Hypoalgesia Induced by Reward Devaluation in Rats

RESEARCH ARTICLE Hypoalgesia Induced by Reward Devaluation in Rats Ana Marı ´a Jime ´nez-Garcı ´a 1,4 , Leandro Ruı ´z-Leyva 1,4 , Cruz Miguel Cenda ´n 1 , Carmen Torres 2 , Mauricio R. Papini 3 , Ignacio Moro ´n 4 * 1Department of Pharmacology, Biomedical Research Center (CIBM) and Institute of Neuroscience, Faculty of Medicine, University of Granada, Campus Ciencias de la Salud, 18016, Granada, Spain, 2Department of Psychology, University of Jae ´n, Campus Las Lagunillas, 23071, Jae ´n, Spain, 3Department of Psychology, Texas Christian University, Fort Worth, TX, 76129, United States of America, 4Department of Psychobiology and Research Center for Mind, Brain, and Behavior (CIMCYC), University of Granada, Faculty of Psychology, Campus Cartuja, 18071, Granada, Spain *[email protected] Abstract Reduced sensitivity to physical pain (hypoalgesia) has been reported after events involving reward devaluation. Reward devaluation was implemented in a consummatory successive negative contrast (cSNC) task. Food-deprived Wistar rats had access to 32% sucrose during 16 sessions followed by access to 4% sucrose during 3 additional sessions. An unshifted control group had access to 4% sucrose throughout the 19 sessions. Pain sensitivity was measured using von Frey filaments (Experiment 1) and Hargreaves thermal stimuli (Experiment 2) in pretraining baseline, 5 min, and 300 min after either the first (session 17) or second (session 18) devaluation session in the cSNC situation. Sucrose consumption was lower in downshifted groups relative to unshifted groups during postshift sessions —the cSNC effect. Hypoalgesia was observed in downshifted groups relative to unshifted controls when pain sensitivity was assessed 5 min after either the first or second devaluation session, regardless of the pain sensitivity test used. Both pain sensitivity tests yielded evidence of hypoalgesia 300 min after the second downshift session, but not 300 min after the first devaluation session. Whereas hypoalgesia was previously shown only after the second devaluation session, here we report evidence of hypoalgesia after both the first and second devaluation sessions using mechanical and thermal nociceptive stimuli. Moreover, the hypoalgesia observed 300 min after the second devaluation session in both experiments provides unique evidence of the effects of reward loss on sensitivity to physical pain 5 hours after the loss episode. The underlying neurobehavioral mechanisms remain to be identified. Introduction Sensitivity to physical pain is influencedby a variety of emotional states [1,2], including stress inducedby immobilization[3] and food deprivation [4]. The emotional modulationof physical PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 1 / 15 a11111 OPEN ACCESS Citation: Jime ´nez-Garcı ´a AM, Ruı ´z-Leyva L, Cenda ´n CM, Torres C, Papini MR, Moro ´n I (2016) Hypoalgesia Induced by Reward Devaluation in Rats. PLoS ONE 11(10): e0164331. doi:10.1371/ journal.pone.0164331 Editor: Yuqing Li, University of Florida, UNITED STATES Received: August 13, 2016 Accepted: September 25, 2016 Published: October 20, 2016 Copyright: This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. Data Availability Statement: All relevant data are within the paper. Funding: This paper was partially supported by grant PSI-2013-44945-P from the Ministry of Economy and Competitiveness, Spain; the PhD Programme in Biomedicine and the Master’s program in Neuroscience and Pain, University of Granada, Spain. MRP’s participation was supported by a Fulbright US Scholar Award. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. pain alsooccursafter an experienceinvolving the devaluationof a large reward, as documented below. In the present experiments,reward devaluation was implemented in terms of the consummatory successive negative contrast (cSNC) task [5]. In a typical cSNC experiment,animals receivefree access for 5 min to a high-valuesucrose solution (typically 32% sucrose) during severaldaily sessions, followed by severalsessions of access to 4% sucrose. Performance during these downshift sessions is compared to the consummatory behavior of animals that have always received access to 4% sucrose (unshifted controls). The cSNC effect involves a suppression of consummatorybehaviorin 32-to-4% sucrose animals followed by a recoveryof levels similar to those of unshifted controls. The initial suppression (typically observedduring the first devaluationsession) and the recoverythat follows (typicallystartingduring the second devaluation session) are dissociablestages of the cSNC effect,as extensively demonstrated by Flaherty’s research on cSNC [5]. Several sources of evidencesuggest that cSNC modulates and is also modulated by physical pain. In one