THE JOURNAL OF EXPERIMENTAL ZOOLOGY 247:99-108 (1988)

Effect of a Light Pulse During the Dark on Photoperiodic Regulation of the Rate of Thyroxine-Induced, Spontaneous, and Prolactin-Inhibited Metamorphosis in Rana pipiens Tadpoles

MARY L. WRIGHT, SUZANNE T. JOREY, LINDA S. BLANCHARD, AND CAROLYN A. BASSO Biology Department, College of Our Lady of the Elms, Chicopee, Massachusetts 01013

ABSTRACT Since Rana pipiens tadpoles injected with thyroxine (TJ early in the dark develop more slowly than those injected in the light, we studied the effect of giving a light pulse of 1 hr early in the dark. Tadpoles injected under a 7.5-W red light bulb in a darkened room with 0.2 pg T4 daily at 2200 hr went through metamorphosis faster on a 12L:3D:lL:8D cycle with a light pulse after injection than on a 12L:12D cycle without a light pulse, and even faster on a 12L:1.5D:lL:9.5D cycle with a light pulse before the injection. Thus a 1-hr light pulse counteracted the metamorphic delay resulting from administration of T4 in the dark, and set in motion the conditions that resulted in a more rapid response to an injection of T4. However, a 1-hr light pulse in the early dark had no effect on growth and development of older or younger untreated tadpoles or those constantly immersed in 30 pg/liter T4.

Larvae on 21L:3D with T4 injection in the dark and on 12L:3D:lL:8D with T4 injection at 0700 hr just before the start of the main light phase progressed faster than 12L:3D:lL:8D with injection at 2200 hr in the dark before only a 1-hr light pulse. Thus the length of the light phase immediately after T4 injection was significant. There was no difference on 12L:12D and 12L:3D:lL:8D cycles in the effectiveness of daily injections of 10 pg prolactin (PRL) in the early dark at 2200 hr in promoting tail growth or antagonizing tail resorption induced by T4 immersion. Under these conditions, PRL utilization did not appear to be inhibited by the light pulse.

Many factors affect the rate of metamorphosis in anuran tadpoles. Environmental variables such as photoperiod, crowding, abundance of food, and time of feeding all modulate metamorphic progress (for reviews see Dent, '68; Kaltenbach, '68; Just et al., '81). Light and the light/dark (LD) cycle apparently interact with, or influence, the hormones that con- trol tadpole development. When the photoperiod was lengthened within a 24-hr day, development was accelerated in Rana temporaria (GuyBtant, '641, Alytes obstetricans (Disclos, '591, and Rana pipiens (Eichler and Gray, '76; Wright, '84) tad- poles, although an inhibitory effect of longer pho- toperiods on larval development was reported for Xenopus Zaevis (Edwards and Pivorun, '86; Del- gad0 et al., '87) and Discoglossus pictus (Delgado et al., '84; Gutikrrez et al., '84).

The effectiveness of thyroxine (T4) in promoting tadpole metamorphosis depends on the time in the LD cycle that the hormone is given. T4 was most effective in promoting both constructive and re-

0 1988 ALAN R. LISS, INC.

gressive metamorphic changes when injected late in the dark phase or in the early or mid light phase, or when tadpoles were immersed in the hormone for a daily 8-hr span at least partly in the light. T4 was least effective if injection was early in the dark phase or if the immersion span was entirely in the dark (Wright et al., '86). Thus there are circadian variations in target tissue respon- siveness to exogenous T4 that depend on the occur- rence of light during T4 immersion or following T4 injection.

The present work was designed to examine fur- ther the relationship of light and photoperiod to anuran larval development. Since tadpoles in- jected with T4 early in the dark developed more slowly than those injected in the light, we studied the effect of a 1-hr light pulse in the early dark on

Address reprint requests to Sr. Mary L. Wright, Biology Depart- ment, College of Our Lady of the Elms, Chicopee, MA 01013.

100 M.L. WRIGHT ET AL.

the progress of spontaneous metamorphosis and that induced by constant T4 immersion or T4 injec- tion before or after the light break. A light pulse in the dark can combine with the main light phase to produce a skeleton photoperiod which consists of two light periods separated by dark (Pittendrigh and Minis, '64). If a skeleton photoperiod is as effective as a single long photoperiod, light may be involved only in the initial stages of T4 utilization, and events set in motion then may not need light to continue. Therefore, an asymmetric skeleton photoperiod was also compared to a single long photoperiod.

Circadian variations in the effectiveness of T4 in promoting metamorphosis could be due to rhythms of hormones that antagonize or synergize with T4. The pituitary hormone prolactin (PRL) antago- nizes tail resorption (Bern et al., '67; Etkin and Gona, '67) and hindlimb growth and development (Wright et al., '79) in metamorphosis. PRL has been reported to more effectively promote regen- eration of newt limbs when administered in the dark (Maier and Singer, '81). In premetamorphic tadpoles, although PRL antagonized hindlimb growth and development just as well at any time in the LD cycle, tail fin growth was faster when PRL was given in the early dark (Wright et al., '86). If PRL was inhibited by light and naturally secreted at night, this hormone might retard the response to T4 given in the dark. Consequently, final experiments studied the effect of a light pulse of 1 hr on the response of tadpoles in spontaneous and T4-induced metamorphosis to PRL injection in the early dark.

