Strain differences in the distribution of arginine-vasopressin- and neuropeptide Y-immunoreactive neurons in the suprachiasmatic nucleus of laboratory rats

BRAIN RESEARCH ELSEVIER

Brain Research 724 (1996) 191-199

Research report

Strain differences in the distribution of arginine-vasopressin- and neuropeptide Y-immunoreactive neurons in the suprachiasmatic nucleus of laboratory rats Franziska Wollnik *, Susanne Bihler Dept. of Biology, University of Konstanz, P.O. Box 5560, M618, D-78434 Konstanz, German>' Accepted 27 February !996

Abstract We have studied the number of arginine-vasopressin (AVP)-immunoreactive (it) somata and the area size of AVP- and neuropeptide Y (NPY)-ir fibers in the suprachiasmatic nuclei (SCN) of three strains of laboratory rats exhibiting a strong unimodal (ACI), a bimodal (BH), and a weak multimodal pattern (LEW) of wheel running activity. In all three strains, AVP-ir somata and fibers were located predominantly in the dorsomedial SCN. Significant strain-differences were found for the area size of AVP-ir fibers as well as for the number and density of AVP-ir somata. The total number of AVP-ir somata was significantly higher in strain ACI (2238 _ 164) than in strains BH (1552_+ 137) and LEW (1426 _+ 110), whereas the mean area of AVP-ir fibers was significantly larger in strain LEW (50779 + 2202 ~ m 2) than in strains ACI (39034 +_ 2095 ~ m 2) and BH (28052 + 1728 /xm2). Consequently, the density of AVP-ir somata was significantly lower in LEW rats, which have a weak multimodal activity pattern, than in BH and ACI rats, which have a bimodal and unimodal activity pattern, respectively. These data suggest that AVP neurons may be part of SCN output pathways controlling circadian activity rhythms. NPY-ir fibers have been identified mainly in the ventral part of the SCN. The mean area of NPY-ir fibers was smallest in BH rats (26 100 _+ 1822 /xm2), which show a rather scattered activity onset, and larger in ACI (29934 + 2468 p~m2) and L E W rats (31 889 + 2728 /xm2), which have rather precise activity onsets. The inbred strains ACI, BH, and LEW may prove to be suitable models to further study distinct neuronal substrates of the SCN functionally correlated with characteristic parameters of circadian rhythms. Keywords: Suprachiasmatic nucleus; Arginine-vasopressin; Neuropeptide Y; Strain difference; Rat

1. I n t r o d u c t i o n The suprachiasmatic nucleus (SCN) of the hypothalamus is considered to be the principal circadian pacemaker in the mammalian brain, driving a wide variety of physiological, endocrine, and behavioral rhythms (for reviews see [26,31,49]). Topographically, the rodent SCN can be separated into two distinct areas [10,13,45,46]: a ventrolateral subdivision receiving direct and indirect visual terminals from the retinohypothalamic [35] and geniculohypothalamic tracts [11,29], and a dorsomedial subdivision from which efferent fibers project to several areas of the hypothalamus [47]. Several neurotransmitters have been identified in the ventrolateral and dorsomedial subdivisions of the SCN [10,46]. The concentration of some of them vary over the course of a day and these changes were found to * Corresponding author. Fax: (49) (7531) 88-3018. 0006-8993/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S 0 0 0 6 - 8 9 9 3 ( 9 6 ) 0 0 3 1 8 - 6

be either endogenously generated or driven by photic cycles [22]. Arginine-vasopressin (AVP), one of the major neuropeptides found in the SCN, is most abundant in the dorsomedial part of the nucleus [10,13,14,45,46]. In vivo, circadian rhythms with peak levels during subjective day and trough levels during subjective night have been demonstrated for AVP release from the SCN into the cerebrospinal fluid [40], and for AVP [22,51] and AVP mRNA [8] content in SCN tissue. In vitro, circadian AVP release has been demonstrated in SCN explants [17], SCN slices [19], and cultured SCN neurons [33]. However, circadian activity rhythms have also been demonstrated in the Brattleboro rat, which due to a genetic AVP-deficiency does not show any AVP-ir neurons in the SCN [34], as well as in the mink, which also has no AVP neurons in the SCN [27]. AVP neurons, therefore, seem to be involved only in the transmission of circadian signals within the

F. Wollnik, S. B i h l e r / Brain Research 724 (1996) 191-199

192

NPY-ir neurons in the SCN of three inbred strains (ACI, BH, and LEW) of rats with distinct circadian locomotor activity patterns [48,50] in order to investigate the relationship between the cellular organization of the SCN and characteristic attributes of circadian function.

SCN and to other hypothalamic areas [14,15,47]. Apparently, they neither control nor constitute part of the pacemaker system within the SCN. Neuropeptide (NPY)-ir fibers are most abundant in the ventral part of the SCN [10,11,13,45,46]. They represent terminals of the geniculohypothalamic tract (GHT), an afferent projection to the SCN originating in the intergeniculate leaflet (IGL), where retinal ganglion cells innervate neuropeptide Y-immunoreactive neurons (for reviews see [29,31]). Under a light/dark (LD) 12:12 h cycle, NPY-ir shows a daily rhythm with two peaks corresponding to the transitions between the two phases of the LD cycle, while under constant darkness NPY-ir shows no or only a weak circadian rhythm [9,42]. Furthermore, NPY levels in the SCN increase steadily for 2 h when rats are exposed to light at the beginning of their subjective day [43], while NPY-ir decreases slightly under continuous light conditions [44]. These results suggest that NPY-ir in the SCN is controlled by photic stimulation. Comparative studies analyzing the number and distribution of AVP-ir neurons within the SCN have demonstrated not only differences between species [10,13,14,46], but also between individual animals of the common vole, Microtus arL,alis, which exhibited a strong, weak, or no expression of circadian rhythmicity [18], as well as between three lines of house mice, Mus domesticus, selected for differences in nest-building behavior [6,7]. The present study analyzed the area and distribution of AVP- and

