Physiological Basis of Behavior jp-rosenfeld@northwestern.edu We study how the brain makes mind and behavior METHODS (in oversimplification terms): 1.Lesions (Ablation): Oldest Method; 2 ways: a) Natural: Take them as they come (bullet holes, tumors; “Neuropsychology”.) b) You do it: chemicals, electricity (cheap). 2. Stimulation: “the opposite of lesions.” Again, electrical or chemical. 3. Recording: a)Imaging (PET, fMRI), b)Electrophysiology (EEG, ERP from neuron populations, or from single neurons) Lesions and Stimulation… ….. Assume that single structures are responsible for single behavior patterns or psychological functions. This is the height of naiveté, to wit……... “Structure A is substrate for anger.” Ergo… a) If you lesion it, no more anger, as in lobotomy. (Usually no more lots of other stuff too.) b) If you stimulate it, you get a guy angry.. Jose Delgado did the reverse….. http://www.youtube.com/watch?v=6nGAr2Ok VqE Brain and CNS are made of cells: Neurons Of course, real neurons come in all shapes & sizes, like the following: Drawing by Santiago Ramón y Cajal of neurons in the pigeon cerebellum. (A) Denotes Purkinje cells, an example of a multipolar neuron. (B) Denotes granule cells which are also multipolar. A) Purkinje Cells, B) Granule Cells Or……. http://www.youtube.com/watch?v =ifD1YG07fB8&feature=related The job of the nervous system is to transmit information from neuron to neuron, and to a target organ like a muscle which initiates behavior. As we were saying… The job of the nervous system is to transmit information from neuron to neuron, and to a target organ like a muscle which initiates behavior. This happens by action potential (“spike”) propagation down a neuron, and, usually, via synaptic transmission to another neuron. What’s a “potential?” The name implies the ability to do work…electrical work, the ability to move charged particles from one level of potential to another, lower level. All cells have a resting potential. That is, a relatively constant difference in electrical potential across the cell membrane. Say +30 mV inside vs. – 10 mv outside. This is where they spend most of their lives…at rest..it’s a good life. “Irritable” tissue—muscle cells or nerve cells--- are different Irritable cells….. …..occasionally show sudden fast changes in cell membrane potential, before recovering the resting level. In neurons, these changes are called “impulses” or action potentials, and they propagate from cell body to end of neuron. Here is a neuron membrane passing from rest to action and back to rest… In the next lecture or two…. ….we will consider how the resting neuronal membrane potential and the action potential are generated. To do this, we need to do some thought experiments… OK, first, let us note that there are 3 passive phenomena mainly influencing the situation within a neuron: (Passive means life is irrelevant to these phenomena.) 1. Membrane (semi-) permeability. 2. Chemical or concentration forces: Particles tend to move away from high and to low concentration. 3. Electrical forces. + “likes” – and “dislikes” +. – “likes” + and “dislikes” – OK, now, back to those thought experiments…… Consider a beaker of water divided by a semi-permeable membrane into 2 compartments, and I put a teaspoon of a monovalent salt into the left compartment. What does the salt ”want to” do? What forces are at work? OK, the concentration forces drive ions to the right. No problem, both positive and negative ions can go through, so do. Now what forces act? ..So another pair go through, which puts system at balance or electrochemical equillibrium…with no difference in potential(=voltage)across the membrane OK, here’s a new situation. What’s different? What forces? OK, only the cation can get through in the first instant. It does. Let’s now analyze forces and predict next moment. The anion “wants to” follow the cation but is too big. (Good example of semipermeability.) What are forces now , and can we predict next moment? We now have electrochemical equilibrium (There is still Fc but balanced by Fe) but with a residual voltage across the membrane In the previous slide, the one permeable little cation is said to be: at its Equilibrium Potential. This is the voltage across the membrane at which the electrical and chemical forces on the ion are in balance. Fc + Fe = 0 or Fe = -Fc The