Current hypotheses of structural pattern formation in the somatosensory system and their potential relevance to humans

Somatosensory nuclei of the manatee brainstem and thalamus

Florida manatees have an extensive, well-developed system of vibrissae distributed over their entire bodies and especially concentrated on the face. Although behavioral and anatomical assessments support the manatee's reliance on somatosensation, a systematic analysis of the manatee thalamus and brainstem areas dedicated to tactile input has never been completed. Using histochemical and histological techniques (including stains for myelin, Nissl, cytochrome oxidase, and acetylcholinesterase), we characterized the relative size, extent, and specializations of somatosensory regions of the brainstem and thalamus. The principal somatosensory regions of the brainstem (trigeminal, cuneate, gracile, and Bischoff's nucleus) and the thalamus (ventroposterior nucleus) were disproportionately large relative to nuclei dedicated to other sensory modalities, providing neuroanatomical evidence that supports the manatee's reliance on somatosensation. In fact, areas of the thalamus related to somatosensation (the ventroposterior and posterior nuclei) and audition (the medial geniculate nucleus) appeared to displace the lateral geniculate nucleus dedicated to the subordinate visual modality. Furthermore, it is noteworthy that, although the manatee cortex contains Rindenkerne (barrel-like cortical nuclei located in layer VI), no corresponding cell clusters were located in the brainstem ("barrelettes") or thalamus ("barreloids").

Neurotrophin receptor (p75) in the trigeminal thalamus of the rat: development, response to injury, transient vibrissa-related patterning, and retrograde transport

We report on the transient, patterned expression of p75 in the ventrobasal (VB) thalamus, the major thalamic relay for somatosensation. We immunostained the brains of developing rats ranging in age from embryonic day (E) 14.5 to postnatal day (PD) 15 with an antibody against p75. To compare p75 expression with the developing synaptic organization within VB, we also immunolocalized the synaptic-vesicle-associated protein, synaptophysin (SYN), on alternate sections. p75-immunoreactivity (IR) was dense and uniform in the ventroposterior medial nucleus (VPM) in the late embryonic and early postnatal periods (E 16.5 to PD 3). In contrast, from PD 4-10, p75-IR in the VPM was patterned, reminiscent of cytochrome-oxidase-stained barreloids, a characteristic feature of the VB in rodents. By PD 14, p75-IR in the VPM was no longer detectable. The ventroposterior lateral nucleus (VPL), in contrast, exhibited no p75-IR. No p75-IR was detected in the ventroposterior lateral nucleus (VPL) at any developmental stage in which VPM could be distinguished from VPL. Light, but clearly patterned SYN-IR, first detectable on PD 2-3, increased in intensity in both VPL and VPM through PD 15. Sectioning the infraorbital nerve on PD 0 resulted in blurred patterns of p75- and SYN-IR within VPM in PD 7-9 rat pups. Removing large portions of the somatosensory cortex on PD 0 resulted in subsequent greatly reduced p75- and SYN-IR within VB. To specify the source of the p75-IR terminals, we stereotaxically injected into the VPM of PD 4-5 rats a monoclonal antibody to p75. One to 2 days later, IR of retrogradely transported p75 antibodies could be traced within axons and cell bodies of neurons associated with the trigeminothalamic pathway through the caudal diencephalon and mesencephalon; labelling was confined to the contralateral trigeminal principal sensory nucleus. The observed, transiently patterned p75-IR in VPM the early postpartum period suggests a role for p75 in synaptogenesis and pattern formation.

C-fiber depletion alters response properties of neurons in trigeminal nucleus principalis

The effects of C-fiber depletion induced by neonatal capsaicin treatment on the functional properties of vibrissa-sensitive low-threshold mechanoreceptive (LTM) neurons in the rat trigeminal nucleus principalis were examined in adult rats. Neonatal rats were injected either with capsaicin or its vehicle within 48 h of birth. The depletion of unmyelinated afferents was confirmed by the significant decrease in plasma extravasation of Evan's blue dye induced in the hindlimb skin of capsaicin-treated rats by cutaneous application of mustard oil and by the significant decrease of unmyelinated fibers in both the sciatic and infraorbital nerves. The mechanoreceptive field (RF) and response properties of 31 vibrissa-sensitive neurons in capsaicin-treated rats were compared with those of 32 vibrissa-sensitive neurons in control (untreated or vehicle-treated) rats. The use of electronically controlled mechanical stimuli allowed quantitative analysis of response properties of vibrissa-sensitive neurons; these included the number of center- and surround-RF vibrissae within the RF (i.e., those vibrissae which when stimulated elicited >/=1 and <1 action potential per stimulus, respectively), the response magnitude and latency, and the selectivity of responses to stimulation of vibrissae in different directions with emphasis on combining both the response magnitude and direction of vibrissal deflection in a vector analysis. Neonatal capsaicin treatment was associated with significant increases in the total number of vibrissae, in the number of center-RF vibrissae per neuronal RF, and in the percentage of vibrissa-sensitive neurons that also responded to stimulation of other types of orofacial tissues. Compared with control rats, capsaicin-treated rats showed significant increases in the response magnitude to stimulation of surround-RF vibrissae as well as in response latency variability to stimulation of both center- and surround-RF vibrissae. C-fiber depletion also significantly altered the directional selectivity of responses to stimulation of vibrissae. For neurons with multiple center-RF vibrissae, the proportion of center-RF vibrissae with net vector responses oriented toward the same quadrant was significantly less in capsaicin-treated compared with control rats. These changes in the functional properties of principalis vibrissa-sensitive neurons associated with marked depletion of C-fiber afferents are consistent with similarly induced alterations in LTM neurons studied at other levels of the rodent somatosensory system, and indeed may contribute to alterations previously described in the somatosensory cortex of adult rodents. Furthermore, these results provide additional support to the view that C fibers may have an important role in shaping the functional properties of LTM neurons in central somatosensory pathways.