2. M a t e r i a l s

Male rats of the inbred strains A C I / Z t m (agouti), B H / Z t m (black hooded), and L E W / Z t m (albino), originally obtained from the Central Animal Laboratory of the Hanover Medical School (Germany), were bred and raised in our laboratory under constant environmental conditions (12 h L:12 h D cycle, lights on at 07.00 h, room temperature 22 _+ 1°C). Three animals of each strain were analyzed for AVP-ir and NPY-ir neurons. Groups of three animals each (one of each strain) were killed and processed simultaneously. For AVP analysis, the animals were killed 3 h before lights out, because AVP content in the SCN is highest between CT4 and CT9 [22,51]. For NPY analysis, the animals were killed at 19.00 h, i.e., at lights out, when NPY-levels are rather high [9,43]. At the day of killing the animals were between 6 - 8 (LEW) and 10-12 (ACI, BH) weeks old. The date was selected such that the animals were standardized according to body weight (ACI: 248 _+ 8 g; BH: 305 _+ 15 g; LEW: 262 _+ 8 g) rather than age.

, Olh I . ,hill IIn ,

IIII II

lllh .

LD 12:12 IO

I l l Oil hill III

ll

Idl

nh

I

I.

1111411. , lillllll

Days

,

lhhh

I

,IIHIINh

lkllllllll

I.

20-

I t

o.ys lulll

I

hl * ,

iiii

llilll.h

DD

Ill dl

,ll

",~k'wr,~.

hl

i I I

II I I

I#

,11 I,h

n

Time of Day (h)

iiI

I

4 ,I

do, ,ll

,2 i

,',;:

ol

'

h

,,,,;

,, ,,

20

i Ill .11 h i

,

12

o l-

l

,,Id

II

2O '

I

|! .h

,|.,!

......... ll I

I I ,I

l

:ie

II II Inl l

I

,

' ' 'l' i

10.

Am

aNN

il li

h

. i , . h

. h. . .e .n

IN I I I

,H ~

.

o

'

;.

,h"',, , ,

ii

i~ 'F

I " ' d,,'# 20.

°,

Id.,., i,Mlhilllll

,I, Ih

Ih

2;~

, ,III. el.I lllhl hh Jill

I:l:J:

,I ,I , II h n I I i h II .

1'2 ' Timeof Day (h)

I.I

MI

II I, ilhll I i ilel~l Illl

leo i

li I 0 II ,,h lh III I hi III h ii ,i hll, hi i I

|~

I .il l.l i Ill .I¢. ,~

III II I I I I II II IU n l i i i i lh|l I

h

2'4

,:'. P ,:~.'

Days

l,,;:,

,

.

I II I III

.

2:~

i

2~

Ii'I I lh U i I II lllh i I I ,, l

..... I I I h "iI

t

'

el, 114 , ,. , ,,

,,, i 10 ,..

.,

..

. .i. . .

1'2

i ,i I

,

,.,

,o.

',=

,

,

I.

lid,

,.

'

LEW ,"'

I leo aO I h t OI

I

10

.,

I iii i ii I t io I . . , . .. i . i | el IIII I h I.

h

.................

,kq.,~ l,l,l~It,,,

20-

I.

i I

l IIi I i, i I i i i II I I I i ii hl Ill

,W,,'m'k".',.,, l.,,

10-

ii h Illil i I I,;I

I

12 1-

IIIl.I

I

24

,dl~:l~:L";,,,,

Days

i ,,,, I, I I, ,

III

i .i ied o t i oil i , . , , | , ol , i I| , J . . . i. h, . i ,J i . iii hill , h.m , I

.

12

Ihl II

I =

,,.i ,.I.

I I i I''1 I I ' , I ,I hll,,

i

10

nilh INII hh ,lllllnl n h ilu Ill is x . l l + III H lhlll hlhlillNllh lill. lldl ll~Nle~ll, Iii I~II III ihll ll~lllil

II i

I h |l , i h |1 , h,II

i ,

,,.

hip I H I I I II Iii

I

,HII l,lkldlh

1-

Days

, i

I

1111 ,11111oll I l l . l l l l l h l l u dlllh lhlh III IJllllh ,ol h l , I H ,II I I I , , l u l , l h .llllgO iii IHk hllllll

20

,ii. i ,,i

+

d n l ,lh H I , + ,inl411h , ,I ,i lllh ll~hl lllhh I l l oil nl , h N l n e l l h IIII d h llollhlh iiii hl h h ikh iiii I I , I I I III imlllhl

r i l l I*

=:

BH

ACI l-

Days

and methods

Ill.,

lllib,h I "2

Time o(

I '

2 4'

Day ( h i

Fig. 1. Representative examples of wheel running activity patterns in ACI, BH, and LEW rats maintained under a light/dark cycle (LD 12:12) and under constant darkness (DD). Numbers on the vertical axes denote days of experiment: numbers and bars on the horizontal axes denote daytime hours and light/dark schedule, respectively. Each 24-h period is divided into 72 bins. The height of each bin represents the number of impulses during such a 20-rain interval (20-200 impulses/bin).

