Neural Cell Properties - Biology Guide PDF

Basic Properties of Neural cells Mature nerve cells are unable to divide and with very few exceptions cannot regenerate their cell processes. In humans 10-100 billion cells are provided at birth, the natural daily loss of about 10,000 cells throughout life causing only a negligible reduction of the...

Basic Properties of Neural cells Mature nerve cells are unable to divide and with very few exceptions cannot regenerate their cell processes. In humans 10-100 billion cells are provided at birth, the natural daily loss of about 10,000 cells throughout life causing only a negligible reduction of the entire cell population. Considering that each cell makes about 1,000 synaptic contacts (resulting in 10- 100 thousand billion total contacts) the enormous complexity of neuronal wiring in the human brain becomes apparent. Basically, nerve cells consist of a cell body (soma or pericaryon) from which two functionally different types of cell processes emanate: dendrites are often highly branched and receive electrical signals from other neurons. In some cases, their surfaces are covered with small protrusions called dendritic spines. Axons (also called neurites) send out electrical signals. In contrast with dendrites, they are of uniform diameter along their entire length, which can vary considerably. Axons can be divided into one or several collaterals forming right angles with the main axon. Their tips are often highly branched, leading to the formation of up to 1,000 nerve endings per axon. Synapses are specialized structures at nerve endings allowing chemical or electrical signal transfer to target cells (for details see below). Besides neurons the population of brain cells additionally comprises glial cells most of which are able to divide throughout life. In contrast with neurons glial cells typically possess a uniform set of cell processes and usually are significantly smaller than neurons. Four major subtypes of glial cells are known (astrocytes, oligodendrocytes, microglia, Schwann cells, ependymal cells) each having its own specialized cellular functions (for details see chapter “Properties of Glial Cells”). There is a close physiological dependence and functional cooperation between neurons and glial cells in the nervous system. The structural (morphological) analysis of neural cells (i.e., neurons and glia) is greatly hampered by the enormous number and structural complexity of these cells in the brain tissue. Staining of thin tissue sections with non-specific dyes (8 μm thickness, 1000 μm = 1 mm) is frequently used for microscopic analysis of brain tissue. By this technique, however, most of the cell processes are not visible due to low contrast of these fine structures and are also cut off and hence cannot be traced over longer distances. However, at the end of the nineteenth century a technique was devised by Camillo Golgi and Ramon y Cajal that allows to completely stain a small subset of the neural cell population with silver salts (silver impregnation). This allowed to discriminate a large number of neuronal cell types mainly on the basis of their dendritic tree shape. Although the neuronal morphologies are assessed by novel, computer-based approaches for a more precise classification, there are classically four basic categories described. - Multipolar neurons – the most predominant in the brain and spinal cord and are inclusive of motor neurons as well as interneurons. These cell types have a single axon extending from one end of the cell body and several dendrites branching as they protrude from the other side of the cell body. - Bipolar neurons – that are only associated with afferent impulses. They have a single axon projecting from one end of the oval cell body and a lone dendritic tree extending from the other end. - Unipolar neurons – have a single axon projecting from the spherical cell body. - Pseudounipolar neurons – have one single process bifurcating close to the cell body that serves both the role of the axon and the role of the dendrite. Shapes of dendritic trees are highly variable from one neuronal type to another within each class, and this diversity contributes to differential processing and computation of synaptic or sensory inputs. Dendritic geometry plays key roles in shaping intrinsic firing patterns, and coincidence detection. Intracellular structures The cell is filled with cytoplasm, which is a thick, gelatinous solution that is enclosed by the cell membrane. It is mainly composed of water, salts, and proteins and provides an environment for many cellular processes to occur, such as protein synthesis, the first stage of cellular respiration (known as glycolysis), etc. The cytoplasm of neural cells contains a highly efficient scaffolding system that is required to stabilize the extensively branched cellular structure and to concomitantly allow for structural flexibility and material transport. This so-called cytoskeleton comprises three major classes of protein fibers of different sizes, structures and chemical composition: Microtubules are long unbranched hollow tubes of approx. 20 nm diameter (1 Mio nanometers = 1mm), that consist of the protein tubulin as alpha and beta heterodimers that are in building mainly 13 protofilaments arranged in a hollow tube. Intermediate filaments. The neuron specific neurofilaments are long solid filaments of 10 nm diameter, that consist of three neuron-specific neurofilament proteins. They have flexible polymer arms standing out that repel neighbouring neurofilaments. The neurofilaments determine the radius of the axon. Glial cells possess glial filaments instead consisting of the glial fibrillary acidic protein (GFAP). Any type of injury or degeneration of the neuronal cells leads to an increase in the concentration of neurofilament light chain (NfL) in the cerebrospinal fluid (CSF) and blood. Therefore, these measurements have recently been used to assess the diagnostic and prognostic value in a variety of diseases of nervous tissue. Mikrofilaments or actin-filaments are small solid filaments of 5 nm diameter, consisting of the globular monomeric protein actin. They are even more abundant in muscle cells, where they support muscle contraction. All three cytoskeletal elements are organized into networks that resist deformation but can reorganize in response to externally applied forces or internal stimuli. In addition, neural cells contain a number cell organelles that are fundamental for cell survival in general and hence occur in any other cell type as well *. In the nucleus the complete genetic material of an organism is stored in the form of deoxyribonucleic acid (DNA). Since in a given cell only a limited set of genes is converted into proteins highly specific control mechanism of gene activation are needed which operate in a cell type and cell stage-specific manner. The process of gene activation is initiated by gene transcription (biosynthesis of messenger RNA), being regulated by proteins called transcription factors.  In the endoplasmic reticulum (ER) protein and lipid biosynthesis are taking place. Protein biosynthesis (translation) requires the presence of ribosomes on the ER surface which associate there with mRNA produced in the nucleus. In the Golgi-complex proteins are sorted and packaged into transport vesicles. Dependent on their individual content some of these vesicles fuse with the plasma membrane, while others deliver their content to predestined intracellular sites. In the mitochondria the energy required for cell survival is generated in a storable form. The complex physiological mechanisms underlying energy production and storage include the biochemical breakdown of glucose ultimately leading to the production of adenosine-tri-phosphate (ATP), the main energy storage molecule of all cells. This process works efficiently only in the presence of oxygen. Nerve cells permanently have high energy consumption and hence are heavily dependent on a constant glucose and oxygen supply via the blood stream. _________________________________ * Neural cells share many basic properties with other cells. For a more detailed description of these basic cell biological features and functions the reader is referred to classical cell biological textbooks. Axonal transport Intracellular transport of material between the cell body and the cell processes is of vital importance for nerve cells. This is particularly obvious in the case of nerve endings which are often quite remote from the soma, where nearly all of the biosynthetic processes are taking place. Regarding the velocity and type of material being translocated two major classes of axonal transport can be discerned: The slow axonal transport shows a velocity range of 1- 6 mm per day. It transports proteins of the cytoskeleton (see above) and supports nerve fiber growth during development and regeneration. The fast axonal transport has a velocity range of 100-400mm per day. It transports larger particles such as vesicles or even mitochondria. Besides an anterograde transport from the cell body to the nerve ending a retrograde transport running in the opposite direction occurs. Fast axonal transport is guided by microtubules and is carried by specialized vehicle proteins that bring their cargo to the desired site. Anterograde transport is affected by the protein kinesin, while retrograde transport relies on the protein dynein. Both are called motor-proteins which use ATP as energy source for transport activity. The molecular mechanism underlying kinesin driven axonal transport has been analyzed in great detail. It provides a striking example of how proteins can do work in the presence of ATP. From the structure of the kinesin molecule valuable evidence about its mode of operation could be deduced. It consists of two very similar subunits (“dimer” structure), each containing a globular motor domain for binding and processing of ATP as well as a rod-like stalk domain that fixes the cargo to the vehicle. Both are linked by a flexible hinge region. Transport activity is initiated by the splitting of an ATP molecule that is bound to one of the motor domains. This leads to a structural reorganization of the kinesin protein that lifts off one of his motor domains from the microtubule surface and moves it forward to become fixed on the microtubule surface again (hand-over hand walking model). The consumption of one molecule of ATP is sufficient for an eight nm step of a single kinesin molecule on the surface of a microtubule towards the nerve ending. As will be outlined later controlled changes of a proteins 3-D structure is a fundamental mechanism to regulate its functional activity.  