0361-9230/90$3.00 + .OO

Brain ResearchEulkdn, Vol. 24, pp. 593-597. e PergamonPress plc, 1990.Printedin the U.S.A.

Lesions Dorsal to the S~prachiasmatic Nuclei Abolish Split Activity Rhythms of Hamsters M. E. HARRINGTON,” G. A. ESKES, P. DICKSON AND B. RUSAK Department of Psychology, Dalhousie University, Halifax, NS B3H 4Jl Canada and *Department of Psychology, Smith College, Northampton, MA 01063

Received 27 November 1989 HARRINGTON, M. E., G. A, ESKES, P. DICKSON AND B. RUSAK. Lesions dorsal to the suprachiasmatic nuclei abolish split activiry ~kyfk~ o~ka~s?e~~. BRAIN RBS BULL 2414) 593-597, 1990.-Constant light exposure fLL) can result in “splitting” of circadian rhythms into two components coupled about 12 hr apart. Splitting has been interpreted as evidence for the presence of two main oscillators or groups of oscillators underlying circadian rhythms. Abolition of splitting after unilateml suprachiasmatic nucleus ablation suggested that each suprachiasmatic nucleus could correspondto one of these componentoscillators. We examined whether lesions outside the suprachiasmatic nuclei (SCN) would abolish split activity rhythms of hamsters in LL. Wheel-running activity was recorded for 3 months after surgery. Tissue damage was assessed by Kliiver-Barrera staining. Damage to areas dorso-caudal to the SCN was able to abolish the split condition. Bilateral damage to the anterior SCN, partial unilateral SCN ablation or unilateral periventricular damage also abolished the split pattern. These results indicate that destruction of one SCN is not essential for the eli~nation of split rhythms since lesions dorsal to the SCN or partiat bilateral SCN damage are also effective. Indirect lesion effects on SCN function or damage to extra-SCN oscillators may account for the loss of the split condition. Circadian rhythms Constant light Suprachiasmatic nuclei

Paraventricular nucleus

Periventricular nucleus

Splitting

CONSTANT light exposure (LL) in Syrian hamsters (16) and rats

to the SCN abolished the split condition in two hamsters (15). In

(3) can result in “splitting” of circadian rhythms into two components coupled approximately 12 hours apart. These two components are often observed to express different circadian periods before they establish a stable phase relation; this observation was critical to the development of a dual-oscillator model for the rodent circadian pacemaker (16). The suprachiasmatic nuclei (SCN) are recognized as the do~n~t light~n~nable pacemaker in mammals (17,20). The fact that this structure is bilaterally symmetric has led to the hypothesis that each suprachiasmatic nucleus contains one of the two putative oscillators comprising the pacemaker (15). Pickard and Turek (15) demonstrated that splitting of the activity rhythm was abolished by unilateral SCN lesions and suggested that each suprac~~matic nucleus may be responsible for generating one of the two activity components expressed during splitting. In another study, dissociation of activity rhythms similar to splitting occurred in LL following unilateral SCN ablation (4). These results indicated that a hamster with only one suprachiasmatic nucleus is capable of generating rhythms with several components. However, the pattern of activity splitting observed in these animals was not identical to the typical splitting pattern seen in intact animals. This difference leaves open the possibility that while one suprachiasmatic nucleus can generate multiple activity components, both SCN are necessary for a stable pattern of two distinct components coupled circa-12 hr apart. There is preliminary evidence that the abolition of split rhythms after unitateral SCN lesions might be attributable to nonspecific SCN or SCN afferent or efferent damage. Lesions directly caudal

one rat, a lesion damaging the optic chiasm rostral to the SCN abolished the split pattern of feeding and drinking rhythms (3). Ablation of the geniculo-hypothalamic tract caused fusion of split activity rhythms in four of seven hamsters (9). The purpose of the present study was to investigate further whether splitting could be abolished by extra-SCN damage by examining the effects of h~~~arnic lesions in and near the SCN on the stability of activity rhythm splitting in Syrian hamsters. METHOD