situation (figure) in a resting neuron is more complicated: Given this situation what are forces on K+, the most permeable cation? More complicated because more ions are involved, and they together affect electrical and chemical forces. You should note that there is a relationship between Fc and Fe… …i.e., the dis-proportionality between the left and right hand concentrations of permeable ions predicts the voltage across membrane at which system is at equilibrium. Thus we have an equation, the Nernst equation, which holds if K+ is sole permeable ion: NERNST EQUATION: E =60 Log (K+o/K+i) NERNST EQUATION with GOLDMAN EXTENTION : Derivation: Total Force on an ion, say K+ = Electrical Force + Concentration (Chemical) Force. Putting in units of Voltage, the total electrochemical force = DV[K+] = ZFE + RT ln (K+i/K+o) (Z= charge/mole,F= valence,E=membrance potential in voltage units, R= gas constant, T= temperature(absolute), ln = log to base e, K+i = inside Potassium concentration, K +o = outside Pot. Conc.) When K+ is at electrochemical equilibrium, DV[K+] = 0 = C1E + C2 ln (Ki+ /K+o), where C1,C2….(all C) are constants. So C1E = - C2 ln (K+i/K+o) = C3 ln (K+o/K+i), and dividing both sides by C1 yields E=C3/C1 ln (K+o/K+i) = C4 ln (K+o/K+i) =C5 log10 (K+o/ K+i ). C5= about 60, so NERNST EQUATION: E =60 Log (K+o/K+i) [Note: Log (x) = Log to base 10] It is noted that this applies when only permeable ion is K+ . Otherwise, one uses the Goldman equation (of which the Nernst is seen to be a special case). E= 60 Log (PK [Ko+] + PNa [Nao+]+Pcl [Cli-]…….)/ (Pk [Ki +] PNa [Nai +] + Pcl [Clo-]…….) Pk= potassium permeability coefficient, Pna= perm coeff for Na, Pcl= perm coeff for Cl. Signs of ions omitted for clarity, but note, cation outside concentrations are in numerators, anion outside concentrations in denominator. Note what happens to equation if all coefficients but Pk go to zero. Sample question: If K+o = 10000 and K+i =10, what is E if K+ is sole permeable ion? In other words, what is the K+ equilibrium potential? This Nernst Equation…. Allows us to see if a neuron is near/at equilibrium and if that ion is sole permeable one. Thus if we calculate the equilibrium potential for K+, we use: E = 60 log ([K+]0/[K+]i) = -80 which is close to but not = to the actual -70. Meaning…? The Equilibrium Potential of Na+.. Using the concentrations of Na+ inside and outside in the equation yields the equilibrium potential for Na+: E = 60 log ([Na+]0/[Na+]i) = +65 mV …but the real resting membrane potential is -70mV Is Na+ at equilibrium? We already guessed that it wasn’t due to Fc and Fe. What do we conclude about PNa or Na+ permeability? Back to the neuron… Thus, the resting membrane potential is there because relatively permeable K+ moves as close to equilibrium as it can… Is there proof? Yes. Scientists have manipulated interior and exterior [K+] and noted the effect on Em, the membrane resting pot. They manipulate exterior [K+] simply by bathing neurons in solutions where [K+] systematically varies. Interior [K+] is manipulated as if neuron were a toothpaste tube. Both can be simultaneously changed. Here are the results… Why the discrepancy?? 1) 90% is due to presence and influence of other ions. So if you used Goldman extension, most of discrepancy would go away: E= 60 Log (PK [Ko+] + PNa [Nao+]+Pcl [Cli-]…….)/ (Pk [Ki +] PNa [Nai +] + Pcl [Clo-]…….) = ~ -79 mV The rest of the discrepancyis due to the… 2) Na+ & K+ Ion exchange mechanism, Or the “Sodium Pump.” This is a dynamic biochemical process that keeps Na+ out and K+ in, as the video will now demonstrate….. http://highered.mcgrawhill.com/sites/0072495855/student_view0/chapter2/a nimation__how_the_sodium_potassium_pump_works. html The whole story http://www.google.com/#q=hodgkin+experiment&hl= en&prmd=iv&source=lnms&tbs=vid:1&ei=kTyeTL2xKY KWnAftvinDQ&sa=X&oi=mode_link&ct=mode&sqi=2&ved=0C AgQ_AU&fp=8a8ae2f39e51403c Measuring Neuronal Voltages Better for fast things (spikes)…. Hodgkin Experiment One stimulation of + 10 mV, a depolarization… You always see the overshoot… …until you see the action potential! 