Proc. Natl. Acad. Sc,. USA Vol. 74, No. 3, pp. 1277-1281, March 1977 Physiological Sciences
Mutual entrainment of bilaterally distributed circadian pacemakers (cockroach/optic lobes/circadian rhythms)
TERRY L. PAGE*, PATRICIA C. CALDAROLAt, AND COLIN S. PITTENDRIGH* Department of Biological Sciences, Stanford University, Stanford, California 94305; and t Stanford University Medical School, Stanford, California 94305
Contributed by Colin S. Pittendrigh, December 6, 1976
ABSTRACT The interactions between the bilaterally distributed components of the circadian system that controls the locomotor activity rhythm of the cockroach Leucophaea maderae were investigated in a series of surgical lesion experiments. Complete excision of one optic lobe (either right or left) or its surgical isolation from the central nervous system had no-effect on the animal's ability to free-run in constant darkness nor was there any indication, as judged by postoperative r values, of any difference between left and right lobe pacemakers. However, these surgical procedures consistently resulted in a significant increase in r over preoperative values while optic nerve section had no effect on r. The proposition is developed that the left and right pacemakers in the two optic lobes are mutally coupled and that the compound pacemaker's period is shorter than either of its constituent pacemakers. It was also found that the integrity of either compound eye is sufficient to assure entrainment of both left and right pacemakers.
The fact that a single eukaryotic cell can be an autonomous circadian pacemaker (1, 2) sharpens several questions about the temporal (circadian) organization of multicellular organisms. Is the rhythmicity they express in a diversity of functions attributable to only one (or a few) cells which, as a pacemaker, impose rhythmicity on other tissues and organs to which they are coupled? Or is the organism to some extent a population of autonomous cellular pacemakers? If so, what mechanisms maintain temporal order among them (3)? Although it is now commonly assumed (4, 5) that a metazoan includes many such autonomous pacemakers, whose order depends on a mixture of mutual and hierarchical entrainment, that assumption is based on very little in the way of hard fact. Clearly, progress depends on having first localized discrete pacemakers in a single organism; only then can their possible interaction be experimentally studied. Circadian pacemakers have now been localized in the nervous system of several bilaterally symmetric animals. They occur in each eye of the mollusc Aplysia (6), in the suprachiasmatic nucleus of the mammalian hypothalamus (7-9), and in the optic lobes of several insects (10-15) and in the midbrain of another (16). The bilaterality of these animals makes several questions experimentally tractable: is there an active pacemaker on each side (left and right) of the same individual? Is there any hemispheric asymmetry in pacemaker behavior? How are they maintained in synchrony? In many natural situations that problem might be trivially solved by their submission to entrainment by a common external (light/dark) cycle. Jacklet (17) and Lickey et al. (18) have already suggested that the left- and right-hand (eye) pacemakers are, in fact, not mutually coupled in Aplysia. But there are obvious ecological situations where that would be intolerable if synchrony between redundant pacemakers is important. Our interest in the cockroach derives from (a) the fact that widely separated redundant pacemakers have been localized in this insect and (b) there is preliminary *
Present address: Hopkins Marine Station, Pacific Grove, Calif. 93950. 1277
evidence to suspect they are mutually coupled (4). Even in prolonged free runs in darkness, lasting up to 100 cycles, we find no evidence of the circadian activity cycle "decomposing" into two separate components such as one would expect were redundant pacemakers uncoupled and free to vary their frequency independently. The observations reported here strongly suggest that there are separate pacemakers in the left and right hemispheres of the cockroach brain, and that they interact in a way compatible with, indeed suggestive of, a mutual entrainment effected by reciprocal excitation. We have found no clear evidence of hemispheric asymmetry. MATERIALS AND METHODS The experimental system is the circadian rhythm of locomotor activity in the cockroach Leucophaea maderae. Individual insects were maintained within delicately balanced Lucite running-wheels whose rotation-caused by the insect moving-is sensed by a magnetic reed switch in the circuit of an event recorder (Esterline Angus). Food and water (replenished every few weeks) were available, ad lib., through the fixed vertical faceplate of the running wheel, The system yields data of the kind illustrated by Fig. 1. The period (r) of the free-running rhythm was assayed as the interval between the onsets of successive circadian activity periods. X for a given free run was derived from the slope of an eye-fitted line through the successive onsets (19). The individual free-running periods reported here represent the average of periods estimated by three people. Surgery was performed under CO2 anesthesia. The entire cephalic complex, optic lobes and midbrain, were exposed by cutting and raising a three-sided flap in the head capsule which is easily replaced and sealed with paraffin wax after surgery is finished. The insects appear unaffected by the CO2 anesthesia; operations were performed with about 95% survival, provided care was taken to remove heat from the light source used in the operations. In the few experiments we report involving optic lobe excision there was significant mortality which we now attribute, not to the operation itself, but the intense (hotter) light source we happened to use at that time. The same operation performed recently with properly filtered low light caused no mortality. Experimental Program. Fig. 2 summarizes the design of our experiments as well as earlier work on which they are based. The pathways given as lines in Fig. 2 are functional connections, not identified anatomical routes. The experiments fall into three series as follows: Series 1. After two transections-one of the retinal nerves at the base of the ommatidia, the other at the base of the optic lobe proximal to the lobula-the entire optic lobe was removed from one side of the brain. This was done on 14 animals; 4 had the left side removed, 10 had the right side removed. The
1278
Physiological Sciences: Page et al.