MATERIALS AND METHODS Rana pipiens tadpoles were obtained by induced

ovulation and subsequent fertilization of eggs from adult frogs supplied by Connecticut Valley Biolog- ical Supply Co., Inc. (Southampton, MA). After hatching, tadpoles were raised in plastic dishpans of 10% Holtfreter's solution occasionally supple- mented with spring water. During each experi- ment, tadpoles were kept in lighted incubators or growth chambers at 22°C with timers set to regu- late the appropriate LD cycle. Feeding of washed, canned spinach took place once daily at 1500 & 1.5 hr. In the T4 experiment with older tadpoles, ani- mals were in the late prometamorphic stages XVII to XVIII (Taylor and Kollros, '46). In all the other experiments, the tadpoles were in premetamorphic stages V-XI. In a single experiment, however, the maximum span was a three-stage range. In each investigation, the groups of tadpoles of the same

clutch were made equivalent by stage at the start. Five of the experiments reported on here were done in the summer, two in the winter, and one in the spring, all within a span of one year.

In the T4 work, tadpoles were treated with DL- Na-T4 by immersion in 30 pg/liter (3.8 x lo-' M) in 10% Holtfreter"~ solution or by daily intraperi- toneal injection of 0.2 pg/0.02 ml 0.7% saline solu- tion at pH 8. Bovine PRL(N1AMDD-bPRL-6,30 IU/ mg, and USDA-bPRL-B-1,13 IU/mg) was dissolved in 0.7%saline solution at pH 8. Tadpoles were in- jected intraperitoneally through the tail muscle with 10 pg PRL daily in a volume of 0.02 ml. Both T4- and PRL-injected tadpoles were kept immersed in 10% Holtfreter's solution, except in the T4-PRL experiment, where some of the PRL-injected ani- mals were also immersed in 30 pg/liter T4. Injec- tions of T4 or PRL during the dark phase were done in a darkened room under a 7.5-W red light bulb so as not to disturb the LD cycle.

In the T4 experiments, changes were monitored in tail and hindlimbs. At appropriate intervals tail height at the widest point, tail length, hindlimb length, and stage were measured. Stage progres- sion is initially based on the morphogenesis of the hindlimb and the appearance of digits, so it is an indicator of hindlimb development until late pro- metamorphosis, when repressive events such as resorption of the anal tail piece and thinning of the skin window over the forelimbs serve as stag- ing criteria (Taylor and Kollros, '46). Although in premetamorphic tadpoles PRL induces a rapid in- crease in tail height and, if used with exogenous T4, retards tail regression, PRL inhibition of hind- limb growth and development takes more time than the length of the PRL experiments (Wright et al., '79). Consequently, hindlimb growth was not monitored in the PRL studies. The data were eval- uated by use of analysis of variance (ANOVA) and/ or Student's t-test, both performed on a computer.

RESULTS In the first experiment, tadpoles were separated

by stage into two equivalent groups of 20 tadpoles each. One group was kept on our standard 12L:12D cycle with light from 0800 to 2000 hr, while the other group received an additional hour of light in the early dark from 2300 to 2400 hr, generating a 12L:3D:lL:BD cycle. T4 injections were given to both groups at 2200 hr, 2 hr after the onset of dark. The extra hour of light shortly after the T4 injec- tion promoted faster metamorphosis. The 12L:3D:lL:8D group was significantly ahead of the 12L:12D group in tail regression (Fig. lA,B) and

PHOTOPERIODIC REGULATION OF METAMORPHIC RATE 101

TIME (DAYS)

Fig. 1. Effect of 1 hr of light in the early dark at 2300 hr on sion (A,B) and hindlimb growth (C) and development (D) than tadpole metamorphosis induced by daily injection of T4 at those on a 12L:12D cycle. Statistically significant differences 2200 hr. There were 20 tadpoles in each group. The tadpoles between the means of the two groups are indicated by aster- on the 12L:3D:lL:8D cycle showed a faster rate of tail regres- isks (* = P < .01; ** = P < .001).

hindlimb growth (Fig. 1C) and development (Fig. 1D). Thus light seemed to be important to the utilization of T4, and 1 hr of it in the early dark counteracted the detrimental effect of giving T4 by injection at this time.

