- Email: [email protected]
Immunohistochemical and ultrastructural studies of the various nuclei of the trigeminal complex in the human newborn
NeuroscienceVol. 45, No. 1, pp. 23-35, 1991 Printed in Grest Britain
0306-4522/91$3.00+ 0.00 Pergamon Press plc 0 1991IBRO
IMMUNOHISTOCHEMICAL AND ULTRASTRUCTURAL STUDIES OF THE VARIOUS NUCLEI OF THE TRIGEMINAL COMPLEX IN THE HUMAN NEWBORN D. T. YEW,* K. M. PANG?and Y. C. MOK~ *Department of Anatomy, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong tGra.l Biology Unit, Prince Phillips Dental Hospital, University of Hong Kong, Hong Kong AImtract-The various nuclei of the trigeminal complex were studied by immunohistochemical (enkephalin localization) and ultrastructural means in the brainstems of eight newborn human babies that died within 24 h after birth. Positive enkephalin neurons were detected in the chief sensory and spinal trigeminal nuclei as well as in some fibers of the trigeminal nerve. Ultrastructurally, two morphologically distinct types of neuron were observed, respectively, in the motor nucleus, the spinal nucleus and the mesencephalic nucleus of the trigeminal complex, whereas three morphologically distinct types of neuron were observed in the chief sensory nucleus. “Glomerulus” formation was a frequently observed feature in the chief sensory nucleus. In the spinal nucleus, rolls of synaptic terminalsstacking up one on top of anotherand synapsing onto the final synaptic element were very much in evidence. Axosomatic, axodendritic, dendrodendritic and dendroaxonic synapses were demonstrated in all the different nuclear areas of the trigeminal complex but axoaxonic synapses were absent in the mesencephalic nucleus. Some of the findings in the present human study were similar to those reported in the rats and cats.
The trigeminal centers in the central nervous system represent a major governing area of the head and neck region. For the last 30 years, extensive studies on the afferent and efferent connections of the trigeminal complex have been clarified in experimental animals and the major connections with the cerebral cortex, cerebellum, thalamus, red nucleus, brainstem and reticular formation as well as with other cranial nerve components have been well documented in the last three decades.1~2~4~9~10~“~‘3~14 However, little is known about the ultrastructure and immunohistochemistry of the trigeminal complex. The existing literature on ultrastructural aspects appears fragmentary and has been centered around animals.5~7~* Recently, a study on the human motor V nucleus has been presented, but the paper only dealt with the morphometry of the neurons and the neuropil component of that particular nucleus6 A more comprehensive consideration of all the nuclear groups of the trigeminal complex of the central nervous system seems to be necessary. The following is a report on some of the immunohistochemical and ultrastructural features of the trigeminal nuclei in the newborn human.
paraformaldehyde for 3 h. After washing overnight in 20% sucrose phosphate.-buffered saline (PBS), the specimens were cut with a cryostat at 3040 pm. Localization ofenkephalin (ENK) was performed by avidin-biotin-peroxidase complex (ABC) methods. The cut sections were immersed in 0.25% Triton X-100 solution with 3% calf serum and then transferred to rabbit anti-ENK sera (dilution 1: 6ooo;Peninsula, U.S.A.) at 37°C for 3 h. Subsequently, they were put into the 4°C refrigerator for 4S-60 h, followed by reaction with biotinylated antibody (dilution 1: 240, Vector Laboratory, California, U.S.A.) at 37°C for 30 min. Then the slides were transferred into the Vectastain ABC reagent (dilution 1:120; Vector Laboratory, California, U.S.A.) at 37°C for 30 min. The resulting sections were then treated with 3,3’-diaminobenxidine (0.05%) plus 0.01% H,O, and 0.05 M Tris-HCl. Blank and adsorption controls were carried out to ensure antigen specificity. For transmission electron microscopy, the tissues were lixed in 2.5% glutaraldehyde (4°C) in 0.1 M cacodylate bulfer at pH 7.4 for 2 h. They were then washed in cold cacodylate buffer and postlixed for 1 h in 1% GsO, in the same buffer. The specimens were dehydrated in graded alcohol and embedded in Spurr’s resin. Thin sections (5Onm) were cut and stained consecutively with uranyl acetate and lead citrate and observed with a JEOL 1OOCX transmission electron microscope. The percentages of the various types of synaptic contact in each nuclear group of the trigeminal complex were computed after counting 100 randomly selected synapses from photomicrographs of each nuclear group. For scanning electron microscopy on block specimens, the tissues were 6xed in 2.5% glutaraldehyde in 0.1 M cacodylate bulfer for 2 h and washed in the same buffer. After washing and dehydrating through graded alcohol, they were put in Freon TF and absolute alcohol mixture for 15 min followed by three changes of Freon TF for 15 min each. They were then critical point dried and coated with gold. Observations were made with a 35CF JSM scanning electron microscope. For scanning electron microscopy on light microscopic sections, specimens were lixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 2 h and postfixed in 1% 0~0,
EXPERIMENTAL PROCEDURES Eight newborn human babies (CR 34-36 cm) that died in the first 24 h of birth were employed. The causes of death were either aspiration syndrome or other forms of respiratory failure. The dead babies were collected within 1 h after death and the regions of the pons and medulla dissected out. For immunohistochemistry, they were fixed in 4% Abbreviations: ABC, avidin-biotin-peroxidase ENK, enkephalin; PBS, phosphate-buffered
complex; saline. 23
in 0.1 M cacodylate buffer. The specimens were then dehydrated in graded alcohol, cleared in xylene and embedded in paraffin wax. Twenty-five sections (25 pm) were cut and then dewaxed. The sections were then put in absolute alcohol for S-10 min followed by critical point drying using liquid CO, and thm sputter coated with gold. Observations were performed with a JSM 840 JEOL scanning electron microscope. The identifi~tion of the different trigeminai nuclear groups from the pons and medullary areas was performed according to DeArmond et al.’ For the spinal V nucleus, the oralis and interpolaris areas were selected and the area caudalis excluded. For the characterization of the various types of synaptic contact. literature on similar subjects5.7.R.‘2 was consulted. RESULTS
Our results indicated were absent
in the motor
that ENK-positive neurons V nucleus (Figs I, 2) but
were present in the chief sensory V and spinal V nuclei (Figs 3-6). The ENK-positive neurons in these two latter nuclei were in general polymorphic but at least one type of positive neuron in the chief sensory V was flask-shaped (Fig. 4). ENK-positive neurons were absent in the mesencephalic V nucleus (Figs 7,8) but ENK-positive fibers were distinctly evident amongst the trigeminal fibers in the mid pons area (Fig. 9).
