Each Spinal Cord Segment Innervates a Dermatome

As the neural tube closes, adjacent mesoderm also segments, here into a series of somites (see Fig. 2-3) that will give rise to skin, muscle, and bone. Each spinal nerve retains its relationship with a somite during development, with the result that spinal cord segments are related systematically to areas of skin, to muscles, and in some instances to bones (e.g., vertebrae). Hence each spinal nerve (except C1, which typically has only a rudimentary dorsal root) innervates a single dermatome (Fig. 10-4). This dermatomal arrangement is particularly apparent in the trunk, where pairs of dermatomes form bands that encircle the chest and abdomen; outgrowth of limb buds during development makes the dermatomal arrangement somewhat more complex in the upper and lower extremities. ^* Similarly, the innervation of skeletal muscles is related systematically to spinal segments (Table 10-1).

Figure 10-4 Cutaneous territories innervated by spinal nerves (dermatomes) and the trigeminal nerve (V₁, V₂, V₃). Co, coccygeal segment.

(Based on Bonica JJ: Applied anatomy relevant to pain. In Bonica JJ, editor: The management of pain, ed 2, Philadelphia, 1990, Lea & Febiger.)

Table 10-1 Innervation of Major Muscles

* Major segments indicated in bold.

Knowledge of the segmental innervation of muscles and cutaneous areas (Table 10-2) can be extremely helpful in diagnosing the site of damage in or near the spinal cord. For example, compression of a dorsal root can cause pain in its dermatome, allowing pain caused by root compression to be differentiated from pain caused by peripheral nerve damage. In addition, the highest level of a sensory or motor deficit may allow deductions about the segmental level of a suspected spinal cord lesion (see Fig. 10-31).

Table 10-2 Dermatomal Levels of Clinical Importance^*

* See Figure 10-4 for additional details.

Figure 10-31 Brown-Séquard syndrome. Damage to the outlined area of the spinal cord (in this example at C8) would affect the indicated tracts, causing ipsilateral spastic paralysis and loss of fine touch and proprioception, and contralateral loss of pain and temperature beginning one or more segments below the level of damage. CST, corticospinal tract; PC, posterior column; STT, spinothalamic tract.

The Spinal Cord Is Shorter Than the Vertebral Canal

The spinal cord approaches its adult length before the vertebral canal does. Until the third month of fetal life, both grow at about the same rate, and the cord fills the canal. Thereafter the body and the vertebral column grow faster than the spinal cord does, so that at the time of birth the spinal cord ends at the third lumbar vertebra. A small additional amount of differential growth in the vertebral column occurs subsequent to this, and by a few months of age the cord ends at about the level of the first lumbar vertebra. However, the spinal nerves still exit through the same intervertebral foramina as they did early in development, and each dorsal root ganglion remains at the level of the appropriate foramen. Proceeding from cervical to sacral levels, the dorsal and ventral roots become progressively longer because they have longer and longer distances to travel before reaching their sites of exit from the vertebral canal (Fig. 10-3A). The lumbar cistern, from the end of the spinal cord at vertebral level L1-L2 to the end of the dural sheath at vertebral level S2, is filled with this collection of dorsal and ventral roots, collectively referred to as the cauda equina (Latin for “horse’s tail”; Fig. 10-5E and F). Hence a needle carefully inserted into the lumbar cistern will pass harmlessly among nerve roots, allowing safe sampling of cerebrospinal fluid (CSF).

Figure 10-5 Formation of the cauda equina. A to E, Cross sections from progressively more caudal levels of a vertebral column in which the subarachnoid space was filled with dyed gelatin (C3, T3, L1, L2, and L3 vertebrae, respectively). The C3 and T3 vertebrae encase spinal cord segments C4 and T4-T5, each adjacent to the dorsal and ventral roots of these segments and suspended by denticulate ligaments. The L1 and L2 vertebrae encase the sacral spinal cord (S) together with a collection of dorsal and ventral roots from lumbar and sacral segments. By the level of the L3 vertebra, the spinal cord has ended and only the cauda equina remains. F, Actual-size view of the caudal end of the spinal cord and the cauda equina, seen from the posterior side after the arachnoid and dura were spread apart. * in A, foramen for the vertebral artery; arrow in C, spinal nerve emerging from the intervertebral foramen.

