Concepts & Trends

Behavioral reactivity to social and nonsocial stimulations: a multivariate analysis of six inbred rat strains

Male rats from six inbred rat strains (Spontaneously Hypertensive Rat, Wistar Kyoto, Brown Norway, Wistar Furth, Fischer 344, and Lewis) have been compared for their behavioral reactivity when placed in several nonsocial (elevated plus-maze, open
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Behavior  Genetics Vol. 27, No. 2. 1997 Behavioral Reactivity  to  Social  and  NonsocialStimulations:  A  Multivariate Analysis  of Six  Inbred  Rat Strains  livi r  Berton 1 3  Andre  Ramos 1 2  Francis  Chaouloff 1  and Pierre Mormede 1 Received 10 Mur. 1996—Final 25 Sept. 1996 Male rats  from  six inbred rat strains (Spontaneously Hypertensive Rat, Wistar Kyoto,Brown Norway, Wistar Furth, Fischer  344,  and Lewis) have been compared for theirbehavioral reactivity when placed in several nonsocial (elevated plus-maze, open field)and social (social interaction in aversive and neutral environment, resident-intruder test,chronic social stress) settings. In addition, a  factorial  analysis was performed to  assess how the variables measured in these  different  tests related to each other. Besides signif-icant strain-related differences in all  tests,  the  factorial  analysis showed that, in nonsocialenvironments, the strains contrasted essentially along two independent behavioral traits,the propensity to approach or avoid an aversive  stimulus  and general motor activity innovel environments (two indices of emotionality). In the social settings, marked  inter- strain differences  were  observed regarding  the  expression  of  aggressive  behaviors  but these differences were not related to the respective levels on the two nonsocial compo-nents of reactivity. Furthermore, large genetic differences were observed in variations ofbody weight induced by a chronic social srressor paradigm. The  factorial  analysis sug-gested a lack of relationship between the  effect  of social  stressors  on body weight and the  measures of emotionality and general activity obtained in the nonsocial  tests.  Con-versely, these variations were influenced  by the  levels  of  aggressiveness  and  sociability.Taken together, these results show (i) that the behavioral variability observed in rats, insocial and nonsocial environments, is  influenced  by genetic factors and (ii) that the be-havioral reactivity to social stimulations is a specific feature, dissociable  from  the levelsof the  different  components of emotionality (approach/avoidance and general activity) asevaluated by the behavioral responses to nonsocial settings. KEY  WORDS:  Inbred  rat  strains; factorial analysis; general activity; fear; anxiety; emotionality; aggressiveness;  chronic  social  stress. INTRODUCTION A  large  number  of  studies  have  pointed out theexistence of  significant  individual  variations  relat- 1  Genetique du  Stress,  INSERM, INRA, Universite de Bor-deaux  2,  Institut Francois  Magendie, 33077 Bordeaux Cedex, France. 2  Departamento  de  Biologia  Celular, Embriologia  e  Genetica,Universidade Federal  de  Santa  Catarina,  88.040-900  Floria-nopolis,  SC,  Brazil. 3  To  whom  correspondence  should  be  addressed  at  Laboratoire  de Genetique  du  Stress,  INSERM—Institut Francois Magendie,33077  Bordeaux Cedex,  France.  Telephone:  (33) Fax:  (33)  E-mail: ing  to social  stressor  effects  (Schuurman,  1981;  Sa-polsky,  1991;  Blanchard  et  al.,  1995,  Haller  et  at., 1995).  Taking  these  individual  differences into  ac- count  is certainly  useful  for the  recognition  of the neurobiological  mechanisms  that  underlie  stress and  fear  disorders.  Accordingly, several  authorshave examined  the physiological correlates of so- cial  rank  in male  rodents  submitted to social  stres-sors.  