NOCICEPTORS

The Cells That Sense Pain

Alan Fein, Ph.D.

NOCICEPTORS: THE CELLS THAT SENSE PAIN

Alan Fein, Ph.D. Professor of Cell Biology University of Connecticut Health Center 263 Farmington Ave. Farmington, CT 06030-3505 Email: afein@neuron.uchc.edu Telephone: 860-679-2263 Fax: 860-679-1269

NOCICEPTORS: THE CELLS THAT SENSE PAIN CONTENTS Chapter 1: INTRODUCTION Chapter 2: IONIC PERMEABILITY AND THE RECEPTOR POTENTIAL

ION CHANNELS

SENSORY STIMULI Chapter 3: TRP CHANNELS

DISCOVERY OF TRP CHANNELS IN DROSOPHILA PHOTORECEPTORS

MAMMALIAN TRP CHANNELS

TASTE AND CHEMESTHESIS

CAPSAICIN

TRPV1 TRPV1 AS A THERAPEUTIC TARGET TRPV2 TRPV3 TRPV4 TRPA1 TRPM8 Chapter 4: CHEMICAL MEDIATORS

SEROTONIN

BRADYKININ 12-HPETE PGE2 PGI2 ATP P2X RECEPTORS P2Y RECEPTORS

PROTEINASE-ACTIVATED RECEPTORS

LOW pH

LYSOPHOSPHATIDIC ACID

NERVE GROWTH FACTOR Chapter 5: Na+, K+, Ca++ and HCN CHANNELS Na+ CHANNELS Nav1.8 Nav1.9 Nav 1.7 Nav 1.3 Nav 1.1 and Nav 1.6 K+ CHANNELS K2P channels ATP-sensitive K+ channels Outward K+ channels Ca++ CHANNELS HCN Channels Chapter 6: NEUROPATHIC PAIN

ANIMAL MODELS OF NERVE INJURY USED TO STUDY NEUROPATHIC PAIN TWO EXAMPLES OF NEUROPATHIC PAIN CAUSALGIA TRIGEMINAL NEURALGIA SOME EXAMPLES OF THE DIVERSITY OF THE TREATMENTS FOR NEUROPATHIC PAIN GABAPENTIN ARTEMIN CANNABINOIDS HCN CHANNELS WHAT IS THE ROLE OF SPONTANEOUS ACTIVITY BOTULINUM TOXIN TYPE A DEMYELINATION, LYSOLECITHIN AND LYSOPHOSPHATIDIC ACID

Chapter 1

INTRODUCTION

Pain is an unpleasant feeling that is an essential component of the bodys defense system. It provides a rapid warning to the nervous system to initiate a motor response to minimize physical harm. Lack of the ability to experience pain, as in the rare condition congenital insensitivity to pain with anhidrosis (Axelrod and Hilz 2003), can cause very serious health problems such as self-mutilation, auto-amputation, and corneal scarring. Up until the twentieth century there was a vigorous and heated debate about the nature of pain. One side held that sensory stimuli, which activate ordinary sense organs, such as those for touch, would initiate pain through the same sense organs if the stimuli were strong enough. The other held that there were a separate set of specialized sense organs specific for pain. It was not until the twentieth century that the debate was settled and it was shown conclusively that there were specialized sensory organs that signaled pain. The word pain comes from the Greek: poin, meaning penalty. Physiologists distinguish between pain and nociception; where nociception refers to signals arriving in the central nervous system resulting from activation of specialized sensory receptors called nociceptors that provide information about tissue damage. Pain then is the unpleasant emotional experience that usually accompanies nociception. The focus of this book is on the nociceptors the specialized sensory receptors that provide information about tissue damage. Historically, to learn something about the stimuli that activate nociceptors large numbers of randomly selected nerve fibers that innervate the skin were typically studied. Large peripheral nerves in mammals are actually compound nerves composed of bundles of thousands of individual nerve fibers enclosed in a loose connective tissue sheath. The conduction velocity with which the individual nerve fibers within a bundle transmit action potentials to and from the nervous system can vary more than 100-fold, making it of interest to know the conduction velocity of the fibers that carry the signal from nociceptors to the brain. The electrical activity of an individual nerve fiber from a nerve bundle can be isolated and recorded from using a variety of methods, one of which is shown in Figure 1-1. In the example given, an intracellular electrode was used to impale the cell body of a sensory neuron in the dorsal root ganglion (DRG) and thereby record its electrical activity. The DRG are comprised of the cell bodies of sensory neurons, and are located lateral to the spinal cord in the vertebral column. These sensory neurons have an axon that projects to peripheral tissues, such as the skin, and are responsible for our sensation of our bodies. The trigeminal ganglion is analogous to the dorsal root ganglia of the spinal cord and is responsible for sensation in the face. The conduction velocity of the impaled neuron in Figure 1-1 was measured by using a brief voltage pulse applied to the extracellular stimulating electrodes to evoke action potentials in the nerve fibers composing the nerve bundle. By knowing the distance from the stimulating electrodes to the recording site, and the time it takes the action potential to reach the recording site following application of the voltage pulse, the conduction velocity can easily be calculated. Many of the afferent (sensory) neurons isolated in this way respond to lowintensity mechanical or thermal stimulation, that is, stimuli that in individuals evoke an

1-1

spinal cord

dorsal root

intracellular recording electrode dorsal root ganglion

ventral root

Lamina I

IB4IB4+ Lamina II

extracellular stimulating electrodes

Figure 1-1. Intense heat from a fire activates the terminals of two nociceptors. Action potentials are propagated along the axons of the nociceptors into the spinal cord and the activity of one of the nociceptors is monitored by an intracellular electrode which impales its cell body which is located in the dorsal root ganglion (DRG). The central terminal of a fiber staining positive for the plant lectin isolectin B4 ( IB4+) is shown terminating in lamina II and that of an IB4- fiber is shown terminating in lamina I. The extracellular stimulating electrodes are connected to a pulse stimulator (not shown) and are used to initiate action potentials in the nerve fibers.

1-2

innocuous or non-painful sensation. In addition, these fibers exhibit the full range of conduction velocities exhibited by the nerve. Some of the neurons recorded this way are distinguished by relatively high thresholds for activation, i.e. they can only be activated by intense (mechanical, thermal or chemical irritant) stimuli that are potentially damaging to tissues. These high threshold neurons are thought to be the primary afferent nociceptors. We have all probably experienced that pain can be caused by thermal, mechanical and chemical stimuli that produce tissue injury. Several possibilities might explain how these different stimuli could result in the sensation of pain. One possibility is that individual nociceptors are sensitive to all of these different stimuli. Another is that there are several different types of nociceptors with each being sensitive to a specific stimulus. As we shall see below it turns out that both possibilities are found in nature: some nociceptors are sensitive to a specific stimulus while others are sensitive to multiple types of stimuli. The nerve fibers (axons) within a compound nerve include both afferent nerves and efferent (motor and autonomic) nerves. The speed at which an individual nerve fiber conducts action potentials is related to the diameter of the fiber. In the larger myelinated fibers, the conduction velocity in meters per second is to a first approximation six times the axon diameter given in microns (see Figure 1-2). The histogram of the distribution of conduction velocities has four peaks: the slowest conducting fibers are unmyelinated and designated C; the faster conducting myelinated fibers are designated A, A and A. The widely held view that is presented in most present day textbooks is that only the smallest diameter and slowest conducting nerve fibers the C- and A-fibers carry the afferent signal from nociceptors that is perceived as pain. Never the less the available evidence, which has been thoroughly reviewed (Lawson 2002; Djouhri and Lawson 2004), suggests that a substantial fraction of the A-fiber nociceptors may conduct in the A conduction velocity range. Hence, to allow for this possibility, the designation used here is that the signal from nociceptors is carried by unmyelinated C-fibers and myelinated Afibers conducting in the A(-) conduction velocity range. It should be kept in mind that the reverse is not true, not all C-fibers and A(-) fibers are nociceptors. The C and A() fibers also carry signals for non noxious innocuous mechanical, warm and cold stimuli. Because of the difference in conduction velocity between the C and the A(-) fibers, the signal from the A(-) fibers arrives at the spinal cord before that from the C-fibers. This raises the possibility that painful stimuli evoke two successive and possibly distinct painful sensations. The evidence supporting the view that C and A(-) fibers signal distinct painful sensations comes from experimental conditions (electrical stimulation and nerve block) where the activity of the A- and C- fibers are studied in isolation. When this is done stimulation of the A-fibers is described as causing a sharp pricking pain sensation and that of the C-fibers a dull, aching burning pain. It is usually stated that for painful stimuli there is a biphasic subjective response: a short-latency pricking pain is followed by a second long latency pain of a burning and less bearable quality. However, the evidence for two successive painful sensations is much less compelling than it is for two distinct painful sensations. In the original report showing that C and A(-) fibers signal distinct painful sensations, it was stated that such a biphasic subjective response to a 1-3

C
number of fibers

Ad Ab Aa

4 24

8 48

12 72

m 16 m 96 m/sec

Figure 1-2. Axon diameters and conduction velocities in a peripheral nerve. Axon diameters are given in micrometers and conduction velocities are given in meters per second. The fibers designated with a C are unmyelinated and those with an A have a myelin coat.

1-4

single transient painful stimulus is often absent in normal subjects (Bishop, Landau et al. 1958). The inability of many normal subjects to experience a first and second pain from one stimulus to the skin surface should not be taken to imply that these two types of pain are artifacts of the experimental conditions under which they were observed. Rather when both are activated simultaneously under normal conditions it is difficult for each to be identified by the observer. Four classes of nociceptors: mechanical, thermal, polymodal, and silent, have been described. Mechanical nociceptors respond to intense pressure while thermal nociceptors respond to extreme hot or cold temperatures (>45C or <5C) and both have myelinated A fibers that conduct impulses at a velocity of 3 to 40 m/s. Collectively, these 2 types of nociceptors are called A(- ) mechano-thermal nociceptors. Polymodal nociceptors respond to noxious mechanical, thermal, and chemical stimuli and have small, unmyelinated C fibers that conduct impulses at a velocity of less than 3 m/s. Remember that the small, myelinated A(- ) fibers carry the nociceptive input responsible for the sharp pricking pain and the small, unmyelinated C fibers carry the nociceptive input responsible for the dull burning pain. Silent nociceptors are activated by chemical stimuli (inflammatory mediators) and respond to mechanical and thermal stimuli only after they have been activated. These nociceptors also have small, unmyelinated C fibers that conduct impulses at a velocity of less than 3 m/s. Nociceptors that respond to noxious temperatures can be divided into those that are unimodal, which are activated by a unique thermal stimulus, and those that are polymodal, which detect painful chemical, mechanical, and thermal stimuli. The basic function of nociceptors is to transmit information to higher-order neurons about tissue damage. Individual receptors can be regarded as an engineers black-box, which transforms tissue damage into an appropriate signal for successive nerve cells. The ultimate function of a nociceptor could be fully described if its input-output relationship alone were given. Here input, of course refers to tissue damage. What about output? One of the central concepts of neurobiology holds that neurons communicate with each other via synapses. The most commonly encountered synapses release chemicals, known as synaptic transmitters. It is by releasing these transmitters that one cell is able to communicate with its postsynaptic neighbors. Because nociceptors are neurons with chemical synapses, their output is encoded in the release of their neurotransmitters: the input-output relationship is simply a conversion of tissue damage into transmitter release. Direct measurement of synaptic transmitter release under physiological conditions is very difficult and has not been accomplished for any nociceptor. It would thus seem that a derivation of the input-output relationship is beyond reach. However, another nearly universal neural property is of assistance: transmitter release is directly controlled by synaptic membrane potential. Therefore, by recording the variation of the membrane potential at the synapse, the nociceptor output could be indirectly surmised. Unfortunately, in most cases, it is technically difficult, if not impossible to record intracellularly from a synaptic terminal. The vast majority of electrophysiological recordings have been carried out on other regions of the cell because these regions are 1-5

more accessible. Electrical activity in nociceptors as in most neurons is associated with propagating action potentials, which occur on a time scale of milliseconds. These action potentials propagate to the synaptic terminal and thereby regulate transmitter release. Two recording techniques are typically used to record nociceptor action potentials: either extracellular electrodes record their occurrence somewhere along the nociceptor axon or they are recorded intracellularly from the nociceptor cell body as illustrated in Figure 1-1. Thus, sensory transduction for nociceptors is typically measured as the conversion of tissue damage into the patterned firing of action potentials. During the past century, the basic framework of sensory transduction for different senses and for many species of vertebrates and invertebrates was established. The typical sensory cell was shown to have a specialized region where sensory receptor molecules detect the stimulus, which for nociceptors is tissue damage. The sensory stimulus causes a conformational change in the receptor molecule, which triggers the transduction process that brings about a change in the membrane potential of the receptor cell. The resulting change in membrane potential is called the receptor potential. In the typical sensory neuron, the part of the cell where sensory transduction takes place is often distant from the synaptic terminal. Therefore, the receptor potential needs to be converted into a series of propagating action potentials, which in turn carry the signal along the axon to the synapse. Unstimulated nociceptors typically fire few or no action potentials, and their response to tissue damage is an increase in the rate of firing of propagating action potentials along the cells axon. These findings are summarized in Figure 1-3, which shows the four most significant regions of an idealized nociceptor, the sensory transduction region, the axon, the cell body, and the synaptic terminal. In parallel with the anatomic regions of a nociceptor shown in Figure 1-3, one can describe a functional scheme for the mechanism of operation of a nociceptor as in Figure 1-4. The cell body, axon and presynaptic terminal of nociceptors should function more or less as they do in other cells. The arrows going from the cell body to the other regions of the nociceptor are meant to indicate that the cell body is necessary to maintain the other regions of the cell, without the cell body the cell would eventually die. The axon conducts the action potential to the synaptic terminal where the transmitter(s) are released. The release of transmitter at the synaptic terminal is subject to modulation by agents released by other neurons and possibly glial cells. Stimuli that cause tissue injury may activate the sensory transduction region of nociceptors either directly or indirectly. For example, a nociceptor may contain heat sensitive receptor molecules in the plasma membrane that respond directly to a damaging heat stimulus or conversely tissue damage by the stimulus might result in the production of a factor that in turn activates the nociceptor. A third possibility is that, because of the injury the tissue becomes inflamed and an extrinsic factor, that activates the nociceptor, enters the damaged region. The box labeled perireceptor events in Figure 1-4 allows for the production, during tissue injury, of factors, which might activate or modulate the nociceptor. Intense noxious stimuli often lead to an increase in the response to subsequent painful stimuli (gain control in Figure 14). After the noxious stimulus is transduced into a receptor potential the response must be transformed or encoded into a series of action potentials, which carry the signal to the synaptic terminal. The current prevailing view is that free nerve endings of A(- ) and C 1-6

receptor potential

stimulus

cell body

presynaptic terminal

sensory transduction region

axon

Figure 1-3. Schematic drawing of a nociceptor showing the four regions of the cell.

1-7

Perireceptor events

Transduction

Gain control

sensory transduction region

Encode

cell body Transmit signal

Axon

Transmitter release

Presynaptic terminal

Modulate transmitter release

Figure 1-4 Functional schematic of nociceptor operation

1-8

fibers constitute the sensory region of nociceptors. There are no specialized structures associated with the nerve endings as there are for other sensory receptors such as mechanoreceptors. Many other sensory receptors can be isolated functionally intact, and the study of isolated photoreceptors, auditory receptors, olfactory receptors, etc. has told us a great deal about their mechanisms of operation. Ideally, to study the pathways involved in the transduction process of nociceptors, a preparation of isolated nociceptors is needed. The free nerve endings of nociceptors are extremely fine and are embedded in a tissue matrix, which if it were to be dissected, to isolate the nociceptors, would release the very molecules that the nociceptor nerve terminal is meant to detect. Because of this lack of accessibility, it is not possible to directly study the nociceptor transduction machinery both in an unstimulated state and in its normal native environment. It should be kept in mind that not all free nerve endings represent the sensory transduction region of nociceptors. Free nerve endings are also responsible for detecting temperature, mechanical stimuli (such as pressure), and information about touch. Because it is not possible to isolate the nociceptors sensory nerve endings in an unstimulated state, studies on isolated nociceptors are often carried out on the cell bodies of nociceptors. For example, the neuronal cell bodies of a dorsal root ganglion are isolated by enzymatic treatment and are cultured before use. The sensory endings are completely removed during the isolation procedure, and it is hoped or assumed that the properties of those terminals are recreated in the cultured cell bodies. In as much as the original ganglion contained more than just nociceptors, only a fraction of the cultured cell bodies will actually be those of nociceptors. This preparation of cultured cell bodies is often used for experiments investigating the cellular and molecular basis of detection of painful stimuli. The uncertainties and assumptions associated with these procedures make it essential that the findings be checked very carefully and shown to resemble what actually occurs in vivo. As mentioned above unstimulated nociceptors typically fire few or no action potentials, and their response to tissue damage is an increase in the rate of firing of action potentials. Since it takes a membrane depolarization to cause an increase in the rate of firing this finding implies that the receptor potential of nociceptors is a membrane depolarization, as shown in Figure 1-3. Thus, the transduction machinery in Figure 1-3, by necessity, must somehow gate ion channels or carriers that can depolarize the plasma membrane of the transduction region. The encoding region (see Figure 1-4) in turn converts the membrane depolarization into an increase in the rate of firing of action potentials. The properties of nociceptors considered so far were elucidated primarily from studies of uninjured tissue. However, intense noxious stimuli resulting in tissue damage often lead to an increase in the response to subsequent painful stimuli, called hyperalgesia, that is, an excessive sensitiveness or sensibility to pain. Hyperalgesia comprises both primary hyperalgesia, an increased sensitivity within the injured area predominantly due to peripheral nociceptor sensitization, and secondary hyperalgesia, an increased sensitivity in the surrounding uninjured area mediated centrally. Those of us that have been injured 1-9

3 normalized response amplitude

2 after

before
40 42 44 46 48 50 stimulus temperature (C)

Figure 1-5. Sensitization of a thermal nociceptor to stimuli that heated an area of the skin to the temperature indicated. (= responses ) Individual to thermal stimuli obtained before the area of skin was burned. (t ) Sensitized responses obtained from the same area after the burn injury.

1-10

can probably remember having experienced this hypersensitivity to pain at the site of injury and the surrounding region. For example, the inflammation due to a sore throat can be so bad that the mere act of swallowing is painful. No description of the properties of nociceptors would be complete without a consideration of nociceptor sensitization resulting from tissue injury. Sensitization is a leftward shift (that is toward lower intensities) of the stimulus-response curve, which relates the magnitude of the neural response to the stimulus intensity. As shown in Figure 1-5, sensitization of a nociceptor is characterized as a decrease in threshold and an increase in the magnitude the response to suprathreshold stimuli. Remember that some but not all nociceptors exhibit sensitization. References cited:

Axelrod, F. B. and M. J. Hilz (2003). "Inherited autonomic neuropathies." Semin Neurol 23(4): 381-90. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15088259 Bishop, G. H., W. M. Landau, et al. (1958). "Evidence for a double peripheral pathway for pain." Science 128(3326): 712-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=13580241 Djouhri, L. and S. N. Lawson (2004). "Abeta-fiber nociceptive primary afferent neurons: a review of incidence and properties in relation to other afferent A-fiber neurons in mammals." Brain Res Brain Res Rev 46(2): 131-45. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15464202 Lawson, S. N. (2002). "Phenotype and function of somatic primary afferent nociceptive neurones with C-, Adelta- or Aalpha/beta-fibres." Exp Physiol 87(2): 239-44. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11856969

1-11

Chapter 2 IONIC PERMEABILITY AND SENSORY TRANSDUCTION In trying to understand nociceptor sensory transduction, it has proven to be instructive to consider the molecular mechanisms utilized for signaling elsewhere in the nervous system. Mechanisms used by the other senses as well as those used for synaptic transmission have turned out to be most useful. Numerous studies have shown that both the sensory receptor potential and the synaptic potential are the result of changes in the ionic permeability of the plasma membrane. That is, they are the result of ions flowing through integral membrane proteins called ion channels. Before considering how the opening and closing of ion channels is regulated during nociceptor signal transduction, it is important to understand the forces that cause ions to flow through these channels. Basically there are two types of forces that drive ionic movement across cell membranes. There is the electric field across the cell membrane which is manifest as an electrical potential across the membrane and there is the concentration gradient for each ionic species. Most cells of the body, including neurons, maintain their cytoplasm at a negative potential with respect to the outside of the cell. For a cell that is in an unstimulated resting state, scientists typically use the term resting potential in referring to this negative potential across the cells plasma membrane. Generally speaking, the resting potential depends on the concentration of ions on the two sides of the plasma membrane and the resting permeability of the cell to these ions. Neglecting the small but not unimportant contribution of other ions we focus our attention on sodium (Na) and potassium (K). If the membrane were exclusively permeable to K, the membrane potential (VM) would be given by VM = EK = (RT/F)ln[(K)o/(K)i] (Equation 2-1)

Where VM is the membrane potential, inside minus outside, EK is the potassium equilibrium potential, (K)o the extracellular potassium activity, (K)i the intracellular potassium activity, R the universal gas constant, T the absolute temperature and F the Faraday constant. Similarly, if the membrane were exclusively permeable to Na, the membrane potential would be given by VM = ENa = (RT/F)ln[(Na)o/(Na)i] (Equation 2-2)

where ENa is the sodium equilibrium potential, (Na)o the extracellular sodium activity, (Na)i the intracellular sodium activity. As shown in Figure 2-1, for most cells including neurons (Na)o is much higher than (Na)i and (K)i is much higher than (K)o. In general, at rest, biological membranes and neuronal membranes in particular are permeable to both Na and K and therefore their resting potential lies somewhere between ENa and EK, the proximity to either of these equilibrium potentials depending on the relative permeability to Na and K. For the typical cell shown in Figure 2-1, VM must therefore lie some where between ENa 65 mV and EK -85 mV. Cells are typically much more permeable to K at rest than Na; therefore the resting potential is always inside

2- 1

Na+ = 140 mM K+ = 5 mM Cl- = 105 mM

outside inside

Na+ = 10 mM K+ = 150 mM Cl- = 10 mM

ENa 65 mV EK -85 mV
Figure 2-1. Extracellular and intracellular concentrations of K , Na and Cl for a typical cell.
+ + -

2- 2

negative and can vary from about -30 mV up to about -80 mV, depending on the degree of permeability to Na. Whenever the membrane potential lies between ENa and EK, Na will tend to leak into and K will tend to leak out of the cell. Unless the cell compensates for the constant loss of K and gain of Na, the ionic concentration gradients will run down, ENa and EK will decrease toward zero, and the membrane potential will disappear. Cells have metabolically dependent enzymes, called pumps, which compensate for this passive leakage by pumping K into and Na out of the cell. This molecule (Na+/K+ ATPase) is a Na-K pump which uses ATP to catalyze the movement of 3 Na ions out of the cell for every two K ions moving inward. There is a net extrusion of one positive charge out of the cell for each cycle of the pump, thus the pump is electrogenic. It is generally agreed that the pump does not directly participate in the generation of electrical signals but rather has its primary effect by maintaining the ionic concentration gradients for Na and K across the cell membrane. Calcium ions have also been found to play an important role in synaptic transmission and sensory transduction and cells have both calcium pumps and exchangers which keep calcium inside the cell, at a much lower concentration, than outside.