In the next experiment, untreated, naturally de- veloping premetamorphic tadpoles were kept on 12L:12D or 12L:3D:lL:8D cycles. After 21 days un- der these different regimens, there were no statis- tically significant differences between the two groups in tail height or length, or in hindlimb growth or development (Table 1). The same results were found when matched groups of these accli- mated tadpoles were then immersed in 30 pglliter T4 while still on 12L:lZD or 12L:3D:lL:8D cycles. There were no statistically significant differences between the two groups in the rate of tail resorp- tion or hindlimb growth and development after 8 days of constant immersion (Table 2). The effective- ness of 1 hr of light in the early dark was also studied in older tadpoles just before the onset of metamorphic climax. One group was kept on a

12L:12D cycle while the other received an addi- tional hour of light from 2200 to 2300 hr, placing it on a 12L:ZD:lL:9D cycle. During an 8-day pe- riod, the tadpoles progressed well into spontaneous metamorphic climax where the endogenous T4 level is high (see White and Nicoll, '81), yet there

TABLE 1. Effect of a light pulse in the early dark on growth and development of untreated prernetamorphic tadpoles'

Lightldark cycle -

Parameter' Day 12L:12D 12L:3D:lL:8D

Tail height (mm) 0 7.5 f 0.1 7.3 f 0.1 21 9.6 k 0.3 9.4 k 0.2

Tail length (mm) 0 23.9 f 0.2 24.2 f 0.2 21 30.8 f 0.9 29.4 f 0.7

Hindlimb length (mm) 0 1.7 f 0.04 1.7 f 0.04 21 3.7 f 0.2 3.1 f 0.2

Stage 0 6.1 + 0.1 6.1 f 0.1 21 10.5 k 0.4 9.9 * 0.2

'There were 30 tadpoles in each group. 'Mean f S.E.

102 M.L. WRIGHT ET AL.

TABLE 2. Effect of a light pulse in the early dark on growth and development of tadpoles constantly immersed in

30 UPII TA'

Lightldark cycle Parameter' Day 12L:12D 12L:3D:lL:8D

Tail height (mm) 0 11.2 f 0.3 11.5 k 0.3 8 7.0 f 0.2 7.6 k 0.2

Tail length (mm) 0 31.7 f 0.7 32.1 + 0.7 8 25.4 f 0.7 25.9 f 0.8

Stage 0 10.5 f 0.2 10.4 f 0.2 8 18.6 k 0.8 17.6 f 1.0

Hindlimb length (mm) 0 4.2 0.2 4.1 + 0.2 8 8.1 f 0.6 7.7 f 0.7

'There were 13 tadpoles in each group. 'Mean 5 S.E.

was no difference in the rate of metamorphosis in the two groups (Fig. 2). Thus an additional hour of light early in the dark had no effect on metamor- phosis induced by T4 immersion-where the stim- ulus of hormone injection was lacking and exogenous T4 was presumably taken up con- stantly-and no effect on the progress of sponta- neous metamorphosis in young or older tadpoles. It seemed that an hour of light in the early dark only made a difference in metamorphic rate when it occurred soon after an injection of exogenous T4.

A subsequent experiment was performed both to repeat and confirm the results showing faster de- velopment with a T4 injection and a light pulse in

HEIGHT LENGTH LENGTH

Fig. 2. Effect of a 1-hr light pulse in the early dark (12L:2D:lL:9D cycle) on late prometamorphic tadpoles in spon- taneous metamorphosis. There were seven tadpoles at initial stages XVII to XVIII in each group. No significant differences in the rate of tail regression, shank growth, or stage progres- sion were found between the two groups.

121: 12D - 2200

i

2200 12L:3D: 11:8D-

12L: 1.5D : 11 : 9.5D - 2230

0800 2000 2300 0800

Fig. 3. Experimental protocol used to compare the effective- ness of T4 injected in the early dark before (12L:3D:lL:8D) or after (12L:1.5D:lL:9.5D) a 1-hr light pulse. Arrows indicate the time of daily T4 injection. Results are in Figures 4 and 5.

the early dark and to investigate the effect of giv- ing the light pulse before the T4 injection, a 12L:1.5D:lL:9.5D cycle. Thus tadpoles' metamor- phic rate was compared when, accompanying an injection of T4 in the early dark at 2200 hr, there was no light, an hour of light after the injection, or an hour of light immediately before the injection, which in this case was given at 2230 hr in the dark (Fig. 3). Tail resorption was faster in the groups receiving a light pulse and lagged behind in the 12L:12D group without the light pulse (Fig. 4). On day 6, the last day of measurement, both experi- mental groups were significantly different from the control 12L:12D group and also from each other, with the 12L:1.5D:lL:9.5D showing a faster rate of tail resorption than the 12L:3D:lL:8D group (Fig. 4). The same results were found with hind- limb growth and stage progression (Fig. 5). These findings confirmed the results of the first experi- ment presented and also showed that metamorpho- sis proceeded even faster when the light pulse was given before the T4 injection rather than an hour after it.