Motor V nucleus. Scanning electron microscopy revealed that there were at least two types of motor cell of different sizes (C-4 pm), but both with numerous projecting processes (Figs 10, 11). Transmission electron microscopic studies pointed out that the neurons had large nuclei (Fig. 12), with two or three sets of Golgi apparatus, some ribosomes and mitochondria (Fig. 13). Various types of synaptic contact were evident, e.g. axodend~tic, axosomatic and dendrodendritic contacts (Fig. 12). Some contacts had standard synaptic thickenings whereas others did not display any thickenings (Fig. 12). Some terminals had dense-core vesicles (Fig. 13). Chief sensory V nucleus. Scanning electron microscopy demonstrated at least three types of neuron in this area. One type of neuron (type I) had a flask-
shaped cell body (Fig. 14). From the apex 01 this neuron, thick terminal processes (dendrites) branched further into primary and secondary branches. Other processes also appeared from the sides. The other two types of neuron had processes arising from the entire circumferences of the cell bodies (Figs 15, 16). One type (type II) (Fig. 15) had more slender processes and the other (type III) had thicker primary processes which further branched into thinner subunits (Fig. 16). Transmission electron microscopy showed that all three types of neuron had prominent Golgi apparatus, ribosomes and mitochondria (Figs 17-19). But only type I neuron had stacks of aligned rough endoplasmic reticulum (Fig. 17) which constituted the very conspicuous “Nissl bodies” in the light microscopy. Well-Denny nucleoli were features of types II and III neurons (Figs 18, 19). Axosomatic synaptic contacts were predominantly seen synapsing onto the flask-shaped neurons (Fig. 20). In this area, afferent sensory fibers were identified (Fig. 21). Numerous synapses contacted the afferent fibers and some of these synapses had typical thickenings whereas others did not display thickenings (Fig. 21). Some of these synapses could be clearly identified as axons. Cross-sectional views showed that each afferent fiber was actually surrounded by synapses all along its circumference (Fig. 22) forming a so-called glomerulus and some of the presynaptic terminals had dense-core synaptic vesicles (Fig. 22). Terminals of these afferent fibers had a very high number of synaptic vesicles and each was surrounded by both axons and dendrites (Fig. 22). Spinal V nucleus. Scanning electron microscopy revealed at least two major categories of neurons-one had a rounder cell body (type I, Fig. 23) and another had a more slender and oblong cell body (type II, Fig. 24). Both types had processes with profuse branching. Transmission electron microscopy revealed that the type I neuron had a cell body with prominent Golgi apparatus, ribosomes as well as an abundant and conspicuous group of rough endoplasmic reticulum, along with a few mitochondria (Fig. 25). The type II neuron had Golgi appar-
Fig. 1. Negative control for ~mmunohist~hemi~1 localization of ENK neurons in the motor nucleus of the V. x 553. Fig. 2. Immunohist~hemic~ localization of ENK in motor nucleus of the V. Note no positive neuron could be seen. x 5.53. Fig. 3. Negative control for immunohistochemical localization of ENK neuron in chief sensory nucleus of the V. x 553. Fig. 4. Immunohistochemical localization of ENK in chief sensory nucleus of the V. Note the presence of some positive neurons. One of these was flask-shaped (arrow). x 553. Fig. 5. Negative control for immunohistochemical localization of ENK neurons in spinal nucleus of the v. x5.53. Fig. 6. Immunohistochemical localization of ENK in the spinal nucleus of the V. Note presence of some polymorphic positive neurons (arrows). x 553. Fig. 7. Negative control for immunohistochemical localization of ENK neurons in the mesencephalic nucleus of the V. x 553. Fig. 8. Immunohist~hemi~l localization of ENK in the rn~en~~~alic nucleus of the V. Note lack of convincing positive neurons. x 553. Fig. 9. Some positive ENK fibers (arrow) were found inside the trigeminal nerve root. x 553.
Figs 1-9. 25
Immunohist~~emist~
and ultr~tructure
atus with a less conspicuous group of rough endoplasmic reticulum (Fig. 26). The rest of the type II neurons had cytoplasm with much lighter contrast, with groups of ribosomes interspacing throughout the cytoplasm (Fig. 27). In the area of the spinal V nucieus, axosomatic, axodendritic and axoaxonic contacts were frequently observed (Fig. 28). Many a time, one would see a number of axons contacting each other before synapsing finally with a dendrite or a neuronal cell body (Fig. 28). Afferent fibers were also seen, being contacted by only a few axons or dendrites (Fig. 29). Since these synaptic contacts were not prevailing in number, formation of glomerulus was not evident. Terminals of afferent fibers were also observed but these seemed to have fewer synaptic vesicles (Fig. 29). h&encephalic V nucleus. Scanning electron microscopy demonstrated two types of neuron, I and II. Type I neuron had an elongated cell body with only a few slender processes (Fig. 30) and type II neuron had a triangular cell body with more branches (Fig. 31). Truly unipolar or pseudounipolar neurons could not be identified. Both types of neuron had prominent rough endoplasmic reticulum with some mit~hond~a, Golgi apparatus and ribosomes (Figs 32,33). Many of the neurons had an indented nucleus (Fig. 33). Axosomatic contacts were observed (Fig. 32) although axodendritic contacts were clearly more abundant (Fig. 34). A comparison between the different proportions of synaptic contact in the different nuclear regions of the trigeminal complex is depicted in Fig. 35. There seemed to be a difference in between the groups although axodendritic contacts constituted the major component in all four nuclear groups (Fig. 35). DISCUSSION
Part of this study, using immunohistochemistry, unveils the presence of ENK-positive neurons in two major sensory centers of the trigeminal nuclear complex, namely the chief sensory and spinal nuclei. The localization of inhibitor neuropeptides like ENK in this central nervous system area of the human has, to the best of our knowledge, not been reported. The occurrence of ENK in this area reiterates the importance of neuropeptide modulation on sensory pathways at the brainstem level. Preliminary studies in our laboratory, however, have failed to locate any substance P-positive neurons in this area. The evaluation of the occurrence of ENK neurons in this area during development is still under investigation in our labora-
of human trigeminal nucleus
27