Basic Properties of the Neuronal Plasma Membrane The entire neuron, like all other cells, is enclosed by a plasma membrane – a double layer mainly composed of phospholipid molecules (for details see below) - acting as a barrier preventing the contents of the cell from mixing with those of the extracellular space. The plasma membranes of most cells consist by half of proteins and by half of lipids. Phospholipids, which form the major part of membrane lipids, are rather complex molecules consisting of two highly hydrophobic fatty acid chains, that are linked via glycerol to the hydrophilic head groups of the molecule. The latter include a phosphate group which depending on the subtype of phospholipid carries another characteristic hydrophilic constituent: e.g., choline in the case of the phospholipid lecithin. In total, phospholipids possess a hydrophilic portion (headgroups) and a hydrophobic portion (fatty acid tails). From this mixed hydrophilic-hydrophobic property the behaviour of phospholipids in the aqueous environment of a cell can be predicted: when dropped on a water surface, phospholipid molecules will evenly spread out in a single molecular layer with the headgroups contacting the water surface and the fatty acid tails oriented to the opposite side. In the plasma membrane, actually, two layers of phospholipid molecules are joined together. The fatty acid groups of each layer are facing each other in the hydrophobic center of the membrane, while the headgroups are aligned on either side of the membrane surface. The core of the plasma membrane thus forms an efficient barrier against the passage of hydrophilic molecules. The second major constituents of the neuronal plasma membrane, the proteins, are much more complex in their biochemical composition, their structure and their functions. The building blocks of proteins are amino acids. They all share two basic structural elements: an amino group and an acidic group, while the rest of the molecule is highly variable. Hence these residual groups alone determine the physico-chemical properties of an amino acid. Some of them are hydrophobic others are hydrophilic. Some are acidic, while others are basic. As a whole twenty different species of amino acids are sufficient to create the huge number of different protein molecules that occur in a living organism. Since proteins are chains of amino acids of variable length and composition their fundamental properties are defined by their amino acid sequence (primary structure of the protein). Under normal physiological conditions a protein will, however, not exist as a long- unfolded chain of amino acids (polypeptide chain) but will adopt higher order structures that are more or less determined by the individual amino acid sequence. Two types of highly regular secondary structures can be observed: An alpha helix is generated when a single polypeptide chain turns regularly about itself to make a rigid cylinder.  A beta sheet is formed when an extended polypeptide chain folds back and forth upon itself, with each section of the chain running in the direction opposite to that of its neighbours. This gives a very rigid structure. Besides a rather irregular random coil structure is quite frequently found as well. This often forms the more flexible parts of a protein. In most proteins all three structural elements coexist in the various regions at different proportions resulting in highly characteristic folding patterns. The next level of structural organization of a protein is called tertiary structure or conformation giving the protein its characteristic overall three-dimensional shape. It is stabilized by chemical bonds between the variable residues of amino acids. As described above for the motor protein kinesin targeted changes in the 3D-shape of a protein (also termed conformational changes) typically coincide with the onset and offset of specific functional activities of a protein. In some proteins, such as kinesin, several polypeptides (so-called subunits) are grouped together to form a functional unit. Kinesin, containing two subunits is called a “dimer”. Proteins consisting of four subunits would be called “tetramers”.  Structural characterization of proteins For the structural analysis of proteins, a wide range of biochemical and biophysical techniques are available. The primary structure is obtained by biochemical methods, either by direct determination of the amino acid sequence from the isolated protein or indirectly - but much more rapidly, from the nucleotide sequence of the corresponding gene or cDNA. To obtain the secondary structure CD-spectroscopy is the method of choice. In many cases the secondary structures of a protein can be predicted from its amino acid sequence by computer. For the determination of the tertiary structure or conformation a combination of very time-consuming and highly demanding biochemical and biophysical techniques are required: initially the protein of interest must be isolated in high purity and relatively large amounts (several milligrams) to allow its crystallization. Unfortunately, this is a poorly predictable process, which is not always successful even in the hands of experts. Subsequently the three-dimensional structure can be solved by x-ray crystallography. Most powerful x-ray beams are generated in synchrotron storage rings where electrons travel close to the speed of light. The parallel beams of x-rays are diffracted by the protein crystal and the characteristic diffraction pattern is recorded on a detector or x-ray film. From this the atomic structure of a protein can be calculated. Functions and Features of Enzymes Enzymes are a special class of proteins which are important for the control of virtually all vital cell functions. As should be expected they comprise an enormously large number of protein families that control a huge diversity of biochemical reactions in a cell. The molecule which is biochemically modified by a given enzyme is called its substrate. The name of an enzyme is determined by adding the ending -ase to the name of its substrate: an enzyme modifying DNA would be called DNAse, one that modifies ATP would be called ATPase. The specificity of an enzyme for its substrate is due to mutual structural matching similar as a key match to its lock. The activity of an enzyme hence can be efficiently controlled by slightly modifying its tertiary structure to induce matching or mismatching of the binding pocket with its substrate, respectively. Many enzymes split their substrate into fragments of various sizes and/or numbers. Digestive enzymes are well known examples. Enzymes that split ATP into ADP and phosphate (ATPases) are particularly important for cells, since they allow the release of stored energy to be used for all sorts of physiological processes (see for example the paragraph about the sodium-potassium pump). Another class of enzymes controls the activity of other proteins (including other enzymes!), instead. They do this by attaching to or removing phosphate residues from their substrate, which induces a conformational change in the substrate protein. Such enzymes are collectively called –kinases. If their substrate is a protein, they are called protein kinases.  