F. Wollnik, S. Bihler/Brain Research 724 (1996) 191-199 2.1. Immunocytochemistry For immunocytochemistry, the animals were deeply anesthetized (Ethrane, Abbott) and transcardially perfused for 10 min with phosphate-buffered saline (0.1 M PBS) containing 1% heparin (Sigma), followed by phosphatebuffered fixative (4% paraformaldehyde in 0.l M phosphate buffer, pH 7.4) for 30 min. The brains were removed and postfixed in the same fixative for 24 h at room temperature and then cryoprotected in 30% sucrose at 4°C. The brains were cut at the level of the hypothalamus into 35 /xm frontal sections and collected in 0.1 M PBS containing 0.1% Triton X-100. The sections were washed for 60 min in 3% goat serum in PBS and for 3 × 10 min in PBS before they were incubated in the primary antisera against AVP (1:4000, Chemicon) and NPY (1:1000, Amersham) for 48 h at 4°C. The sections were then processed according to the conventional avidin-biotin peroxidase complex (ABC, Vector Labs) protocol and stained with diaminobenzidine (DAB, Sigma). The sections of one rat of each strain were treated simultaneously under identical conditions. 2.2. Data collection and analysis The area of the SCN containing AVP-ir and NPY-ir fibers was measured using a computerized image analysis system (Softlmaging). A data filter and a threshold level were determined such that only prominently stained cells and fibers were detected. Using this criterion, we achieved a standardized optimal visualization and quantification of the area of AVP-ir and NPY-ir fibers for each of the 2 1 - 2 4 sections analyzed in each brain. In addition to

193

measuring the area, the number of AVP-ir cells was counted in every section containing the SCN. Added together these counts yielded the total number of AVP-ir neurons in the SCN. To calculate the density of AVP-ir neurons in the SCN we divided the total number of AVP-ir cells through the sum of the measured areas. Like Bult et al. [6] and Gerkema et al. [18], we discerned six rostro-caudal levels of the SCN. For each parameter analyzed, SCN sections were assigned to these six levels and estimates were obtained for each animal and each level based on one or the mean of two to three sections. Strain differences were verified by two-way A N O V A using rostro-caudal SCN levels and strains as independent variables, followed by either T u k e y ' s HSD tests or using planned comparisons by contrasts.

3. Results The three rat strains used in this study exhibit distinct circadian locomotor activity patterns [48,50]. ACI rats have a unimodal activity pattern characterized by high activity levels and well-defined onsets and offsets of activity (Fig. 1). BH rats, on the other hand, have a bimodal activity pattern with blurred onsets and offsets and expanded times of activity. Finally, LEW animals have a multimodal activity pattern with three distinct bouts of activity about 4 to 6 h apart and a lower overall activity level. Spectral harmonic analyses revealed characteristic strain differences for the strength of the circadian component and detected additional rhythmic components in the activity pattern of BH (12 h and 6 h) and LEW rats (6 h, 4.8 h, and 4 h). Strain differences were also found for the

Table 1 Circadian rhythm parameters of three inbred strains of rats (mean -+ S.E.M.) Strain LD 12:12 + Activity level (impulsesper day) F(2,27) = 19.62, P < 0.001 Duration of activity (min per day) F(2,27) = 25.86, P < 0.001 Onset of activity(min before lights out) F(2,24) = 33.21, P < 0.001 Offset of activity (min before lights on) F(2,24) = 140.46, P < 0.001 DD * TDD(h) F(2,45) = 20.8, P < 0.001 Activity level (impulsesper cycle) F(2,45) = 40.5, P < 0.001 Duration of activity(min per cycle) F(2,45) = 52.33, P < 0.001

ACI

BH

LEW

1393 ± 591 a

799 _+271 b

314 _+ 154 ~

334±66a

143+40 b

336_+91a 15 _+24 a

25 + 17 ~

120--+42 b

30 -+ 66 a

-- 50 -+ 35 b

360 ± 60 c

24.37 + 0.05 a 2043 _+340 413 _+41 a

a

24.25 + 0.05 b

24.05 + 0.05 ~

862 + 126 b

608 _+98 c

326 ± 31 b

201 _+ 15 c

Differences in circadianrhythm parameters between strains were tested for significanceby a one-wayanalysisof variance(ANOVA), followed by Tukey's HSD tests. Means with different superscripts are significantlydifferent from each other at least at the P __<0.05 level. Data from [48]; * data from [50].

F. Wollnik, S. Bihler / Brain Research 724 (1996) 191-199

194

ACI

BH

LEW

A AVP

B NPY

Fig. 2, Photomicrographs illustrating the differences in the density and area size of AVP-ir (A) and NPY-ir (B) neurons in the medial portion (levels 3 - 4 ) of the SCN of ACI, BH, and L E W rats. ACI rats had significantly larger numbers of AVP-ir somata than BH and L E W rats. AVP-ir neurons were located predominantly in the dorsomedial subdivision of the SCN, while NPY-ir fibers were located exclusively in the ventral subdivision of the SCN. The area containing NPY-ir fibers was significantly smaller in BH rats than in ACI and BH rats. Bar 100 /xm.

overall level, the duration, and the onset and offset of activity (Table 1). AVP-ir cells were found in the SCN of all three strains (Fig. 2A). Both somata and fibers were located mostly in the dorsomedial subdivision of the SCN, although some

AVP-ir fibers could also be found in the ventrolateral subdivision. Comparing the number of AVP-ir somata at six rostro-caudal levels of the SCN (Fig. 3A, Table 2), two-way ANOVA revealed significant differences between strains (F(2,176) = 15.18, P < 0.001), levels of the SCN

Table 2 Number of AVP-ir somata (means ± S.E.M.) at six rostro-caudal levels of the SCN assessed in three inbred strains of laboratory rats (ACI, BH, LEW) Strain

n

ACI

3

BH

3

LEW

3

Number of AVP-ir somata at six rostro-caudal levels of the SCN (1 = rostral, 6 = caudal) 1

