第一部分 幼年型粒单核细胞白血病DNA甲基化水平与临床特征相关性及地西他滨治疗短期疗效分析

背景

幼年型粒单核细胞白血病（juvenile myelomonocytic leukemia, JMML）是一种罕见的侵袭性儿童骨髓增殖性肿瘤，其特点是造血干细胞的异常克隆性增殖和RAS通路的过度激活。JMML患者的自然病程和转归具有异质性，多数未经治疗患者的中位总生存时间不足2年。研究表明，在JMML患者中，基因组DNA甲基化水平与预后密切相关，多项大型JMML队列DNA甲基化水平研究的荟萃分析达成了JMML患者DNA甲基化分组的国际共识，但目前不同DNA甲基化亚组的临床特征及预后仍存在种族差异。近年来，去甲基化药物（hypomethylating agents, HMAs）被尝试引入JMML治疗方案，JMML患者接受阿扎胞苷（azacytidine, AZA）治疗可获得部分/完全的血液学、细胞遗传学和分子学缓解。同为HMAs的地西他滨（decitabine）被广泛应用于成人髓系血液肿瘤的治疗，但在JMML治疗中的报道较少，目前尚未明确两种HMAs（地西他滨及AZA）在JMML治疗中的优劣性。本研究旨在寻找初诊时DNA甲基化水平与临床指标及预后的相关性，并回顾性探讨去甲基化药物地西他滨为主的药物治疗对改善造血干细胞移植前JMML患者病情的短期疗效。

方法

回顾性分析2013年12月至2020年10月期间我中心收治的26例JMML患者的临床资料，包括性别、年龄、全血细胞计数、胎儿血红蛋白（fetal hemoglobin, HbF）、骨髓穿刺、染色体核型、融合基因（需排除BCR/ABL和KMT2A重排者）等基线资料和实验室检查结果。

对初诊时有单个核细胞样本冻存的24例患者进行DNA提取，使用Illumina Infinium MethylationEPIC Beadchip（850k）平台定量全基因组DNA甲基化水平。根据JMML患者甲基化的国际共识，标准偏差大于0.25的位点被纳入DNA甲基化亚组的分类指标，采用无监督分层聚类法进行分组。

其中10例患者给予去甲基化药物地西他滨20 mg/m²·d ´5天治疗，肿瘤负荷较高者在地西他滨前加用阿糖胞苷50~100 mg/m²·d ´3~5天或依托泊苷50 mg/m²·d ´3~5天。每次治疗间隔为4周，3 ~ 4个疗程后桥接造血干细胞移植。根据国际专家共识评估患者对治疗的反应，疗效评估时间为下一周期治疗开始前一天。

结果

26例JMML患者中位发病年龄为18.0个月（范围：0.1 ~ 153），其中男性19例，女性7例。诊断时中位血小板计数为38´10⁹/L（范围：10~277），12例血小板计数低于33´10⁹/L。诊断时单核细胞计数中位数为4.8´10⁹/L（范围：0.5~15.4），所有患者单核细胞计数均高于同年龄正常值上限。23例患者中有16例HbF高于同年龄正常值上限（3例患者诊断时未检测HbF水平）。22例患者携带至少一种经典的JMML相关突变，4例无上述基因突变的患者被划分为五阴性组。

无监督分层聚类结果将24例诊断时有骨髓样本留存的JMML患者的基因组DNA甲基化水平分为两个亚组：13例低甲基化组和11例高甲基化组。不同甲基化亚组在年龄、血小板、HbF、PTPN11基因突变及≥2个体细胞突变的差异均有统计学意义（P<0.05）。相关性分析显示，高甲基化与确诊时年龄≥24月龄、血小板<33´10⁹/L、HbF高于同年龄正常上限、PTPN11突变和≥2个体细胞突变呈正相关（P<0.05）。

