Development of predictive assay for progression of slow-onset type 1 diabetes and its application to disease prevention

日本1型糖尿病（T1D）是一种异质性疾病，其特征在于在发病模式或发病年龄方面具有不同的临床特征。据估计，成人发作T1D的患病率与儿童期发作T1D的患病率相似，约三分之二的成人发作T1D被归类为“缓慢发作T1D”。尽管抗谷氨酸脱羧酶65（GAD 65）抗体是慢发型T1D患者未来胰岛素缺乏的良好预测标志物，但5年时的阳性预测值约为65%，三分之一的GAD 65抗体阳性患者不需要长期胰岛素治疗。因此，为了开发未来胰岛素缺乏症的高度预测标记物，我们使用随机噬菌体展示技术进行了抗AD65抗体的精细表位作图。人胰岛GAD噬菌体随机展示文库的制备第一年，我们制备了人胰岛GAD噬菌体随机展示文库，并用来自慢发型T1D患者和健康对照者的混合血清筛选该文库。扩增与患者或对照血清反应的单个克隆后，进行PCR直接测序分析，以确定GAD65分子的特异性DNA序列.抗GAD表位特异性抗体检测的建立利用第一年获得的结果，我们试图通过夹心ELISA方法鉴定与慢发型T1D患者胰岛素缺乏进展相关的特异性表位。由于随机噬菌体抗体库系统结合ELISA法检测GAD抗体的信噪比较高，难以准确定位GAD抗体的表位，本研究尝试利用其他检测系统进行精细的表位定位。

Japanese type 1 diabetes (T1D) is a heterogeneous disease characterized by the different clinical features in terms of its mode of onset or age at disease onset. It is estimated that the prevalence of adult-onset T1D is similar to that of childhood-onset T1D, and that around two-thirds of adult-onset T1D are categorized in "slow-onset T1D". Although anti-glutamic acid decarboxylase 65 (GAD65) antibodies are good predictive marker for future insulin deficiency in slow-onset T1D, positive predictive value at 5 years is about 65% and one-thirds of GAD65 antibody-positive patients do not need insulin therapy for long period. Therefore, in order to develop the highly predictive marker for future insulin deficiency in slow-onset T1D we performed a fine epitope mapping of anti-AD65 antibodies using random phage display technology.1. Preparation of human islet GAD random phage display libraryIn the first year we prepared a human islet GAD random phage display library, and screened this library with pooled sera from patients with slow-onset T1D and healthy control subjects. After amplifying the individual clone that reacted with sera from patients or controls, the PCR-direct sequence analysis was performed to determine the specific DNA sequence for GAD65 molecule.2. Development of anti-GAD epitope-specific antibody measurementWith the results obtained in the first year, we tried to identify the specific epitopes associated with the progression to insulin deficiency in slow-onset T1D patients by sandwich ELISA method. There were many difficulties to identify the diabetes-specific epitopes for GAD antibodies using random phage display library system with ELISA because of the high Noise/Signal ratio.We are trying to do the fine epitope mapping using other detection systems.