Adult male golden hamsters (Lak:LVG; Charles River Laboratories) were taken from a colony room kept under a 1ight:dark cycle (lights on 0600 hr to 2ooO hr) and were housed in individual plastic cages (46 X 25 X 20 cm) equipped with running wheels (17 cm in diameter) under continuous light (50 iux at cage level). Cages were cleaned and food and water replenished approximately every seven to ten days at various times during the day. For some hamsters, wheel rotations activated a pen on an Esterline Angus event recorder. Each hamster’s wheel-running record was then cut into segments representing 24 hr and each day’s record was pasted below the previous day’s. For other hamsters, wheel rotations were recorded by an IBM Model 30 computer and Dataquest software (Data Sciences, Roseville, MN). Actograms were produced by fit performing a log conversion of the data and then plotting each day’s data on a 24-hr scale with 15 quantitization levels on the Y-axis. Animals were housed under LL until they developed an activity rhythm split into two components in approximate antiphase

593

594

HARRINGTON.

17

24

ESKES. DICKSON AND RUSAK

TABLE 1 MEASURES OF THE PERIOD OF THE FREE-RUNNING RHYTHM OF HAMSTERS ACTIVITY RHYTHMS BEFORE AND AFI-ER FUSION OF SPLIT RHYTHMS

Animal

B14 R94 RI8 B17 RIO R106 R89 B213 Bl5

FIG. 1. Double-plotted wheel-running activity record of a hamster (Bl7) with lesion damage to the medial paraventricular hypothalamic nuclei immediately dorsal to the suprachiasmatic nuclei (see Fig. 3B). Shortly the day of surgery (indicated by an arrow on the left hand side of the record) the split pattern disappeared and a unimodal rhythm was observed. This animal showed an unusually short duration activity time and a rhythm with a short period after the lesion.

Prelesion Period

Postlesion Period

23.61 24.04 23.94 24.01 24.06 24.15 24.29 24.11 24.16

23.72 23.72 23.76 23.80 24.00, 24.01, 24.23, 24.25, 24.40

then then then then

24.13 24.6, then 24.00 24.04, then 24.18 24.475

Several measures of the postlesion period are included in four cases. These represent relatively steady-state periods shown at various intervals after the lesion, between which periods changed gradually over an interval of several weeks.

following

(circa-12 hr apart; mean * SE = 75 + 5 days). After their activity had shown a stable split pattern (83 2 9 days), hamsters received lesions aimed at the periventricular zone dorsal to the SCN. Surgery was performed under sodium pentobarbital anesthesia (80 mgikg). Hamsters were kept in their cages under LL until fully anesthetized. They were removed to an adjacent room for surgery. Hamsters were positioned in a Kopf stereotaxic instrument with the incisor bar 2.0 mm below the interaural line. Radiofrequency lesions were performed using a Grass LM4 lesion maker. Current was passed for 15 seconds through an electrode constructed from a stainless steel insect pin insulated except for 0.5 mm at the tip. The coordinates used were: 0 to 0.7 mm anterior to bregma, 0.2 to 0.3 mm lateral to the midline, and 7.5 to 8.2 mm ventral from dura. Sham-operated hamsters (n=3) were treated exactly the same as hamsters receiving lesions except that no current was passed. At the end of surgery, the hamster was returned to its cage while still anesthetized. Animals were housed under LL for 90 ? 5 days following surgery. A total of 40 operated hamsters were studied. At the end of the experiment hamsters were given a lethal dose of sodium pentobarbital and were perfused with an intracardiac rinse of 0.9% saline, followed by 10% formalin. Tissue was sectioned through the hypothalamus at 40 microns and stained with the Khiver-Barrera technique for cells and fibers (11). Stained sections were projected on standard drawings of the paraventricular and suprachiasmatic area to assess the extent of lesion damage; assessments were made by a single observer without knowledge of the behavior of the animals.