Synaptophysin immunoreactivity and distributions of calcium-binding proteins highlight the functional organization of the rat's dorsal column nuclei

The mammalian dorsal column nuclei (DCN) are principally composed of the cuneate (CN) and gracile (GN) nuclei. Data presented here support previously published anatomical and functional evidence that the longitudinal organization of the CN and GN reflect the complex role of the DCN in somatosensory processing. The CN is organized longitudinally into three parts. Within the middle portion of this nucleus, primary afferent projections and cuneothalamic cells are concentrated. Although traditional cytoarchitectonic analyses had failed to detect this tripartite organization in rats, we found evidence for it, with a functional middle region, extending approximately 0.2-0.9 mm caudal to the obex, characterized by precise somatotopy of primary afferent terminations and corresponding somatotopy of cytochrome oxidase (CO) blotches. Additional evidence is presented here consistent with a functionally distinct middle region within the rat's CN: (1) patches of dense synaptophysin (a synaptic-vesical-associated protein)-immunoreactivity (SYN-IR) are limited to the middle CN region, coincident with the dense CO blotches; (2) neurons immunoreactive for the calcium-binding proteins calbindin-D28 (CB), calretinin (CR) and parvalbumin (PV) are concentrated in the middle CN region. Furthermore, in adult rats subjected to perinatal forepaw removal, (1) the patterns of SYN-IR in the middle region of the CN are disrupted, as had previously been shown for the patterns of CO blotches; (2) in contrast, however, distributions of CN cells with PV-, CB- and CR-IR are unaffected. Evidence for a tripartite division in the GN is also presented, based on the distributions of cells with PV-, CB- and CR-IR.

Timecourse of development of the wallaby trigeminal pathway. I. Periphery to brainstem

The development of the vibrissae and their innervation and the maturation of the brainstem trigeminal sensory nuclei have been studied in the wallaby, Macropus eugenii, from birth to adulthood. At birth, developing vibrissal follicles consist of solid epidermal pegs surrounded by dermal condensations. The developing follicles and adjacent skin are innervated by trigeminal afferents. Ten days after birth the follicle contains a dermal papilla and the deep vibrissal nerve can be recognised. A hair cone is present at postnatal day (P) 30 and hairs are apparent on the skin surface by P35. By P63 the deep vibrissal nerve can be seen innervating Merkel cells in the outer root sheath; in addition, the first signs of the blood sinus can be recognised. Innervation of the inner conical body and lanceolate and lamellated receptors supplying the mesenchymal sheath and waist region are not seen until P119, when the follicle resembles that seen in the adult. At birth, central processes of the trigeminal ganglion cells have entered the trigeminal tract and extend from the rostral pons to the upper cervical cord. Labelling with a carbocyanine dye at P0 shows afferents extending medially from the tract into the trigeminal subnuclei at all levels. At this stage the trigeminal nuclei appear as areas of increased cell density in the lateral brainstem. By P30-40 the four subnuclei can be distinguished on the basis of shape, cytoarchitecture, and succinic dehydrogenase reactivity. Adult morphology is not fully established until P210. In mature animals, nucleus principalis contains closely packed, polymorphic cells, frequently aligned parallel to thick fibre bundles that traverse the nucleus obliquely. Subnuclei oralis and interpolaris contain sparsely distributed, medium to large cells, randomly oriented, as well as prominent rostrocaudally directed fibre bundles. Subnucleus caudalis consists of the marginal layer, substantia gelatinosa, and magnocellular layers as described in other species. Patches of increased succinic dehydrogenase or cytochrome oxidase reactivity, presumably corresponding to the vibrissae, are present in subnuclei principalis, interpolaris, and caudalis in developing and adult animals, although the pattern is less clear than in rats. The brainstem patches are first seen at P40, approximately 6 weeks before the corresponding vibrissal-related pattern develops in the cortex. This suggests that the onset of patch formation may be regulated independently at different levels of the pathway.