ION CHANNELS Molecular biology has provided us with a basic understanding of the relationship between the structure and function of ion channels in general. Channel proteins have amino acid sequences that extend across the lipid bilayer of the plasma membrane from the inside to the outside of the cell. They contain a specialized region called the P- or pore region which forms a channel or pore, that provides a path through which ions such as Na+, K+, Ca2+, and Cl- can pass through the membrane. The salient feature of the ion channels that underlie the receptor potential and the synaptic potential is that they undergo a transition from a closed to an open state (see Figure 2-2) that is regulated or gated by changes to the channel that result from the sensory stimulus or the synaptic transmitter. Two well understood mechanisms used to gate these channels are shown in Figures 2-3 A & B. For synaptic transmission (Figure 2-3A) the synaptic transmitter (i.e. ligand) binds to extracellular sites on the ion channel in the post synaptic membrane and gates it open. These ligand-gated channels are also sometimes referred to as ionotropic channels. For many sensory stimuli an intracellular second messenger, generated by the sensory transduction process (Figure 2-3B) gates the channel open. For the purposes of simplicity, the channel shown in Figure 2-3 A is shown with two external binding sites and that in Figure 2-3B with two internal binding sites, although in nature channels often have more than two binding sites. Channels are typically not the property of a single protein molecule, but rather are the result of the noncovalent binding of several subunits facing one another to form the pore region. Channels can be either homomeric, in which all the subunits are identical, or heteromeric, that is having non identical subunits with different properties. The pore region can be selective for either, Na+, K+, Ca2+, and Cl-. Additionally some channels are

2- 3

closed outside
inside

open

inside

Figure 2-2. Transition of a membrane channel pore region from the closed to the open state.

2- 4

agonist binding

second messenger binding

+ + ++

+ + + +

Figure 2-3. Several mechanisms of ion channel gating, (A) Binding of and extracellular agonist gates the channel open. (B) Binding of an intracellular second messenger gates the channel open. (C) Charge movement within the channel protein due to membrane depolarization gates the channel open.

2- 5

+ + + +

+ + + +

voltage gating

found to allow all three cations (Na+, K+ and Ca2+) to pass through their pore region, such channels are referred to nonselective cation channels. The majority of ion channels have three, four, or five subunits, arranged in circular symmetry, forming a single aqueous pore at the axial intersection (see Figure 2-4A for an example with four subunits). And as shown in the figure each subunit has only a single pore domain. In contrast, as shown in Figure 2-4B, K+ leak channels are composed of two pore domain K+ channels (K2P channels). Potassium leak channels are essential to neuromuscular function because they are responsible for cells being more permeable to K+ at rest than to Na+, they typically stabilize the cells membrane potential at hyperpolarized voltages below the firing threshold of nerves and muscles. In these channels, as shown in Figure 2-4B, each subunit has two pore domains arranged in tandem. Except for those cases where binding of the ligand to the ionotropic channel actually decreases the permeability of the channel, and decreases membrane conductance, the transmitter usually opens the channel, allowing ions to flow through it, thus increasing the conductance of the cells membrane for ions. The response to the ligand turns off when the ligand unbinds and diffuses away (or is broken down), the channel then shifts back to its closed conformation. Surprisingly, molecular biology has revealed a multiplicity of genes for ionotropic receptors that appear to have essentially identical functions. For example, the nicotinic acetylcholine receptor found in neurons which typically has five subunits (it is pentameric), consisting of only two types of subunits, alpha and beta (2 alphas and 3 betas). It turns out that there are at least 8 genes that encode alpha subunits and 4 that encode beta. Thus there are a large number of different possible combinations of alpha and beta subunits in one animal, the function of which is not understood. The tacit assumption is that these different genes evolved because they sub serve different functions. Two obvious possibilities are that they have different affinities for acetylcholine and therefore open at different concentrations or they have slightly different ionic permeability properties. Unlike the ionotropic receptors, where the receptor and the channel are the same molecule, the receptor molecule for the metabotropic receptor gates the channel indirectly, that is the receptor is a separate molecule from the ion channel that underlies the receptor potential. The metabotropic receptors can be classified into two types: the Gprotein-coupled receptors (GPCRs) and receptor tyrosine kinases. The family of GPCRs are coupled to an effector molecule via a guanosine nucleotide-binding protein (a Gprotein), hence their name. Activation of the effector component typically requires the participation of several other proteins in addition to the G-protein. Usually the effector molecule is an enzyme that produces a diffusible second messenger, for example, cyclic adenosine monophosphate (cAMP), cyclic guanosine monophosphate (cGMP), arachidonic acid, diacylglycerol, or an inositol polyphosphate. These second messengers can either directly gate the ion channel (see Figure 2-3B) or can trigger a further biochemical cascade. For example the second messenger might mobilize calcium ions from intracellular stores and the elevated intracellular calcium might directly gate an ion channel. Another possibility is that the second messenger activates specific protein

2- 6

A
1 2 3 4 5 P 6

3 4 5

2 6

B
2 3 1 P1 2 3 P2 4 P2 4 P1 1

Figure 2-4. Schematic illustration of the structure of ion channels with one pore domain (A) or two pore domains (B).

2- 7

kinases (phosphate transferring proteins) that phosphorylate the ion channel or other cellular proteins (thereby altering their activities) and either initiating or modulating the receptor potential. In some instances, the G protein of the second messenger can act directly on an ion channel. The receptor tyrosine kinases might gate the ion channel directly or indirectly via phosphorylation, that is they transfer a phosphate group to the channel or other cellular proteins. The channels found in both the encoding region (see Figure 1-4) and the axon, which convert the receptor potential into a train of propagating action potentials are gated by membrane depolarization (see Figure 2-3C). Both voltage gated Na+ and K+ channels play an important role in the generation and propagation of action potentials. The voltage gated Ca2+ channels play an important role at the presynaptic terminal where they function in the release of the synaptic transmitter. SENSORY STIMULI Before delving into the molecular mechanisms underlying nociceptor signal transduction, it is helpful to briefly consider the types of stimuli that occur during tissue damage, with an emphasis on stimuli that one might reasonably assume to be involved in signal transduction. First there are the stimuli themselves, such as mechanical tissue deformation, and either increases or decreases in tissue temperature. These stimuli might directly regulate ion channels (see Figure 2-2) in the nociceptor plasma membrane thereby giving rise to the receptor potential. Next there are the local changes in the extracellular milieu resulting from release and or exposure of molecules from the damaged tissue. That is, molecules normally found either inside cells or in the cell membrane might now be found in or exposed to the extracellular space where they can bind to receptors in the plasma of the nociceptor. Finally there are the molecules that enter the damaged region, as part of the bodies inflammatory response to injury, where they can bind to receptors in the nociceptor plasma membrane. Any of these three possibilities might reasonably be expected to participate in nociceptor signal transduction. Ideally one would like to identify the specific role in nociceptor signal transduction if any, of all the substances that appear in damaged tissue during painful stimuli. However as pointed out in Chapter 1, it is not possible to isolate the nociceptors sensory nerve endings in an unstimulated state and study how they respond to painful stimuli. Rather, as we shall see one is forced to use indirect methods. For example, the neuronal cell bodies of a dorsal root ganglion are often used after being isolated and cultured.

2- 8

Chapter 3 TRP CHANNELS The TRP channel family is of interest because several members have been implicated in nociceptor signal transduction. I have strived to limit the consideration of TRP channels to their role in nociception resulting in a superficial consideration of all their known properties. Discovery of TRP channels in Drosophila photoreceptors The trp mutant was originally isolated from Drosophila photoreceptors in which the light response decayed to baseline during prolonged illumination, hence the name transient receptor potential (see Figure 3-1). The gene was found to encode a Ca2+-selective channel responsible for the major component of the light response. Two other members of the TRP family of proteins (TRPL and TRP) were subsequently found for Drosophila photoreceptors, which are believed to be responsible for the residual light response in the trp mutant. The Drosophila phototransduction cascade is initiated when light strikes rhodopsin, which leads to stimulation of a heterotrimeric G protein that activates PLC. It appears that channel activation, which leads to an influx of cations into the photoreceptor, is mediated in vivo by DAG or its metabolites, polyunsaturated fatty acids (PUFAs). Using the terminology discussed in chapter 2, the TRP family in Drosophila are second messenger gated ion channels activated downstream from the G-protein coupled receptor (GPCR) rhodopsin. Mammalian TRP channels Based on sequence homology numerous members of the TRP channel family have been identified in vertebrates; the mammalian members of this family have been classified into 6 subfamilies: TRPC (Canonical), TRPV (Vanilloid), TRPM (Melastatin), TRPP (Polycystin), TRPML (Mucolipin) and TRPA (Ankyrin). Mammalian TRP channels are permeable to cations, and have 6 transmembrane domains flanked by intracellular N and C-terminal regions. Four subunits are thought to assemble as homo-and/or heterotetramers to form functional channels. Although TRP channels may be weakly voltage-dependent they lack the voltage sensor of voltage-gated channels (see Figure 23C). Taste and Chemesthesis The sense of taste (gustation) is the ability to perceive the flavor of substances such as food. Taste sensations include sweet, salty, sour, bitter and umami (savory). Umami is the taste that occurs when foods with the amino acid glutamate are eaten. In contrast to the sense of taste, the trigeminal nerve conveys information about irritating and noxious molecules that come into contact with the mouth. Chemesthetic sensations are defined as those that occur anywhere in the body when chemicals activate receptors for other senses. Thus the sensations transmitted to the brain when pain fibers of the trigeminal nerve are activated by noxious molecules would be described as chemesthetic sensations. The burn

3-1

trp

light stimulus

Figure 3-1 Comparison of the time course of the Drosophila photoreceptor light response (receptor potentilal) for the wild type and the trp mutant. Note that the trp mutant light response decays to baseline during the light stimulus, hence the name transient receptor potential, while the wild type photoreceptor exhibits a maintained depolarization during the stimulus.

3-2

50 mV

wild type

from chili pepper and the cooling from the menthol in mouthwash are examples of chemesthesis. Capsaicin Capsaicin (8-methyl-N-vanillyl-6-nonenamide) the pungent ingredient of hot chili peppers that gives them their burning sensation or piquancy was first isolated in the nineteenth century. Interest in the sensory effects of capsaicin has a very long history. Christopher Columbus described the eating of chili peppers by natives in the New World more than 500 years ago and Wilbur Scoville developed a test and a scale in 1912 to measure the hotness or piquancy of chili peppers, and the Scoville scale is still in use today. The effects of capsaicin are best understood in terms of its excitatory and desensitizing, actions on polymodal nociceptors. Electrophysiological studies reveal that capsaicin depolarizes DRG neurons (see Figure 1-1) and decreases their input resistance in a concentration dependent manner, suggesting that the specific excitatory effect of capsaicin on nociceptive neurons involves an increase in membrane permeability to ions such as sodium and/or calcium. Subsequent studies showed that the ionic permeability pathway discriminated poorly between cations, with divalent cations being relatively more permeable than monovalent cations. The discovery of resiniferatoxin an ultra potent capsaicin analog that mimics the cellular actions of capsaicin, and of the potent capsaicin antagonist, capsazepine, strongly suggested the existence of a specific capsaicin receptor. TRPV1 The capsaicin receptor (TRPV1) was cloned using a calcium influx assay of non-neuronal cells transfected with cDNA constructed from dorsal root ganglia RNA (Caterina, Schumacher et al. 1997). Electrophysiological analysis proved that the cloned receptor was similar to the native capsaicin-receptor of sensory neurons in several ways. Capsaicin-evoked currents were reversible upon ligand removal and lower concentrations of resiniferatoxin evoked maximal responses that persisted after ligand removal. The activation curves for capsaicin-currents from both native channels and the cloned receptor showed Hill coefficients of 2 suggesting the existence of more than one capsaicin binding site. TRPV1 channels are not only activated by capsaicin but also by elevated temperatures and by protons at pH below 6.5, confirming earlier studies showing that currents evoked by heat, low pH and capsaicin were commonly found in the same sensory neuron. Interestingly, TRPV1 is the only member of the TRPV channel family that is activated by capsaicin: knockout of TRPV1 in man and mouse result in capsaicin insensitivity (Caterina, Leffler et al. 2000; Park, Lee et al. 2007). We will consider the effects of low pH on TRPV1 along with its effect on other ion channels in a later chapter. Earlier studies showed that heat-evoked and capsaicin-evoked currents were commonly found in the same sensory neuron. This raised the possibility that TRPV1 was a temperature detector that enabled thermal nociceptors to respond to a range of hot temperatures. If this were so then in animals in which TRPV1 was knocked out one

3-3

would expect a deficit in responding to those hot temperatures that activate TRPV1. Paradoxically mice in which TRPV1 is knocked out exhibited deficits in their response to hot temperatures above 50C whereas TRPV1 is activated at temperatures at or above 42C. Although these animals were impaired in the detection of painful heat it is clear that either they normally use an alternative mechanism, other than TRPV1, to detect noxious heat at temperatures below 50C, or they have a backup mechanism. Remember that intense noxious stimuli resulting in tissue damage often lead to an increase in the response to subsequent painful stimuli, referred to as hyperalgesia, and that primary hyperalgesia is due to peripheral nociceptor sensitization or hypersensitivity. Interestingly, TRPV1 knockout mice exhibited little thermal hypersensitivity in the setting of tissue inflammation whereas wild type mice exhibited normal hypersensitivity. We will return to this finding in a later chapter when we consider the mechanisms of hyperalgesia. We are all probably familiar with the burning pain produced by the application of alcoholic tinctures, such as tincture of iodine, to skin wounds. The burning sensation, raises the possibility that ethanol might be activating TRPV1. To test this idea the effect of ethanol on isolated neurons from the trigeminal or dorsal root ganglia as well as TRPV1-expressing HEK293 cells was investigated (Trevisani, Smart et al. 2002). It was found that ethanol activated TRPV1 and potentiated responses to capsaicin and other activators of TRPV1; supporting the notion that alcohol causes a burning sensation by activating TRPV1. The uncertainties and assumptions associated with using isolated and cultured trigeminal and DRG neurons as well as cells made to express TRPV1 make it essential that these findings be checked very carefully and shown to reflect what happens in vivo. TRPV1 AS A THERAPEUTIC TARGET In contrast to the hyperalgesia (excessive sensitiveness or sensibility to pain) following intense noxious stimuli, exposure to capsaicin can result in a subsequent desensitization. Whereas desensitization to comparatively low doses of capsaicin may be specific for capsaicin and its congeners desensitization to higher doses is associated with a loss of responsiveness to other chemical, heat and noxious (high threshold) mechanical stimuli. This cross desensitization of noxious stimuli by capsaicin suggests the use of capsaicin or an analog of it as an analgesic. Of course the ultimate goal, not yet achieved, is to find an analog of capsaicin that induces analgesia without first causing pain. Capsaicin desensitization is well documented, with the extent of desensitization depending on the capsaicin concentration, how frequently it is applied and for how long. Capsaicin induced desensitization has been observed both by recording the activity of DRG neurons as well as by monitoring behavioral (pain) reactions. With low doses of capsaicin given at appropriate time intervals, desensitization does not necessarily take place so that painful excitation can be reproduced with each capsaicin application. With higher doses or prolonged exposure desensitization ensues and consecutive applications of capsaicin become less effective or fail to produce any effect.

3-4

Recently a novel method for producing analgesia using capsaicin in combination with a membrane impermeable local anesthetic (QX-314) has been described (Binshtok, Bean et al. 2007). QX-314 is a positively charged blocker of voltage gated sodium channels, that inhibits action potentials when applied intracellularly but fails to block when applied extracellularly. The idea was to introduce QX-314 intracellularly to pain sensing neurons through the open TRPV1 channel, thereby avoiding the motor and tactile effects that occur with the extracellular application of local anesthetics such as lidocaine. We have all probably experienced the inhibition of motor control and tactile senses with the use of local anesthetics during dental procedures. One limitation of the combination treatment is the same as with the use of capsaicin alone and that is the capsaicin itself causes a painful sensation, which with the combination treatment is expected to last until the QX-314 takes effect. Inhibition of TRPV1 would seem to be a simple approach for producing analgesia. However the situation is not that simple; following the identification of TRPV1 in nociceptors a variety of cell types including keratinocyes, pancreatic cells, endothelial cells, lymphocytes, macrophages and cells from different regions of the brain were shown to also express TRPV1. Its presence in all these cell types in different parts of the body suggests that TRPV1 is normally stimulated by an endogenous ligand (endovanilloid) and not by thermal stimulation. In this context it is important to point out that there is accumulating evidence suggesting that activation of TRPV1 by its endogenous ligand is essential for the maintenance of internal body temperature. Capsaicin in addition to eliciting the sensation of burning also causes hypothermia in a variety of animals and introduction of TRPV1 antagonists leads to hyperthermia in rats, mice and monkeys. One suggestion arising from these studies is that the TRPV1 channels, which function in the regulation of body temperature, are tonically activated via an endogenous ligand. Because TRPV1 antagonists cause hyperthermia it is unlikely that they can developed for use systemically as stand alone agents for the treatment of pain. One possibility that needs to be considered is that an endovanilloid is produced during tissue damage and thereby mediates nociceptor activation. For example, the putative endovanilloid N-Arachidonoyldopamine (NADA) was identified as an endogenous molecule in the mammalian nervous system occurring in several brain nuclei and the dorsal root ganglion. It was originally studied because it activated cannabinoid receptors and it was subsequently found to potently activate the TRPV1 receptor. As would be expected for an endogenous ligand of the TRPV1 receptor NADA was found to increase the frequency of action potential firing of spinal nociceptive neurons and enhance the response to thermal stimuli. Further work is needed to determine whether NADA or another endovanilloid is in fact the normal activator of TRPV1 channels. TRPV2

3-5

TRPV2 is closely related to the capsaicin receptor TRPV1, with which it shares 49% sequence identity, however TRPV1 is activated by capsaicin and responds to temperatures above 43C whereas TRPV2 does not respond to capsaicin and responds to temperatures at or above 52C. TRPV2 is expressed in a variety of tissues including various regions of the brain, spinal cord and sensory ganglia. Its expression in tissues that are never exposed to temperatures as high as 52C suggests that TRPV2 is normally activated by stimuli other than noxious heat in these regions of the body. However, based on its similarity to TRPV1 and its ability to detect high heat stimuli at or above 52C TRPV2 would appear to be a likely candidate for sensing noxious heat at or above 52C in nociceptors. In a study using an ex vivo preparation of spinal cord, thoracic and upper lumbar DRGs, dorsal cutaneous nerves, and dorsolateral trunk skin on one side, the role of TRPV2 in sensing noxious heat was studied. In the ex vivo preparation the sensory endings of the nociceptor remain attached to the cell bodies in the dorsal root ganglion. The nociceptors in the ex vivo preparation would be expected to have properties closer to those of cells in vivo compared to DRG cells isolated by enzymatic treatment and cultured before use (see chapter 1). Using the ex vivo preparation in animals for which TRPV1 was knocked out it was shown that a major population of nociceptors in which TRPV2 was not expressed as determined by immunohistological staining, responded normally to noxious heat from 31C to 52C. This study shows that there is a population of nociceptors that under normal conditions do not require the presence of functional TRPV1 or TRPV2 to detect noxious heat. Until a mouse in which TRPV2 is knocked out has been reported the role of TRPV2 in noxious heat perception will remain undetermined. TRPV3 Camphor is a naturally occurring compound currently used as an active ingredient in an anti itch gel currently on the market. Camphor has been shown to increase the perceived intensity of the cutaneous sensations produced during heating of the skin consistent with its activation of TRPV3. TRPV3 is closely related to TRPV1 and TRPV2 with which it shares 43% and 41% sequence identity respectively. TRPV3 has a unique threshold: It is activated at innocuous temperatures with an activation threshold around 33 to 35C and exhibits increasing responses at higher noxious temperatures. As reported for TRPV1 knockout mice TRPV3 knockout mice exhibited behavioral deficits in their response to hot temperatures at or above 50C. This finding suggests that TRPV3 and TRPV1 have overlapping functions in noxious heat sensation. However, in contrast with TRPV1 knockout mice, deficits in heat hyperalgesia were not observed in TRPV3 knockout mice. TRPV4 3-6

TRPV4 is a calcium permeable nonselective cation channel that shares 40% amino acid identity with TRPV1. It exhibits remarkable gating properties being activated by hypotonic solutions, by certain phorbol ester derivatives and by innocuous temperatures in the range of 27C to 34C. Activation by hypotonic solutions suggests that it serves as a sensor for osmolarity and/or mechanical stretch associated with cellular swelling. Additionally TRPV4 is activated by a process involving the cytochrome P450 epoxygenase dependent formation of epoxyeicosatrienoic acids: submicromolar concentrations of 5' ,6' -epoxyeicosatrienoic acid activates TRPV4. These findings indicate that TRPV4 can be activated by a range of physical and chemical stimuli which may or may not share a common mechanism. TRPV4 knockout mice exhibited several abnormalities in physiological functions that were commensurate with the known gating properties of the channel. TRPV4 knockout mice exhibited abnormalities in osmotic regulation and a marked reduction in the sensitivity of the tail to pressure. Whether TRPV4 plays a role in thermal hyperalgesia is controversial. One study found that TRPV4 knockout mice exhibited a reduced hyperalgesia from 3545C but not at 50C (Todaka, Taniguchi et al. 2004) while a different group failed to find a role for TRPV4 in thermal hyperalgesia (Lee, Iida et al. 2005). Also the first study failed to find alterations in acute thermal behavior in TRPV4 knockout mice whereas the latter study found longer withdrawal latencies during acute tail heating at 45C 46C, suggesting some role of TRPV4 for heat nociception. We shall consider the role of TRPV4 in detecting noxious mechanical stimuli in chapter 4 when we consider the role of protease activated receptor 2 (PAR2) activation in mechanical hyperalgesia.

Table 3-1 Effects of knocking out TRPV channels in Mice Deficits in response to noxious Temperatures Deficits in hyperalgesia in in knockout mice knockout mice TRPV1 >50C exhibited little thermal hypersensitivity TRPV2 no reported knockout no reported knockout TRPV3 >50C not observed TRPV4 controversial controversial

In summary, of the TRPV channels studied with genetic knockout studies (TRPV1, TRPV3, and TRPV4) TRPV1 and TRPV3 have been shown to play a role in thermal nociception. TRPV1 is necessary for thermal hyperalgesia, whereas, TRPV3 is dispensable for this function. Until further studies are conducted, the role of TRPV4 will remain controversial. These findings make it clear that combined knockouts of these channels (TRPV1/TRPV3, TRPV3/TRPV4, and TRPV1/TRPV3/TRPV4) are desirable if we are to understand more fully the role of TRPV channels in nociception. A knockout 3-7

mutant for TRPV2 has not been reported, and its role in noxious heat perception remains to be evaluated.