tory. Localization of immunopositive ENK fibers in the trigeminal nerve itself further indicates the possibility of peripheral extension of this inhibitor peptide to more distally located regions. Our scanning electron microscopic and transmission electron microscopic studies have demonstrated rno~holo~~~ly the presence of two types of neuron in the motor nucleus, three types of neuron in the chief. sensory nucleus, two types of neuron in the spinal V nucleus and two types of neuron in the mesencephalic V nucleus. Although all these neurons have the basic cellular components of rough endoplasmic reticulum, Golgi apparatus, ribosomes and mit~hond~a, the abundance of these organelles in the different types of neuron is different. In the study of the human motor V nucleuq6 the amount of Nissl substance has been used as an index for assessing the progressiveness of neuronal development. However, care must be taken to ensure that the same morphological categories of neuron are compared. There have been very few ultrastructural studies on the, trigeminal nuclear complex. The existing ones include Gobel and Dubner’s study’ of the chief sensory nucleus and Kerr’s studies’** on the subnucleus caudalis of the spinal V nucleus. All of these studies have been based on rats and cats. In Gobel and Dubner’s study, they have discovered large neurons in the chief sensory nucleus which are typified by the presence of mainstem (terminal) dendrites which further branch to form primary and secondary branches. These neurons reported in their study appear to closely resemble our flask-shaped neuron in the chief sensory nucleus of our human study. The formation of the “glomerulus” with a central incoming afferent fiber that is surrounded by and synapsing with axons and dendrites of neighboring neurons in situ, has been observed in both the animal studies and our human study. In the study of the spinal V nucleus of cats,‘” glomeruli formation has not been a feature as synaptic contacts from neighboring neurons in situ do not seem to surround the afferent fibers. This is confirmed in our human study as well. On the other hand, rolls of synaptic terminals stacking up one on top of another and finally synapsing onto a final postsynaptic element (e.g. cell body or dendrite) have also been Seen in Kerr’s studies of the cat as well as the present study in the human, although he has focused on the subnucleus caudalis of the spinal V and we have focused on the subnuclei oralis and interpolaris of the spinal V.
Figs IO and Il. Scanning electron microscopy on 2%pm-thick sections showing two types of neuron in the motor V nucteus. Both had cell bodies (B) with numero~ processes. x 510. Fig. 12. Transmission electron microscopy showing axosomatic contact (2) between motor cell body (S), and axon (A) and axodendritic contact (1) between axon (A) and dendrite (D). Note also presence of axoaxonic contact (3) between two axons (A), and dendrodendritic contact {4) between two dendrites (D). de, synapse with dense-core vesicles; N, nucleus of motor c&l of the V. x 11,050. Fig. 13. Cytoplasmic inclusions of a standard motor neuron of the V. Ri, ribosomes; M, mitochondria. Note very prominent Golgi apparatus (G). R, rough endoplasmic reticulum. x 11,050.
28
Figs 14-16. Scanning microscopy of a 25-pm-section showing morphology of three types of neuron in the chief sensory nucleus of the V. In Fig. 14, the neuron was flask-shaped (termed as type I) with major terminal dendrites (arrow) arising from one end of the cell which further divided into primary and secondary dendrites. Neurons in Figs 15 and 16 (termed as types II and III) had branches coming from the entire circumferences of their cell bodies. Type II neurons had some slender branches (arrow). x 2210. Fig. 17. Transmission electron microscopy of a part of the flask-shaped neuron (type I) showing ribosomes (Ri), some rough endoplasmic reticulum (R), Golgi apparatus (G), mitochondria and nucleus (N). x 11,050. Fig. 18. Transmission efectron microscopy of a part of a type II neuron showing two branches (Bl and B2). Well-developed Golgi apparatus (G), mitochondria (M) and nucleus (N). Ri, ribosomes. Note nucleus had a clear nucleolus. x 10,200. Fig. 19. Transmission electron microscopy of a type III neuron showing ribosome (Ri), limited rough endoplasmic reticulum (R), some mitochondria (M) and two Golgi apparatuses (G). Arrows point at the branching points. One distinct branch (B) was observed. x I 1,050. Fig. 20. In the area of the chief sensory nucleus of the V, axosomatic contacts between axon (A) and the cell body (S) of a flask-shaped neuron were observed. x 8500. Fig. 21. An alIerent fiber (F) was seen coming into the chief sensory nucleus of the V. Numerous synaptic contacts were observed along the afferent fiber (arrows). t, contact with synaptic thickening; and arrow without t, contact without synaptic thickening. 1 and 2 were axon terminals with dense-core vesicles. x 11,050. Fig. 22. Two typical glomeruli in the chief sensory nucleus. One glomer~us had a central al&rent fiber (F) surrounded by eight or more synapses. Another glomerulus had a fiber terminal (TF) with synaptic vesicles contacting axons (A) and dendrites (D). x 11,050. Figs 23 and 24. Two types of neuron in the spinal V nuclear region. The neuron (type I) in Fig. 23 had a rounder body (B) and the neuron (type II) in Fig. 24 had a more oblong body (B). x 2210. Fig. 25. Transmission electron microscopy showing a type of I neuron in the spinal V region showing well-develop rough endoplasmic reticulum (R) and Golgi apparatus (G). M, ~t~hondria; Ri, ribosome; N, nucleus. x I 1,050. Figs 26 and 27. Transmission electron microscopy showing two subgroups of type II neurons. One (Fig. 26) had prominent Golgi apparatus (G), a small amount of rough endoplasmic reticulum (R) and a few mitochondria (M). N, nucleus; B, branch. The other neuron (Fig. 27) had less cytoplasmic contacts and had ribosomes (Ri), mitochondria (M) and little rough endoplasmic reticulum (R). x 11,050. Fig. 28. The various synapses in the spinal V region. An axosomatic contact (a) was shown between axon 1 and the body of neuron (S). Also note the chain of axonic contacts I-4 finally ending on a dendrite (D) with a synapse (b) between axon 4 and D. Other axon-axon contacts were indicated in the pairs 5-6 and 8-9. 7, astrocytic process having close contact with axon 6. x 11,050. Fig. 29. In the spinal V region, afferent fibers (F) were seen. Axons (A) and dendrites (D) also synapse on these afferent fibers but no glomerulus was formed. Terminals of these fibers (TF) were also observed but these had less synaptic vesicles than those in the chief sensory nucleus. x 11,050. Figs 30 and 3 1.Scanning electron microscopy of block material (Fig. 30) and of a 25-pm-section (Fig. 31) showing two types of neuron in the mesencephalic V region. Type I (Fig. 30) had few processes and a longer cell body (B). Type II (Fig. 31) had a more triangular body (B). x 850. Fig. 32. In the region of the mesencephalic V nucleus. Axosomatic contacts (arrows) between axons (A) and body of neuron (S). Axodendritic contact between axon (A) and dendrite (D) was also seen. x I 1,050. Fig. 33. The cytopiasmic content of a standard mesencephalic neuron showing an indented nucleus (N), well-developed rough endoplasmic reticulum (R) and Golgi apparatus (G). There were aiso many mitochondria (M) and some ribosomes (Ri). x 11,050. Fig. 34. The basic types of synaptic contact in the mesencephalic V region showing axodendritic contacts (2.3) and axon-dendrite-axon contacts (If and complex dendrodendritic contacts (dendritedendrite dendrite) (4). A, axon; D. dendrite. x 11,050.