which is atypical for hamsters housed in LL (see Table 1). These four hamsters did not appear to differ from the other five in duration of activity time. Splitting was eliminated in hamsters with extensive lesion damage to the medial portion of the paraventricular hypothalamic nuclei (PVN) immediately dorsal to the SCN, but with little or no SCN damage (n=4; see Figs. 1, 3 and 5). Splitting was also eliminated in hamsters with complete unilateral SCN damage and some PVN damage (n= l), partial unilateral SCN damage (n = l), bilateral damage to the periventricular zone and anterior SCN (n = 1)) or unilateral damage in the periventricular zone (n = 2, see Fig. 4). All hamsters with some SCN damage (see Figs. 3B, D, 4A, C and D) showed postlesion rhythms with shorter periods than prelesion rhythms. Two hamsters, both with largely unilateral extra-SCN damage (see Fig. 4B and E), showed

24

RESULTS Lesions

in nine hamsters

rapidly

abolished

the split condition

(see

Figs. 1,2 and 5). One month after rhythm fusion, the duration of the activity time of eight of these nine hamsters appeared normal for hamsters in the unsplit condition (median= 10.5 hr; range = 7-12 hr). The hamster whose record is shown in Fig. 1 had an unusually short duration of activity (approximately 2 hr). After restoration of a unimodal activity pattern, four hamsters expressed free-running activity rhythms with periods shorter than 24 hr,

FIG. 2. Double-plotted wheel-mnmng activity record of a hamster (B15) with unilateral lesion damage in the periventricular zone (see Fig. 4E). A rhythm with a longer period (approximately 24.4 hr) was observed after fusion of the split rhythm. Day of surgery is indicated by the arrow.

EXTItA-SCN LESIONS ABOLISH SPLIT RHYTHMS

E[G. 3. Representative coronal brain sections, arranged from anterior to posterior (top to bouom), showing histological results for four hamsters demonstrating fused rhythms after surgery. The SCN and the PVN are outlined with dashed lines. Dark stippling indicates areas of complete tissue destruction while ligher stippling indicates areas with heavy gliosis. These four hamsters all showed damage to the periventticular zone and the PVN dorso-caudal to the SCN. Animals: Bi4 (A), B17 (B), RIO (C) and R89 (D), FIG. 5. Double-plotted running-wheel activity record for a hamster (R89) with lesion damage to the medial paraventricular hypothalamic nuclei (see Fig. 3D). The activity rhythm of this hamster appeared to be unimodal immediately after surgery (see arrow). A second component appeared after several weeks, eventually fusing with the initial component.

FfG. 4. Histological results for five hamsters demons~ating fused postsurgery rhythms. See Fig. 3 for further exphmation. Animals: R94 (A), B213 (B), R18 (C), R106 (D), and B15 (E).

FIG. 6. Histological results for seven of the fiiteen hamsters that maintained the split condition throughout the experiment. All of these animals had tissue damage to the ~ven~cu1~ zone and PVN. See Fig. 3 for further expknration. Animals: B18 (A), B207 (B), Rl20 (C), R102 (D), B13 (E), B2 (F) and B8 (G).

HARRINGTON.

596

FIG. 7. Representative brain sections showing lesion damage of the three

hamsters with partial SCN lesions showing apparently arrhythmic runningwheel activity after surgery. See Fig. 3 for further explanation. Animals: R119 (A), RlOO (B) and R116 (C).