采用去甲基化药物地西他滨为主治疗的10例JMML患者，中位治疗疗程为3个周期，1个周期后总体反应率为70.0%，3个周期后为71.4%。治疗后白细胞和单核细胞计数显著降低（P<0.05），脾脏体积也呈降低趋势（治疗前后差异无统计学意义），12个月无事件生存率为80.0±12.6%。

结论

JMML患者高甲基化与确诊时年龄≥24个月、PTPN11基因突变和≥2个体细胞突变等高危预后因素显著相关。去甲基化药物地西他滨为主的治疗可降低JMML患者的肿瘤负荷，改善临床状况。

第二部分 JMML疾病中PTPN11基因突变的小分子靶向治疗探索

中文摘要

背景

幼年型粒单核细胞白血病（juvenile myelomonocytic leukemia, JMML）是一种罕见的儿童骨髓增殖性肿瘤，占儿童白血病的2%~3%，在14岁以下儿童中的年发病率约为1.2/10⁶。JMML患者通常对白血病常规化疗方案反应较差，目前造血干细胞移植（hematopoietic stem cell transplantation, HSCT）仍是唯一可能治愈该疾病的治疗方案。然而，约有半数患者在HSCT前发生死亡，接受HSCT的患者5年无事件生存率（event-free survival rate, EFS）也仅有44~53%，现在仍缺乏治疗JMML的特效药物。超过90%的JMML患者存在五种经典的RAS通路相关基因突变/缺失，包括PTPN11、NRAS、KRAS、NF1及CBL基因，其中PTPN11基因的体细胞突变最为常见。JMML患儿的预后与其基因突变的类型有一定相关性，携带PTPN11基因突变的患儿其5年EFS明显低于无PTPN11基因突变的患儿。目前亟需针对JMML患者中PTPN11基因突变者的靶向治疗药物。PTPN11基因编码的SHP2蛋白酪氨酸磷酸酶催化位点的高度保守性和携带正电荷的特性，使得该靶点长久以来难以成药。近年来出现的SHP2变构抑制剂也不能有效抑制发生功能获得性（gain of function, GOF）PTPN11基因突变细胞系的增殖，限制了该类药物在JMML患者中的应用。蛋白水解靶向嵌合体（proteolysis targeting chimeras, PROTAC）作为一种新型药物形式，以其可以靶向降解不可成药的致病蛋白、可能克服传统小分子药物耐药性等的优势，在多种疾病的治疗中已开展相关临床试验研究，可能成为一种用于具有PTPN11基因GOF突变患者的有潜力的治疗方案。基于以上研究背景，本研究拟构建可模拟JMML患者PTPN11基因突变疾病状态的细胞系模型，并利用该模型对新型SHP2蛋白变构抑制剂及PROTAC等针对PTPN11基因突变的小分子靶向药物的体外作用疗效及机制进行探索。

方法

利用具有PTPN11^E76K和PTPN11^D61G突变的慢病毒载体质粒，在TF-1细胞系中进行相应基因突变的过表达，对构建的过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞系，通过撤除生长因子后计数、蛋白免疫印迹实验（Western Blot, WB）、实时荧光定量反转录聚合酶连锁反应（quantitative real-time polymerase chain reaction, qRT-PCR）检测是否构建成功。

通过MTT检测法检测现有的部分SHP2变构抑制剂、20种新型SHP2变构抑制剂、靶向SHP2蛋白的PROTAC药物（SHP2-PROTAC）对过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞系的杀伤作用。

使用1μM、5μM、10μM浓度的SHP2-PROTAC和SHP2变构抑制剂SHP 099处理过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞系24小时后，利用流式细胞分析仪检测细胞凋亡和细胞周期变化情况。

利用WB方法验证SHP2-PROTAC处理后是否对过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞系内SHP2蛋白进行了靶向降解。