Membrane Proteins Membrane proteins are particularly important in nerve cells, since they control physiological interactions and signalling with other cells. Since proteins are highly hydrophilic molecules the question arises how their association with the hydrophobic parts of phospholipid membranes can be accomplished. While it is easily understood that proteins can be fixed on either membrane surface as so-called peripheral membrane proteins the question remains how proteins can become deeply embedded into the hydrophobic core as so-called integral membrane proteins. As mentioned above, the building blocks of proteins, the amino acids widely differ in their side chain properties. Among the twenty amino acids used for protein biosynthesis several possess highly hydrophobic side chains. This allows to inserting such stretches of a polypeptide chain into the core of a membrane that are composed of hydrophobic amino acids. By computer-assisted screening of the amino acid sequence of protein stretches comprising at least 20 hydrophobic residues in series can be identified and related to membrane-spanning segments of a protein (transmembrane segments). Some proteins contain only a single transmembrane segment, while others exhibit more then 10. In several cases the number of these segments is useful to classify protein families. Bioelectricity Electricity generally requires the separation of charges to take its rise. Technical electricity is based on the movement of electrons and hence works extremely fast. In living cells electrical signalling involves the movement of ions, instead, including both negative (i.e., anions such as chloride) as well as positive charges (i.e., cations such as sodium, potassium, or calcium ions). Since ions are enveloped by water molecules, they are highly hydrophilic and hence cannot easily pass the lipid bilayer of a plasma membrane. Membrane potentials are built up by an unequal distribution of charges on both sides of a membrane. The neuronal plasma membrane collects charges on either surface similar like a capacitator. Due to its large surface area and small thickness, it possesses a relatively high capacitance (C= area/distance). During the pioneering days of electrophysiology, the squid has proven as a valuable animal model: due to its relatively large size (> 0.5 mm in diameter) the squid giant axon allowed to record membrane potentials with rather coarse technical equipment. Most of the basic information about electrical signalling in nerve cells was gained by experiments performed on this system in the middle of the 20th century by Alan Hodgkin and Andrew Huxley. The ionic composition of the squid axonal cytoplasm (intracellular compartment) in fact greatly differs from those of the extracellular body fluids (blood or sea water): while the sodium ion + concentration (Na ) is approx. eight-fold higher in the extracellular fluid than in the axoplasm the reverse is true for potassium (K+), which occurs in an about twenty-fold higher concentration in the axoplasm as compared to the extracellular space. Chloride is enriched in the extracellular fluid, similar as sodium. To maintain such steep gradients of ionic concentrations active transport mechanisms are necessary. The unequal distribution of sodium and potassium is accomplished by a so-called sodium-potassium pump which is fuelled by ATP (about two thirds of the total energy consumption of a nerve cell is due to this process!). This pump is a large membrane protein with ten transmembrane segments. The intracellular portion harbours an ATP-binding site. This site is able to split ATP enzymatically into ADP and phosphate, hence it acts as an ATPase. In total the transport process is an unequal exchange of sodium for potassium ions: for three sodium ions transported out of the cell two potassium ions are carried in. The molecular mechanism underlying this ion exchange process has been -and still is being- analysed in great detail. The splitting of bound ATP leads to a conformational change in the intracellular part of the sodium-potassium pump so that sodium becomes occluded in a binding pocket close to the membrane. Subsequent removal of ADP opens the binding pocket and releases sodium into a hydrophilic tunnel between the transmembrane segments of the pump protein. Ensuing removal of the phosphate residue that was split of from ATP evokes another conformational change which now occludes potassium ions in a binding pocket at the extracellular face. Binding of a new ATP causes potassium release into the cytoplasm. The process is constantly repeated as long as enough ATP is available. The unbalanced exchange of sodium for potassium makes (3:2) makes only a small contribution to the membrane potential. The major part comes from a selective permeability of the membrane for potassium ions but not for sodium (see below). The Resting Membrane Potential Since the plasma membrane is selectively leaky to potassium ions, potassium diffuses out of the cell, following the concentration gradient previously built up by the sodium-potassium pump. The efflux of these cations creates a significant electrical potential across the membrane that finally prevents further outward diffusion of potassium. When both forces (chemical ion gradient and the counteracting electrical potential) are at equilibrium a stable membrane potential is attained, which is called resting potential. The resting potential seems to be mainly determined by the outward diffusion of potassium ions. This hypothesis can be tested by a mathematical approach. The Nernst Equation The Nernst equation allows for calculating electrical potentials (E ) that are generated by the x diffusion of ions.  E = RT/zF ln [X+] /[X+] x o i where R is the gas constant, T the absolute temperature in degrees Kelvin, z is the charge of the + + ion, and F is the Faraday constant, while [X ] is the extracellular and [X ] the intracellular o i concentration of ion X. For monovalent ions (such as sodium or potassium) at room temperature the equation reduces to E = 58 mV log [X+] /[X+] x 10 o i Insertion of the extracellular and intracellular potassium ion concentrations as measured for the squid giant axons leads to E = 58 mV log (20/400) x 10 E = 58 (-1.3) mV = -75 mV x This value is quite close to the resting potential determined experimentally for the squid giant axon albeit being slightly less negative (-70 mV). Accordingly, the resting potential in fact is mainly due to the diffusion of potassium ions, but not exclusively. Other ions, including sodium and chloride make a certain contribution, too. To consider these ions properly the Nernst equation was extended by Goldman, Hodgkin and Katz as follows: + + - p [Na ] + p [K ] + p [Cl ] Na o K o Cl i ____________________________ E = RT/zF ln x + + - p [Na ] + p [K ] + p [Cl ] Na i K i Cl o where p is the permeability factor for a given ion. The Action Potential Experimental manipulation of the resting potential in principle can be done in two ways: injecting current by a microelectrode will further increase the potential difference across the membrane, leading to hyperpolarization of the membrane. If the polarity of the microlectrode is reversed before current injection this will lead to a decrease in the potential difference, an effect called depolarization. If a short current pulse is applied that leads to only a small depolarization the membrane potential will passively follow and thereafter rapidly return to the resting potential again. The situation is very different, however, with larger depolarizing stimuli. At around -40 mV depolarization a critical strength, or threshold, is reached. Beyond this threshold, one observes a strongly overshooting depolarization that is superimposed on the passive response and lass for several milliseconds. Typically, the depolarization goes beyond 0 mV leading to a change in polarity, i.e., the inside of the neuronal membrane becomes briefly positive relative to the extracellular side. This active response is called action potential. An important property of action potentials is that they are all-or-none events: if the stimulus strength is further increased the action potential will retain its size. As a consequence, information about stimulus intensity must be encoded by a different mechanism. If a given stimulus is lasting longer a whole series of action potentials will be evoked. With increasing stimulus intensity, the delay between consecutive action potentials decreases as a whole leading to a higher frequency of action potentials (frequency code). Refractory period If pairs of stimuli are applied to a nerve at varying temporal intervals the minimal interval required to achieve a normal response can be determined. These experiments show that for several milliseconds after firing an action potential, it is impossible to evoke another action potential, no matter how large the depolarizing stimulus is. In other words, the axon is refractory to stimulation during this period. This absolute refractory period is followed by a relative refractory period during which the second action potential is of smaller size than the first one. In subsequent chapters it will be shown that the refractory period is important to maintain unidirectional signal propagation along a nerve fiber. Furthermore, the molecular basis of electrical unresponsiveness will be addressed.  