MSU LIBRARIES ——_ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wi11 be charged if book is returned after the date stamped be10w. (3-7}! It” .14 ‘l..-' ' u a $41 1,1 NEURAL CONTROL OF CIRCADIAN RHYTHMS By Michael Howard Brown A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1985 ACKNOWLEDGEMENTS I would like to thank Drs. Neil R. Carlson and Friedrich K. Stephan for providing computer software and advice on computer interfacing. I am also indebted to Dr. Antonio A. Nunez for help with the research and advice on the manuscript and to Drs. Lynwood G. Clemens and Ralph Levine for serving as committee members. This research was funded in part by NIMH grant NH 37877 awarded to Antonio A. Nunez. ii TABLE OF CONTENTS Page LIST OF TABLES ......................................... iv LIST OF FIGURES ........................................ v LIST OF ABBREVIATIONS .................................. vii INTRODUCTION ............................... ............ 1 EXPERIMENT 1 Method .. ...... . ..... ....... ..... .............. ...... 5 Results ............................................. 12 Discussion .......................................... 2A EXPERIMENT 2 Method .............................................. 26 Results ............................................. 28 GENERAL DISCUSSION ..................................... A2 LIST OF REFERENCES ..................................... A7 iii LIST OF TABLES Page TABLE 1: POSTSURGICAL LEVELS OF ACTIVITY AND DRINKING EXPRESSED AS PER CENT OF BASELINE. ................ 14 TABLE 2: RESULTS OF CHI-SQUARE PERIODOGRAM ANALYSIS OF SELECTED '10 DAY BLOCKS OF ACTIVITY AND DRINKING RECORDS. 0.0.00000...OOOOOOIOOOCCOCOOOOO0.0.0.0.... 15 TABLE 3: SEMINAL VESSICLE WEIGHTS (EXPRESSED AS PER CENT OF BODY WEIGHT) AND POSTSURGICAL LEVELS OF ACTIVITY AND DRINKING (EXPRESSED AS PER CENT OF BASELINE). ........................................ 29 TABLE 4: RESULTS OF CHI-SQUARE PERIODOGRAM ANALYSIS OF SELECTED 10 DAY BLOCKS OF ACTIVITY AND DRINKING RECORDS OF BLINDED RATS. .......................... 30 iv LIST OF FIGURES Page Figure 1: Representative doubleplotted event records and periodograms for data sets containing: A, high amplitude rhythms with low amplitude "noise" components; B, low or zero amplitude rhythms; and C, multiple circadian periodicities. .............. 9 Figure 2: Doubleplotted activity and drinking records for a sham-operated control rat (arrows on these and subsequent event records indicate day of surgery). ......................................... 16 Figure 3: Photomicrographic representation of the location of a parasagittal knife cut that damaged the Optic chiasm and activity and drinking records showing an increase in period length of the rhythms after surgery. ............................ 18 Figure A: Photomicrograph of the glial scar resulting from a horizontal cut that damaged the SCN and activity and drinking records showing a loss of rhythmicity after surgery. ........................ 20 Figure 5: Schematic representation of the location of combined horizontal and bilateral parasagittal cuts and activity and drinking records showing that neither cut abolished the rhythms. ........... 22 Figure 6: Photomicrograph showing the typical location of parasagittal cuts in blinded animals and activity and drinking records showing that this cut did not abolish the rhythms or change the period length of the rhythms. ..................... 32 Figure 7: Photomicrograph depicting the scar tissue that resulted from a cut that damaged the ventral part of the SCN and activity and drinking records showing a decrease in period length of the rhythms after surgery. .................................... 35 Figure 8: Photomicrographic representation of the location of a horizontal cut that damaged the dorsal part of the SCN and event records showing that after surgery the activity rhythm persisted while the drinking rhythm was temporarily abolished. ........................................ 37 Figure 9: Photomicrograph showing the location of a horizontal cut that did not damage the SCN and activity and drinking records showing that neither sham surgery (indicated by first arrow) nor the horizontal cut (indicated by second arrow) abolished the rhythms. ............................ 