Chen et al. 10.1073/pnas.0800897105. |
Fig. 6. Representative field of area between cell bodies in 3-week cultured rat hippocampal neurons after high-pressure freezing, freeze substitution, and plastic embedding. Synapses, typically on spines (*), are plentiful.
Fig. 7. Binding domain of ABR monoclonal antibody to PSD-95. (A) Different regions from PSD-95 tagged with GST, anti-GST used as a loading control. (B) PDZ1 domain of PSD95 is targeted by the PSD-95 antibody.
Movie 1. Contiguous virtual sections, each 1.5 nm thick, displayed in X-Y plane through a tomographic reconstruction of a mushroom-shaped hippcampal spine. Each individual structure in the PSD was segmented to generate the 3D renderings, such as that shown in Fig. 1 of paper. The plane of virtual section shown here was selected to be approximately parallel to some vertical filaments (arrowheads) to display their full length in a single section. More typically only parts of vertical filaments and other structures are captured in a single section (arrow). Each structural element must be segmented by tracing it through adjacent sections (arrowheads indicate same filament in successive sections), usually in three different orthogonal projections, to be sure that it is a single, continuous structure. QuickTime Format. (Scale bar: 100 nm.)
Table 1. Dimensions of NMDAR- and AMPAR-type structures
Extracellular domains | Cytoplasmic domains | |||||
Height nm ± SD (range, N) | Length nm ± SD (range, N) | Width nm ± SD (range, N) | Height nm ± SD (range, N) | Length nm ± SD (range, N) | Width nm ± SD (range, N) | |
AMPAR (tomography) | 9 ±1 (8-11, n = 14) | 17 ± 3 (13-24, n = 14) | 7 ± 1 (5-9, n = 14) | 5 ±1 (3-6, n = 18) | 18 ± 3 (14-25, n = 18) | 10 ± 1 (8-13, n = 18) |
AMPAR (measured*) | 10-12 | 16 | 8 | |||
(predicted**) | ~3 | 6-7 | 6-7 | |||
NMDAR (tomography) | 9 ± 1 (6-11, n = 16) | 14 ± 3 (11-20, n = 16) | 8 ± 1 (6-11, n = 16) | 16±4 (10-24, n = 14) | 20 ± 2 (13-22, n = 14) | 14 ± 2 (11-17, n = 14) |
NMDAR (predicted**) | ~10 | ~20 | ~20 | |||
*Based on intact AMPAR structure from single-particle EM (1).
**Calculated (2) and SI Table 2.
1. Nakagawa T, Cheng Y, Ramm E, Sheng M, Walz T (2005) Structure and different conformational states of native AMPA receptor complexes. Nature 433:545-549.
2. Uversky VN (2002) What does it mean to be natively unfolded? Eur J Biochem 269:2-12.
Table 2. Stokes radius of cytoplasmic domains of gutamate receptors
Receptor domains | Number of residues | Molecular weight (kDa) | Stokes radius (native* nm) | Stokes radius (unstructured* nm) |
NR2A-c tail | 627 | 70 | 5.0 | |
NR2B-c tail | 644 | 72 | 5.2 | |
GluR1-c tail (long form) | 81 | 8.5 | 1.6 | |
GluR2-c tail (short form) | 50 | 5.7 | 1.4 |
*Protein domain conformation was predicted from its primary sequence by IUPred (1).
Stokes radius under various conformation was deduced from its mass based on the formulas by Uverskey et al. (2, 3).
1. Dosztanyi Z, Csizmok V, Tompa P, Simon I (2005) IUPred: Web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content. Bioinformatics (Oxford, England) 21:3433-3434.
2. Uversky VN (2002) What does it mean to be natively unfolded? Eur J Biochem 269:2-12.
3. Uversky VN, Vassilenko KS (2002) Natively unfolded proteins: A point where biology waits for physics. Protein Sci 11:739-756.
SI Methods
Selection of Dendritic Spines for EM Tomography.
Only dendritic spines free of apparent ice damage and showing both a prominent PSD in cross-section view and a complement of synaptic vesicles in the presynaptic terminal were selected for EM tomography. Eight spine heads from unstimulated spines were reconstructed. Because all their PSDs manifested similar structural details, we focused our efforts on rendering all of the individual molecular structures in a selected PSD. Our approach is built on including every individual structure in the tomogram, so it could be representative of other systems, such as slice cultures or intact, perfused brain.Classification of Spines.