Xal~ D'([C+;'~ 25
Proc. Natl. Acad. Sci. USA 74 (1977) lamina medulla
DD O T X-
D D-
:.
-
optic
optic lobe
0NX DD
tR. 23.9
2
6
'C 23.7
presence of a free-running rhythm and its period were assayed before and after the operation. Series 2. The optic lobe on one side (left and right separately) was isolated from the midbrain by one transection proximal to the lobula. This operation was performed on 25 animals; in 12 the left lobe was isolated, and in 13 the right lobe was isolated. The isolated lobe was left in place. Again, the presence of a free-running rhythm and its period were assayed before and after the operation. Series 3. Control and experimental animals were exposed, after an assayed free-run in darkness, to a light cycle whose period (T) was 25 hr (4 light; 21 dark). In the experimental animals (n = 11) one side of the cephalic complex was isolated from its ipsilateral eye by transection of the retinal nerve. The entrainment behavior of the experimental and control animals was then compared. In some of the experimental animals, after weeks of entrainment, the second optic nerve was then cut and the animal returned to the 25-hr light cycle. The adequacy of the first retinal nerve transection and absence of regeneration were demonstrated by finding that the animal free-ran in the light cycle after transection of the second retinal nerve.
RESULTS AND DISCUSSION Both Optic Lobes Contain Pacemaking Elements. The intact animal free-runs in constant darkness at 250, with a circadian period (r) whose average value is 23.73 hr i 0.20, (Fig. 2-2; Table 1). The free-running rhythm is wholly unimpaired when the two eyes are isolated from the brain and optic lobes (Fig. 2-3; see refs. 20 and 21): the eyes themselves therefore cannot contain the pacemaker. The rhythm also persists after removal of either left or right optic lobe (Fig. 2-5 and 6) or its isolation from the midbrain by transection proximal to the lobula (Figs. 1 and 2-7 and 8; Table 1). However, when the midbrain is isolated from both optic lobes, by their transection (Fig. 2-4) the animal is aperiodic. This aperiodicity is the basis for the now general view (10-14) that the pacemaker is in the optic lobes. That conclusion is further encouraged by the finding (13, 14) that lesions of a relatively small part of the lobe, near the lobula, are sufficient to destroy effective pacemaker
T-25
vre -T-24
rC 024 4 Aperiodic
Tu25
9
* T 25
-
1 TI25
Zrj23.9
T r; =25 @023.7 COMPONENTJ
aj239
C
T*24
T-24
FIG. 1. Illustrative raw data from experimental series 2 and series 3. Upper. The first 2 weeks of the record the intact animal (L. maderae) was free-running at 250 in constant darkness (DD), with a period (T) of 23.63 hr. On day 15 the optic tract was cut (OTX) proximal to the lobula. The subsequent free run has a longer period (T = 24.45 hr). Lower. The animal (L. maderae) had its left optic nerve (retinal nerve) completely transected (DNX) 12 days before the beginning of the record, which shows a free run in DD at 250. On the day indicated, the animal was exposed to a light cycle whose period (T) was 25 hr (4 light, 21 dark). The rhythm entrains normally (see text).
