lack coincidence detection
capability
Whole-cell NMDA-induced currents, measured in organotypic
hippocampal cultures from NR1R/ animals
expressing exclusively the mutant NR1 subunits, were comparable in
magnitude with those from wild-type animals (Table 1). Single et al.
(2000) reported previously that they could not detect agonist-induced
currents in enucleated patches from acute hippocampal slices from
NR1R/ animals. The agonist-induced
currents measured by us in the whole-cell configuration were
insensitive to Mg2+, even at high
concentrations, as expected from previous work on recombinant receptors
containing the NR1 N598R subunit (Burnashev et al., 1992; Schoepfer et
al., 1994). Currents from NR1R/+ animals
were partly sensitive to Mg2+, consistent
with a mixture of wild-type plus mutant receptors and/or receptors
containing both wild-type and mutant NR1 subunits (Behe et al., 1995).
Recombinant NR1R/NR2 receptor channels are
impermeable for Ca2+ (Burnashev et al.,
1992), and the Ca2+ reversal potential in
NR1R/+ animals is shifted as predicted
(Single et al., 2000). In summary, we conclude that
NR1R/ animals express NMDARs with
impaired coincidence-dependent Ca2+
signaling at the cell surface.
Absence of gross anatomical alterations in the brains of
NR1R/ animals
NR1R mutant mice differ from
NR1/ knock-out mice in their motor and
respiratory behavior during the postnatal hours (see Results) (Single et al., 2000). However, like
NR1/ knock-out mice (Forrest et al.,
1994; Li et al., 1994), NR1R mutant mice
did not appear to be developmentally retarded. Indeed, we could not
detect any obvious morphological or neuroanatomical abnormalities in
NR1R mutant mice other than in the
brainstem trigeminal nuclei (see below).
Previously, it has been speculated that NMDAR-mediated
Ca2+ influx is relevant for migration of
cerebellar granule cells (Komuro and Rakic, 1993, 1998). Migration of
neocortical neurons has been analyzed in NR1 knock-out mice and was
shown not to be affected by the lack of NMDARs (Messersmith et al.,
1997). Although we have not performed a detailed analysis of the
migration of neurons in NR1R mutant mice,
it seems likely from our data that the NR1 N598R mutation alters NMDAR
function without interfering with neuronal migration. A region-specific
activation of the NR1Rneo allele in our
mouse model could elucidate this problem.
Exclusive expression of NR1R impairs
whisker-related patterning
NR1R/ mice did not show the normal
patterning into whisker-related rows or barrelettes in the brainstem
trigeminal nuclei of newborn mice, when assessed by two independent
techniques. Neither histochemical barrelettes, as revealed by CO
staining, nor barrelette boundaries, as revealed by TN-C
immunohistochemistry, could be detected in these mice. CO
histochemistry revealed slight differences in the BSTC between
NR1/ and
NR1R/ mice, whereas TN-C
immunohistochemistry showed equally amorphous nuclei. The
NR1//NR1R/
difference may be explained by the fact that
NR1R/ mice should have the NMDAR
signaling complex in place, which is absent in
NR1/ mice. Thus,
NR1R/ mice would have the molecular
machinery for barrelette formation; however, their signaling pathways
are not stimulated in the coordinated manner that is ultimately
required for whisker-related pattern formation.
It is still conceivable that the lack of patterning of the postsynaptic
cells in the trigeminal nuclei of NR1R/
mice is secondary to altered input into this nucleus. However, this is
unlikely. First, both methods revealed normal-sized trigeminal nuclei,
either as positively stained (CO) or negatively stained (TN-C) areas.
Second, DiI labeling revealed topologically correct targeting of
primary afferents in NR1R/ mutant mice,
similar to the findings in NR1/
knock-out mice (Li et al., 1994). Third, the absence of NR2 messages in
trigeminal ganglion cells implies that no functional NMDARs are
expressed in the primary afferents (Watanabe et al., 1994); therefore,
an unaltered input activity pattern is expected. Fourth, the
establishment of active synaptic connections between the
whisker-related primary afferents and brainstem neurons is not
dependent on the presence of NMDARs (Li et al., 1994). Together, this
is evidence that the lack of barrelette formation in
NR1R/ mutants is not attributable to
altered input.
Lack of NMDARs in the forebrain impairs whisker-related patterning in
the barrel cortex (Iwasato et al., 2000), whereas expression of mutant
NR1R subunits in addition to wild-type NR1
subunits (NR1R/+) does not (Single et al.,
2000). We found comparable results in the BSTC.
NR1R/+ pups showed whisker-related
patterns at a somewhat reduced prevalence, i.e., in the nVi of
approximately one-half of the pups. These results show that the
theoretically expected (Behe et al., 1995) fourfold reduced number of
purely wild-type receptors in NR1R/+ mice
is sufficient for whisker-related pattern formation, in accordance with
Iwasato et al. (1997).