验证SHP2-PROTAC的细胞杀伤作用是否依赖E3泛素连接酶组分Cereblon蛋白（CRBN）及蛋白酶体：分别加用10μM浓度的CRBN配体Lenalidomide及1μM浓度的蛋白酶体抑制剂Bortezomib处理后，利用MTT及WB方法检测SHP2-PROTAC对过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞系杀伤作用及SHP2蛋白降解作用的变化。

结果

PTPN11^E76K和PTPN11^D61G突变的过表达成功使TF-1细胞发生恶性转化，脱离生长因子依赖，SHP2蛋白及其下游磷酸化ERK蛋白表达增高。

MTT药筛结果显示，20种新型SHP2变构抑制剂中，除编号#9、#13、#14、#20的化合物外，其他新型SHP2变构抑制剂对PTPN11基因突变细胞的杀伤活性均显著优于两种SHP2变构抑制剂SHP 099及RMC-4550。

TDS 0167型SHP2-PROTAC可以有效杀伤过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞，促进细胞凋亡并呈现明显的剂量依赖性，将细胞周期阻滞在G0/G1期，使胞内SHP2蛋白及其下游磷酸化ERK蛋白表达降低。

加用CRBN配体Lenalidomide或蛋白酶体抑制剂Bortezomib后，TDS 0167型SHP2-PROTAC对过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞的杀伤作用减弱，细胞内SHP2蛋白的降解减少。

结论

利用过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞模型，筛选到部分新型SHP2变构抑制剂对GOF型PTPN11基因突变细胞的体外杀伤效果优于现有SHP2变构抑制剂。

TDS 0167型SHP2-PROTAC药物能够有效杀伤过表达PTPN11^E76K和PTPN11^D61G突变的TF-1细胞，并对细胞内SHP2蛋白进行靶向降解，这种作用依赖于泛素-蛋白酶体系统（CRBN蛋白与蛋白酶体）。

Correlation analysis of methylation level and clinical characteristics at diagnosis of juvenile monocytic leukemia and short-term efficacy of decitabine treatment

ABSTRACT

Background

As a rare, aggressive pediatric myeloproliferative neoplasm, juvenile myelomonocytic leukemia (JMML) is characterized by abnormal clonal proliferation of hematopoietic stem cells and activation of RAS pathway. The development and outcome of JMML are heterogeneous, and the median overall survival (OS) time of most untreated patients is less than 2 years. Studies show that the methylation level of JMML patients is closely related to prognosis, and patients with high methylation level have relatively poor prognosis. Meta-analyses of studies on DNA methylation levels in large JMML cohorts have reached an international consensus on DNA methylation grouping in JMML patients. However, there are still racial differences in the clinical characteristics and prognosis of different DNA methylation subgroups. In recent years, hypomethylating agents (HMAs) have been attempted to be introduced into JMML regimens, as abnormally elevated DNA methylation levels may be a negative prognostic factor for JMML patients. Particularly notable, several studies have shown that azacytidine (AZA) can induce encouraging hematologic, cytogenic, and molecular response in JMML patients. However, since two types of HMAs (decitabine or AZA) are available, the dust has not yet settled which HMA is superior for JMML. Decitabine, which has also been widely used in the treatment of adult myeloid hematologic malignancy, is rarely reported in JMML treatment. This study aims to find clinical indicators that are associated with different methylation levels and prognosis, as well as the short-term effect of decitabine-dominant therapy on JMML patients before hematopoietic stem cell transplantation.

Methods:

The clinical data of 26 JMML patients admitted to our center from December 2013 to October 2020 were retrospectively analyzed, including but not limited to complete blood count with blood smear, fetal hemoglobin (HbF), bone marrow aspiration, chromosome karyotyping, fusion genes (especially excluding of BCR/ABL and KMT2A rearrangement), and next-generation sequencing.