postlesion rhythms with longer periods than prelesion rhythms (see Table 1 and Fig. 2). One hamster (see Fig. 5) expressed an app~ntly unim~al rhythm with an unusually short duration activity time during the first two weeks after surgery. At that time a second component seemed to gain expression and gradually fused with the first component. Simultaneously, the period of the rhythm changed and the duration of activity time increased. In 1.5 hamsters, the split condition was maintained for the duration of the observation period after surgery. Seven of these hamsters had lesion damage to the periventricular zone and PVN that did not approach the dorso-caudal border of the SCN (see Fig. 6). Splitting was also maintained in hamsters with partial bilateral anterior SCN damage (n = I), hypothalamic damage outside the PVN and periventricular zone (n=4), and in hamsters with sham lesions (n = 3). In eleven other hamsters the responses to lesions were ambiguous. The loss of the split condition was very gradual in three hamsters. Of these three animals, one had partial unilateral SCN damage, another had bilateral SCN damage sparing the dorsocaudal SCN, and the third had largely unilateral damage to the medial PVN. Four hamsters developed unimodal rhythms but showed a rapid return to the split condition; one had bilateral damage to the periventricular area with unilateral damage of the caudal SCN, another hamster had damage in the periventricular area dorsal and anterior to the SCN, the third hamster had a bilateral medial PVN lesion, and the fourth showed unilateral damage to the dorso-caudal SCN. For five hamsters the behavioral response to the lesion was unclear. Four hamsters became apparentIy physic after surgery. One of these hamsters had a complete SCN lesion, while the remaining three hamsters had partial SCN damage (see Fig. 7). DISCUSSION

These results demonstrate that unilateral SCN damage is not required to abolish splitting. Symmetric damage to regions dorsocaudal to the SCN with little or no damage to the SCN can also restore unimodal rhythms, Lesions that appeared to differ little in location were differentially effective in abolishing the split condition (see Figs. 3 and 6), suggesting that the loss of splitting does not depend on destruction of a unique neural substrate. These data are open to several interpretations. The behavioral effects might be due to direct or indirect stimulation of SCN cells. If this hypothesis were correct, it would be expected that electrical

ESKES. DICKSON

AND RUSAK

stimulation of the SCN or of SCN afferents might alter split rhythms. The effect of such stimulation on split rhythms might be expected to vary with the phase of the rhythm at which the stimulation was applied [cf. (19)]. The phase of the rhythm at which surgery was performed was not controlled in this study. Effective lesions may alter intrinsic SCN organization whether or not the SCN was included in the area of primary tissue damage. Since both anterograde degeneration and retrograde degeneration are possible consequences of brain lesion damage (22) even lesions in extra-SCN areas may alter neural organization of the SCN. Behavioral changes could also be attributable to the loss of SCN efferents or afferents. This hy~thesis could be tested by studying animals with SCN-area lesions produced by specific neurotoxins. SCN efferent fibers coursing dorsally through the subparaventricular area to the paraventricular nucleus of the thalamus and to more caudal hypothalamic nuclei (27) were probably damaged by many lesions. Lesions may have damaged retino-hypothalamic tract terminals which, although densely clustered in the SCN, also spread into the surrounding hypothalamic nuclei (10). The geniculo-hypothalamic tract, originating from cells in the intergeniculate leaflet and ventral lateral geniculate nucleus (8, 14, 24) may have also been damaged by the hypothalamic lesions. Geniculo-hypothaIamic tract damage can cause fusion of split rhythms (9). Damage to either the retino-hypothalamic tract or the geniculo-hypothalamic tract mav have altered photic input to the SCN. A reduction in perceived’light intensity would be exnected to alter split rhythms (3,7) ,. If the effective lesions abolished the split condition by suppressing the expression of one component in a manner similar to “masking” (1). it might be expected that the unsuppressed component would show no change in either free-running period or activity time duration. This was not observed in any animals; therefore, this explanation is not plausible. If instead these lesions could be thought of as removing one oscillator in a dual-oscillator system such as that modeled by Pittendrigh and Daan (16), one might expect to observe a change in the rhythm’s period, with the remaining oscillator expressing its intrinsic period. If each of the two oscillators have different intrinsic periods, one would expect the fused rhythms of a group of animals to have a bimodal dis~ibution of periods. There was a weak suggestion of such a distribution in our data. Four of nine hamsters showed fused rhythms with periods between 23.7 and 23.8. The remaining 5 hamsters had fused rhythms with initial periods between 24.0 and 24.4. It is likely that a dual oscillator model is too simplistic and that the circadian system consists of multiple oscillators, both inside and outside the SCN (2, 5, 13, 18, 21, 23, 25). In this view, the split condition can be thought of as a highly probable arrangement under LL of a population of oscillators, many, but not all, of which are in the SCN. in this case, it would be very unlikely that a specific anatomical locus could be identified as the substrate for the oscillators underlying each component of a split rhythm. It has previously been reported that there is a strong positive correlation between the volume of undamaged SCN tissue and the activity rhythm period length after partial SCN lesions (4). In our sample a decrease in the rhythm’s period after fusion was strongly associated with SCN damage. A decrease in period length might be due to the loss of some oscillators in a multiple coupled oscillator system or to altered coupling between intact oscillators. Split rhythms only develop in LL in nocturnal animals. The appearance of split rhythms may be related to the lengthening of the free-running period usually observed in LL [Pickard, personal communication; but see (12)]. A decrease in period may alter the circadian system of the hamster so that it is no longer likely to show split rhythms. Unilateral SCN lesions which caused fusion of

