1. NMDA receptors
    1. Several experiments have shown that synaptic strengthening occurs when molecules of the neurotransmitter bind with postsynaptic receptors located in a dendritic spine that is already depolarized.
    2. Kelso et al (1986) found that if they artificially depolarized the CA1 neurons and then stimulated the axons that formed synapses with them, the synapses became stronger. The stimulation and depolarization had to occur together, see Fig. 4-13.
    3. The strongest evidence implicating NMDA receptors in LTP comes from the research with drugs that block NMDA receptors, such as AP5.
      1. AP5 prevents the establishment of LTP in field CA1
      2. but AP5 has no effect on LTP that has already been established.
      3. Thus, although the activation of NMDA receptors is necessary for LTP, transmission in the potentiated synapses involves non-NMDA receptors (primarily, AMPA receptors).
  2. Calcium ions
    1. The entry of calcium ions through the ion channels controlled by NMDA receptors is an essential step in LTP.
    2. Lynch et al (1984)
      1. Injected EGTA directly into hippocampal pyramidal cells
      2. EGTA binds with Ca and makes it insoluble, destroying its biological activity
      3. Found the EGTA blocked the establishment of LTP in the injected cells.
    3. Malenka et al (1988) demonstrated not only is Ca necessary for LTP, but it is sufficient.
      1. Injected nitr-5 directly into CA1 pyramidal cells then exposed them to ultraviolet light
      2. When nitr-5 is exposed to ultraviolet light, it releases its hold on Ca2+
      3. The increase of intracellular concentration of free Ca2+ triggered LTP between these cells and their input, see Fig. 4-14.
    4. The concept of dentritic spike
      1. When an action potential is triggered in the axon of the pyramidal cell, the backwash of depolarization across the cell body triggers a dendritic spike, which is propagated up the trunk of the dendrite.
    5. Yuste and Denk (1995) use a special technique, two-photon fluorescence microscopy to visualize individual spines on dendrites of pyramidal cells in hippocampal slices.
      1. Injected a special fluorescent dye that permitted observation of Ca influx directly into CA1 pyramidal cells
      2. Found that when the activity of individual synapses and depolarization of the dendrite (the action potential was triggered in the axon) occurs at the same time, the amount of Ca that entered was much greater than the sum of amounts in the two individual events, see Fig. 4-15.
    6. Magee and Johnston (1997) measured both Ca influx into individual dentrites and EPSP produced
      1. Also found that when synapses became active at the same time a dendritic spike has been triggered, Ca "hotspots" occurred near the activated synapses
      2. EPSP of the activated synapses became larger
      3. Infusion of a small amount of TTX onto the base of the dendrite prevented the formation of dendritic spikes and LTP, see Fig. 4-16. This confirmed that the dendritic spikes were necessary for LTP
  3. Protein kinases (CaM-KII, PKA, PKC, tyrosine kinase, etc.)
    1. CaM-KII is present in especially high concentration in the postsynaptic thickening
    2. Silva et al (1992)
      1. produced a targeted mutation of gene responsible for the production of CaM-KII,
      2. LTP in CA1 can not be produced
    3. These kinases might activate biochemical processes that produce postsynaptic changes that cause the insertion of AMPA receptors into the postsynaptic membrane and change the physical structure of the synapses.
  4. Protein synthesis
    1. The activated protein kinases also trigger protein synthesis.
    2. Frey et al (1988) used anisomycin (to block protein synthesis)
      1. If the drug was administered before, during, or immediately after the prolonged stimulation
        1. LTP occurred, but disappeared a few hours later
      2. If the drug was administered 1 hour after the prolonged stimulation
        1. LTP persisted
      3. Therefore, protein synthesis
        1. is not required for the earlier stages of LTP (lasting an hour or so).
        2. is necessary for establishing the later phase of LTP.
        3. is accomplished with an hour of stimulation.
    3. Protein synthesis takes place in dendrites.
      1. Dendrites contain all they need to synthesize proteins
  5. Retrograde messenger involved in LTP
    1. Nitric oxide (NO, a soluble gas) is used as a messenger in many parts of the body. NO lasts only a short time before it is destroyed.
    2. Drugs that block NO syntheses prevented the establishment of LTP in hippocampal slices
    3. NO presumably diffuses out of the dentritic spine, back to the terminal button, which eventually lead to an increase of the release of glutamate
  6. Mechanisms of synaptic plasticity (the increases in synaptic strength), see Fig. 4-17
    1. The increase of synaptic strength could be possibly caused by
      1. Presynaptic changes
        1. increased release of transmitter substance
      2. Postsynaptic changes
        1. an increased number of receptors,
        2. an increased ability of the receptors to activate changes in the permeability of the postsynaptic membrane
        3. increased communication between the region of the postsynaptic membrane and the rest of the neuron
      3. Both presynaptic and postsynaptic changes
        1. an increased number of synapses
    2. Each of these possibilities has some supporting evidence
    3. For example, several studies have found that NMDA-mediated LTP increased the number of postsynaptic AMPA receptors
      1. Tocco et al (1992), used autoradiography method
      2. incubated the brain slices with radioactive ligands for NMDA and AMPA receptors
      3. Found increased sensitivity (or numbers) of AMPA receptors, but not NMDA receptors
  7. Structure changes at synapses
    1. Hosokawa et al (1995) used a confocal microscope to observe individual dendritic spines of CA1 pyramidal cells
      1. A subpopulation of small spines grew in length and changed their orientation toward the shaft of the dendrite away from the perpendicular, see Fig. 4-18.
    2. Several studies have found that LTP increased the number of "perforated synapses" (the dendritic spine and terminal button elongate, two active zones appear), see Fig. 4-18.
    3. Buchs et al (1996) used a special stain to label Ca in dendritic spines and found that after LTP, most of the labelled spines formed perforated synapses with the presynaptic terminals.
    4. Edwards (1995) suggested that development of perforated synapses is a means by which the number of postsynaptic AMPA receptors increases, see Fig. 4-19.
  8. Explanation for the classical properties of LTP, see Fig. 4-20.
    1. Ca2+ influx acts locally, resulting input-specific LTP, but it is difficult to explain input-specificity in the later stage of LTP (protein produced could in principle travel to any synapses within the cell). How to modify only the appropriate synapses.
    2. Only when synaptic input is strong enough (more input axons stimulated) to depolarize the postsynaptic membrane, giving rise to cooperativity.