Ionic mechanisms underlying the generation of action potentials The sodium hypothesis A most obvious explanation for the fast membrane depolarization that coincides with an action potential would be a sudden change in the permeability of the neuronal plasma membrane for sodium, leading to a sodium influx according to the sodium concentration gradient. To prove this hypothesis the extracellular sodium concentration was lowered experimentally. Under these conditions action potentials started with a considerable delay and were of smaller size. By gradually lowering the extracellular sodium concentration this effect became more and more pronounced. The voltage clamp technique In order to directly measure the ionic currents underlying an action potential a novel technique was applied, termed voltage clamp technique. It consists of an electronic feedback system that holds the membrane potential constant at a voltage chosen by the investigator. In its simplest form the voltage clamp consists of two separate electrodes, one connected to a voltage and the other connected to a current-passing amplifier. A negative feedback loop is created by adding a feedback amplifier, which compares the voltage set by the experimenter (so-called command voltage) with the measured membrane voltage. As mentioned above during an action potential the membrane voltage normally changes in a depolarizing direction. During a patch clamp experiment this change in membrane voltage is carefully compensated by current injection via the current- passing amplifier under the control of the feedback amplifier. The size of the compensating current is equal to those of the ionic current that flows through the cellular membrane during an action potential. Ionic currents underlying an action potential By such an approach Hodgkin and Huxley were able to record membrane currents of the squid giant axon that occurred after setting the membrane potential to 10mV, being well above the threshold for an action potential. They found a biphasic current response consisting of an early inward current that after a few milliseconds was superimposed by a delayed outward current. By ionic substitution experiments they could show that the early inward current was carried by sodium ions, while the remaining outward current was carried by potassium ions. As a whole the characteristic of the sodium current fitted very well with the membrane depolarization occurring during the rising phase of an action potential while the potassium current suited best with the ensuing membrane repolarization. A most remarkable property of the sodium current was its spontaneous inactivation a few milliseconds after its initial rise. Pharmacology of ionic currents In order to better characterize the protein pores through which sodium and potassium ions flow during an action potential a selective blockade of either current was desired. The poison of the buffer fish (Tetrodotoxin, TTX) proved as a very useful pharmacological tool for this purpose. It precisely fits into the opening of the sodium channel protein, but is too large to penetrate the pore. By plugging the pore TTX selectively prevents sodium influx into the cell. As a consequence, action potentials are no longer generated ultimately leading to the death of the animal that has taken up the poison. For the experimenter the application of TTX allows to study the potassium current in isolation, since potassium ions flow through separate channels, that are not affected by TTX. A selective blockade of potassium channels is possible by application of the small organic molecule tetra-ethyl-ammonium (TEA), which allows to recording sodium currents in isolation. By these experiments it was proven that at least two different types of protein channels occur in the neuronal plasma membrane one selectively permeable for sodium the other for potassium. Isolation and functional characterization of the sodium channel protein Channel-specific toxins such as TTX were also of great help for the biochemical isolation and purification of the sodium channel protein. The electric organ of the electric eel proved as most suitable starting material, since it was relatively enriched of voltage-gated sodium channels. With the aid of radioactively-labeled toxin a high molecular weight protein could be identified that made up the pore-forming complex (α-subunit) of the voltage-gated sodium channel which was accompanied by two accessory β-subunits of much lower molecular size. With the purified channel protein available it became possible to clone and sequence the cDNA encoding the protein by methods of gene technology. The amino acid sequence of the α-subunit was deduced from the nucleotide sequence of its cDNA and subsequently screened for transmembrane segments (see chapter “Membrane Proteins”). It turned out that the channel protein possessed four large domains of similar structure, each including six transmembrane segments (S1-S6). One of these (S4-segment) appeared quite unusual, since it contained several positively charged amino acid residues. Furthermore, the region connecting S5 and S6 revealed a hairpin loop structure. In order to verify the electrophysiological properties of the isolated channel protein and to compare it with those of native sodium channels the cDNA was converted into mRNA by in vitro- transcription. The resulting mRNA was injected into unfertilized eggs (“oocytes”) of frogs. The frog oocytes “accepted” the foreign mRNA and produced large amounts of sodium channel protein. By subsequent electrophysiological recoding the voltage-gated sodium current of the frog oocytes could be compared with those occurring naturally in neuronal cells. This technique additionally allowed to performing all sorts of genetic manipulations of the sodium channel cDNA before expressing it in frog oocytes. By this approach the specific functional roles of several of parts of the channel protein could be experimentally elucidated. First of all, exchanging the positively charged amino acids of the S4 segments for neutral ones abolished the voltage-sensitivity of the channel protein, emphasizing that S4 represents the voltage-sensor of the channel. Secondly it was shown that the intracellular loop connecting the third and the fourth domain of the channel protein is responsible for the spontaneous channel inactivation. Thirdly the four hairpin loops connecting the S5 and S6 segments together build the wall of the hydrophilic channel pore. In the case of potassium channels the lack of a rich source of channel protein made a different approach necessary. In the fruit fly Drosophila, a mutation was identified that showed an unexpected behaviour under ether anaesthesia. Unlike normal flies this mutant, called Shaker, shook its legs and wings, suggesting some neurological deficit. Electro-physiological recording revealed a prolonged duration of action potentials, as would be expected if the potassium channel conductance was reduced. Voltage-clamp studies in fact revealed a strong reduction of the potassium outward current. Molecular biological analysis of the mutated gene locus (making use of a technique called “positional cloning”) allowed to isolating and sequencing the underlying gene. The deduced amino acid sequence of the encoded potassium channel showed remarkable similarities with those of the sodium channel. It included, however, only a single large domain comprising six transmembrane segments (to form a functional potassium channel, however, four of these subunits are grouped together). As described above for the sodium channel the S4- segment was rich in positively charged amino acids and there was a hairpin loop between S5 and S6. As a whole, both channel classes were obviously sharing basic structural elements. More close inspection of the Shaker locus revealed the presence of several other potassium channel genes, encoding various families of potassium channels with different electrophysiological properties. Corresponding genes were later identified in mouse and human as well. Single channel analysis By the invention of the patch clamp technique (developed by Erwin Neher and Bert Sakmann) it became possible to directly measure the activity of single ion channels (“single channel recording”). In contrast with traditional methods using sharp microelectrodes to penetrate the cell membrane, the surface of a patch clamp glass electrode (patch pipette) is quite smooth. The patch pipette is gently placed against the cell surface (attached patch). Subsequent patch clamp recording can be done in one of three different configurations. In the first case (whole-cell mode) the experimenter applies suction through the pipette to disrupt the membrane patch under the electrode, thus fusing the cellular cytoplasm with the fluid in the patch pipette. By this technique the activity of all ion channels in the plasma membrane of a cell is collectively measured. To obtain single channel recording the patch pipette is retracted from the cell surface thus pulling apart a small piece of membrane that ideally contains only a single ion channel (inside-out and right side-out configuration, respectively). Under these conditions typical current fluctuations become visible reflecting the opening and closing of an individual ion channel. In the case of voltage-gated channels the open-probability increases with the degree of membrane depolarization. By applying various command voltages and plotting the current against the applied voltage the single channel conductance of an ion channel can be determined (measured in Pico-Siemens, 1 pS = 1pAmpere/1Volt).  