When  the  intensity  of social  interactions  is adequately maintained at a high level, dominant an-imals show locomotor hyperactivity,  increased  go-nadic  and  sympatho-adrenomedullary  function  and   OOOI-8Z4IW/0300-OI55S12.50/0  C 1997  Plenum  Publishing;  Corporation  156 Berton, Ramos,  Chaouloff,  and  Mormede the  development  of  cardiovascular pathologies (Ely and  Henry, 1978; Mormede  et  al. 1990; Blanchard at  al. 1995;  Me  Kittrick  et  al. 1995). Conversely, in  subordinates,  the  increase  in  pituitary-adreno- cortical  activity more  likely  results  in  digestive  andimmune  dysfunctions  (Ely  and  Henry, 1978; Comsa  et  al. 1982; Raab  et  al. 1986; Mendl  et al. 1992; Blanchard  et  al. 1995). Recently,  studies have shifted  from the  simple dominant-subordinate  classification  of the  individ- uals  to  focus  on the  behavioral strategies used  byanimals  to  deal with  the  social environment.  It has thus  been shown that some individuals primarily flee or  adopt  an  offensive  strategy when  faced  with an unfamiliar  conspecific, whereas others primarily freeze  (Schuurman,  1981).  These alternative behav- ioral  strategies, showing  a  considerable stabilityacross time {Blanchard  et  al. 1988),  are  generallyassociated  with  distinct neuroendocrine profiles(Sapolsky, 1991;  Haller  et  al. 1995). In  rodents, Benus  et al.  (1991) have proposed that  individual differences  in  reactivity  to  social stimulations  reflect heritable alternative strategies to  cope  with  environmental challenges,  likely  to be expressed  equally  in  nonsocial  situations.  Some  in- vestigators have actually shown that aggressive  an- imals  react  by  active avoidance  or  exploration when  they  are  placed  in an  aversive nonsocial set- ting  whereas nonaggressive ones rather respond  byimmobility  or  withdrawal (Svare  and  Leshner,1973; Benus  et  al. 1991; Mendl  et  al. 1992).  On the  basis  of  these results,  it can be  suggested that the  individual reactivity  to  social stimulation  re- flects the  general level  of  emotional reactivity rather  than  a  specific behavioral trait. However,some results have shown that social behavior,  and more  specifically aggressive behavior,  can be  mod- ulated  by  pharmacological treatments (Mos  et  al. 1992) or  environmental manipulation (Albonetti and  Farabollini, 1994), which  do not  alter  the  emo- tional  reactivity  of the  animals. Furthermore, sev- eral  inbred strains  of  rats known  to  contrast  in aversive nonsocial situations have been shown  not to  differ  when placed  in a  social setting  (Potegal and  Myers, 1989; Hendley  et  al. 1992; Lemaire and  Mormede, 1995). In  response  to  this  uncertainty,  the aim of the present study  was to  clarify  the  relationshipsamong individual reactivity  to  social stimulations,certain  indices  of  emotionality  (or  fear) evaluated by  nonsocial procedures,  and general  activity.  The influence  of  genetic factors  on the  determination  ofthe  type  and  intensity  of  stress responses  has  been widely  demonstrated. This genetic component  is likely  to be  partly responsible  for the  frequent  in-terindividual  variations observed  in  response  to so- cial  and  nonsocial stimulations (Eleftheriou  et  al. 1974;  Serova  and  Koslova, 1992; Popova  et  al. 1993; Castanon  and  Mormede, 1994). Conse- quently, the  comparison  of rat  strains  differing  in their  genetic background permits investigation  ofthe  putative association between emotionality  and the  behavioral reactivity  to  social stimulations.  In this  study  we  have  chosen  to  characterize  the be- havior of six  inbred  rat  strains submitted  to  widelyused  tests  of  anxiety/fear (elevated plus-maze  and open  field) and  placed  in  several social settings (so- cial  interaction  in  aversive  and  neutral environ- ment,  resident-intruder test  and  chronic social stress).  Each animal  was  characterized  in  this  set of  tests,  and in  order  to  investigate  the  multidimen-sional nature  of the  behaviors evaluated,  a  factorialanalysis  was  performed. METHODS Animals A  total  of 72  male rats  from  six  inbred strains was  used  in  these  experiments.  The  strains, namely, Spontaneous  Hypertensive Rats (SHR/NIco), Wis- tar  Kyoto (WKY/NIco), Brown Norway (BN/OrlIco), Wistar Furth (WF/Ico), Fischer  344 (F344/Ico),  and  Lewis (LEW/NIco), were pur-chased from IFFA CREDO (1'Arbresle, France). Upon  arrival  at our  laboratory,  the  5-week-old ratswere housed  as  triads  in  collective plastic cages  (28 X  43 X 18 cm)  with food  and  water available  ad libitum  under  a  12-h  light—dark  schedule (lights  on at  0700). Procedure Three  weeks  following  arrival,  the  animalswere submitted  to the first set of  tests,  i.e. those relating to  nonsocial behaviors.  