TRPA1 Heterologously expressed TRPA1 ion channels are activated by irritant compounds from mustard seed, wasabi, horseradish, winter green, cinnamon, garlic, vehicle exhaust fumes and tear gas, all of which elicit a painful burning or pricking sensation. TRPA1 is expressed in DRG neurons and in the inner ear; however, TRPA1 is apparently not essential for the initial detection of sound by hair cells. The role of TRPA1 as a sensor of noxious cold has been controversial; mouse TRPA1 when expressed in CHO cells is activated at temperatures starting near 17C, which is close to the threshold of noxious cold for humans (15C) (Story, Peier et al. 2003). The controversy arose when the rat and human orthologues of TRPA1 expressed in either a human embryonic kidney (HEK293) cell-line or Xenopus oocytes were not activated by cold (Jordt, Bautista et al. 2004). Subsequently, another group was unable to elicit cold activation of heterologously expressed mouse TRPA1 channels in HEK293 cells (Nagata, Duggan et al. 2005). Yet a fourth study (Sawada, Hosokawa et al. 2007) found that mouse TRPA1 expressed in HEK293 cells is a cold-activated channel, which supports the previous findings that TRPA1 responds to noxious cold. The controversy also extended to TRPA1 knockout mice. Nociceptive behavioral responses to contact with a cold surface or to acetoneevoked evaporative cooling were evaluated by two different groups (Bautista, Jordt et al. 2006; Kwan, Allchorne et al. 2006); with the former finding a lack of involvement of TRPA1 in the acute detection of cold and the latter finding a reduced sensitivity to cooling. These contradictory findings regarding the cold activation of TRPA1 appear to have been resolved by subsequent work. A study in mice in which all sensory neurons expressing the tetrodotoxin resistant voltage activated sodium channel (NaV1.8, see Chapter 5) were eliminated, showed resistance to noxious cold, assayed using a cold plate at 0C (Abrahamsen, Zhao et al. 2008). This finding was similar to what was observed in the TRPA1 knockout mice using a cold plate at 0C (Kwan, Allchorne et al. 2006). Significantly, the NaV1.8 knockout mice also exhibited a significant reduction in the expression of TRPA1 in DRG neurons and a lack a TRPA1-mediated nociceptive response to formalin (see Chapter 4). Furthermore, a later study provided a plausible explanation for the discrepancies in the earlier work described above and concluded that TRPA1 acts as a cold sensor in vitro and vivo (Karashima, Talavera et al. 2009). In contrast to the debate over the role of TRPA1 as a sensor of noxious cold its role in bradykinin evoked nociceptor excitation and pain hypersensitivity was not controversial. Bradykinin (BK) is a peptide containing nine amino acid residues (nonapeptide) that is released into inflamed tissues where it induces pain and mechanical and thermal hypersensitivity. Bradykinin injections in TRPA1 knockout mice were much less painful and showed little or no evidence of thermal or mechanical hypersensitivity following the injections. Both consequences are expected if TRPA1 mediates the actions of bradykinin. We will more thoroughly consider the effects of bradykinin on TRPA1 along with its effect on TRPV1 and other ion channels in chapter 4.

3-8

TRPM8 The ability of recombinant TRPM8 to be activated by cold is widely accepted. TRPM8 is activated by cooling agents such as menthol or at temperatures below 26C. Additionally, three independent studies using TRPM8 knockout mice (Bautista, Siemens et al. 2007; Colburn, Lubin et al. 2007; Dhaka, Murray et al. 2007) indicate that TRPM8 is involved in sensing noxious cold. Pain-induced by evaporative cooling of the paw was measured by observing licking and flinching responses of the stimulated paw in normal and TRPM8 knockout mice (Bautista, Siemens et al. 2007): the knockout mice displayed significantly reduced behavior compared to the normal mice. A similar result was found by others (Dhaka, Murray et al. 2007) who additionally found that injection of icilin, a synthetic compound that activates TRPM8 and, to a much lesser extent, TRPA1, into the hind paw of wild-type mice causes the rapid induction of hind paw withdrawal when the mice are placed on a 1C cold plate and that this behavior is completely ablated in TRPM8 knockout mice, suggesting that TRPM8 activation can elicit a nociceptive-like response. The third group (Colburn, Lubin et al. 2007) also found a reduced nociceptive response to evaporative cooling of the paw in TRPM8 knockout mice. Furthermore, these authors also found that following constriction injury caused by ligation of the sciatic nerve normal mice exhibited an enhanced sensitivity to acetone with protracted licking and shaking of the paw whereas TRPM8 knockout mice exhibited no significant increase in the response to evaporative cooling of the paw. These data plainly indicate that TRPM8 is involved in sensing noxious cold. TRPM8 knockout mice retain a number of cold sensitive neurons indicating that TRPM8 is not the only receptor activated by cold. Combined knockouts of TRPA1 and TRPM8 might help clarify the relative role of TRPA1 and TRPM8 in the detection of noxious cold. References cited:

Abrahamsen, B., J. Zhao, et al. (2008). "The cell and molecular basis of mechanical, cold, and inflammatory pain." Science 321(5889): 702-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18669863 Bautista, D. M., S. E. Jordt, et al. (2006). "TRPA1 mediates the inflammatory actions of environmental irritants and proalgesic agents." Cell 124(6): 1269-82. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16564016

3-9

Bautista, D. M., J. Siemens, et al. (2007). "The menthol receptor TRPM8 is the principal detector of environmental cold." Nature 448(7150): 204-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17538622 Binshtok, A. M., B. P. Bean, et al. (2007). "Inhibition of nociceptors by TRPV1-mediated entry of impermeant sodium channel blockers." Nature 449(7162): 607-10. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17914397 Caterina, M. J., A. Leffler, et al. (2000). "Impaired nociception and pain sensation in mice lacking the capsaicin receptor." Science 288(5464): 306-13. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10764638 Caterina, M. J., M. A. Schumacher, et al. (1997). "The capsaicin receptor: a heatactivated ion channel in the pain pathway." Nature 389(6653): 816-24. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9349813 Colburn, R. W., M. L. Lubin, et al. (2007). "Attenuated cold sensitivity in TRPM8 null mice." Neuron 54(3): 379-86. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17481392 Dhaka, A., A. N. Murray, et al. (2007). "TRPM8 Is Required for Cold Sensation in Mice." Neuron 54(3): 371-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17481391 Jordt, S. E., D. M. Bautista, et al. (2004). "Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1." Nature 427(6971): 260-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14712238 Karashima, Y., K. Talavera, et al. (2009). "TRPA1 acts as a cold sensor in vitro and in vivo." Proc Natl Acad Sci U S A 106(4): 1273-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=19144922 Kwan, K. Y., A. J. Allchorne, et al. (2006). "TRPA1 contributes to cold, mechanical, and chemical nociception but is not essential for hair-cell transduction." Neuron 50(2): 277-89. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16630838

3-10

Lee, H., T. Iida, et al. (2005). "Altered thermal selection behavior in mice lacking transient receptor potential vanilloid 4." J Neurosci 25(5): 1304-10. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15689568 Nagata, K., A. Duggan, et al. (2005). "Nociceptor and hair cell transducer properties of TRPA1, a channel for pain and hearing." J Neurosci 25(16): 4052-61. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15843607 Park, J. J., J. Lee, et al. (2007). "Induction of total insensitivity to capsaicin and hypersensitivity to garlic extract in human by decreased expression of TRPV1." Neurosci Lett 411(2): 87-91. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17110039 Sawada, Y., H. Hosokawa, et al. (2007). "Cold sensitivity of recombinant TRPA1 channels." Brain Res 1160: 39-46. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17588549 Story, G. M., A. M. Peier, et al. (2003). "ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures." Cell 112(6): 819-29. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12654248 Todaka, H., J. Taniguchi, et al. (2004). "Warm temperature-sensitive transient receptor potential vanilloid 4 (TRPV4) plays an essential role in thermal hyperalgesia." J Biol Chem 279(34): 35133-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15187078 Trevisani, M., D. Smart, et al. (2002). "Ethanol elicits and potentiates nociceptor responses via the vanilloid receptor-1." Nat Neurosci 5(6): 546-51. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11992116

3-11

Chapter 4 CHEMICAL MEDIATORS One of the long standing goals of pain research is to identify the chemical mediators released into injured or diseased tissues that are responsible for the activation and sensitization of nociceptors. Pain scientists distinguish two aspects of sensitization: allodynia (pain resulting from a normally innocuous stimulus) and hyperalgesia (an enhanced response to a normally painful stimulus). As mentioned in chapter 2 these mediators interact with ion channels in the plasma membrane of the nociceptor utilizing mechanisms used for signaling elsewhere in the nervous system. These mediators either act directly on ligand-gated ion channels (ionotropic channels) or indirectly via metabotropic receptors that are either G-protein-coupled receptors (GPCRs) or receptor tyrosine kinases. SEROTONIN Injection of serotonin (5-Hydroxytryptamine, 5-HT) produces pain and hyperalgesia in humans. In-vivo the source of this 5-HT in humans is platelets, which are known to play an important role in inflammation. In support of this idea is the finding that cutaneous injection of platelets causes acute pain and hyperalgesia (Schmelz, Osiander et al. 1997). Unlike the nerve terminal of 5-HT containing neurons, the platelet cannot synthesize 5HT, rather it relies upon 5-HT uptake. Interestingly, the protein responsible for the human platelet 5-HT uptake is identical to that for the brain 5-HT transporter, and the selective serotonin reuptake inhibitors, used to treat depression, significantly reduce the 5-HT concentration in the platelets of depressed patients (Maurer-Spurej, Pittendreigh et al. 2004). One might reasonably assume that in these patients there will be a significant reduction in the amount of 5-HT released from platelets into damaged inflamed tissue resulting in reduced 5-HT induced pain. The mammalian family of serotonin receptors is very large consisting of fourteen different receptor subtypes, grouped into seven families. Although 5-HT is known to play an important role in nociception, there is only a limited appreciation of the 5-HT receptor subtypes involved in this process, and how they interact with each other and other chemical mediators of nociception. Intraplantar injection in the rat of 5-HT or the 5-HT2A receptor agonist (-methyl 5-HT) significantly reduced the paw-withdrawal latency to radiant heat stimulation. Furthermore, pretreatment with the 5-HT2A receptor antagonist (ketanserin), attenuated the behavioral response following the injection of 5-HT (Tokunaga, Saika et al. 1998). These findings strongly suggest that the 5-HT2A receptor subtype is involved in 5-HT-induced hyperalgesia in acute injury and inflammation. These findings were extended by showing that 5-HT2A receptor inhibition in rats: by local injection (intra plantar) of sarpogrelate blocked primary thermal hyperalgesia (Sasaki, Obata et al. 2006), and systemic injection of sarpogrelate blocked complete Freunds adjuvant (CFA) induced thermal hyperalgesia (Okamoto, Imbe et al. 2002), by local injection of ketanserin, produced dose-dependent inhibition of carrageenan-evoked hyperalgesia (Wei, Chen et al. 2005). Taken together these results suggest that 5-HT has a central role in hyperalgesia resulting from tissue injury by activating 5-HT2A receptors at nociceptor nerve terminals.

4-1

The 5-HT2A receptor is a G-protein coupled receptor, and its activation leads to depolarization of the resting membrane potential of acutely isolated rat DRG neurons. In cells exhibiting a 5-HT2A-mediated response, 5-HT and -methyl 5-HT depolarized the resting membrane potential and decreased the membrane permeability (measured electrically as an increase of the slope of the current voltage relationship of the membrane) (Todorovic, Scroggs et al. 1997). In order to understand the different mechanisms by which different ion channels depolarize the plasma membrane we will consider the idealized situation illustrated in Figure 4-1. The two situations illustrated in Figure 4-1 are for a cell that has a resting permeability to both Na+ and K+. In both the situations illustrated in Figure 4-1 the membrane depolarization results from an increase in the ratio PNa/PK (i.e. the permeability to sodium increases relative to that for potassium). Ideally, this can happen either when PNa increases and PK remains the same as illustrated on the left of Figure 4-1 (increased permeability causes a decreased slope resistance) or when PNa remains the same and PK decreases (decreased permeability causes an increased slope resistance), as illustrated on the right of the Figure 4-1. In the acutely isolated cells that showed a 5-HT2A-mediated depolarization the reversal potential (Erev) for the depolarization was linearly related to the logarithm of the extracellular potassium concentration [K+]out, indicating the depolarization resulted from a decrease in the resting K+ permeability (Todorovic, Scroggs et al. 1997). Please remember from chapter 2 that K2P channels are thought to be responsible for the cells resting K+ permeability. Therefore it seems reasonable to speculate that the decrease in K+ permeability results from the closing of K2P channels in the DRG cells studied. In an heterologous expression system the excitatory effects of a G-protein coupled receptor have been shown to occur via inhibition of some K2P channels (Chemin, Girard et al. 2003). The TREK-1 channel is a member of the K2P channel family and is extensively colocalized with TRPV1 (Alloui, Zimmermann et al. 2006) making it a candidate for one of the channels responsible for the resting K+ permeability of nociceptors. Moreover, in animals in which TREK-1 is knocked out animals are more sensitive to low threshold mechanical stimuli and display an increased thermal and mechanical hyperalgesia in conditions of inflammation(Alloui, Zimmermann et al. 2006). It would be interesting to determine whether TREK-1 or some other K2P channel is involved in the 5-HT2A-mediated response. If the closing of K+ channels can cause pain as these findings indicate then another conclusion to be drawn from this work is that the opening of K+ channels in nociceptors is potentially an important mechanism in antinociception. We will return to this idea when we consider the role of K+ channels in the antinociception induced by opioid receptor agonists. In contrast to all the other serotonin receptors which are G-protein coupled receptors, the 5-HT3 receptor is a ligand-gated cation channel consisting of five monomers forming a central pore region (see Figure 2-4A). Five monomer subtypes, the 5-HT3AE subunits, have been identified and functional homomeric 5-HT3A receptors and heteromeric 5HT3A/B receptors were found to be expressed in neurons. In DRG neurons exhibiting a 5HT3 receptor response, 5-HT and 2-methyl-5-HT produced depolarization with decreased input resistance. Moreover, the reversal potential (Erev) for the depolarizing response

4-2

65 membrane potential (mV)

[ENa]

measured electrically as a decreased slope resistance

measured electrically as an increased slope resistance

PNa increased PK unchanged

PNa unchanged PK decreased

PNa/PK increases

-85 [EK]

Figure 4-1. The same depolarizing receptor potential can be generated by either an increase in sodium permeability or a decrease in potassium permeability.

4-3

became less negative when the extracellular K+ concentration was raised to 10 mM and the depolarization was converted to hyperpolarization in a Na+ free solution, indicating that the 5-HT3 response resulted from an increased permeability to Na+ and K+. Studies in 5-HT3A-knock out mice lead the authors to the interpretation that 5-HT3 receptors are not involved in acute pain but are required for persistent pain. The formalin test is widely used for evaluating the effects of analgesic compounds in laboratory animals. The noxious stimulus in the formalin test in mice is an injection of a dilute formalin solution under the skin of the dorsal surface of a hind paw. The response is the amount of time the animals spend licking the injected paw. There are two distinct periods of licking, an early phase lasting the first 5 min and a late phase lasting from 20 to 30 min after the injection. It is generally thought that the early phase is due to a direct effect on nociceptors. Using the formalin test they found that the first-phase of pain behavior did not differ in wild-type and mutant mice. In contrast, the second phase of pain behavior was significantly reduced in the mutant animals, indicating that 5-HT3 receptors are important for persistent pain (Zeitz, Guy et al. 2002). Moreover, they also observed a significant reduction of the second-phase behavior in the formalin test following intrathecal (inside the spinal canal) administration of a 5-HT3-receptor antagonist, suggesting that the 5-HT3 receptors affected are in the spinal cord. BRADYKININ Intradermal injection of bradykinin in humans produces a dose-dependent pain and a heat hyperalgesia, indicating that bradykinin both excites and sensitizes nociceptors (Manning, Raja et al. 1991). Bradykinin is a polypeptide formed in the blood; it causes contraction of non-vascular smooth muscle, is a potent vasodilator of certain blood vessels, increases vascular permeability and most importantly for our purposes is involved in the mechanism of pain. Local inflammation following tissue damage triggers the release of bradykinin (the nonapeptide H-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-ArgOH) produced by kallikrein mediated enzymatic cleavage of kininogen at the site of tissue injury and inflammation. Kallikreins are serine proteases that liberate kinins (bradykinin and kallidin) from the kininogens (High-molecular weight kininogen and low-molecular weight kininogen). Human tissue kallikrein preferentially releases the decapeptide Lys- bradykinin (kallidin) from kininogens. Prekallikrein is the precursor of kallikrein and it can only activate kinins after being activated during tissue injury and inflammation. Once formed bradykinin is degraded by two enzymes carboxypeptidase-N, also known as kininase-1, and angiotensin converting enzyme (ACE), also called kininase-2. Kininase-1 transforms bradykinin and kallidin into their active metabolites, des-Arg9-bradykinin and Lys-des-Arg9- bradykinin (i.e. bradykinin and kallidin without their C-terminal arginine residues). ACE removes the C-terminal dipeptide from bradykinin or Lys- bradykinin, which leads to their inactivation. ACE inhibitors lead to an increase in bradykinin due to decreased degradation and also to a decrease in angiotensin (a vasoconstrictor), for which they are used in the treatment of hypertension. It has been suggested that some of the

4-4

blood pressure lowering effects of ACE inhibitors may be due to their effects on bradykinin. The actions of bradykinin are mediated through two G-protein-coupled receptors, denoted B1 and B2. Bradykinin activates B2 receptors while B1 receptors exhibit higher affinity for des-Arg9-bradykinin (i.e. the B1 receptor is selective for kinin metabolites without the C-terminal arginine residue). B2 receptors are constitutively expressed in DRG neurons and are thought to be the predominant functional bradykinin receptor subtype in non-traumatized tissues. On the other hand B1 receptors are not normally constitutively expressed to a significant extent, but are up regulated during chronic inflammation. Since the B1 receptor is not present in non-inflamed tissues it is an appealing target for the development of antagonists as they would be expected to cause few unwanted side effects. Since the B2 receptor is a G-protein-coupled receptor there needs to be an ion channel that is activated down stream from B2. Also given that B2 receptor activation causes thermal hyperalgesia and that TRPV1 knockout mice exhibit little thermal hypersensitivity in the setting of tissue inflammation, it is plausible that TRPV1 might be the ion channel that is acted upon down stream from bradykinin. In two studies (Kollarik and Undem 2004; Rong, Hillsley et al. 2004) with TRPV1 knockout mice, bradykinin elicited an action potential discharge in C-fibers, and in both studies there was no difference in the initial response in the knockout mice compared to the normal mice. In one of the studies (Kollarik and Undem 2004) the response in the knockout mice was less persistent than in normal mice, while in the other there was no significant difference. These findings suggest that TRPV1 contributes to, but is not required for, B2 receptor mediated nociceptor excitation. On the other hand, injection of bradykinin intraplantarly produced substantial thermal hypersensitivity in wild-type mice but not in TRPV1 knockout mice, demonstrating that TRPV1 is necessary for the development of bradykinin-induced thermal hypersensitivity in vivo (Chuang, Prescott et al. 2001). Activation of most cells by bradykinin is mediated by phospholipase C and/or phospholipase A2, therefore these are the biochemical pathways that are likely to mediate between B2 and TRPV1. The models in Figures 4-2A & B summarize how phospholipase C and phospholipase A2, respectively, are thought to couple B2 to TRPV1. Much of the evidence supporting this model comes from experiments using isolated neuronal cell bodies of a dorsal root ganglion or from heterologous expression systems. It is important to keep in mind that the uncertainties and assumptions associated with these procedures make it essential that the findings be checked very carefully and shown to resemble what actually occurs in vivo. As illustrated in Figure 4-2A, B2 is coupled to the enzyme phospholipase C- via the guanosine nucleotide-binding protein Gq/11. Heterotrimeric G proteins (G) of the Gq/11 family stimulate phospholipase C- via Gq/11 family members (Gq, G11, G14 and G15/16). Five different - and 12 -subunits have been described allowing for numerous coexpression possibilities for , and subunits. The exact subunit composition for Gq/11 coupled to B2 is unknown and is indicated as Gq/11 in Figure 4-2A. Binding of bradykinin to B2 leads to the activation of Gq/11 via exchange of GTP 4-5

TRPV1

A
BK

PIP2

DAG

B2

PLC-b Gq/11a b g IP3

PKCe

B
G?

NSAID PLA2 AA LOX BK

TRPV1 COX

IP PGI2 PGE2 EP GS PKA

B2
12-HPETE

ATP AC cAMP

NSAID AA PLA2

COX

PGE2

TRPA1
EP

G? BK

GS PIP2 ATP AC PKA P

B2

PLC-b Gq/11a DAG b IP g 3

cAMP
Ca2+

Ca2+

ER

Figure 4-2. Scheme showing the biochemical pathways that have been implicated in the modulation of TRPV1 (A&B) and TRPA1 (C) by bradykinin (BK). See text for further details

4-6

for GDP in the nucleotide-binding pocket and dissociation of the subunits. In turn the activated Gq/11 with bound GTP activates the enzyme PLC- which hydrolyzes PIP2 (phosphatidylinositol (4,5)-bisphosphate) to form IP3 (inositol 1,4,5-trisphosphate) and DAG (diacylglycerol). The Gq/11 -subunit is an intrinsic GTPase, hydrolysing the terminal phosphate of GTP to restore GDP to the nucleotide-binding pocket leading to the reassociation of the Gq/11 with the subunits thereby returning Gq/11 to the inactive state. As illustrated in Figure 4-2A the available evidence has so far suggested two possible mechanisms by which hydrolysis of PIP2 modulates the activity of TRPV1: either by phosphorylation of TRPV1 through the activation of PKC (protein kinase C) via DAG, and/or through depletion of PIP2 which normally acts to inhibit TRPV1. Briefly, evidence from isolated DRG neurons and HEK293 cells expressing TRPV1 and B2 has shown that stimulation with bradykinin, or activation of PKC, lowers the threshold temperature for activation of TRPV1 currents (Cesare and McNaughton 1996; Sugiura, Tominaga et al. 2002). Moreover, inhibition of PKC results in a 70% decrease in the sensitization of TRPV1 by bradykinin (Cesare, Dekker et al. 1999). This finding does not rule out a role for other isoforms of PKC from playing a role in TRPV1 sensitization by bradykinin. Since PKC is normally activated by DAG it is reasonable to assume that this is the mechanism of PKC activation in nocicepotors (see Figure 4-2A). Turning to the other mechanism: in a study using heterologous expression systems, it was shown that decreasing the level of PIP2 in the plasma membrane mimics the potentiating effects of bradykinin on TRPV1 (Chuang, Prescott et al. 2001). The role of PIP2 in regulating TRPV1 may be more complicated than shown in Figure 4-2A: the evidence suggests a model in which PIP2 has both an inhibitory and activating effects on TRPV1 (Lukacs, Thyagarajan et al. 2007). A concomitant activating and inhibitory effect is the result of a bell-shaped dependence of TRPV1 channel activity on PIP2 levels. If resting PIP2 levels are high, that is to the right of the peak of the bell-shaped dose response curve, a moderate decrease in PIP2 levels will result in increased TRPV1 channel activity, whereas further decreases to PIP2 levels to the left of the peak will result in channel inhibition. Remember from chapter 3 that exposure of TRPV1 containing nociceptors to high doses of capsaicin is associated with a loss of responsiveness to capsaicin as well as other chemicals, heat and noxious mechanical stimuli. It has been suggested that high doses of capsaicin maximally activate TRPV1 causing a large influx of calcium which activates PLC thereby depleting PIP2 causing a profound inhibition of TRPV1 which is responsible for the loss of responsiveness to capsaicin and other stimuli. Cumulative evidence from a number of cell types has shown that GPCRs can couple to PLA2, however which G-protein is used to couple B2 to PLA2 in nociceptors has not yet been determined, hence the question mark between B2 and PLA2 in Figure 4-2B. Activated PLA2 catalytically hydrolyzes phospholipids releasing arachidonic acid. Two important pathways for arachidonic acid metabolism are the cyclooxygenase (COX) and 12-lipoxygenase (LOX) pathways. The COX pathway forms intermediate compounds which are then converted into biologically active compounds, which include the prostaglandins PGE2 and PGI2, while the LOX pathway produces 12- HPETE. Thus the

