Articles

Article

Original Article

Exp Neurobiol 2012; 21(1): 16-22

Published online March 31, 2012

https://doi.org/10.5607/en.2012.21.1.16

© The Korean Society for Brain and Neural Sciences

Repeated Short-term (2h×14d) Emotional Stress Induces Lasting Depression-like Behavior in Mice

Kyoung-Shim Kim1, Hye-Joo Kwon2,3, In-Sun Baek4 and Pyung-Lim Han2,3,4*

1Laboratory Animal Center, Korea Research Institute of Bioscience and Biotechnology, Daejeon 305-860, 2Brain Disease Research Institute, 3Department of Brain and Cognitive Sciences, 4Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Korea

Correspondence to: *To whom correspondence should be addressed.
TEL: 82-2-3277-4130, FAX: 82-2-3277-3419
e-mail: plhan@ewha.ac.kr

Chronic behavioral stress is a risk factor for depression. To understand chronic stress effects and the mechanism underlying stress-induced emotional changes, various animals model have been developed. We recently reported that mice treated with restraints for 2 h daily for 14 consecutive days (2h-14d or 2h×14d) show lasting depression-like behavior. Restraint provokes emotional stress in the body, but the nature of stress induced by restraints is presumably more complex than emotional stress. So a question remains unsolved whether a similar procedure with "emotional" stress is sufficient to cause depression-like behavior. To address this, we examined whether "emotional" constraints in mice treated for 2h×14d by enforcing them to individually stand on a small stepping platform placed in a water bucket with a quarter full of water, and the stress evoked by this procedure was termed "water-bucket stress". The water-bucket stress activated the hypothalamus-pituitary-adrenal gland (HPA) system in a manner similar to restraint as evidenced by elevation of serum glucocorticoids. After the 2h×14d water-bucket stress, mice showed behavioral changes that were attributed to depression-like behavior, which was stably detected >3 weeks after last water-bucket stress endorsement. Administration of the anti-depressant, imipramine, for 20 days from time after the last emotional constraint completely reversed the stress-induced depression-like behavior. These results suggest that emotional stress evokes for 2h×14d in mice stably induces depression-like behavior in mice, as does the 2h×14d restraint.

Keywords: emotional stress, anxiety, depression, behavior

Depression is a prevailing form of psychiatric illness worldwide, but the underlying mechanism is not clearly elaborated. Recent progresses in the establishment of depression animal models give us a hope to unravel the molecular mechanism of depression and to develop advanced anti-depression strategies [1, 2]. Mice or rats exposed to a variety of stressors over a prolonged period display behavioral changes that parallel depression-like symptoms, such as decreased social interaction, decreased intake of sucrose, increased immobility time in helpless conditions such as in the forced swim test and tail suspension test [3-5]. These stress-induced behavioral changes are generally reversed by chronic, but not acute, treatment with antidepressant drugs, such as imipramine or fluoxetine [4, 6], supporting the usefulness of stress-induced animal models in depression studies.

Regarding animals models, various stressors are delivered repeatedly or continuously in mice or rats for a certain period of time to recapitulate depression pathophysiology in human. Among rodent models induced by "continuous" behavioral constraints are the chronic mild stress (CMS) model [3], chronic social defeat stress [4] and prolonged isolation stress [7], while among animal models induced by "repeated" constraints are animal models with repeated restraints [8-10]. Each of these models, including others, may have their own features. However, any specific model does not have all advantages over others. For examples, the social defeat stress model introduced by Berton et al. [4] have subject animals experience an offensive aggression physically for 10 min and then emotionally and contextually through olfactory, auditory and visual stimuli by placing mice in a close proximity for the rest of the day and this procedure was repeated for consecutive 10 days. The social defeat stress model produced by this experimental procedure has a feature to provide highly robust depression-like behavioral phenotypes and molecular changes [4, 11, 12], whereas this model requires a stable provision of proper offensive animals, for example aged CD1 mice [4, 11, 13]. Given the criticism that the 10-d continuous stress exposure in the chronic social defeat stress model might be too severe, "repeated" social defeat stress paradigms have been used [14-16]. Depression-like behavioral phenotypes can be also induced in animal models treated with repeated stress, such as with 2h×10d [17], 2h×14d restraint [10], or 6h×21d restraint [18, 19]. Repeated restraints have a superior feasibility in handling experimental procedures compared to other models, although the concept of effective doses of stress threshold and optimized stress strength have not been established.

