Brain Research, 302 (1984) 371-377 371 Elsevier

BRE 10073

A Retinohypothalamic Pathway in Man: Light Mediation of Circadian Rhythms

ALFREDO A. SADUN ~, JUDITH D. SCHAECHTER 1 and LOIS E. H. SMITH 2

lNeuro-Ophthalmology Unit, Howe Laboratory, Massachusetts Eye and Ear Infirmary, and 2Department of Ophthalmology, Children's Hospital Harvard Medical School Boston, MA (U.S.A.)

(Accepted November 1st, 1983)

Key words: suprachiasmatic nucleus - - retinohypothalamic pathway - - human hypothalamus - - circadian rhythms - - diurnal cycles - - paraphenylenediamine

It has been proposed that, in animals, a retinohypothalamic pathway exists which mediates the synchronization of the diurnal light- dark cycle with the central neural components regulating endogenous rhythms. There have been numerous anatomic, physiologic and behavioral investigations to substantiate this proposed connection in experimental animals.

Morphologic investigation of a retinohypothalamic tract in man has awaited the development of a technique capable of axonal trac- ing in the human brain. The paraphenylenediamine method was applied to 7 post-mortem human brains. Degenerated axons were found in the suprachiasmatic nuclei of the hypothalamus in each of the 4 patients who had incurred prior optic nerve damage. The ret- inosuprachiasmatic pathway may be the anatomical substrate for the integration of retinal light information with endogenous rhythms in man.

INTRODUCTION

In 1862 Wagner proposed that nerve fibers deriv-

ing from the optic chiasm entered the hypothalamus

in man31. Since then, the existence of a retinohypo-

thalamic connection has been a controversial issue.

Support for this pathway was gained when investiga-

tors found changes in circadian rhythms in experi- mental animals following certain lesions of the visual

pathways. Bilateral transection of the optic nerve in

the rat resulted in desynchronization between endog- enous circadian rhythms and the diurnal l ight-dark cycle21, 34. Yet, more distal destruction of the retinal

ganglion projections, by transection of the primary

or accessory optic tracts, did not alter circadian

rhythms23, 24. It was, therefore, hypothesized that there exists a retinofugal pathway terminating in the

region of the optic chiasm which is involved in the

regulation of endogenous rhythms in animals.

Evidence of a retinohypothalamic pathway has also been provided by earlier morphological studies in a large number of animal species, including those of Herrick 16 in amphibia and by Jacobs and Mor- gane 19 in cetacea. However , Armstrong3 failed to

confirm the pathway in reptiles. Subsequent studies employing silver impregnation

methods for identifying degenerated axons have

done little to settle the controversy. B1/.imcke5 and Rieke 27 used the ammoniacal silver methods of Biels-

chowsky and Glees to demonstrate optic fiber termi-

nations in the periventricular system of the third ven-

tricle, supraoptic nucleus and tuberal nuclei in the cat and rat, respectively. Polyak 26 utilized the Marchi

method, following a lesion of the retinal ganglion

cells, and noted the existence of a direct retinohypo- thalamic connection in rat and guinea-pig, but not in

cat and pigeon. Some investigators, using the Nauta

method, failed to demonstrate the retinohypotha- lamic pathway in the cat, pigeon and monkey6,11. 32.

However, Sousa-Pinto and Castro-Correia 33 used the

Fink-Heimer modification of the Nauta method to

demonstrate a retinohypothalamic pathway in the rat. In a comprehensive study, Kiernan20 employed a

variety of degeneration tracing methods in 5 mam- mals, including rat, rabbit and an amphibian, yet he failed to demonstrate optic fibers terminating in the hypothalamus of these animals.

The development of the more contemporary trac-

Correspondence: A. A. Sadun, Massachusetts Eye and Ear Infirmary, 243 Charles Street, Boston, MA 02114, U.S.A.

0006-8993/84/$03.00 © 1984 Elsevier Science Publishers B.V.

372

ing method of autoradiography allowed for a direct

retinohypothalamic pathway to be firmly established in the rat 25. Further autoradiographic studies have

identified this retinal pathway to the suprachiasmatic

nucleus of the medial hypothalamus in several mam- mals, including cat, tree shrew and monkey 22.