In the T4 injection experiments above, the light pulse in the early dark could have combined with the main light phase to give an asymmetric skele- ton photoperiod, producing acceleration of meta- morphosis by a long photoperiod effect. With the light pulse serving as the initiator of the photope- riod, a 12L:3D:lL:8D cycle with T4 injection at 2200 hr might be equivalent to a 21L:3D cycle. The last T4 experiment was performed to examine

PHOTOPERIODIC REGULATION OF METAMORPHIC RATE 103

10

c r P w

6

4 121:12D

.-121:3D:IL : 8 D

***

I I 1 4 6 TIME (DAYS)

Fig. 4. Effect on tail regression of a light pulse before (12L:1.5D:lL:9.5D) or after 12L:3D:lL:8D) T4 injection in the early dark. There were 20 tadpoles in each group. Tail resorp- tion was fastest when the light pulse came before the T4 injection and slowest with no light pulse a t all. Statistically significant differences among the three groups using ANOVA are indicated by asterisks on the graphs (* = P < .05; ** = P < .01; *** = P < 0.005). In both tail height (A) and tail length (B), t-tests showed that by the last day the two groups with the light pulse were each significantly different from the 12L:12D control (P < .001) and from each other (P < .01 for tail height and P < ,051 for tail length).

K

this possibility. A 12L:3D:lL:8D cycle with T4 in- jection at 0700 hr, 1 hr before the start of the principal photoperiod, was also studied (Fig. 6).

Tail resorption was significantly faster in the 21L:3D cycle and the 12L:3D:lL:8D cycle with T4 injection at 0700 hr than in the 12L:3D:lL:8D group with T4 at 2200 hr (Fig. 7). There were no significant differences between the 21L:3D and the 12L:3D:lL:8D group with T4 at 0700 hr. In hind- limb parameters, there were no significant differ- ences in the rate of limb growth among the three groups, although by day 7 there was a trend for the 21L:3D group to be ahead of the other two (Fig. 8A). There was little variability within each group in stage progression, so from day 4 on, the 21L:3D group forged significantly ahead of the other two

IA

xx

XVlll

XVI

w XIV W a c vI XI1

x

Vll l

I I I I 0 2 4 6

TIME ( D A Y S )

Fig. 5. Effect on the hindlimb of a light pulse before (12L:l.5D:lL:9.5D) or after (12L:3D:lL:8D) T4 injection in the early dark. There were 20 tadpoles in each group. Tadpoles progressed further in hindlimb growth and stage in the groups with a light pulse in the early dark, while the 12L:12D control lagged behind. Statistically significant differences among the three groups using ANOVA are indicated by asterisks (* = P < .05; ** P < .025; *** = P < .005). In hindlimb length (A), t-tests indicated significant differences from the 12L:12D con- trol on days 2 and 4 (P < .05) for the 12L:3D:lL:8D group and on days 4 and 6 (P < .01) for the 12L:1.5D:lL:9.5D group. In stage progression (R), such differences occur in the 12L:3D:lL:8D group on days 2 (P < .02), and 4 and 6 (P < .001) and in the 12L:1.5D:lL:9.5D group on days 4 (P < .01) and 6 (P < ,001). The two experimental groups receiving light pulses in the early dark were not significantly different from each other in hindlimb length and differed on days 2 and 6 (P < .02) in stage.

in hindlimb development (Fig. 8B). On day 7, how- ever, the picture changed and became essentially the same as with tail resorption. Both the 12L:3D:lL:8D A.M. T4 injection and the 21L:3D groups were significantly ahead of the 12L: 3D:lL:gD group with T4 injection at 2200 hr. But by day 7, regressive events such as appearance of the skin window and emergence of the forelimbs become the determinants of stage progression rather than the development of the hindlimb. These results indicate that the 12L:3D:lL:8D group

104 M.L. WRIGH!I' ET AL.

211:3D -

'5""

A

.-121:3D:IL:8D-T4 2200 A--21L;3D ~ - - -12L:3D: lL :8D-T4 0700 **

I I

1 2 1 : 3 0 : 1 L : 8 0 - B

- 0800 2000 2 3 0 0 0800 1 25-

Fig. 6. Experimental protocol used to determine if the skel- eton photoperiod on a 12L:3D:lL:8D cycle resulted in a rate of development similar to a long photoperiod beginning at the light pulse (21L3D cycle). The three LD cycles shown were compared. The two 12L:3D:lL:8D cycles differed only in the time of T, injection (indicated by arrows). The 12L:3D:lL:8D cycle with T4 at 0700 hr was used to assess the effectiveness of T4 given before the main light phase instead of before the

I

short light pulse. 0 2 4 6 7

20- a

d 3 15- b *

I

TIME (DAYS)

with T4 injection at 2200 hr did not have a func- tional asymmetric skeleton photoperiod of 21 hr, since it developed more slowly than the 21L:3D group. The actual length of the light phase follow- ing a T4 injection may be more important. If a long photoperiod follows administration of T4 such as in the 21L:3D and the 12L:3D:lL:8D cycle with T4 at 0700 hr, metamorphosis is accelerated in compari- son to when only a 1-hr light-pulse immediately follows T4 injection.