39 vi LIST OF ABBREVIATIONS OC optic chiasm PVN paraventricular nucleus of the hypothalamus SCN suprachiasmatic nucleus SON supraoptic nucleus 3V third ventricle 1 period length arrows on the photomicrographs indicate glial scar tissue arrows on the event records indicate day of surgery calibration bars on the photomicrographs represent 100 pm vii ABSTRACT NEURAL CONTROL OF CIRCADIAN RHYTHMS By Michael Howard Brown The suprachiasmatic nucleus (SCN) Of the hypothalamus and its efferent projections are necessary for the generation Of circadian rhythms. TO further investigate the role Of SCN connections in the generation Of behavioral rhythms, male Long-Evans rats were housed in constant conditions and given horizontal knife cuts aimed dorsal to the SCN, bilateral parasagittal cuts lateral to the SCN, or sham surgery. Rhythms in locomotor activity and drinking behavior were monitored using a microcomputer. Horizontal cuts that spared the SCN failed to abolish rhythms. Effects Of horizontal cuts that damaged the SCN ranged from changes in period length Of the rhythms to abolition Of drinking rhythms. Parasagittal cuts did not damage the SCN or abolish rhythms. These and previous results indicate a functional redundancy in SCN efferent connections and that neural circuits within the SCN may be more important than SCN efferent connections in the generation Of behavioral circadian rhythms. INTRODUCTION Animals display daily fluctuations in a wide variety Of behavioral and physiological functions. These daily variations have been called circadian rhythms (from the Latin words circa--about and dies-—day) and persist even under constant environmental conditions. The persistence Of "free-running" ryhthms in the absence Of external time cues implies that the animal has an endogenous oscillator or "biological clock". Normally the timekeeping mechanism is synchronized by the light/dark cycle and the rhythm shows a period (average time between activity onsets on consecutive days) Of 2A hours. However, in the absence Of external time cues, the period Of the rhythm is close to but usually not exactly 2“ hours. Total or nearly total lesions Of the suprachiasmatic nucleus Of the hypothalamus (SCN) in rodents abolish free- running circadian rhythms in a variety of behavioral and physiological functions (Moore & Eichler, 1972; Stephan & Zucker, 1972; Moore & Klein, 197“). Anatomical studies have revealed that SCN neurons send efferent projections in caudal, lateral, dorsal, and rostral directions (Stephan et al., 1981; Berk & Finkelstein, 1981; Swanson & Cowan, 1975). Hypothalamic knife cuts which cut all of these projections 1 2 (i.e. surgical isolations Of the SCN) are equivalent tO SCN lesions in that they abolish rhythms in drinking, activity, sleep, and brain temperature (Stephan & Nunez, 1977). Furthermore, Inouye and Kawamura (1979) found that after isolation Of the SCN from the rest Of the brain, circadian rhythms in multiple unit activity persisted within an "island" of tissue containing the SCN but not in other areas of the hypothalamus. More selective hypothalamic cuts have also been used tO elucidate the functional significance of different SCN efferent projections. Retrochiasmatic cuts (placed caudal to the SCN) in a coronal plane disrupt rhythms in adrenal corticosterone secretion and pineal serotonin N-acetyltrans- ferase activity (Moore & Eichler, 1972; 1976; Moore & Klein, 197“) but not drinking (Nunez & Stephan, 1977), feeding (Nishio et a1., 1979), or locomotor activity (Nunez & Casati, 1979). It has also been shown that coronal cuts anterior to the nucleus and bilateral parasagittal cuts dO not abolish the nocturnal pattern Of drinking in rats kept on a light/dark cycle (Nunez & Stephan, 1977). However, persistence Of the nocturnal pattern of drinking in animals with anterior or parasagittal cuts could have been due to a "masking" effect of light, as these animals were never Observed under constant conditions. Additionally, the parasagittal cuts used in previous experiments were shallow and fibers projecting dorso-laterally