Among all the mammals studied to date, the retino-

hypothalamic projection has maintained some con-

sistent features. Each suprachiasmatic nucleus has

been shown to receive a bilateral retinal projection,

with predominantly contralateral innervation. Addi-

tionally, the ventral and lateral portions of the nuclei

are most densely innervated. Evidence of the retino-

hypothalamic pathway in the many mammals studied

to date suggests the possibility of its existence in man

as well. Documentation of the retino-suprachiasmatic

pathway in man was not feasible until a technique ca- pable of tracing neural pathways in the human brain

was found. Fortunately, the recently developed para-

phenylenediamine (PPD) method reliably stains de- generated axons and axon preterminals in the visual

system in human autospy brain tissue 29.3°. Thus, a di-

rect investigation of human visual neuroanatomy is

now possible.

MATERIALS AND METHODS

Autopsy brain specimens were obtained from 7 pa- tients, 4 of whom had documented optic nerve dam-

age prior to death. A 72-year-old woman had an an-

eurysm of the left carotid artery which had com-

pressed the left optic nerve. She was followed for 4 years during which progressive visual field defects

were documented, with subsequent loss of visual acu-

ity and optic atrophy. Further tissue specimens were

obtained from a 62-year-old woman who developed a

melanoma in the right eye, which was enucleated 7 years prior to her death. Brain tissue was also ob- tained from a 45-year-old male with marked loss of vision and documented bilateral optic atrophy from retinitis pigmentosa diagnosed 10 years prior to death. Similarly, tissue was received from a 63-year- old male blinded and with bilateral atrophy due to a central artery occlusion which occurred 7 years prior to death. Three control autospy brains without known visual or neurological damage were also ex-

amined.

The brain tissue was processed by the PPD meth- od 30. This is a modification of a technique which em-

ploys PPD as a stain for degenerating axonsS and

axon preterminals17, 28. The PPD method is adapted

for use in human brain tissue obtained at autopsy. It

reliably stains degenerated axons even after long sur- vival periods in man 30.

The brain specimens, obtained at autospy 4-24 h after death, are routinely fixed in 10% formalin for

1-20 days. The brains are cut into blocks with the

smallest width less than 0.5 cm thick. These brain

specimens are then placed into a fixative solution of 2% glutaraldehyde, 2% paraformaldehyde, 2.5% di-

methylsulfoxide (DMSO) in 0.1 M sodium cacodyl-

ate buffer (pH 7.4). The brain blocks are stored in

fixative at 4 °C for periods ranging from a few days to

several weeks. These blocks are then further sec-

tioned to slices of tissue approximately 0.5 mm by 4

mm by 6 mm. These tissue slices are stored in fixative in the cold for another 2-4 days.

These small tissue blocks are repeatedly rinsed in

0.1 M sodium cacodylate and then placed overnight

in 0.5% osmium tetroxide in a 0.1 M sodium cacodyl-

ate solution. After several rinses in distilled water,

the tissue is dehydrated through a graded series of al- cohols and propylene oxides before infiltration with

Embed 812 (previously Epon 812). After polymeri-

zation at 60 °C for 2 days, the blocks are trimmed of

excess plastic and semi-thin sections 1-2 ~m thick are

cut on the ultramicrotome (Sorvall MT2B). The sec- tions are placed on distilled water drops, dried at ap-

proximately 90 °C, kept warm overnight at 50 °C and

cooled before staining. The slides are stained by im- mersion in a solution of 1% PPD in methanol for 5-10

min and then rinsed and washed in 95% ethanol for

2-5 min. The slides are dried on a warm plate and al- lowed to cool before mounting. The staining intensity

does not vary with PPD concentration or with the du- ration of immersion in the PPD solution. Rather, the darkness of the stain can be varied only by the thick- ness of the section. Ethanol does not decolorize PPD, so the rinse time is not critical.

The sections are examined with a Zeiss 16-stand- ard light microscope. The degenerated axons are vis- ible with a 'high dry' objective (40 ×) or an oil immer- sion objective (63 ×).

373

Fig. 1A: cross-section of normal, right optic nerve shows fascicles of retinal ganglion cell axons separated by septae (s). The PPD method stains the myelin sheaths of intact axons, producing dark rings. (PPD; 1400 x). B: cross-section of lesioned, left optic nerve from a patient (same as in A) who incurred complete unilateral transection of this optic nerve one year prior to death. The degener- ated axons (arrows) are seen as dark profiles. Normal axons are not seen. (PPD; 1400 x).