4 stimulations….. (Bucking Currents) Voltage Gated Ion Channels. The first channels to open are Na+ channels. The first critical finding by Hodgkin was that the greater the depolarization across the membrane, the greater the Na+ permeability (PNa), so Na+ rushes in (why?) which further depolarizes membrane. Hodgkin Cycle (inner wheel) Why doesn’t it inevitably lead to a spike? Because although the first effect of depolarization in opens Na+ channels so Na+ rushes in…. There is a delayed effect of the stimulation/depolarization, which is to more slowly open K+ channels. What does K+ “want to” do? Inner wheel PLUS Outer wheel The outer wheel … ….puts the brakes on the inner wheel. Whether you get a spike or re-polarization depends on the race between the inner and outer wheels. See James Stewart version of “Flight of the Phoenix.” http://www.youtube.com/watch?v=XYUnEOxU2LE&f eature=related Theory of the Action potential--Hodgkin http://www.afodor.net/HHModel.h tm Evidence: 1. The spike top = ~ +60 mV which is close to the equilibrium potential for Na+. 2. The spike top can be manipulated: …as in these experiments. Well this is nice… But this evidence is about one instant in time—the peak of the spike, but….. Hodgkin model is about the whole epoch: from resting potential to spike top, and back to baseline, and concerns in and out ionic movements, Na+ in and K+ out later, and causal permeability changes. In other words… The Hodgkin model is dynamic concerning changes during time. For this, one needs special methods, the voltage clamp or current injector. Remember this? Let’s add superscientist who monitors scope and injects equal & opposing current:This clamps voltage. Superscientists don’t exist…but Analog amplifiers do…… In other words… Even during a naturally occurring spike, the clamp keeps voltage constant by injecting ions as necessary. An Experiment with a voltage clamp or current injector This PROVES even the dynamic parts of Hodgkin-Huxley Model: During the action potential, there is an early inward flow which is Na+ rushing in. ….and a delayed outward flow which is K+ leaking back out and helping to restore resting potantial, aided by NA-K pump. Kymograph: NM preparation By the way….. A Nerve is a bundle of axons… Old Principles (from 1930s) explained: The Kymograph told us of “All or none” means either you got the whole spike (-70 to +60= 130) or none. All because Na+ goes to its equilibrium potential. “Refractory periods” harder to activate neuron = Na+ conductance inactivated. “Threshold”= when Inner wheel outruns outer wheel & brakes fail. The spike propagates undiminished (all or none) to end. How? What if you stimulate axon in middle? You get propagation in both directions. Orthodromic (natural) vs. Antidromic (experimental) conduction. To test where a cell body of a long axon is. Determinants of Conduction Speed Physical: temperature, diameter. Physiological: Nodes, Myelin. Continuous vs Saltatory Conduction Continuous vs Saltatory Conduction Strength-Duration Curve What happens if you depolarize, but don’t get a spike? You get: Cable properties: decay over time and distance. If you looked 3 times at one place: Why does it decay without a spike? And why doesn’t spike? Here’s a resting axon: We introduce a sub threshold stimulation. What happens? The positivity put in repels other positives up to membrane, which causes outside +’s to be repelled—Capacitative current. Capacitative flow means no actual physical movement.. Pretty soon, the inside of the membrane would have a full capacity of + charge. Meanwhile, more + charge has accumulated further down and acts like a depolarizer..BUT.. Meanwhile…. So the membrane has capacitance & resistance, as does the axoplasm..and we can represent membrane like this… Why is a spike undiminished and propagated? Because the initial and each subsequent change in potential is all the way to the Na+ equilibrium potential. Enough na+ rushes in at each segment to FULLY depolarize (to +60) the next segment. In living mammals, the usual way of activating a neuron… ….is not with stimulators causing threshold or subthreshold polarizations. It is with : Synaptic Transmission Some notes: Presynaptic terminals are boutons de terminaux or terminal buttons. Post synaptic structures are usually either the dendritic branches (“arborizations”) or the cell body itself. Pre synaptic spikes lead to release & flow of transmitter from terminals. There is both excitatory and inhibitory transmitter making depos & hyperpos, respectively, (EPSP, IPSP), at postsynaptic neurons. Postsynaptic EPSP and IPSP: Types of