Absence of whisker-related patterning in the BSTC has been described
previously in animals lacking either all NMDARs (Li et al., 1994) or
the NR2B-containing subpopulation (Kutsuwada et al., 1996). Because
normal patterning is observed in animals lacking the NR2D subunit
(Ikeda et al., 1995), these data indicate that the NR2B subunit is
critical for whisker-related patterning and cannot be compensated for
by the NR2D subunit. We found that protein levels of the NR1 and all of
the NR2 subunits (including NR2B) were unchanged in
NR1R mutant animals, whereas NR2B subunit
levels were downregulated in NR1 knock-out mice (Forrest et al., 1994).
Thus, we can rule out the possibility that the observed lack of
barrelettes in NR1R/ mice was
attributable to secondary effects on NR2B function, such as protein
level and synaptic localization. This is important because direct
interaction between the C termini of NR2 subunits and MAGUK
(membrane-associated guanylate kinase) proteins [via PDZ (postsynaptic
density-95/Discs large/zona occludens-1), SH3 (Src homology 3), and
guanylyl kinase domains] is assumed to play a major role in
anchoring NMDARs in the postsynaptic density (for review, see Scannevin
and Huganir, 2000; Sheng and Pak, 2000). For example, genetically
engineered mice carrying a deletion of the cytoplasmic C terminus of
the NMDAR subunit NR2B express functional NMDARs at normal levels, but
the amount of the truncated subunit at synapses was decreased and
whisker-related patterns fail to form in the BSTC (Mori et al.,
1998).
Consequences of impaired NMDAR-mediated coincidence detection and
Ca2+ influx
The NR1 N598R mutation results in
Ca2+-impermeable NMDARs (Burnashev et al.,
1992) and thus abolishes the primary intracellular signal that normally
follows coincidence detection (Bliss and Collingridge, 1993). NMDARs
are concentrated in the postsynaptic membrane of excitatory synapses,
in which they associate via the C termini of their subunits with a
multiprotein scaffolding and signaling complex (Husi et al., 2000).
Many interactions of NMDAR subunits with proteins in this complex seem
to be independent of Ca2+ influx through
the NMDAR channel, e.g., interaction with intermediate filaments, adapter protein Yotiao, MAGUK proteins. However,
some protein interactions as well as some signaling events are
regulated by Ca2+ influx through the NMDAR
channel, e.g., -actinin-2 binding or activation and translocation of
Ca2+/calmodulin kinase II (for
review, see Kennedy, 2000; Scannevin and Huganir, 2000; Sheng and Pak,
2000; Bayer and Schulman, 2001). The exact molecular mechanisms
that ultimately lead to the absence of whisker-related patterning that
we observed remain to be elucidated.
We studied the expression levels of the subunits of the major
neurotransmitter receptors and we did not detect any obvious alteration
of their protein levels. RNA expression profiling using cDNA array
technology also failed to reveal obvious differences. The RNA
expression profile of NR1R/+ mice, labeled
transgenic model by Specht and Schoepfer (2001), their
Figure 1, is virtually identical to NR1+/+
"wild-type littermate" controls. However, the expression of mutant NMDARs leads to altered map formation. From that, we conclude that the
signaling cascades downstream of coincidence detection and
Ca2+ influx act in a highly specific
manner without major reprogramming of the neural gene expression
pattern. Therefore, most of these signals must be local in nature. This
concept of localized signaling cascades fits well with the recent
description of spatially well defined signaling complexes (Husi et al.,
2000). Highly localized signaling could be an underlying principle of
neuronal information processing.
In summary, we showed that NMDARs with impaired coincidence-dependent
calcium signaling disturb pattern formation in the brainstem of newborn
mice. Region-specific activation of the NR1 N598R allele will provide
more detailed insights into the role of NMDAR-mediated signaling in
more mature animals and/or regions other than the BSTC, e.g., the
primary somatosensory cortex.
| |
FOOTNOTES |
Received Sept. 6, 2002; revised Dec. 10, 2002; accepted Dec. 16, 2002.
*
Y.R. and M.K. contributed equally to this work.
This work was supported by a Wellcome Trust Senior Fellowship (R.S.).
Y.R. and M.A.N. were supported by a Wellcome prize studentship. C.M.T.
was supported by a Wellcome Traveling fellowship. G.F.R. is recipient
of an Emmy Noether fellowship (Deutsche
Forschungsgemeinschaft). We thank K. Rajewsky for DNA plasmids
(floxed neo and HSV tk) and E14 ES cells, F. Schwenk and K. Rajewsky
for the generous gift of Cre Deleter mice, and C. M. Becker, A. Faissner, M. Herkert, and S. J. Moss for antibodies.