4-7

lipid mediators PGE2, PGI2 and 12-HPETE are hypothesized to be produced in the nociceptor in response to bradykinin. Although we are considering the role of PGE2, PGI2 and 12-HPETE in mediating the effects of bradykinin, it should be kept in mind that these lipid mediators do not have to be produced in the nociceptor where they act; they can also be produced by other nearby cells during tissue injury (not necessarily in response to bradykinin), and then diffuse to the nociceptor; this is referred to as paracrine signaling. First consider the evidence supporting the AA LOX 12-HPETE pathway in Figure 4-2B. Experiments with cultured DRG neurons and cutaneous nerve fibers in the in vitro skin-nerve preparation demonstrated that bradykinin, acting via B2 receptors, excite sensory nerve endings by activating TRPV1 via the production of LOX metabolites of arachidonic acid (Shin, Cho et al. 2002). Moreover bradykinin directly stimulated the production of 12-HPETE which was shown in an expression system to directly activate TRPV1 (Hwang, Cho et al. 2000). It is unclear to what extent this pathway normally contributes to B2 receptor mediated nociceptor excitation given the evidence, discussed above, from TRPV1 knockout mice, that TRPV1 activation is not required for nociceptor activation by bradykinin (Kollarik and Undem 2004; Rong, Hillsley et al. 2004). We need to consider whether PGE2 by itself causes pain and thermal hyperalgesia. Intraplantar injection of PGE2 into the hindpaw of mice produced a dose dependent short duration paw licking (nociceptive) behavior when compared with control animals (Kassuya, Ferreira et al. 2007). Additionally there was a reduction of paw withdrawal latency (thermal hyperalgesia) following intraplantar PGE2 injection which was significantly diminished in TRPV1 knockout mice (Moriyama, Higashi et al. 2005). Furthermore, using an isolated skin nerve preparation the bradykinin induced thermal hyperalgesia was shown to be mediated by cyclooxygenase activation (Petho, Derow et al. 2001). The diversity of actions of PGE2 are believed to result from its interaction with a family of G-protein-coupled (prostanoid E receptors), EP receptors, designated EP1 EP4, all of which are found in DRG neurons (Southall and Vasko 2001). The identity of the EP receptor(s) which couple PGE2 to thermal hyperalgesia is still controversial. The activation of EP receptors by PGE2 can stimulate PKA, PKC and mitogen-activated protein kinases (MAPKs). In cells transfected with TRPV1 heat activated currents were greatly potentiated by activation of PKA (Rathee, Distler et al. 2002) and the potentiation was greatly reduced in cells transfected with TRPV1 mutants having mutations at the PKA phosphorylation sites. Furthermore, disruption of the PKA anchoring protein AKAP150 in mice diminishes the decrease in paw withdrawal latency, to thermal stimuli, induced by intraplantar PGE2 injection but does not affect thermal sensitivity under basal conditions (Schnizler, Shutov et al. 2008). These findings strongly suggest that PGE2induced thermal hyperalgesia is mediated in part by PKA. As illustrated in Figure 2B EP is coupled to PKA via the G protein (GS) which activates the enzyme adenylate cyclase (AC). AC catalyzes the conversion of adenosine triphosphate (ATP) to cyclic adenosine monophosphate (cAMP) which in turn activates PKA. In a separate study thermal hyperalgesia induced by intraplantar injection of PGE2 was unaffected by knockout of PKC (Khasar, Lin et al. 1999).

4-8

In contrast to the family of EP receptors for PGE2, there is only a single IP receptor for PGI2. Using the acetic acid-induced writhing test, where injection of a dilute acetic acid solution intraperitoneally causes writhing responses in wild-type mice, IP knockout mice showed a markedly reduced writhing response compared to the wild-type mice (Murata, Ushikubi et al. 1997). Furthermore, thermal hyperalgesia induced by intraplantar PGI2 injection was significantly diminished in TRPV1 knockout mice and IP knockout mice, indicating that PGI2 sensitizes TRPV1 via IP receptors (Moriyama, Higashi et al. 2005). Moreover, in DRG neurons PGI2 -induced potentiation of capsaicin activation of TRPV1 was suppressed by a PLC inhibitor and also by a PKC inhibitor suggesting the involvement of a PLC-PKC dependent pathway (Moriyama, Higashi et al. 2005). In Figure 4-2 this is indicated by a dotted line going from IP in Figure 4-2B to Gq/11 in Figure 4-2A. It is important to check that the model in Figure 4-2A & B describes the actual in vivo coupling between the B2 receptor and TRPV1. The mechanism of bradykinin-induced nociception (paw licking) was tested by injecting bradykinin into the paw of mice in the presence of different enzyme inhibitors. Selective inhibitors of phospholipase C, PKC, PLA2 or LOX markedly decreased the nociception caused by BK but not that of capsaicin (Ferreira, da Silva et al. 2004). As far as they go these findings are consistent with the models presented in Figure 4-2A & B. TRPV1 is not the only TRP channel that plays a role in the activation and sensitization of nociceptors by bradykinin; TRPA1 has also been implicated. In TRPA1 knockout mice the bradykinin-induced response of DRG neurons was significantly reduced but not absent and comparable to that of TRPV1 knockout mice (Bautista, Jordt et al. 2006). In a behavioral study it was found that following subcutaneous intraplantar bradykinin injection in mice, wild-type mice spent almost three times as long tending to the affected paw as TRPA1 knockout mice (Kwan, Allchorne et al. 2006). These findings suggest that TRPA1 activation plays a role in the acute pain caused by bradykinin. With respect to pain hypersensitivity there was no evidence of bradykinin-induced thermal hypersensitivity with intraplantar bradykinin injection in TRPA1 knockout mice (Bautista, Jordt et al. 2006). Thus both TRPV1 and TRPA1 are necessary for the development of bradykinin-induced thermal hyperalgesia. Colocalization studies using antibodies to TRPA1 and TRPV1 showed that all TRPA1-positive neurons also expressed TRPV1 (Bautista, Movahed et al. 2005), indicating that the biochemical pathways shown in Figure 4-2A & B are available to modulate TRPA1. Although the experimental evidence is not as extensive as for TRPV1, we can suggest (see Figure 42C) that to some extent the same pathways that modulate TRPV1 also affect TRPA1. Using a heterologous expression system and dorsal root ganglia neurons both the PLC and PKA pathways were shown to potentiate currents carried by TRPA1 (Wang, Dai et al. 2008). Therefore in Figure 4-2C the pathways for activation of PLC and PKA are the same as in Figure 4-2A & B. However, the PKC inhibitor did not prevent the potentiation by bradykinin of the currents carried by TRPA1 and the PKC activator did not potentiate the TRPA1 response (Wang,

4-9

Dai et al. 2008). These findings indicate that PKC activation does not contribute to the sensitization of TRPA1, which is different from the mechanism for TRPV1 sensitization, hence PKC is not present in Figure 4-2C. Intracellular calcium is an important intracellular messenger and the release of calcium from intracellular stores has been shown to directly activate TRPA1 (Jordt, Bautista et al. 2004; Zurborg, Yurgionas et al. 2007). Hence, the inositol 1,4,5-trisphosphate (IP3) induced calcium release pathway is included in Figure 4-2C. Further studies will be required to elucidate the exact nature of the functional interaction between TRPV1 and TRPA1. One possibility is that TRPV1 and TRPA1 combine to form hetero multimeric channels. The analgesic action of nonsteroidal anti-inflammatory drugs (NSAIDs), such as aspirin, is produced through their inhibition of cyclooxygenase (COX), the enzyme that makes prostaglandins. Based on the molecular model in Figure 4-2 we can conclude that one of the NSAID sites of action is the COX activated by bradykinin in nociceptors. ATP It was mentioned in chapter 1 that during tissue damage the release of components of the cell cytoplasm are likely candidates to act on nociceptors to cause pain. ATP (Adenosine5'-triphosphate) is an important source of intracellular energy where it is produced during cellular respiration and consumed by many cellular processes. Human experiments have shown that delivery of ATP into skin causes pain in a dose dependent manner (Hamilton, Warburton et al. 2000). Since ATP is membrane impermeable, receptors for ATP located in the nociceptor plasma membrane are needed to detect the ATP released from damaged cells into the extracellular space. ATP targets two distinct receptor subtypes of the P2receptor family: ATP activates ionotropic P2X receptors and metabotropic P2Y receptors. Currently, seven different P2X receptor subtypes and eight P2Y subtypes have been identified. P2X RECEPTORS Normally, P2X2 and P2X3 are expressed by small sensory neurons of the DRG. The experimental findings suggest that ATP-induced currents in DRG neurones are mediated largely by homomeric P2X3 receptors and heteromeric P2X2/ P2X3 receptors. In P2X3 knockout mice there was a significant loss of approximately 90% of DRG neurons responsive to ATP (Cockayne, Hamilton et al. 2000). A small residual sustained response to ATP was seen in some DRG neurons from P2X2/P2X3 double knockout mice indicating the presence of low levels of other P2X subunits or P2Y receptors in some neurons (Cockayne, Dunn et al. 2005). The early phase of formalin-induced pain behavior was significantly reduced in P2X3 knockout mice, although responses to other noxious stimuli were normal (Cockayne, Hamilton et al. 2000; Souslova, Cesare et al. 2000). In contrast, the early phase of formalin-induced pain behavior was not attenuated in P2X2 knockout mice (Cockayne, Dunn et al. 2005). These in-vivo findings taken together with the in-vitro findings

4-10

discussed above support the notion that the ionotropic P2X3 receptor signals acute pain from damaged tissue. A-317491 (Jarvis, Burgard et al. 2002) is a non-nucleotide antagonist that has high affinity and selectivity for blocking P2X3 and P2X2/ P2X3 channels and was found to produce a similar reduction of formalin-induced pain as was produced in the knockout mice. Moreover, as in the knockout mice responses to noxious mechanical and thermal stimuli were normal. Taken together, the in-vitro findings, the findings from the knockout mice and those with A-317491 have provided strong evidence that P2X3-containing channels contribute to nociception. Interestingly, TRPA1 knockout mice exhibit a significant reduction in all phases of formalin-induced pain behavior including the early phase (McNamara, Mandel-Brehm et al. 2007). The attenuation of the early phase of formalin-induced pain behavior in TRPA1 knockout mice is much greater than in P2X3 knockout mice. At the present time it is unclear what relationship if any exists between P2X3 and TRPA1 receptors. Formaldehyde, the active ingredient in formalin, is a fixative that covalently cross-links proteins in a nonspecific manner. This cross-linking leads to a variety of effects, including general tissue damage. The thinking has been that the tissue damage releases ATP from cells and the released ATP activates P2X3 receptors. In contrast the finding with TRPA1 indicates that there is a direct effect of formalin on TRPA1 rather than an indirect effect through nonspecific tissue damage (McNamara, Mandel-Brehm et al. 2007). Remember from chapter 3 that TRPA1 channels are activated by a variety of irritant compounds; the mechanism of formalin activation is likely to be similar to these compounds, which are thought to induce covalent modification of TRPA1. P2Y RECEPTORS DRG neurons also express P2Y receptors and these receptors have been implicated in the potentiation of pain: extracellular ATP injection has been shown to induce thermal hyperalgesia in mice and the ATP induced thermal hypersensitivity was lost in TRPV1 knockout mice (Moriyama, Iida et al. 2003). ATP-induced thermal hyperalgesia was preserved in P2Y1 knock out mice (Moriyama, Iida et al. 2003) while P2Y2 knockout mice did not show significant ATP-induced thermal hyperalgesia (Malin, Davis et al. 2008). Additionally, P2Y2 knockout mice show deficits in noxious heat (but not cold) sensation compared with wild type mice. The obvious next question is how activation of P2Y2 leads to thermal hypersensitivity that is lost in TRPV1 knockout mice. A simple explanation would be that activation of P2Y2 causes TRPV1 to be modified in such a way that its thermal sensitivity is increased. The best available evidence suggests that P2Y2 is a GPCR that is coupled to the enzyme phospholipase C- via the guanosine nucleotide-binding protein Gq/11. As shown in Figure 4-3 the available evidence has so far suggested two possible mechanisms by which hydrolysis of PIP2 modulates the activity of TRPV1: either modulation of TRPV1 occurs by phosphorylation through the activation of PKC (protein kinase C) via DAG or through depletion of PIP2 which normally acts to inhibit TRPV1.

4-11

ATP P2Y2
GTP

ATP

TRPV1

Gq/11a PLC-b 2 PIP g GDP b

ATP PKCe

IP3

GDP

DAG

g b

Figure 4-3. Activation of P2Y2 by ATP causes thermal hypersensitiviy via phosphorylation of TRPV1 or release from inhibition by PIP2 . See text for further details.

4-12

PROTEINASE-ACTIVATED RECEPTORS Proteases in the circulation which are generated during tissue injury have been shown to activate a family of G-protein coupled, proteinase-activated receptors (PARs). These PARs play a role in hemostasis, inflammation, and pain. The proteases cleave extracellular N-terminal domains of the PARs to expose tethered ligands that bind to and activate the cleaved receptors (see Figure 4-5). Four PARs have been identified by molecular cloning: PAR1, PAR2, PAR3 and PAR4. Of these, PAR1 and PAR2 are present on DRG neurons and have been shown to play a role in neurogenic inflammation, that is, inflammatory symptoms that result from the release of substances from primary sensory nerve terminals (see below). Before considering the role of PAR1 and PAR2 in neurogenic inflammation and pain we need to consider which proteases activate PAR1 and PAR2 and where the proteases come from. PAR2 is activated by the serine proteases tryptase and trypsin, although trypsin is able to activate PAR2, trypsin itself is not present in most tissues, thus, it is probably not the endogenous enzyme that activates PAR2. Conversely, tryptase is released during human mast cell degranulation and is able to cleave PAR2 in cells normally expressing PAR2 or in cells transfected with the receptor. Therefore tryptase is a likely candidate for the enzyme that activates PAR2 (see Figure 45). Thrombin appears to be the most likely agonist to activate neuronal PAR1. Over 100 years ago it was documented that antidromic stimulation of afferent sensory nerve fibers, resulting in conduction of action potentials opposite to the normal direction, results in erythema (reddening of the skin). The finding that peripheral activation of afferent sensory neurons is able to produce manifestations of an inflammatory response is referred to as neurogenic inflammation. Destruction of capsaicin sensitive neurons greatly decreases neurogenic inflammation produced by antidromic stimulation of afferent sensory fibers, implicating the capsaicin sensitive nociceptors in neurogenic inflammation. As described in chapter 1 painful stimuli result in the generation of a train of action potentials in nociceptors that are conducted by their axons to the spinal cord, and after processing by the brain a sensation of pain occurs. On the other hand, the action potentials will also retrogradely (opposite to the usual direction) invade the arborizations (branching terminal processes of the nociceptor) as illustrated in Figure 4-4. The resulting depolarization of the terminal causes the release of the neuropeptides, substance P (SP) and calcitonin gene-related peptide (CGRP), which in turn act on target cells in the periphery such as mast cells and vascular smooth muscle producing inflammation, which is characterized by redness, warmth and swelling (Schmelz and Petersen 2001; Richardson and Vasko 2002). In general CGRP does not induce heat hyperalgesia of nociceptors, however in appropriately chosen mouse strains, CGRP does induce heat hyperalgesia (Mogil, Miermeister et al. 2005). Therefore, depending on the mouse strain, neurogenic inflammation can cause hyperalgesia. Presumably there is a similar variability in the ability of CGRP to induce heat hyperalgesia in the nociceptors of humans. As mentioned above PAR2 is activated by the serine proteases tryptase and trypsin, additionally a short synthetic peptide, (SLIGRL, seryl-leucyl-isoleucyl-glycyl--arginylleucinamide) based on the tethered ligand sequence, has been shown to activate the receptor and thereby mimic the effects of the proteases (see Figure 4-5). Injection of the

4-13

region of neurogenic inflammation

blood vessel

noxious stimulus

mast cell

to the spinal cord

Figure 4-4. Schematic drawing of the process of neurogenic inflammation. A noxious stimulus causes the depolarization of the terminal of a nociceptor thereby initiating propagating action potentials in the axon. The direction of propagation is shown by the red arrows. Action potentials propagate along the axon towards the spinal cord and also invade the nearby branching terminal processes (arborizations) of the nociceptor. See text for further details.

4-14

SLIGRL TRYPTASE

TFLLR THROMBIN

(TREK-1?) K2P PAR1 Gs G i AC

TRPV1

TRPV4

PAR2
P

ERK1/2

capsaicin cAMP 4a PDD PKA PKC PKD

Figure 4-5. Mechanisms of nociceptor activation and inhibition by the protease receptors 2 and 1 respectively. See text for further explanation.

4-15

synthetic SLIGRL in the rat paw has been shown to induce neurogenic inflammation that depends on the release of CGRP and SP (Steinhoff, Vergnolle et al. 2000). Several studies have attempted to exclude the possibility that the nociceptive behavior and hyperalgesic effects of PAR2 activation might be secondary to the neurogenic inflammatory response. Intraplantar injection of the PAR2 agonist SLIGRL, at sub inflammatory doses, was found to induce thermal hyperalgesia and biting and licking (nociceptive) behavior. Remember that TRPV1 is necessary for the hyperalgesia resulting from intense noxious stimuli that result in tissue damage (see Chapter 3) and for the development of bradykinin-induced thermal hypersensitivity. Likewise, it was found that PAR2-induced hyperalgesia results from sensitization of TRPV1 (Amadesi, Nie et al. 2004) and was abolished by the capsaicin receptor antagonist capsazepine (Kawao, Shimada et al. 2002). Although capsazepine inhibited the thermal hyperalgesia it did not inhibit the PAR2-induced nociception (time spent in licking and biting the injected paw) (Kawao, Shimada et al. 2002). Further studies in DRG neurons and transfected HEK293 cells indicated that PAR2 activates PKC and PKA in DRG neurons (see Figure 4-5), and thereby sensitizes TRPV1 (Amadesi, Cottrell et al. 2006). This finding bears similarity to the sensitization of TRPV1 by bradykinin acting via PKC and PKA (see Figure 4-2). Remember that although capsazepine inhibited PAR2-induced thermal hyperalgesia it did not inhibit the immediate pain, implying the existence of another pathway that excites the nociceptor. In a study of nociceptive neurons innervating the mouse colon, SLIGRL superfusion for 3 min caused a sustained depolarization which was associated with an increased input resistance that lasted up to 60 min. and was blocked by the PKC inhibitor, calphostin, and the ERK1/2 (extracellular signal-regulated kinase ) inhibitor PD98059 (Kayssi, Amadesi et al. 2007). The membrane depolarization and increase in input resistance following PAR2 activation probably result from the closing of K+ channels open at the resting membrane potential. As discussed earlier, K2P channels are thought to be responsible for the cells resting K+ permeability and are therefore likely to be the K+ channels that are closed following PAR2 activation (see Figure 4-5). Remember that TREK-1 is a K2P channel expressed in nociceptors and TREK-1 knockout animals are hyper-sensitive to mechanical stimuli (Alloui, Zimmermann et al. 2006). Perhaps TREK1 or some other K2P channel is involved in the PAR2-mediated depolarization. In addition to blocking the resting K+ permeability, SLIGRL markedly suppressed (55%) the delayed rectifier K+ currents (Kayssi, Amadesi et al. 2007). Membrane depolarization opens the delayed rectifier K+ channel which tends to bring the membrane potential back to the resting potential, therefore suppression of the delayed rectifier would be expected to enhance and prolong membrane depolarization caused by other ion channels. Intraplantar injection in the rat of a sub inflammatory dose of 10 g of SLIGRL induced a prolonged thermal and mechanical hyperalgesia (Vergnolle, Bunnett et al. 2001) while injection of a much lower dose produced only thermal hyperalgesia (Kawabata, Kawao et al. 2001). Subsequently, it was hypothesized that PAR2-mediated mechanical hyperalgesia requires sensitization of TRPV4 (Grant, Cottrell et al. 2007). They found that PAR2 mediated mechanical hyperalgesia was not observed in TRPV4 knockout mice. Moreover, intraplantar injection of the TRPV4 agonist 4-Phorbol 12,13-

4-16

didecanoate (4-PDD) (Watanabe, Davis et al. 2002) produced mechanical hyperalgesia in normal mice which was not observed in TRPV4 knockout mice (Grant, Cottrell et al. 2007). Lastly, they found that antagonists of phospholipase C and protein kinases A, C and D inhibited PAR2-induced sensitization of TRPV4 Ca2+ signals and currents. In Figure 4-5 I have attempted to summarize the different mechanisms, just described, by which PAR2 activation is thought to cause pain and both thermal and mechanical hyperalgesia. As mentioned above thrombin appears to be the most likely agonist to activate PAR1. Just as for PAR2 which is activated by a short synthetic peptide based on the tethered ligand sequence of PAR2, PAR1 can be selectively activated by the synthetic peptide (TFLLR) corresponding to the PAR1 receptor's tethered ligand (see Figure 4-5). Intraplantar injection of the PAR1 agonist TFLLR, at sub inflammatory doses, increased the threshold and withdrawal latency for both mechanical and thermal stimuli (Asfaha, Brussee et al. 2002). The available evidence from other systems suggests that PAR1 is negatively coupled to adenylate cyclase (AC) via the inhibitory G-protein (Gi), thereby inhibiting the activation of PKA. Thus, the PAR-1 induced analgesia might possibly be explained by inhibition of adenyl cyclase as shown in Figure 4-5. In addition to PAR1, PAR4 has been identified as another protease activated receptor important for analgesia (Asfaha, Cenac et al. 2007). Intraplantar injection of a short synthetic peptide (GYPGKF) based on the tethered ligand sequence of PAR4 increased nociceptive threshold in response to thermal and mechanical noxious stimuli. Moreover, co-injection of GYPGKF with carrageenan significantly reduced the resulting hyperalgesia and allodynia. For the sake of simplicity PAR4 has been omitted from Figure 4-5. LOW pH Injection of acidic solutions (pH 5.0 6.0) cause intense burning pain (Issberner, Reeh et al. 1996) and a substantial decrease in synovial fluid pH (6.6-7.4) is found in inflamed joints (Treuhaft and McCarty 1971). There is no consensus concerning the identity of the specific molecular receptor that is activated by low pH in nociceptors. A decrease in extracellular pH has two effects on TRPV1: first extracellular protons lower the threshold for TRPV1 activation by capsaicin and heat, and second acidification below pH 6.0 directly opens the channel. It is unlikely that TRPV1 is the sole sensor for extracellular protons because individual nerve fibers in rat skin which fire action potentials in response to protons do not always fire in response to capsaicin. Moreover, in DRG neurons from TRPV1 knockout mice the response to protons was reduced but not eliminated (Caterina, Leffler et al. 2000). With the discovery and subsequent cloning of acid sensing ion channels (ASICs), they became candidates for the sensor of extracellular protons. Normally a significant fraction of DRG neurons from mice exhibit ASIC-like transient currents in response to protons (pH 5.0) in contrast, for transgenic mice expressing a dominant-negative form of ASIC3, none of the neurons exhibited ASIC-like transient currents (Mogil, Breese et al. 2005). Surprisingly, in behavioral tests the transgenic mice

4-17

were found to be more sensitive to a number of pain modalities than the wild type mice, making it unlikely that they are direct transducers of nociceptive stimuli. The sensitivity to protons of human dorsal root ganglion (hDRG) neurons was examined by rapid acidification of the extracellular fluid from pH 7.35 to 6.0 causing a prolonged depolarization of the membrane potential in all 40 cells tested (Baumann, Burchiel et al. 1996). Surprisingly, the membrane depolarization was associated with a decrease in membrane conductance in 27 of the 40 cells tested rather than the increase in conductance that would be expected with activation of TRPV1 or ASICs (see Figure 4-1). Subsequent ion substitution experiments indicated that the decrease in conductance upon acidification was due to a decrease of the background K+ permeability (Baumann, Chaudhary et al. 2004). In as much as K2P channels are thought to be responsible for the cells resting K+ permeability, it is possible that an acid sensitive K2P channel is involved. Two K2P channels (TASK-1 and TASK-2) and an inward rectifying K+ channel (Kir2.3) that are blocked by protons have been identified in DRG neurons (Baumann, Chaudhary et al. 2004). Inward rectifying K+ channels are not perfect rectifiers; they can pass some outward current in the voltage range above the resting potential, therefore their inhibition would be expected to enhance and prolong the membrane depolarization caused by other ion channels. The effects of protons considered above might involve the direct effect of protons on several different ion channels considered above (TRPV1, ASICs, TASK-1, TASK-2, Kir2.3). However, it is possible that the effects of protons are mediated by a receptor molecule that is actually separate from the ion channel that is being modulated. Recently, proton-sensing G-protein coupled receptors have been identified (Ludwig, Vanek et al. 2003), and subsequently shown to be present in the small-diameter DRG neurons responsible for nociception (Huang, Tzeng et al. 2007). Using a Xenopus oocyte expression system, external acidosis was shown to profoundly down-modulate human TREK-1 activity (Cohen, Sagron et al. 2009). The authors were able to distinguish a rapid and slow component of the decrease in TREK-1 currents resulting from exposure to low external pH. The fast component resulted from protonation of extracellular residues on TREK-1. While the slow component, of TREK-1 desensitization, was mediated by a proton-sensitive GPCR which appeared to activate phospholipase-C. There are a number of ways in which activation of phospholipase-C could potentially modulate TREK-1 activity (refer to Figure 4-2 for several examples). The original paper describing protonsensing G-protein coupled receptors showed that the receptor was inactive at pH 7.8, and fully activated at pH 6.8 (Ludwig, Vanek et al. 2003). Thus it only takes weak-to moderate extracellular acidification to fully activate the proton-sensitive GPCR, making them ideal receptors for the detection of extracellular acidification.