lobe
Ta24
Tz24
LD 4:21
retina nerve
-
12
rC23.9
re
-T=25
T-25
T 25
FIG. 2. Experimental designs and principal results (L. maderae). The cephalic complex (eyes, optic lobes, and midbrain) is represented schematically: the lobula is indicated at the proximal end of each lobe. T = period, in hours, of an external light/dark cycle. T = period of free-running rhythm; Tc = that driven by the compound pacemaker (both left and right lobes); TL and TR = the periods of rhythms driven by the left and right lobes, respectively. rT is the period of the entrained rhythm; it equals T. Number of animals (n) given for each schematic represents total in published literature plus those reported here. 1. The system entrained by a light cycle with T = 24 hr (e.g., 12 light/12 dark); TO = T (20). 2. The system free-running in constant darkness at 250; TC is 23.7 hr (see Table 1) (21). 3. The system freerunning in light:dark 12:12 (250) after both compound eyes have been isolated by transection of each retinal nerve (22, 23) (n = 37). 4. The system after transection (or complete removal) of both optic lobes proximal to the lobula is totally aperiodic (11, 13-15) (n = 64). 5 and 6. Excision of entire optic lobe (left and right, respectively); a rhythm persists with T = 23.9 hr (see Table 1). For 5, n = 4, and for 6, n = 10 in this paper; see also refs. 13 and 14. 7 and 8. Isolation of the midbrain from left and right lobes, respectively, by transection proximal to lobula: r = 23.9 hr (Table 1). For 7, n = 11, and for 8, n = 14 in this paper; see also ref. 11. 9. The intact system readily entrains to a light cycle of T = 25 hr (4 light; 21 dark) (n = 11 in this paper). 10. So do animals in which the coupling to light has been destroyed on one side by transection of the retinal nerve. These animals show no 23.7-hr component in the periodicity of their activity (n = 11 in this paper). 11. One possible explanation is that both oscillators receive parallel input from each eye (hypothetical). 12. Another is mutual coupling of the left and right pacemakers, which also explains why Tc < TL and TR (hypothetical).
activity. However, the cautioi originally raised by Nishiitsutsuji-Uwo and Pittendrigh (10) remains appropriate: the results summarized by Fig. 2-3 through 8 do not demand, as unique explanation, the conclusion that a physically discrete group of cells (or single cell) within each lobe is an autonomous pacemaker. Two other possibilities must at least be considered. First, even if the pacemaker were located exclusively in the midbrain, periodicity could still be lost after its bilateral isolation (by transection of both lobes) from the rest of the central nervous system; the lobes, in this case, would be no more than essential output pathways. This possibility is also not rigorously eliminated by the finding that microlesions (13, 14) near the lobula are sufficient to cause arrhythmia although it renders it somewhat less plausible. However, our finding, reported in detail below, that pacemaker period changes after removal of either lobe makes it unlikely in the extreme that the lobe is
Proc. Natl. Acad. Sci. USA 74 (1977)
Physiological Sciences: Page et al.
Table 2. Statistical tests on effects of optic nerve section (ONX) and optic lobe isolation or removal (OLX) on period (r)
Table 1. Effects of optic lobe removal on free-running period*
n
One lobe (left) Series 1 Series 2 Combined One lobe (right) Series 1 Series 1 Combined One lobe (all cases) Series 1 Series 2 Combined
T
2 lobes
T 1 lobe
Ar
SD of AT
Sample
n
rleft lobe Vs. Tut lobe
23, 16
rbefore vs- rafter ONX
10 23.69 ± 0.19 23.89 ± 0.26 0.20 0.21 13 23.78 ± 0.19 24.01 ± 0.28 0.19 0.18 23 23.74 ± 0.19 23.95 ± 0.28 0.19 0.19 4 23.80 ± 0.13 23.97 ± 0.26 0.16 0.22 12 23.68 ± 0.23 23.96 ± 0.24 0.25 0.31 16 23.71 ± 0.21 23.96 ± 0.24 0.23 0.29 14 23.72 ± 0.18 23.91 ± 0.25 0.19 0.20 25 23.73 ± 0.21 23.98 ± 0.26 0.22 0.25 39 23.73 ± 0.20 23.96 ± 0.26 0.21 0.23
* Average values +SD.