(A to E, from Key A, Retzius G: Studien in der anatomie des nervensystems und des bindegewebes, vol 1, Stockholm, 1875, Norstad. F, courtesy Dr. Norman Koelling, University of Arizona College of Medicine. Adapted from Nolte J, Angevine JB Jr: The human brain in photographs and diagrams, ed 3, St. Louis, 2007, Mosby.)

Each of the first seven cervical nerves leaves the vertebral canal above the corresponding vertebra; for instance, the first cervical nerve leaves between the occiput and the first cervical vertebra (the atlas), the second leaves between the first and second cervical vertebrae (the atlas and the axis), and so on. However, because there are only seven cervical vertebrae, the eighth cervical nerve leaves between the seventh cervical and first thoracic vertebrae, and each of the subsequent nerves leaves below the corresponding vertebra.

The meningeal coverings of the spinal cord were described in Chapter 4 (see Fig. 4-13). The cord is suspended within an arachnoid-lined dural tube by the denticulate ligaments (Fig. 10-6A), which are extensions of the pia-arachnoid, similar to but more substantial than arachnoid trabeculae. In addition, the caudal end of the cord is anchored to the end of the dural tube by the filum terminale (Fig. 10-6B), an extension of the pial covering of the conus medullaris. The filum terminale then acquires a dural outer layer and in turn is anchored to the coccyx.

Figure 10-6 Meningeal suspension of the spinal cord. A, Anterior surface of the lower cervical and upper thoracic cord, revealed by cutting open the spinal dural sheath. B, Posterior view of the filum terminale traveling among the roots of the cauda equina, emerging from the spinal dural sheath, and crossing the sacrum on its way to an anchor point on the coccyx.

(Courtesy Dr. Norman Koelling, University of Arizona College of Medicine. Adapted from Nolte J, Angevine JB Jr: The human brain in photographs and diagrams, ed 3, St. Louis, 2007, Mosby.)

The Posterior Horn Contains Sensory Interneurons and Projection Neurons

The posterior horn consists mainly of interneurons whose processes remain within the spinal cord and of projection neurons whose axons collect into long, ascending sensory pathways. This area of gray matter contains two prominent parts, the substantia gelatinosa and the body of the posterior horn, both present at all spinal levels.

The substantia gelatinosa is a distinctive region of gray matter that caps the posterior horn (Fig. 10-8). In myelin-stained preparations this region looks pale compared with the rest of the gray matter because it deals mostly with finely myelinated and unmyelinated sensory fibers that carry pain and temperature information. Between the substantia gelatinosa and the surface of the cord is a relatively pale-staining area of white matter called Lissauer’s tract. ^* This tract stains more lightly than the rest of the white matter because it contains the finely myelinated and unmyelinated fibers with which the substantia gelatinosa deals.

Figure 10-8 Cross sections of the spinal cord at various levels; note the large lateral extensions of the anterior horns in C5, C8, and L5. C, Clarke’s nucleus; DR, dorsal root; FC, fasciculus cuneatus; FG, fasciculus gracilis; IL, intermediolateral cell column; L, Lissauer’s tract; SG, substantia gelatinosa.

The body of the posterior horn consists mainly of interneurons and projection neurons that transmit various types of somatic and visceral sensory information. In this respect it functionally overlaps parts of the intermediate gray matter.

The Anterior Horn Contains Motor Neurons

The anterior horn contains the cell bodies of the large motor neurons that supply skeletal muscle (Fig. 10-9). These alpha motor neurons, also referred to as lower motor neurons, ^* are the only means by which the nervous system can exercise control over body movements, whether voluntary or involuntary; a number of different parts and pathways of the nervous system can influence these lower motor neurons, but they alone can elicit muscle contraction. Destruction of the lower motor neurons supplying a muscle or interruption of their axons therefore causes complete paralysis of that muscle. Lower motor neuron lesions cause paralysis of a type called flaccid paralysis, indicating that the muscle is limp and uncontracted. Reflex contractions can no longer be elicited, and the muscle slowly atrophies (owing to a lack of trophic factors normally delivered to it by motor axons; see Chapter 24). This occurs, for example, in poliomyelitis (a viral disease that attacks the motor neurons of the anterior horn) and in injuries in which ventral roots are damaged.