Correspondence should be addressed to Ralf Schoepfer, Laboratory for
Molecular Pharmacology, Department of Pharmacology, University College
London, Gower Street, London WC1E 6BT, UK. E-mail:
r.schoepfer{at}ucl.ac.uk.
A. J. Annala's present address: Pharmacology Department,
University of California, Los Angeles School of Medicine, Los Angeles, CA 90095.
P. E. Chen's present address: Department of Neuroscience,
University of Edinburgh, Edinburgh EH8 9JZ, UK.
I. Dean's present address: Department of Physiology, University
College London, London WC1E 6BT, UK.
M. Kneussel's present address: Centre for Molecular Neurobiology,
D-20251 Hamburg, Germany.
M. A. Nassar's present address: Department of Biology, University
College London, London WC1E 6BT, UK.
| |
References |
-
Bayer KU,
Schulman H
(2001)
Regulation of signal transduction by protein targeting: the case for CaMKII.
Biochem Biophys Res Commun
289:917-923[ISI][Medline].
-
Behe P,
Stern P,
Wyllie DJ,
Nassar M,
Schoepfer R,
Colquhoun D
(1995)
Determination of NMDA NR1 subunit copy number in recombinant NMDA receptors.
Proc R Soc Lond B Biol Sci
262:205-213[Medline].
-
Ben-Ari Y
(2001)
Developing networks play a similar melody.
Trends Neurosci
24:353-360[ISI][Medline].
-
Bliss TV,
Collingridge GL
(1993)
A synaptic model of memory: long-term potentiation in the hippocampus.
Nature
361:31-39[Medline].
-
Bourne HR,
Nicoll R
(1993)
Molecular machines integrate coincident synaptic signals.
Cell [Suppl]
72:65-75.
-
Brandon NJ,
Delmas P,
Kittler JT,
McDonald BJ,
Sieghart W,
Brown DA,
Smart TG,
Moss SJ
(2000)
GABAA receptor phosphorylation and functional modulation in cortical neurons by a protein kinase C-dependent pathway.
J Biol Chem
275:38856-38862[Abstract/Free Full Text].
-
Brandon NJ,
Delmas P,
Hill J,
Smart TG,
Moss SJ
(2001)
Constitutive tyrosine phosphorylation of the GABA(A) receptor gamma2 subunit in rat brain.
Neuropharmacology
41:745-752[Medline].
-
Burnashev N,
Schoepfer R,
Monyer H,
Ruppersberg JP,
Gunther W,
Seeburg PH,
Sakmann B
(1992)
Control by asparagine residues of calcium permeability and magnesium blockade in the NMDA receptor.
Science
257:1415-1419[Abstract/Free Full Text].
-
Crair MC
(1999)
Neuronal activity during development: permissive or instructive?
Curr Opin Neurobiol
9:88-93[ISI][Medline].
-
Cull-Candy S,
Brickley S,
Farrant M
(2001)
NMDA receptor subunits: diversity, development and disease.
Curr Opin Neurobiol
11:327-335[ISI][Medline].
-
Erzurumlu RS,
Jhaveri S
(1992)
Trigeminal ganglion cell processes are spatially ordered prior to the differentiation of the vibrissa pad.
J Neurosci
12:3946-3955[Abstract].
-
Erzurumlu RS,
Kind PC
(2001)
Neural activity: sculptor of "barrels" in the neocortex.
Trends Neurosci
24:589-595[ISI][Medline].
-
Faissner A,
Kruse J
(1990)
J1/tenascin is a repulsive substrate for central nervous system neurons.
Neuron
5:627-637[ISI][Medline].
-
Faissner A,
Steindler D
(1995)
Boundaries and inhibitory molecules in developing neural tissues.
Glia
13:233-254[ISI][Medline].
-
Forrest D,
Yuzaki M,
Soares HD,
Ng L,
Luk DC,
Sheng M,
Stewart CL,
Morgan JI,
Connor JA,
Curran T
(1994)
Targeted disruption of NMDA receptor 1 gene abolishes NMDA response and results in neonatal death.
Neuron
13:325-338[ISI][Medline].
-
Fox K,
Schlaggar BL,
Glazewski S,
O'Leary DD
(1996)
Glutamate receptor blockade at cortical synapses disrupts development of thalamocortical and columnar organization in somatosensory cortex.
Proc Natl Acad Sci USA
93:5584-5589[Abstract/Free Full Text].
-
Goodman CS,
Shatz CJ
(1993)
Developmental mechanisms that generate precise patterns of neuronal connectivity.
Cell [Suppl]
72:77-98.
-
Handyside AH,
O'Neill GT,
Jones M,
Hooper ML
(1989)
Use of BRL-conditioned medium in combination with feeder layers to isolate a diploid embryonal stem cell line.
Roux Arch Dev Biol
198:48-55
TOP
ABSTRACT
Introduction
