【作者】 王志青；

【导师】 姜泊；

【作者基本信息】 南方医科大学 ， 消化系病学， 2013， 博士

【摘要】 背景／目的：Peutz-Jeghers综合征（Peutz-Jeghers Syndrome, PJS, MIM#175200）是一种以皮肤粘膜黑斑、胃肠道多发息肉为特征的罕见常染色体显性遗传病。胃肠道多发息肉在患者青少年时期常常引起肠套叠、肠梗阻、消化道出血；随着年龄增长,PJS患者不仅消化道息肉有恶变潜能,还可伴发生殖系统和其他许多器官的良性或恶性肿瘤,其家族的癌症发病率也较普通人群为高。目前尚无干预措施能根治和预防PJS发病及息肉恶变。因此,研究PJS相关致病基因、息肉发生及恶变的途径和机制具有非常重要的理论和现实意义。PJS致病基因STK11/LKB1基因编码的丝氨酸／苏氨酸蛋白激酶是迄今发现的唯一一个抑癌作用的蛋白激酶,其属于AMPKK家族成员,通过对AMPK的变构调节,实现对能量代谢的调控；还可能通过转录因子SPI对VEGF产生负向调控,抑制细胞的增殖,也可能通过调控细胞周期,使细胞周期停滞G1期、促进细胞凋亡,抑制细胞增殖；在诱导神经细胞极性的形成中,通过与STRAD蛋白形成复合物而活化后,KB1S431A过量表达（一种丝氨酸-丙氨酸突变）,抑制轴分化；另有报道,STK11基因在非小细胞肺癌的极性调控中,可能通过JNK信号通路,对转录因子AP-1、c-Jun、JunD和ATF2进行调控,提高转录活性参与细胞极性形成。然而,关于PJS息肉形成及恶变机制的研究仍处于初级阶段。近年来,有报道PJ息肉为腺瘤样息肉,可发生癌变；另有报道错构瘤本身也可能演变为腺瘤和癌,即存在着由错构瘤→腺瘤→腺癌的演变过程；还有学者认为有两种机理导致PJS患者恶变,即错钩瘤→腺瘤→腺癌途径和de nove的恶变途径。然而,PJS错构瘤性息肉发生不典型增生/腺瘤样变的频率较低,因此,研究PJS息肉恶变的途径和机制一直是该领域的瓶颈。随着对PJS疾病的深入研究,探讨STK11/LKB1基因的突变类型、突变位点与PJS患者发生恶性肿瘤的关系,已成为该领域基础研究和临床研究的焦点。Lim W等报道PJS患者STK11/LKB1基因第三号外显子突变,发生恶性肿瘤风险高；Mehenni H则认为该基因第六号外显子突变与恶性肿瘤相关；Schumacher发现其羧基端和VIB-VIII功能区的错义突变与恶性肿瘤的关联更为密切,而移码缺失与剪切位点突变被认为与恶性肿瘤的发生无关,乳腺癌的发生主要由截短突变引起。然而,这些发现并没有被其他独立的PJS人群所证实。目前,PJS致病基因STK11/LKB1在不同研究人群中显示出较大差异的突变率（50%-90%）,基因型和表型的关系尚无定论,且PJS患者息肉恶变风险、恶变途径尚存争议。这些群体间的差异,不仅由于潜在的群体遗传学分层,样本量偏小也是重要原因,导致研究结果存在一定程度的分歧、无法在大规模人群中重复。可见,增加标本量,建立PJS家系生物标本库,做好随访工作,是进行这一罕见遗传病研究迫切需要解决的问题。本课题探讨建立PJS家系生物标本库的标准化程序,研究并制定规范化的组织及血液样品等生物标本采集、处理、保存操作流程,建立功能完善、信息丰富的PJS家系生物标本库。在此基础上,对PJS患者进行临床病理特征分析,探讨PJ息肉的病理演变过程,评估中国PJS患者发生恶性肿瘤的风险；并进行分子遗传学分析,绘制STK11/LKB1基因突变图谱,寻找中国PJS患者STK11/LKB1基因可能存在的突变热点区域及基因型和表型的关系；应用Protein Pathway Array技术进行PJ息肉蛋白质组学分析,识别与PJ息肉发生发展可能相关的信号传导通路蛋白以及潜在的分子治疗靶点。方法：1、借鉴国内外标本库建立的标准化程序,收集全国范围内PJS患者及家系正常人的血液标本和息肉组织标本,采用收集-反馈-修改的模式对标本收集的操作流程不断完善,并健全各种管理和质量控制措施。