synapses: There is also ephaptic transmission, direct electrical activation of one neuron by another… These are called “tight junctions,” (or “gap junctions”) as they are only 40Ao apart. Ephapses that make spikes are mostly in invertebrates. But there are sub-threshold ephaptic influences in mammals. Better diagram of A-D, A-S Axosomatic micrographs Properties of synaptic information transmission (contrasted with spike propagation down axon): 1. S.T. is unidirectional in CNS because transmitter is presynaptic. (In axons, you can have ortho & antidromic, i.e., bi-directional information flow.) 2. You can have repetitive discharge in S.T. With presy stim., big flow of (pre-sy) transmitter can fire the post-synaptic cell repeatedly. (Unlike 1:1 spike on axon) 3. Related, in S.T.,there is not necessarily “frequency following.” (Yes on axon, if you stimulate the axon and record from axon.) More properties of synaptic information transmission (contrasted with spike propagation down axon): 4. Synapses have low safety factor. Means likely loss of information due to susceptibility to drop in 02 concentration, drugs, etc. in synapse (vs. axon). 5.Reasons for delay in information flow. In S.T. it is number of synapses. (In axon, it’s diameter, presence(absence) of myelin.) More properties of synaptic information transmission (contrasted with spike propagation down axon): 6. There can be either Inhibition or Excitation at synapse. (Only excitation for spike flow. All or none.) 7. Synaptic potentials are graded/additive. (Only all or none spike.) 8. Synaptic potentials are local like sub-threshold depos. (Spikes propagate undiminished.) 7. & 8. have major implication: There is no such thing as a unitary EPSP (or IPSP). You may hear a phrase like “the EPSP” (or “the IPSP”) but the size of EPSP or IPSP depends on amount of excitatory or inhibitory transmitter simultaneously reaching post synaptic membrane. This was most persuasively shown.. …by the great Australian neurophysiologist, Sir John Eccles of Canberra (later Buffalo). Eccles’ classic demonstrations of Spatial Summation: Temporal Summation So he actually demonstrated that….. …the EPSP size depends on the temporally and spatially integrated positive (EPSP, depolarizing) and negative (IPSP, hyperpolarizing) input. Same is true of IPSP…. Eccles did another brilliant demonstration: There was more: The second hump= spike could be further “dissected”: The inflection points (when 2nd derivative =0) required use of analog computers. The Initial Segment of the axon, also called the axon hillock, has a lower threshold than the rest of the axon, and it is a relay booster when necessary, which is good for safety factor. There is electron microscopic evidence that I.S. membrane has different structure (slightly) than main axon membrane. Mechanisms of PSPs: EPSP: Excitatory transmitter increases both Na+ and K+ permeabilities. It has an “equilibrium potential” ~ 0. This is, of course, a depolarization. Tends to cause a spike if big and widespread enough. Mechanisms of PSPs: IPSP: involves increase in permeability to CL- and especially, K+. It has an “equilibrium potential” ~ -89, which is hyperpolarizing away from -70. This puts the neuron away from -70, and harder to excite than when at rest = inhibition. The “dynamic duo from Deutchland”: Westecker & Deeke This was the inference implied by reduced EPSP, but has since been confirmed with release experiments. That is, they do the D&W clamp manipulation and look not only at EPSP size, but actual amount of transmitter released into synapse. Note… This clamp, at a more positive voltage level, is not really an excitatory process, though it does move neuron in depolarizing direction. The voltage is clamped, after all. Why is this important? Is there a parallel in nature? What were those axo-axonic synapses all about? But in nature, we have no voltage clamps, but we do have axoaxonic synapses: Note again…. On the axon, this is not a true excitatory process, since it will never lead to axon spike; it’s like a chemically mediated, clamp. The spike is generated earlier, at green arrow. Whither all-or none? Is this a violation of all or none?