374

RESULTS

Sections through the optic nerves, optic chiasms, optic tracts and the suprachiasmatic nuclei (SCN) of the hypothalamus were stained with PPD. Our crite- rion for degenerated axons is the appearance of dark brown circular profiles. Preterminal degeneration is thought to be represented by smaller, homogeneous- ly pale brown circular profiles adjacent to neuron so- mata or large dendrites.

In examination of sections through intact optic nerves, we noted many dark rings. These PPD- stained myelin sheaths were seen encircling the unstained cytoplasm of normal axons (Fig. 1A). Few, if any, degenerated profiles were seen in these undamaged optic nerves. In contradistinction, many degenerated axons were observed in the optic nerves from patients who had a lesion involving the retinal ganglion cells or the retinal ganglion cell axons (Fig. 1B).

Application of the PPD method to the decussating fibers of the optic chiasms from the patients with uni- lateral optic nerve damage show degenerated axons interspersed among normal axons. Similarly, both degenerated and normal axons were observed inter- mingled within the optic tracts from patients with uni- lateral optic nerve damage.

The SCN of the hypothalamus were cut in multiple coronal planes, and processed by the PPD method. Degeneration was observed bilaterally throughout the SCN from those 4 patients who sustained either unilateral or bilateral optic nerve damage prior to death. The degenerated axon profiles found in the SCN were somewhat smaller than those noted in the optic nerves (Fig. 3). Degenerated axon pretermi- nals were noted, at high magnification, adjacent to the neuron somata and dendrites of the SCN (Fig. 4). Very few, if any, degenerated axons and no degener- ated axon preterminals were found in the SCN from the normal (control) patients (Fig. 2).

DISCUSSION

The present morphological investigation provides evidence for the existence of a retino-suprachiasma- tic pathway in man. This study employed a recently developed method utilizing paraphenylenediamine (PPD), an agent which stains by a different principle

than does silver in standard impregnation methods. Reduced-silver methods stain the neurofibrillar

components of nerve cells, allowing for the evalu- ation of fibrillar changes that occur during degenera- tion14. Optimal silver impregnation of neurofibrillar degeneration of axons is transient, and is typically applied 4-7 days after the lesion 12. Therefore, silver impregnation techniques, such as the Nauta and Fink-Heimer methods, are difficult to apply to hu- man tissue.

PPD, on the other hand, chelates osmium which precipitates in lipids. Thus, PPD marks lipid el- ements which accumulate in degenerating axons. It probably also stains the lipid debris from pre-existing axons, found much later in the course of degenera- tion, and which we presently term degenerated ax- ons.

While the exact biochemical and morphological course of neuronal degeneration remains unknown, the conventional notion has been that products of de- generation are not seen in brain tissue after long sur- vival periods. However, there is an increasing body of evidence that remnants of degeneration can be found in brain tissue of certain species, including man, years after injury4,13,3°. Evans 9 described in

bird and rabbit CNS the time course of the degenera- tion process, and noted a persistence of lipid com- pounds as part of the stained material.

We postulate that in certain species lipid remnants become ensheathed by oligodendrocytes and, in this state, may be preserved for years. It is possible that each PPD-stained dark profile represents the compil- ation of several degenerated axons within one sheath. This may explain the apparent large size of some PPD-stained degenerated profiles in compari- son to normal myelinated axons.

The PPD technique reliably stains both degenerat- ing and degenerated axons in animals 17 and man 4,3°. Comparison of the PPD technique with the Fink- Heimer method and electron microscopy has been made in cat 28. This showed the PPD method to be simpler and less capricious than the Fink-Heimer technique. Moreover, the PPD method was powerful enough to resolve degenerated axon terminals17,28, further demonstrating the sensitivity and reliability of this relatively unconventional staining technique.

Utilizing the PPD method, degeneration was noted bilaterally in the human suprachiasmatic nu-

375

Fig. 2. Section through the suprachiasmatic nucleus from a normal (control) patient. Note neuronal somata and absence of degener- ated profiles. Rings of myelin are seen ensheathing normal axons. (PPD; 600 x ). Fig. 3. Section through the suprachiasmatic nucleus from a patient with a damaged ipsilateral optic nerve. Note degenerated axons (arrows) throughout field. (PPD; 600 x). Fig. 4. Higher magnification of a section through the suprachiasmatic nucleus from a patient (same as in Fig. 3) with a damaged ipsilat- eral optic nerve. Note degenerated axons (large arrows) and degenerated preterminals (P) adjacent to neuronal somata and dendrites (PPD; 1000 ×).