Among many questions remained to be solved, it is a challenging question to know whether depression-related behavioral changes can be stably produced in animals after what extents of "emotional" stressor are repeatedly treated. Therefore, in the present study, we examined whether lasting depression-related behavioral changes is produced by exposing mice to emotional stressor repeatedly for 2h×14d.

Animals

Male C57BL/6 mice at 7 weeks of age were purchased from Daehan BioLink, Inc. (Eumsung, Chungbuk, Korea). They were housed in pair in clear plastic cages at a temperature (22~23℃)- and humidity (50~60%)-controlled environment with a 12 hour-light/dark cycle (light on at 7 a.m.) and were allowed to lab chow and water freely. They were adapted to the new environment for 1 week before use in experiments.

Treatment of water-bucket stress

To deliver water-bucket stress, 8-week-old mice weighting -22~23 g at 8 weeks of age were placed on a small stepping round platform (3.5 cm diameter) in a bucket (42 cm diameter × 55 cm height) with a quarter full of water. The depth of water was 17.5 cm, and the platform was elevated 1 cm above the water surface (Fig. 1A). The water-bucket stressor was delivered to animals for 1~3 h daily from 10 A.M. for 14 days on a scheduled procedure. Control mice were placed in pair in their original cages undisturbed at home environment. After each session of water-bucket stress, animals were returned to their home environment by housing them in pair in normal plastic cages with free access to food and water. Imipramine was intraperitoneally injected each day at indicated schedule. Animals were handled in accordance with The Guideline of Animal Care at Ewha Womans University School of Medicine.

Corticosterone measurement

Blood was collected from the heart of sacrificed mice at indicated time points in all cases, and centrifuged at 1,500 g for 15 min to obtain serum. Plasma corticosterone levels were assessed using a competitive enzyme immunoassay kit (R&D Systems, Minneapolis, MN, USA) as recently described [20].

Behavioral assessments

Behavioral assessments were carried out using a computerized video-tracking system, SMART (Panlab S.I., Barcelona, Spain) as previously described [10]. The mice were all placed in the same environment 30 min prior to the behavioral tests. Behavioral analyses were performed by following the sequences indicated.

The elevated plus maze (EPM) test

The EPM apparatus consisted of four arms (30×7 cm each) made of white Foamex, which was elevated 50 cm above the floor and placed at right angles to each other. Two of the arms had 20 cm high walls (enclosed arms), while the other two had no walls (open arms). For the EPM test, each mouse was initially placed at the center of the platform and left to explore the arms for 5 min. The number of entries into the open and enclosed arms and the time spent in each arm were recorded. Entry into each arm was scored as an event if the animal placed all paws into the corresponding arm.

The open field test (OFT)

Locomotor activity was measured in the open field made of a white Foamex chamber (45×45×40 cm). Each mouse was placed individually at the center of the open field, and the locomotion was recorded for the indicated period. The horizontal locomotor activity was judged by the distance the animal moved.

The tail suspension test (TST)

Mice were individually suspended 50 cm above the floor by the tail using an adhesive tape (approximately 1 cm from the tip of the tail) for 6 min. Immobility time during the last 5 min was recorded.

The forced swim test (FST)

Mice were placed in a transparent Plexiglas cylinder (height: 27 cm, diameter: 15 cm) containing water at a temperature of 22~23℃ and a depth of 15 cm so that they could not escape and could not touch the bottom. Mice were placed for 6 min in the water-filled cylinder and immobility time during the last 5 min was recorded. Immobility time was the total duration of animals remained floating with all limbs motionless.

Statistical analysis

Two-sample comparisons were carried out using the Student t-test, and multiple comparisons were made using one-way ANOVA followed by the Newman-Keuls multiple range test. All data were presented as the means±S.E.M. and statistical difference was accepted at the 5% level unless otherwise indicated.

A new depression model induced by emotional stress

Subject mice were individually enforced to stand on a small stepping platform (3.5 cm in diameter) placed in the middle of a bucket with a quarter full of water (Fig. 1A) and thereby to confront with potential drowning stress and isolation stress for hour(s). The emotional stress evoked by this procedure was termed "water-bucket" (or "small-island") stress.