Figs 14-19. 29
30
D. T.
YEW
et ui.
Figs 20-22
Figs 23-27. MC45/I-4
31
32
D. T.
YEW
et al.
Figs 28 and 29.
Figs 30-34. 33
D. T. YEW et ul.
34 Motor Nuclaur
Sensory Nucleus
SpinalNucleus
Mesencophalic Nucleus
Contents Fig. 35. A comparison between Percentage of the various types of synaptic contact in the motor V, sensory V (chief sensory), spinal V and mesencepbalic V nuclear regions. A/S, axosomatic contact; A/D, axodendritic contact; D/D, dendrodendritic contact; A/A, axoaxonic contact; D/A, dendroaxonic contact.
For the ultrastructural aspects of the motor and mesencephalic V nuclei, no comparative studies have been demonstrated. However, in the study of the latter, we have not been able to locate any unipolar neuron in the mesencephalic area. The question of whether this type of neuron in newborn is indeed absent in the newborn or that they are appearing in such a small quantity that they become hard to locate needs further investigation employing more specimens. The presence of multipolar neurons in the mesencephalic V has also been documented in the human by Pearson.” As far as synaptic contacts are concerned, axosomatic, axodendritic, dendroaxonic and dendrodendritic synapses have all been observed in all the four major trigeminal nuclei of the human. Axoaxonic contacts, on the other hand, have not been identified in the mesencephalic nucleus of our human specimen. In the studies of the chief sensory V and the spinal V nuclei in the cats and rats,‘*‘*’ different hinds of synaptic contact from those listed in our human studies have been observed. Axosomatic contacts, in the spinal V nucleus, have only been identified in the
substantia gelatinosa of the subnucleus caudalis of the spinal V nucleus in animals.’ The motor V nucleus in the human has been studied by Hamano et aL6 employing morphometric means. They have indicated that them are four developmental stages for the motor V nucleus. Stage 1 (1618 weeks gestation) represents the stage of early differentiation of neuronal cells. Stage 2 (21-27 weeks gestation) signifies a stage of increase in Nissl substance and neuronal processes. Stage 3 (30-40 weeks gestation) represents a stage of maturation (and increase) of neuropil which is the tissue around the neuron. Since a correlated analysis on the other V nuclei has not been done, it is uncertain whether these results are also applicable to the other nuclei. Nevertheless, this presents an interesting view of looking at development from another angle. Furthermore, this paper points out that there has been very little change in the neurons of the trigeminal V motor nucleus in the postnatal period, and if this assumption is accepted then the neuronal characteristics in the different trigeminal nuclei of our newborn human may well represent a similar general pattern in the adult human as well.
REFERENCES
1. Carpenter M. M. and Hanna G. R. (1961) Fiber projections from the spinal trigeminal nucleus in the cat. J. camp. Neural. 117, 117-131. 2. Crosby E. C., Humphrey T. and Latter E. W. (1962) Correlative Anatomy of the Nervous System. Macmillan, New York. 3. DeArmond S. J., Fusco M. M. and Dewey M. M. (1976) Structure of the Human Bruin, 2nd edn. Oxford University Press, New York. 4. Edwards S. G. (1972) The ascending and descending projections of the red nucleus in the cat: an experimental study using an autoradiographic tracing method. Brain Res. Is, 45-63. 5. G&e1 S. and Dubner P. (196Q) Fine structure studies of the main sensory trigeminal mrcleus in the cat and rat. J. camp. Newok 137, 459-494. 6. Hamano S., Goto N. and Nara T. (1988) Development of human motor trigeminal nucleus. Ped. i%urosci. 14,230-235. 7. Kerr F. W. L. (1970) The fine structure of the subnucleus caudalis of the trigeminal nerve. Brain Res. 23, 129-145.