from the SCN may have survived the cuts. Nunez and Stephan (1977) reported that 3 knife cuts which partially isolated the nucleus, severing fibers projecting dorsally, laterally, and caudally, but not rostrally abolished rhythms in drinking and activity in rats housed under a light/dark cycle or in constant dim light. Dark (1980), however, reported that similar partial isolation Of the SCN did not abolish entrainment Of drinking to a light/dark cycle or phase shifting when the light/dark cycle was phase delayed by four hours but did abolish free- running drinking rhythms when the rats were placed in constant dim light. From these results it can be concluded that: (i) the SCN is necessary for the generation Of normal circadian rhythms, (ii) rhythms are mediated by SCN efferent projections, (iii) rostral projections from the SCN alone are not sufficient to maintain behavioral rhythms in constant environmental conditions. The evidence indirectly suggests that either there is functional redundancy in the circuitry mediating behavioral rhythms (i.e. the information is carried by more than one set Of fibers and it is not necessary for all Of them to be intact tO maintain rhythms) or behavioral rhythms are mediated by dorsal or dorso-lateral projections from the SCN. These conclusions, however, are tentative as no one has yet reported Observations on animals with cuts that interrupt only dorsal projections from the SCN or on animals housed in constant conditions after receiving bilateral parasagittal cuts. The goal Of the research presented here was to determine the effects of deep parasagittal cuts and u dorsal cuts on free-running behavioral rhythms in order to provide evidence tO evaluate these tentative conclusions, and clarify the hypothalamic circuitry involved in the endogenous generation Of behavioral rhythms. EXPERIMENT 1 TO determine whether fibers projecting dorsally out Of the SCN are critical for the generation Of activity and drinking rhythms, knife cuts in a horizontal plane aimed dorsal to the SCN were made in rats housed in constant dim light and rhythms Of the animals were monitored before and after surgery. In additional animals, bilateral parasagittal cuts were made tO determine whether laterally or dorsolaterally projecting efferents Of the SCN are critical for the generation Of free-running rhythms. Method 912.1222 929 Maine Male Long-Evans rats (Blue Spruce Farms) were individually housed in cages with access to activity wheels, and given ad lib access to food and water. These animals were housed under constant dim illumination. The light intensity, as measured by reflectance Off of a white card using a Pentax Spotmeter, was 7 foot-candles. After 17-28 days Of adaptation to these conditions, 16-u8 days Of baseline data were collected. After baseline data collection, ryhthms Of the animals were monitored for 20-25 consecutive days following , the different surgical manipulations. 2222 Eellsszien 229 énelxeie Drinking spout licks were measured using an Optoelectronic drinkometer composed Of an infrared light emitting diode (LED) aimed at a phototransistor. The phototransistor and LED were placed beside the end Of the drinking spout such that the tongue Of the animal interrupted the light beam during each spout lick. Activity was monitored by a similar LED/phototransistor pair placed in a position such that a cam on the axle Of the wheel interrupted the light beam once during each wheel revolution. The phototransistors were connected to Optoisolated controller boards (Mullen Computer Products) inside a OAK Northstar Horizon computer. The computer also contained a real time clock board (Mountain Hardware, Inc.). A program written in BASIC, with assembly language subroutines for time-critical functions such as accessing the clock and reading the input ports, was used tO count responses per 12 minute interval. Each day 119 intervals (23 hours, 48 minutes) Of data were collected and then written on 5.25 inch floppy disks (Verbatim Corporation) for permanent storage. Data for interval 120 Of each day were not collected but were estimated by averaging interval 119 with interval 1 Of the next day. For visual inspection