Mice (C57BL/6, 7-weeks old, male) treated with water-bucket stress for 1 or 2 h daily tended to weigh slightly more than the control for the first 1~3 days. Thereafter they gradually weighed less than the control mice until day 14. Thus, mice with water-bucket stress showed body weight changes in a reverse S-shape pattern over time (Fig. 1B), which is sharply contrast to that induced by a restraint [10]. Mice treated with 1 or 2 h of water-bucket stress ate more food than the control during the initial 1~3 days, and this tendency, though weaken thereafter, was continued to the end of the recording (Fig. 1C). Mice treated with 1 or 2 h of water-bucket stress took more water than the control during the stress period, but their water intake returned to the control level after stress (Fig. 1D).

Corticosterone levels

The water-bucket stress increased plasma corticosterone levels with ~400 ng/ml of the peak value reached ~60 min after the start of stressor exposure (Fig. 2A), indicating that the water-bucket stress potently activates the HPA axis, but its activation level is slightly weaker than that induced by the 2-h restraint [10]. After the cessation of water-bucket stress, enhanced corticosterone levels slowly returned downwards to the baseline (Fig. 2A). When this daily 2-h water-bucket stress was reapplied for next days, plasma corticosterone levels at the end of each stress session were gradually declined (Fig. 2B).

Repeated (2h×14d) emotional stress induced anxiety-and depression-like behavior

Next, we examined whether repeated water-bucket stress induces behavioral changes. In the elevated plus maze (EPM) test, mice treated with water-bucket stress for 2h×14d, but not for 1h×14d, showed lesser percentages of entries and time spent in the open arm than the control mice (Fig. 3A~C). In the open field test, however, the locomotor activities of mice in all stressed groups were comparable to the non-stressed control mice (Fig. 3D). In the forced swim test (FST) which was performed 2 weeks after the last stress session (Fig. 3A), mice treated with water-bucket stress for 2h×14d or 1h×14d showed a significant increase in immobility time (Fig. 3E). We observed that depression-like behavior in the FST was detected a month after the last stressor exposure (data not shown).

The depression-like behavior induced by the 2h×14d emotional stress lasted for more than 3 weeks

Next, we examined whether depression-like behavior induced by 2h×14d water-bucket stress was robustly persisted for 3 weeks and whether treatment of stress-mice with the anti-depressant, imipramine, for 3 weeks, gave anti-depression effects. The forced swim tests performed on post-stress day 21 (3 weeks after the last stressor exposure) revealed that mice with water-bucket stress showed enhanced immobility time in the TST (Fig. 4A, B). Similarly, the FST performed on post-stress day 22 showed that the 2h×14d stress mice displayed enhanced immobility time (Fig. 4C). In contrast, administration of imipramine in the 2h×14d stress-mice for 21~22 days from post-stress day 1 reversed depression-like phenotype in the TST and FST (Fig. 4A~C). These results suggest that 2h×14d water-bucket stress produces longlasting changes in depression-related behavior.

The present study demonstrates that emotional stress evoked by placing mice individually to stand on confined small island in a bucket with a quarter full of water (so named water-bucket stress) for 2 h daily for 14 consecutive days (2h×14d) effectively produced a lasting change in depression-related behavior. This behavioral alteration was rescued by treatment with imipramine for 14~21 days of post-stress. We recently reported that restraint stress treated for 2 h daily for 14 days stably produced lasting changes in depression-like behavior [10]. Similar to the result of the previous study, the 2h×14d water-bucket stress activated the HPA axis and evoked anxiety- and depression-like behavioral responses. Though this 2h×14d water-bucket stress model should be strengthened by further characterization in the future, this model has a few attractive features as described below.