LYSOPHOSPHATIDIC ACID Lysophosphatidic acid (LPA, 1-acyl-sn-glycerol-3-phosphate), the simplest glycerolphospholipid, has one mole of fatty acid per mole of lipid: where the fatty acid can be either saturated or unsaturated, depending on the tissue. The pathways which might give

4-18

rise to the production of LPA are illustrated in Figure 4-6. LPA is rapidly produced and released from activated platelets and is more abundant in serum (1-5 M) than in plasma. Where plasma is the liquid portion of the blood that is separated from the blood cells and serum is the leftover fluid after agitating the plasma to precipitate a clot. Seven G proteincoupled receptors (LPA1 LPA7) specific for lysophosphatidic acid have so far been identified. Intraplantar injection of LPA in mice provokes painful responses (Renback, Inoue et al. 1999). Subsequently, the effects of LPA on TREK-1 ion channels were investigated using a Xenopus oocyte expression system (Cohen, Sagron et al. 2009). At a physiological concentration of 1 M, LPA dramatically decreased TREK-1 currents; the effect persisted for minutes and was reversible upon washout. Pre-treatment with U73122, an inhibitor of phospholipase-C, completely blocked the LPA-induced decrease in TREK-1 currents. Further experiments indicated that phospholipase-C was probably being activated by endogenous Xenopus LPA receptors. The effects of LPA on voltage gated sodium currents in rat dorsal root ganglion (DRG) neurons were also investigated (Seung Lee, Hong et al. 2005). LPA suppressed tetrodotoxin sensitive sodium currents while increasing tetrodotoxin insensitive sodium currents. The tetrodotoxin insensitive currents will be considered more fully in chapter 5 where we discuss voltage gated channels. NERVE GROWTH FACTOR Nerve growth factor (NGF) is a trophic factor that promotes the survival of nociceptors during development. For our purposes though, the important findings are that the highaffinity NGF receptor (trkA) is expressed on nociceptors in adult animals and that NGF levels are elevated in inflamed skin. Using a synthetic trkA-IgG fusion molecule to sequester NGF it was found that administration of trkA-IgG together with carrageenan blocked the hyperalgesia resulting from carrageenan induced inflammation (McMahon, Bennett et al. 1995). Additionally, the TrkA receptor mediates the hyperalgesia caused by NGF because NGF can still induce thermal hyperalgesia in mice in which the low affinity p75 neutrophin-receptor is knocked out (Bergmann, Reiter et al. 1998). These findings strongly suggest that NGF is involved in regulating the sensitivity to pain in adult animals. Moreover, thermal hyperalgesia of the hind paw developed within minutes of intraplantar NGF injection indicating that gene transcription was not involved (Lewin, Ritter et al. 1993). In order to examine the direct effects of NGF on nociceptors from adult animals experiments were carried out on dissociated DRG neurons in culture. A rapid enhancement of the capsaicin induced current was observed with NGF application in these isolated DRG neurons (Shu and Mendell 1999; Shu and Mendell 2001). These findings clearly indicate that one of the targets of NGF is TRPV1. Subsequent experiments, using DRG neurons and expression systems, to try and elucidate the signaling pathways involved in sensitization by NGF have so far yielded conflicting results.

4-19

PLD

PC

PIP2

PLC

phosphatidic acid PLA2

DAG kinase PA phosphohydrolase

DAG + IP3

DAG lipase monoacylglycerol kinase

lysophosphatidic acid

monoacylglycerol

PLD phospholipase D PC phosphatidyl choline PLC phospholipase C PIP2 phosphatidylinositol (4,5)-bisphosphate DAG diacylglycerol PA phosphatidic acid IP3 inositol 1,4,5-trisphosphate

Figure 4-6. Pathways for the production of lysophosphatidic acid.

4-20

SUMMARY In this chapter I have tried to review briefly in words the evidence for the chemical mediators thought to play a role in pain. In the diagram of Figure 4-7 I have summarized the ion channels and receptors (G-protein coupled receptors and receptor tyrosine kinase) thought to be involved in carrying out the actions of the chemical mediators considered in this chapter. Finally, in Table 4-1, for each chemical mediator I have listed the GPCR, the ion channels and the second messenger pathways thought to play a role.

4-21

TABLE 4-1
Ion Channels Cheimical Mediator Acute Pain Hyperalgesia GPCR Ionotropic Metabotropic TRPV1 Bradykinin yes yes B2 TRPA1 TRPV1 ATP yes yes P2Y2 P2X2, P2X3 Serotonin yes yes 5-HT2A 5-HT3 tryptase trypsin protons (low pH) yes yes yes yes PAR2 Proton Sensitive GPCR ASIC LPA PGE2 PGI2 12-HPETE NGF yes yes yes yes -----------yes yes ---yes LPA1 EP IP K2P (?TREK-1) TRPV1 TRPA1 TRPV1 TRPV1 TRPV1 PLC PKA PKA PLC, PKC ? TRPV1 TRPV4 K2P (?TREK-1) K2P (?TREK-1) TRPV1 PKA, PKC PKA, PKC, PKD PKC, ERK1/2 PLC K2P (?TREK-1) Second Messenger Pathways PLC- DAG PKC PIP2 PLA2 COX PGE2 PKA LOX 12-HPETE PLC- PIP2 IP3 Ca2+ PLA2 COX PGE2 PKA PLC- DAG PKC PIP2

4-22

IN

PGE 2
EP

P GI

TO N

BR

SE RO

IP

AD

YK

TR YP

TA S

5-

HT

2A

IN I

PA R2

TRPV1

pH G P CR

H+

TRPA1

LPA G P CR

LPA

TRPV4

K2P
ATP

P2Y2

trk

P2X 32X 3 /P 2X 2 P

H 5-

T3
NI O OT R SE N

Figure 4-7. Channels and receptors involved in nociceptor signal transduction

F NG
P AT

GPCR

ION CHANNEL

RECEPTOR TYROSINE KINASE

References cited:

Alloui, A., K. Zimmermann, et al. (2006). "TREK-1, a K+ channel involved in polymodal pain perception." Embo J 25(11): 2368-76. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16675954 Amadesi, S., G. S. Cottrell, et al. (2006). "Protease-activated receptor 2 sensitizes TRPV1 by protein kinase Cepsilon- and A-dependent mechanisms in rats and mice." J Physiol 575(Pt 2): 555-71. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16793902 Amadesi, S., J. Nie, et al. (2004). "Protease-activated receptor 2 sensitizes the capsaicin receptor transient receptor potential vanilloid receptor 1 to induce hyperalgesia." J Neurosci 24(18): 4300-12. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15128844 Asfaha, S., V. Brussee, et al. (2002). "Proteinase-activated receptor-1 agonists attenuate nociception in response to noxious stimuli." Br J Pharmacol 135(5): 1101-6. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11877315 Asfaha, S., N. Cenac, et al. (2007). "Protease-activated receptor-4: a novel mechanism of inflammatory pain modulation." Br J Pharmacol 150(2): 176-85. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17179954 Baumann, T. K., K. J. Burchiel, et al. (1996). "Responses of adult human dorsal root ganglion neurons in culture to capsaicin and low pH." Pain 65(1): 31-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8826487 Baumann, T. K., P. Chaudhary, et al. (2004). "Background potassium channel block and TRPV1 activation contribute to proton depolarization of sensory neurons from humans with neuropathic pain." Eur J Neurosci 19(5): 1343-51. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15016092 Bautista, D. M., S. E. Jordt, et al. (2006). "TRPA1 mediates the inflammatory actions of environmental irritants and proalgesic agents." Cell 124(6): 1269-82. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16564016

4-24

Bautista, D. M., P. Movahed, et al. (2005). "Pungent products from garlic activate the sensory ion channel TRPA1." Proc Natl Acad Sci U S A 102(34): 12248-52. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16103371 Bergmann, I., R. Reiter, et al. (1998). "Nerve growth factor evokes hyperalgesia in mice lacking the low-affinity neurotrophin receptor p75." Neurosci Lett 255(2): 87-90. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9835221 Caterina, M. J., A. Leffler, et al. (2000). "Impaired nociception and pain sensation in mice lacking the capsaicin receptor." Science 288(5464): 306-13. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10764638 Cesare, P., L. V. Dekker, et al. (1999). "Specific involvement of PKC-epsilon in sensitization of the neuronal response to painful heat." Neuron 23(3): 617-24. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10433272 Cesare, P. and P. McNaughton (1996). "A novel heat-activated current in nociceptive neurons and its sensitization by bradykinin." Proc Natl Acad Sci U S A 93(26): 15435-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8986829 Chemin, J., C. Girard, et al. (2003). "Mechanisms underlying excitatory effects of group I metabotropic glutamate receptors via inhibition of 2P domain K+ channels." Embo J 22(20): 5403-11. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14532113 Chuang, H. H., E. D. Prescott, et al. (2001). "Bradykinin and nerve growth factor release the capsaicin receptor from PtdIns(4,5)P2-mediated inhibition." Nature 411(6840): 957-62. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11418861 Cockayne, D. A., P. M. Dunn, et al. (2005). "P2X2 knockout mice and P2X2/P2X3 double knockout mice reveal a role for the P2X2 receptor subunit in mediating multiple sensory effects of ATP." J Physiol 567(Pt 2): 621-39. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15961431

4-25

Cockayne, D. A., S. G. Hamilton, et al. (2000). "Urinary bladder hyporeflexia and reduced pain-related behaviour in P2X3-deficient mice." Nature 407(6807): 10115. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11069181 Cohen, A., R. Sagron, et al. (2009). "Pain-associated signals, acidosis and lysophosphatidic acid, modulate the neuronal K(2P)2.1 channel." Mol Cell Neurosci 40(3): 382-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=19130888 Ferreira, J., G. L. da Silva, et al. (2004). "Contribution of vanilloid receptors to the overt nociception induced by B2 kinin receptor activation in mice." Br J Pharmacol 141(5): 787-94. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14967737 Grant, A. D., G. S. Cottrell, et al. (2007). "Protease-activated receptor 2 sensitizes the transient receptor potential vanilloid 4 ion channel to cause mechanical hyperalgesia in mice." J Physiol 578(Pt 3): 715-33. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17124270 Hamilton, S. G., J. Warburton, et al. (2000). "ATP in human skin elicits a dose-related pain response which is potentiated under conditions of hyperalgesia." Brain 123 ( Pt 6): 1238-46. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10825361 Huang, C. W., J. N. Tzeng, et al. (2007). "Nociceptors of dorsal root ganglion express proton-sensing G-protein-coupled receptors." Mol Cell Neurosci 36(2): 195-210. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17720533 Hwang, S. W., H. Cho, et al. (2000). "Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances." Proc Natl Acad Sci U S A 97(11): 6155-60. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10823958 Issberner, U., P. W. Reeh, et al. (1996). "Pain due to tissue acidosis: a mechanism for inflammatory and ischemic myalgia?" Neurosci Lett 208(3): 191-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8733302

4-26

Jarvis, M. F., E. C. Burgard, et al. (2002). "A-317491, a novel potent and selective nonnucleotide antagonist of P2X3 and P2X2/3 receptors, reduces chronic inflammatory and neuropathic pain in the rat." Proc Natl Acad Sci U S A 99(26): 17179-84. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12482951 Jordt, S. E., D. M. Bautista, et al. (2004). "Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1." Nature 427(6971): 260-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14712238 Kassuya, C. A., J. Ferreira, et al. (2007). "Intraplantar PGE2 causes nociceptive behaviour and mechanical allodynia: the role of prostanoid E receptors and protein kinases." Br J Pharmacol 150(6): 727-37. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17310141 Kawabata, A., N. Kawao, et al. (2001). "Peripheral PAR-2 triggers thermal hyperalgesia and nociceptive responses in rats." Neuroreport 12(4): 715-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11277570 Kawao, N., C. Shimada, et al. (2002). "Capsazepine inhibits thermal hyperalgesia but not nociception triggered by protease-activated receptor-2 in rats." Jpn J Pharmacol 89(2): 184-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12120762 Kayssi, A., S. Amadesi, et al. (2007). "Mechanisms of protease-activated receptor 2evoked hyperexcitability of nociceptive neurons innervating the mouse colon." J Physiol 580(Pt.3): 977-91. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17289784 Khasar, S. G., Y. H. Lin, et al. (1999). "A novel nociceptor signaling pathway revealed in protein kinase C epsilon mutant mice." Neuron 24(1): 253-60. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10677042 Kollarik, M. and B. J. Undem (2004). "Activation of bronchopulmonary vagal afferent nerves with bradykinin, acid and vanilloid receptor agonists in wild-type and TRPV1-/- mice." J Physiol 555(Pt 1): 115-23. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14634201

4-27

Kwan, K. Y., A. J. Allchorne, et al. (2006). "TRPA1 contributes to cold, mechanical, and chemical nociception but is not essential for hair-cell transduction." Neuron 50(2): 277-89. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16630838 Lewin, G. R., A. M. Ritter, et al. (1993). "Nerve growth factor-induced hyperalgesia in the neonatal and adult rat." J Neurosci 13(5): 2136-48. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8478693 Ludwig, M. G., M. Vanek, et al. (2003). "Proton-sensing G-protein-coupled receptors." Nature 425(6953): 93-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12955148 Lukacs, V., B. Thyagarajan, et al. (2007). "Dual regulation of TRPV1 by phosphoinositides." J Neurosci 27(26): 7070-80. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17596456 Malin, S. A., B. M. Davis, et al. (2008). "Thermal nociception and TRPV1 function are attenuated in mice lacking the nucleotide receptor P2Y(2)." Pain. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18343036 Manning, D. C., S. N. Raja, et al. (1991). "Pain and hyperalgesia after intradermal injection of bradykinin in humans." Clin Pharmacol Ther 50(6): 721-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=1752117 Maurer-Spurej, E., C. Pittendreigh, et al. (2004). "The influence of selective serotonin reuptake inhibitors on human platelet serotonin." Thromb Haemost 91(1): 119-28. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14691577 McMahon, S. B., D. L. Bennett, et al. (1995). "The biological effects of endogenous nerve growth factor on adult sensory neurons revealed by a trkA-IgG fusion molecule." Nat Med 1(8): 774-80. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=7585179 McNamara, C. R., J. Mandel-Brehm, et al. (2007). "TRPA1 mediates formalin-induced pain." Proc Natl Acad Sci U S A 104(33): 13525-30. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17686976

4-28

Mogil, J. S., N. M. Breese, et al. (2005). "Transgenic expression of a dominant-negative ASIC3 subunit leads to increased sensitivity to mechanical and inflammatory stimuli." J Neurosci 25(43): 9893-901. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16251436 Mogil, J. S., F. Miermeister, et al. (2005). "Variable sensitivity to noxious heat is mediated by differential expression of the CGRP gene." Proc Natl Acad Sci U S A 102(36): 12938-43. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16118273 Moriyama, T., T. Higashi, et al. (2005). "Sensitization of TRPV1 by EP1 and IP reveals peripheral nociceptive mechanism of prostaglandins." Mol Pain 1: 3. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15813989 Moriyama, T., T. Iida, et al. (2003). "Possible involvement of P2Y2 metabotropic receptors in ATP-induced transient receptor potential vanilloid receptor 1mediated thermal hypersensitivity." J Neurosci 23(14): 6058-62. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12853424 Murata, T., F. Ushikubi, et al. (1997). "Altered pain perception and inflammatory response in mice lacking prostacyclin receptor." Nature 388(6643): 678-82. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9262402 Okamoto, K., H. Imbe, et al. (2002). "5-HT2A receptor subtype in the peripheral branch of sensory fibers is involved in the potentiation of inflammatory pain in rats." Pain 99(1-2): 133-43. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12237191 Petho, G., A. Derow, et al. (2001). "Bradykinin-induced nociceptor sensitization to heat is mediated by cyclooxygenase products in isolated rat skin." Eur J Neurosci 14(2): 210-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11553274 Rathee, P. K., C. Distler, et al. (2002). "PKA/AKAP/VR-1 module: A common link of Gs-mediated signaling to thermal hyperalgesia." J Neurosci 22(11): 4740-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12040081

4-29

Renback, K., M. Inoue, et al. (1999). "Lysophosphatidic acid-induced, pertussis toxinsensitive nociception through a substance P release from peripheral nerve endings in mice." Neurosci Lett 270(1): 59-61. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10454146 Richardson, J. D. and M. R. Vasko (2002). "Cellular mechanisms of neurogenic inflammation." J Pharmacol Exp Ther 302(3): 839-45. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12183638 Rong, W., K. Hillsley, et al. (2004). "Jejunal afferent nerve sensitivity in wild-type and TRPV1 knockout mice." J Physiol 560(Pt 3): 867-81. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15331673 Sasaki, M., H. Obata, et al. (2006). "Peripheral 5-HT2A receptor antagonism attenuates primary thermal hyperalgesia and secondary mechanical allodynia after thermal injury in rats." Pain 122(1-2): 130-6. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16527395 Schmelz, M., G. Osiander, et al. (1997). "Intracutaneous injections of platelets cause acute pain and protracted hyperalgesia." Neurosci Lett 226(3): 171-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9175594 Schmelz, M. and L. J. Petersen (2001). "Neurogenic inflammation in human and rodent skin." News Physiol Sci 16: 33-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11390944 Schnizler, K., L. P. Shutov, et al. (2008). "Protein kinase A anchoring via AKAP150 is essential for TRPV1 modulation by forskolin and prostaglandin E2 in mouse sensory neurons." J Neurosci 28(19): 4904-17. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18463244 Seung Lee, W., M. P. Hong, et al. (2005). "Effects of lysophosphatidic acid on sodium currents in rat dorsal root ganglion neurons." Brain Res 1035(1): 100-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15713282 Shin, J., H. Cho, et al. (2002). "Bradykinin-12-lipoxygenase-VR1 signaling pathway for inflammatory hyperalgesia." Proc Natl Acad Sci U S A 99(15): 10150-5.

