A putative retinohypothalamic projection containing substance P in the human

BRAIN RESEARCH ELSEVIER

Brain Research 659 (1994) 249-253

Short communication

A putative retinohypothalamic projection containing substance P in the human 1 Robert Y. Moore *, Joan C. Speh Departments of Psychiatry, Neurology and Neuroscience, Center for Neuroscience, A lzheimer Disease Research Center and the Center for the Neural Basis of Cognition, BST W1656, Universi~ of Pittsburgh, Pittsburgh, PA 15261, USA Accepted 28 June 1994

Abstract

The retinohypothalamic tract (RHT) is the principal pathway mediating the entraining effects of light on the circadian pacemaker, the suprachiasmatic nucleus (SCN). In the rat, the RHT has two components, one which projects to the SCN and the intergeniculate leaflet of the thalamus and has no known peptide content and one which projects to the SCN and, perhaps, to the olivary pretectal nucleus and contains substance P (SP). Both terminate predominantly in a zone of the SCN that contains vasoactive intestinal polypeptide (VIP)-producing neurons. In the human, there is a similar dense axonal plexus of SP-immunoreactive axons in the SCN located largely in the area occupied by VIP-immunoreactive neurons and distinct from other SP-immunoreactive axons in the area. We propose that this SP plexus represents a component of the RHT in thc human brain.

Keywords: Circadian rhythm; Entrainment; Suprachiasmatic nucleus; Intergeniculate leaflet; Retinal ganglion cell

The retinohypothalamic tract (RHT) is sufficient to maintain entrainment of circadian function [7] and transection of the R H T as it enters the suprachiasmatic nucleus (SCN) results in loss of entrainment of the locomotor activity rhythm [6]. The R H T originates from a distinct subset of ganglion cells that project to the SCN and to the intergeniculate leaflet (IGL) of the lateral geniculate complex [1,16,20]. There are recent data, however, which indicate that there is another set of retinal ganglion cells that produces substance P (SP) and projects to the SCN and, perhaps, to the olivary pretectal nucleus but not to the I G L in the rat [11,12,24,25]. The R H T is present in virtually all mammals including primates [2,13,14,17]. Evidence for an R H T in the human has been found using a method believed to show degenerating axons [21] and using DiI [3]. In the first study [21], degenerating axons were described in the SCN at very long intervals after loss of an eye and,

* Corresponding author. Fax: (1) (412) 648-8376. t A preliminary report of these data was presented at the 1993 Society for Neuroscience Meeting. 0006-8993/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0 0 0 6 - 8 9 9 3 ( 9 4 ) 0 0 7 8 1 - 0

in the second [3], labeled axons were described entering the SCN from the optic chiasm. However, the exact distribution and the organization of the terminal plexus was not discernible using these techniques. The description of SP-immunoreactive (SP + ) R H T axons in the rodent suggested that the human R H T also might be demonstrable with antisera to SP. The intent of the present study was to determine whether there is a SP + axonal plexus in the human SCN that conforms in distribution to what would be expected of the R H T as this has been shown in other mammals. Ten hypothalami were obtained from brains of individuals without known neurological or psychiatric disease from routine postmortem examination. The brains were from individuals of both sexes (six male, four female; age 22-75 years). Postmortem intervals were 5-20 hours and the brains were fixed by immersion in 10% buffered formaldehyde solution for 1-2 weeks. The hypothalamus was removed as a single block from each, immersed in graded sucrose solutions and sectioned on a freezing microtome at 40 g m in the coronal plane. Series of sections from all hypothalami were prepared with antisera to vasopressin (VP, Peninsula Labs.), vasoactive intestinal polypeptide (VIP, IncStar) and SP. Two antisera against SP were used in the

250

R. Y. More. ,I.(.Speh / Brain Research 65Q (1994) 249 253

study; one was obtained from Dr. Susan Leeman, Boston University School of Medicine and the other from Dr. Koichi Takatsuji, Osaka Prefectural College of Nursing. The two antisera gave essentially identical results. Immunoreactivity is blocked with each antiserum by absorbing the antiserum with authentic peptide in concentrations of 10-50 g M . Standard immunohistochemical techniques using the avidin-biotin procedure of Hsu et al. [4] were employed; the details of these have been described elsewhere [15,17]. The appearance of the human SCN demonstrated in our material p r e p a r e d with antisera to VP and VIP has been described previously [15]. The SCN is the most rostral nucleus evident in sections through the hypothalamus. The length of the SCN in the brains examined in this study is approximately 4 mm. For the rostral 1 mm, the SCN is a sparse collection of cells immediately dorsal to the optic chiasm and lateral to the third ventricle in a small pyramid of tissue that forms the base of the lamina terminalis. In material stained with antisera to VIP, this most rostral extent of the SCN is characterized by immunoreactive neurons and an associated axonal plexus in the ventral SCN and extending into the dorsal third of the chiasm as pockets of neurons and axons (Fig. 1). At these levels and continuing for the remainder of the SCN, the VIP +