Immunohistochemistry
and ultrastructure
of human trigeminal nucleus
3.5
8. Kerr F. W. L. (1970) The organization of primary at&rents in the subnucleus caudalis of the trigeminal-a light and electron microscope study of degeneration. Brain Res. 23, 147-165. 9. Kuyper H. G. J. M. and Lawrence D. G. (1967) Cortical projections to the red nucleus and brainstem in the rhesus monkey. Brain Res. 4, 151-188. 10. Miller R. A. and Strominger N. L. (1977) An experimental study of the e&rent connections of the superior peduncle in the rhesus monkey. Brain Res. 133, 237-250. 11. Pearson A. A. (1949) The development and connections of the mesencephalic root of the trigeminal nerve in men. J. camp. Neurol. 90, 1-46. 12. Peters A., Palay S. L. and Webster H. (1976) The Fine Structure of Nervous System: The Nervous and Supporting Cells. W.B. Saunders, Philadelphia. 13. Szentagothai J. (1948) Anatomical considerations of monosynaptic reflex arc. J. Neurophysiol. 11, 445-454. 14. Tovik A. (1957) The ascending fibers from the main trigeminal sensory nucleus-an experimental study in the cat. Am. J. Anat. 100, 14. (Accepted
1 May 1991)
0306-4522/91$3.00+ 0.00 Pergamon Press plc 0 1991IBRO
IMMUNOHISTOCHEMICAL AND ULTRASTRUCTURAL STUDIES OF THE VARIOUS NUCLEI OF THE TRIGEMINAL COMPLEX IN THE HUMAN NEWBORN D. T. YEW,* K. M. PANG?and Y. C. MOK~ *Department of Anatomy, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong tGra.l Biology Unit, Prince Phillips Dental Hospital, University of Hong Kong, Hong Kong AImtract-The various nuclei of the trigeminal complex were studied by immunohistochemical (enkephalin localization) and ultrastructural means in the brainstems of eight newborn human babies that died within 24 h after birth. Positive enkephalin neurons were detected in the chief sensory and spinal trigeminal nuclei as well as in some fibers of the trigeminal nerve. Ultrastructurally, two morphologically distinct types of neuron were observed, respectively, in the motor nucleus, the spinal nucleus and the mesencephalic nucleus of the trigeminal complex, whereas three morphologically distinct types of neuron were observed in the chief sensory nucleus. “Glomerulus” formation was a frequently observed feature in the chief sensory nucleus. In the spinal nucleus, rolls of synaptic terminalsstacking up one on top of anotherand synapsing onto the final synaptic element were very much in evidence. Axosomatic, axodendritic, dendrodendritic and dendroaxonic synapses were demonstrated in all the different nuclear areas of the trigeminal complex but axoaxonic synapses were absent in the mesencephalic nucleus. Some of the findings in the present human study were similar to those reported in the rats and cats.
The trigeminal centers in the central nervous system represent a major governing area of the head and neck region. For the last 30 years, extensive studies on the afferent and efferent connections of the trigeminal complex have been clarified in experimental animals and the major connections with the cerebral cortex, cerebellum, thalamus, red nucleus, brainstem and reticular formation as well as with other cranial nerve components have been well documented in the last three decades.1~2~4~9~10~“~‘3~14 However, little is known about the ultrastructure and immunohistochemistry of the trigeminal complex. The existing literature on ultrastructural aspects appears fragmentary and has been centered around animals.5~7~* Recently, a study on the human motor V nucleus has been presented, but the paper only dealt with the morphometry of the neurons and the neuropil component of that particular nucleus6 A more comprehensive consideration of all the nuclear groups of the trigeminal complex of the central nervous system seems to be necessary. The following is a report on some of the immunohistochemical and ultrastructural features of the trigeminal nuclei in the newborn human.
paraformaldehyde for 3 h. After washing overnight in 20% sucrose phosphate.-buffered saline (PBS), the specimens were cut with a cryostat at 3040 pm. Localization ofenkephalin (ENK) was performed by avidin-biotin-peroxidase complex (ABC) methods. The cut sections were immersed in 0.25% Triton X-100 solution with 3% calf serum and then transferred to rabbit anti-ENK sera (dilution 1: 6ooo;Peninsula, U.S.A.) at 37°C for 3 h. Subsequently, they were put into the 4°C refrigerator for 4S-60 h, followed by reaction with biotinylated antibody (dilution 1: 240, Vector Laboratory, California, U.S.A.) at 37°C for 30 min. Then the slides were transferred into the Vectastain ABC reagent (dilution 1:120; Vector Laboratory, California, U.S.A.) at 37°C for 30 min. The resulting sections were then treated with 3,3’-diaminobenxidine (0.05%) plus 0.01% H,O, and 0.05 M Tris-HCl. Blank and adsorption controls were carried out to ensure antigen specificity. For transmission electron microscopy, the tissues were lixed in 2.5% glutaraldehyde (4°C) in 0.1 M cacodylate bulfer at pH 7.4 for 2 h. They were then washed in cold cacodylate buffer and postlixed for 1 h in 1% GsO, in the same buffer. The specimens were dehydrated in graded alcohol and embedded in Spurr’s resin. Thin sections (5Onm) were cut and stained consecutively with uranyl acetate and lead citrate and observed with a JEOL 1OOCX transmission electron microscope. The percentages of the various types of synaptic contact in each nuclear group of the trigeminal complex were computed after counting 100 randomly selected synapses from photomicrographs of each nuclear group. For scanning electron microscopy on block specimens, the tissues were 6xed in 2.5% glutaraldehyde in 0.1 M cacodylate bulfer for 2 h and washed in the same buffer. After washing and dehydrating through graded alcohol, they were put in Freon TF and absolute alcohol mixture for 15 min followed by three changes of Freon TF for 15 min each. They were then critical point dried and coated with gold. Observations were made with a 35CF JSM scanning electron microscope. For scanning electron microscopy on light microscopic sections, specimens were lixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 2 h and postfixed in 1% 0~0,
EXPERIMENTAL PROCEDURES Eight newborn human babies (CR 34-36 cm) that died in the first 24 h of birth were employed. The causes of death were either aspiration syndrome or other forms of respiratory failure. The dead babies were collected within 1 h after death and the regions of the pons and medulla dissected out. For immunohistochemistry, they were fixed in 4% Abbreviations: ABC, avidin-biotin-peroxidase ENK, enkephalin; PBS, phosphate-buffered
complex; saline. 23
in 0.1 M cacodylate buffer. The specimens were then dehydrated in graded alcohol, cleared in xylene and embedded in paraffin wax. Twenty-five sections (25 pm) were cut and then dewaxed. The sections were then put in absolute alcohol for S-10 min followed by critical point drying using liquid CO, and thm sputter coated with gold. Observations were performed with a JSM 840 JEOL scanning electron microscope. The identifi~tion of the different trigeminai nuclear groups from the pons and medullary areas was performed according to DeArmond et al.’ For the spinal V nucleus, the oralis and interpolaris areas were selected and the area caudalis excluded. For the characterization of the various types of synaptic contact. literature on similar subjects5.7.R.‘2 was consulted. RESULTS
Our results indicated were absent
in the motor
that ENK-positive neurons V nucleus (Figs I, 2) but
were present in the chief sensory V and spinal V nuclei (Figs 3-6). The ENK-positive neurons in these two latter nuclei were in general polymorphic but at least one type of positive neuron in the chief sensory V was flask-shaped (Fig. 4). ENK-positive neurons were absent in the mesencephalic V nucleus (Figs 7,8) but ENK-positive fibers were distinctly evident amongst the trigeminal fibers in the mid pons area (Fig. 9).