4-30

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12097645 Shu, X. and L. M. Mendell (1999). "Nerve growth factor acutely sensitizes the response of adult rat sensory neurons to capsaicin." Neurosci Lett 274(3): 159-62. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10548414 Shu, X. and L. M. Mendell (2001). "Acute sensitization by NGF of the response of smalldiameter sensory neurons to capsaicin." J Neurophysiol 86(6): 2931-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11731549 Souslova, V., P. Cesare, et al. (2000). "Warm-coding deficits and aberrant inflammatory pain in mice lacking P2X3 receptors." Nature 407(6807): 1015-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11069182 Southall, M. D. and M. R. Vasko (2001). "Prostaglandin receptor subtypes, EP3C and EP4, mediate the prostaglandin E2-induced cAMP production and sensitization of sensory neurons." J Biol Chem 276(19): 16083-91. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11278900 Steinhoff, M., N. Vergnolle, et al. (2000). "Agonists of proteinase-activated receptor 2 induce inflammation by a neurogenic mechanism." Nat Med 6(2): 151-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10655102 Sugiura, T., M. Tominaga, et al. (2002). "Bradykinin lowers the threshold temperature for heat activation of vanilloid receptor 1." J Neurophysiol 88(1): 544-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12091579 Todorovic, S. M., R. S. Scroggs, et al. (1997). "Cationic modulation of 5-HT2 and 5-HT3 receptors in rat sensory neurons: the role of K+, Ca2+ and Mg2+." Brain Res 765(2): 291-300. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9313902 Tokunaga, A., M. Saika, et al. (1998). "5-HT2A receptor subtype is involved in the thermal hyperalgesic mechanism of serotonin in the periphery." Pain 76(3): 34955. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9718253

4-31

Treuhaft, P. S. and D. McCarty (1971). "Synovial fluid pH, lactate, oxygen and carbon dioxide partial pressure in various joint diseases." Arthritis Rheum 14(4): 475-84. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=5564921 Vergnolle, N., N. W. Bunnett, et al. (2001). "Proteinase-activated receptor-2 and hyperalgesia: A novel pain pathway." Nat Med 7(7): 821-6. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11433347 Wang, S., Y. Dai, et al. (2008). "Phospholipase C and protein kinase A mediate bradykinin sensitization of TRPA1: a molecular mechanism of inflammatory pain." Brain 131(Pt 5): 1241-51. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18356188 Watanabe, H., J. B. Davis, et al. (2002). "Activation of TRPV4 channels (hVRL2/mTRP12) by phorbol derivatives." J Biol Chem 277(16): 13569-77. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11827975 Wei, H., Y. Chen, et al. (2005). "The contribution of peripheral 5-hydroxytryptamine2A receptor to carrageenan-evoked hyperalgesia, inflammation and spinal Fos protein expression in the rat." Neuroscience 132(4): 1073-82. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15857711 Zeitz, K. P., N. Guy, et al. (2002). "The 5-HT3 subtype of serotonin receptor contributes to nociceptive processing via a novel subset of myelinated and unmyelinated nociceptors." J Neurosci 22(3): 1010-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11826129 Zurborg, S., B. Yurgionas, et al. (2007). "Direct activation of the ion channel TRPA1 by Ca2+." Nat Neurosci 10(3): 277-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17259981

4-32

Chapter 5 Na+, K+, Ca++ and HCN Channels After the noxious stimulus is transduced into a receptor potential the response must be transformed or encoded into a train of action potentials, which carry the signal to the synaptic terminal culminating in neurotransmitter release (i.e. information is conveyed to the spinal cord). It is during this process that voltage gated sodium, potassium, calcium and hyperpolarization-activated, cation nonselective (HCN) channels make their contribution. To the extent that these channels are enriched in nociceptors in comparison to other regions of the nervous system and the body they may serve as targets for the development of analgesic or anesthetic drugs. Na+ CHANNELS Voltage-gated sodium channels are essential for encoding the receptor potential into a series of action potentials and for conducting the action potentials along the axon. Voltage-gated sodium channels are composed of a pore-forming alpha subunit and at least one auxiliary beta subunit. -subunits are multifunctional: they modulate channel gating, regulate the level of channel expression, and function as cell adhesion molecules interacting with the extracellular matrix and the cytoskeleton. The family of pore-forming -subunits has nine known members named Nav1.1 through Nav1.9; not all of these are present in DRG neurons. Neurons differ in the shape of their action potentials and also in the rate and regularity at which they fire action potentials. Generally speaking nociceptors fire action potentials having a longer duration (several milliseconds) and a relatively slow rate typically in a range less than 10 per second. One can distinguish two general classes of voltage-gated sodium channels based on their sensitivity to tetrodotoxin (TTX) a potent neurotoxin that blocks action potentials in nerves by binding to the pore of voltage-gated sodium channel -subunits. Not all subunits are sensitive to TTX, therefore we can distinguish between TTX sensitive (TTXS) and TTX resistant (TTX-R) sodium channels. Nociceptive sensory neurons have been shown to express both TTX-R and TTX-S sodium channels. Of the voltage-gated sodium channels, the TTX-S channels Nav1.1, Nav1.3, Nav1.6 and Nav1.7, and the TTX-R channels Nav1.8 and Nav1.9, have been implicated in the functioning of nociceptors in both normal and pathological states. Characterization of the contribution of specific voltage-gated sodium channels to the functioning of DRG neurons is limited by the lack of selective channel blockers. Generation of knockout mice for specific voltage-gated sodium channels provides an alternative solution to this problem and these animals can be characterized behaviorally. Furthermore electrophysiological recordings can be made from these animals to further characterize the contribution of specific channels to the detection of painful stimuli. The TTX-R voltage gated sodium channels Nav1.8 and Nav1.9 are expressed predominantly in small DRG neurons, which include nociceptive cells, and it has been

5-1

suggested that they play an important role in pain mechanisms. Additionally the preferential expression of Nav1.8 and Nav1.9 in small DRG neurons suggests that these channels are good targets for the pharmacologic treatment of pain. Nav1.8 Nav 1.8 is expressed predominantly in small diameter nociceptive afferents and Nav 1.8 knockout mice exhibit a pronounced increase in the threshold to noxious mechanical stimuli (Akopian, Souslova et al. 1999; Matthews, Wood et al. 2006). Mice in which diphtheria toxin A was used to kill neurons expressing Nav1.8 (DTA mice) also showed a pronounced increase in the threshold to noxious mechanical stimuli (Abrahamsen, Zhao et al. 2008). Interestingly, in these studies responses to noxious heat were similar to those in normal mice and the development of inflammatory hyperalgesia was delayed. These findings might be explained if Nav 1.8 voltage-gated sodium channels were specifically localized to mechano sensitive nociceptors and determined their excitability. In this context TTX-R sodium channels have been shown to be present at sufficiently high densities in the peripheral terminals of nociceptors to be the determinant of their excitability (Brock, McLachlan et al. 1998). Alternatively, knockout of Nav 1.8 might disrupt the mechano sensing apparatus located in the nociceptor terminal. Further studies of the exact role of Nav 1.8 in mechano sensitive nociception and the development of highly specific Nav 1.8 channel blockers would help to clarify the situation. In addition to the increase in the threshold to noxious mechanical stimuli, Nav 1.8 knockout mice also exhibited a loss of sensitivity to noxious cold stimuli (Matthews, Wood et al. 2006; Zimmermann, Leffler et al. 2007). The loss of sensitivity to noxious cold probably occurs because Nav 1.8 appears to be the only voltage gated sodium channel that remains functional at very cold temperatures (Zimmermann, Leffler et al. 2007). Additionally in the cold much lower currents were needed to trigger Nav1.8 generated action potentials than at 30 C for TTX treated nociceptor terminals (Zimmermann, Leffler et al. 2007). The decreased threshold for triggering action potentials at reduced temperatures probably results from an increase in the input membrane resistance due to the closure of background potassium (K2P) channels (Maingret, Lauritzen et al. 2000; Reid and Flonta 2001). Remember as discussed in chapter 4 (see Figure 4-1) the closure of background potassium channels will depolarize the membrane and furthermore it will also increase the input membrane resistance, thus less inward current initiated by noxious stimuli will be needed to trigger action potentials. With respect to the TRP channels thought to be cold sensors, DTA mice (Abrahamsen, Zhao et al. 2008) exhibit a reduced expression of the cold sensor TRPA1 and they exhibit an almost complete suppression of the second phase of the formalin response, which has been ascribed to activation of TRPA1 (McNamara, Mandel-Brehm et al. 2007). In contrast the expression of the cold sensor, TRPM8 appears to be normal in DTA mice. Based on these results one can make a tentative model for the transduction of noxious cold in mouse mechano-cold nociceptors as illustrated in Figure 5-1. Exposure to a noxious cold temperature around 0 C results in the activation of TRPA1 and the inhibition of the background potassium channels (K2P), the resulting depolarization excites Nav 1.8 and action potentials propagate to the spinal cord. In the model of Figure

5-2

TRPA1
opens

NaV 1.8
action potentials

cold

K2P
closes

excites

Figure 5-1. In a mechano-cold nociceptor exposure to a cold temperature of around OOC opens TRPA1channels and closes K2P channels, which in turn activate NaV1.8 channels thereby exciting the nociceptor.

5-3

5-1 the mechano sensing apparatus, which is upstream from Nav 1.8 has been left out for the purpose of simplicity. Nav1.9 Most DRG neurons that express Nav 1.8 also express Nav 1.9, although a very small number of DRG cells express either only Nav 1.8 or Nav 1.9 (Amaya, Decosterd et al. 2000). For mice in which Nav 1.9 has been knocked out the behavioral response to noxious mechanical, noxious heat and noxious cold stimulation were no different than in wild type mice (Priest, Murphy et al. 2005; Amaya, Wang et al. 2006). Additionally, experiments, using a skin nerve preparation, indicated that mechanical and thermal pain thresholds were the same for NaV 1.9 knockout mice and wild type mice (Priest, Murphy et al. 2005). In contrast hypersensitivity resulting from peripheral inflammation produced by intra plantar injection of complete Freunds adjuvant is substantially reduced in the knockout mice (Amaya, Wang et al. 2006), as is the hypersensitivity resulting from intra plantar injection of formalin or carrageenan (Priest, Murphy et al. 2005). The threshold for painful stimuli has been shown to be reduced by multiple inflammatory mediators (such as PGE2, bradykinin, ATP and 5-HT) acting through multiple intracellular signaling pathways (for example PLC-, PKA and PKC) (see chapter 4). In chapter 4 evidences were presented that these intracellular pathways targeted the ion channels TRPV1, TRPA1, TRPV4 and K2P to produce the hypersensitivity. The evidence presented above suggests that NaV 1.9 might also be a target for these inflammatory mediators and intracellular signaling pathways. In this context thermal hypersensitivity resulting from intra plantar injection of PGE2 was significantly reduced in NaV 1.9 knockout mice compared to wild type mice (Priest, Murphy et al. 2005; Amaya, Wang et al. 2006). There is as yet no consensus as to the intracellular signaling pathway that targets NaV 1.9. Because the inflammatory mediators PGE2, bradykinin, ATP and 5-HT mentioned above act through G-protein coupled receptors the hypothesis that activation of G-proteins was required to regulate NaV 1.9 was tested (Baker, Chandra et al. 2003). They found that the amplitude of a TTX-R Na+ current, attributed to NaV 1.9, recorded from small DRG neurons having less than a 25 m diameter, was increased more than 3fold by 500 m intracellular GTP- S, a non-hydrolysable analog of GTP. Presumably, GTP- S is working by binding to a G-protein and keeping the G-protein to which it is bound in the active state because the GTP- S is resistant to hydrolysis to GDP (see Chapter 4). A later study showed that the TTX-R Na+ current being up regulated was indeed NaV 1.9 (Ostman, Nassar et al. 2008). To reiterate, although these findings indicate the involvement of a G-protein they do not point to the intracellular signaling pathway that is activated by the G-protein. One possibility is that the activated G-protein, with bound GTP- S, works by interacting directly with NaV 1.9. Nav 1.7 In 2006 a group of individuals from three families were reported upon, who exhibited a congenital inability to perceive pain and were otherwise apparently normal (Cox, Reimann et al. 2006). The loss of the ability to sense pain was shown to result from nonsense mutations in the gene SCN9A, which encodes the -subunit of the TTX-S 5-4

voltage-gated sodium channel Nav1.7 that is expressed at high levels in nociceptive DRG neurons. Before these findings in humans Nav 1.7 knockout mice were found to die shortly after birth, therefore Nav 1.7 had to be deleted in a subset of neurons, therefore yielding less useful information than the findings from the global knockout of Nav1.7 in humans (Nassar, Stirling et al. 2004). Obviously Nav 1.7 is an excellent target for the development of new analgesics for the treatment of pain. There is now also evidence that point mutations in SCN9A can result in an increase in pain sensations. Patients, with the painful inherited neuropathy, inherited erythromelalgia (sometimes called erythermalgia), experience episodes of severe chronic burning pain, skin redness, and swelling of the extremities, ears and face. Patients with primary erythermalgia have mutations in SCN9A which encodes the -subunit of Nav1.7 (Yang, Wang et al. 2004). The electrophysiological properties of mutant Nav1.7 channels having the same mutations found in patients with erythermalgia have been studied (Cummins, Dib-Hajj et al. 2004). The mutant channels exhibited hyper-excitability brought about by a hyperpolarizing shift in activation and by a slowing in deactivation and inactivation. Nav 1.3 Neuropathic pain, or "neuralgia", sometimes develops following injury or disease that damages a nerve. A variety of changes occur both in nociceptors and also in the central nervous system after nerve damage, among these changes hyper-excitability of DRG neurons is well documented after injury to DRG peripheral axons. It has been proposed that hyper excitable DRG neurons might contribute to neuropathic pain (for example see (Devor 2001)) and that upregulation of Nav 1.3 contributes to the abnormal hyper excitability of injured DRG neurons (Black, Cummins et al. 1999). Nav 1.3 is a TTX-S channel expressed in neurons throughout the embryonic nervous system that is substantially down regulated in adult animals. However, normal levels of neuropathic pain behavior develops in Nav 1.3 knockout mice (Nassar, Baker et al. 2006) suggesting that Nav 1.3 expression is not necessary for the development of neuropathic pain. Nav 1.1 and Nav 1.6 Although Nav 1.1 and Nav 1.6 are expressed in nociceptors their specific roles if any, in nociception and pain sensation is as yet not clear. K+ CHANNELS K2P channels In chapter 4 we considered a number of situations in which chemical mediators such as serotonin, tryptase etc. appear to excite (depolarize) nociceptors by blocking the resting background K+ permeability (K2P channels). There is yet another way in which these background K+ channels may play a role in nociception, it turns out that the members of the TREK subfamily (TREK-1, TREK-2 and TRAAK) of the K2P channels are activated

5-5

by both mechanical and thermal stimuli (for example see (Kang, Choe et al. 2005)). Moreover, these channels are colocalized with the TRP channels, TRPV1, TRPV2 and TRPM8 in trigeminal ganglion neurons (Yamamoto, Hatakeyama et al. 2009). The activity of TREK-1, TREK-2 and TRAAK is very low at around 24 C and they become very active at 37 C (Kang, Choe et al. 2005). Additionally membrane stretch greatly increases the activity of TREK-1, TREK-2 and TRAAK at 37 C (Kang, Choe et al. 2005). The activity of all three channels continues to increase up to 42 C the maximum temperature for which accurate measurements could be made in the expression systems used. Remember from chapter 3 that TRPV1 is activated at temperatures above 42 C. Assuming that the activity of these K2P channels continues to increase above 42 C then their activation in nociceptors would tend to counteract excitation by increasing the membrane K+ permeability. Consequently, it was suggested that painful mechanical and thermal stimuli in nociceptors would be expected to activate members of the TREK subfamily of K2P channels, thereby tuning or counteracting excitation (Honore 2007; Noel, Zimmermann et al. 2009). That is, activation of a nociceptor by noxious mechanical or thermal stimuli is a balance between depolarization caused by activation of a mechano or thermo sensitive excitatory ion channels and hyperpolarization caused by activation of a TREK K2P channel family members. ATP-sensitive K+ channels For the reason that early work showed that glibenclamide, a blocker of ATP-dependent potassium channels (KATP), antagonizes morphine analgesia (Ocana, Del Pozo et al. 1990) this section begins with a discussion of morphine and its receptors. The pain relieving properties of morphine, the principal active ingredient in opium has been known for centuries. Morphines analgesic effect is primarily attributed to the activation of opioid receptors within the central nervous system. However, morphine acts both in the central nervous system and the peripheral nervous system to cause analgesia. For our purposes we are only interested in its action on nociceptors. A variety of experiments have suggested the existence of at least three types of opioid receptors, , and . Morphine's analgesic effects are dramatically reduced in mice lacking the opioid receptor suggesting that -receptors primarily mediate analgesia (Loh, Liu et al. 1998). Additionally, in behavioral studies in the rat the analgesic effects of peripherally administered morphine also appear to be mediated by the -receptor but are not readily detectable in normal tissue; the effects are only apparent during hyperalgesia when receptors are sensitized (Levine and Taiwo 1989). Utilizing a rat skin-nerve preparation the effects of morphine on the response properties of single nociceptors innervating normal and inflamed skin have been compared (Wenk, Brederson et al. 2006). Morphine had no significant effect on the response of nociceptors to mechanical or thermal stimuli in normal skin. However, nociceptors innervating inflamed skin exhibit lower thresholds for noxious mechanical stimuli and responses to noxious mechanical and thermal stimuli were elevated, and peripherally administered morphine inhibited the activity of cutaneous nociceptors under these conditions of inflammation. These findings indicate that morphine acting on opioid receptors located in the sensory transduction region of the nociceptor (see Figure 1-3) mediate analgesia during local inflammation.

5-6

It has been known for some time that high blood glucose levels antagonize morphine analgesia and it was suggested that the effect might be due to elevated intracellular ATP levels (Singh, Chatterjee et al. 1983). Subsequently, as mentioned above, it was shown that glibenclamide, a blocker of ATP-dependent potassium channels (KATP), antagonizes morphine induced analgesia in mice (Ocana, Del Pozo et al. 1990), suggesting that morphine acts by activating KATP channels. ATP-dependent potassium channels are a type of potassium channel composed of four regulatory sulfonylurea receptor (SUR) subunits and four ATP sensitive pore forming inwardly rectifying potassium channel (Kir) subunits. KATP channels act as metabolic sensors and when intracellular ATP levels are high the channels are closed. In capsaicin sensitive rat DRG neurons, diazoxide a KATP channel agonist, reversed the sensitization produced by PGE2, indicating that activation of KATP can reverse the enhanced excitability of DRG neurons (Chi, Jiang et al. 2007). The opioid receptors are G-protein coupled receptors, which raises the question of which intracellular second messenger pathway mediates the action of morphine on KATP. It has been suggested that morphine activates KATP via the nitric oxide-cGMP-PKG pathway (Granados-Soto, Rufino et al. 1997; Sachs, Cunha et al. 2004). Outward K+ channels Remember that the closing of background potassium channels will, in addition to depolarizing the membrane, will also increase the input membrane resistance; thus less exogenous current will be needed to trigger action potentials. Another way to increase the effectiveness of inward currents initiated by noxious stimuli in triggering action potentials is to inhibit depolarization activated potassium currents (KV). In rat DRG neurons PGE2 and the stable PGI2 analog carbaPGI2 suppressed the sustained type of voltage gated outward K+ current (Nicol, Vasko et al. 1997). Ca++ CHANNELS Opioid -receptors have also been localized to the synaptic terminal of DRG neurons in the spinal cord (Besse, Lombard et al. 1990; Arvidsson, Riedl et al. 1995). Additionally, morphine applied to the spinal cord reduced substance P release evoked by sciatic nerve stimulation (Go and Yaksh 1987). Moreover, opioids were found to suppress excitatory but not inhibitory synaptic transmission into adult rat spinal cord neurons (Kohno, Kumamoto et al. 1999). Taken together these findings imply that opioids play a role in regulating transmitter release from nociceptor terminals. Two different mechanisms could contribute to the presynaptic inhibition of transmitter release by opioids. First, as discussed above morphine could activate KATP channels in the terminal, by this means hyperpolarizing the presynaptic terminal and decreasing the terminals input resistance. This would reduce the ability of invading action potentials to depolarize the terminal and activate the calcium channels which in turn would reduce the resulting calcium influx and consequent transmitter release. Second, the opioids could directly inhibit the voltage gated calcium channels in the synaptic terminal. Using Xenopus oocytes expressing neuronal calcium channels and opioid receptors, activation of the morphine receptor with a synthetic enkephalin (DAMGO) resulted in a rapid inhibition of calcium currents (Bourinet, Soong et al. 1996). In rat DRG neurons -opioid receptors are negatively

5-7

coupled to three types of high-threshold calcium channels thought to play a role in synaptic transmission (Rusin and Moises 1995). It has been suggested that the opioid receptors inhibit the calcium channels via a direct action of the G-protein on the calcium channel and not via a second messenger pathway (see review by (Dolphin 1995)). In conclusion the available evidence is consistent with opiates inhibiting transmitter release by activating KATP channels and/or inhibiting calcium channels in the synaptic terminal. HCN Channels The membrane currents resulting from the activation of hyperpolarization-activated, cation nonselective (HCN) channels were first described in isolated DRG neurons more than 25 years ago (Mayer and Westbrook 1983). It wasnt until the last few years that the possible role of HCN channels in peripheral neuropathic pain (or peripheral neuralgia) became apparent (see review by (Jiang, Sun et al. 2008)). Neuropathic pain is discussed in the following chapter 6; here we will consider the general properties of HCN channels, their role in regulating firing frequency in some DRG neurons and also those properties that will prove useful in understanding their possible role in neuropathic pain. HCN channels are activated by hyperpolarizing voltage steps to potentials more negative than, -50 to -60 mV, which is near the resting potentials of most cells. These hyperpolarization-activated cation currents, Ih, were originally termed If for funny and Iq for queer because unlike the majority of voltage gated channels they were activated by hyperpolarization rather than depolarization. As shown in Figure 5-2A a hyperpolarizing voltage step activates a slowly developing inward current, Ih, the amplitude of which increases with increasing hyperpolarization (not shown). If the magnitude of Ih is plotted against membrane potential (I/V relationship) it is apparent that there is a region of anomalous inward rectification. By inward rectification it is meant that the channels pass current (positive charge) more easily into the cell (inward direction) and poorly in the outward direction. When first discovered inward rectification was called anomalous rectification to distinguish it from the more commonly encountered outwardly rectifying currents. HCN channels are not the only inwardly rectifying channels; remember that KATP channels have pore forming inwardly rectifying potassium channel (Kir) subunits. Ih channels are permeable to both Na+ and K+ ions while Kir channels are permeable to K+ ions. On the one hand, activation of Kir will tend to inhibit the firing of action potentials, because the membrane potential will move towards EK ( 85 mV). In contrast activation of HCN channels will tend to excite the nociceptor: depending on the permeability ratio of PK/PNa the membrane potential will move towards a value between EK and ENa, which will depolarize the membrane toward the threshold for firing action potentials. As illustrated in Figure 5-2B, Ih is an inward current activated by hyperpolarization beyond the resting potential, which is manifest as the depolarizing sag of the membrane potential during hyperpolarizing current. Since the Ih current does not inactivate at a given voltage, a sustained inward Ih current will play a role in determining the resting potential and input resistance. Because the net membrane current at the resting potential is zero, the Ih inward current must be balanced by an outward current. This current might

5-8

membrane potential (mV)

-20 -40

-100

membrane current (pA)

Ih inactivate
100

Ih activate
0.5 s

B
membrane current membrane potential (mV) -65

Ih inactivate

-75

sag

-85

Ih activate

Figure 5.2. Ih activation and inactivation under voltage clamp (A) and current clamp (B) conditions. In (A) Ih is activated by a voltage step from -40 to -100 mV and inactivates when the voltage returns to -40mV. In (B) a hyperpolarizing current pulse activates Ih causing a depolarizing sag during the membrane hyperpolarization.
5-9

be contributed by a resting background K+ current carried by K2P channels. Participation of Ih in determining the resting potential has the consequence that it may play a role in determining the neurons excitability. When Ih is present in a cell together with a low threshold voltage gated sodium channel and an appropriate amount of background potassium channels there is the potential for the rhythmic firing of action potentials. At a permissive resting potential the after hyperpolarization following an action potential may be sufficient to activate Ih which may be sufficient to depolarize the membrane back to the action potential threshold (Robinson and Siegelbaum 2003). It is this potential for rhythmic firing that is thought to play a role in neuropathic pain (see chapter 6). A family of four mammalian genes encodes the HCN14 channel subunits responsible for Ih currents. Each HCN subunit is composed of six transmembrane segments, with a voltage sensor and a pore forming P region. One important feature of HCN channels is their cyclic nucleotide binding domain which underlies their regulation by cAMP and does not require protein phosphorylation. HCN2 and HCN4 are strongly modulated by increased concentrations of cAMP, with the voltage of activation shifted up to more positive potentials by 1020 mV, while HCN1 and HCN3 channels are relatively insensitive to cAMP. Consequently, with increased cAMP a hyperpolarizing voltage step activates HCN2 and HCN4 more completely. Remember from chapter 4 that PGE2 and PAR2 activate PKA via a rise in cAMP. This raises the possibility that these chemical mediators might sensitize nociceptors by shifting the voltage of activation of HCN2 or HCN4 channels via a rise in cAMP. This possibility was tested in DRG neurons using a variety of techniques including ZD7288, a potent Ih blocker (not selective among HCN1-4) (Momin, Cadiou et al. 2008). These authors found that there was a population of DRG neurons, with smaller diameter cell bodies, that had a cAMP-sensitive Ih. Moreover, they found for these small nociceptive neurons that elevating cAMP levels shifted the voltage activation curve of Ih to more depolarized potentials and caused a steady depolarization of the resting membrane potential which was blocked by ZD7288. These findings suggest that modulation of Ih via a rise in cAMP plays an important role in the nociceptor sensitization caused by PGE2 (Momin, Cadiou et al. 2008).