nothing but an output channel. The issue then would be semantic: if activity of an output pathway were feeding back on the pacemaker and affecting its period, in what sense is it not
of that pacemaker? Second, although we now confidently conclude that each lobe (see below) contributes to pacemaker behavior, we still cannot exclude the possibility that each requires connections with, and feedback from, other cells in the midbrain to sustain its oscillation. This residual uncertainty does not impair the validity of the analysis we pursue here. The involvement of each lobe in determining pacemaker period implies a bilateral redundancy of pacemaker elements, which is what concerns us: are the bilaterally distributed elements symmetric in their behavior, and how do they interact? It is only for ease of expression that, in what follows, we make the simplifying assumption that the whole of each (left and right) pacemaker is contained within the corresponding optic lobe. Sufficiency and Equality of Left and Right Lobes. While several earlier papers (10, 12-14) have shown that a single lobe can pace a rhythm, the question of bilateral symmetry, or systematic asymmetry has not been addressed. Table 1 records the behavior of 39 animals from which one lobe had been removed entirely (series 1; n = 14) or isolated from the midbrain by transection of the entire lobe proximal to the lobula (series 2; n = 25). In 23 of the 39 cases the animals were driven by the left lobe, and in 16 by the right lobe. A clear circadian rhythm persisted in all 39 animals driven by a single lobe, and there was no indication, as judged by the period of these rhythms, that left (r = 23.95 + 0.28 hr) and right (7 = 23.96 ± 0.24 hr) pacemakers were different (Table 2)*. Compound Eyes Are Exclusive Pathway Coupling Pacepart
t
There was, as noted in Materials and Methods, substantial mortality in the series 1 experiments in which tlhe entire optic lobe was excised from one side. Of the 41 animals used, only 16 survived and gave data usable in Table 1; 31 died within 10 days after surgery during which activity, when present, was aperiodic. Twenty of the 31 deaths occurred in animals from which the left lobe had been excised. We cannot exclude the possibility that this significant asymmetry in the incidence of deaths may reflect a handedness in the nervous system, but neither can we exclude the handedness of the operator. We believe the effect is most likely due to the hot light source used in those
experiments coming from the left of the preparation.
1279
Tbefore Vs. rafter OLX ATONX vs.- ATOLX *
25 39 25, 25
Statistic
p
t =-0.004*
> 0.45
F =Ot F 19.50t F = 14.07t
>0.20
Mutual entrainment of bilaterally distributed circadian pacemakers (cockroach/optic lobes/circadian rhythms)
TERRY L. PAGE*, PATRICIA C. CALDAROLAt, AND COLIN S. PITTENDRIGH* Department of Biological Sciences, Stanford University, Stanford, California 94305; and t Stanford University Medical School, Stanford, California 94305
Contributed by Colin S. Pittendrigh, December 6, 1976
ABSTRACT The interactions between the bilaterally distributed components of the circadian system that controls the locomotor activity rhythm of the cockroach Leucophaea maderae were investigated in a series of surgical lesion experiments. Complete excision of one optic lobe (either right or left) or its surgical isolation from the central nervous system had no-effect on the animal's ability to free-run in constant darkness nor was there any indication, as judged by postoperative r values, of any difference between left and right lobe pacemakers. However, these surgical procedures consistently resulted in a significant increase in r over preoperative values while optic nerve section had no effect on r. The proposition is developed that the left and right pacemakers in the two optic lobes are mutally coupled and that the compound pacemaker's period is shorter than either of its constituent pacemakers. It was also found that the integrity of either compound eye is sufficient to assure entrainment of both left and right pacemakers.
