目的·优化关节腔内注射碘乙酸钠（monosodium iodoacetate，MIA）的方法，缩短实验时长，评估构建不同缺损深度大鼠颞下颌关节骨关节炎（temporomandibular joint osteoarthritis，TMJOA）模型的可行性。方法·将18只雄性Sprague-Dawley大鼠随机分为6组。对照组每只双侧关节腔内注射50 μL生理盐水，其余5组按照0.5、1.0、1.5、2.0、2.5 mg/50 μL注射MIA50 μL，分别记为NS组、MIA 1~5组，7 d后取颞下颌关节髁突进行观察分析：采用体视显微镜观察髁突表面形态；使用小动物计算机断层扫描仪（micro computed tomography，micro-CT）扫描并分析大鼠髁突软骨下骨参数［骨体积分数（bone volume/total tissue volume，BV/TV），骨小梁间距（trabecular separation/spacing，Tb.Sp），骨小梁数量（trabecular bone number，Tb.N）］；采用苏木精-伊红（hematoxylin-eosin，HE）染色和番红O-固绿染色进行髁突标本的组织学观察，并采用改良Mankin评分法进行评分。结果·与对照组相比，建模7 d后，体视显微镜下见MIA1、MIA2组髁突表面出现微小改变，MIA3、MIA4、MIA5组髁突骨质明显破坏，且破坏分布均匀。Micro-CT三维重建后见MIA3、MIA4、MIA5组髁突骨质明显吸收，MIA1、MIA2组髁突表面局部破坏。骨参数分析后发现MIA4、MIA5组的BV/TV显著降低（P=0.039，P=0.019）、Tb.Sp显著增大（P=0.030，P=0.003），MIA5组Tb.N明显减少（P=0.004）。HE和番红O-固绿染色结果显示：MIA1组纤维层欠光滑，蛋白多糖（红染区）变化不明显；MIA2组结构层次不清，出现无细胞区，蛋白多糖开始减少；MIA3组结构紊乱，细胞数量明显减少；MIA4、MIA5组软骨层几乎不可见。MIA3、MIA4、MIA5组改良Mankin评分均高于对照组（P=0.008，P<0.001，P<0.001）。结论·建模时间为1周时，1.5 mg/50 μL是诱导软骨呈典型TMJOA表现而软骨下骨松质无显著破坏的中间剂量，可用作构建以软骨缺损为主要病变的动物模型，2.0 mg/50 μL MIA可用于建立以软骨下骨破坏为主要特征的大鼠TMJOA模型。

Objective ·To optimize the method of injecting monosodium iodoacetate (MIA) into the joint cavity, shorten the experimental duration, and evaluate the feasibility of constructing temporomandibular joint osteoarthritis (TMJOA) models in rats with different degrees of injury. Methods ·Eighteen male Sprague-Dawley rats were randomly divided into 6 groups. The control group was injected with 50 μL of normal saline into the bilateral articular cavities of each rat, and the other 5 groups were injected with MIA at concentrations of 0.5, 1.0, 1.5, 2.0 or 2.5 mg/50 μL. The groups were recorded as the NS group, and MIA 1~5 groups. After 7 days, the condyles of the temporomandibular joint (TMJ) were collected for observation and analysis. The surface morphology of the condyles was observed under a stereomicroscope. Micro-computed tomography (micro-CT) was used to scan and analyze the subchondral bone of the condyles [bone volume/total tissue volume (BV/TV), trabecular separation/spacing (Tb.Sp), and trabecular bone number (Tb.N)]. Hematoxylin-eosin (HE) and safranin O-fast green staining were used for histological observation of the condyle specimens, and the modified Mankin scores were used for evaluation. Results ·Compared with the control group, under a stereomicroscope, minor changes were observed on the condylar surface of the MIA1 and MIA2 groups at 7 days after modeling, while obvious and evenly distributed bone destruction was observed in the condyles of the MIA3, MIA4, and MIA5 groups. Micro-CT three-dimensional reconstruction showed obvious bone resorption in the condyles of the MIA3, MIA4 and MIA5 groups, and local destruction of the condylar surface in the MIA1 and MIA2 groups. Bone parameter analysis revealed a significant decrease in BV/TV (P=0.039, P=0.019) and a significant increase in Tb.Sp in the MIA4 and MIA5 groups (P=0.030, P=0.003); Tb.N was significantly reduced in the MIA5 group (P=0.004). HE and safranin O-fast green staining results showed that the fibrous layer in the MIA1 group was not smooth, and the proteoglycan content (red-stained area) did not change significantly. The structure of the MIA2 group was unclear, cell-free areas appeared, and proteoglycan began to decrease. The structure of the MIA3 group was disordered and the number of cells was significantly reduced. The cartilage layer in the MIA4 and MIA5 groups was almost invisible. The modified Mankin scores in the MIA3, MIA4 and MIA5 groups were higher than those in the control group (P=0.008, P<0.001, P<0.001). Conclusion ·When the modeling time is 1 week, 1.5 mg/50 μL is an intermediate dose that induces typical TMJOA manifestations in cartilage without significant destruction of subchondral bone, and can be used to construct an animal model with cartilage defects as the main lesion.A dose of 2.0 mg/50 μL MIA can be used to establish a rat TMJOA model with subchondral bone destruction as the main feature.