Figure 10-9 A single motor neuron from the lumbar spinal cord of an adult cat. A marker substance (horseradish peroxidase) was injected from the intracellular tip of a microelectrode, and the neuron was subsequently reconstructed from a series of sections. The extent and complexity of the dendritic trees of real neurons are obviously different from those of the “cartoon” neurons in most of the diagrams in this book.

(Modified from Ulfhake B et al: J Comp Neurol 278:69, 1988.)

Alpha motor neurons occur in longitudinally oriented, cigar-shaped groups, each group innervating an individual muscle. Hence in cross sections they appear to be arranged in clusters (Fig. 10-10), separated from one another by areas of interneurons; the clusters that innervate axial muscles are medial to those that innervate limb muscles. In the cervical and lumbar enlargements, which innervate the limbs, the anterior horns are enlarged laterally to accommodate the additional motor neurons (Fig. 10-8). Smaller gamma motor neurons are interspersed with alpha motor neurons in all such groups. They innervate the intrafusal muscle fibers of muscle spindles, so they are also referred to as fusimotor neurons.

Figure 10-10 Clusters of motor neurons (arrows) in the anterior horn at S4.

Two columns of motor neurons in the anterior horn of the cervical cord are recognized as separate entities. The spinal accessory nucleus extends from the caudal medulla to about C5. The axons of these motor neurons emerge from the lateral surface of the spinal cord just posterior to the denticulate ligament as a separate series of rootlets that form the accessory nerve (see Fig. 3-17). The phrenic nucleus, containing the motor neurons that innervate the diaphragm, is located in the medial portion of the anterior horn in segments C3 to C5. This makes injuries to the upper cervical spinal cord a matter of grave concern, because destruction of the descending pathways that control the phrenic nucleus and other respiratory motor neurons renders a patient unable to breathe.

The Intermediate Gray Matter Contains Autonomic Neurons

The gray matter that is intermediate to the anterior and posterior horns has some characteristics of both and also contains the spinal preganglionic autonomic neurons. In addition, at some levels it includes a distinctive region called Clarke’s nucleus.

The preganglionic sympathetic neurons for the entire body lie in segments T1 through L3, most of them located in a column of cells called the intermediolateral cell column, which forms a pointy lateral horn on the spinal gray matter (Fig. 10-8). Their axons leave through the ventral roots. Cells in a corresponding location in segments S2 to S4 constitute the sacral parasympathetic nucleus but do not form a distinct lateral horn. Their axons leave through the ventral roots and synapse on the postganglionic parasympathetic neurons for the pelvic viscera.

Clarke’s nucleus (or the nucleus dorsalis) is a rounded collection of large cells located on the medial surface of the base of the posterior horn from about T1 to L2. It is particularly prominent at lower thoracic levels (Fig. 10-8). This is an important relay nucleus for the transmission of information to the cerebellum and may also play a role in forwarding proprioceptive information from the leg to the thalamus. Because of its prominent role in sensory processing, it is treated by many as part of the posterior horn.

The remainder of the intermediate gray matter is a collection of various projection neurons, sensory interneurons, and interneurons that synapse on motor neurons.

Spinal Cord Gray Matter Is Arranged in Layers

In 1952 Rexed devised a system for subdividing the gray matter of the cat’s spinal cord into layers, or laminae. The same system has since been applied to the cords of other mammals, including humans (Fig. 10-7B). Lamina I (also called the marginal zone) is a thin layer of gray matter that covers the substantia gelatinosa, lamina II is the substantia gelatinosa, and laminae III through VI are the body of the posterior horn; lamina VII roughly corresponds to the intermediate gray matter (including Clarke’s nucleus) but also includes large extensions into the anterior horn; lamina VIII comprises some of the interneuronal zones of the anterior horn, whereas lamina IX consists of the clusters of motor neurons embedded in the anterior horn; lamina X is the zone of gray matter surrounding the central canal.