References cited: Abrahamsen, B., J. Zhao, et al. (2008). "The cell and molecular basis of mechanical, cold, and inflammatory pain." Science 321(5889): 702-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18669863

5-10

Akopian, A. N., V. Souslova, et al. (1999). "The tetrodotoxin-resistant sodium channel SNS has a specialized function in pain pathways." Nat Neurosci 2(6): 541-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10448219 Amaya, F., I. Decosterd, et al. (2000). "Diversity of expression of the sensory neuronspecific TTX-resistant voltage-gated sodium ion channels SNS and SNS2." Mol Cell Neurosci 15(4): 331-42. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10845770 Amaya, F., H. Wang, et al. (2006). "The voltage-gated sodium channel Na(v)1.9 is an effector of peripheral inflammatory pain hypersensitivity." J Neurosci 26(50): 12852-60. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17167076 Arvidsson, U., M. Riedl, et al. (1995). "Distribution and targeting of a mu-opioid receptor (MOR1) in brain and spinal cord." J Neurosci 15(5 Pt 1): 3328-41. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=7751913 Baker, M. D., S. Y. Chandra, et al. (2003). "GTP-induced tetrodotoxin-resistant Na+ current regulates excitability in mouse and rat small diameter sensory neurones." J Physiol 548(Pt 2): 373-82. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12651922 Besse, D., M. C. Lombard, et al. (1990). "Pre- and postsynaptic distribution of mu, delta and kappa opioid receptors in the superficial layers of the cervical dorsal horn of the rat spinal cord." Brain Res 521(1-2): 15-22. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=2169958 Black, J. A., T. R. Cummins, et al. (1999). "Upregulation of a silent sodium channel after peripheral, but not central, nerve injury in DRG neurons." J Neurophysiol 82(5): 2776-85. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10561444 Bourinet, E., T. W. Soong, et al. (1996). "Determinants of the G protein-dependent opioid modulation of neuronal calcium channels." Proc Natl Acad Sci U S A 93(4): 1486-91. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8643659

5-11

Brock, J. A., E. M. McLachlan, et al. (1998). "Tetrodotoxin-resistant impulses in single nociceptor nerve terminals in guinea-pig cornea." J Physiol 512 ( Pt 1): 211-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9729630 Chi, X. X., X. Jiang, et al. (2007). "ATP-sensitive potassium currents reduce the PGE2mediated enhancement of excitability in adult rat sensory neurons." Brain Res 1145: 28-40. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17320840 Cox, J. J., F. Reimann, et al. (2006). "An SCN9A channelopathy causes congenital inability to experience pain." Nature 444(7121): 894-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17167479 Cummins, T. R., S. D. Dib-Hajj, et al. (2004). "Electrophysiological properties of mutant Nav1.7 sodium channels in a painful inherited neuropathy." J Neurosci 24(38): 8232-6. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15385606 Devor, M. (2001). "Neuropathic pain: what do we do with all these theories?" Acta Anaesthesiol Scand 45(9): 1121-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11683663 Dolphin, A. C. (1995). "The G.L. Brown Prize Lecture. Voltage-dependent calcium channels and their modulation by neurotransmitters and G proteins." Exp Physiol 80(1): 1-36. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=7734129 Go, V. L. and T. L. Yaksh (1987). "Release of substance P from the cat spinal cord." J Physiol 391: 141-67. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=2451003 Granados-Soto, V., M. O. Rufino, et al. (1997). "Evidence for the involvement of the nitric oxide-cGMP pathway in the antinociception of morphine in the formalin test." Eur J Pharmacol 340(2-3): 177-80. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9537812

5-12

Honore, E. (2007). "The neuronal background K2P channels: focus on TREK1." Nat Rev Neurosci 8(4): 251-61. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17375039 Jiang, Y. Q., Q. Sun, et al. (2008). "Characteristics of HCN channels and their participation in neuropathic pain." Neurochem Res 33(10): 1979-89. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18461446 Kang, D., C. Choe, et al. (2005). "Thermosensitivity of the two-pore domain K+ channels TREK-2 and TRAAK." J Physiol 564(Pt 1): 103-16. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15677687 Kohno, T., E. Kumamoto, et al. (1999). "Actions of opioids on excitatory and inhibitory transmission in substantia gelatinosa of adult rat spinal cord." J Physiol 518 ( Pt 3): 803-13. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10420016 Levine, J. D. and Y. O. Taiwo (1989). "Involvement of the mu-opiate receptor in peripheral analgesia." Neuroscience 32(3): 571-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=2557556 Loh, H. H., H. C. Liu, et al. (1998). "mu Opioid receptor knockout in mice: effects on ligand-induced analgesia and morphine lethality." Brain Res Mol Brain Res 54(2): 321-6. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9555078 Maingret, F., I. Lauritzen, et al. (2000). "TREK-1 is a heat-activated background K(+) channel." Embo J 19(11): 2483-91. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10835347 Matthews, E. A., J. N. Wood, et al. (2006). "Nav 1.8-null mice show stimulus-dependent deficits in spinal neuronal activity." Mol Pain 2: 5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16478543 Mayer, M. L. and G. L. Westbrook (1983). "A voltage-clamp analysis of inward (anomalous) rectification in mouse spinal sensory ganglion neurones." J Physiol 340: 19-45.

5-13

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=6887047 McNamara, C. R., J. Mandel-Brehm, et al. (2007). "TRPA1 mediates formalin-induced pain." Proc Natl Acad Sci U S A 104(33): 13525-30. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17686976 Momin, A., H. Cadiou, et al. (2008). "Role of the hyperpolarization-activated current Ih in somatosensory neurons." J Physiol 586(Pt 24): 5911-29. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18936078 Nassar, M. A., M. D. Baker, et al. (2006). "Nerve injury induces robust allodynia and ectopic discharges in Nav1.3 null mutant mice." Mol Pain 2: 33. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17052333 Nassar, M. A., L. C. Stirling, et al. (2004). "Nociceptor-specific gene deletion reveals a major role for Nav1.7 (PN1) in acute and inflammatory pain." Proc Natl Acad Sci U S A 101(34): 12706-11. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15314237 Nicol, G. D., M. R. Vasko, et al. (1997). "Prostaglandins suppress an outward potassium current in embryonic rat sensory neurons." J Neurophysiol 77(1): 167-76. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=9120557 Noel, J., K. Zimmermann, et al. (2009). "The mechano-activated K+ channels TRAAK and TREK-1 control both warm and cold perception." Embo J 28(9): 1308-18. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=19279663 Ocana, M., E. Del Pozo, et al. (1990). "An ATP-dependent potassium channel blocker antagonizes morphine analgesia." Eur J Pharmacol 186(2-3): 377-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=2127024 Ostman, J. A., M. A. Nassar, et al. (2008). "GTP up-regulated persistent Na+ current and enhanced nociceptor excitability require NaV1.9." J Physiol 586(4): 1077-87. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18096591

5-14

Priest, B. T., B. A. Murphy, et al. (2005). "Contribution of the tetrodotoxin-resistant voltage-gated sodium channel NaV1.9 to sensory transmission and nociceptive behavior." Proc Natl Acad Sci U S A 102(26): 9382-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15964986 Reid, G. and M. Flonta (2001). "Cold transduction by inhibition of a background potassium conductance in rat primary sensory neurones." Neurosci Lett 297(3): 171-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11137755 Robinson, R. B. and S. A. Siegelbaum (2003). "Hyperpolarization-activated cation currents: from molecules to physiological function." Annu Rev Physiol 65: 45380. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12471170 Rusin, K. I. and H. C. Moises (1995). "mu-Opioid receptor activation reduces multiple components of high-threshold calcium current in rat sensory neurons." J Neurosci 15(6): 4315-27. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=7540671 Sachs, D., F. Q. Cunha, et al. (2004). "Peripheral analgesic blockade of hypernociception: activation of arginine/NO/cGMP/protein kinase G/ATP-sensitive K+ channel pathway." Proc Natl Acad Sci U S A 101(10): 3680-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14990791 Singh, I. S., T. K. Chatterjee, et al. (1983). "Modification of morphine antinociceptive response by blood glucose status: possible involvement of cellular energetics." Eur J Pharmacol 90(4): 437-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=6350025 Wenk, H. N., J. D. Brederson, et al. (2006). "Morphine directly inhibits nociceptors in inflamed skin." J Neurophysiol 95(4): 2083-97. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16339007 Yamamoto, Y., T. Hatakeyama, et al. (2009). "Immunohistochemical colocalization of TREK-1, TREK-2 and TRAAK with TRP channels in the trigeminal ganglion cells." Neurosci Lett 454(2): 129-33. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=19429069

5-15

Yang, Y., Y. Wang, et al. (2004). "Mutations in SCN9A, encoding a sodium channel alpha subunit, in patients with primary erythermalgia." J Med Genet 41(3): 171-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14985375 Zimmermann, K., A. Leffler, et al. (2007). "Sensory neuron sodium channel Nav1.8 is essential for pain at low temperatures." Nature 447(7146): 855-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17568746

5-16

Chapter 6 NEUROPATHIC PAIN As mentioned in the Introduction the capacity to experience pain has a protective function: it warns us against existing or impending damage to the body and evokes responses that minimize the damage. That is, acute or nociceptive pain is a necessary protective mechanism; in contrast chronic pain serves no obvious survival or helpful function. Among the different types of chronic pain is peripheral neuropathic pain (or peripheral neuralgia) the essential feature of which is pain resulting from a wound or damage to a primary nociceptor. Neuropathic pain is often intense and unrelenting and resistant to relief by available therapies. Chronic pain without any evidence of a lesion or damage to the primary nociceptor as in a migraine is not considered neuropathic pain. The injury may be in any part of the nociceptor and may be the result of any of a number of possible insults to the nociceptor. Although the insult in peripheral neuropathic pain is to the primary nociceptor the changes underlying the neuropathic pain syndrome may include changes to the peripheral nervous system, the spinal cord and central nervous system. In this chapter we will focus on those changes thought to occur in the peripheral nervous system. Symptoms of peripheral neuropathic pain can include persistent or paroxysmal pain, burning, prickling, itching or tingling that is independent of any obvious stimulus. There can also be abnormally heightened sensations such as allodynia (pain resulting from a normally innocuous stimulus) and hyperalgesia (an enhanced response to a normally painful stimulus). Intuitively, one might think that when an afferent nerve is injured it would fail to transmit information to the spinal cord. That is, one might reasonably expect a loss in sensations rather than a heightened or persistent sensation. That is what happens when a telephone line is cut; one cannot make or receive telephone calls. To the extent that there may be some loss of sensation associated with peripheral neuropathic pain the analogy to the telephone line holds true. However, the enhanced response and the presence of pain in the absence of a stimulus imply that there is something fundamentally different between a damaged neuronal axon and a cut telephone cable. The question then becomes what kinds of changes occur when a nerve is damaged that might give rise to neuropathic pain? Functionally the spinal roots are classically divided into dorsal roots for sensory transmission and ventral roots for motor transmission. The ventral roots are thought to be composed of the axons of myelinated motor neurons. However, in humans and other mammals on the order of one third of all axons in the ventral roots are unmyelinated, have their cell bodies in the dorsal root ganglion, and are predominantly nociceptive (see Figure 6-1). This probably explains why dorsal rhizotomy, a procedure in which the spinal nerve root between the dorsal root ganglion (DRG) and the spinal cord is severed, sometimes fails to provide relief from chronic pain. Furthermore, these types of lesions where the dorsal roots are severed have not been found to cause neuropathic pain in humans. Therefore one can say that not all lesions to nociceptors result in neuropathic pain.

6-1

spinal cord

dorsal root dorsal root ganglion

ventral root

severed nerve degenerating stump

Figure 6-1. Schematic drawing showing two nociceptors in the dorsal root ganglion. For the one shown in red its axon enters the spinal cord through the dorsal root, and for the one shown in blue its axon enters the spinal cord through the ventral root. Also the nerve formed by the two roots is shown as being severed distal to where the nerve leaves the spinal canal.

6-2

Animal models of nerve injury used to study neuropathic pain When a nerve bundle is severed distal to the DRG (as illustrated in Figure 6-1) the proximal stump that is the portion still connected to the cell bodies seals off, and under the most favorable conditions nerve fibers sprout and will regenerate and reform the appropriate peripheral connections. However if the regeneration is blocked somehow, as in an amputation, the sprouts can form a tangled mass or neuroma. The part of the axons, of the severed nerve bundle, separated from the nerve cells nucleus degenerate in a process called Wallerian degeneration also known as anterograde (occurring in the normal or forward direction) degeneration. When only a fraction of the nerve fibers in a nerve bundle are damaged and undergoing Wallerian degeneration the remaining intact axons will be exposed to the products from the degenerating axons and also to agents released from the surrounding tissue that is responding to the degenerating axons. In animals in which an experimental neuroma is made by axotomy (severing) of the sciatic nerve, abnormal spontaneous activity can be recorded from dorsal roots and the nerve fibers just above the neuroma. The spontaneous discharge is also ectopic: originating in an abnormal place, rather than the normal location the peripheral nerve ending. The spontaneous activity may be related to the development of the persistent pain or paroxysmal pain that occurs with neuropathic pain (see discussion below). After sciatic nerve axotomy a large number of genes are either up or down regulated in DRG neurons (Costigan, Befort et al. 2002; Xiao, Huang et al. 2002). These changes in gene expression most likely lead to an increase in the excitability of the DRG neurons which is the simplest explanation for the abnormal spontaneous activity. One would predict that an up regulation of excitatory voltage gated Na+ and Ca++ channels and a down regulation of K+ channels, which oppose the excitatory channels, might be responsible for the increased excitability of DRG neurons following peripheral nerve injury. Remember from chapter 5 that it was suggested that up regulation of Nav 1.3 contributes to the abnormal hyper-excitability and spontaneous activity of injured DRG neurons (Black, Cummins et al. 1999). However, normal levels of neuropathic pain behavior develops in Nav 1.3 knockout mice (Nassar, Baker et al. 2006) suggesting that increased Nav 1.3 expression is not necessary for the development of neuropathic pain. The effects of axotomy of the sciatic nerve on Ca++, K+ and HCN channel currents has been studied in isolated rat DRG neurons from roots (L4 and L5) that give rise to the sciatic nerve (Abdulla and Smith 2001). Currents carried by all three channel types were decreased in DRG neurons following sciatic axotomy, suggesting that increases in neuronal excitability are associated with decreases in K+ channel currents. One of the problems in studying neuropathic pain experimentally and treating it clinically is that it can be caused by various neuropathies; consequently the symptoms and treatments may depend upon the particular cause and on which nerve, or nerves, are involved. Mechanical nerve injury can result from acute or chronic nerve compression or more severely by the partial or complete severing of a nerve. Several mechanical nerve injury models (illustrated in Figure 6-2) have been developed and have proved useful in the experimental study of neuropathic pain in rodents. As illustrated in Figure 6-2 nerve roots can be either ligated and or severed. The sciatic nerve can be partially ligated or a

6-3

dorsal root ganglia spinal nerve ligation

sciatic nerve partial sciatic ligation

severed spinal nerve

chronic constriction

severed nerves

Figure 6-2. Several different ways for mechanically injuring nerves innervating the rodent foot. These procedures have in common that some of the nerve fibers innervating the foot remain intact and allow for testing hyperalgesia and allodynia.

6-4

chronic constriction can be placed around the sciatic. Also nerves emanating from the sciatic can be severed. These procedures have in common that the innervation of the foot remains partially intact allowing tests for allodynia and hyperalgesia to be done. It is important to keep in mind that allodynia and hyperalgesia may occur in the nociceptor and/or more centrally, and it is therefore necessary to determine whether the hypersensitivity in a nerve injury-induced neuropathic pain state is in whole or in part the result of changes in the properties of nociceptors. In rats with unilateral ligation of L5/L6 lumbar spinal nerves, which produced allodynia and hypersensitivity in the ipsilateral leg, single fiber recordings were made distal to the ligation from the nerves innervating the leg (Shim, Kim et al. 2005). Response thresholds to mechanical stimuli were lower and the magnitude of responses to suprathreshold mechanical stimuli was greater for both the C- and A-fibers in these animals than in sham operated animals. Only C-fibers were responsive to heat stimuli and their thresholds were lower in the animals with nerve ligation than in the sham operated animals. These findings indicate that following nerve injury nociceptors innervating the skin become sensitized to both mechanical and thermal stimuli, thereby providing evidence that nociceptor sensitization can contribute to neuropathic pain. Mechanical trauma is not the only way a nerve can be injured. Injury can also occur as a result of a metabolic disorder (diabetes), infection (postherpetic neuralgia), autoimmune disorders, or a chemically induced nerve trauma. Complicating the situation further is that about a third of peripheral neuropathies are considered to be idiopathic that is resulting from an obscure or unknown cause. Even though chronic pain with allodynia may occur in most types of neuropathy it is very unlikely that a single molecular change will uniformly characterize neuropathic pain states. A couple of briefly described examples to illustrate the diversity of neuropathic pain syndromes are given below. Two examples of neuropathic pain CAUSALGIA The term causalgia was first used by the civil war physician Silas Weir Mitchell in 1864 to describe the intense burning pain and marked sensitivity to vibration or touch in the distribution of an injured peripheral nerve following military injuries. In the 1930-1940s causalgia was associated with the sympathetic nervous system and it is now generally accepted that pain associated with sympathetic efferent function is classified as sympathetically maintained pain. Under normal conditions activity in postganglionic sympathetic fibers does not produce pain, nor is it capable of activating nociceptors; however, after nerve injury nociceptors in the injured nerve can become excited by epinephrine and by stimulating the sympathetic trunk. However, it is not clear whether there is a direct sympathetic effect on the nociceptor or if the effect is indirect. Sympathectomy is an effective treatment especially for patients who show a positive response to sympathetic block. TRIGEMINAL NEURALGIA

6-5

Trigeminal neuralgia, also called tic douloureux, is a chronic pain condition characterized by sudden bursts (paroxysms) of facial pain in the areas supplied by the trigeminal nerve: the cheeks, jaw, teeth, gums, lips and less frequently around the eye or forehead. These episodes of pain can be triggered by a light touch around the mouth or face or even by talking or eating. It may be caused by irritation or stimulation of the trigeminal nerve by a blood vessel pressing on it as it exits the brainstem. In some cases it is the associated with multiple sclerosis or a tumor. The first choice treatment option for trigeminal neuralgia is medication: anticonvulsants such as carbamazepine (trade name Tegretol) are generally effective. For those patients unable to tolerate the side effects of the medications or those that become resistant to medication surgery is the next option. Some examples of the diversity of the treatments for neuropathic pain GABAPENTIN The anti epileptic drug gabapentin (trade name Neurontin) was initially synthesized to mimic the chemical structure of the neurotransmitter gamma-aminobutyric acid (GABA). However gabapentin is not currently thought to act on the same brain receptor as GABA. Its use as an analgesic for neuropathic pain resulted from clinical case reports of its analgesic effects in patients with well documented histories of neuropathic pain. Pregabalin (trade name Lyrica) was designed as a more potent successor to gabapentin and like gabapentin was found to be useful for the treatment of neuropathic pain. A high affinity binding protein for [3H]gabapentin was subsequently isolated and then identified as the 2-1 subunit of a voltage gated calcium channel (Gee, Brown et al. 1996). The expression of the 2-1 subunit was increased 17-fold in the DRG ipsilateral to nerve injury but not in the contralateral DRG (Luo, Chaplan et al. 2001). The increased expression of the 2-1 subunit was found to precede the onset of allodynia in the experimental animals and to diminish while animals were recovering. These results imply that up regulation of 2-1 in nociceptors could potentially play a role in the development of neuropathic pain following nerve injury. If the analgesic effects of gabapentin and pregabalin result from their binding to the 2-1 subunit of the voltage gated calcium channel and if the binding could somehow be blocked or eliminated the analgesic effect should be greatly diminished. It turns out that substitution of alanine for arginine at position 217 in the 2-1 molecule prevents gabapentin and pregabalin binding (see (Field, Cox et al. 2006)). Utilizing gene targeting techniques a mutant mouse was produced having alanine at position 217 in the 2-1 molecule (Field, Cox et al. 2006). The mutant mice exhibited normal pain responses, however the analgesic effect of pregabalin during the late phase of formalin induced pain or allodynia following chronic ligature constriction of the sciatic nerve was lost, thereby conclusively demonstrating that the analgesic actions of pregabalin are mediated via the 2-1 subunit of voltage gated calcium channels. It should be kept in mind that the mutation in the 2-1 subunit greatly decreased pregablin binding throughout the nervous system and not just in the DRG ipsilateral to the nerve injury.

6-6

In the study (Luo, Chaplan et al. 2001) showing that the increased expression of the 2-1 subunit preceded the onset of allodynia following nerve injury; the injury was produced by ligation of the L5/L6 lumbar spinal nerves at a point distal to their DRGs and proximal to their union to form the sciatic nerve. In a subsequent study (Luo, Calcutt et al. 2002) three types of mechanical nerve injury were utilized to determine whether 2-1 subunit up-regulation in the DRG correlated with the induced allodynia. In addition to the ligation of L5/L6 as described above, L5/L6 were transected at the same location and thirdly the sciatic nerve on one side was constricted by a series of ligatures around the nerve. In all three situations there was a significant increase in the expression of the 2-1 subunit in the DRG and a corresponding tactile allodynia, which was inhibited by gabapentin. These authors also studied animals exhibiting tactile allodynia resulting from diabetic neuropathy and toxic neuropathy induced by vincristine (an agent used in cancer chemotherapy which works by inhibiting microtubule assembly and its main side-effect is peripheral neuropathy). In these animals there was no significant change in the expression of the 2-1 subunit in the DRG, suggesting that the level of the 2-1 subunit in the DRG is not the determining factor for the tactile allodynia during all neuropathic pain states. Interestingly, in animals with diabetic neuropathy there was a tactile allodynia that was inhibited by gabapentin suggesting the involvement of 2-1 subunits in locations other than the DRG. Although gabapentin and pregabalin act via the 2-1 subunit of a voltage gated calcium channel, and the expression of this subunit is up regulated in the DRG, there is as yet no compelling evidence that their site of action as an analgesic for neuropathic pain is on the nociceptor and not in the spinal cord or in the CNS. ARTEMIN Current therapies for the treatment of neuropathic pain have been described as inadequate since in many cases they are only of limited benefit and they have a high incidence of undesirable side effects. In the continuing search for new therapies, artemin, one of the members of the glial cell derived neurotrophic factor (GDNF) family was considered. It signals through the GDNF family receptor GFR3, which complexes with the tyrosine kinase receptor RET. GFR3 expression in adults is largely restricted to small diameter DRG cells which have unmyelinated axons many of which are nociceptors. The effects of subcutaneous administered artemin on thermal and tactile hypersensitivity as a result of spinal nerve ligation in rats were examined. It was found that systemic, intermittent artemin administration produced a dose dependent reversal of nerve ligation induced thermal and tactile hypersensitivity which was reestablished after cessation of artemin administration (Gardell, Wang et al. 2003). In a subsequent study the effects of artemin following nerve injury were specifically studied on C fibers because the expression of the GFR3 receptor is predominantly found in small diameter DRG neurons having unmyelinated axons (Bennett, Boucher et al. 2006). They showed that artemin protected against injury induced changes in the histochemistry and electrophysiological properties of C fibers.