2

3

4

5

6

44.0 +7.2 51.7 + 11.7 32.7 +6.8

73.7 ± 18.9 72.9 ± 18.1 67. I ± 10.8

148.7 + 11.4 112.3 + 15.7 103.5 +9.4

160.2 + 13.1 99.2 _+ 18.4 111.2 ± 12.7

136.8 ± 13.1 68.7 _+ 10.7 50.5 +_ I 1.0

54.6 _+ 11.3 30.5 ±6.0 47.6 _+6.9

Total number

2238.0 + 163.7 1552.0 + 137.4 1425.7 + 109.6

Two-way ANOVA: 15.18 Fs,rains(2,176) - 15.18, P _< 0.001 FLe,.eL~(5,176) = 28.62, P < 0.001 Fst,.mn~,t.evel~(1,176) = 30.14, P < 0.001 '~ Statistical comparisons between groups: a ACI - BH + L E W BH - L E W

r

Differences between strains were tested for significance by a two-way analysis of variance (ANOVA) using strains and rostro-caudal levels as independent variables, followed by either Tukey's HSD test or planned comparisons by contrasts. The lower part of the table summarizes the statistical comparisons between strains ( - , no significance; * P _< 0.05; * * * P < 0.01). a Planned comparison ACI vs. BH + LEW.

F. Wollnik, S. Bihler/Brain Research 724 (1996) 191-199

(F(5,176) = 28.62, P < 0.001), and strain-level interactions (F(10,176) = 2.63, P < 0.01). With the exception of the first level, ACI rats had the largest number of AVP-ir somata throughout the rostro-caudal extension of the SCN. The mean total numbers of AVP-ir somata in the SCN of the three strains were 2238.0_+ 163.7 (ACI), 1552.0+ 137.4 (BH), and 1425.7 _+ 109.6 (LEW), revealing a 1.6fold difference between strain ACI and the other two strains (Fig. 4A). Two-way ANOVA of the area containing AVP-ir fibers revealed significant differences between strains (F(2,158) = 57.12, P < 0.001), levels of the SCN (F(5,158) = 19.44, P < 0 . 0 0 1 ) , and strain-level interactions (F(10,158)= 2.19, P < 0.05, Table 3). As expected, the area was largest in the medial part of the SCN. In contrast to the number of AVP-ir neurons, which was highest in the ACI strain, the

A

195

total

number of

~00-

AVP-ir c o m e t s

Qtt

25002000150010005000

B

ACI

r

i

BH

LEW

mean area of AVP-Ir fibers

607 r^E

'

mmm

r--

eem

i

[xl0Spm2] ' 1

IINNN 1

ACI

A

BH

LEW

n u m b e r of AVP-Ir neurone

200-

~--

C

ACI

•

BH

eml

5-

1,50-

[xl0"=pm2]

density of AVP-Ir s o m s t e

6-

~ ~r~.

I~IIW

4100-

32-

50-

10

I

I

0

I

ACI

D

B

a r e a of A V P - i r fibers

80-

•

ACI

m'

BH

[xlOalAm2]

I

BH

I

mean area of NPY-ir fibers

LEW [xl03pm;]

60-

50T

tllltt

40-

60-

3020-

40-

102O

C

a r e a of N P Y - I r fibers

~

60

ACI

_

•

40-_

*

[xlOSltm=]

BH

i

LEW

. L

BH

~

20-

0

I

Fig. 4. M e a n s _+ S.E.M. calculated for (A) the total n u m b e r of A V P - i r somata, (B) the area containing A V P - i r fibers, (C) the density of A V P - i r somata, a n d (D) the area containing NPY-ir fibers in the S C N o f ACI, BH, and L E W rats. Significant differences: * P _< 0.05, * * * P < 0.01.

0

C

ACI

I

I

rostral medial caudal indexof frontalsectionsthroughthe SCN Fig. 3. Measurements (means-S.EM.) obtained for (A) the number of A V P - i r somata, (B) the area c o n t a i n i n g A V P - i r fibers, and (C) the area c o n t a i n i n g N P Y - i r fibers in the S C N o f ACI, BH, and L E W rats at six different levels f r o m rostral to caudal. Significant differences: * P _< 0.05, • ** P _< 0.01.

mean area of AVP-ir fibers was significantly larger in LEW rats (50779-t-2202 /xm 2) than in ACI (39034_+ 2095 /xm 2) and BH rats (28052 _+ 1728 /xm2), resulting in an AVP-ir neuron density 2.3-times higher in ACI than in LEW rats (Fig. 4C). NPY-ir fibers were located predominantly in the ventral subdivision of the SCN in all three strains (Fig. 2B). Comparing the area of NPY-ir fibers at six rostro-caudal levels of the SCN (Fig. 3C, Table 4), two-way ANOVA revealed significant differences only between strains (F(2,144)= 8.38, P < 0 . 0 0 1 ) and levels of the SCN (F(5,144) = 68.55, P < 0.001), while strain-level interac-

F. Wollnik, S. Bihler / Brain Research 724 (1996) 191-199

196

Table 3 Area containing AVP-ir fibers (means ± S.E.M.) at six rostro-caudal levels of the SCN assessed in three inbred strains of laboratory rats (ACI, BH, LEW) Strain

Area size of AVP-ir fibers ( / x m 2) at six rostro-caudal levels of the SCN (1 = rostral, 6 = caudal) Mean area (/+m e)

n

ACI

3

BH

3

LEW

3

1

2

3

4

5

6

35760 ± 3563 19083 ± 3381 35704 ± 4728

47305 ± 5135 31416 ± 3138 51761 ± 2299

50909 ±4978 38234 ± 3909 63146 + 4350

48797 ± 3378 37639 ± 5734 67012 + 4939

39100 ± 3518 28136 ± 3630 62581 ± 2836

20639 ± 3122 21498 + 2855 45485 ± 2676

39034 ± 2095 28052 ± 1728 50779 ± 2202

Two-way ANOVA: Fstrains(2,158) = 57.12, P < 0.001 FLe,el~(5,158) = 19.44, P < 0.001 Fstr,in~. Levels(10,158) = 2.19, P < 0.05 Statistical comparisons between groups: ACI-BH * BH-LEW ACI-LEW -