Since a 1-hr light pulse was more effective if given before a T4 injection than after it, light might set in motion the conditions that result in a more rapid response to T4. The hormone PRL, which antagonizes metamorphosis, might affect the tis- sue response to T4. An experiment was performed to determine if a light pulse in the early dark interfered with the effect of PRL injected at 2200 hr into tadpoles in spontaneous metamorphosis. There were no signxicant differences in the rate of tail growth or stage progression on 12L:12D and 12L:3D:lL:8D cycles (Table 3). A further experi- ment examined whether PRL, injected at 2200 hr in the dark, was more effective in antagonizing metamorphosis induced by constant immersion in T4 with or without a light pulse in the early dark. Although PRL did not affect stage progression in

Fig. 7. Comparison of metamorphic progress when T4 was injected in the dark at 2200 hr before a short light pulse, at 0700 hr before the main light phase, or on a 21L:3D cycle with a long light phase after the injection (see Fig.6 for protocol). There were 20 tadpoles in each group. In tail regression, the group that received a short light pulse only after T4 (12L:3D:1L:8D-T4 2200 hr) lagged behind the groups with a longer light phase after injection. Asterisks on the graphs indicate significant differences among the three groups using ANOVA ( * = P < .01; ** = P < .005). With t-tests, the 12L:3D:lL:8D group with T4 injection at 2200 hr was signifi- cantly different from the group with T4 injection at 0700 hr in tail height (A) on day 7 (P < .001) and in tail length (B) on days 6 and 7 (P < ,011, and from the 21L:3D group only in tail length on day 6 (P < .Ol), although the difference was nearly significant on day 7 in both parameters. There were no signif- icant differences between the 12L.3D.lLBD group with T4 at 0700 hr and the 21L:3D group.

TABLE 3. Effect of intravenous injection of atriopeptin III at a dose o f 30 uglkg on the plasma renin concentration

(PRC) and hematocrit (HT) of the freshwater turtle'

Control Experiment Recovery (-20 min) (20 min) (80 min)

PRC 3.18 + 0.36 3.30 ? 0.40 3.14 +- 0.42

HT 29.1 k 1.0 28.9 & 0.7 28.3 2 0.3 (ngAIimllhr)

'Values are the means * SEM of six experiments. Experiments were conducting according to experimental protocol I in Materials and Methods.

c PHOTOPERIODIC REGULATION OF METAMORPHIC RATE 105 A

#~ this short experiment, it retarded tail resorption to the same extent on both the 12L:lZD and 12L:3D:lL:8D cycles (Table 4). Thus there was no evidence that a short light pulse in the early dark interfered with the utilization of PRL adminis- tered at that time.

I DISCUSSION I Previous work (Wright et al., '86) indicated that

on a 12L:lZD cycle, T4 was more effective in pro- moting metamorphic changes when injected late in the dark, or in the early or mid light phase, and least effective when administered early in the dark. These findings showed that there was circadian variation in the response of the tissues to T4 that depended on the occurrence of light following T4 injection. The results presented here demonstrated further that a 1-hr light pulse before or after a T4 injection counteracted the metamorphic delay re- sulting from administration of T4 in the early dark and restored a higher level of tissue sensitivity and responsiveness to the hormone. Since both sponta- neous (Eichler and Gray, '76) and T4-induced (Wright et al., '88) metamorphosis were acceler-

B .-12L:3D : 11: 8 D - T4 2200 A- -21 L: 3 D

xx-

xvIII-

XVI-

n - - I Z L : 3 D : l L : 8 D - T 4 0700

Y

l3 XIV- 2 XII-

IA

X-

VIII-

0 2 4 6 7 TIME (DAYS)

Fig. 8. Comparison of limb growth and stage progression ated by long photoperiods, the light pulse in the when T4 injection in the dark was followed by a short light pulse (12L:3D:lL:8D-T4 2200) or a long light phase (21L:3D and 12L3D:lL:8D-T4 0700 groups). There were 20 tadpoles by a long photoperiod effect. in each group. There were no statistically significant differ-

dark might have the response to T4

An extended light phase combined with a short ences among the different regimens in the rate of hindlimb pulse in the dark forms an asymmetric skeleton

the same in the three groups, the variability was small so that with ANOVA there were significant differences on the last 3 between the Onset Of the main photoperiod and the days as indicated by asterisks on the graphs (* = p < .05; ** end Of the pulse or Of the interval between the = P < .005). At first the 21L:3D group was ahead of the other onset of the pulse and the end of the main photo- two groups, but on day 7, both s o u p s with longer light periods period (Pittendrigh and Minis, '64). With such a after T4 were significantly ahead of the group with Only a l-hr cycle, long photoperiod effects may be obtained light pulse after the T4 injection. With t-tests, the with two shorter periods of light. The activity cycle 12L:3D:lL:8D group with T4 a t 2200 hr was significantly different from the 21L:3D group on days 6 and 7 (P < .01) and of rodents (Pittendrigh and Daan, '76) and the from the other 12L:3D:lL:8D group on day 7 (P < .02). There growth and performance of ram lambs (Schan-

growth (A). Although stage Progression S3Xled to be photoperiod whose length consists of the interval

TABLE 4. Effect of a light pulse in the early dark on PRL antagonism of metamorphic changes induced by constant immersion in 30 pglliter T4'