neurons and axons are surrounded by a population of VP + neurons and axons. There is a dense plexus o1: fine, varicose SP + axons that overlaps the zone occupied by the VIP + elements. Over the next 1 ram, the SCN elongates dorsally. As this occurs, the area occupied by VIP + neurons and axons and the SP + axonal plexus, expands dorsally but VIP + and SP + elements are no longer present in the chiasm. The SP + axons form a dense plexus in the ventral SCN (Fig. 2) with scattered axons extending into the VP + area and through it into the adjacent anterior hypothalamic area. At rostrocaudal levels of SCN from 2 - 3 m m from the rostral border, the nucleus continues to expand dorsally and in the mediolateral direction. As this occurs, both the VIP + and SP + areas expand (Fig. 1). At these levels, however, SP + neurons and an associated plexus appear in the anterior hypothalamic area and this gradually extends into the SCN. The axons in this plexus tend to be coarser than those that are present more rostrally in the SCN but in the caudal 1 mm of SCN the distinction between the intrinsic SCN plexus and that arising from the anterior hypothalamic area becomes lost. No SP + neurons are noted within the SCN at any level. The SP + plexus is present in the SCN in all ten brains examined. The density of the plexus varies but this occurs in association with changes

i

,.,<+"c"--

"

:

Fig, I. Diagrams of coronal sections through the human S C N showing the distributionof the VIP + neurons and axonal plexus (A-C) and of the SP + axonal plexus (D-F) at three levels from the rostral SCN to approximately 3 mm (C,F) caudal to the rostral pole of the nucleus. The drawings were made from adjacent sections to illustrate the overlap of the SP + plexus with the VIP + perikarya and plexus. Abbreviations: 3V, third ventricle; OC optic chiasm; OVLT, organum vasculosum lamina terminalis; SCN, suprachiasmatic nucleus; SON, supraoptic nucleus.

R.Y. More, J.C. Speh /Brain Research 659 (1994) 249-253

251

Fig. 2. Darkfield photomicrographstaken of adjacent sections at the level of the human SCN shown in Fig. 1B,E. The section in A was prepared with an antiserum to VIP and shows immunoreactive cells and axons. The bright material in the optic chiasm is artifact. The section in B was prepared with an antiserum to SP and shows a dense plexus of immunorcactive axons. The asterisks in each figure mark blood vessels. The marker bar = 40/zm.

in the intensity of the VP staining and the intensity of the SP staining in the anterior hypothalamic area. These differences in the quality of the immunocytochemistry appear to be a function of the length of the postmortem interval and premortem agonal factors that affect fixation. The R H T projection to the SCN is sufficient for the maintenance of entrainment [7] and it may be necessary [6]. In the most studied mammals, rodents, the R H T has three components: (1) a major projection terminating in the SCN; (2) lesser projections to the anterior hypothalamic area -subparaventricular zone and retrochiasmatic area that appear to arise, in large part, from the projection to the SCN; (3) a smaller projection to the lateral hypothalamic area [5,9]. The pattern of R H T projections in non-human primates is similar (R.Y. Moore, unpublished results) and, although there are reports of R H T projections in the human, these provide little detailed information. For example, Sadun and collaborators [21] report evidence for axonal debris in the SCN of individuals that suf-

fered enucleation many months prior to death and Friedman et al. [3] show DiI-labeled fibers leaving the optic chiasm to enter the SCN. In neither instance, however, does the material permit a detailed description of the distribution of retinal afferents or the pattern of terminal innervation of the SCN. In the course of analyzing the chemical neuroanatomy of the human SCN [15], we prepared a series of sections through ten brains with antisera to SP. We observed a plexus of SP + axons in the SCN in this material but did not appreciate its significance until Takatsuji and colleagues [24,25] demonstrated a dense plexus of SP + axons in the ventrolateral SCN of the rat which disappeared after bilateral enucleation. It is worth noting, however, that this plexus in the rat has not been observed by all investigators. Larsen [8] in a very detailed paper shows scattered immunoreactive perikarya in the dorsomedial SCN and a loose and even plexus over the entire nucleus. Similarly, Otori et al. [19] show only a scattered SP + plexus in the SCN that they report unaffected by enuc[eation. In contrast,