2、分析61个中国PJS家系133例PJS患者的临床病理资料,探讨PJ息肉的病理演变过程,评估中国PJS患者发生恶性肿瘤的风险。3、联合应用Sanger测序和MLPA检测技术,对52个PJS家系（遗传性家系25个,散发性家系27个）共计116例PJS患者、95例PJS家系正常人,进行STK11/LKB1基因突变筛查,并分析基因型和表型的关系。4、选取明确STK11/LKB1基因突变检测结果的28例PJS新鲜息肉组织和35例结肠癌癌旁正常组织（无STK11/LKB1基因突变）,应用Protein Pathway Array技术筛选与PJ息肉相关的蛋白表达谱,并用Western Blot和I雌予以鉴定。5、统计方法：数据处理采用SPSS17.0软件包,计数资料分别计算例数和所占比例,构成情况间的差别用χ2检验或Fisher’s精确检验,应用Student’s t-test和Significant Analysis of Microarray（SAM）进行差异蛋白质的筛选。以P<0.05,Q<5为有统计学差异。结果：1、初步建立中国PJS家系生物标本库通过制定标准化的标本收集流程,我们收集了61个PJS家系（28个遗传性家系和33个散发性家系）,133例PJS患者以及95例家系正常成员的生物标本,包括全血、血清、新鲜息肉组织、息肉蜡块组织以及正常粘膜组织,采集了完整的临床病理资料,并记录长期跟踪随访信息（尤其是息肉切除、息肉病理及恶性肿瘤发生的信息）,初步建立了PJS家系生物标本库。定期随机抽取样本提取DNA和蛋白进行检测,证明所收集的标本具有较高质量。2、中国PJS患者临床病理特征的分析分析133例PJS患者的临床病理资料,发现85例（64%,85/133）患者经历过至少一次外科剖腹手术（1-5次）,第一次手术的平均年龄为17.29岁（3-68岁）,多发息肉或体积较大息肉引起的小肠套叠或肠梗阻是导致外科手术的主要原因。16例（12%,16/133）患者的息肉出现错构瘤伴轻/中度不典型增生,患者平均年龄30.4岁,息肉直径均在3厘米以上,2枚息肉位于胃底,9枚分布于小肠,1枚位于直肠,其余息肉分布于结肠。25例PJS患者（18.8%,25/133）合并了27例恶性肿瘤,消化道肿瘤占66.7%（18/27）,其次为乳腺癌和妇科肿瘤,占25.9%（7/27）,另有白血病和肺癌各一例；其中2例PJS患者分别合并两种不同的恶性肿瘤,一例合并小肠癌、宫颈癌,另一例合并小肠癌、乳腺癌；患者发生恶性肿瘤的平均年龄为37.4岁（27-68岁）。3、Sanger测序法检测STK11/LKB1基因突变应用Sanger测序对52个PJS家系进行STK11/LKB1基因的突变筛查,27例先证者（51.9%,27/52）中检测到25种不同的STK11/LKB1基因突变,包括8个错义突变和17个截短突变,突变均与家系表型共分离,且在50例正常对照中未发现相同突变；其中,10种突变（40%,10/25）与PJS患者合并恶性肿瘤相关,包括胃肠道恶性肿瘤、胰腺癌、肝癌、肺癌、宫颈癌及白血病；14种（56%,14/25）突变为首次报道的新的病理性突变,包括四个缺失突变（c.151-162del12, c.308₃17del10, c.834₈35delTG c.898-906del9）、三个插入突变（c.402-403dupTG, c.892-893insC, c.402-403dupTG）.两个无义突变（c.783C>G,c.904C>T）、两个剪切位点突变（IVS2-1G>A,IVS5+2T>C）以及三个错义突变（c.862G>A, c.866T>G,c.891G>C）。值得注意的是,8个突变（8/27,29.6%）位于STK11/LKB1基因第七号外显子,是STK11/LKB1基因九个外显子中最短的一个,仅编码19个氨基酸,疑似中国PJS患者的突变热点。4. MLPA技术检测STK11/LKB1基因大片段缺失/重复应用MLPA技术检测25例Sanger测序未发现STK11/LKB1基因突变的PJS先证者,发现8例（15.4%,8/25）先证者存在该基因的大片段缺失,其中4例为1号外显子缺失（c.