First, the 2h×14d water-bucket depression model is based on very emotional stress. Among depression models developed on the basis of emotional stress or emotional stress-related are prey-predator conflict stress model [21-23], territorial conflict stress model [24-26], isolation stress [7], conditioned emotional stress (CES) model [27, 28], chronic mild stress (CMS) model [3, 29], social defeat stress models [4,-30,32], and restraint stress model [8, 9]. To induce depression-like behavioral changes, intended stressor(s) in these models was delivered to subject animals repeatedly or chronically, from days to weeks to months depending models. Because applied stressors and stressor-treatment procedures in these models have their own unique features, the outcome of the behavioral changes and molecular and biochemical changes in the brain might have some distinct features. In particular, because sources of stressors used in these models are varied, the neural system responding to various perceptual modalities might be complexly impinged. In the water-bucket stress model, stressor modality is relatively simple and very emotional. In addition, a stable induction of depression-like behavior was provoked by the defined procedure, and its tied behavioral correction by treatment with imipramine supports for the usefulness of the 2h×14d stress paradigm to generate a depression model in mice. Future studies to reveal associated neural circuit for evoking emotional stress will be achieved using this model.

Second, the 2h×14d water-bucket stress procedure is composed of relatively mild stress, but this 2h×14d stress stably produces relatively long-lasting changes in depression-related behavior. The water-bucket increased plasma corticostrone level rapidly (Fig. 2), but the peak level was slightly lower than that induced by restraint and time-dependent decline upon repeated stress was also milder than that induced by restraint [10], suggesting that water-bucket stress activates HPA axis slightly milder than restraint stress does. The time-dependent decline in plasma corticostrone levels implies that the feedback-loop actively worked out in the body, but its requirement in the 2h×14d water-bucket stress model was less strongly active. Concerning the chronic social defeat stress model introduced by Berton et al. [4], it was produced by a continuous 10-days stress exposure, which consisted of daily 10-min defeat stress followed by enforcing subject mice to closely stay with an aggressor for the remainder of the day. Regarding isolation stress model, mice exposed to isolation stress for 3 months show anxiety- and anhedonia-like symptoms [7]. Isolation stress is a chronic mild stress, but the experimental procedure takes too long or require big patient to repeat in a regular laboratory. While the chronic mild stress (so called CMS) is given by multiple combinations of various unpredictable stressors within the CMS schedule [3, 29]. For examples, mice or rats are exposed to 85 dB noise for 3 h, water deprivation for 20 h, 45 degree cage tilting, wet cage-floor, repeated cold stress (4℃), empty cage, cage shaking, light-on for all day, and/or low intensity stroboscopic illumination (300 flashes/min) for 9 h, are combinatorially and continuously administered for a period of 4~8 weeks. These stressors as a whole appear not to be milder compared to isolation stress and water-bucket stress, although the total strength of stress among different models might be difficult to compare precisely, particularly when stress expose is prolonged.

Third, the 2h×14d water-bucket stress paradigm produces depression-like behavior which is long-lasting. Clinically used antidepressants usually need to be treated for 1~3 weeks to obtain antidepressant effects [33, 34]. Recent studies suggest that depression is produced by epigenetic, biochemical and structural alterations in the brain [4, 11, 12, 35] Because the 2h×14d water-bucket stress model is developed to have a window of more than 3 weeks from the day after the last stress session until behavioral tests, this model permits an opportunity to evaluate the efficacy of potential anti-depressants and their underlying mechanisms. Moreover, the 2h×14d water-bucket stress paradigm produces long-lasting behavioral changes, which will help us unravel the mechanisms underlying depression and develop novel strategies for anti-depression.

In conclusion, the present study demonstrated that the 2h×14d water-bucket stress, though mild and emotional, causes a lasting change of depression-like behavior, an effects that are produced by the 2h×14d restraint. Moreover, the present study demonstrated that repeated water-bucket stress model can be used to test of the efficacy of potential anti-depression drugs.