252

R. ~ More. J.( . 5peh / Brain Research 659 (/994) 249-253

Mikkelsen and Larsen [11] demonstrate a SP + plexus in the area innervated by the R H T in the middle two-thirds of the SCN. In their study, as well as that of Larsen [8], SP + perikarya also are identified and Mikkelsen and Larsen [11] show that a population of neurons in the dorsomedial SCN, corresponding to that demonstrated by immunocytochemistry, contains SP message. It appears to us that the difference between the Takatsuji and Mikkelsen studies and those of Otori et al. [19] and Larsen, [8] is one of sensitivity of the antisera employed. In earlier work in the rat, we used a commercial antiserum and observed only a sparse SP + plexus distributed evenly over the SCN (R.Y. Moore and J.C. Speh, unpublished results). However, with both the Leeman and Takatsuji antisera, we observe a dense plexus in the ventrolateral SCN with the distribution shown by both Takatsuji et al. [24] and Mikkelsen and Larsen [11]. This plexus differs from that shown by anterograde transport of lectins [5,9,22] in that it appears to be present only over the rostral part of the SCN, approximately the rostral three-fourths of the nucleus. Thus, we conclude with Takatsuji and colleagues [24,25] and Mikkelsen and Larsen [11], that there is a SP + projection from retinal ganglion cells which terminates in the ventrolateral SCN and represents a portion of the total R H T shown in anterograde transport studies [5,9,22]. Collaterals of these SP + axons may also innervate the olivary pretectal nucleus [12,25]. The presence of SP + perikarya in the SCN also has been reported in the hamster [18] and in the monkey [10]. SP + neurons have been reported in the intergeniculatc leaflet [23] and it is possible that these give rise to some of the plexus in the SCN that is independent of the retinal innervation. On the basis of the conclusion that there arc SP + retinal ganglion cells in the rat projecting to the SCN, we re-examined all of our human brains in which series of sections through the SCN were prepared with SP antisera. In our view, this material provides convincing evidence that there is a SP + component of the R H T in the human. All of the brains prepared with the SP antisera show a plexus of varicose axons in the region of the SCN where one would expect to find the R H T terminal zone. That is, in rodents [2,5,9,22] and in the monkey (R.Y. Moore, unpublished results), the densest terminal zone of the R H T overlies the region occupied by VIP + cell bodies and surrounded by VP + cell bodies. In the human, the distribution of V1P + and VP + cell bodies is very similar to that in other m a m mals [15]. In the zone in which V1P + perikarya are present, there is a relatively dense plexus of fine, varicose SP + axons. This extends from the very rostral SCN, where many of the neurons of the nucleus are e m b e d d e d in the optic chiasm, to at least the caudal third of the SCN. In the rostral half of the SCN, SP + fibers appear to enter the nucleus from the ventral

border, the optic chiasm, to ramify and form a dense plexus within the VIP + cell zone. At these levels. scattered SP + axons extend beyond the SCN into the adjacent A H A in a pattern like the retinohypothalamic projection in the rat [5,9] and there is only a very sparse plexus of coarse, varicose SP + axons in the A H A that arises from neurons intrinsic to it. In the A H A around the caudal half of the SCN, there are increasing numbers of SP + neurons with an associated plexus that appears in the dorsal A H A and these extend caudally to surround the SCN. In the caudal SCN, SP + axons extend into the periphery of the nucleus and, at the most caudal levels, it is impossible to distinguish the plexus within the SCN from that invading it from adjacent hypothalamus. One interpretation of these data is that, as in the rat, there is a component of the human R H T that is SP + and innervates the rostral two-thirds to threefourths of the nucleus but is replaced caudally by an intrinsic hypothalamic plexus. Another possible interpretation is that there is an intrinsic SP-containing neuronal population in the SCN that gives rise to the SP + axonal plexus. We think the latter is unlikely because we have no difficulty showing VIP + , VP + , neurotensin and neuropeptide Y-immunoreactive neurons in the human SCN [15,17], and SP + neurons in other areas and we do not see any SP + neurons in the SCN in any of our cases. The interpretation of normal material is often difficult, but it appears to us reasonable to conclude that there is a SP + component of the human R H T that innervates the SCN and adjacent anterior hypothalamus. This work was supported by a Grant from the AFOSR. We are grateful to Koichi Takatsuji and Susan Leeman for their kind gifts of SP antisera and to Nadine Suhan for skilled technical assistance. [1] Card, J.P., Whealy, M.E., Robbins, A.K., Moore, R.Y. and Enquist, L.W., Two alpha herpes virus strains are transported differentially in the rodent visual system, Neuron, 6 (1991) 957-969. [2] Cassone, V.M., Speh, J.C., Card, J.P. and Moore, R.Y., Comparative anatomy of mammalian suprachiasmatic nucleus, J. Biol. Rhythms, 3 (1988) 71-91. [3] Friedman, D.I., Johnson, J.K., Chorsky, R.E. and Stopa, E.G., Labeling of human retinohypothalamic tract with the carbocyanine dye, DiI, Brain Res., 560 (1991) 297-~12. [4] Hsu, S.M., Raine, L. and Fanger, H., The use of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures, J. Histochem. Cytochem., 29 (1981) 577 580. [5] Johnson, R.F., Moore, R.Y. and Morin, L.P., Retinohypothalamic projections in the rat and hamster demonstrated using cholera toxin, Brain Res., 462 (1988) 301-312. [6] Johnson, R.F., Moore, R.Y. and Morin, L.P., Loss of entrainment and anatomical plasticity after lesions of the hamster retinohypothalamic tract, Brain Res., 460 (1988) 297-313. [7] Klein, D.C. and Moore, R.Y.. Pineal N-acetyltransferase and