Motor V nucleus. Scanning electron microscopy revealed that there were at least two types of motor cell of different sizes (C-4 pm), but both with numerous projecting processes (Figs 10, 11). Transmission electron microscopic studies pointed out that the neurons had large nuclei (Fig. 12), with two or three sets of Golgi apparatus, some ribosomes and mitochondria (Fig. 13). Various types of synaptic contact were evident, e.g. axodend~tic, axosomatic and dendrodendritic contacts (Fig. 12). Some contacts had standard synaptic thickenings whereas others did not display any thickenings (Fig. 12). Some terminals had dense-core vesicles (Fig. 13). Chief sensory V nucleus. Scanning electron microscopy demonstrated at least three types of neuron in this area. One type of neuron (type I) had a flask-
shaped cell body (Fig. 14). From the apex 01 this neuron, thick terminal processes (dendrites) branched further into primary and secondary branches. Other processes also appeared from the sides. The other two types of neuron had processes arising from the entire circumferences of the cell bodies (Figs 15, 16). One type (type II) (Fig. 15) had more slender processes and the other (type III) had thicker primary processes which further branched into thinner subunits (Fig. 16). Transmission electron microscopy showed that all three types of neuron had prominent Golgi apparatus, ribosomes and mitochondria (Figs 17-19). But only type I neuron had stacks of aligned rough endoplasmic reticulum (Fig. 17) which constituted the very conspicuous “Nissl bodies” in the light microscopy. Well-Denny nucleoli were features of types II and III neurons (Figs 18, 19). Axosomatic synaptic contacts were predominantly seen synapsing onto the flask-shaped neurons (Fig. 20). In this area, afferent sensory fibers were identified (Fig. 21). Numerous synapses contacted the afferent fibers and some of these synapses had typical thickenings whereas others did not display thickenings (Fig. 21). Some of these synapses could be clearly identified as axons. Cross-sectional views showed that each afferent fiber was actually surrounded by synapses all along its circumference (Fig. 22) forming a so-called glomerulus and some of the presynaptic terminals had dense-core synaptic vesicles (Fig. 22). Terminals of these afferent fibers had a very high number of synaptic vesicles and each was surrounded by both axons and dendrites (Fig. 22). Spinal V nucleus. Scanning electron microscopy revealed at least two major categories of neurons-one had a rounder cell body (type I, Fig. 23) and another had a more slender and oblong cell body (type II, Fig. 24). Both types had processes with profuse branching. Transmission electron microscopy revealed that the type I neuron had a cell body with prominent Golgi apparatus, ribosomes as well as an abundant and conspicuous group of rough endoplasmic reticulum, along with a few mitochondria (Fig. 25). The type II neuron had Golgi appar-
Fig. 1. Negative control for ~mmunohist~hemi~1 localization of ENK neurons in the motor nucleus of the V. x 553. Fig. 2. Immunohist~hemic~ localization of ENK in motor nucleus of the V. Note no positive neuron could be seen. x 5.53. Fig. 3. Negative control for immunohistochemical localization of ENK neuron in chief sensory nucleus of the V. x 553. Fig. 4. Immunohistochemical localization of ENK in chief sensory nucleus of the V. Note the presence of some positive neurons. One of these was flask-shaped (arrow). x 553. Fig. 5. Negative control for immunohistochemical localization of ENK neurons in spinal nucleus of the v. x5.53. Fig. 6. Immunohistochemical localization of ENK in the spinal nucleus of the V. Note presence of some polymorphic positive neurons (arrows). x 553. Fig. 7. Negative control for immunohistochemical localization of ENK neurons in the mesencephalic nucleus of the V. x 553. Fig. 8. Immunohist~hemi~l localization of ENK in the rn~en~~~alic nucleus of the V. Note lack of convincing positive neurons. x 553. Fig. 9. Some positive ENK fibers (arrow) were found inside the trigeminal nerve root. x 553.
Figs 1-9. 25
Immunohist~~emist~
and ultr~tructure
atus with a less conspicuous group of rough endoplasmic reticulum (Fig. 26). The rest of the type II neurons had cytoplasm with much lighter contrast, with groups of ribosomes interspacing throughout the cytoplasm (Fig. 27). In the area of the spinal V nucieus, axosomatic, axodendritic and axoaxonic contacts were frequently observed (Fig. 28). Many a time, one would see a number of axons contacting each other before synapsing finally with a dendrite or a neuronal cell body (Fig. 28). Afferent fibers were also seen, being contacted by only a few axons or dendrites (Fig. 29). Since these synaptic contacts were not prevailing in number, formation of glomerulus was not evident. Terminals of afferent fibers were also observed but these seemed to have fewer synaptic vesicles (Fig. 29). h&encephalic V nucleus. Scanning electron microscopy demonstrated two types of neuron, I and II. Type I neuron had an elongated cell body with only a few slender processes (Fig. 30) and type II neuron had a triangular cell body with more branches (Fig. 31). Truly unipolar or pseudounipolar neurons could not be identified. Both types of neuron had prominent rough endoplasmic reticulum with some mit~hond~a, Golgi apparatus and ribosomes (Figs 32,33). Many of the neurons had an indented nucleus (Fig. 33). Axosomatic contacts were observed (Fig. 32) although axodendritic contacts were clearly more abundant (Fig. 34). A comparison between the different proportions of synaptic contact in the different nuclear regions of the trigeminal complex is depicted in Fig. 35. There seemed to be a difference in between the groups although axodendritic contacts constituted the major component in all four nuclear groups (Fig. 35). DISCUSSION
Part of this study, using immunohistochemistry, unveils the presence of ENK-positive neurons in two major sensory centers of the trigeminal nuclear complex, namely the chief sensory and spinal nuclei. The localization of inhibitor neuropeptides like ENK in this central nervous system area of the human has, to the best of our knowledge, not been reported. The occurrence of ENK in this area reiterates the importance of neuropeptide modulation on sensory pathways at the brainstem level. Preliminary studies in our laboratory, however, have failed to locate any substance P-positive neurons in this area. The evaluation of the occurrence of ENK neurons in this area during development is still under investigation in our labora-
of human trigeminal nucleus
27
