1. Rubenstein, J.L.R. & Merzenich, M.M. Model of Autism: Increased Ratio of Excitation/Inhibition in Key Neural Systems. Genes, Brains and Behavior 2:2555-267 (2003).
2. Welsh et al. Is autism due to Brain Desynchronization? Int. J. Devel. Neuroscience 23:253-263 (2005).
3. Purcell et al., Postmortem Brain Abnormalities of the Glutamate Neurotransmitter System in Autism. Neurology 57:1618-1628 (2001).
4. Zuo et al. Neurodegeneraton in Lurcher Mice Caused by a Mutation in d2 Glutamate Receptor Gene. Nature 388:769-773 (1997).
5. Sobolesky et al., X-ray Structure, Symmetry and Mechanism of AMPA-subtype Glutamate receptor. Nature 462: 745-757 (2009).
www.rcsb.org Rutgers Center for Structural Biology. Database of biological structures.
Concise answers to the Study Guide questions and molecular images resulting from the modeling exercises are available on the Proteopedia website. Simply type in ‘autism’ in the search bar.
1. Autism & the Excitatory/Inhibitory Ratio. How do Rubenstein & Merzenich justify the statement that a more weakly inhibited cortex is functionally more poorly differentiated? Describe the developmental trajectory of the rat auditory cortex. What is a tonographic map? What is meant by ‘a critical window’ during development of the tonographic map? What experiments did Merzenich and colleaques perform to validate their hypothesis? These authors implicate the procedural memory (‘stimulus response’) in the learning of language. What brain circuits are thought to be involved in this form of learning? What are the mechanisms in local cortical circuits that could give rise to an altered balance in the E/I ratio (page 259)? The authors comment on tuberous sclerosis (page 262) and speculate on how mutations in tuberin (TSC1) and hamartin (TSC2) might impair speech acquisition in humans. Based on Study Guide No. 2, how would you reframe this argument? (HINT: substitute auditory cortex for visual cortex). What are the possible routes to new therapies for improving language comprehension in autistic individuals Page 264)?
2. Neurophysiology & Neuropathology. Describe the basic architecture and circuitry of the cerebellum (see Chapter 19. Purves et al. NEUROSCIENCE, 4th Ed. Sinauer 2008). The Purkinje cell is the central player in the modulation of movement by the cerebellum. The Purkinje dendritic arbor receives two types of excitatory input (parallel and climbing fibers) and delivers an inhibitory projection to the deep cerebellar nuclear cells, which transmit information through the thalamus to the motor cortex. Note that the cerebellar nuclear cells receive excitatory inputs from both the mossy and climbing fibers so that, in effect, the output of the deep cerebellar nuclear cells is a modulated signal reflecting ‘error corrections’ to motor plans. Describe the parallel fiber system (granule cells). How many synapses does each Purkinje cell make with the parallel fiber system? What is the origin of the climbing fibers that synapse with Purkinje cells? (HINT: see paper by Welsh et al.)? The inferior olive (IO), a brainstem structure, acts as a ‘pacemaker’ for the cerebellum. Give a brief description of the desynchronization theory of autism put forth by Welsh and collaborators.
A general question is: how does the cerebellum learn motor patterns? This takes place at Purkinje Cell-Parallel Fiber synapses. The synaptic strength of these connections is weakened by the endocytosis of AMPA-type glutamate receptors. The net effect is to weaken the inhibitory influence on the deep cerebellar nuclei. Thus, the signals projected to motor cortex are more greatly influenced by the information conveyed by climbing fibers (the inferior olive) than by the corrections projected from the molecular layer. What has been ‘learned’ by the brain in this process?
One of the first observations (Bauman & Kemper, 1988) establishing that autism is a neurobiological condition, as opposed to being a consequence of poor parenting, was the reduction in the number of Purkinje cells in post mortem tissues compared with cerebella from normal controls. Using more recent methods (micro-arrays for cDNA, RT-PCR for mRNA, Western blots (immunoprecipitation) for protein levels, and autoradiography for AMPAR subunits in post-mortem brains, Purcell et al. make a strong case for AMPAR loss in Purkinje cells. What is the most compelling image in their paper? Is this consistent with the proposal by Rubenstein & Merzenich or is some other AMPAR associated mechanism involved?
3. Mouse Model. The ‘Lurcher’ mouse, characterized by its motor disturbances (ataxia), arises from a semi-dominant single base mutation that substitutes an alanine with a threonine in a subunit of the AMPA-type receptor. The cerebellum of the Lurcher mouse displays Purkinje cell loss. What is the major electrophysiological difference between wild-type and Lurcher Purkinje cells? (HINT: examine the channel conductance). The mutation occurs in the third alanine position of the conserved SYTANLAAF signature motif for ionotrophic GluR family. Explain the meaning of the data collected from Lurcher mice presented in Figure 3 of Zuo et al. What is the apparent reason for the death of Purkinje cells in the second and third postnatal week?