Of the records, the data were "doubleplotted". This procedure involves printing the data for a "8-hour period on each line and repeating the second 7 half Of the line on the first half Of the next line. Therefore, consecutive days follow each other vertically with each day repeated. Responses were added over 3 consecutive 12 minute intervals and total reSponses for each 36 minute interval were represented by one printed character. For activity records, a dark rectangle was printed for each 36 minute interval in which there were 3 or more responses while a blank space was printed if there were fewer than 3 responses. For drinking records, a threshold Of 15 responses/36 minutes was used. For further analysis, individual records were divided into blocks Of 10 days each. A chi-square periodogram (Sokolove & Bushell, 1978) was used to test for the presence of rhythms in the circadian range (23-26 hours). For each test period, P, within the range tested, the periodogram program yields a statistic, QP’ which is related to the root mean square amplitude of oscillations in the data set at a period Of P. TO aid in interpretation Of the data, plots Of Q versus P are made. In a "monte carlo" study, Sokolove and P Bushell determined that for rhythmic data such a plot shows a peak QP value at a test period at or near the period Of the cyclicity in the data while for data that are random with respect to time, QP distributes approximately as a chi- square with P-1 degrees of freedom. This provides a convenient significance test for the presence Of rhythms. An interplolation procedure (Enright, 1965) was used to Obtain QP values at a resolution Of 0.1 hours. In many cases 8 the animals showed rhythms that were readily distinguishable visually and appeared to have clear onset and termination Of activity with a stable period length. When this occured the periodograms generally yielded clearly distinguishable significant peaks at test periods that seemed tO correSpond well with the "visually apparent" period. An example Of a doubleplot and correSponding periodogram of this type of data is shown in Figure 1A. In other cases, blocks of the data appeared to be arhythmic or to have low amplitude rhythms with high amplitude "noise" (i.e. noncircadian) components. As illustrated in Figure 1B., periodograms computed on "noisy" data showed a reduced peak or nO peak at all. In a few cases the data visually appeared to have clear rhythms but with multiple components such as different periods for onset and termination Of activity. When this occured, periodogram analysis Often resulted in multiple peaks as shown in Figure 1C. In these cases a second estimate Of period length was Obtained by visually Obtaining and drawing a best fitting straight line across the onsets and terminations Of activity. In these cases the period Of the rhythm was defined as the test period Of the periodogram peak that most closely corresponded with the visual estimate based on onset Of activity. For each 10 day block of data, the average number Of responses per 12 minute interval was also computed by the periodogram program, and postsurgical activity and drinking levels were computed as per cent Of baseline. Figure 1: Representative doubleplotted event records and periodograms for data sets containing: A, high amplitude rhythms with low amplitude "noise" components; B, low or zero amplitude ryhthms; and C, multiple circadian periodicities. o- 0.! Figure 1. 10 ' U . 0'; !I '3 II III I go; on out on IS°FZI (C’ttl OO'QKI .t'le IS'CI! 00"! 06"! CD'S! OL'SZ 0"Sl OC'BZ OO‘SZ Ot'SZ 08": OI’SZ 00"! 06"! 00"! 0"98 00"! Ofi’OZ 0"02 OC'O! 03"! 01"! 00"! OO'SZ OI'tl 00°C! 06"! 00"! OI'SZ oz':z 01"! 00"! 06°68 OI'BZ Ol'fll OQ'EZ OB'SZ 09": 02°82 OI'SC OO'BZ 06°92 OI'OZ 01'08 09"! 03"! 09"! 0t'98 08°92 OI'OZ 00“.! 00"! 