Axonal Signal Propagation An action potential is not a local event but it moves along an axon to finally reach its nerve ending. What are the mechanisms driving it away from its initial site of generation? Since a quite similar task is performed by electrical cables in technical systems the question arises if axons might work in a comparable manner. Cable properties of an axon The passive properties of an axon (so called cable properties) can only be studied if the active physiological mechanisms are experimentally suppressed. This can be most easily achieved by local cooling of a nerve fiber. Under these conditions a local depolarization applied to a nerve fiber still passively spreads to both sides (electronical spread). Similar as in electrical cables the spread of electrotonical potentials along an axon is dependent on the resistance and the capacitance of the fiber. Due to the relatively high internal resistance of axons, the electrotonic potential, however, is decaying exponentially so that it will move only a few millimetres away from its original site. This distance is determined by the size of the fiber. Larger fibers that have a lower resistance allow for a wider spread of the potential. Active signal propagation What are the active mechanisms that amplify the signal allowing it to travel over longer distances? Under normal physiological conditions voltage-gated sodium channels residing in the membrane regions ahead of a depolarized site will become activated if a certain threshold of depolarization is achieved. They will then produce a full-size action potential at the novel site (obeying the all or none law). Sodium channels residing in membrane areas behind the site of original stimulation normally are still inactivated (refractory period!), hence the signal is conducted only in one direction. By constantly repeating passive electrotonical spread and re-amplification via sodium channel activation, the signal continuously travels along the axon, without any loss of signal strength. By this mechanism of continuous impulse conduction maximal velocities of up to 25m/sec can be achieved. A well-known example is the squid giant axon, which has a diameter of more than half a mm. Since the conduction velocity increases only with the square root of the fiber diameter, increasing the fiber size at a certain point becomes inefficient and costs a lot of space in the nervous tissue. Myelination An alternative mechanism, which occurs in vertebrate species only (comprising fish, amphibia, reptiles, birds, and mammals), is provided by surrounding the nerve fiber with an insulating sheath (so called myelin sheath). This sheath is periodically interrupted at the nodes of Ranvier. Only at these sites action potentials can be generated, since voltage-gated sodium channels are heavily clustered there. Hence in a myelinated nerve fiber the signal does not move continuously along the axon membrane but jumps from one node to the next (saltatory impulse conduction). This type of signal propagation is not only faster but also less energy consuming than the continuous mode of conduction, since energy consuming ion transport (sodium-potassium ATPase) is limited to the nodes of Ranvier. The insulating sheath of myelin enhances impulse conduction by two basic mechanisms: the leakage current across the axonal membrane is reduced the membrane capacitance is decreased Taken together both factors allow for a faster and wider spread of electrotonic potentials until the next node of Ranvier. In myelinated nerve fibers there is a linear relationship between axon diameter and conduction velocity. The myelin sheath is not a product of the neuron itself but is generated by two highly specialized populations of glial cells. In the peripheral nerve fibers myelin is generated by Schwann cells, while in the central nervous system (CNS: brain and spinal cord) this task is performed by oligodendrocytes. While a Schwann cell produces only a single myelin segment, each of the oligodendrocytes sends out several cell processes to form a multitude of myelin segments on different nerve fibers. In both cases myelin represents a modified plasma membrane of the glial cell being wrapped around the axon in many turns. By squeezing out the cytoplasm between the overlapping membranes a compact multilayered membrane sheath is formed. This structure is strongly stabilized by a set of myelin-specific adhesion proteins. Myelin proteins Membrane apposition at the cytoplasmic side is accomplished by a set of fairly low molecular weight basic proteins (MBP) that associate at the intracellular membrane surfaces. Furthermore, a highly hydrophic proteolipid protein (PLP) with four transmembrane segments is involved in the extracellular membrane apposition. While MBP is common to both Schwann cells and oligodendrocytes PLP is unique to myelin and myelin-forming cells of the CNS. In the myelin of peripheral nerve fibers PLP is replaced by the P -glycoprotein. 0 P possesses a single transmembrane segment. Its extracellular domain has an antiparallel β-sheet 0 structure as is typical for many cell adhesion molecules, in general. P provides for a stable contact 0 at the extracellular membrane faces. Multiple sclerosis Multiple sclerosis is the most frequent demyelination disease in the Western world. It is an autoimmune disease caused by auto-antibodies against myelin-specific proteins, in particular against myelin basic protein (MBP). According to the molecular mimicry theory it is assumed that certain viruses (in particular the measle virus) can carry proteins that share structural similarities with MBP into an infected organism, which subsequently produces antibodies against these protein structures. Due to the blood-brain barrier these antibodies normally do not have access to MBP in the brain. However, once a local and transient break of the barrier occurs, myelin in the vicinity of the afflicted blood vessels become destroyed by immune cells from the blood stream. Properties and Functions of Glial Cells Apart from oligodendrocytes and Schwann cells which are responsible for myelination in the central and peripheral nervous system, respectively, three additional types of glial cells occur in the CNS (brain and spinal cord). Astrocytes Astrocytes have many cell processes that radiate out from the cell body in all directions, giving them a star-like appearance, and possess a cell-type specific cytoskeleton protein, the glial fibrillary acidic protein (GFAP). Their cell branches are in close contact with blood vessels on the one side and with nerve cells on the other. Hence, they provide nutrients coming from the blood stream (e.g. glucose) to nerve cells and they perform metabolic waste removal. A most vital function is the removal of certain neurotransmitters (such as glutamate and GABA, see below) from the extracellular space, which may become harmful to nerve cells if they stay there for a too long time. Furthermore, astrocytes balance out the extracellular potassium concentration thus regulating the excitability of nerve cells. More recently a participation in neuronal signal processing has been reported. Microglia Microglial cells are small ramified cells that occur throughout the adult central nervous system. Following injury, they become activated and start to divide and migrate to the site of injury and probably transform there into macrophage-like cells, being capable of clearing damaged cells and debris by phagocytosis. In similarity with lymphocytes activated microglia express a variety of immunomodulator peptides and surface receptors involved in immune defence. The capacity of microglia to exert neurotoxic or neuroprotective effects depends on their phenotypic polarization. Specifically, M1 microglia enhance pro-inflammatory responses through the release of substantial quantities of inflammatory cytokines. This process contributes to tissue damage. Conversely, M2 microglia can facilitate tissue repair and regeneration, provide protection to neurons. Ependymal cells Ependymal cells are forming the lining of the fluid-filled spaces in the brain and spinal cord that is the ependyma. Their main function is to secrete, circulate the cerebrospinal fluid that fills the ventricles of the central nervous system. Neurotransmission The neuromuscular junction is a specialized type of synapse, connecting the nerve ending of a motor nerve with a skeletal muscle fiber. Due to its relatively large size and convenient experimental access, it has served as a model system to analyze the basic factors and mechanisms underlying chemo-electrical signal transfer from a nerve terminal to a target cell (neurotransmission). The pioneering experiments done by Sir Bernard Katz revealed that shortly after the arrival of an action potential at the nerve ending a small potential is generated in the neuromuscular junction that triggers the formation of a muscle action potential. It is called endplate potential and it can be efficiently blocked by the poison curare. By carefully titrating the amount of curare applied to the neuromuscular junction the endplate potential can be recorded in isolation from the muscle action potential. It was thus shown that the endplate potential - unlike the action potential - is a graded potential, since its amplitude varies with the strength of nerve fiber stimulation. Ultrastructural studies of the neuromuscular junction revealed the presence of numerous small vesicles in the nerve terminal. These contain a signalling molecule (the neurotransmitter acetylcholine, ACh). Upon arrival of an action potential at the nerve terminal the vesicles fuse with the plasma membrane (exocytosis) and release their content, the Ach, into the extracellular space between the nerve ending and the muscle surface (synaptic cleft). The neurotransmitter