The fact that a single eukaryotic cell can be an autonomous circadian pacemaker (1, 2) sharpens several questions about the temporal (circadian) organization of multicellular organisms. Is the rhythmicity they express in a diversity of functions attributable to only one (or a few) cells which, as a pacemaker, impose rhythmicity on other tissues and organs to which they are coupled? Or is the organism to some extent a population of autonomous cellular pacemakers? If so, what mechanisms maintain temporal order among them (3)? Although it is now commonly assumed (4, 5) that a metazoan includes many such autonomous pacemakers, whose order depends on a mixture of mutual and hierarchical entrainment, that assumption is based on very little in the way of hard fact. Clearly, progress depends on having first localized discrete pacemakers in a single organism; only then can their possible interaction be experimentally studied. Circadian pacemakers have now been localized in the nervous system of several bilaterally symmetric animals. They occur in each eye of the mollusc Aplysia (6), in the suprachiasmatic nucleus of the mammalian hypothalamus (7-9), and in the optic lobes of several insects (10-15) and in the midbrain of another (16). The bilaterality of these animals makes several questions experimentally tractable: is there an active pacemaker on each side (left and right) of the same individual? Is there any hemispheric asymmetry in pacemaker behavior? How are they maintained in synchrony? In many natural situations that problem might be trivially solved by their submission to entrainment by a common external (light/dark) cycle. Jacklet (17) and Lickey et al. (18) have already suggested that the left- and right-hand (eye) pacemakers are, in fact, not mutually coupled in Aplysia. But there are obvious ecological situations where that would be intolerable if synchrony between redundant pacemakers is important. Our interest in the cockroach derives from (a) the fact that widely separated redundant pacemakers have been localized in this insect and (b) there is preliminary *
Present address: Hopkins Marine Station, Pacific Grove, Calif. 93950. 1277
evidence to suspect they are mutually coupled (4). Even in prolonged free runs in darkness, lasting up to 100 cycles, we find no evidence of the circadian activity cycle "decomposing" into two separate components such as one would expect were redundant pacemakers uncoupled and free to vary their frequency independently. The observations reported here strongly suggest that there are separate pacemakers in the left and right hemispheres of the cockroach brain, and that they interact in a way compatible with, indeed suggestive of, a mutual entrainment effected by reciprocal excitation. We have found no clear evidence of hemispheric asymmetry. MATERIALS AND METHODS The experimental system is the circadian rhythm of locomotor activity in the cockroach Leucophaea maderae. Individual insects were maintained within delicately balanced Lucite running-wheels whose rotation-caused by the insect moving-is sensed by a magnetic reed switch in the circuit of an event recorder (Esterline Angus). Food and water (replenished every few weeks) were available, ad lib., through the fixed vertical faceplate of the running wheel, The system yields data of the kind illustrated by Fig. 1. The period (r) of the free-running rhythm was assayed as the interval between the onsets of successive circadian activity periods. X for a given free run was derived from the slope of an eye-fitted line through the successive onsets (19). The individual free-running periods reported here represent the average of periods estimated by three people. Surgery was performed under CO2 anesthesia. The entire cephalic complex, optic lobes and midbrain, were exposed by cutting and raising a three-sided flap in the head capsule which is easily replaced and sealed with paraffin wax after surgery is finished. The insects appear unaffected by the CO2 anesthesia; operations were performed with about 95% survival, provided care was taken to remove heat from the light source used in the operations. In the few experiments we report involving optic lobe excision there was significant mortality which we now attribute, not to the operation itself, but the intense (hotter) light source we happened to use at that time. The same operation performed recently with properly filtered low light caused no mortality. Experimental Program. Fig. 2 summarizes the design of our experiments as well as earlier work on which they are based. The pathways given as lines in Fig. 2 are functional connections, not identified anatomical routes. The experiments fall into three series as follows: Series 1. After two transections-one of the retinal nerves at the base of the ommatidia, the other at the base of the optic lobe proximal to the lobula-the entire optic lobe was removed from one side of the brain. This was done on 14 animals; 4 had the left side removed, 10 had the right side removed. The
1278
Physiological Sciences: Page et al.
Xal~ D'([C+;'~ 25
Proc. Natl. Acad. Sci. USA 74 (1977) lamina medulla
DD O T X-
D D-
:.
-
optic
optic lobe
0NX DD
tR. 23.9
2
6
'C 23.7
presence of a free-running rhythm and its period were assayed before and after the operation. Series 2. The optic lobe on one side (left and right separately) was isolated from the midbrain by one transection proximal to the lobula. This operation was performed on 25 animals; in 12 the left lobe was isolated, and in 13 the right lobe was isolated. The isolated lobe was left in place. Again, the presence of a free-running rhythm and its period were assayed before and after the operation. Series 3. Control and experimental animals were exposed, after an assayed free-run in darkness, to a light cycle whose period (T) was 25 hr (4 light; 21 dark). In the experimental animals (n = 11) one side of the cephalic complex was isolated from its ipsilateral eye by transection of the retinal nerve. The entrainment behavior of the experimental and control animals was then compared. In some of the experimental animals, after weeks of entrainment, the second optic nerve was then cut and the animal returned to the 25-hr light cycle. The adequacy of the first retinal nerve transection and absence of regeneration were demonstrated by finding that the animal free-ran in the light cycle after transection of the second retinal nerve.
