目的：利用NMR的代謝組學(xué)方法比較慢性輕度不可預(yù)知應(yīng)激抑郁模型（CUMS）與慢性束縛應(yīng)激抑郁模型（CRS），為抗抑郁中藥新藥的研究提供模型參考。方法：選用SD雄性大鼠制作CUMS模型與CRS模型，以大鼠曠場行為、體重以及糖水偏愛作為常規(guī)指標(biāo)，并采用代謝組學(xué)技術(shù)對(duì)實(shí)驗(yàn)21 d大鼠尿液進(jìn)行NMR數(shù)據(jù)采集，運(yùn)用SPSS，MatLab7.5，SIMCA-P軟件對(duì)數(shù)據(jù)進(jìn)行分析。結(jié)果：行為學(xué)顯示，與正常組相比，CUMS組除水平運(yùn)動(dòng)外，各種常規(guī)指標(biāo)差異均有統(tǒng)計(jì)學(xué)意義（P＜0.05），CRS組大鼠水平穿越格數(shù)、豎立次數(shù)和理毛次數(shù)等指標(biāo)直觀上都減少，中央格停留時(shí)間延長，糞便粒數(shù)增加，但是無統(tǒng)計(jì)意義（P＞0.05）；NMR代謝組學(xué)分析結(jié)果顯示，散點(diǎn)圖中CUMS組與正常組各樣品散點(diǎn)比較集中，CRS組點(diǎn)較分散，介于正常與CUMS組之間。CUMS模型和CRS模型共有的標(biāo)志物有纈氨酸、亮氨酸、異亮氨酸、甘氨酸、酪氨酸、肌酸，CUMS模型特有標(biāo)志物為脯氨酸，CRS模型特有標(biāo)志物為天冬酰胺及精氨酸。結(jié)論：行為學(xué)及代謝組學(xué)均顯示CUMS組造模比較成功，CRS組在行為學(xué)指標(biāo)上較差，但運(yùn)用代謝組學(xué)技術(shù)顯示造模成功，說明代謝組學(xué)方法比行為學(xué)更靈敏，由于代謝組學(xué)技術(shù)比常規(guī)方法具有靈敏、整體、直觀性的優(yōu)點(diǎn)，因此可用于抑郁模型的篩選。

Objective: To evaluate the chronic unpredictable mild stress (CUMS) model and chronic restraint stress (CRS) model based on metabonomics using NMR. Methods: A total of 18 male Sprague-Dawley (SD) rats were randomly divided into the control group, CUMS group, and CRS group. Taking open-field test, body weight, and sucrose preference test as general indexes during the experimental period, the 21d urine samples were analyzed by NMR. Moreover the data were processed by the SPSS, SIMCA-P, and MATLAB software. Results: The results of behavior changes showed that compared with those of rats in control group, there was significant difference in CUMS group except crossing scores, while those of rats in the CRS group showed a significant difference in dwell time, sucrose preference and consumption. Scores plot of PCA for urine showed that the scatter points of three groups were separated well among groups. Points of CUMS and control group were clustered within groups, while the CRS group scatter points were far away from each other. The shared biomarkers of CUMS model and CRS model have valine, leucine, isoleucine, glycine, tyrosine, creatine, and CUMS model specific biomarker is proline while CRS model specific bioimarkers are aspartic acid and arginine. Conclusion: The behavior changes and metabonomic results show that the CRS model is bad in behavior changes but well in metabonomics, while the CUMS model is successful in each index, which suggests that metabolomics could be more sensitive than behavior test, and has the advantage of sensitive, ensemble, and intuitive, so it can be used for depression model screening.

科技部國際合作項(xiàng)目（2008DFA30430）；國家自然科學(xué)基金項(xiàng)目（300772759）；山西省中醫(yī)藥代謝組學(xué)國際合作研究聯(lián)合實(shí)驗(yàn)室建設(shè)（2008081043）