experiment[6], sensitivity to physical pain was assessedin terms of the paw-withdrawal latency in the hot plate after reward devaluation.The results showed reduced pain sensitivity (i.e.,increasedpaw-withdrawal latency) after the second 32-to-4% sucrose downshift session,but not after the first one, relative to unshiftedcontrols. Conversely, a subcutaneousformalin injectionin a hind paw before the first and second downshift sessions enhanced the consummatory suppression induced by either a 32-to-4% or a 16-to-4% sucrose downshift [7]. These studies add to growing evidencesuggestingthat situations that actually or potentially involve tissue damage share common underlyingmechanisms with situations involving reward devaluation.Basedon extensive evidence,Papini et al. [8] suggested a connectionbetweenphysical pain (tissue damage) and psychologicalpain (reward loss) that invites furthercomparisons betweenthesetwo sets of phenomena. For example, opioid ligands known to modulate physical pain also modulate cSNC, either during the first downshift session ([D-Pen2,D-Pen5]enkephalin, a selectivedelta-receptor agonist; [9]), during the second downshift session (U50,488H, a selectivekappa-receptor agonist; [10]), or duringboth sessions (morphine;[11]). Two opioid-receptor antagonists enhance the cSNC effecteither selectivelyduring the first downshift session (naltrindole, a selectivedelta receptor antagonist) or during both the first and second downshift sessions (naloxone, a nonselectiveopioid receptor antagonist) [12]. Similarly, lesions of brain areas involved in pain processing, such as the anterior cingulate cortex, also affectrecoveryfrom reward downshift from the seconddownshift session onward [13]. The goalof theseexperimentswas to lookfor evidenceof the modulationof physical pain by reward devaluation during the first and second downshift sessions by using different techniques to assess pain sensitivity:the von Frey test and the Hargreaves test. These tests estimate pain thresholds in terms of paw-withdrawal latency applying a localizedmechanicalor thermalstimulus. Both are relatively easy to apply, and have beenextensively used with rodents [14,15,16,17] and even humans [18]. Additionally, these experiments were designedto provide evidenceof the postsession time course of the effectby measuringpain sensitivity 5 and 300 min after the end of the reward devaluationsession.The von Frey test consists of filaments varyingin thickness that are pressed against the skinof a hind paw. Detectionof the mechanicalpressure results in paw withdrawal, thus providing an objectivemeasure of mechanicalhypoor hyperalgesia. The Hargreaves test estimates pain sensitivity also in terms of paw withdrawal, but afterapplying a localizedthermal stimulus [16]. These techniques differin the type of nociceptivestimulus used—mechanicalor thermal. Moreover, these techniques stimulate only a small patch of skin in a hind leg and, therefore, are probably more sensitive than other techniques, such as the hot plate test previously used in a similar experiment[6]. An additional advantage of the von Frey and Hargreaves tests for situations involving repeated testing is that there is little evidencethat they induce aversive conditioningin intact animals [19]. By contrast, just a single exposure Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 2 / 15 Competing Interests: The authors have declared that no competing interests exist. using the hot plate test has beenshown to induce aversive conditioning [20]. Thus, using the von Frey test (Experiment1) and Hargreaves test (Experiment2) would tend to minimizea possible interaction betweenreward devaluation and aversive conditioning.Based on previous results [6], a hypoalgesiceffect was expectedafter the seconddownshift session, but not after the first downshift session, at least when testing occurred5 min after the session. Testing 300 min after both sessions was expectedto yield little or no evidenceof changes in pain sensitivity. While there is no available evidencedescribingthe postsessiontime course of pain sensitivity effectsafter reward devaluation,posttrainingdrug manipulations in the cSNC task typically failed to show effectsafter a 180-min interval [21,22]. Materials and Method Subjects In Experiment1, the subjectswere 41 male Wistar rats purchased from Harlan Laboratories (Barcelona, Spain). In Experiment2, the subjectswere 40 male Wistar rats purchased from Charles Rivers (Les Oncins, France). A parvoviruscontamination in rats from Harlan Laboratories explains the switch of vendors; however, quarantine (one week) and habituation to the laboratory were done as usual. Rats were approximately 75 days old at the beginningof each experiment.The mean (±SEM) ad lib weight of all the rats was 281.2 (±2.1) g. Rats were housed individuallyin