RESULTS AND DISCUSSION Both Optic Lobes Contain Pacemaking Elements. The intact animal free-runs in constant darkness at 250, with a circadian period (r) whose average value is 23.73 hr i 0.20, (Fig. 2-2; Table 1). The free-running rhythm is wholly unimpaired when the two eyes are isolated from the brain and optic lobes (Fig. 2-3; see refs. 20 and 21): the eyes themselves therefore cannot contain the pacemaker. The rhythm also persists after removal of either left or right optic lobe (Fig. 2-5 and 6) or its isolation from the midbrain by transection proximal to the lobula (Figs. 1 and 2-7 and 8; Table 1). However, when the midbrain is isolated from both optic lobes, by their transection (Fig. 2-4) the animal is aperiodic. This aperiodicity is the basis for the now general view (10-14) that the pacemaker is in the optic lobes. That conclusion is further encouraged by the finding (13, 14) that lesions of a relatively small part of the lobe, near the lobula, are sufficient to destroy effective pacemaker
T-25
vre -T-24
rC 024 4 Aperiodic
Tu25
9
* T 25
-
1 TI25
Zrj23.9
T r; =25 @023.7 COMPONENTJ
aj239
C
T*24
T-24
FIG. 1. Illustrative raw data from experimental series 2 and series 3. Upper. The first 2 weeks of the record the intact animal (L. maderae) was free-running at 250 in constant darkness (DD), with a period (T) of 23.63 hr. On day 15 the optic tract was cut (OTX) proximal to the lobula. The subsequent free run has a longer period (T = 24.45 hr). Lower. The animal (L. maderae) had its left optic nerve (retinal nerve) completely transected (DNX) 12 days before the beginning of the record, which shows a free run in DD at 250. On the day indicated, the animal was exposed to a light cycle whose period (T) was 25 hr (4 light, 21 dark). The rhythm entrains normally (see text).
lobe
Ta24
Tz24
LD 4:21
retina nerve
-
12
rC23.9
re
-T=25
T-25
T 25
FIG. 2. Experimental designs and principal results (L. maderae). The cephalic complex (eyes, optic lobes, and midbrain) is represented schematically: the lobula is indicated at the proximal end of each lobe. T = period, in hours, of an external light/dark cycle. T = period of free-running rhythm; Tc = that driven by the compound pacemaker (both left and right lobes); TL and TR = the periods of rhythms driven by the left and right lobes, respectively. rT is the period of the entrained rhythm; it equals T. Number of animals (n) given for each schematic represents total in published literature plus those reported here. 1. The system entrained by a light cycle with T = 24 hr (e.g., 12 light/12 dark); TO = T (20). 2. The system free-running in constant darkness at 250; TC is 23.7 hr (see Table 1) (21). 3. The system freerunning in light:dark 12:12 (250) after both compound eyes have been isolated by transection of each retinal nerve (22, 23) (n = 37). 4. The system after transection (or complete removal) of both optic lobes proximal to the lobula is totally aperiodic (11, 13-15) (n = 64). 5 and 6. Excision of entire optic lobe (left and right, respectively); a rhythm persists with T = 23.9 hr (see Table 1). For 5, n = 4, and for 6, n = 10 in this paper; see also refs. 13 and 14. 7 and 8. Isolation of the midbrain from left and right lobes, respectively, by transection proximal to lobula: r = 23.9 hr (Table 1). For 7, n = 11, and for 8, n = 14 in this paper; see also ref. 11. 9. The intact system readily entrains to a light cycle of T = 25 hr (4 light; 21 dark) (n = 11 in this paper). 10. So do animals in which the coupling to light has been destroyed on one side by transection of the retinal nerve. These animals show no 23.7-hr component in the periodicity of their activity (n = 11 in this paper). 11. One possible explanation is that both oscillators receive parallel input from each eye (hypothetical). 12. Another is mutual coupling of the left and right pacemakers, which also explains why Tc < TL and TR (hypothetical).