  1. Pittenger C, Duman RS. Stress, depression, and neuroplasticity: a convergence of mechanisms. Neuropsychopharmacology 2008;33:88-109.
    Pubmed
  2. Nestler EJ, Hyman SE. Animal models of neuropsychiatric disorders. Nat Neurosci 2010;13:1161-1169.
    Pubmed
  3. Willner P. Chronic mild stress (CMS) revisited: consistency and behavioural-neurobiological concordance in the effects of CMS. Neuropsychobiology 2005;52:90-110.
    Pubmed
  4. Berton O, McClung CA, Dileone RJ, Krishnan V, Renthal W, Russo SJ, Graham D, Tsankova NM, Bolanos CA, Rios M, Monteggia LM, Self DW, Nestler EJ. Essential role of BDNF in the mesolimbic dopamine pathway in social defeat stress. Science 2006;311:864-868.
    Pubmed
  5. Greene J, Banasr M, Lee B, Warner-Schmidt J, Duman RS. Vascular endothelial growth factor signaling is required for the behavioral actions of antidepressant treatment: pharmacological and cellular characterization. Neuropsychopharmacology 2009;34:2459-2468.
    Pubmed
  6. Covington HE, Maze I, LaPlant QC, Vialou VF, Ohnishi YN, Berton O, Fass DM, Renthal W, Rush AJ, Wu EY, Ghose S, Krishnan V, Russo SJ, Tamminga C, Haggarty SJ, Nestler EJ. Antidepressant actions of histone deacetylase inhibitors. J Neurosci 2009;29:11451-11460.
    Pubmed
  7. Wallace DL, Han MH, Graham DL, Green TA, Vialou V, Iñiguez SD, Cao JL, Kirk A, Chakravarty S, Kumar A, Krishnan V, Neve RL, Cooper DC, Bolaños CA, Barrot M, McClung CA, Nestler EJ. CREB regulation of nucleus accumbens excitability mediates social isolation-induced behavioral deficits. Nat Neurosci 2009;12:200-209.
    Pubmed
  8. Radulovic M, Weber C, Spiess J. The effect of acute immobilization stress on the abundance of corticotropinreleasing factor receptor in lymphoid organs. J Neuroimmunol 2000;103:153-164.
    Pubmed
  9. Pham K, Nacher J, Hof PR, McEwen BS. Repeated restraint stress suppresses neurogenesis and induces biphasic PSA-NCAM expression in the adult rat dentate gyrus. Eur J Neurosci 2003;17:879-886.
    Pubmed
  10. Kim KS, Han PL. Optimization of chronic stress paradigms using anxiety- and depression-like behavioral parameters. J Neurosci Res 2006;83:497-507.
    Pubmed
  11. Krishnan V, Han MH, Graham DL, Berton O, Renthal W, Russo SJ, Laplant Q, Graham A, Lutter M, Lagace DC, Ghose S, Reister R, Tannous P, Green TA, Neve RL, Chakravarty S, Kumar A, Eisch AJ, Self DW, Lee FS, Tamminga CA, Cooper DC, Gershenfeld HK, Nestler EJ. Molecular adaptations underlying susceptibility and resistance to social defeat in brain reward regions. Cell 2007;131:391-404.
    Pubmed
  12. Wilkinson MB, Xiao G, Kumar A, LaPlant Q, Renthal W, Sikder D, Kodadek TJ, Nestler EJ. Imipramine treatment and resiliency exhibit similar chromatin regulation in the mouse nucleus accumbens in depression models. J Neurosci 2009;29:7820-7832.
    Pubmed
  13. Tsankova NM, Berton O, Renthal W, Kumar A, Neve RL, Nestler EJ. Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action. Nat Neurosci 2006;9:519-525.
    Pubmed
  14. Bailey MT, Engler H, Powell ND, Padgett DA, Sheridan JF. Repeated social defeat increases the bactericidal activity of splenic macrophages through a Toll-like receptor-dependent pathway. Am J Physiol Regul Integr Comp Physiol 2007;293:R1180-R1190.
    Pubmed
  15. Becker C, Zeau B, Rivat C, Blugeot A, Hamon M, Benoliel JJ. Repeated social defeat-induced depression-like behavioral and biological alterations in rats: involvement of cholecystokinin. Mol Psychiatry 2008;13:1079-1092.
    Pubmed
  16. Fanous S, Hammer RP, Nikulina EM. Short- and long-term effects of intermittent social defeat stress on brainderived neurotrophic factor expression in mesocorticolimbic brain regions. Neuroscience 2010;167:598-607.
    Pubmed
  17. Wood GE, Norris EH, Waters E, Stoldt JT, McEwen BS. Chronic immobilization stress alters aspects of emotionality and associative learning in the rat. Behav Neurosci 2008;122:282-292.