R.Y. More, J.C. Speh /Brain Research 659 (1994) 249-253 hydroxyindole-o-methyltranferase: control by the retino hypothalamic tract and the suprachiasmatic nucleus, Brain Res., 174 (1979) 245-262. [8] Larsen, P.J., Distribution of substance P-immunoreactive elements in the preoptic area and the hypothamus of the rat, Z Comp. Neurol., 316 (1992) 287-313. [9] Levine, J.D., Weiss, M.L., Rosenwasser, A.M. and Miselis, R.R., Retinohypothalamic tract in the female albino rat, J. Comp. Neurol., 306 (1991) 344-360. [10] Mick, G., Najimid, M., Girard, M. and Chayvialle, J.-A., Evidence for substance P containing subpopulation in the primate suprachiasmatic nucleus, Brain Res., 573 (1992) 311-317. [11] Mikkelsen, J.D. and Larsen, P.J., Substance P in the suprachiasmatic nucleus of the rat: an immunohistochemical and in situ hybridization study, Histochemistry, 100 (1993) 3-16. [12] Miquel-Hidalgo, J.-J., Senba, E., Takatsuji, K. and Tohyama, M., Effects of eye enucleation on substance P-immunoreactive fibers of some retinorecipient nuclei of the rat in relation to their origin from the superior colliculus, Neuroscience, 44 (1991) 235-243. [13] Moore, R.Y. and Lenn, N.J., A retinohypothalamic projection in the rat, J. Comp. Neurol., 146 (1972) 1-14. [14] Moore, R.Y., Retinohypothalamic projection in mammals: a comparative study, Brain Res., 51 (1973) 403-409. [15] Moore, R.Y., The organization of the human circadian timing system, Prog. Brain Res., 93 (1992) 101-115. [16] Moore, R.Y. and Card, J.P., The intergeniculate leaflet: an anatomically and functionally distinct subdivision of the lateral geniculate complex, J. Comp. Neurol., 344 (1994) 403-430. [17] Moore, R.Y., Organization of the primate circadian system, J. BioL Rhythms, 8 (1993) $3-$9.

253

[18] Morin, L.P., Blanchard, J. and Moore, R.Y., lntergeniculate leaflet and suprachiasmatic nucleus organization and connections in the golden hamster, Vis. Neurosci., 8 (1992) 219-230. [19] Otori, Y., Tominaga, K., Fukuhara, C., Yang, C., Yamazaki, S., Cagampang, F.R.A., Okamura, H. and Inouye, S.-I.T., Substance P-like immunoreactivity in the suprachiasmatic nucleus of the rat, Brain Res., 619 (19931 271-277. [20] Pickard, G.E., Bifurcating axons of retinal ganglion cells terminate in the hypothalamic suprachiasmatic nucleus and the inlergeniculate leaflet of the thalamus, Neurosci. Lett., 55 (1985) 211-217. [21] Sadun, A.A., Schaechter, J.D. and Smith, E.E.N., A retinohypothalamic pathway in man: light mediation of circadian rhythms, Brain Res., 302 (1984) 37-377. [22] Speh, J.C. and Moore, R.Y., Retinal afferents as demonstrated with unconjugated cholera toxin, Soc. Neurosci. Ahstr., 22 11992) 876. [23] Takatsuji, K. and Tohyama, M., The organization of the rat lateral geniculate body by immunohistochemical analysis of neuroactive substances, Brain Res., 48 (19891198 209. [24] Takatsuji, K., Miguel, Hidalgo, J.J.M. and Tohyama, M., Substance P-immunoreactive innervation from the retina I(7 the suprachiasmatic nucleus in the rat, Brain Res., 568 (19911 223 229. [25] Takatsuji, K. and Tohyama, M., The development and innervation of two neuropeptides (substance P and neuropeptide Y) immunoreactive fibers in rat suprachiasmatic nucleus. In H. Nakagawa, Y. Oomura and K. Nagai (Eds.), New Functional

Aspects" of the Suprachiasmatic Nucleus ~[" the Hypothalamus, John Libbey, London, 1993, pp. 45-52.