subsequently diffuses to the postsynaptic muscle membrane and specifically binds to a receptor protein that is heavily clustered at this site (acetylcholine receptor, AChR). The AChR is a chemically gated ion channel that opens as soon as Ach has bound to it. As a consequence, cations (mainly sodium) flow into the muscle cell, producing a small membrane depolarization, that corresponds to the endplate potential described above. Finally, a muscle action potential is elicited by voltage-gated sodium channels in the muscle membrane, opening in response to the endplate potential. Synaptic transmission at a cholinergic synapse (like the neuromuscular junction) is terminated by enzymatic degradation of the neurotransmitter by the enzyme acetylcholine- esterase (AChE), which is localized in the postsynaptic membrane. Regulation of exocytosis and neurotransmitter release The fusion of synaptic vesicles with the plasma membrane (exocytosis) is functionally linked to the arrival of an action potential at the nerve ending by voltage-gated calcium channels. When these channels open calcium ions flow into the nerve terminal, due to the steep concentration gradient. This initiates a complex cascade of molecular activities that ultimately lead to the attachment of synaptic vesicles to the presynaptic plasma membrane and to fusion of the lipid bilayers: In the resting state synaptic vesicles are docked to microfilaments within a nerve terminal via a protein called synapsin. The rise in the intracellular calcium concentration induces a conformational change in the synapsin protein, allowing synaptic vesicles to detach from the microfilaments. In the next step vesicles become attached to the intracellular surface of the presynaptic membrane. On the molecular level this is accomplished by the vesicle protein synaptobrevin on the one side and the presynaptic membrane protein syntaxin on the other. Both proteins align their α-helices in parallel to form a very compact and stable coiled-coil structure. The ensuing formation of a fusion complex requires the presence of the presynaptic membrane protein synaptotagmin which is believed to act as a calcium sensor. The proteins, which are responsible to mediate vesicle fusion are the so-called SNAREs. They can be divided into two categories: vesicle or v-SNAREs, which are incorporated into the membranes of transport vesicles during budding, and target or t-SNAREs, which are associated with nerve terminal membranes. Properties of Acetylcholine Receptors Pharmacological properties A most convenient way to characterize a neurotransmitter receptor in the living tissue is by studying the effect of certain drugs on its physiological activity. Some drugs will mimic the effect of the original neurotransmitter, they are called agonists. Others will block the physiological response of the receptor even in the presence of the neurotransmitter, they are called antagonists. By testing several agonists and antagonists it turned out that two different subtypes of acetylcholine receptors occur in the skeletal and the heart muscle, respectively. The AChR in the skeletal muscle is blocked by curare (antagonist) and activated by nicotine (agonist), while those in the heart muscle is blocked by atropine and activated my muscarine, instead. Both receptor subtypes are named according to their agonist: hence the skeletal muscle receptor is called nicotinic AChR (nAChR), while those in heart is called muscarinic AChR (mAChR). Structure and function of the nAChR protein In order to characterize this protein biochemically, it was necessary to find a tissue that was highly enriched in it (see also: isolation of the voltage-gated sodium channel). The electric organ of the electric ray (Torpedo marmorata) proved as a most suitable source for isolating the nAChR. The snake toxin peptide α−bungarotoxin helped to identify the receptor protein during the isolation procedure. Biochemical studies ultimately showed that the nAChR is an oligomeric protein consisting of four different polypeptides termed α,β,γ,δ. In the native protein complex the α- subunit occurs twice, hence the AChR is a pentamer (i.e., it contains a total of five subunits) with a stoichiometry of 2αβγδ. Since ACh binds to the α-subunits, each AChR can bind two molecules of the neurotransmitter. The next step was the cloning of each of the four subunits by recombinant DNA technology, allowing to deduce their amino acid sequences. Subsequent hydrophobicity analysis revealed in each case four transmembrane segments (M1 –M4). By electron microscopy the 3D-structure of the AChR was determined at least at low resolution. As was expected it forms an ionic pore with a large extracellular funnel-like structure. The hydrophilic pore is lined by the five M2 segments of the individual subunits. The two ACh-binding pockets are formed by the extracellular portions extending from the M1 segments of the α− and γ- as well as the α- and δ- subunits. Binding of two molecules of ACh induces a twisting movement within the M2-segments, which opens the ionic pathway through the membrane. Structure and function of the mAChR While the nAChR is a chemically-gated ion channel (belonging to the family of ionotropic receptors) the mAChR operates in a fundamentally different manner. First of all, it is not a pore- forming protein, but consists of only a single subunit with seven transmembrane segments. Secondly on the intracellular side it is linked to a GTP-binding protein (G-protein), a protein complex of three different subunits (guanosine-triphosphate, GTP, is closely related adenosine- triphosphate, ATP). The mAChR therefore is a member of the G-protein coupled receptor family (also called metabotropic receptor family). Binding of a single ACh molecule induces a conformational change in the receptor protein that ultimately leads to a dissociation of the G-protein complex. The membrane anchored βγ-subunit moves laterally within the membrane until it reaches a potassium channel to which it binds. The channel is opened and the ensuing potassium efflux will hyperpolarize the postsynaptic membrane thus creating an inhibitory postsynaptic potential (IPSP).  Amino acid neurotransmitters Besides acetylcholine certain amino acids can operate as neurotransmitters, as well. These are glutamate, gamma-aminobutyric acid (GABA), and glycine. The family of glutamate receptors comprises both metabotropic and ionotropic receptors. The former are G-protein coupled receptors that come in at least eight different subtypes most of which induce ipsps. Ionotropic receptors can be subdivided pharmacologically into NMDA-receptors (sensitive to the agonist N-methyl-D- aspartate) and non-NMDA-receptors. As expected, non-NMDA-receptors are gated by glutamate and due to the influx of sodium ions induce an excitatory postsynaptic potential. NMDA-receptors by contrast are unresponsive to glutamate alone, since they have a magnesium block that keeps their ionic pathway blocked. This block is removed by depolarization of the postsynaptic membrane. Hence NMDA-receptors require membrane depolarization prior to or at least concomitant with the release of glutamate since they are both chemically-gated and voltage-gated ion channels. From this it is clear that synapses containing NMDA-receptors alone are functionally silent. Hence NMDA-receptors normally co-localize and co-operate with non- NMDA-receptors at glutamatergic synapses. Another striking feature of NMDA-receptors is their high permeability for calcium. Influx of calcium trough NMDA-receptors elicits postsynaptic activities leading to a use-dependent change of synaptic efficiency as observed during learning processes (see chapter “Synaptic Plasticity”). As compared with the nicotinic AChR ionotropic glutamate receptors slightly deviate in their membrane topology, in that they possess only three transmembrane segments that are connected via a hairpin loop element resembling those found in voltage-gated potassium channels. GABA-receptors comprise ionotropic and metabotropic receptors as well. The former are called GABA -receptors while the latter are termed GABA -receptors. GABA -receptors have A B B + seven transmembrane domains and similar as metabotropic AChR open K -channels via the liberated βγ-subunits of G-proteins, ultimately leading to an ipsp. GABA -receptors are chemically-gated chloride channels, which induce ipsps via a chloride A influx into the postsynaptic cells. In their basic structural arrangement, they closely resemble ionotropic AChR in that they are composed of five subunits (pentamer) each of which possesses four transmembrane segments. The involvement of GABA-receptors in a neuronal circuit is most conveniently demonstrated by the application of specific drugs. The plant toxin bicuculline is a specific GABAergic antagonist that competes with GABA for its binding pocket. In living organisms  it causes severe convulsions since it abolishes all inhibitory synaptic activity. Alternatively, the toxin picrotoxin is frequently used which directly binds to the ion channel portion of the GABA-receptor and not to its binding pocket. Hence it is called an allosteric antagonist. Furthermore, GABA -receptors are the targets for a variety of therapeutically relevant drugs such A as barbiturates (components of sleeping pills) and benzodiazepines (“Valium”) which