* -

For methods and notations, see Table 2.

tions proved significant only when BH rats were compared with ACI and L E W rats combined ( F ( 1 , 1 4 4 ) = 12.97, P < 0.001). With the exception of the last level, BH rats had the smallest area of NPY-ir fibers throughout the rostro-caudal extension of the SCN. The mean areas of NPY-ir fibers in the three strains were 29 934 _+ 2468 p,m 2 (ACI), 26 100 ___ 1822 /xm 2 (BH), and 31 899 _+ 2728 /xm 2 (LEW), revealing a 1.3-fold difference between strains BH and L E W (Fig. 4D).

4. Discussion The present study found significant differences in the distribution of AVP-ir and NPY-ir neurons in the SCN of

the inbred rat strains ACI, BH and LEW. The first difference concerned the number and density of AVP-ir somata. They were highest in ACI rats, which have a strong unimodal activity pattern with a high overall level of activity, intermediate in BH rats, which have a weak bimodal activity pattern, and lowest in LEW rats, which have a rather weak multimodal activity pattern and a low overall level of activity. The second difference concerned the area size of SCN containing NPY-ir fibers. It was significantly smaller in BH rats than in animals of the ACI and LEW strain. Our values of area size of NPY-ir fibers are in general agreement with measurements reported in a similar quantitative study on the SCN of ordinary Sprague-Dawley rats [9]. This study demonstrated a diurnal rhythm for the area

Table 4 Area containing NPY-ir fibers (means ± S.E.M.) at six rostro-caudal levels of the SCN assessed in three inbred strains of laboratory rats (ACI, BH, LEW) Strain

n

ACI

3

BH

3

LEW

3

Area size of NPY-ir fibers ( / x m 2) at six rostro-caudal levels of the SCN (1 = rostral, 6 = caudal) 1

2

3

4

5

6

4121 ± 728 6263 ± 1321 4534 ± 592

20149 ± 2072 20177 ± 1473 20900 ± 2524

38811 ± 2798 31839 ± 739 38139 ±4969

50872 ± 2248 35941 ± 2015 54518 + 5866

43038 ± 3475 36510 + 2333 48995 ±5485

20524 ± 2804 21335 ± 5246 31168 ± 3021

Two-way ANOVA: Fstrain~(2,144) = 8.38, P < 0.001 FLeve~(5,144) = 68.55, P < 0.001 Fstrain~. Levels(l,144) = 12.97, P < 0.001 a Statistical comparisons between groups: a BH-ACI + L E W ACI-LEW For methods and notations, see Table 2. a Planned comparison BH vs. ACI + LEW.

Mean area ( / x m 2)

29934 ± 2468 26100 ± 1822 31899 ± 2728

F. Wollnik, S. Bihler/Brain Research 724 (1996) 191-199

size of NPY-ir, with maximum sizes occurring at the transition from light to dark. In the present study, all animals were killed at this transition time of the lighting regime. The mean NPY-ir area sizes determined for ACI and LEW rats were similar to the area size found in Sprague-Dawley rats, while the area size measured in BH rats was significantly smaller. The NPY-ir fibers of the SCN are known to originate in the IGL, which receives photic information from the retina [29,31]. Although neither the IGL nor the GHT are necessary prerequisites for the photic entrainment of the circadian pacemaker system [16,20,21,37], the following findings suggest their involvement in mediating both photic and non-photic information to the SCN. First, IGL lesions alter the rate and phase angle of entrainment [24,36,37] and reduce the period lengthening effect of constant light [20,37] and the occurrence of splitting under constant light [21]. Second, electrical stimulation of the GHT [39] and NPY injections into the SCN [1] induce phase shifts of free-running activity rhythms very similar to those observed after dark pulses [5] and induced activity [32]. Third, IGL lesions abolish the phase shifting effect of non-photic stimuli such as benzodiazepine administration [25] and induced activity by novel wheel stimulus [23]. Fourth, treatment with antiserum to NPY enhances lightinduced phase advances [3] and attenuates behaviorally induced phase shifts [4]. Due to the demonstrated strain differences in the area size of NPY-ir fibers in the SCN the three rat strains ACI, BH and LEW provide a useful model for the study of the physiological function of NPY projections to the SCN. We plan to further analyze these strains for possible differences in, for example, the rate of reentrainment and the effect of non-photic stimuli on the circadian pacemaker system. In addition to strain differences found for the size of the area containing NPY-ir fibers, the present study also demonstrated differences in the number and density of AVP-ir neurons. In general, our results on the distribution of AVP-ir somata and fibers are in good agreement with previous studies on small laboratory animals [7,10,13,18,45,46]. In all three strains, AVP-ir cells were clustered predominantly in the dorsomedial part of the SCN. Rats of the ACI strain had the highest number of AVP-ir somata within the SCN. Furthermore, only ACI rats showed additional AVP-ir somata in the ventrolateral division of the SCN, similar to the distribution normally reported for rats [45] and mice [7,13]. In BH and LEW rats, AVP-ir somata were located exclusively in the dorsomedial subdivision of the SCN, a distribution normally found in golden hamsters [10]. The functional role of AVP in the SCN is still not known, but a number of studies suggest that AVP-ir neurons may be involved in the efferent control of circadian activity rhythms. One possible function of AVP is that of a neurohormone transmitting temporal information from the SCN to the cerebrospinal fluid (CSF) and thus to