All  animals under- went  each test (described below) only once, with an  interval  of 7  days between  two  tests. Animalswere then individually housed during  one  week  be- fore  the  beginning  of a  second  set of  tests,  i.e. those  concerning social  aspects  of  their behavior. All  tests,  which  were video recorded and analyzed  Behavioral  Reactivity  to  Social/Nonsocial Situations157 a  posteriori,  were  carried  out between 1330 and 1800.  Within  each  session,  the animals were as- signed  to the  order  of testing at random. Nonsocial  Behavior  Tests Open  Field  Novel  Environment/Dim Light) The  apparatus, made  of  wood,  had a  whitefloor  of  100 x 100 cm divided by black lines into 25  squares  of 20 x 20 cm. The  walls,  40 cm  high, were  also painted white. The test  room  had was dimly  lit,  with  7 lux  being  measured inside  the ap- paratus.  Each  rat was  placed  in the  center  of the open  field and its  behavior  was  filmed  by a  video- camera  for 5 min. The parameters evaluated were the  number of  outer  squares  (those  adjacent to the walls),  and the  number  of  inner squares,  crossed, these  two measures being referred to as outer and inner  locomotion, respectively.  The  number  of fe-cal  boli  (defecation) was counted at the end of eachtest at which time the whole  area  was cleaned using a wet  sponge  and a dry  paper towel (the cleaning procedure  was also the same for all other tests). Elevated  Plus-Maze As  described  by Chaouloff  et at  (1994), the apparatus  was made of Perspex, with  four  elevatedarms  (66 cm  from  the  floor)  45 cm  in  length  and10 cm in  width.  The  arms  were arranged  in a  cross- like  disposition,  with  two  opposite  arms  being  en- closed  (by walls  50  cm high) and two being open, having  at  their  intersection a central square plat- form  (10 x 10 cm)  which gave  access  to any of the  four  arms.  All floor  surfaces were black  and the central  platform was under an illumination of 70lux.  Each  rat was  placed  in the  central platform facing  an  open  arm and its  behavior  was  recorded for  5  min.  The  recorded variables were number  ofentries and  time spent  (with  all  four  paws) insideeach arm and number of end exploring behaviors in  open  arms. Social  Behavior  Tests Social  Interaction Aversive Environment) The  apparatus was the same as that described for  the  open  field  test,  with the difference that the light  intensity was  raised  to 250 lux in order toincrease the  aversiveness  of the environment. The animals,  after having  been  socially isolated  for 7days,  were placed in pairs (same strain but with no previous  common social experience) in two  oppo- site  corners of the test  arena.  Using the same re-cording procedures  as for the  open  field and the elevated plus-maze  tests,  the following behaviorswere observed  for 10  min: general locomotion (to-tal number  of  squares crossed)  and  total time  of social interaction. The latter comprised the follow- ing  classes  of  active interaction:  sniffing,  following, grooming,  crawling under,  and  crawling  over,  as described by File  (1980). Social Interaction  Neutral  Environment) After  2 weeks of  individual  caging, two ani- mals  (of the  same strain with  no  previous common social experience) were  placed  in the  test  arena  for 10  min.  In  order  to  minimize aversive  aspects  of confrontation  with environmental novelty,  the  test environment was  similar  in  every respect  to the home  cages (45 x 30 x 20 cm,  with  a sawdust- covered floor) under dim light (8 lux). The en-counters were videotaped and social behaviors ofboth subjects were then ethologically  analyzed. The following  behavioral parameters, based on behavior lists  by  Grant  and  Mackintosh (1963)  and by  Fer-nandez-Espejo and Mir (1990), were considered:(i) frequency and total duration of the sequences  biting attack, boxing,  aggressive allogrooming,  on  top,  and  fighting,  and (ii) total duration of nonaggressive interac-tions including the behavioral sequences sniffing,  allogrooming, crawling  over,  and crawling  under. Resident-Intruder  Test After  3 weeks of individual housing, each ex-perimental animal  was  confronted  with  a  collec-tively  housed  intruder  rat  (Wistar, IFFA CREDO) of  the  same  sex and  age.  