activity. However, the cautioi originally raised by Nishiitsutsuji-Uwo and Pittendrigh (10) remains appropriate: the results summarized by Fig. 2-3 through 8 do not demand, as unique explanation, the conclusion that a physically discrete group of cells (or single cell) within each lobe is an autonomous pacemaker. Two other possibilities must at least be considered. First, even if the pacemaker were located exclusively in the midbrain, periodicity could still be lost after its bilateral isolation (by transection of both lobes) from the rest of the central nervous system; the lobes, in this case, would be no more than essential output pathways. This possibility is also not rigorously eliminated by the finding that microlesions (13, 14) near the lobula are sufficient to cause arrhythmia although it renders it somewhat less plausible. However, our finding, reported in detail below, that pacemaker period changes after removal of either lobe makes it unlikely in the extreme that the lobe is
Proc. Natl. Acad. Sci. USA 74 (1977)
Physiological Sciences: Page et al.
Table 2. Statistical tests on effects of optic nerve section (ONX) and optic lobe isolation or removal (OLX) on period (r)
Table 1. Effects of optic lobe removal on free-running period*
n
One lobe (left) Series 1 Series 2 Combined One lobe (right) Series 1 Series 1 Combined One lobe (all cases) Series 1 Series 2 Combined
T
2 lobes
T 1 lobe
Ar
SD of AT
Sample
n
rleft lobe Vs. Tut lobe
23, 16
rbefore vs- rafter ONX
10 23.69 ± 0.19 23.89 ± 0.26 0.20 0.21 13 23.78 ± 0.19 24.01 ± 0.28 0.19 0.18 23 23.74 ± 0.19 23.95 ± 0.28 0.19 0.19 4 23.80 ± 0.13 23.97 ± 0.26 0.16 0.22 12 23.68 ± 0.23 23.96 ± 0.24 0.25 0.31 16 23.71 ± 0.21 23.96 ± 0.24 0.23 0.29 14 23.72 ± 0.18 23.91 ± 0.25 0.19 0.20 25 23.73 ± 0.21 23.98 ± 0.26 0.22 0.25 39 23.73 ± 0.20 23.96 ± 0.26 0.21 0.23
* Average values +SD.
nothing but an output channel. The issue then would be semantic: if activity of an output pathway were feeding back on the pacemaker and affecting its period, in what sense is it not
of that pacemaker? Second, although we now confidently conclude that each lobe (see below) contributes to pacemaker behavior, we still cannot exclude the possibility that each requires connections with, and feedback from, other cells in the midbrain to sustain its oscillation. This residual uncertainty does not impair the validity of the analysis we pursue here. The involvement of each lobe in determining pacemaker period implies a bilateral redundancy of pacemaker elements, which is what concerns us: are the bilaterally distributed elements symmetric in their behavior, and how do they interact? It is only for ease of expression that, in what follows, we make the simplifying assumption that the whole of each (left and right) pacemaker is contained within the corresponding optic lobe. Sufficiency and Equality of Left and Right Lobes. While several earlier papers (10, 12-14) have shown that a single lobe can pace a rhythm, the question of bilateral symmetry, or systematic asymmetry has not been addressed. Table 1 records the behavior of 39 animals from which one lobe had been removed entirely (series 1; n = 14) or isolated from the midbrain by transection of the entire lobe proximal to the lobula (series 2; n = 25). In 23 of the 39 cases the animals were driven by the left lobe, and in 16 by the right lobe. A clear circadian rhythm persisted in all 39 animals driven by a single lobe, and there was no indication, as judged by the period of these rhythms, that left (r = 23.95 + 0.28 hr) and right (7 = 23.96 ± 0.24 hr) pacemakers were different (Table 2)*. Compound Eyes Are Exclusive Pathway Coupling Pacepart
t
There was, as noted in Materials and Methods, substantial mortality in the series 1 experiments in which tlhe entire optic lobe was excised from one side. Of the 41 animals used, only 16 survived and gave data usable in Table 1; 31 died within 10 days after surgery during which activity, when present, was aperiodic. Twenty of the 31 deaths occurred in animals from which the left lobe had been excised. We cannot exclude the possibility that this significant asymmetry in the incidence of deaths may reflect a handedness in the nervous system, but neither can we exclude the handedness of the operator. We believe the effect is most likely due to the hot light source used in those
experiments coming from the left of the preparation.
1279
Tbefore Vs. rafter OLX ATONX vs.- ATOLX *
25 39 25, 25
Statistic
p
t =-0.004*
> 0.45
F =Ot F 19.50t F = 14.07t
>0.20