CANNABINOIDS

6-7

Marijuana was widely used for medicinal purposes (including analgesia) in the United States before it was classified as a Schedule I drug, which classified it as a drug with high abuse risk and no accepted medical use. However, the passage of referenda in several states has allowed the use of marijuana for medicinal purposes. Animal studies have indicated that cannabinoids produce analgesic effects at peripheral sites as well as spinal and supra spinal sites. Nevertheless, the use of cannabinoids as analgesics in humans is hindered by their potential for adverse effects, such as hallucinations, euphoria, or dysphoria, in the patients who use them. What is needed is a cannabinoid agonist that produces analgesia but having minimal adverse effects. The effects of cannabinoids are mediated via binding to two G protein coupled receptors (CB1 and CB2) that inhibit adenylate cyclase leading to decreased cAMP levels in most tissues and cells. In order to determine the extent to which CB1 receptors located in nociceptors contributed to analgesia in inflammatory and neuropathic pain CB1 was deleted in nociceptors in the peripheral nervous system of mice, while preserving its expression in the central nervous system (Agarwal, Pacher et al. 2007). These authors used a model of neuropathic pain in which they lesioned two of the three terminal branches of the sciatic nerve (tibial and common peroneal nerves) leaving the remaining sural nerve intact. They examined the analgesic effects of systemic administration of the cannabinoid receptor agonist WIN 55,212-2 on the response latency to thermal stimuli and the mechanical threshold in animals following nerve lesion. WIN 55,212-2 significantly increased response latency to thermal stimuli and raised the mechanical threshold, however the effects were significantly weaker in the knockout mice. They concluded that CB1 receptors expressed on nociceptors mediates a significant proportion of the cannabinoid induced analgesia produced in neuropathic pain. They reached the same conclusion for the analgesia produced by WIN 55,212-2 during inflammatory pain. These findings strongly argue for the development of peripherally acting analgesics based on synthetic cannabinoids that do not cross the blood-brain barrier. Finally, it is worth considering how CB1 receptor agonists which inhibit adenylate cyclase leading to decreased cAMP levels might cause analgesia for inflammatory pain and neuropathic pain. First consider the inflammatory mediators discussed in chapter 4 that activate the enzyme adenylate cyclase (AC) via the stimulatory G-protein (GS). As shown in Figure 6-3 the AC activators are PGE2 and PAR2. Activated AC catalyzes the conversion adenosine triphosphate (ATP) into cyclic adenosine monophosphate (cAMP) which in turn modulates the activity of four ion channels (HCN, TRPV1, TRPV4, TRPA1). The activity of TRPV1, TRPV4 and TRPA1 is modulated by phosphorylation via the cAMP activated protein kinase (PKA). On the other hand HCN channels have a cAMP binding domain which regulates the channel and does not require protein phosphorylation. Both PAR1 and cannabinoids (CB) inhibit AC via the inhibitory Gprotein Gi, and thereby they will tend to inhibit nociceptor activation by the inflammatory mediators PGE2 and PAR2 and in that way they inhibit inflammatory pain. How then might CB1 receptor agonists cause analgesia for neuropathic pain? One possible mechanism shown in Figure 6-3 is via the HCN channels which have been implicated in neuropathic pain (see below).

6-8

SLIGRL TRYPTASE

TFLLR THROMBIN

HCN2 or 4
PAR1 Gs Gi AC ATP cAMP PGE2 EP Gs AC AC Gi ATP ATP CB1 PKA
P
+ + + + + + + +

TRPV1

PAR2

PKA P

TRPV4

TRPA1

CB PKA
P

Figure 6-3. Regulation of nociceptor function by cAMP and PKA. PAR2 and PGE2 activate AC via Gs thereby leading to an increase in cAMP concentration and thereby to activation of PKA. In contrast PAR1 and CB1 inhibit AC via Gi thereby leading to a decrease in cAMP concentration and a decrease in the activity of PKA.

6-9

HCN CHANNELS The procedure of placing a ligature around a spinal nerve to produce a neuropathy was originally developed to have an animal model for neuropathic pain in which there was chronic pain with allodynia and hyperalgesia (Bennett and Xie 1988). Using this model it was shown that A-fibers appear to mediate mechanical allodynia and hypersensitivity while C-fibers the thermal hyperalgesia (Shir and Seltzer 1990). One mechanism that could give rise to the chronic pain of the neuropathic pain state is the presence of spontaneous activity which could activate and/or sensitize spinal neurons and thereby contribute to pain, as well as allodynia and hyperalgesia. As mentioned above, in animals in which an experimental neuroma is made by severing the sciatic nerve high levels of spontaneous ectopic neuronal activity can be recorded. In order to determine the type of fibers contributing to spontaneous activity during neuropathic pain in which there was chronic pain with allodynia and hyperalgesia a simple experiment was performed (Kajander and Bennett 1992). Ligatures were placed around the sciatic nerve and at different times post operatively nerve fiber recordings were made from the lumbar roots which contribute to the sciatic nerve (see Figure 6-4). Stimulating electrodes were placed distal and proximal to the site of ligation. The conduction velocity of spontaneously active and silent axons could be determined using the proximal stimulating electrodes. The distal stimulating electrodes were used to determine which fibers conducted through the ligation site. Spontaneous discharges were observed in 35% of A fibers (55 fibers 89% of which did not conduct through the ligation site), 15% of A fibers (20 fibers 65 % of which did not conduct through the ligation site) and only 3% of C fibers (2 fibers which did conduct through the ligation site). A subsequent experiment determined that the spontaneous activity could some times originate at or near the site of ligation (Tal and Eliav 1996) in addition to the DRG. These findings indicate that after mechanical trauma many of the injured nerves A and A fibers and some C fibers become spontaneously active, and the spontaneous activity can originate at the DRG or at or near the site of injury. Furthermore, some of the fibers having spontaneous activity still innervated the region affected by the damaged nerve. Many of the spontaneous active A and A fibers described above exhibited a regular rhythmic firing pattern strongly suggesting the possibility that the firing pattern results from an underlying pacemaker current. As discussed in chapter 5, pacemaker currents (Ih) carried by HCN channels are found in DRG neurons. In rats in which spinal nerves L5 and L6 were ligated the resulting tactile allodynia was dose dependently suppressed by ZD7288 a drug originally thought to be a specific Ih blocker (not selective among HCN14) (Chaplan, Guo et al. 2003). Furthermore, ZD7288 decreased spontaneous discharges from A and A fibers. Finally nerve injury increased the pacemaker currents in large DRG neurons and the resting membrane potential of these neurons was significantly more positive than controls. These results suggest that increased Ih plays a role in the tactile allodynia of neuropathic pain. Moreover these findings support the idea that spontaneous discharges in DRG neurons play a causal role in neuropathic pain. In disagreement with the behavioral and electrophysiological findings described above; both HCN mRNA and protein was decreased in DRG neurons on the same side as the ligation.

6-10

recording electrode and amplifier extracellular proximal stimulating electrodes extracellular distal stimulating electrodes

DRG

chronic constriction

Figure 6-4. Experimental system used for determining the types of nerve fibers firing spontaneously during neuropathic pain.See text for further details.

6-11

In a subsequent study abundant axonal accumulation of HCN channel protein was found at the sites of injury along with a slight decrease in DRG neuronal cell bodies (Jiang, Xing et al. 2008). These findings suggest that accumulation of HCN channels at the axonal site of injury gives rise to the spontaneous ectopic firing of action potentials which contributes to the mechanical allodynia of neuropathic pain. The conclusions linking Ih to the induction of the tactile allodynia of neuropathic pain depended on the specificity of ZD7288 for Ih which has been called into question with the discovery of the nonspecific effects of ZD7288. We await future results before concluding that Ih plays a causal role in the initiation of tactile allodynia in neuropathic pain. The role for HCN1 in neuropathic pain was investigated by ligation of the sciatic nerve in HCN1 knockout mice (Momin, Cadiou et al. 2008). Following nerve ligation mechanical hyperalgesia and cold allodynia was present in control animals and in the HCN1 knockout mice the mechanical hyperalgesia was similar however there was a significant decrease in the cold allodynia of more than 50% in the knockout mice. These findings suggest a causal role for HCN1 in the induction of cold allodynia in neuropathic pain. What is the role of spontaneous activity? It was mentioned earlier that following nerve injury nociceptors innervating the skin become sensitized to both mechanical and thermal stimuli. Thereby providing evidence that nociceptor sensitization can contribute to the neuropathic pain state. Since allodynia and hyperalgesia can result from changes that occur centrally, it might be that spontaneous activity leads to centrally mediated hyperalgesia. It has been shown that electrical stimulation of C fibers in humans can lead to hyperalgesia, indicating that electrical activity in C fibers is sufficient to produce centrally mediated hyperalgesia (Klede, Handwerker et al. 2003). Although ongoing spontaneous activity might therefore be sufficient to produce mechanical hyperalgesia it is not necessary; L5 ganglionectomy in which the afferents are removed resulted in mechanical hyperalgesia comparable to that for spinal nerve ligation (Sheth, Dorsi et al. 2002). The possibility that the spontaneous ongoing pain of neuropathic pain is caused by spontaneous firing of nociceptive neurons has been studied in rats (Djouhri, Koutsikou et al. 2006). Spontaneous foot lifting behavior as a result of nerve damage was used as an indicator of spontaneous pain (Choi, Yoon et al. 1994). A correlation was found between spontaneous foot lifting and the firing rate of C nociceptors following nerve injury and complete Freunds adjuvant treatment, a finding consistent with the possibility of a causal relationship between the two (Djouhri, Koutsikou et al. 2006). Although there may be situations in which chronic neuropathic pain is the result of spontaneous firing of nociceptive afferents, this should not be taken to imply the converse that all cases of ongoing neuropathic pain result from spontaneous firing of nociceptors. But what about the spontaneous activity of A afferents described above, could the spontaneous firing of these neurons be the cause of chronic pain as well as secondary allodynia and hyperalgesia? Remember from chapter 1 that a substantial fraction of the A-fiber nociceptors appear to conduct in the A conduction velocity range (Lawson

6-12

2002; Djouhri and Lawson 2004). Thus the spontaneous firing of A fibers may very well be the cause of ongoing pain, as well as secondary allodynia and hyperalgesia. At the present time this question remains to be unanswered. In summary, multiple sites are altered following nerve injury. Abnormalities can occur in both injured and uninjured nociceptors innervating the affected region. These effects include spontaneous activity, as well as allodynia and hyperalgesia. Central effects specifically sensitization following nerve injury can also occur, though their mechanisms are not considered here. BOTULINUM TOXIN TYPE A Botulinum toxin type A (trade name BOTOX) binds to the pre-synaptic nerve terminal where it is internalized into the cell and then interferes with vesicle docking thereby inhibiting acetylcholine release and muscle contraction. This has made BOTOX useful for the treatment of medical disorders arising from excessive muscle contraction which are sometimes painful. However, a dissociation between muscle relaxation and analgesia was sometimes observed clinically; suggesting that BOTOX might have analgesic effects independent of its muscle relaxing effects. The use of BOTOX as an analgesic agent was tested in rats in which ligation of the sciatic nerve was performed to induce neuropathic pain. A single ipsilateral intra plantar non toxic dose of BOTOX injected either 5 or 12 days after ligation of the sciatic nerve was able to significantly reduce mechanical allodynia for at least 3 weeks (Luvisetto, Marinelli et al. 2007). These findings strongly support the suggestion that BOTOX has analgesic effects independent of its muscle relaxing effects. Double blind placebo controlled studies for the use of BOTOX to treat neuropathic pain have been carried out in patients with post traumatic/post operative pain or post herpetic neuralgia (Ranoux, Attal et al. 2008) and another group with painful diabetic neuropathy (Yuan, Sheu et al. 2009). In both studies BOTOX significantly reduced neuropathic pain for periods lasting up to 3 months. DEMYELINATION, LYSOLECITHIN AND LYSOPHOSPHATIDIC ACID Neuropathic pain (i.e. spontaneous pain, hyperalgesia, and allodynia) is also associated with human peripheral demyelinating neuropathies such as some types of Charcot-MarieTooth disease and Guillain-Barre syndrome. Traumatic nerve injury, such as that resulting from placing a ligature around a nerve will also lead to demyelination of the injured nerve. Consequently, it is reasonable to consider whether or not demyelination might contribute to the development of the neuropathic pain state. The possible contribution of demyelination to the development of neuropathic pain was studied using the demyelinating agent lysolecithin (lysophosphatidyl choline) applied to peripheral nerves (Wallace, Cottrell et al. 2003). These authors found that topical application of lysolecithin caused focal demyelination, without any morphological or immunological indications of axonal loss. Functionally they found the occurrence of low frequency

6-13

spontaneous action potentials, with no significant peripheral allodynia or hyperalgesia but with central mechanical allodynia and thermal hyperalgesia. These findings suggest that demyelination, and not axonal damage, of afferent A-fibers induces central neuropathic pain. At about the same time that the work with lysolecithin described above was being done another independent group was investigating the role of lysophosphatidic acid (LPA) in neuropathic pain (Inoue, Rashid et al. 2004). These authors found that intrathecal injection of LPA induced behavioral allodynia and hyperalgesia with demyelination in the dorsal root similar to that found for animals after nerve ligation. Keep in mind that the intrathecal space surrounds the spinal cord and the dorsal root ganglion. Moreover they found that mice lacking one of the LPA receptors (LPA1) did not develop behavioral allodynia, hyperalgesia and demyelination after nerve injury. They concluded that receptor-mediated LPA signaling is crucial in the initiation of neuropathic pain. In order to clarify the situation with regard to lysolecithin and LPA the same group examined the effects of lysolecithin in LPA1 knockout mice (Inoue, Xie et al. 2008). They found that in contrast to normal mice those lacking the LPA1 receptor did not develop behavioral allodynia and hyperalgesia after intrathecal injection of lysolecithin. They concluded that lysolecithin is converted to LPA which then activates the LPA1 receptor to initiate neuropathic pain. This work was extended by the injection of LPA into the trigeminal ganglia of rats (Ahn, Lee et al. 2009). LPA injection into a trigeminal ganglia induced mechanical allodynia both ipsilateral and contralateral to the injection site and mechanical hyperalgesia was only observed ipsilateral to the injection site. The impetus for studying the role of LPA in neuropathic pain (Inoue, Rashid et al. 2004) was the earlier finding by the same group that intrathecal injection of botulinum toxin C3 (BoTXC3) before peripheral nerve injury inhibited the development of hyperalgesia in mice (Ye, Inoue et al. 2000). BoTXC3 inhibits the RhoA/Rho kinase (ROCK) pathway by ADP ribosylation of RhoA, and RhoA is activated by LPA signaling through the Gprotein G12/13 (Inoue, Rashid et al. 2004). Furthermore, they showed that the induction of mechanical allodynia and thermal hyperalgesia by intrathecal LPA injection was dose dependently inhibited by BoTXC3 and also by Y-27632, a reversible inhibitor of ROCK. A possible role for the RhoA/Rho kinase pathway in injured neurons was indicated by the finding that neuronal RhoA mRNA and the proportion of L5 DRG neurons that express RhoA rises following distal axotomy (Cheng, Webber et al. 2008). As mentioned above botulinum toxin type A has also been found to inhibit neuropathic pain, moreover it has been shown to work in humans. Botulinum toxin type A inhibits acetylcholine release at peripheral cholinergic synapses by proteolytically cleaving the SNAP-25 protein which is essential for release of transmitter. However, it has been shown that botulinum toxin type A also targets RhoB for degradation by proteasome (Ishida, Zhang et al. 2004). RhoB like RhoA is activated by LPA signaling through its Gprotein coupled receptor. Taken together these findings strongly point to the LPAGproteinRho/Rhokinase pathway as a potential therapeutic target for the treatment of neuropathic pain.

6-14

References cited: Abdulla, F. A. and P. A. Smith (2001). "Axotomy- and autotomy-induced changes in Ca2+ and K+ channel currents of rat dorsal root ganglion neurons." J Neurophysiol 85(2): 644-58. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11160500 Agarwal, N., P. Pacher, et al. (2007). "Cannabinoids mediate analgesia largely via peripheral type 1 cannabinoid receptors in nociceptors." Nat Neurosci 10(7): 8709. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17558404 Ahn, D. K., S. Y. Lee, et al. (2009). "Intratrigeminal ganglionic injection of LPA causes neuropathic pain-like behavior and demyelination in rats." Pain 146(1-2): 114-20. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=19665300 Bennett, D. L., T. J. Boucher, et al. (2006). "Artemin has potent neurotrophic actions on injured C-fibres." J Peripher Nerv Syst 11(4): 330-45. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17117942 Bennett, G. J. and Y. K. Xie (1988). "A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man." Pain 33(1): 87-107. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=2837713 Black, J. A., T. R. Cummins, et al. (1999). "Upregulation of a silent sodium channel after peripheral, but not central, nerve injury in DRG neurons." J Neurophysiol 82(5): 2776-85. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10561444 Chaplan, S. R., H. Q. Guo, et al. (2003). "Neuronal hyperpolarization-activated pacemaker channels drive neuropathic pain." J Neurosci 23(4): 1169-78. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12598605 Cheng, C., C. A. Webber, et al. (2008). "Activated RHOA and peripheral axon regeneration." Exp Neurol 212(2): 358-69. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18554585

6-15

Choi, Y., Y. W. Yoon, et al. (1994). "Behavioral signs of ongoing pain and cold allodynia in a rat model of neuropathic pain." Pain 59(3): 369-76. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=7708411 Costigan, M., K. Befort, et al. (2002). "Replicate high-density rat genome oligonucleotide microarrays reveal hundreds of regulated genes in the dorsal root ganglion after peripheral nerve injury." BMC Neurosci 3: 16. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12401135 Djouhri, L., S. Koutsikou, et al. (2006). "Spontaneous pain, both neuropathic and inflammatory, is related to frequency of spontaneous firing in intact C-fiber nociceptors." J Neurosci 26(4): 1281-92. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=16436616 Djouhri, L. and S. N. Lawson (2004). "Abeta-fiber nociceptive primary afferent neurons: a review of incidence and properties in relation to other afferent A-fiber neurons in mammals." Brain Res Brain Res Rev 46(2): 131-45. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15464202 Field, M. J., P. J. Cox, et al. (2006). "Identification of the alpha2-delta-1 subunit of voltage-dependent calcium channels as a molecular target for pain mediating the analgesic actions of pregabalin." Proc Natl Acad Sci U S A 103(46): 17537-42. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17088553 Gardell, L. R., R. Wang, et al. (2003). "Multiple actions of systemic artemin in experimental neuropathy." Nat Med 9(11): 1383-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=14528299 Gee, N. S., J. P. Brown, et al. (1996). "The novel anticonvulsant drug, gabapentin (Neurontin), binds to the alpha2delta subunit of a calcium channel." J Biol Chem 271(10): 5768-76. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8621444 Inoue, M., M. H. Rashid, et al. (2004). "Initiation of neuropathic pain requires lysophosphatidic acid receptor signaling." Nat Med 10(7): 712-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15195086

6-16

Inoue, M., W. Xie, et al. (2008). "Lysophosphatidylcholine induces neuropathic pain through an action of autotaxin to generate lysophosphatidic acid." Neuroscience 152(2): 296-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18280050 Ishida, H., X. Zhang, et al. (2004). "Botulinum toxin type A targets RhoB to inhibit lysophosphatidic acid-stimulated actin reorganization and acetylcholine release in nerve growth factor-treated PC12 cells." J Pharmacol Exp Ther 310(3): 881-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15140914 Jiang, Y. Q., G. G. Xing, et al. (2008). "Axonal accumulation of hyperpolarizationactivated cyclic nucleotide-gated cation channels contributes to mechanical allodynia after peripheral nerve injury in rat." Pain 137(3): 495-506. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18179873 Kajander, K. C. and G. J. Bennett (1992). "Onset of a painful peripheral neuropathy in rat: a partial and differential deafferentation and spontaneous discharge in A beta and A delta primary afferent neurons." J Neurophysiol 68(3): 734-44. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=1331353 Klede, M., H. O. Handwerker, et al. (2003). "Central origin of secondary mechanical hyperalgesia." J Neurophysiol 90(1): 353-9. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12843313 Lawson, S. N. (2002). "Phenotype and function of somatic primary afferent nociceptive neurones with C-, Adelta- or Aalpha/beta-fibres." Exp Physiol 87(2): 239-44. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11856969 Luo, Z. D., N. A. Calcutt, et al. (2002). "Injury type-specific calcium channel alpha 2 delta-1 subunit up-regulation in rat neuropathic pain models correlates with antiallodynic effects of gabapentin." J Pharmacol Exp Ther 303(3): 1199-205. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12438544 Luo, Z. D., S. R. Chaplan, et al. (2001). "Upregulation of dorsal root ganglion (alpha)2(delta) calcium channel subunit and its correlation with allodynia in spinal nerve-injured rats." J Neurosci 21(6): 1868-75. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11245671

6-17

Luvisetto, S., S. Marinelli, et al. (2007). "Anti-allodynic efficacy of botulinum neurotoxin A in a model of neuropathic pain." Neuroscience 145(1): 1-4. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17218063 Momin, A., H. Cadiou, et al. (2008). "Role of the hyperpolarization-activated current Ih in somatosensory neurons." J Physiol 586(Pt 24): 5911-29. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18936078 Nassar, M. A., M. D. Baker, et al. (2006). "Nerve injury induces robust allodynia and ectopic discharges in Nav1.3 null mutant mice." Mol Pain 2: 33. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=17052333 Ranoux, D., N. Attal, et al. (2008). "Botulinum toxin type A induces direct analgesic effects in chronic neuropathic pain." Ann Neurol 64(3): 274-83. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=18546285 Sheth, R. N., M. J. Dorsi, et al. (2002). "Mechanical hyperalgesia after an L5 ventral rhizotomy or an L5 ganglionectomy in the rat." Pain 96(1-2): 63-72. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=11932062 Shim, B., D. W. Kim, et al. (2005). "Mechanical and heat sensitization of cutaneous nociceptors in rats with experimental peripheral neuropathy." Neuroscience 132(1): 193-201. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=15780478 Shir, Y. and Z. Seltzer (1990). "A-fibers mediate mechanical hyperesthesia and allodynia and C-fibers mediate thermal hyperalgesia in a new model of causalgiform pain disorders in rats." Neurosci Lett 115(1): 62-7. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=2216058 Tal, M. and E. Eliav (1996). "Abnormal discharge originates at the site of nerve injury in experimental constriction neuropathy (CCI) in the rat." Pain 64(3): 511-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=8783316 Wallace, V. C., D. F. Cottrell, et al. (2003). "Focal lysolecithin-induced demyelination of peripheral afferents results in neuropathic pain behavior that is attenuated by cannabinoids." J Neurosci 23(8): 3221-33.

6-18

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12716929 Xiao, H. S., Q. H. Huang, et al. (2002). "Identification of gene expression profile of dorsal root ganglion in the rat peripheral axotomy model of neuropathic pain." Proc Natl Acad Sci U S A 99(12): 8360-5. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=12060780 Ye, X., M. Inoue, et al. (2000). "Botulinum toxin C3 inhibits hyperalgesia in mice with partial sciatic nerve injury." Jpn J Pharmacol 83(2): 161-3. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=10928331 Yuan, R. Y., J. J. Sheu, et al. (2009). "Botulinum toxin for diabetic neuropathic pain: a randomized double-blind crossover trial." Neurology 72(17): 1473-8. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dop t=Citation&list_uids=19246421

6-19