This terminology has proved useful for experimental anatomists and physiologists because the histological differences among the laminae correspond to functional differences (Table 10-3). For example, the functional dichotomy between large- and small-diameter peripheral nerve fibers is maintained to a great extent in the patterns of termination of these fibers in the spinal gray matter: there are prominent (though not exclusive) terminations of pain and temperature afferents in laminae I and II, tactile afferents from cutaneous nerves in lamina III, and Ia muscle spindle afferents in laminae VI, VII, and IX.

Table 10-3 Important Subdivisions of Spinal Cord Gray Matter

Muscle Stretch Leads to Excitation of Motor Neurons

All skeletal muscles (except perhaps some muscles in the head) contract to at least some extent in response to being stretched. The reflex arc responsible for this contraction utilizes the simplest possible route through the CNS because it involves only two neurons and a single intervening synapse. It is therefore sometimes referred to as the monosynaptic reflex or the myotatic reflex (from two Greek words meaning “muscle stretch”). The afferent limb of the arc is a Ia afferent with its associated muscle spindle primary ending. Central processes of the Ia afferent make synapses within the spinal cord directly on the alpha motor neurons that innervate the muscle containing the stimulated spindle (Fig. 10-11).

Figure 10-11 Stretch reflex. Striking the patellar tendon activates muscle spindle primary endings, which then monosynaptically excite alpha motor neurons that innervate the stretched muscle.

The stretch reflex is commonly used for clinical testing purposes. Tapping the patellar tendon, as in the familiar knee-jerk reflex, stretches the quadriceps slightly. Ia endings in quadriceps muscle spindles are excited and in turn excite quadriceps alpha motor neurons; these cause the quadriceps to contract, completing the reflex. Similarly, tapping the Achilles tendon stretches the gastrocnemius slightly, thereby causing a reflex contraction. Testing a variety of stretch reflexes can provide valuable clinical information about the integrity not only of peripheral nerves but also of predictable spinal cord segments (Table 10-4). Because stretch reflexes are usually elicited by tapping a tendon, they are often referred to as deep tendon reflexes (sometimes abbreviated as DTRs). One should remember that even though the reflex is studied in this manner, the responsible receptors are actually in the muscles attached to the tapped tendons.

Table 10-4 Deep Tendon Reflexes Commonly Tested Clinically

Stretch reflexes are thought to be important for the constant automatic corrections we perform during movements and postures (although other reflexes may in fact be even more important for this function). As an example, when we stand still and upright, we actually sway to and fro a bit. Each time we sway in one direction, some muscles are stretched, and the resulting reflex contraction helps return us toward the desired position.

Muscle Tension Can Lead to Inhibition of Motor Neurons

Stimulation of a Ib fiber from a Golgi tendon organ has an effect that varies, depending on the position and activity of the limb at the time of stimulation. It sometimes has an effect opposite to that of stimulating a Ia fiber: the alpha motor neurons that innervate the muscle connected to that tendon organ are inhibited. This effect is a form of autogenic inhibition and involves an inhibitory interneuron between the afferent and efferent fibers (Fig. 10-12). Under other circumstances (e.g., stimulating a tendon organ attached to a weight-supporting muscle), excitation of the motor neurons can result (again, through an interneuron).

Figure 10-12 Reflex connections of Golgi tendon organs. Contraction of a muscle activates the Golgi tendon organ (GTO) in its attached tendon. Under some conditions, the Ib afferents then activate inhibitory interneurons that inhibit the motor neurons to that muscle (autogenic inhibition). Under other conditions, the opposite effect is noted, mediated by excitatory interneurons.

The normal role of reflexes mediated by Golgi tendon organs is not yet completely understood. It was thought for a time that autogenic inhibition initiated by these receptors is protective in nature, preventing muscles from developing excess tension. However, in view of the great sensitivity of tendon organs to actively generated tension, it is clear that this reflex is activated long before hazardous levels are reached. Therefore it is now thought that Golgi tendon organs contribute to fine adjustments in the force of muscle contraction during ordinary motor activities and that other receptors initiate additional forms of autogenic inhibition at higher tension levels.