EXTRA-SCN

597

LESIONS ABOLISH SPLIT RHYTHMS

split rhythms also caused decreased rhythm periods in 8 of 11 hamsters (15); two hamsters showed no change and one hamster showed an increased period. In the present study, 6 hamsters showed a decrease in period, while 3 showed an increase in period after fusion of split rhythms. From these results it appears that a decrease in period is not necessary for fusion of split rhythms. Three hamsters showed apparently arrhythmic running-wheel activity after partial SCN lesions (see Fig.7). Since arrhythmicity was determined by visual inspection of the activity records, residual low-amplitude circadian rhythms might have been detected using other analytic tools (6). Arrhythmicity after partial SCN damage has been reported previously [see (26)]. One animal that showed severely disrupted activity patterns under these conditions sustained very little SCN damage, which was confined to anterior portions of the nuclei (see Fig. 7). It is likely that long-term housing under LL contributes to the disruption of activity patterns after partial damage to the SCN.

In summary, this study has shown that unilateral SCN damage is not necessary for the abolition of the split condition. There does not appear to be a unique site in the hamster brain destruction of which causes loss of the split condition. Further studies using techniques that allow more specific and localized alterations of neural function than lesions cause may help clarify the neural mechanisms involved in generating and maintaining the split activity pattern. ACKNOWLEDGEMENTS

We are very grateful to D. Goguen, M. Jain, M. Prince, T. Rahmani, D. Burger, M. Pufall and J. Needles for their skilled technical assistance and to J. Meijer for helpful discussions. This research was supported by grants from Medical Research Council of Canada (MA8929), Natural Science and Engineering Research Council of Canada (AO305), Daihousie University RDFS, Smith College Committee on Faculty Compensation and Development, and National Institutes of Health (NS26496).