The DNA methylation level of 24 patients’ whole genome at diagnosis were detected. An Illumina Infinium Methylation EPIC Beadchip (850k) platform were used to quantify the genome-wide DNA methylation. According to the international consensus of methylation in JMML patients, CpG sites with standard deviations greater than 0.25 were included, and unsupervised hierarchical clustering based on the minimum distance to the centroid was used in stratification.

Ten patients were treated with decitabine at a dose of 20 mg/m²·d for 5 days, after chemotherapy with cytarabine (50-100 mg/m²·d ´3~5 days) or etoposide (50 mg/m²·d ´3~5 days). The interval of each treatment was 4 weeks, bridging to transplantation after three to four courses. The response evaluation time was the day before the start of next cycle of therapy according to the international experts' agreement.

Results:

The median age of onset was 18.0 months (range: 0.1-153), including 19 males and 7 females. Median platelet count at diagnosis was 38´10⁹/L (range: 10-277), while twelve patients with platelet count less than 33´10⁹/L. Median monocytes count at diagnosis was 4.8´10⁹/L (range: 0.5-15.4), while all patients had age-adjusted monocytes count elevation. Sixteen in 22 patients had age-adjusted HbF elevation (three patients’ HbF level at diagnosis were not available). Twenty-two patients harbored at least one of canonical JMML mutations, while four patients with none of those mutations were defined as quintuple-negative group.

Genomic DNA methylation levels at diagnosis of 24 patients were classified into two subgroups: 13 cases in the low methylation (LM) subgroup and 11 cases in the high methylation (HM) subgroup. There were significant differences in age, platelets, fetal hemoglobin, PTPN11 gene mutation and ≥2 somatic mutations in different methylation subgroups (P<0.05). Correlation analysis showed that hypermethylation was positively correlated with age ≥24 months at diagnosis, PLT <33´10⁹/L, age-adjusted HbF elevation, PTPN11 mutation and ≥2 somatic mutations (P<0.05).

For patients treated with decitabine-based therapy, the median treatment course was 3 cycles, and the overall response rate was 70.0% after one cycle and 71.4% after three cycles. White blood cell and monocyte counts were significantly lower after treatment, and spleen volume was also lower, though not significantly lower. The 12-month event-free survival rate was 80.0±12.6%.

Conclusion:

JMML patients with hypermethylation were significantly correlated with high-risk prognostic factors such as age ≥24 months at diagnosis, PTPN11 gene mutation and ≥2 somatic mutations. Decitabine-dominant therapy was beneficial for reducing tumor burden and improving clinical condition in JMML patients.

Investigation of small-molecule targeted therapy for JMML with PTPN11 mutation

ABSTRACT

Background

Juvenile myelomonocytic leukemia (JMML) is a rare childhood myeloproliferative neoplasm. It accounts for 2% to 3% of childhood leukemia. The annual incidence of JMML is 1.2 in 10⁶ children under the age of 14. JMML patients usually respond poorly to conventional chemotherapy regimen. Hematopoietic stem cell transplantation (HSCT) is the only treatment option that can cure this disease. However, about half of the patients died before HSCT, and the 5-year event-free survival rate (EFS) of patients receiving HSCT was only 44~53%. There is still a lack of specific drugs for JMML treatment. More than 90% of JMML patients have five classic mutations related with RAS pathway, including PTPN11, NRAS, KRAS, NF1 and CBL genes, among which somatic mutations of PTPN11 gene were the most common one. The prognosis of JMML children was correlated with the type of gene mutation. The 5-year EFS of children with PTPN11 gene mutation was significantly lower than those without PTPN11 gene mutation. There is an urgent need for therapies targeting PTPN11 mutations in JMML patients. Due to the highly conserved and positively charged catalytic site of SHP2 protein encoded by PTPN11 gene, targeting this protein is difficult for a long time. Moreover, the SHP2 allosteric inhibitors in recent years also cannot effectively inhibit the proliferation of PTPN11 mutant cell lines with gain of function (GOF) mutations, which limits the application of such drugs in JMML patients. Proteolysis targeting chimeras (PROTAC), as a new drug form, has been used in clinical trials for the treatment of a variety of diseases due to its advantages of targeting and degrading unconsumable pathogenic proteins and overcoming the resistance of traditional small molecule drugs, which make it as a promising therapeutic option for patients with PTPN11 GOF mutations. Based on the above research background, this study intends to construct a cell line model that can resemble the disease state of PTPN11 gene mutation in JMML patients, and use this model for a preliminary exploration on the in vitro efficacy and mechanism of novel SHP2 allosteric inhibitors, PROTAC and other small-molecule targeted drugs targeting PTPN11 mutations.