197

other brain regions [41]. The other possible function of AVP is that of a neurotransmitter in neuronal output pathways from the SCN to other hypothalamic areas [14,15,47]. In the light of this alternative, it is rather tempting to assume that the distinct circadian activity patterns of the three inbred strains are causally related to the observed strain differences in the number and density of AVP-ir neurons in the SCN. The possible correlation between AVP-ir neurons in the SCN and circadian clock function has already been investigated in three recent studies [6,7,18]. House mice selected for differences in nest-building behavior showed almost identical differences in the number and distribution of AVP-ir neurons in the SCN and associated differences in parameters of circadian rhythmicity [6,7]. The number of AVP-ir neurons was significantly larger in mice with a strong unimodal activity pattern and a high level of activity than in mice with a weak multimodal activity pattern. Furthermore, the activity pattern of mice with a relatively small number of AVP-ir neurons resembles conspicuously the multimodal activity pattern of our LEW rats, while the activity pattern of mice with a relatively large number of AVP-ir neurons is almost identical to the unimodal activity pattern of our ACI rats. Similar to the present study these studies demonstrated a positive correlation between the number of AVP-ir neurons and the strength of the circadian activity rhythm. These findings are, however, in conflict with the results of a study that investigated the number and density of AVP-ir neurons in the SCN of voles with a great interindividual variability in the expression of circadian activity rhythms [18]. Here, the number of AVP-ir neurons was largest in voles with no circadian activity rhythm, intermediate in voles with a weak rhythm, and smallest in voles with a strong rhythm, suggesting a negative correlation between the number of AVP-ir neurons and the strength of circadian rhythmicity. These divergent results could, of course, be explained by species differences, but it has also been speculated that the large numbers of AVP-ir neurons found in voles with no or only a weak circadian rhythm may be the result of reduced AVP release rather than increased AVP synthesis [18]. This hypothesis is supported by the observation that in contrast to other species, the number of AVP-ir cells in voles did not vary over the day. Also conflicting with our hypothesis of an AVP-correlated expression of circadian rhythms is the demonstration of circadian rhythms in Brattleboro rats [34], which suffer from a genetic AVP deficiency, and minks [27], which have no AVP-ir neurons in the SCN. Although these examples show that AVP cannot be indispensable for the generation of circadian rhythms, it may still play a functional role in the circadian organization of other species or genetically intact animals. Several studies have demonstrated AVP-ir neurons to be part of neuronal output pathways from the SCN to other hypothalamic areas [14,15,47]. One such pathway runs via

198

F. Wollnik, S. Bihler/Brain Research 724 (1996) 191-199

the paraventricular nucleus and the superior cervical ganglion to the pineal gland, where norepinephrine released from post-ganglionic sympathetic nerve endings controls the circadian rhythm of melatonin secretion [38]. Since the output of nocturnal melatonin is generally proportional to the length of night, melatonin is the carrier of important temporal information about the photoperiod. In mammals, this information is used not only to control seasonal rhythms such as the rhythm of reproduction, but according to a number of recent studies also to synchronize and couple circadian rhythms (reviewed in [2,12]). The circadian system of adult rats, for example, can be entrained by daily injections of melatonin, provided the time of injection coincides with the animal's activity onset. Injections of melatonin can also accelerate the reentrainment of activity rhythms and synchronize disrupted components of a circadian rhythm under constant light conditions. The mammalian SCN contains a considerable amount of melatonin receptors [30], and it is presently assumed that melatonin affects the coupling of some or all circadian oscillators within the SCN [12]. Based on these and the present findings we are inclined to assume a causal relationship between the quantity of AVP-ir neurons, melatonin secretion from the pineal, and the dissociation of circadian activity patterns. Under this hypothesis, the bimodal and multimodal activity patterns of BH and LEW rats, respectively, are the result of a desynchronization of multiple circadian oscillators caused by the reduction or complete loss of melatonin feedback on the SCN. The reduction of melatonin, in turn, is due to the weaker signal to the pineal produced by a smaller number of AVP-ir neurons. The hypothesis is further supported by the finding that certain pharmacological substances such as moclobemide and desipramine, which affect the noradrenergic transmitter system and thereby activate the pineal, increase the activity of BH and LEW rats and induce a unimodal activity pattern in these animals [50]. Further studies, however, are necessary to test this hypothesis. First, we need to verify the causal relationship between the density of AVP-ir neurons in the SCN and the amplitude and duration of melatonin secretion in the pineal, or the density of melatonin receptors in the SCN. Second, we need to verify the causal relationship between the number of AVP-ir neurons in the SCN and the strength of behavioral circadian rhythmicity. The latter could be tested, for example, by specifically altering or blocking AVP neuronal function using antisense application [28], or by employing methods of quantitative genetic analysis on crosses of the three inbred strains. In conclusion, the data presented in this paper support the hypothesis that differences in the cellular organization of the SCN are reflected in various aspects of circadian behavior. The defined genetic background of commonly available inbred strains of laboratory rats with distinct activity patterns provides a powerful tool to further investi-

gate the neuronal mechanisms and physiological organization of circadian rhythms in mammals.

Acknowledgements This study was supported by a grant from the Deutsche Forschungsgemeinschaft (Wo 354/3-3). The authors thank Andreas Herrmann for reviewing the manuscript.