REFERENCES 1. Aschoff, J. Exogenous and endogenous components in circadian rhythms. In: Cold Spring Harbor Symposia on Quantitative Biology, vol. 25. Baltimore: Waverly; 1960:11-27. 2. Boulos, Z.; Morin, L. P. Entrainment of split circadian activity rhythms in hamsters. J. Biol. Rhythm. l:l-15; 1985. 3. Boulos, Z.; Terman, M. Splitting of circadian rhythms in the rat. J. Comp. Physiol. [A] 134:75-83; 1979. 4. Davis, F. C.; Gorski, R. A. Unilateral lesions of the hamster suprachiasmatic nuclei: Evidence for redundant control of circadian rhythms. J. Comp. Physiol. [A] 154:221-232; 1984. 5. Davis, F. C.; Menaker, M. Hamsters through time’s window: Temporal structure of hamster locomotor rhythmicity. Am. J. Physiol. 239:R149-R155; 1980. 6. Dowse, H. B.; Hall, J. C.; Ringo, J. M. Circadian and ultradian rhythms in period mutants of Drosophila melanogaster. Behav. Genet. 17:19-35; 1987. 7. Earnest, D. J.; Turek, F. W. Splitting of the circadian rhythm of activity in hamsters: Effects of exposure to constant darkness and subsequent m-exposure to constant light. J. Comp. Physiol. [A] 145:405+11; 1982. 8. Harrington, M. E.; Nance, D. M.; Rusak, B. Double-labeling of neuropeptide Y-immunoreactive neurons which project from the geniculate to the suprachiasmatic nuclei. Brain Res. 410:275-282; 1987. 9. Harrington, M. E.; Rusak, B. Ablation of the geniculo-hypothalamic tract alters circadian activity rhythms of hamsters housed under constant light. Physiol. Behav. 42: 183-189; 1988. 10. Johnson, R. F.; Moore, R. Y.; Morin, L. P. Loss of entrainment and anatomical plasticity after lesions of the hamster retinohypothalamic tract. Brain Res. 460:297-313; 1988. 11. Kltiver, H.; Barrera, E. A method for the combined staining of cells and fibers in the nervous system. J. Neuropathol. Exp. Neurol. 12:400X)3; 1953. 12. Morin, L. P. Age-related changes in hamster circadian period, entrainment, and rhythm splitting. J. Biol. Rhythm. 3:237-248; 1988. 13. Morin, L. P.; Cummings, L. A. Splitting of wheelrunning rhythms by castrated or steroid treated male and female hamsters. Physiol. Behav. 29665-675; 1982. 14. Pickard, G. E. The afferent connections of the suprachiasmatic

15. 16.

17.

18.

19. 20. 21.

22.

23.

24.

25. 26.

27.

nucleus of the golden hamster with emphasis on the retinohypothalamic projection. J. Comp. Neurol. 211:65-83; 1982. Pickard, G. E.; Turek, F. W. The suprachiasmatic nuclei: Two circadian clocks? Brain Res. 268:201-210; 1983. Pittendrigh, C. S.; Daan, S. A functional analysis of circadian pacemakers in nocturnal rodents V. Pacemaker structure: A clock for all seasons. J. Comp. Physiol. 106:333-355; 1976. Rosenwasser, A. M. Behavioral neurobiology of circadian pacemakers: A comparative perspective. Prog. Psychobiol. Physiol. Psychol. 13:155-226; 1988. Rusak, B. The role of the suprachiasmatic nuclei in the generation of circadian rhythms in the golden hamster, Mesocricerus aurafus. J. Comp. Physiol [A] 118:145-164; 1977. Rusak, B.; Groos, G. Suprachiasmatic stimulation phase shifts rodent circadian rhythms. Science 215:1407-1409; 1982. Rusak, B.; Zucker, I. Neural regulation of circadian rhythms. Physiol. Rev. 59449-526; 1979. Satinoff, E.; Presser, R. A. Suprachiasmatic nuclear lesions eliminate circadian rhythms of drinking and activity, but not of body temperature, in male rata. J. Biol. Rhythm. 3:1-22; 1988. Schoenfeld, T. A.; Hamilton, L. W. Secondary brain changes following lesions: A new paradigm for lesion experimentation. Physiol. Behav. 18:951-967; 1977. Stephan, F. K. Circadian rhythm dissociation induced by periodic feeding in rats with suprachiasmatic lesions. Behav. Brain Res. 7:81-98; 1983. Swanson, L. W.; Cowan, W. M.; Jones, E. G. An autoradiographic study of the efferent connections of the ventral lateral geniculate nucleus in the albino rat and the cat. J. Comp. Neural. 156:143-164; 1974. Terman, M.; Terman, J. A circadian pacemaker for visual sensitivity? Ann. NY Acad. Sci. 453:147-161; 1985. van den Pol, A. N.; Powley, T. A fine-grained anatomical analysis of the role of the rat suprachiasmatic nucleus in circadian rhythms of feeding and drinking. Brain Res. 160:307-326; 1979. Watts, A. G.; Swanson, L. W.; Sanchez-Watts, G. Efferent projections of the suprachiasmatic nucleus: I. Studies using anterograde transport of Phaseolus vulgaris leucoagglutinin in the rat. J. Comp. Neurol. 258:204-229; 1987.