The  intruder  rat was in- troduced into the resident home cage 10 s  afterfood  and  water removal, with  the  confrontation lasting  10 min. The test took  place  in the dimly litobservation room (8 lux). The encounters were vid-eotaped  and the  latency  to the first  fight,  the  totalnumber  of fights and the  respective  number  of fights  initiated  by the resident and the intruder wereevaluated. The percentage of animals attacking theintruder  in  each strain  was  also calculated.  158 Berton Ramos Chaouloff and Mormede Table  I. Strain Comparison in the Nonsocial Behavior Tests Variable  Strain  effect  SHR WKY BN WF LEW Open field Outer  squares Inner  squares Total  squares Defecations tf(5)  = tf(3)  = 19.38**33.02*'* H 5) = 22.88*** W(3)  = 23.28*** 786.0 83 0.9 ± 6 abc ±  1.2 a ±6 ab ±0.4  b 41  ±7 d 1  .9 ± 0.2 b 43  ±6 c 2.3  ±0.7  b 79 10.1 89 5.6 ±5ab ±2.0  a ±6a ±0.7  a80  ±9 a 10.5  ±2.0  a 91  ± 10 a 1.9  ±0.6  b 63  ± 7 bed 1.8  t.  1.3 b 65  ± 7 be 2.2  ± 0.5 b Plush  maze Time open  arms Time  closed arms End  exploring Closed arm  entries Open  arm  entries Total  entries A/(5)  = W(5)  =//(5)  = W(5)  =H(5)  = //(5)  = 12.17* 21.29*** 9.37  NS 12.16* 13.72* 12.95* 14.0 187 0.2 5.0 I.S6.S ±4.1  a ± 15 a ±0.) ± 0.5 be ±0.3 a ±  0.7 ab 2.8 ± 1 .5 be 225 ± 1 1 b 0.1  ±0.1 4.5 ± 0.7 be 0.3 ±0.1  b 4.8  +  0.7 b 9.8 198 0.56.8 1.2 8.0 ±2.9  a ± 10 b ±0.2 ± 0.5 ab ±0.3 a ±0.7b 8.5  ± 2.9 ab 196  ±9 b0.4 ± 0.2 8.2 ± 1.0 a 1.2  ±0.4 a 9.5 ± 1.2 a 1.6  £ 0.7 c 272  ± 9 a0.0  ± 0.0 4.3 ± 1.0 c 0.3  ±0.1 b 4.6 ± J.I b 55 8.0 63 0.9 13.2 220 0.6 6.4 1.6 8.0 F344 ± 9 cd ±  1.4 a ±9 be ±0.4  b ±  5.0 ab ± 19 b ±0.2 ±  1.4 abc ±0.6  a ± 1.9 ab   Mean  ± SEM of all  variables  are  represented  for  each  strain  {n  =  12/strain).  The  strain  effect  has  been tested  by the  Kruskal- Wallis  test.  *p  <  .05;  **p <  .001; ***p  <  .0001)  and an  adapted  t-test  has  been  used  to  perform  the  pairwise  comparisons. Two  strains  with  a common subsequent letter do not  differ  significantly  at the .05 level. Chronic  Social Stress The social stress procedure used here is the one  described by Mormede  el al.  (1990) as adaptedfrom Taylor  et al.  (1987).  The protocol is based onthe presence  of  sexually active females  and on group  instability.  It has  been shown that this par- adigm  promotes  a  high  level  of  social interactions between  males  that  is  associated with both  a  sig- nificant  body  weight  loss  and a sympathetic andpituitary-adrenocortical axis activation.  At thebeginning  of the stress period all the males were housed  three  (of  different  strains)  per  cage  (28 X43 X 18 cm) in the  presence  of  three females. Then the  composition of the social group was changed every  day by  randomized assignment  of  malesamong  cages  without changing the females. All an- imals  were changed daily  into  clean cages during the  last hour  of the  light  period  and  this procedure was  maintained during one week. Since the varia-tions  of  body weight during this  procedure have been  reported  to  reflect  the  physiological state  ofanimals  (Mormede  et  al.,  1990, Lemaire and Mor-mede, 1995), we have chosen this measure as an index  of the impact of  social  stressors. Statistics The  different  measures obtained in behavioraltests were first analyzed  individually.  Becausesome of the variables did not meet strict normality criteria, the Kruskall-Wallis  test  was  used  to  assess the  strain  effect,  whereas pairwise comparisons were  carried  out by an  adapted t-test (Crunch Soft-ware, 1991,  p.  534).  