polycarbonate cages with ad lib water, in a room with constant temperature (24°C) and humidity (50–60%). Animals were housed under a 12:12 h cycle of light: darkness (lightson at 08:00 h) and fooddeprived to 82–85% of theirad lib weights throughout the experiment.Deprivation levels were maintained by providing rat chow at least 5–50 min after the end of all behavioral testing. In Experiment1, rat chow was from Harlan, Mucedola, Italy; in Experiment2, rat chow from Envigo (formerly Harlan), Barcelona, Spain. Water was continuously available in the home cage (these rats were never water deprived).These animals participated in a previous experimentin an instrumental successive negative contrast (iSNC) task in a runway situation, with solid foodpelletsas reward. These animals also had postsession access to either 2% ethanol or water. Assignment to the newconditionswas matched as far as possible for prior experience.In Experiment1, the numbers of rats with downshifted/unshiftedprior experiencewere 6/4, 4/6, 6/5, and 6/4, respectively, for Groups 32/17, 4/17, 32/18, and 4/18. In Experiment2, the equivalent numbers were 6/4, 6/4, 4/6, and 4/6, respectively, for Groups 32/ 17, 4/17, 32/18, and 4/18. As for prior ethanol/water exposure,the numbers for thesame groups were 7/3, 7/3, 8/3, and 8/2 in Experiment1, and 5/5, 5/5, 5/5, and 5/5 in Experiment2. The experimentalprotocols were approved by the University of Granada Research Ethics Committee. Apparatus All the behavioral procedures (cSNC and pain sensitivity measures) were conductedin the same lab room. The cSNC procedure was identical for both experiments. cSNC training was carriedout infour boxes made of clear Plexiglas and measuring30 x 30 x 15 cm (L x H x D). A graduated cylinder(in 0.01-ml units) containing the sucrose solution was introduced to the insidein the center of one of the lateral walls of the box.The amount of sucrosesolution consumedby each animal in each session (in milliliters)was obtainedby subtractingthe amount of sucrose solution recordedafter the session from the amount provided before the session. Animals licked the solutions from a metallic sipper tube protruding 3 cm inside the box. Sucrose solutions were prepared weight/weightby mixing 32 g of commercial sugar for every 68 g of distilledwater (32% sucrose)and 4 g of sugar for every96 g of distilledwater (4% sucrose). Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 3 / 15 The von Frey and Hargreaves tests were conducted in two boxes, also made of clear Plexiglas, and measuring20 x 20 x 24 cm (L x H x D). For the von Frey test, the floorwas made of aluminum bars. For the Hargreaves test the floorwas made of glass. Pain sensitivitywas determined by measuringthe paw-withdrawal response to a punctate mechanical or thermal stimulation of one of the hindpaws. In the von Frey test, stimulation involved applying one of a range of 9 von Frey filaments (Touch-Test SensoryEvaluators, North CoastMedical, CA, USA) ranging from 0.4 to 10 g (3.92–98.1mN) to a hind paw. For the Hargreaves test, radiant heat (42–43°C through the glass floor, for a maximum of 13 s) was applied by a plantar test apparatus (Ugo Basile,Comerio, Italy). Temperature was kept constant throughout the experiment. Procedure To the authors’ knowledge,this is the firstexperimentexploringthe effectsof reward downshift in the cSNC situation on physical pain assessedwith the von Frey filamentsand the Hargreaves thermal test. The protocols implemented here combine testing parameters usedseparately in prior experimentsinvolving cSNC, von Frey testing,and Hargreaves testing. cSNC traininglasted 19 daily sessions: 16 preshift sessions followed by 3 postshiftsessions. Four groups were included:32/17 (n= 10), 4/17 (n= 10), 32/18 (n= 11 in Experiment1, n= 10 in Experiment2), and 4/18 (n= 10). In group labels, the first number refers to the sucrose concentration administered during preshift sessions (32 or 4% sucrose;all animals had access to 4% sucroseduringpostshiftsessions),whereas the secondnumber refers to the session when the von Frey test or Hargreaves test was administered, either after the first postshift session (17) or the second postshiftsession (18). Each day, the animal rack was moved into the experimentalroom, rats were allowed 15 min in theirhome cage to settled,and then they were placedin the contrast box. The contrast session lasted 5 min from the first contact with the sipper tube. Session duration for each box was measured manually with digital clocks (Digital Onstart 100). Animals were run in squads of 4 and the order of squads variedacross days. At the end of all sessions, everyday, contrast boxes were wipedwith a wet paper towel and feceswere removed whenpresent. In Experiment1, Von Frey testing was conducted 5 and 300 min after session 17 for Groups 