    Pubmed
  18. Hunter RG, Bellani R, Bloss E, Costa A, Romeo RD, McEwen BS. Regulation of CART mRNA by stress and corticosteroids in the hippocampus and amygdala. Brain Res 2007;1152:234-240.
    Pubmed
  19. Hunter RG, Bellani R, Bloss E, Costa A, McCarthy K, McEwen BS. Regulation of kainate receptor subunit mRNA by stress and corticosteroids in the rat hippocampus. PLoS One 2009;4:e4328.
    Pubmed
  20. Lee KW, Kim JB, Seo JS, Kim TK, Im JY, Baek IS, Kim KS, Lee JK, Han PL. Behavioral stress accelerates plaque pathogenesis in the brain of Tg2576 mice via generation of metabolic oxidative stress. J Neurochem 2009;108:165-175.
    Pubmed
  21. Adamec RE, Shallow T. Lasting effects on rodent anxiety of a single exposure to a cat. Physiol Behav 1993;54:101-109.
    Pubmed
  22. Adamec R. Does long term potentiation in periacqueductal gray (PAG) mediate lasting changes in rodent anxiety-like behavior (ALB) produced by predator stress? Effects of low frequency stimulation (LFS) of PAG on place preference and changes in ALB produced by predator stress. Behav Brain Res 2001;120:111-135.
    Pubmed
  23. Figueiredo HF, Bodie BL, Tauchi M, Dolgas CM, Herman JP. Stress integration after acute and chronic predator stress: differential activation of central stress circuitry and sensitization of the hypothalamo-pituitary-adrenocortical axis. Endocrinology 2003;144:5249-5258.
    Pubmed
  24. Tornatzky W, Miczek KA. Behavioral and autonomic responses to intermittent social stress: differential protection by clonidine and metoprolol. Psychopharmacology (Berl) 1994;116:346-356.
    Pubmed
  25. Koolhaas JM, Meerlo P, De Boer SF, Strubbe JH, Bohus B. The temporal dynamics of the stress response. Neurosci Biobehav Rev 1997;21:775-782.
    Pubmed
  26. Bartolomucci A, Palanza P, Costoli T, Savani E, Laviola G, Parmigiani S, Sgoifo A. Chronic psychosocial stress persistently alters autonomic function and physical activity in mice. Physiol Behav 2003;80:57-67.
    Pubmed
  27. Gomita Y, Kataoka Y, Ichimaru Y, Ueki S. Methamphetamine mortality to emotional stimuli administered in the form of affective communication. Life Sci 1983;32:941-947.
    Pubmed
  28. Katsura M, Mohri Y, Shuto K, Tsujimura A, Ukai M, Ohkuma S. Psychological stress, but not physical stress, causes increase in diazepam binding inhibitor (DBI) mRNA expression in mouse brains. Brain Res Mol Brain Res 2002;104:103-109.
    Pubmed
  29. Willner P. Validity, reliability and utility of the chronic mild stress model of depression: a 10-year review and evaluation. Psychopharmacology (Berl) 1997;134:319-329.
    Pubmed
  30. Siegfried B, Frischknecht HR, Waser PG. Defeat, learned submissiveness, and analgesia in mice: effect of genotype. Behav Neural Biol 1984;42:91-97.
    Pubmed
  31. Huhman KL, Solomon MB, Janicki M, Harmon AC, Lin SM, Israel JE, Jasnow AM. Conditioned defeat in male and female Syrian hamsters. Horm Behav 2003;44:293-299.
    Pubmed
  32. Avgustinovich DF, Kovalenko IL, Kudryavtseva NN. A model of anxious depression: persistence of behavioral pathology. Neurosci Behav Physiol 2005;35:917-924.
    Pubmed
  33. Garattini S, Samanin R. Biochemical hypotheses on antidepressant drugs: a guide for clinicians or a toy for pharmacologists?. Psychol Med 1988;18:287-304.
    Pubmed
  34. Krishnan V, Nestler EJ. The molecular neurobiology of depression. Nature 2008;455:894-902.
    Pubmed
  35. LaPlant Q, Vialou V, Covington HE, Dumitriu D, Feng J, Warren BL, Maze I, Dietz DM, Watts EL, Iñiguez SD, Koo JW, Mouzon E, Renthal W, Hollis F, Wang H, Noonan MA, Ren Y, Eisch AJ, Bolaños CA, Kabbaj M, Xiao G, Neve RL, Hurd YL, Oosting RS, Fan G, Morrison JH, Nestler EJ. Dnmt3a regulates emotional behavior and spine plasticity in the nucleus accumbens. Nat Neurosci 2010;13:1137-1143.
    Pubmed