Light/dark cycle 12L:12D

Parameter' Day T4 TCPRL

Tail height (mm) 0 9.1 f 0.1 9.0 f 0.1 7 7.1 f 0.2 8.9 f 0.2*

7 25.0 * 0.4 27.5 + 0.6*

7 19.1 + 0.1 18.9 f 0.1

Tail length (mm) 0 25.0 f 0.2 25.2 k 0.3

Stage 0 5.0 f 0.0 5.0 f 0.0

12L:3D:lL:8D T4 T4-PRL

9.2 + 0.2 9.1 f 0.1 6.9 0.2 8.8 k 0.2*

25.4 0.3 25.0 + 0.3 24.5 0.4 27.2 + 0.5*

5.0 f 0.0 5.0 f 0.0 18.9 f 0.1 18.9 k 0.1

'There were 20 tadpoles in each group. 'Mean f S.E. *T4-PRL significantly different from T4 group on same lightidark cycle (P < ,001).

106 M.L. WRIGHT ET AL.

period cycles, but the elevation of hormone levels in ram lambs stimulated by long photoperiods was not completely mimicked by a skeleton photope- riod (Schanbacher et al., '85). In the present work, metamorphic rate did not increase over 12L:12D controls either in naturally developing tadpoles or in those constantly immersed in T4 on the asym- metric skeleton photoperiods found in the 12L:3D:lL:8D or 12L:2D:lL:9D cycles. Thus, long photoperiod effects were not observed in the ab- sence of a surge in T4 produced by injection.

However, in tadpoles injected with T4 at 2200 hr, early in the dark, progress was faster on a 12L:3D:lL:8D than on a 12L:12D cycle. The as- sumption was tested that a surge in T4 due to injection near the time of the 1-hr light pulse pro- duced a long photoperiod effect on 12L:3D:lL:8D, making it equivalent to 21 h r of light if the short pulse was the initiator of the photoperiod. The experimental findings did not support this hypoth- esis, since metamorphic rate on a 12L:3D:lL:8D cycle with T4 injection in the early dark before the light pulse was slower than on a 21L:3D cycle with T4 injection in the dark before the long photope- riod. A skeleton photoperiod was not as effective as a single long photoperiod. When the rate of metamorphosis was compared in two groups of tad- poles on 12L:3D:lL:8D cycles, one group receiving T4 injections daily at 0700 hr before the start of the main light phase, and the other at 2200 hr before the beginning of the nocturnal light pulse, progress was faster in the group injected with T4 at 0700 hr. In general, then, metamorphosis was faster when a long light phase rather than a short pulse followed T4 administration. The timing of the T4 injection relative to the pulse and the main light phase seemed to be the important factor.

The altered LD cycles used in the present work might have changed the levels or rhythms of hor- mones such as PRL and corticosteroids that modu- late the action of T4 in metamorphosis (see White and Nicoll, '81). Thus T4 would have been injected into a different hormonal milieu in each case. A change in the amounts of these hormones does not seem likely, since there was no difference in growth and development of control tadpoles, not injected with T4, on 12L:3D:lL:8D and 12L:12D cycles. But the phases of the rhythms of these or other hor- mones implicated in metamorphosis might have been reentrained by the altered LD cycle. Joseph ('74) showed that the time relations between T4 and PRL injections were important in PRL's abil- ity to retard metamorphosis. PRL injected 20 hr after the injection of T4 was most effective. How-

ever, there was little difference between injection of PRL at 0 or 4 hr after T4 or at 8 or 12 hr after T4. If the 12L:3D:lL:8D cycle phase shifted the rhythm of a hormone interacting with T4 by only a few hours, it might account for the lack of effect on tadpoles in spontaneous metamorphosis in the present work. On the other hand, the introduction of a T4 injection into the altered hormonal environ- ment on 12L:3D:lL:8D might have a different ef- fect than on a 12L:12D cycle. Meier ('75) has suggested that metamorphosis could be regulated by changing relationships in the rhythms of secre- tion of various metamorphic hormones. The results presented here provide additional evidence that the effectiveness of T4 in promoting metamorpho- sis is linked to the LD cycle, which might synchro- nize such changes.

Since metamorphic rate was greater when T4 was administered in the dark phase with a light pulse, or on long photoperiod cycles, it was possible that PRL, which antagonizes metamorphosis, was inhibited by light. In two experiments, PRL was injected in the early dark at 2200 hr to groups of tadpoles on 12L:lZD and 12L:3D:lL:8D cycles. Since the extra hour of light on the 12L:3D:lL:8D cycle made no difference to the progress of sponta- neous metamorphosis or to that induced by con- stant T4 immersion when PRL was not injected, any differences observed in these experiments would have been due to the influence of light on PRL utilization. However, there was no difference in the effectiveness of PRL in promoting tail growth of tadpoles in spontaneous metamorphosis or in antagonizing tail resorption induced by con- stant immersion in T4, on the two cycles. Thus the responsiveness of the tissues to PRL was not af- fected by the light pulse. These findings are in accord with previous work with premetamorphic tadpoles (Wright et al., '86) that showed little dif- ference when PRL was injected at various times in a 12L:12D cycle. However, the results do not rule out the possibility that the altered LD cycle af- fected the time of endogenous PRL release.