—the fact that base-to-peak spike size may be reduced? Not really. The spike still goes to the Na+ equilibrium potential, +66. That is the constant “all” of all or none. We now just learn that peak is constant, not necessarily base-peak. Is there a parallel pre-synaptic excitation? Pre-synaptic Excitation: Synaptic Transmission is Chemical. Evidence: 1) LOGICAL/INDIRECT: a) 2 kinds of PSPs exist and can be explained by 2 kinds of chemical transmitters. (Only 1 kind of pre-synaptic spike.) b) Delay in pathways are consistent with diffusion times, not with electrical conduction=186,000 mps. LOGICAL/INDIRECT: (Cont.) c) Post synaptic membrane is not electrically excitable, which axon is, but axon is not chemically excitable (except at axo-axonal synapses). Direct Evidence: Loewi’s Experiment OK, so what are the chemicals which are transmitters and where in CNS are they? McLennan’s 5 criteria The 5 criteria a candidate chemical must meet to be dubbed a transmitter were developed to parallel the processes that happen in synaptic transmission: 1. The first thing to happen is the synthesis of the transmitter in presynaptic neuron. Therefore, McLennan #1 is that presynaptic terminals should contain precursors and enzymes of biosynthesis. Next…. 2. After transmitter is made is must be mobilized (by spike)and diffuse into synapse. So McLennan # 2 is that presynaptic spikes must be followed by release of substance. 3. Then there is a chemical reaction with neurotransmitter receptor, leading to a PSP. So McLennan # 3 is that topical application of substance must produce PSP. The PSP now does its job: ex or inh. Now…. 4) The transmitter has arrived at the post synaptic side and must be degraded back to its precursors, so McLennan #4 is that the enzymes of degradation must be found in the post-synaptic cell. …By the way, for some transmitters, much of what is released is reabsorbed by presynaptic cell, so never gets there (across cleft) for degradation… The last criterion 5. McLennan #5: Consistent in vivo/in vitro effects, meaning if a substance is known in the test tube to block a process—like synthesis of Ach– then it must be shown to block Cholinergic processes in behaving animals. Example: The following reaction works if Choline Acetylase is active. Acetic Acid +Choline ---- ACH By the way…. Most candidate neurotransmitters have not met all 5 criteria and are therefore called “putative” neurotransmitters. Next we will talk about those that are pretty well established. These tend to be in the peripheral N.S. because that’s where it was easiest to dissect pathways. Things are tougher in the brain. 1. Cholinergics: Ach. Where is Ach found? 1. Neuromuscular junction (ala Loewi’s “Vagusstoffe”) 2. Specific loci in Autonomic Nervous System (ANS). (later) 3. Nucleus Basalis: Alzheimers? 4. Pontine nuclei regulating sleep. 5. Cerebral Cortex—putative. Carlson’s adrenergic pathways (The point of this slide is to show there are many ACH paths in brain.) 2. Catecholamines: a)Epinephrine (adrenalin), b)Norepinephrine (noradrenaline), c)Dopamine. Mixtures of a+b used to be called “Sympathin.” Norepinephine We will see it is found all over brain so has many roles in psychological processes, especially motivation, reinforcement, emotion. Specifically: 1. post-ganglionic terminals in Sympathetic N.S. and released from adrenal gland where it can act everywhere (“fight or flight.”) Norepinephine (cont.; The point of this slide is to show there are many ACH paths in brain.) 2. Dorsal ascending noradrenergic tract. 3. Ventral ascending noradrenergic tract. These tracts connect pontine nuclei of origin to virtually the whole forebrain Carlson’s version (Doesn’t show the true origin of VANB!!) Dopamine: 1. Afferent collaterals to reticular formation Dopamine 2. Nigro-Striate Path Indolamines a) Serotonin or 5hydroxy Tryptamine (5-HT). Found: 1. Descending raphe-spinal fibers going from nucleus raphe magnus in medulla to spinal cord. Inhibitory. 