Expression data from multiple myeloma plasma cells before and after co-culture with osteoclasts. Homo sapiens

To understand how interactions of myeloma cells with osteoclasts and mesenchymal stem cells in the bone marrow affect the clinical course of myeloma, we used microarrays to study changes in gene expression in freshly isolated myeloma plasma cells following co-cultures with osteoclasts (8 experiments) or with mesenchymal stem cells (13 experiments). Interaction with osteoclasts induced changes in the expression of 675 genes, and interaction with mesenchymal stem cells induced changes in the expression of 296 genes. Expression of only 58 genes commonly and similarly changed in both co-culture systems. Among these, we identified genes associated with overall, progression-free, and post-relapse survival, and developed survival prediction models. Gene expression data from 347 patients treated with total therapy 2 protocol, 433 with total therapy 3, and 98 patients who received various treatments (91 of them high-dose therapy with autologous stem cell support) were used for the analysis. Good predictive models were developed only for post-relapse survival, using genes involved in interaction with osteoclasts or with mesenchymal stem cells. The best predictive model used expression of first relapse of 33 probesets whose expression changed in myeloma cells following interaction with osteoclasts, with hazard ratios of 24, 20, and 12 for patients who relapsed following total therapy 2, total therapy 3 and the various other treatments, respectively. Among the probesets used for prediction, only 10, representing 8 genes, were commonly changed after both co-culture systems. These could present favorable target for therapy. Global gene expression profiling of primary multiple myeloma plasma cells (MMPCs) before and after interaction with osteoclasts (OCs) was done using Affymetrix microarrays. Overall design: Eight MMPC and OC co-culture experiments were performed using MMPC isolated from 8 patients and OC prepared from 8 different patients.

为明确骨髓内骨髓瘤细胞与破骨细胞（osteoclasts）、间充质干细胞（mesenchymal stem cells）的相互作用对骨髓瘤临床病程的影响，本研究采用微阵列（microarrays）技术，对分别与破骨细胞（共8组实验）或间充质干细胞（共13组实验）共培养后，新鲜分离的骨髓瘤浆细胞的基因表达变化展开分析。与破骨细胞的相互作用可诱导675个基因的表达发生显著改变，而与间充质干细胞的相互作用则可导致296个基因的表达出现变化。仅58个基因在两种共培养体系中均发生了方向一致且程度相似的表达改变。我们在上述共表达差异基因中筛选出与总生存期、无进展生存期及复发后生存期相关的基因，并据此构建了生存预测模型。本研究纳入了三类患者的基因表达数据用于分析：347例接受全疗程2（Total Therapy 2）方案治疗者、433例接受全疗程3（Total Therapy 3）方案治疗者，以及98例接受各类抗肿瘤治疗者（其中91例采用了联合自体干细胞支持的大剂量治疗方案）。仅基于与破骨细胞或间充质干细胞相互作用相关的差异基因，我们成功构建了针对复发后生存期的优质预测模型。最优预测模型纳入了33个探针组（probesets）的表达量，这些探针组对应的基因在骨髓瘤细胞与破骨细胞相互作用后发生了表达改变。该模型针对接受全疗程2、全疗程3及其他各类治疗后复发的患者，其风险比分别为24、20和12。在用于构建预测模型的探针组中，仅有10个（对应8个功能基因）在两种共培养体系中均发生了表达改变，此类基因或可成为多发性骨髓瘤治疗的潜在靶点。本研究通过Affymetrix微阵列技术，完成了原代多发性骨髓瘤浆细胞（primary multiple myeloma plasma cells, MMPCs）与破骨细胞（osteoclasts, OCs）相互作用前后的全基因组表达谱分析。总体实验设计：利用从8例患者体内分离得到的MMPCs，以及从另外8例不同患者体内制备的OCs，共开展8组MMPC与OC的共培养实验。