The mutability of low-affinity nerve growth factor receptor (p75NGFR) expression in the rat cuneate nucleus following perinatal injury and adult deafferentations: comparisons with cytochrome oxidase

In normal adult rats, intense immunostaining for the 75 kDa low-affinity receptor for nerve growth factor and other neurotrophins (p75NGFR) is concentrated in the middle region of the cuneate nucleus (CN), distributed in a blotchy pattern similar to that of cytochrome oxidase (CO) activity. In the adult rats, partial dorsal rhizotomies (centered around the 7th and 8th cervical spinal segments) resulted in the complete disappearance of p75NGFR-like immunoreactivity within the ipsilateral CN, but did not affect the distribution of the CO blotches. Perinatal (postnatal day 1-8) damage to the ipsilateral forepaw and subsequent rearing to adulthood also resulted in significant disruption of the topographical expression of p75NGFR-like immunoreactivity within the CN, as well as--as previously reported--disruption of the CO blotches. Although the patterns of staining in intact adult rats are similar for CO staining and for p75NGFR-like immunoreactivity within the CN, the CO staining appears to be primarily associated with postsynaptic cells, while the p75NGFR-like immunostaining appears to be associated with primary afferent terminals.

Organization of the proximal, orbital segment of the infraorbital nerve at multiple intervals after axotomy at birth: a quantitative electron microscopic study in rat

Although much is known of the central consequences of infraorbital nerve (ION) transection at birth, little is known about the effects of this lesion on the organization of the ION itself. To advance our understanding of how deafferentation alters the developing trigeminal neuraxis, 19 newborn rats were subjected to left ION section and perfused 1, 2, 4, 7, 17, or 90 days later. Left IONs were removed in the orbit proximal to the nerve injury site, and axon numbers, types, and fasciculation patterns were assessed with light and electron microscopic methods. Complete axon counts demonstrated that the axotomized ION contained an average (+/- SD) of 13,945 +/- 10,335, 14,112 +/- 3,501, 16,531 +/- 1,904, 9,045 +/- 1,465, 7,018 +/- 4,212, and 8,672 +/- 1,030 axons at the above-listed ages, respectively. These values are well below the 33,059 axons in the normal adult ION (Jacquin et al. [1984] Brain Res. 290:131-135) and the 42,219 axons in the newborn ION (Renehan and Rhoades [1984] Brain Res. 322:369-373). The axotomized ION also contained lower than normal percentages of myelinated axons (26.7% +/- 6.3% on postnatal day 90 vs. 59.7% +/- 6.2% in normal adults). Unmyelinated fibers constituted the vast majority of the remaining fiber types; degenerating fibers never accounted for > 1.6% of all the axons. The number of fascicles making up the axotomized ION overlapped significantly with those found in the normal newborn and adult ION. We conclude that 1) extensive, though variable, axon elimination occurs proximally within one day of the lesion; 2) the 74% reduction in fiber number seen at 90 days is not reliably achieved until postnatal day 7; 3) the higher than normal proportion of unmyelinated axons in the injured ION may underly many of the known effects of neonatal ION injury on the developing whisker-barrel neuraxis; 4) gross changes in ION fasciculation patterns are not prerequisite to injury-induced pattern alterations in the developing trigeminal system.

Enhanced cytochrome-oxidase staining of the cuneate nucleus in the rat reveals a modifiable somatotopic map

Existing cytochrome oxidase (CO)-staining techniques were modified to enhance sensitivity and contrast in order to examine patterns of CO-activity in the dorsal column nuclei (DCN) of adult Long-Evans rats. Within a rostrocaudally limited region in the middle of the cuneate nucleus (CN) distinctive blotches of intense CO-activity were observed. The CO-staining was maximally differentiated approximately 0.3-0.7 mm caudal to the obex. No CO-blotches were observed anywhere else in the DCN. Transganglionic labelling (WGA-HRP) demonstrated that some of the CO-blotches in the rat CN are related to the terminal projection fields of primary afferents from the skin of the forepaws. The corresponding location of primary afferent termination fields and CO-staining patterns supports a tripartite rostrocaudal division in the rat CN, similar to that described by other investigators in cats, monkeys and raccoons. Comparing the patterns of CO-staining to (1) the cytoarchitecture (Nissl-stained sections), or to (2) the dendritoarchitecture (distribution of microtubule-associated protein 2 (MAP2) or to (3) the organization of retrogradely labelled (WGA-HRP/HRP) cuneothalamic cells, revealed no topographical organization corresponding to the CO-blotches. Postnatal (at least up to 11 days postpartum) forepaw deafferentation or removal disrupted the CO-staining pattern in the CN.