act as allosteric agonists enhancing the effect of GABA. Since these drugs all increase synaptic inhibition in the brain, they have a sedative or even anaesthetic effect in living organisms. Glycine-receptors closely resemble GABA -receptors. They are also chemically-gated chloride A channels and exhibit a similar overall molecular structure. They are more abundant in spinal cord tissue than in brain. Biogenic Amines Biogenic amines form another important class of neurotransmitter molecules. They are produced by nerve cells using the amino acid tyrosine as starting material. By enzymatic removal of the acidic group (-COOH) and subsequent addition of a hydroxyl group (-OH) the neurotransmitter dopamine is generated. Dopamine binds to G-protein coupled receptors D1-D5. The level of dopamine available in nerve terminals is controlled by the enzyme monoamineoxidase (MAO), which inactivates the neurotransmitter in the presynapse. Dopaminergic neurons are most abundant in two midbrain regions, the substantia nigra (“black nucleus”) and the ventral tegmental area. The latter is part of the mesolimbic dopaminergic pathway that is involved in the processing of rewarding stimuli and drug addiction. Dopamine receptors are also important targets of antipsychotic drugs. Furthermore, one of the most widespread neurodegenerative diseases, parkinsonism, is caused by a gradual degeneration of dopaminergic neurons in the substantia nigra. The second member of this group, noradrenaline (= norepinephrine) is generated from dopamine by the addition of another hydroxyl group (-OH). Neurons in the locus coeruleus (“blue nucleus”) represent a major source of noradrenaline in the brain. They send out numerous axon collaterals into the thalamus and throughout the cortex. The widespread distribution of noradrenergic nerve terminals in the forebrain explains the profound influence of noradrenaline on a variety of higher brain functions including multimodal alertness and motivation.  Electrical Synapses In contrast with chemical synapses that use a neurotransmitter substance to pass over neuronal signals to a target cell electrical synapses directly convey signals from one cell to the other. Hence, they do not have synaptic vesicles and the extracellular space between pre-and postsynaptic membrane is greatly diminished. Direct electrical coupling of pre-and postsynaptic membranes is achieved by densely packed membrane channels, which allow an unimpeded passage of charged or uncharged particles up to a size of approx. 1000 daltons molecular weight. These channels, termed connexons are formed by six subunits of the protein connexin, which is a four transmembrane domain protein. Electrical synapses operate more quickly than chemical synapses and, in many cases, allow current flow in both directions. As a whole they provide a means to synchronize electrical signalling in a network of neurons. Synaptic Integration Nerve cells generate action potentials at a specific site, the axon initial segment. In this region the threshold to fire an action potential is lowest, probably due to a strong accumulation of voltage- gated sodium channels there. If a single epsp is elicited at a synapse somewhere on a dendrite this local depolarization will passively spread across the neuronal membrane to finally reach the axon hillock. Since on its way much of its original voltage will decay, a single epsp usually will not suffice to overcome the firing threshold at the axon hillock. If several epsps occur in fast succession at a synapse (temporal summation) the depolarizations may superimpose thus greatly increasing the chance to generate an action potential at the axon hillock. Alternatively, several epsps might be generated simultaneously at various synapses (spatial summation) to overcome the firing threshold there. As a whole, the somatic and dendritic membrane surface of a nerve cells serves to integrate all synaptic potentials arriving at a time to ultimately decide if the target cell itself will elicit an action potential or not. In this complex process not only epsps but also ipsps have to be considered. Trafficking of receptors/channels at chemical and electrical synapses The estimated half-life for the turnover of gap junction channels is approximately 1 to 3 hours, which aligns with the turnover rates observed for other ion channels and synaptic receptors. This dynamic process may play a role in modulating electrical transmission in electrical synapses and contribute to the plasticity of chemical synapses. Postsynaptic densities (PSDs) serve as a framework that regulates this trafficking. At glutamatergic synapses, PSD95 and calcium/calmodulin-dependent protein kinase II (CaMKII) are key components of PSDs. In the case of electrical synapses, Zonula occludens protein 1 (ZO1) plays a structural role in the scaffold responsible for the organization and transport of connexons, which are delivered as unpaired hemichannels in vesicles. These hemichannels are then inserted at the edge of the gap junction plaque and align with hemichannels on the opposing membrane. Synaptic Plasticity The ability of synapses to change their response to incoming signals in a use-dependent manner is called synaptic plasticity. It is typically accompanied by an increase (or decrease) in the size of the epsp generated in response to a given stimulation. In pioneering experiments, it was found that the epsps of a synapse significantly increase after high frequency stimulation (posttetanic potentiation). The effect, however, lasts for only a few minutes. It is due to an accumulation of calcium in the nerve terminal during tetanic stimulation. In more sophisticated experiments performed by the laboratory of Eric Kandel the molecular basis of synaptic plasticity was investigated. By using the sea hare (Aplysia californica) as a model organism they could throw a bridge between a “learned” animal behaviour and the underlying cellular and molecular processes. The gills of the sea hare (actually a snail!) are very sensitive to mechanical stimulation and quickly retract when being touched (gill withdrawal reflex). Repeated weak stimulation of this kind, however, gradually attenuates the response (a process called habituation). If the habituated animal receives a strong, even toxic stimulus of the antennae, the gill withdrawal response gets sensitized. This can be explained by the wiring of sensory fibers in the Aplysia nervous system. The sensory fibers from the antennae make axo-axonic synapses on the nerve terminals of sensory fibers coming from the gills. These synapses use serotonin as neurotransmitter. Serotonin release activates the G-protein coupled serotonin receptor at the postsynapse (which is the nerve terminal membrane of the “gill fibers”). As usual for G-protein coupled receptors, the G-protein subsequently dissociates, releasing α- as well as βγ-subunits. The further chain of events is as follows: G-protein α-subunits activate the enzyme adenylyl cyclase. the enzymes start to produce cyclic adenosine monophosphate (cAMP) from ATP cAMP activates the enzyme protein kinase A protein kinase A phosphorylates potassium channels, leading to current reduction. the repolarization of action potentials arriving at the “gill fiber” synapse is delayed. neurotransmitter release is enhanced. As long as the potassium channels retain phosphorylation synaptic transmission between the gill sensory fiber and the motoneuron activating the gill muscle is enhanced.  Studies with Mutant Flies The fruit fly Drosophila has proven as a most convenient model organism to analyze the molecular basis of cellular processes, since mutations can be rather conveniently generated and isolated. Learning mutants (“stupid flies”) were identified by an odour discrimination task. It was found that deficits in this learning task in many cases correlated with mutations of genes encoding proteins controlling the cAMP-level, such as adenylyl cyclase (AC, increases the level of cAMP) or phosphodiesterase (PDE, decreases the level of cAMP), emphasizing the key role of cyclic AMP in learning processes. More recently is was found that cAMP in addition to its effect on ion channel phosphorylation (se above) can also affect gene expression through its binding to the regulatory parts of transcription factors in the cell nucleus. This could lead to long term changes in the nervous system such as the new outgrowth of nerve fibers and the formation of additional synapses. Long term Potentiation The hippocampus is a forebrain structure (telencephalon) that is closely associated with long term memory formation, since its removal abolishes declarative memory formation in humans. Therefore, this brain region has become a model system to analyze the cellular and molecular basis of learning processes in higher animal species. Since the majority of internal connections of hippocampal neurons are arranged within a plane it is possible to cut out tissue sections (< 1/2 mm thick) from the hippocampus without damaging its internal circuitry. Owing to the highly ordered arrangement of neurons in discrete layers electrophysiological recording of defined cells within a tissue section is feasible. Most frequently the synaptic connections between CA3 and CA1 neurons are being studied (CA stands for cornu ammonis = ammons horn). CA3 cells are