32/17 and 4/17, and after session 18 for Groups 32/18 and 4/18. Animals receivedone additional postshiftsession (19) to determinewhether von Frey testing affected consummatory behavior on the following session.To minimizenovelty effectswith the von Frey boxes, animals were familiarizedbefore the critical test sessions according to the following schedule. Baselinemeasurements were conducted two days before the start of iSNC training,duringa previous phase not reported here (see Subjectsfor a description of previous experience).In addition, rats were exposedto the von Frey boxes for 5 min per session,5 min after each of eight preshift sessions during the current experiment.Duringthese 8 box-exposure sessions no measurements were taken with von Frey filaments. In every test, each filament was applied three times for 2–3 s, separated by 5-s intervals using the up-down paradigm[23]. Testing started with the 2-g (19.6 mN) von Frey filament (i.e.,the middle of the range). The filament was manually pressed against the paw’s plantar surface with sufficientforce to cause a depression in the skin. The paw chosen for stimulation was always the same for a given animal and it was counterbalancedfor right and left hind paw within each group (5/5 for three groups and 6/5 for Group 32/18). In each consecutivetest, if there was no response to the filament, a stronger stimulus was then selected;if there was a positiveresponse, a weakerone was then used.The response to the Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 4 / 15 filament was considered positive when immediate withdrawal or shaking of the paw was observed.Observerswere blind with respectto the contrast assignment of the subject(i.e., 32% vs. 4%). In Experiment2, thermal pain sensitivity was assessedwith a technique describedin [24] with slight modifications.Hargreaves tests were conducted 5 and 300 min after session 17 for Groups 32/17 and 4/17, and after session 18 for Groups 32/18 and 4/18. Animals receivedone additional postshift session (19) to determinewhether Hargreaves testing affectedconsummatory behavior on the following session.To minimizenovelty effectswith the Hargreaves test boxes, animals were familiarizedbefore the critical test sessions according to the following schedule.Baselinemeasurements were conductedtwo days before the start of iSNC training (see Subjects).In addition, rats were exposedto the Hargreaves test boxes for 5 min per session, 5 min after each of eight preshift sessions during Experiment2. A video camera (Sony Handycam HD) was placed over the Hargreaves test boxes. During these sessions, no video recordings or pain measurements were taken. Pain sensitivity was assessedfor each animal after session 17 in Groups 32/17 and 4/17, or after session 18 in Groups 32/18 and 4/18. The Hargreaves test started with 5 min of habituation to the boxes. Duringthis period,toiletpaper was placed on the floorto collecturine and feces. After 5 min, the toilet paper was removed and a beamof radiant heat was focusedto the plantar surfaceof a hindpaws with a plantar test apparatus, until the rat made a withdrawal response. Paw withdrawal interrupted the light reflectedfrom the paw onto a photocell and automatically turned off the light and the timer. The latency of the withdrawal response (as an indirectmeasure of the heat-induced pain threshold) was thus recorded automatically. The intensity of the light was adjusted at the start of Experiment2 such that average baseline latency was about 13 s. This intensity was neverchanged. Each rat was tested twice alternately on each hind paw. All the latencies recorded were averaged to obtain a single latency measure per animal. Typically, a measurement was derivedfrom four recordings,two from each hind paw. At least 1 min was allowed betweenconsecutive measurements in the same paw. A cut-off latency time of 30 s was usedto avoid skin damage and minimize pain. A Hargreaves test usually lasted 15 min. Statistics In the cSNC situation, the dependent variable was the total amount of sucrose consumedin each session (in milliliters).In Experiment 1, pain sensitivity in the von Frey test was expressed as a mechanical threshold producing a response in 50% of the trials.This paw withdrawal threshold value was calculated using the following formula: 50% threshold ðgÞ ¼ 10ðXfþkdÞ=10;000 where X f is the value (in log units) of the final von Frey filament used,κis the tabular value (see Appendix in [23]) for the pattern of positive/negativeresponses, and δis the mean difference (in log units) betweenstimuli. In Experiment2, pain sensitivity in the Hargreaves test was expressed in terms of the mean paw-withdrawal latency (in seconds)over measurements taken from each paw. All data analyses were computed with the IBM SPSS Statistics 21 package. The specific analysis of variance used is describedin the Results section.Interactions were analyzed with pairwiseLSD tests derivedfrom the main analysis. Normality was assessed with the Kolmogorov-Smirnov test whenevera significant effectwas detected in downshift sessions and in the von Frey and Hargreaves tests to minimizeType I error. The alpha value was set at the 0.05 level in all statistical analyses. Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 5 / 15 Results Experiment 1 A Contrast (32% vs. 4% sucrose) x von Frey (session 17 vs. 18) x Session (1–16) analysis of preshift consummatory performance indicated that 32% groups consumed significantlymore sucrose than 4% groups, F(1, 37) = 28.65, p<0.001, and also there was a significant increase in consumption across sessions, F(15, 555) = 87.15, p<0.001. Sucrose consumption during the last preshift session (16) and the three postshift sessions (17–19) is presented in Fig 1. Consumption was higherin groups exposedto 32% sucrosethan to 4% sucroseon session 16, the last preshift session,F(1, 37) = 20.80, p<0.001, but there was no differencebetweengroups assignedfor von Frey testing on session 17 or 18, or interaction betweenthese two factors, Fs<1.91, ps>0.17. An analysis of postshift sessions data indicated that there was a weak cSNC effectlastinga singlesession.This analysis yieldeda significantinteractionbetweencontrast and postshiftsession,F(2, 74) = 5.68, p<0.006, and LSD pairwisecomparisons confirmedthat downshifted groups consumed significantly less sucrose on session 17 than unshiftedcontrols, F(1, 37) = 4.48, p<0.05. The differenceswere not significant for postshift sessions 18 and 19. There was no evidenceof deviations from normality on sessions 16–19 (statistics: <0.18, ps>0.09). Thus, consummatory behavior showed no evidenceof a downshift effecton session 18, before the von Frey test was administeredin Groups 32/18 and 4/18. Additionally, there was no evidencethat von Frey testing in one day affectedconsummatory behavior in the contrast box the following day. Fig 1. Mean (±SEM) sucrose consumption (ml) during the last preshift session (16) and either session 17 or 18 (Post) depending on the groups. 32: animals exposed to reward devaluation from 32% to 4% sucrose during postshift sessions. 4: animals exposed to an unshifted reward condition, always receiving access to 4% sucrose throughout the experiment. The asterisk reflects a significant difference between both downshifted groups vs. both unshifted controls (see text for details). doi:10.1371/journal.pone.0164331.g001 Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 6 / 15 Fig 2 shows the mean mechanicalpain thresholds in groups tested on baseline session, and after session 17 (top) or 18 (bottom). Whereas the baselinemeasurementwas obtained at the same postsessiontime for all animals,the postshiftmeasurements were obtainedafter the contrast session. Thus, these three values (baseline,5 min, and 300 min) are separated by different time interval.Becausethe relevant comparisons are betweendownshifted and unshiftedgroups tested equally on any given day, except for their prior history, these results were analyzed separately with one-way designs. Baseline measurements did not differbetweendownshifted and unshifted groups for both test day conditions, Fs<1. Two main outcomes are observedin Fig 2. First, in groups tested after the first downshift session (32/17, 4/17), pain thresholds increase after 5 min, but they decreasedafter 300 min to match the values of unshiftedcontrols. This was confirmedstatistically. Relative to unshiftedcontrols, downshifted animals exhibited hypoalgesia5 min after session 17, F(1, 18) = 8.61, p<0.01, but not 300 min after that session, F<1. Second,in groups tested after the second downshifttrial, pain thresholds also increasedin downshiftedanimals relative to unshifted controls after 5 min, F(1, 19) = 14.28, p<0.002, but, unexpectedly, this group differenceremained significanteven 5 hours after the end of the seconddownshift session, F(1, 19) = 13.58, p<0.003. No evidenceof deviations from normality was detected on any of the significanteffectsshown in Fig 2 (statistics: <0.23, ps>0.18). The results reportedabove were not dependenton group differencesin feedingmotivation, as assessed in terms of bodyweight. The mean (±SEM) weights across sessions 1–19 were 239.6 (5.5), 240.0 (5.6), 232.4 (3.2), and 231.6 (4.1) g for Groups 32/17, 32/18, 4/17, and 4/18, respectively. A Contrast x von Frey analysis detectedno effects,all Fs<1. A similar analysis on the mean weights during postshiftsessions 17–19 also detectedno effects,all Fs<1. Experiment 2 The results of the cSNC task were analyzed with Contrast(32% vs. 4% sucrose)x Hargreaves test (session 17 vs. 18) x Session analyses. Data from four individualsessions, one from