376

cleus (SCN) of the hypothalamus. This suggests that

some retinal ganglion cell axons branch off at the lev-

el of the optic chiasm to project directly to both SCN,

which lie just dorsal to the chiasm. Other ret inofugal

fibers continue past to form the pr imary optic tracts.

A direct re t inohypothalamic connect ion has been

proposed in animals to media te the synchronizat ion

of the diurnal l ight -dark cycle with the central neural

components regulating circadian rhythms. This path-

way in animals is demons t ra ted by the neuroanatomi-

cal studies previously discussed22. It is also suppor ted

by numerous behavioral 7 and physiological 23.24 stud-

ies in various exper imental animals.

Environmenta l i l lumination has been demon-

strated to be a major regulator of circadian rhythms

in animals. Non-physiological light condit ions, such

as constant lightness or darkness, have been shown to

significantly alter the normal ly expressed circadian

rhythms. The per iod of hamster estrous 2, mela tonin

secretion in rats 1 and locomotor activity in spar-

rows 10 proved to be sensitive to cycles of environ-

mental lighting. Similarly, in the human, abnormali-

ties of the neuroendocr ine regulatory system have

been at t r ibuted to the exclusion of diurnal l igh t -dark

cycles due to blindness. For example , it has been

noted that blind women experience menarche at an

earl ier age than normal sighted women 15. It has also

been shown that blind persons have differences in

their water , carbohydra te and insulin balances TM as

compared to sighted persons. Thus, it is possible that

the absence of a retinal input to the hypothalamus in

a blind person is the pr imary defect resulting in the

disruption of endocrine rhythms.

It is likely that in man, as in o ther mammals , the di-

urnal l ight -dark cycle acts to entrain the endogenous

circadian rhythms. The ret ino-suprachiasmatic path-

way, identif ied by the PPD method, provides a neu-

roanatomical substrate for the integrat ion of retinal

light information with those hypothalamic neuroen-

docrine mechanisms which may regulate circadian

rhythms in man.

ACKNOWLEDGEMENTS

We thank Drs. T. Hedley-Whyte , D. Miller and E.

P. Richardson for their assistance and expert ise in

obtaining appropr ia te neuropathological specimens.

Suppor ted in part by a grant from the Nat ional Socie-

ty to Prevent Blindness ( A . A . S . ) and the H E E D

Foundat ion Fellowships (A .A .S . and L .E .H .S . ) .

Presented in part at the 1982 Annua l Meet ing of the

Associat ion for Research in Vision and Ophthalmol-

ogy, Sarasota, Florida.

REFERENCES

1 Adler, J., Lynch, H. J. and Wurtman, R. J., Effect of cyclic changes in environmental lighting and ambient tempera- ture on the daily rhythm in melatonin excretion by rats, Brain Research, 163 (1979) 111-120.

2 Alleva, J. J., Waleski, M. V. and Alleva, F. R., A biologic- al clock controlling the estrous cycle of the hamster, Endo- crinology, 88 (1971) 1308-1379.

3 Armstrong, J. A., An experimental study of the visual path- ways in a snake (Natrix natrix), J. Anat., 85 (1951) 275-288.

4 Beatty, R. M., Sadun, A. A., Smith, L.E.H., Vonsattel, J. P. and Richardson, E. P., Jr., Direct demonstration of transsynaptic degeneration in the human visual system: a comparison of retrograde and anterograde changes, J. Neu- rol. Neurosurg. Psychiat., 45 (1982) 143-146.

5 Blfimcke, S., Zur frage einer nervenfaserbindung zwischen retina and hypothalamus. I. Anatomische and experimen- telle Untersuchungen an Meerschweinchen und Katzen, Z. Zellforsch., 48 (1958) 261-282.

6 Cowan, W. M., Adamson, I. and Powell, T. P. S., An ex- perimental study of the avian visual system, J. Anat., 65 (1961) 545-563.

7 Critchlov, V., The role of light in the neuroendocrine sys- tem. In A. Nalbonov (Ed.), Advances in Neuroendocrino-

logy, Univ. of Illinois Press, Urbana, IL, 1963, pp. 377-401.

8 Estable-Puig, J. F., Bauer, W. C. and Blumberg, J. M., Paraphenylenediamine staining of osmium fixed plastic em- bedded tissue for light and phase microscopy, J. Neuropath. exp. Neurol., 24 (1965) 531-535.