connected with CA1 neurons through the so-called Schaffer collaterals, the nerve terminals of which are excitatory and release glutamate. The pioneering observation was made in 1973 by Bliss and Lomo who showed that a high-frequency stimulation of Schaffer collaterals produces a significant increase in the epsp. Since the effect lasts for hours it is called long term potentiation (LTP). On a molecular level NMDA-receptors play a key role in the induction of LTP. As outlined above (chapter amino acid neurotransmitters) these receptors require postsynaptic membrane 2+ depolarization that removes a Mg -block to become responsive to glutamate. High-frequency stimulation evokes depolarization of an extended portion of the dendritic tree of CA1 neurons. Under these conditions glutamate release from Schaffer collateral synapses coincides with postsynaptic membrane depolarization. The NMDA-receptor hence can be considered as a coincidence detector molecule. Since NMDA-receptor ion channels are permeable to calcium the calcium concentration in the postsynaptic cell increases. Calcium acts as second messenger that induces several metabolic effects in the target cell: 1. Postsynaptic effects: Calcium activates an enzyme called calcium-calmodulin kinase II (CaM-kinase II). This enzyme first of all phosphorylates itself (autophosphorylation) which stabilizes its own activated state (explaining the long-term effect of LTP). Secondly it induces the insertion of additional non-NMDA receptors into the postsynaptic membrane, leading to an increase in the epsp even after a weaker stimulation. Activation of glutamate receptors in the spine also induces actin-dependent modulation of spine morphology. Glutamate contributes to the initial actin-dependent spine motility by actin reorganization and also to events that lead to spine stability. 2. Presynaptic effects: Calcium additionally activates an enzyme that produces the gaseous neurotransmitter nitric oxide (NO). Gases like NO can easily penetrate membranes and hence can reach the presynaptic nerve terminal (retrograde messenger). NO is believed to enhance neurotransmitter release by still unknown mechanisms. Basic Architecture of the Spinal Cord The spinal cord is part of the central nervous system. It is surrounded by the vertebral column and exhibits a segmental organization. The spinal cord is connected to its target areas on the body surface by peripheral nerves. On cross sections a central grey matter, harbouring most of the neurons, is surrounded by white matter largely consisting of myelinated nerve fibers. On the dorsal side pairs of ganglia occur (dorsal root ganglia) that contain the cell bodies of pseudo-unipolar ganglion cells. These send out long sensory fibers into the skin, muscles, and joints to detect mechanical stimuli, such as touch, stretch, pain, and temperature. Sensory signals are subsequently transferred by the axons of pseudo-unipolar cells into the grey matter of spinal cord via the dorsal roots. In many cases the information is directly targeted to motoneurons residing in the ventral part of the grey matter (ventral horn). If their axons project back to muscles of the body region that had been previously stimulated a reflex arc is activated. Owing to this direct connection between sensory input and motor output, often involving only a single synaptic contact, reflexes are very fast and cannot be influenced voluntarily. The knee jerk is a famous example for such a monosynaptic reflex. It is initiated by a week hammer hit on the surface of the patella. A slight dislocation of this bone pulls the tendon to which it is connected. This activates a sensory structure (stretch receptor) within the adjoining muscle. The stretch signal enters the spinal cord via the dorsal roots as described above and is directly transferred onto a motoneuron that activates the extensor muscle in the leg. To rule out concomitant activation of the antagonistic flexor muscle motoneurons commanding this muscle are inhibited by the same signal via a circuitry called afferent collateral inhibition (for details see transparency). Ascending and descending connections In addition to this local processing of signals on the level of a single spinal cord segment there are extensive connections with the brain, in particular with cortical areas involved in voluntary movement control (motor cortex). These are provided by ascending and descending pathways that exhibit a specific and highly ordered arrangement in the white matter: ascending fibers transporting sensory signals related to temperature and pain traverse the midline of the spinal cord to ascend on the opposite side (contralateral projection) to the brain, while those conveying signals about touch remain on the same side (ipsilateral projection). Descending fibers from the motor cortex are also connected in an ipsilateral fashion. As a result, a half-sided transection of the spinal cord produces the so-called Brown-Séquard syndrome, which is typically evoked by motor cycle accidents. In this syndrome the injured person suffers from paralysis and loss of touch sensation in the leg of the injured side, while it experiences loss of temperature and pain sensation in the opposite leg. Rhythmic movement control In addition to the execution of reflexes the spinal cord plays an important role in regulating rhythmic movements such as swimming, walking, and running, thus relieving higher brain regions from detailed movement control (servo function). Rhythmic movements are typically based on a periodic and alternating activation of antagonist muscle groups of a limb. This periodic activity can be traced back to the activity of corresponding neurons in the ventral roots which show a wiring diagram that is designed to create periodic activity patterns (half-centre model, for details see transparency). The central pattern generator of the spinal cord can be even nailed down to the molecular level. By the coordinated activity of certain ion channels a periodic interruption of action potential firing is provided: thereby the concerted interaction of voltage-gated calcium-channels and calcium- activated potassium channels plays a key role (for details see transparency). The Autonomic Nervous System The autonomic nervous system controls the activity of inner organs such as heart, lung, stomach, liver etc. It operates independently from voluntary control and is of key relevance for the survival of an organism since it coordinates vital visceral activities such as digestion, heart beat or the control of body temperature. Structurally it is closely interlocked with the spinal cord and peripheral nerves and it exhibits also a reflex arc organization. As compared with the above- mentioned somatic reflex arc an autonomic reflex arc exhibits the following specializations: the cell bodies of the autonomic neurons in the spinal cord are localized in the intermediate region of grey matter. Their axons leave the spinal cord via the ventral roots joining spinal nerves. their axons do not project directly to the target organs but terminate in autonomic ganglia first from there so-called postganglionic fibers emanate to contact a certain inner organ each inner organ receives a dual innervation through both a sympathetic and a parasympathetic branch. dual innervation allows for a better fine-tuning, coordination, and adaptation of inner organ activities to changing environmental and emotional conditions. typically, those organs preparing the organism for fight and flight are activated by the sympathetic branch and inhibited by the parasympathetic branch. organs preparing the organism for rest and digest are activated by the parasympathetic pathway and inhibited by the sympathetic pathway. The sympathetic ganglia are lined up as a chain along the spinal cord. Parasympathetic ganglia usually lie close to the target organ, or embedded in the tissue of the target organ instead. Sympathetic ganglia occur in the breast and lumbal region of the spinal cord, while parasympathetic fibers emanate from the nucleus of the nervus vagus in the brainstem as well as from the sacral region of spinal cord. In the parasympathetic pathway both preganglionic and postganglionic fibers are cholinergic. The target neurons of the former possess nicotinic AChRs, while the targets of the latter use muscarinic AChRs, instead. In the sympathetic pathway preganglionic fibres release acetylcholine which binds to nicotinic AChRs. Postganglionic fibers use noradrenaline instead. The third branch of the autonomic nervous system is the enteric nervous system (ENS). The enteric nervous system (ENS) is large, complex and uniquely able to orchestrate gastrointestinal behaviour independently of the central nervous system (CNS). It is situated within the wall of the gastrointestinal tract and in human contains 200-600 million neurons, distributed in many thousands of small ganglia, the great majority of which are found in two plexuses, the myenteric and submucosal plexuses.  The myenteric plexus forms a continuous network that extends from the upper esophagus to the internal anal sphincter and is located between the longitudinal and circular layers of muscle. Submucosal ganglia and connecting fiber bundles form plexuses in the small and large intestines buried in the submucosa, but not in the stomach and esophagus.