each group, were lost due to technical difficulty;in these cases,the group average was substituted for the missing value. Preshift consummatory performance(sessions 1–16) showed that rats consumed significantlymore 32% sucrose than 4% sucrose,as shown by significanteffectsfor contrast, F(1, 36) = 84.33, p<0.001, and for the contrast by session interaction,F(15, 540) = 5.94, p<0.001. There was also a significant increase in consumption across preshift sessions, F(15,540) = 73.22, p<0.001. Other effectswere not significant.Fig 3 shows the performance during the last preshift session, session 16; an analysis of just this session indicated a significant contrast effect,F(1, 36) = 11.85, p<0.002. Fig 3 also shows the results of the three postshift sessions, sessions 17–19. A similar analysis yielded a significant contrast by session interaction,F(2, 72) = 9.64, p<0.001. In addition,there were significantmain effectsfor contrast and Hargreaves test, Fs(1, 36) >4.74, ps<0.04. The source of the contrast by session interactionwas a significantlylower sucrose consumption of downshiftedgroups compared to unshifted controls on sessions 17 and 18, as indicated by LSD pairwisecomparisons, Fs(1, 36) >11.61, ps>0.003. Across all sessions of training,there were no significant interactions involving the Hargreaves test factor, which indicated that group assignments were not biased.However, during postshift sessions, the groups tested for physical pain after session 17 consumed significantlymore sucrose than the groups tested after session 18. Notice that since the triple interaction was nonsignificant,F<1, the sizeof the cSNC effectwas similar in both sets of downshifted-unshiftedgroups. There was no evidenceof deviations from normality on sessions 16–19 (statistics: <0.24, ps>0.12). Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 7 / 15 Fig 2. Threshold force (g) for paw withdrawal in the von Frey test during baseline sessions and either 5 or 300 min after sessions 17 (top) or 18 (bottom), depending on the groups. 32: 32-to-4% sucrose downshift. 4: unshifted controls always exposed to 4% sucrose. The asterisks reflect a significant difference between the corresponding downshifted vs. unshifted groups (see text for details). doi:10.1371/journal.pone.0164331.g002 Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 8 / 15 Fig 4 shows the paw-withdrawal latency in the Hargreaves test for downshifted and unshiftedgroups tested after cSNC session17 (top) or session 18 (bottom).As in Experiment 1, measurements were separated by different timeintervalsand, therefore, the results were analyzed separately for baseline,5 min, and 300 min tests. Becauseof technical difficulties,data from two animals in Group 32/17, in the 5-min Hargreaves test, were lost, therefore leavingan n= 8 for this measure;however, thoseanimalsdid produce data for baselineand 300-min tests, thus leaving an n= 10 for both of these measures. In addition,one of the two measurements in either the right or left paw was lost in 9 animals from Groups 32/17 and 4/17, in the 5-min test; in all these cases,the secondmeasure taken from the correspondingpaw was usedin place of the average (the average of two measures was used when both were available). No differenceswere observedin terms of baselinelatencies either after session 17 or 18, Fs (1, 18) <1.17, ps>0.29. Pain sensitivity increased in downshifted groups relative to unshifted controls, as measured5 min after session 17 or 18. Thus, latencies were longer in Group 32/17 than 4/17, F(1, 16) = 6.43, p<0.03, and in Group 32/18 than 4/18, F(1, 18) = 38.26, p<0.001. As in Experiment1, the Hargreaves test yielded differenteffectsat the 300-min interval dependingon the session. After the first downshift event, latencies in Groups 32/17 and 4/17 had becomenondifferential,F<1. However, after the seconddownshift event, paw-withdrawal latency was still longer in Group 32/18 than in Group 4/18, F(1, 18) = 20.43, p<0.001. None of the significanteffectsshown in Fig 4 were basedon samples that deviatedfrom normality (statistics:<0.26, ps>0.05). The results reportedabove were not dependent on group differencesin feedingmotivation, as assessed in terms of bodyweight. The mean (±SEM) weights across sessions 1–18 (weights Fig 3. Mean (±SEM) sucrose consumption (ml) during the last preshift session (16) and postshift sessions (17–19). 32: animals exposed to a 32-to-4% sucrose downshift during postshift sessions. 4: animals exposed to an unshifted reward condition, receiving access to 4% sucrose throughout the experiment. The asterisks reflect significant differences between downshifted vs. unshifted groups (see text for details). doi:10.1371/journal.pone.0164331.g003 Hypoalgesia and Reward Devaluation PLOS ONE | DOI:10.1371/journal.pone.0164331 October 20, 2016 9 / 15