9 Evans, D. H. L. and Hamlyn, L. H., A study of silver de- generation methods in the central nervous system, J. Anat., 90 (1956) 193-203.

10 Gaston, S. and Menaker, M., Pineal function: the biologic- al clock in the sparrow? Science, 160 (1968) 1125-1127.

11 Giolli, R. A., An experimental study of the accessory optic system in the Cynomolgus monkey, J. comp. Neurol., 121 (1963) 89-107.

12 Glees, P. and Nauta, W. J. H., A critical review of studies on axonal and terminal degeneration, Mschr. Psychiac Neurol., 129 (1955) 74-91.

13 Graf6, M. R. and Leonard, C. M., Successful silver impreg- nation of degenerating axons after long survivals in the hu- man brain, J. Neuropath. exp. Neurol., 39 (1980) 555-574.

14 Guillery, R. W., Light- and electron-microscopical studies of normal and degenerating axons. In W. J. H. Nauta and S. O. E. Ebbeson (Eds.), Contemporary Research in Neu- roanatomy, Springer-Verlag, New York, 1969, pp. 77-105.

15 Hasan, W., Chandra, P. and Prasad, J. N., Continuous illu-

mination and oogenesis in albino rabbits, Ind. J. Physiol. Pharmacol., 25 (1981) 279-284.

16 Herrick, C. J., The hypothalamus of Necturus, J. comp. Neurol., 59 (1934) 375-429.

17 Hollander, H. and Vaaland, J. L., A reliable staining meth- od for semi-thin sections in experimental neuroanatomy, Brain Research, 10 (1968) 120-126.

18 Hollwich, F., The influence of light via the eyes on animals and man, Ann. N. Y. Acad. Sci., 117 (1964) 105-131.

19 Jacobs, S. and Morgane, P. J., Retinohypothalamic con- nections in cetacea, Nature (Lond.), 203 (1964) 778-780.

20 Kiernan, J. A., On the probable absence of retino-hypotha- lamic connections in five mammals and an amphibian, J. comp. Neurol., 131 (1967) 405--408.

21 Klein, D. C. and Weller, J. L., Indole metabolism in the pi- neal gland: a circadian rhythm in N-acetyltransferase, Sci- ence, 169 (1970) 1093-1095.

22 Moore, R. Y., Retinohypothalamic projection in mam- mals: a comparative study, Brain Research, 49 (1973) 403-409.

23 Moore, R. Y. and Eichler, V. B., Loss of a circadian adren- al corticosterone rhythm following suprachiasmatic lesion in the rat, Brain Research, 42 (1972) 201-206.

24 Moore, R. Y. and Klein, D. C., Visual pathways and the central neural control of a circadian rhythm in pineal sero- tonin N-acetyltransferase activity, Brain Research, 71 (1974) 17-33.

377

25 Moore, R. Y. and Lenn, N. J., A retinohypothalamic pro- jection in the rat, J. comp. Neurol., 146 (1972) 1-14.

26 Polyak, S., The Vertebrate Visual System, Univ. of Chicago Press, Chicago, 1957, pp. 288-389.

27 Rieke, W. O., Optico-hypothalamic pathways in the rat, Anat. Rec., 130 (1958) 363-364.

28 Sadun, A. A., Differential distribution of cortical termi- nations in the cat red nucleus, Brain Research, 99 (1975) 145-151.

29 Sadun, A. A. and Smith, L. E. H., Retinal projections to the human hypothalamus. In Ass. Res. Vis. Ophthal., Ab- stracts, 1982.

30 Sadun, A. A., Smith, L. E. H. and Kenyon, K. R., Par- aphenylenediamine: new method for tracing human visual pathways, J. Neuropath. exp. Neurol., 42 (1983)200-206.

31 Scharrer, E., Photo-neuro-endocrine systems: general con- ceptsAnn. N.Y. Acad. Sci., 117 (1964) 13-22.

32 Singleton, M. C. and Peele, T. L., Distribution of optic fi- bers in the cat, J. comp. Neurol., 125 (1965) 303-328.

33 Sousa-Pinto, A. and Castro-Correia, J., Light microscopic observations on the possible retinohypothalamic projection in the rat, Exp. Brain Res., 11 (1970) 515-527.

34 Stephan, F. K. and Zucker, I., Circadian rhythms in drink- ing behavior and locomotor activity are eliminated by hypo- thalamic lesions, Proc. nat. Acad. Sci. U.S.A., 69 (1972) 1583-1586.