tory. Localization of immunopositive ENK fibers in the trigeminal nerve itself further indicates the possibility of peripheral extension of this inhibitor peptide to more distally located regions. Our scanning electron microscopic and transmission electron microscopic studies have demonstrated rno~holo~~~ly the presence of two types of neuron in the motor nucleus, three types of neuron in the chief. sensory nucleus, two types of neuron in the spinal V nucleus and two types of neuron in the mesencephalic V nucleus. Although all these neurons have the basic cellular components of rough endoplasmic reticulum, Golgi apparatus, ribosomes and mit~hond~a, the abundance of these organelles in the different types of neuron is different. In the study of the human motor V nucleuq6 the amount of Nissl substance has been used as an index for assessing the progressiveness of neuronal development. However, care must be taken to ensure that the same morphological categories of neuron are compared. There have been very few ultrastructural studies on the, trigeminal nuclear complex. The existing ones include Gobel and Dubner’s study’ of the chief sensory nucleus and Kerr’s studies’** on the subnucleus caudalis of the spinal V nucleus. All of these studies have been based on rats and cats. In Gobel and Dubner’s study, they have discovered large neurons in the chief sensory nucleus which are typified by the presence of mainstem (terminal) dendrites which further branch to form primary and secondary branches. These neurons reported in their study appear to closely resemble our flask-shaped neuron in the chief sensory nucleus of our human study. The formation of the “glomerulus” with a central incoming afferent fiber that is surrounded by and synapsing with axons and dendrites of neighboring neurons in situ, has been observed in both the animal studies and our human study. In the study of the spinal V nucleus of cats,‘” glomeruli formation has not been a feature as synaptic contacts from neighboring neurons in situ do not seem to surround the afferent fibers. This is confirmed in our human study as well. On the other hand, rolls of synaptic terminals stacking up one on top of another and finally synapsing onto a final postsynaptic element (e.g. cell body or dendrite) have also been Seen in Kerr’s studies of the cat as well as the present study in the human, although he has focused on the subnucleus caudalis of the spinal V and we have focused on the subnuclei oralis and interpolaris of the spinal V.
Figs IO and Il. Scanning electron microscopy on 2%pm-thick sections showing two types of neuron in the motor V nucteus. Both had cell bodies (B) with numero~ processes. x 510. Fig. 12. Transmission electron microscopy showing axosomatic contact (2) between motor cell body (S), and axon (A) and axodendritic contact (1) between axon (A) and dendrite (D). Note also presence of axoaxonic contact (3) between two axons (A), and dendrodendritic contact {4) between two dendrites (D). de, synapse with dense-core vesicles; N, nucleus of motor c&l of the V. x 11,050. Fig. 13. Cytoplasmic inclusions of a standard motor neuron of the V. Ri, ribosomes; M, mitochondria. Note very prominent Golgi apparatus (G). R, rough endoplasmic reticulum. x 11,050.
28
Figs 14-16. Scanning microscopy of a 25-pm-section showing morphology of three types of neuron in the chief sensory nucleus of the V. In Fig. 14, the neuron was flask-shaped (termed as type I) with major terminal dendrites (arrow) arising from one end of the cell which further divided into primary and secondary dendrites. Neurons in Figs 15 and 16 (termed as types II and III) had branches coming from the entire circumferences of their cell bodies. Type II neurons had some slender branches (arrow). x 2210. Fig. 17. Transmission electron microscopy of a part of the flask-shaped neuron (type I) showing ribosomes (Ri), some rough endoplasmic reticulum (R), Golgi apparatus (G), mitochondria and nucleus (N). x 11,050. Fig. 18. Transmission efectron microscopy of a part of a type II neuron showing two branches (Bl and B2). Well-developed Golgi apparatus (G), mitochondria (M) and nucleus (N). Ri, ribosomes. Note nucleus had a clear nucleolus. x 10,200. Fig. 19. Transmission electron microscopy of a type III neuron showing ribosome (Ri), limited rough endoplasmic reticulum (R), some mitochondria (M) and two Golgi apparatuses (G). Arrows point at the branching points. One distinct branch (B) was observed. x I 1,050. Fig. 20. In the area of the chief sensory nucleus of the V, axosomatic contacts between axon (A) and the cell body (S) of a flask-shaped neuron were observed. x 8500. Fig. 21. An alIerent fiber (F) was seen coming into the chief sensory nucleus of the V. Numerous synaptic contacts were observed along the afferent fiber (arrows). t, contact with synaptic thickening; and arrow without t, contact without synaptic thickening. 1 and 2 were axon terminals with dense-core vesicles. x 11,050. Fig. 22. Two typical glomeruli in the chief sensory nucleus. One glomer~us had a central al&rent fiber (F) surrounded by eight or more synapses. Another glomerulus had a fiber terminal (TF) with synaptic vesicles contacting axons (A) and dendrites (D). x 11,050. Figs 23 and 24. Two types of neuron in the spinal V nuclear region. The neuron (type I) in Fig. 23 had a rounder body (B) and the neuron (type II) in Fig. 24 had a more oblong body (B). x 2210. Fig. 25. Transmission electron microscopy showing a type of I neuron in the spinal V region showing well-develop rough endoplasmic reticulum (R) and Golgi apparatus (G). M, ~t~hondria; Ri, ribosome; N, nucleus. x I 1,050. Figs 26 and 27. Transmission electron microscopy showing two subgroups of type II neurons. One (Fig. 26) had prominent Golgi apparatus (G), a small amount of rough endoplasmic reticulum (R) and a few mitochondria (M). N, nucleus; B, branch. The other neuron (Fig. 27) had less cytoplasmic contacts and had ribosomes (Ri), mitochondria (M) and little rough endoplasmic reticulum (R). x 11,050. Fig. 28. The various synapses in the spinal V region. An axosomatic contact (a) was shown between axon 1 and the body of neuron (S). Also note the chain of axonic contacts I-4 finally ending on a dendrite (D) with a synapse (b) between axon 4 and D. Other axon-axon contacts were indicated in the pairs 5-6 and 8-9. 7, astrocytic process having close contact with axon 6. x 11,050. Fig. 29. In the spinal V region, afferent fibers (F) were seen. Axons (A) and dendrites (D) also synapse on these afferent fibers but no glomerulus was formed. Terminals of these fibers (TF) were also observed but these had less synaptic vesicles than those in the chief sensory nucleus. x 11,050. Figs 30 and 3 1.Scanning electron microscopy of block material (Fig. 30) and of a 25-pm-section (Fig. 31) showing two types of neuron in the mesencephalic V region. Type I (Fig. 30) had few processes and a longer cell body (B). Type II (Fig. 31) had a more triangular body (B). x 850. Fig. 32. In the region of the mesencephalic V nucleus. Axosomatic contacts (arrows) between axons (A) and body of neuron (S). Axodendritic contact between axon (A) and dendrite (D) was also seen. x I 1,050. Fig. 33. The cytopiasmic content of a standard mesencephalic neuron showing an indented nucleus (N), well-developed rough endoplasmic reticulum (R) and Golgi apparatus (G). There were aiso many mitochondria (M) and some ribosomes (Ri). x 11,050. Fig. 34. The basic types of synaptic contact in the mesencephalic V region showing axodendritic contacts (2.3) and axon-dendrite-axon contacts (If and complex dendrodendritic contacts (dendritedendrite dendrite) (4). A, axon; D. dendrite. x 11,050.