For the LD cycle to affect metamorphosis, it is necessary that information about the environmen- tal illumination reach the responding organs or tissues. The pineal gland, which secretes the hor- mone melatonin on a rhythmic basis, is believed to integrate photoperiodic input with neuroendocrine functions in amphibians as well as other verte- brates (Korf and Oksche, '86). Treatment of anuran tadpoles with pineal extracts (Addair and Chides- ter, '28) or melatonin (Gutierrez et al., '84; Ed- wards and Pivorun, '86; Delgado et al., '87) or by

PHOTOPERIODIC REGULATION OF METAMORPHIC RATE 107

pinealectomy (%my and Disclos, '70) affected growth and development, although the results var- ied with the species and the experimental conditions.

Melatonin is found in highest concentration in the pineal and plasma during the dark phase of the LD cycle (Underwood, '82). Short light pulses given at night have been shown to depress pineal melatonin levels in hamsters (Reiter et al., '86) and bright light suppressed nighttime melatonin pro- duction in humans (Lewy et al., '87). A model con- structed for the action of bright light in the evening on humans predicted that light would phase delay the melatonin rhythm (as discussed in Arendt and Broadway, '87). Melatonin is generally inhibitory to both thyroid and interrenal function (Johnson, '82) and may also inhibit PRL, depending on the time in the LD cycle that it is given (Blask et al., '81). Thus melatonin could mediate an effect of the LD cycle on hormones that interact with T4 in metamorphosis. Since melatonin may affect the thyroid axis at any level from the brain to the serum proteins that bind T4 (Johnson, '821, how- ever, a "direct" effect of light, acting through the pineal, on thyroid hormone binding, metabolism, or function cannot be ruled out. Future research in this laboratory will include investigation of a pos- sible role of melatonin in mediating the effect of light on the rate of the metamorphic process and perhaps on the endocrine factors that control metamorphosis

ACKNOWLEDGMENTS This research was supported by a Penta Corpo-

ration Grant of Research Corporation. Thanks are due to the National Institute of Arthritis, Metabo- lism and Digestive Diseases, the National Insti- tute of Diabetes and Digestive and Kidney Diseases, and the National Hormone and Pituitary Program of the University of Maryland School of Medicine for gifts of prolactin.

LITERATURE CITED Addair, J., and F.E. Chidester (1928) Pineal and metamorpho-

sis: The influence of pineal feeding upon the rate of meta- morphosis in frogs. Endocrinology, 12:791-796.

Arendt, J., and J. Broadway (1987) Light and melatonin as zeitgebers in man. Chronobiol. Int., 4:273-282.

Bern, H.A., C.S. Nicoll, and R.C. Strohman (1967) Prolactin and tadpole growth. Proc. SOC. Exp. Biol. Med., 126.518-520.

Blask, D.E., C.A. Leadem, and B.A. Richardson (1981) Nutri- tional status, time of day, and pinealectomy: Factors influ- encing the sensitivity of the neuroendocrine-reproductive axis of the rat to melatonin. Horm. Res., 14:104-113.

Delgado, M.J., P. Gutierrez, and M. Alonso-Bedate (1984) Growth response of premetamorphic Rana ridibunda and

Discoglossus pictus tadpoles to melatonin injections and pho- toperiod. Acta Embryol. Morphol. Exp., 5:23-39.

Delgado, M.J., P. Gutierrez, and M. Alonso-Bedate (1987) Mel- atonin and photoperiod alter growth and larval development in Xenopus laeuis tadpoles. Comp. Biochem. Physiol. [A],

Dent, J.N. (1968) Survey of amphibian metamorphosis. In: Metamorphosis-A Problem in Developmental Biology. W. Etkin and L.I. Gilbert (eds). Appleton-Century-Crofts, New York, pp. 271-311.

Disclos, P. (1959) Influence regulatrice de la lumiere sur le developpement de Wards d'dlytus obstetricans Law. soumis a un traitement thyroxinien. C.R. Acad. Sci. Paris, 249:1277- 1279.

Edwards, M.L.O., and E.B. Pivorun (1986) Melatonin modula- tion of metamorphosis in Xenopus laevis. Am. Zool., 26:122 (abstract).

Eichler, V.B., and L.S. Gray (1976) The influence of environ- mental lighting on the growth and prometamorphic devel- opment of larval Rana pipiens. Dev. Growth Differ., 18:177- 182.

Etkin, W.B., and A.G. Gona (1967) Antagonism between pro- lactin and thyroid hormone in amphibian development. J.

Gutierrez, P., M.J. Delgado, and M. Alonso-Bedate (1984) In- fluence of photoperiod and melatonin administration on growth and metamorphosis in Discoglossus pictus larvae. Comp. Biochem. Physiol. [A], 79:255-260.