In  order  to  assess  the  infra- strain  effect  of chronic social stress, the Friedmantest was used. The Spearman correlation  coefficient was  used to test the association between pairs ofvariables. In  order  to  explore  the  relationships between the  respective  variables measured in the  different tests,  all variables were then submitted to a prin-cipal-component analysis with a varimax rotation.This statistical tool was used to learn more pre- cisely  if distinct biological phenomena were beingmeasured,  and to  determine  the way  variables  from different  tests  relate to each other. Only factors with  eigenvalues greater  than  one were kept. Based on the  factor loadings  of  each variable,  factor scores  were calculated (Crunch Software, 1991,  p. 477) for each strain, allowing graphic representa- tion  of the strains on each factor. RESULTSNonsocial Behavior  Tests  Table  1) Open  Field Significant  differences among strains were found  for all  measures  of ambulation (total  crossed squares,  p <-  .001; inner squares,  p <  .0001; outersquares,  p <  .001)  and for  defecation  p <  .001).BN, SHR and WF strains had the highest scores for  the  total  and  outer locomotion, whereas  the lowest  scores  were registered by WKY animals.  Behavioral  Reactivity  to  Sodal/Nonsocial  Situations 159 Fig.  1.  Social interaction  in an  aversive environment. Rats were placed  by  pairs  of the  same  strain  in the  open  field  during  10 min.  The  squares  crossed  and the  duration  of  social  interactions  in  each strain  are  represented.  Two  strains  with  a  common  letter do not  differ  significantly  at the .05  level. The  inner  ambulation was the lowest in WKY and LEW  rats,  whereas defecation was higher in BN rats  than  in all  other  strains.  No  significant  corre- lation  was  observed between defecation  and  loco- motion. Elevated  Plus-Maze Significant  differences  were  found  among strains  for all  variables evaluated  in  this test (open and  closed arm entries, total entries, and time inopen arms,  p <  .05 for  all;  time  in  closed  arms,  p <  .001)  except  for the end  exploring  of  open arms. The LEW and WKY  strains entered less  often,  andspent  less  time  in, the open arms  (two  indices of fear)  than  the  other strains, with  LEW  animalsspending  more  time  in the  enclosed  arm.  The  num-ber of enclosed arms entries (an index of generalmotor  activity)  was maximal in WF animals andthe lowest in SHR and WKY strains. The  time  spent  in  open arms  was  found  to be correlated  only  with  the  inner  locomotion  in theopen  field  r =  .39,  p  <  .001), whereas  the  number of  enclosed  arm  entries  was  correlated with  all lo- comotion variables  of the  open  field,  particularly with  the total number of squares crossed  r =  .46, p  <  .001). Social Behavior  Tests Social  Interaction in an Aversive Environment  Fig.  1) Rat  strain  had a  marked  influence  on the  var- iables evaluated (locomotion  and  duration  of  socialinteractions,  p <  .0001).  WKY,  LEW and  F344 rats  had the  lowest  scores  for  these  two  variables, the  scores  being maximal  in BN  rats. Although  theduration of  social interactions  and  locomotion werepositively correlated  r -  .61,  p <  .001),  it is  note- worthy  that  SHR  animals  had one of the  lowestinteraction  scores  in  spite  of  their high locomotion. Social Interaction in a Neutral Environment Fig. 2) Significant  differences  among strains were found  in the total duration of aggressive and  non- aggressive  interactions  p  <  .0001  for  each  vari-able).  WF  animals  interacted aggressively  for a longer duration whereas the BN strain was char-acterized  by the  longest duration  of  nonaggressive interaction.  Strains also  differed  significantly  in their  expression  of the  different  behavioral  postures considered (allogrooming,  p <  .05;  boxing,  biting attack,  on  top,  and fighting, p  <  .0001).  Altogether, SHR r  WKY and WF animals expressed all the ag-gressive postures more  frequently,  as  opposed  to the  less aggressive  BN and  F344  strains. Durations  of aggressive and nonaggressive in-teractions were  found  to  be  negatively correlated  r  —  -.54,  p  <  .001)  and  were independent  of body  weights. Resident-Intruder test Fig.  3) Significant  differences among strains were ob-served neither  in the  total number  of fights nor in the latency of the first fight. Although  offensive
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!