00": Ol'tz 06“! Ot'tt 11 §BE§SEX Each animal was randomly assigned to one Of the following surgical conditions. Horizontal knife cuts; After baseline data collection, surgery was performed under sodium pentobarbital (Nembutal, Abbott Laboratories) anesthesia (A0 mg/kg body weight, injected intraperitoneally) The surgery involved placing the rats in a stereotaxic apparatus (KOpff Instruments) then removing a flap of skull, retracting the superior sagittal sinus and lowering a Scouten (Scouten et al., 1981) retractable wire microknife through the brain to a point aimed midway between the SCN and the paraventricular nucleus of the hypothalamus (PVN). The stereotaxic coordinates were 2 or 1.5 mm posterior to the bregma and 7 or 7.1 mm ventral to the dura at the midline Of the animal with the incisor bar 7.5 mm below ear bar zero. The wire was then extended 2 mm from the barrel and the whole assembly rotated 3600 in each direction to make a circular cut aimed at severing all Of the dorsal projections from the SCN. Animals were then Observed for a minimum Of two weeks to determine whether the cut disrupted rhythms in drinking and activity. éiIEEszel Bereaegissel Eeiis ease; After baseline data collection, surgery aimed at cutting lateral projections was performed on a second group Of animals. The knife was lowered to the base Of the brain 3 mm posterior to the bregma and 1 mm lateral tO the midline on each side Of the 12 animal. The blade was then extended 2 mm in the rostral direction and raised 3 mm to make a deep cut in a parasagittal plane. Animals were then Observed for a minimum Of two weeks tO determine whether this type Of cut had any effect on free-running rhythms. Sham surgery; Sham surgery consisted Of making an incision through the skin and tissue on top Of the skull and suturing this wound. aieseiegz At the end Of the experiment, all lesioned animals were sacrificed by overdose Of sodium pentobarbitol and perfused transcardially with physiological saline followed by 10% (w/v) formaldehyde. The brains were removed and stored in 10% formaldehyde/30% (w/v) sucrose. Frozen 50 um coronal sections were then made Of the preoptic area and anterior hypothalamus. Every other section was mounted and stained with cresyl violet acetate and the slides coverslipped. A Zeiss microscope equiped with a camera system was used to verify the location of the lesions. For presentation Of histological results, selected sections were photographed using Kodak Technical Fan Film 2A15. Prints Of these sections were made on Kodak Polycontrast Rapid II paper. §E§2l2§ With the exception of 2 animals (see below), the rats exhibited clear circadian rhythms in drinking and activity before surgery. The results presented here are for animals that recovered from the brain surgery and for which 13 histological evaluation Of the damage was possible. After surgery, activity levels were reduced in 11 out Of these 12 animals. Postsurgical activity and drinking responses expressed as per cent Of baseline are presented in Table 1. Results Of the periodogram analysis for selected 10 day blocks Of activity and drinking records are presented in Table 2. §eee §2£§2£2 £2332 Activity and drinking data for a representative sham- Operated control rat are presented in Figure 2. All Of the sham-Operated control animals exhibited clear rhythms in drinking and activity before surgery. Two of the animals became hypoactive after surgery and statistical verification of postsurgical activity rhythms was not possible. Visual inspection Of the activity records however, revealed bouts Of activity with circadian periodicities and in phase with the presurgical rhythms. While the third sham-Operated rat showed a relatively low activity level, it displayed an activity rhythm that was readily apparent visually although the periodogram peak for the presurgical data sample was not statistically significant. Periodogram peaks for the postsurgical activity record Of this animal were statistically significant. Drinking records for all of these animals showed rhythms with circadian periodicities that were evident both visually and by significant periodogram peaks. In all of these cases the drinking rhythms persisted after surgery with no changes in period length. 111 O TABLE 1: POSTSURGICAL LEVELS OF ACTIVITY AND DRINKING EXPRESSED AS PER CENT OF BASELINE ACTIVITY DRINKING TYPE OF ANIMAL FIRST SECOND FIRST SECOND SURGERY NUMBER 10 DAYS 10 DAYS 10 DAYS 10 DAYS HZ 101 1U.6u* 11.46' 188.27 159.98 HZ 102 28.65 18.56* 450.51 164.72 HZ 110 6u.62* 56.00 132.92 28.31 H2 112 15.59“ 3.97' 101.93 113.78 HZ 113 43.02 “8.0” 127.25 151.uu HZ 115 119.55 27.82 109.