The segment of PSD in our reconstruction (Fig. 1) is 120 nm by 400 nm, so the whole PSD is estimated to be ~400 nm × 400 nm, giving it an area of 0.16 μm2. The mean lengths of PSDs is 340 ± 102 nm (ranging 230-530 nm) in the eight tomographic reconstructions. Therefore, we expect areas of these PSDs to range from 0.05-0.28 μm2, which fits closely the range in area between the thin and mushroom types of spine, as derived from serial EM (1). We used published criteria (1) to classify the spine in Fig. 1 as a mushroom-type spine. As originally classified, the neck width (dn) in mushroom spines is much less than the width of the spine head (dh), whereas in stubby spines dn » L (length of the spine). In our example
= 0.24, and 
= 0.36, making this spine a better fit to a mushroom type. Furthermore, the area of its PSD (0.16 μm2) fits that expected of a mushroom spine with a macular PSD (1). Dimensions of PSD-95 Family Members and Glutamate Receptors.
The PSD-95 family members share many structural similarities, including three PDZ domains, one SH3, and one GK domain, permitting estimates of the dimensions of a generic family member. Each of the three PDZ domains in these MAGUKs corresponds to a 4-nm globular protein (2), the first two tandem PDZ domains are separated by only a few amino acids (3), and the SH3 and GK domains associate with each other forming a 4 nm × 8 nm structural module (4). Fully extended, the molecule would be ~20 nm long and 4 nm in diameter.For estimating sizes and cytoplasmic domains of glutamate receptors, their conformation can be predicted from their primary sequences obtained with GCG-Lite (http://helixweb.nih.gov/emboss_lite/). Molecular weight from primary sequences was calculated at http://bioinformatics.org/sms/prot_mw.html. Amino acid abundance was calculated with SAPS (http://molbio.info.nih.gov/molbio /gcglite/saps.html). To predict whether glutamate receptor c-tails are natively unstructured, the entire sequence of a receptor subunit was put into the following predictors that calculate the extent of disorder: IUPred (http://iupred.enzim.hu/), FoldIndex (http://bip.weizmann.ac.il/fldbin/findex), and PONDR (http://www.pondr.com/). Estimates of the size of the cytoplasmic side of glutamate receptors was based on calculations of Stokes radius (5, 6).
Estimation of the Number of AMPAR-Type Structures in a PSD.
Twenty-nine AMPAR-type structures were identified in a 120-nm slab of reconstructed PSD. Therefore, a typical 400-nm PSD would have to include at least 29 AMPA receptors. If, however, we assume that the density of AMPAR-type structures is constant around the entire PSD, then their total number would be ~100. Thus, we estimate the number of AMPAR-type structures in a PSD to be ~30-100.Epitope Mapping.
To map the epitope recognized by ABR monoclonal antibody to PSD-95 (MA1-046; Affinity BioReagents), DNA constructs encoding domains of PSD-95 gene (GenBank NM_019621) corresponding to amino acid residue to N terminal (1-64), PDZ1 (65-151), PDZ2 (155-249), N teriminal + PDZ1 (1-151), and also the full-length PSD95 were PCR cloned into the pDEST-24 vector, which includes a GST tag (Invitrogen). All constructs were confirmed by sequencing. Constructs from PSD-95 were transformed into BL21 (DE3)- competent cells (Novagen). Bacteria were cultured and induced by IPTG. Protein extraction and GST purification were done with a BugBuster GST bind purification kit (70794-3; Novagen). The protein samples were boiled in SDS/PAGE sample buffer and then separated by SDS/PAGE and transferred onto a nitrocellulose membrane. For analysis, the ABR anti-rat PSD-95 monoclonal antibody and goat anti-GST (27-4577-01; Amersham Pharmacia) were used as primary antibodies; secondary antibodies were goat anti-mouse IgG and rabbit anti-goat IgG conjugated to horseradish peroxidase. Bound antibodies were detected by enhanced chemiluminiscence (Amersham Pharmacia Biosciences).1. Harris KM, Jensen FE, Tsao B (1992) Three-dimensional structure of dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: Implications for the maturation of synaptic physiology and long-term potentiation. J Neurosci 12:2685-2705.
2. Doyle DA, et al. (1996) Crystal structures of a complexed and peptide-free membrane protein-binding domain: Molecular basis of peptide recognition by PDZ. Cell 85:1067-1076.
3. Long JF, et al. (2003) Supramodular structure and synergistic target binding of the N-terminal tandem PDZ domains of PSD-95. J Mol Biol 327:203-214.
4. McGee AW, et al. (2001) Structure of the SH3-guanylate kinase module from PSD-95 suggests a mechanism for regulated assembly of MAGUK scaffolding proteins. Mol Cell 8:1291-1301.
5. Uversky VN, Vassilenko KS (2002) Natively unfolded proteins: A point where biology waits for physics. Protein Sci 11:739-756.
6. Uversky VN (2002) What does it mean to be natively unfolded? Eur J Biochem 269:2-12.