Clinically, autogenic inhibition may be manifested in a phenomenon called the clasp-knife response. In certain pathological conditions that follow damage to descending motor pathways, the resistance of muscles to manipulation is greatly increased. Thus one would have considerable difficulty flexing the leg of an individual with such a condition. If sufficient force is applied, however, the leg slowly flexes until at some point all resistance suddenly disappears and the leg collapses in flexion, like a clasp knife snapping shut. This collapse of resistance was once attributed to autogenic inhibition initiated by Golgi tendon organs, but here too, other receptors play the major role.

Painful Stimuli Elicit Coordinated Withdrawal Reflexes

Whereas stretch reflexes and autogenic inhibition are initiated by muscle or tendon receptors and primarily involve the muscle stretched or tensed, the flexor reflex is initiated by cutaneous receptors and involves a whole limb. A familiar example is withdrawal from a painful stimulus; after accidentally touching something painfully hot or sharp, we automatically remove the offended hand from that vicinity by flexing the arm to which it is attached.

The flexor reflex pathways in the spinal cord are normally held in a somewhat inhibited state by descending influences from the brainstem, so that only noxious stimuli result in a strong reflex. If these descending influences are removed, either surgically in experimental animals or as a result of some pathological condition, reflex flexion can result from harmless tactile stimulation. This indicates that most or all cutaneous receptors feed into the pathway, but ordinarily only nociceptors have a powerful enough influence to cause a reflex withdrawal.

Because the flexor reflex involves an entire limb, its pathway must spread over several spinal segments to include the motor neurons innervating all the various flexor muscles of that limb. This spreading occurs in two ways. First, all primary afferent fibers bifurcate on entering the spinal cord, and their processes then extend one or more segments in both rostral and caudal directions. Second, the flexor pathway includes at least one interneuron, which itself may have processes extending over several segments (Fig. 10-13).

Figure 10-13 Flexor reflex. This reflex involves several segments, and all connections are polysynaptic. In the example shown, a nociceptive fiber from the foot enters the spinal cord at S1 and activates (through at least one interneuron) motor neurons to iliopsoas and hamstring muscles.

Although this reflex is usually called the flexor reflex, the term withdrawal reflex is also used and is perhaps more appropriate. The reflex is not an all-or-none phenomenon for a given limb; rather, it shows different patterns, depending on which portion of the limb is stimulated (each pattern being appropriate to withdraw the stimulated area). It would be imprudent to flex a lower extremity when a painful stimulus was applied to the anterior surface of the thigh because this would drive the thigh into the stimulus. In such a situation, it would make much more sense to activate the extensors, which is in fact what happens. Modification of the reflex response so that it reflects the area being stimulated is called local sign.

Reflexes Are Accompanied by Reciprocal and Crossed Effects

So far, this has been a simplified description of reflex circuits, including only the most direct and dominant motor effects. However, these reflexes also include weaker influences on other muscles of the same limb and even of contralateral limbs.

It would clearly be easier to shorten a stretched muscle if the motor neurons to its synergists were excited and those to its antagonists inhibited. This reciprocal inhibition actually does occur and is a general principle in all reflexes: reflex activity in a given muscle produces similar activity in its ipsilateral synergists and the opposite activity in its ipsilateral antagonists (Fig. 10-14). Thus the standard tap on the patellar tendon causes not only excitation of quadriceps motor neurons but also inhibition (through an interneuron) of motor neurons to the hamstring muscles. If one extensor muscle of the thigh were selectively stretched, its motor neurons would be monosynaptically excited, as would those of all the other thigh extensors. After stimulation of a Golgi tendon organ the pattern may be just the reverse: if tension is applied to the patellar tendon during certain phases of a movement, the quadriceps is inhibited and the hamstring muscles are excited, both actions occurring through interneurons. Finally, the flexor reflex is accompanied by inhibition of the extensors of that limb.

Figure 10-14 Reciprocal inhibition. Striking the patellar tendon initiates a stretch reflex, as in Figure 10-11. It also causes inhibition, through an interneuron, of the motor neurons to the antagonist hamstring muscles.