“? 129.49 PS 108 38.56' 22.72* 106.31 29.94 PS 109 26.93 19.77 209.87 117.7“ PS 111 20.91’ ----- 1ON.86 97.23 PS 113*“ 1o3.u5 ----- 120.66 ----- PS 115** 121.30 ----- 57.05 ----- SH 105 N1.N9' ----- 122.u9 95.33 SH 106 11.77' o.u7* 232.37 351.66 SH 114 62.58 56.13 119.u5 84.96 * Activity levels were weighted according to the number of days that the animals ran. ** A 10 day sample Of data after horizontal cuts were made was used in calculating post-parasagittal cut data. .3335 uncuwcsnuna L8 to»: no: mmIi. 13 act mama «O 03.8» a can ms >3 :0 358.com an: 33.5.0. .33 #3393969 15333 Lou 0:333 an. Own: no: onus-$3 3:0 ”3:378: .83.... @3338 33 «O aoIoo 233 0.3.5” mac 9 < . £398» 333338 vacuum 23 3...: on» no no»: 0.8: 3.45 can amIom 23o so...“ mama 3 >3 :0 coscougon on: 3033 .915 280 50.5 :93» «Lo: 33 2.333 + .333023 2.3 5 258 v.83 3.23 3:33» M: a. 86 u a an 38:23» no: a .030: 633.850 mm .3098 579. name :0 39.8.8“— nu: {macaw .33: 33.850 8 uncaxo mo.o v m an 33:23» ago: 33: o .mw.~m, m.=~ =m.owp o.:m .ka.~.. a.=~ -.kk. _.mm .m.. we IIII tmwowop wosm mp.mop mosm 39.00? oocm mmomwp Noam mm: mm om.s.~ o.=~ am.~.~ k.=~ km.ow~ m.=~ .... .... .. .... PP. ma .am.~o. m.=~ .. .... mo.m- m.=m A_.ma~ A.=~ mo.mo~ a.a~ o..mmm c.s~ so. me ma.m- =.=~ .mm.km. 0.3m m.o.~ m.s~ mm.mm~ k.=~ Po.kmm m.a~ mm..=m m.=~ no. we ”m ms.mm. m.=~ mm.mk_ a.=~ No.Pm, p.3m mm.oo~ :.=~ ok.a- =.s~ .... +3.. zm mo..a~ m.=~ oe.sm. k.s~ mm.=mm k.=~ ...... .... so.oas m.=~ so. mm mp.m.~ o.=~ mm.oop o.=~ no.0o. o.=~ .... .... km.m~. ~.=~ mo, zm sm.sm. o.=~ ..o.oop m.=~ m~.~k. m.mm -.ks. ..m~ .. .... m=.oc~ =.m~ +m_. N: m..mo. m.=~ .oo.==. 0.:m .. .... mm.mm. ~.=~ m~.ok~ ..=~ Fc.mm~ a.s~ +m_. N: .m.op. o.=N .mo.km. p.2m .o.om~ m.=m ~m.oo. m.m~ wo.mo~ m.=~ eo.mm~ k.=~ ~.. N: on IIII cc IIII m=.cwp P.2m cmm.mop p.2N om.wwp >.m~ pm.omN p.a~ opp N2 waoahp woMN tmwoomp o.mm hm.NON 0.:N tpooomp o.mN cmN.N.~— comm ma.o>~ moam Now N: .mwommp moam nopfihp 0.:N mmomsm ooaN a=.mom =J~N mN.mNN 0.:N For N: ac . ao . aa a aa . n.a . ac, . mamas: »¢mom=m sasze Lo ma»h “m-mm mama “m-m_ mama p.-. mung “m-mm m.nn “w-m_ mung n.-. mama 023.253 uhH>Hho< mnmoumm 025.235 az< ~HH>HHU< mo mxuoam wNuuoqnou .nunaamcm Hmuwmusmunon one Hanumusnoua Lou pom: 0L0: amplmwp com .omp1ppp n>mu each mama can amp amp :0 vosuouuoa no: zuompan .opm an; Lou .nanaamcm Hmoumusnunon on» In own: cum: mo1om namo scum mumo new a: Imp so come one: muse HmucoNHLos .mpm can NPN mama Low .naoo kucoango: Lou ocuaonmo ma own: no: auomIan emcm Imago nouooaaoa AmmImw n>mvv name we cameo» Imp or c + .amumooNIoa 0:» cu venom 0L9: nxmoa acuNHmn M: II .mo.o u I so acoaINNcmgn no: a .oouoc onazuosuo no unouxo mp1ep name no vuauoauon no: anon .uouoc unaxuonuo no unooxo mo.c v m an unmouuucmun III: moans» a o~.mms m.=~ ~m.mmm :.=~ o:..em m.=N o=.NmN =.=~ mo.moo o.=~ :m.:FN m.=~ :.N :m o.o~m o.=~ om.omm N.=m Pm.=o~ o.=~ NN.oN: F.=~ :~.o.N m.=~ m:.moN ..sm m.~ :m om.e.m m.=~ ~_.=:. P.=~ mm.om~ m.=~ -.=o= N.:~ om..om _.sm =~.omm :.m~ N.~ :m I. .11. :N.me_ N.=~ .Nm.:m. :.=m 9:.omm ..aw :..o.~ m.=~ =m.-m :.:~ cpm m: .N..om :.=~ No..~m =.=~ oN.P.m m.=N mo.omo m.=~ oe.:m~ m.=~ =:.m~m m.=~ com me II 1111 II 1111 om.oa~ P.=m Fm.opz m.=~ ab.~m~ ~.=N om.ooo o.=~ mom mm NINNNE as... $3.: 0.3 .239 SN 8.2m mém 8.3: ....N mmdmw 3a mmm N: I:=.om. ~.=~ .. 11.1 m:...m m.=m No.m:. ..aw mN.oo~ :.=~ om.m~m o.=~ INN N: I. .111 I. 1111 Pm..a~ N.=N m:.=om P.m~ mm.mpm m.=~ om.m:: m.:~ omm N: .111 .111 mm..m~ o.=~ :o.:o~ m.=~ ::.m:. ..aw =~.=mm m.=~ m.~ N: 1.11 .I 1111 N..N- N.=N .11. ~0.oe. m.=~ N:.m~= o.=~ +o.~ N: .111 so.~o. =.=N o.o~m o.=m 1.1111 .11. ~:.mem N.=~ NN.oN: ..sm +mpm N: .111 ..:.::. ~.=~ mm.N.m m.=N 1111 :~.N.~ ~.=~ -.:o= N.=N .mpm N: =m.:- o.mm Na.0N. N.mN =N.m=~ ..sm ::.m:: m.m~ m..m=m :.m~ :m.~em ..sw :om N: Pm.oo~ N.=~ .N~.mo_ N.=N om.:N~ N.=~ m~.=oo o.:~ :~.Nom m.=~ mm.=oo m.=~ :om N: :c.:=m :.mm ::.ocm m.=~ :o.::~ o.=m o..oom :.m~ :~.N.m ~.=~ N...oe F.=N mom N: :o P :o D :o p :0 p :o A :a p ::::=: N:No:=m scszI no we»: .mmumm maao mm1mp muea o.-. mmNao¢ mh¢m amoqum mo womoumm quasza 02¢ NHH>HHU< no mxuoqm M49 op omhumqmm mo mamrq