Methods

1. Using lentiviral vector plasmids with mutations of PTPN11^E76K and PTPN11^D61G to overexpress corresponding gene mutations in TF-1 cell lines. By cell number counting after removal of the growth factor, Western Blot (WB), and quantitative real-time polymerase chain reaction (qRT-PCR) to detect whether the construction of TF-1 cell lines overexpressing GOF mutations of PTPN11 gene is successful.

2. MTT assay was used to detect the killing effect of some established SHP2 allosteric inhibitors, 20 novel SHP2 allosteric inhibitors, and PROTAC drugs targeting SHP2 protein (SHP2-PROTAC) on cell lines overexpressing GOF mutation of PTPN11 gene.

3. After 24h treatment with 1μM, 5μM, and 10μM of SHP2-PROTAC and SHP2 allosteric inhibitors SHP 099, the proportion of apoptosis and cell cycle arrest of PTPN11 GOF-mutant cells was detected by flow cytometry.

4. WB method was used to verify whether SHP2-PROTAC carried out targeted degradation of intracellular SHP2 protein.

5. Verify whether the killing effect of SHP2-PROTAC depends on the component of E3 ubiquitin ligase (Cereblon protein, CRBN) and proteasome: The CRBN ligand Lenalidomide and proteasome inhibitor Bortezomib were used to detect the activity of overexpressing PTPN11 GOF-mutant cell lines and the change of SHP2 protein expression after SHP2-PROTAC by MTT and WB.

Results

1. Overexpression of PTPN11^E76K and PTPN11^D61G mutations successfully led to malignant transformation of TF-1 cells, including the independent of growth factor, and increased expression of SHP2 protein and the phosphorylation of downstream ERK protein.

2. The results of MTT screening showed that, except for compounds #9, #13, #14 and #20 of the 20 novel SHP2 allosteric inhibitors, the killing activity of other novel SHP2 allosteric inhibitors on PTPN11 mutant cells was significantly better than that of the two established drugs, SHP 099 and RMC-4550.

3. The SHP2-PROTAC of TDS 0167 could effectively kill TF-1 cells overexpressed with mutations of PTPN11^E76K and PTPN11^D61G, promote their cell apoptosis in a dose-dependent manner, arrest their cell cycle in G0/G1 phase, and reduce the expression of SHP2 protein and its downstream phosphorylated ERK protein.

4. After addition of CRBN ligand Lenalidomide or proteasome inhibitor Bortezomib, the killing effect of SHP2-PROTAC on TF-1 cells overexpressed with PTPN11^E76K and PTPN11^D61G mutation was inhibited, and the degradation of SHP2 protein in the cells was reduced.

Conclusions

1. We successfully constructed a cell line model that can resemble the disease state of PTPN11 gene mutation (PTPN11^E76K and PTPN11^D61G) in JMML patients, and found novel SHP2 allosteric inhibitors had better killing effect on PTPN11 GOF-mutant cells in vitro than existing SHP2 allosteric inhibitors.

2. SHP2-PROTAC, TDS 0167, depending on the ubiquitin-proteasome system (CRBN protein and proteasome), could effectively kill TF-1 cells that overexpress PTPN11^E76K and PTPN11^D61G mutations, and degrade the cellular SHP2 protein.