Figs 14-19. 29
30
D. T.
YEW
et ui.
Figs 20-22
Figs 23-27. MC45/I-4
31
32
D. T.
YEW
et al.
Figs 28 and 29.
Figs 30-34. 33
D. T. YEW et ul.
34 Motor Nuclaur
Sensory Nucleus
SpinalNucleus
Mesencophalic Nucleus
Contents Fig. 35. A comparison between Percentage of the various types of synaptic contact in the motor V, sensory V (chief sensory), spinal V and mesencepbalic V nuclear regions. A/S, axosomatic contact; A/D, axodendritic contact; D/D, dendrodendritic contact; A/A, axoaxonic contact; D/A, dendroaxonic contact.
For the ultrastructural aspects of the motor and mesencephalic V nuclei, no comparative studies have been demonstrated. However, in the study of the latter, we have not been able to locate any unipolar neuron in the mesencephalic area. The question of whether this type of neuron in newborn is indeed absent in the newborn or that they are appearing in such a small quantity that they become hard to locate needs further investigation employing more specimens. The presence of multipolar neurons in the mesencephalic V has also been documented in the human by Pearson.” As far as synaptic contacts are concerned, axosomatic, axodendritic, dendroaxonic and dendrodendritic synapses have all been observed in all the four major trigeminal nuclei of the human. Axoaxonic contacts, on the other hand, have not been identified in the mesencephalic nucleus of our human specimen. In the studies of the chief sensory V and the spinal V nuclei in the cats and rats,‘*‘*’ different hinds of synaptic contact from those listed in our human studies have been observed. Axosomatic contacts, in the spinal V nucleus, have only been identified in the
substantia gelatinosa of the subnucleus caudalis of the spinal V nucleus in animals.’ The motor V nucleus in the human has been studied by Hamano et aL6 employing morphometric means. They have indicated that them are four developmental stages for the motor V nucleus. Stage 1 (1618 weeks gestation) represents the stage of early differentiation of neuronal cells. Stage 2 (21-27 weeks gestation) signifies a stage of increase in Nissl substance and neuronal processes. Stage 3 (30-40 weeks gestation) represents a stage of maturation (and increase) of neuropil which is the tissue around the neuron. Since a correlated analysis on the other V nuclei has not been done, it is uncertain whether these results are also applicable to the other nuclei. Nevertheless, this presents an interesting view of looking at development from another angle. Furthermore, this paper points out that there has been very little change in the neurons of the trigeminal V motor nucleus in the postnatal period, and if this assumption is accepted then the neuronal characteristics in the different trigeminal nuclei of our newborn human may well represent a similar general pattern in the adult human as well.
REFERENCES
1. Carpenter M. M. and Hanna G. R. (1961) Fiber projections from the spinal trigeminal nucleus in the cat. J. camp. Neural. 117, 117-131. 2. Crosby E. C., Humphrey T. and Latter E. W. (1962) Correlative Anatomy of the Nervous System. Macmillan, New York. 3. DeArmond S. J., Fusco M. M. and Dewey M. M. (1976) Structure of the Human Bruin, 2nd edn. Oxford University Press, New York. 4. Edwards S. G. (1972) The ascending and descending projections of the red nucleus in the cat: an experimental study using an autoradiographic tracing method. Brain Res. Is, 45-63. 5. G&e1 S. and Dubner P. (196Q) Fine structure studies of the main sensory trigeminal mrcleus in the cat and rat. J. camp. Newok 137, 459-494. 6. Hamano S., Goto N. and Nara T. (1988) Development of human motor trigeminal nucleus. Ped. i%urosci. 14,230-235. 7. Kerr F. W. L. (1970) The fine structure of the subnucleus caudalis of the trigeminal nerve. Brain Res. 23, 129-145.
Immunohistochemistry
and ultrastructure
of human trigeminal nucleus
3.5
8. Kerr F. W. L. (1970) The organization of primary at&rents in the subnucleus caudalis of the trigeminal-a light and electron microscope study of degeneration. Brain Res. 23, 147-165. 9. Kuyper H. G. J. M. and Lawrence D. G. (1967) Cortical projections to the red nucleus and brainstem in the rhesus monkey. Brain Res. 4, 151-188. 10. Miller R. A. and Strominger N. L. (1977) An experimental study of the e&rent connections of the superior peduncle in the rhesus monkey. Brain Res. 133, 237-250. 11. Pearson A. A. (1949) The development and connections of the mesencephalic root of the trigeminal nerve in men. J. camp. Neurol. 90, 1-46. 12. Peters A., Palay S. L. and Webster H. (1976) The Fine Structure of Nervous System: The Nervous and Supporting Cells. W.B. Saunders, Philadelphia. 13. Szentagothai J. (1948) Anatomical considerations of monosynaptic reflex arc. J. Neurophysiol. 11, 445-454. 14. Tovik A. (1957) The ascending fibers from the main trigeminal sensory nucleus-an experimental study in the cat. Am. J. Anat. 100, 14. (Accepted
1 May 1991)