4. Structural Analysis of the AMPA glutamate receptor. Chemical signaling across synapses by glutamate (excitatory) or GABA (inhibitory) is mediated by ligand-gated ion channels spanning the lipid bilayer and embedded in the post-synaptic density. There are two types of GluR receptors: AMPA/Kainate and NMDA. There is a third class, metabotropic (mGluR) receptors that activate G-coupled signals intracellualrly. We will take up this class later this term when we discuss Fragile-X syndrome.
AMPA receptors respond quickly to convert glutamate binding energy into the work of opening an ionic channel. The influx of positive ions raises the resting potential of the target neuron, eventually leading to depolarization of the membrane. A central question is how does ligand binding open channel pores? Secondly, what is the mechanism of de-sensitization of the channel to further stimulation? AMPA receptors are known to recycle in response to long-term depression (LTD) in cerebellar circuits.
NMDA receptors are critically important for learning and memory. They function as ‘coincidence detectors’ allowing Ca++ entry into the cell only when glutamate binds and the membrane voltage changes. The entry of Ca++ leads to changes in synaptic strength via CAMKII and CREB. In the LTP experimental paradigm (organotypic slices or cultured neurons) NMDA signaling is the key step initiating post-synaptic remodeling. NMDA receptors have a Mg++ dependent desensitization mechanism.
AMPAR crystallizes in space group P1 (one complete receptor in the unit cell). The analysis of the structure relies heavily on the superposition of previously determined structures, which are homologous to subunits in the full-length receptor. The GluA2cryst structure is an ‘antagonist-bound, closed’ state, implying that we should be able to derive a mechanism for ligand-gated pore opening or, at a minimum, explain what closes the channel.
The GluA2cryst structure is a homo-tetramer, like the K+ channel structures studied last week. However, it does not possess overall four-found symmetry about a central axis passing through the pore. Describe the quasi-symmetry of the four subunits for each of the major domains (ATD, LBD, TMD). Use superposition (‘alignment’ option in SPDPV) to pinpoint the largest difference in conformation between the ‘A-type’ subunit and the ‘B-type’ subunit for each of the domains. The ATD determines the types of association that a given receptor can form and the kinetics of activation. A separate crystal structure for the ATD dimer has been obtained (Jin et al, 2009). Use superposition (‘alignment’ option in SPDPV) to pinpoint the largest difference in conformation between these dimers and the GluA2cryst dimers. What is the overall RMSD?
What is meant by the term ‘domain swapping’? In the GluA2cryst structure this results in a gross difference in the overall surface presentation view of the ‘A-type’ compared with the ‘C-type’ subunit. Create an image showing this.
The LBD domain possesses the classical ‘clam-shell’ structure consisting of two subdomains (S1 and S2). Describe the unusual topology of this domain (HINT: observe how the chain leaves the domain and passes several times through the plasma membrane before re-emerging to form S2.). The structure of the LBD has been independently determined with a number of different ligands (including the glutamate agonist) bound in the clam-shell cleft. Use superposition (‘alignment’ option in SPDPV) to pinpoint the conformational changes that result from agonist versus antagonist binding. Examine the clam-shell hinges and comment on the kinds of amino acids found there.
The ‘closed state’ TMD domain of GluA2crys appears to have four fold symmetry and be very homologous to the classical closed state K+ channel structure (KcsA) we studied last week. Use superposition (‘alignment’ option in SPDPV) to determine the rms difference between the trans-membrane helices, the pore helix, and the channel strand. What closes off the entrance to the channel in the GluA2cryst structure? (HINT: M3 ‘cuff’). Where is the Q/R site in the GluA2cryst structure? Compare the structures of the polypeptide chains that link the end of helix M3 to the S2 domain of the LBD for the ‘A-type” and ‘B-type’ subunits.
Study Figure 11. Explain how the superposition of the glutamate-bound LBD domain (‘open-state’) onto the antagonist-bound LBD domain (‘closed state’) gives a crucial piece of information about the mechanism of channel activation. (HINT: what happens to the channel-blocking C-terminal ends of the M3 helices?). The ala-to-thr mutation in the Lurcher mouse occurs on the M3 helix. Does the channel opening mechanism derived from this structural comparison explain why the ion channel is constituently open?
- Return to A Structural Biologist Looks at Autism: A Course of Study -