Guybtant, R. (1964) Actions comparees de la lumiPre et de l'obscurith sur la croissance et la mbtamorphose de thtards de Rana temporaria. Ann. Sci. Univ. Besancon-Zoll. Physiol.,

Johnson, L. (1982) The pineal gland as modulator of the adre- nal and thyroid axes. In: The Pineal Gland, Vol. 111: Extra- Reproductive Effects. R. Reiter (ed). CRC Press, Florida, pp.

Joseph, M.M. (1974) Circadian systems in the hormonal regu- lation of amphibian metamorphosis. In: Chronobiology. L.E. Scheving, F. Halberg, and J. Pauly (eds). Igaku Shoin Ltd.,

Just, J.J., J. Kraus-Just, and D.A. Check (1981) Survey of chordate metamorphosis. In: Metamorphosis-A Problem in Developmental Biology, 2nd ed. L.I. Gilbert and E. Frieden (eds). Plenum Press, New York, pp. 265-326.

Kaltenbach, J.C. (1968) Nature of hormone action in amphib- ian metamorphosis. In: Metamorphosis-A Problem in De- velopmental Biology. w. Etkin and L.I. Gilbert (eds). Appleton-Century-Crofts, New York, pp. 399-441.

Korf, H.W., and A. Oksche (1986) The pineal organ. In: Verte- brate Endocrinology: Fundamentals and Biomedical Impli- cations. P.K. Pang and M.P. Schreibman (eds). Academic Press, New York, pp. 105-145.

Lewy, A.J., R.L. Sack, L.S. Miller, and T.M. Hoban (1987) Antidepressant and circadian phase-shifting effects of light. Science, 235:352-354.

Maier, C.E., and M. Singer (1981) The effect of prolactin on the rate of forelimb regeneration in newts exposed to photo- period extremes. J. Exp. Zool., 215:395-397.

Meier, A.H. (1975) Chronophysiology of prolactin in the lower vertebrates. Am. Zool., 15:905-916.

Pittendrigh, C.S., and S. Daan (1976) A functional analysis of circadian pacemakers in nocturnal rodents I. The stability and lability of spontaneous frequency. J. Cornp. Physiol.,

Pittendrigh, C.S., and D.H. Minis (1964) The entrainment of

86:417-421.

EXP. ZOO^., 165:249-258.

19:77-98.

107-152.

Tokyo, pp. 142-146.

106:223-252.

108 M.L. WRIGHT ET AL. circadian oscillations by light and their role as photoperiodic clocks. Am. Nat., 98.261-294.

Reiter, R.J., B.N. Joshi, T. Heinzeller, and F. Nurnberger (1986) A single 1- or 5-second light pulse at night inhibits hamster pineal melatonin. Endocrinology, 118:1906-1909.

RBmy, C., and P. Disclos (1970) Influence de l’epiphysectomie sur le dBveloppement de la thyroyde et des gonades chez les Wards d‘Alytes obstetricans. C.R. Soc. Biol. (Paris), 164:1989- 1993.

Schanbacher, B.D., and J.D. Crouse (1981) Photoperiodic reg- ulation of growth: A photosensitive phase during the light dark cycle. Am. J. Physiol., 241:El-E5.

Schanbacher, B.D., W. Wer, J.A. Nienaber, and G.L. Hahn (1985) Twenty-four-hour profiles of prolactin and testoster- one in ram lambs exposed to skeleton photoperiods consist- ing of various light pulses. J. Reprod. Fertil., 73:37-43.

Taylor, A.C., and J.J. Kollros (1946) Stages in the normal development of Rana pipiens larvae. Anat. Rec., 94:7-23.

Underwood, H. (1982) The pineal and circadian organization in fish, amphibians, and reptiles. In: The Pineal Gland, Vol. 111: Extra-Reproductive Effects.R.J. Reiter (ed). CRC Press,

Florida, pp. 1-25. White, B.A., and C.S. Nicoll (1981) Hormonal control of am-

phibian metamorphosis. In: Metamorphosis-A Problem in Developmental Biology, 2nd ed. L.I. Gilbert and E. Frieden (eds). Plenum Press, New York, pp. 363-396.

Wright, Sr. M.L. (1984) Effect of time of thyroxine (T4) admin- istration, photoperiod, and daylength on the rate of tadpole metamorphosis. Am. Zool., 24:118. (abstract).

Wright, M.L., E.K. Frim, V.A. Bonak, and C. Baril (1986) Metamorphic rate in Rana pipiens larvae treated with thy- roxine or prolactin at different times in the lightidark cycle. Gen. Comp. Endocrinol., 63.51-61.

Wright, M.L., S.J. Jorey, Y.M. Myers, M.L. Fieldstad, C.M. Paquette, and M.B. Clark (1988) Influence of photoperiod, daylength, and feeding schedule on tadpole growth and de- velopment. Dev. Growth Differ., 30:315-323.

Wright, Sr. M.L., M.A. Majerowski, S.M. Lukas, and P.A. Pike (1979) Effect of prolactin on growth, development, and epi- dermal cell proliferation in the hindlimb of the Ranapipiens tadpole. Gen. Comp. Endocrinol., 39.53-62.