References [1] Albers, H.E. and Ferris, C.F., Neuropeptide Y: role in light-dark entrainmentof hamster circadianrhythms, Neurosci. Lett., 50 (1984) 163-168. [2] Armstrong, S.M., Melatonin and circadian control in mammals, Experientia, 45 (1989) 932-93. [3] Biello, S.M., Enhanced photic phase shifting alter treatment with antiserum to neuropeptide Y, Brain Res., 673 (1995) 25-29. [4] Biello, S.M., Janik, D. and Mrosovsky, N., Neuropeptide Y and behaviorally induced phase shifts, Neuroscience, (1994) 273-279. [5] Boulos, Z. and Rusak, B., Circadianphase response curves for dark pulses in the hamster, J. Comp. Physiol., 146 (1982) 411-417. [6] Bult, A., Hiestand, L., Van der Zee, E.A. and Lynch, C.B., Circadian rhythms differ between selected mouse lines: a model to study the role of vasopressin neurons in the suprachiasmaticnuclei, Brain Res. Bull., 32 (1993) 623-627. [7] Bult, A., Van der Zee, E.A., Compaan, J.C. and Lynch, C.B_ Differences in the number of arginine-vasopressin-immunoreactive neurons exist in the suprachiasmatic nuclei of house mice selected for differences in nest-building behavior, Brain Res.. 578 (1992) 335 338. [8] Cagampang, F.R.A., Yang, J., Nakayama, Y., Fukuhara, C. and Inouye, S.-I.T., Circadian variationof arginine-vasopressinmessenger RNA in the rat suprachiasmatic nucleus, Mol. Brain Res., 24 (1994) 179-184. [9] Calz~, L., Giardino, L., Zanni, M.. Velardo, A., Parchi. P. and Marrama. P., Daily changes of neuropeptide Y-like immunoreactivity in the suprachiasmaticnucleusof the rat, Regul. Pept., 27 (1990) 127 137. [10] Card, J.P. and Moore, R.Y., The suprachiasmatic nucleus of the golden hamster: immunohistochemicalanalysis of cell and fiber distribution, Neuroscience, 13 (1984) 415-431. [11] Card. J.P. and Moore, R.Y., The organization of visual circuits influencingthe circadian activity of the suprachiasmatic nucleus. In D.C. Klein, R.Y. Moore and S.M. Reppert (Eds.), Suprachiasmatic Nucleus." The Mind's Clock, Oxford University Press, New York, 1991, pp. 51-76. [12] Cassone, V.M., Melatonin and suprachiasmatic nucleus. In D.C. Klein, R.Y. Moore and S.M. Reppert (Eds.), Suprachiasmatic Nucleus: The Mind's Clock, Oxford UniversityPress. New York, 1991, pp. 309 323. [13] Cassone, V.M., Speh, J.C., Card, J.P. and Moore, R.Y., Comparative anatomy of the mammalianhypothalamic suprachiasmatic nucleus, J. Biol. Rhythms, 3 (1988) 71-91. [14] Castel, M., Feinstein,N., Cohen, S. and Harari, N., Vasopressinergic innervationof the mouse suprachiasmaticnucleus: an immuno-electron microscopic analysis, J. Comp. Neurol., 298 (1990) 172 187. [15] Daikoku, S., Hisano, S. and Kagotani, Y., Neuronal associations in the rat suprachiasmatic nucleus demonstrated by immunoelectron microscopy, J. Comp. Neurol.. 325 (1992) 559-57 I.

F. Wollnik, S. Bihler/Brain Research 724 (1996) 191-199 [16] Dark, J.G. and Asdourian, D., Entrainment of the rat's activity rhythm by cyclic light following lateral geniculate nucleus lesions, Physiol. Behav., 15 (1975) 295-301. [17] Earnest, D.J. and Sladek, C.D., Circadian rhythms of vasopressin release from individual rat suprachiasmatic explants in vitro, Brain Res., 382 (1986) 129-133. [18] Gerkema, M.P., Van der Zee, E.A. and Feitsma, L.E., Expression of circadian rhythmicity correlates with the number of argininevasopressin-immunoreactive cells in the suprachiasmatic nucelus of common voles, Microtus arL,alis, Brain Res., 639 (1994) 93-101. [19] Gillette, M.U. and Reppert, S.M., The hypothalamic suprachiasmatic nuclei: circadian patterns of vasopressin secretion and neuronal activity in vitro, Brain Res. Bull., 19 (1987) 135-139. [20] Harrington, M.E. and Rusak, B., Lesions of the thalamic intergeniculate leaflet alter hamster circadian rhythms, J. Biol. Rhythms, 1 (1986) 309-325. [21] Harrington, M.E. and Rusak, B., Ablation of the geniculo-hypothalamic tract alters circadian activity rhythms of hamsters housed under constant light, Physiol. Behac., 42 (1988) 183-189. [22] Inouye, S.-I.T., Shinohara, K., Tominaga, K., Takeuchi, J., Nagasaki, H., Isobe, Y., Fukuhara, C., Otori, Y., Yang, J., Cagampang, F.R.A., Yamazaki, S. and Tokumasu, A., Circadian rhythms in peptides and their precursor messenger RNAs in the suprachiasmatic nucleus. In H. Nakagawa, Y. Oomura and K. Nagai (Eds.) New Functional Aspects of the Suprachiasmatic Nucleus of the Hypothalamus, John Libbey, 1993, pp. 219-233. [23] Janik, D. and Mrosovsky, N., Intergeniculate leaflet lesions and behaviorally-induced shifts of circadian rhythms, Brain Res,, 651 (1994) 174-182. [24] Johnson, R.F., Moore, R.Y. and Morin, L.P., Lateral geniculate lesions alter circadian activity rhythms in the hamster, Brain Res. Bull., 22 (1989) 411-422. [25] Johnson, R.F., Smale, L., Moore, R.Y. and Morin, L.P., Lateral geniculate lesions block circadian phase-shift responses to a benzodiazepine, Proc. Natl. Acad. Sci. USA, 85 (1988) 5301-5304. [26] Klein, D.C., Moore, R.Y. and Reppert, S.M. (Eds.), Suprachiasmatic Nucleus: The Mind's Clock, Oxford University Press, New York, 1991, 467 pp. [27] Larsen, P.J. and Mikkelsen, J.D., The suprachiasmatic nucleus of the mink (Mustela L,ison): apparent absence of vasopressin-immunoreactive neurons, Cell. Tissue Res., 273 (1993) 239-247. [28] Meeker, R., LeGrand, G., Ramirez, J., Smith, T. and Shih, Y.-H., Antisense vasopressin oligonucleotides: uptake, turnover, distribution, toxicity and behavioral effects, J. Neuroendocrinol., 7 (1995) 419-428, [29] Moore, R.Y. and Card, J.P., Intergeniculate leaflet: an anatomically and functionally distinct subdivision of the lateral geniculate complex, J. Comp. Neurol., 344 (1994) 403-430. [30] Morgan, P.J. and Williams, L.M., Central melatonin receptors: implications for a mode of action, Experientia, 45 (1989) 955-965. [31] Morin, L.P., The circadian visual system, Brain Res. Rev., 647 (1994) 102-127. [32] Mrosovsky, N., Phase response curves for social entrainment, J. Comp. Physiol., 162 (1988) 35-46. [33] Murakami, N., Takamure, M., Takahashi, K., Utunomiya, K., Kuroda, H. and Etoh, T., Long-term cultured neurons from rat suprachiasmatic nucleus retain the capacity for circadian oscillation of vasopressin release, Brain Res., 545 (1991) 347-350.