2.Other places I leave to your reading… The Raphe-Spinal System of analgesia By the way, both catecholamines and indolamines are Monamines Inhibitory amino acid Transmitters: GABA (Pretty exclusively inhibitory.) GLYCINE ( +/-) GLUTAMATE (+/-) Everywhere in mammals. For example, remember this? There are many other substances with varying “degrees of putativity.” Such as the peptide neuromodulators*, Enkephalins and Endorphins (synthetic opiates), Substance “P” Nitric Acid (NO) * These mediate synaptic transmission but not restricted to cleft. OK, so what are some drugs whose effects are mediated at synapses or axons? (A Minicourse in Psychopharmacology) Check web site. Brief Neuroanatomy. (Good idea to study each day’s lecture each day) NOTE: I will have the opportunity to sneak in an introduction to the topic of neural coding here and there. It’s a bit more elaborate.. One needs to fill in details of “CNS” By evolution-driven organization.. …we mean that the higher the animal, the greater the degree of encephalization: Sharks get only up to pons, fish up to midbrain, some reptiles have a tiny cortex with few myelinated fibers. This is also true for the inside out organization…see next slide: BrainStem inside-out organization… Cortical projection systems Early specific and late reticular evoked eeg potentials or event related potentials, which you get in response to any sensory stimulus— like that fingertip tap… Cortical Reticular Arousal : stimulate r.f. or any sensory path This change from synchronous (rhythmic) high amplitude, low frequency to arhythmic, low amplitude, high frequency is alpha blocking (or arousal of EEG) Slightly more realistic…. Other stuff & illustrative perspective: Side View of Brainstem with Cerebellum Brain Stem Auxilliary( NO= not to memorize) Visual System NO Another view (NO) My way Different view Auditory pathways NO Acoustic receptor cartoon (no) Somatic Sensation 1)Conscious Proprioception (aka dorsal or posterior column system) 2) Unconscious Proprioception (cerebellar—we’ll talk about later) 3) Pain & Itch a) below the neck b) above the neck NOTE: Here’s where I will also sneak in an introduction to the topic of neural coding along the way. Neural Coding means how the CNS represents sensory, motor, and psychological processes in any CNS activity: spike pattern, rate; brain wave frequencies, and derivatives (more later). Dorsal (posterior)Column Sense NO Neural coding: Mountcastle & associates showed that the firing rate of joint angle receptors is a logarithmic or power function of the joint angle and the speed of rotation to it: (These experiments were done on anaesthtized, sometimes curarized, always restrained cats & monkeys.) The question was: What part(s) of this complex system mediates facial pain? Top (main, oralis) or bottom(caudalis) middle (interpolaris) or all or two of the above? We are sneaking also here into parallel topic of neural (sensory) coding The predominant view: submodality segregation as in spinal cord Pain caudally (s. caudalis) and nonpain rostrally (main, oralis). (In cord, pain is in spinothalamic tract, ventral, and non-pain is in dorsal column.) This is a labelled line code: place signals sensation/perception. Clinical (surgical) evidence was supportive: caudal tractotomy relieved facial pain as in Tic Doloreaux (trigeminal neuralgia). But it should have been obvious that saggital knife cuts in TT would likely cut the nucleus,rostrally There was also a “within-line code” suggested by Khayyatt experiment. First accidentally motivated study: What is the effect of pure trigeminal nuclear lesions of rostral subnuclei (main, oralis) on orofacial pain? Motivated by our accidental finding that stimulation rostrally was very aversive. If stimulation activates pain, then lesions should remove pain. How to measure facial pain in a rat? Face rub latency, inversely. Knife cut experiment: How to measure non-pain stimulation of face? Remember this? Tested sites on the face: Facial Pain Results: Non-painful facial stimulation results: So we turned the Dubner hypothesis upside down… In the rat, the top is for pain and the bottom is for non pain. Vyklyky in Czeckoslavakia found similar result in cat: Lesions of caudalis did NOT prevent aversive conditioning with tooth pulp stimulus! (“Rosenfeld, Ha! Now there are two of us!”) Evidence for conscious and unconscious proprioception: 1. Spinocerebellar fibers go to cerebellum, not cortex—the substrate for consciousness. Unfortunately, evidence to contrary was found in ‘63. 2. Better Evidence: Coolest Evidence: Cortex Transition…… We are now entering the topic of neural coding formally. How does CNS activity encode sensory, motor, and Psychological (cognitive, emotional, perceptual etc) events? First we need to examine what the CNS events are, and that requires us to go to the next powerpoint on the web site: “EEG,ERPs, & relation to single neuronal activity.”