199

[34] Peterson, G.M., Watkins, W.B. and Moore, R.Y., The suprachiasmatic hypothalamic nuclei of the rat. VI. Vasopressin neurons and circadian rhythmicity, Beha~'. Neural Biol., 29 (1980) 236-245. [35] Pickard, G.E., The afferent connections of the suprachiasmatic nucleus of the golden hamster with emphasis on the retinohypothalamic projection, J. Comp. Neurol., 211 (1982) 65-83. [36] Pickard, G.E., Entrainment of the circadian rhythm of wheel-running activity is phase shifted by ablation of the intergeniculate leaflet, Brain Res., 494 (1989) 151-154. [37] Pickard, G.E., Ralph, M.R. and Menaker, M., The intergeniculate leaflet partially mediates effect of light on circadian rhythms, J. Biol. Rhythms, 2 (1987) 35-56. [38] Reiter, R.J., The melatonin rhythm: both a clock and a calendar, Experientia, 49 (1993) 654-664. [39] Rusak, B., Meijer, J.H. and Harrington, M.E., Hamster circadian rhythms are phase-shifted by electrical stimulation of the geniculohypothalamic tract, Brain Res., 493 (1989) 283-291. [40] Schwartz, W.J., Coleman, R.J. and Reppert, S.M., A daily vasopressin rhythm in rat cerebrospinal fluid, Brain Res., 263 (1983) 105-112. [41] Schwartz, W.J. and Reppert, S.M., Neural regulation of the circadian vasopressin rhythm in cerebrospinal fluid: a pre-eminent role for the suprachiasmatic nuclei, J. Neurosci., 5 (1985) 2771-2778. [42] Shinohara, K., Tominaga, K., Fukuhara, C., Otori, Y. and Inouye, S.-I.T., Processing of photic information within the intergeniculate leaflet of the lateral geniculate body: assessed by neuropeptide Y immunoreactivity in the suprachiasmatic nucleus of rats, Neuroscience, 56 (1993) 813-822. [43] Shinohara, K., Tominaga, K., Isobe, Y. and Inouye, S.-I.T., Photic regulation of peptides located in the ventrolateral subdivision of the suprachiasmatic nucleus of the rat: Daily variations of vasoactive intestinal polypeptide, gastrin-releasing peptide, and neuropeptide Y, J. Neurosci., 13 (1993) 793-800. [44] Takatsuji, K. and Tohyama, M., Differential effects on substance P immunoreactivity in the rat suprachiasmatic and olivary pretectal nuclei, NeuroReport, 4 (1993) 647-650. [45] van den Pol, A.N., The suprachiasmatic nucleus: morphological and cytochemical substrates for cellular interaction. In D.C. Klein, R.Y. Moore, S.M. Reppert (Eds.), Suprachiasmatic Nucleus: The Mind's Clock, Oxford University Press, New York, 1991, pp. 17-50. [46] van den Pol, A.N. and Tsujimoto, K.L., Neurotransmitters of the hypothalamic suprachiasmatic nucleus: Immunocytochemical analysis of 25 neuronal antigens, Neuroscience, 15 (1985) 1049-1086. [47] Watts, A.G., The efferent projections of the suprachiasmatic nucleus: anatomical insights into the control of circadian rhythms. In D.C. Klein, R.Y. Moore, S.M. Reppert (Eds.), Suprachiasmatic Nucleus: The Mind's Clock, Oxford University Press, New York, 1991, pp. 77-106. [48] Wollnik, F., Strain differences in the pattern and intensity of wheel running activity in laboratory rats, Experientia, 47 (1991) 593-598. [49] Wollnik, F., Neural control of circadian rhythms in mammals, Verh. Dtsch. Zool. Ges., 85.2 (1992) 231-246. [50] Wollnik, F., Effects of chronic administration and withdrawal of antidepressant agents on circadian activity rhythms in rats, Pharmacol. Biochem. BehaL'., 43 (1992) 549-561. [51] Yamase, K., Takahashi, S., Nomura, K., Haruta, K. and Kawashima, S., Circadian changes in arginine vasopressin level in the suprachiasmatic nuclei in the rat, Neurosci. Lett., 130 (1991) 255-258.