-1114-?₂90+?del）,1例为3号外显子缺失（c.375-?₄64+?del）,1例为3-9号外显子缺失（c.375-?₁365+?del）,1例为4-6号外显子缺失（c.465-?₈62+?del）,1例为6号外显子缺失（c.735-?862+?del）;三个PJS家系合并恶性肿瘤的发生,包括结肠癌、胰腺癌、乳腺癌和宫颈癌。5、分析基因型和表型的关系遗传性PJS家系STK11/LKB1基因突变率为80%（20/25）,高于散发性家系的突变率55.6%（15/27）（p=0.06）；在116例PJS患者中,18例（21.7%,18/83）携带STK11/LKB1基因突变的患者发生恶性肿瘤,5例（15.2%,5/33）未携带基因突变的患者合并恶性肿瘤（p=0.462）；此外,5个插入突变和3个（3/4）无义突变所在的8个PJS家系中均合并了恶性肿瘤的发生,而所有剪切位点突变（5个）和小片段缺失突变（4个）所在的家系中未合并恶性肿瘤。6, STK11/LKB1基因第Ⅺ功能区突变与PJ息肉发生不典型增生相关将检测出的27个STK11/LKB1基因突变位点按照STK11蛋白不同的功能区（Ⅰ-Ⅺ）进行分组,发现10个（37%,10/27）突变位于STK11蛋白第Ⅺ功能区（277-309氨基酸残基）,其中9个（90%,9/10）突变与胃肠道错构瘤性息肉发生不典型增生相关,且携带不典型增生息肉的患者平均年龄26.8岁（16-44岁）；然而,在STK11蛋白第Ⅰ-Ⅹ功能区的17个基因突变中,仅有2个（11.8%,2/17）突变与PJ息肉发生不典型增生相关；两组相比较,有显著性差异（p=0.0001）。随后,回顾性分析了HGMD数据库中PJS患者STK11/LKB1基因突变与恶性肿瘤的关系,发现76.9%（10/13）的Ⅺ功能区突变与恶性肿瘤相关。7.p-p38MAPK蛋白在STK11/LKB1基因Ⅺ功能区突变的PJ息肉中表达显著升高为进一步探讨STK11/LKB1基因第Ⅺ功能区突变与恶性肿瘤相关的机制,应用Protein Pathway Array技术对PJ息肉组织和正常肠粘膜组织进行蛋白质组学分析,比较13例携带STK11/LKB1基因Ⅰ-Ⅹ功能区突变的PJ息肉与4例Ⅺ功能区突变的PJ息肉中蛋白表达的差异,发现Ⅰ-Ⅹ功能区突变的PJS息肉中p-p38MAPK蛋白表达高于正常肠粘膜1.83倍,而Ⅺ功能区突变的息肉中该蛋白表达高于正常肠粘膜3.23倍,两组比较,p-p38MAPK蛋白表达差异具有统计学意义（p=0.003）,并用Western Blot验证了该结果。8. Galectin-3蛋白可能成为抑制PJ息肉生长的分子治疗靶点除p-p38MAPK蛋白外,应用Protein Pathway Array技术对28例PJS息肉组织和35例正常肠粘膜组织进行蛋白表达的筛选中还发现39种差异蛋白,13种蛋白在PJ息肉组织中呈高表达,26种蛋白呈低表达,并用Western Blot和IHC验证了Galectin-3,GSTP1, NQO1,COX-2及ICAM-1蛋白的表达,肯定了Protein pathway array实验结果的可靠性。其中,Galectin-3（半乳糖凝集素-3）蛋白在PJ息肉中的表达是正常粘膜的7倍,在PJS错构瘤和发生不典型增生的息肉中均呈胞浆强阳性表达,且其抑制剂低分子柑橘果胶（MCP）毒副作用极小,故该蛋白有望成为抑制PJ息肉生长的分子治疗靶点。9、应用IPA （Ingenuity Pathway Analysis）数据分析平台预测PJ息肉发生机制综合PJ息肉中筛选出的差异蛋白,我们分析STK11/LKB1基因突变后可能激活p38MAPK蛋白,从而通过调节炎症反应（COX-2, ICAM-1）和代谢途径（Galectin-3, GSTP1, NQO1）引起PJ息肉的形成,且p-p38MAPK蛋白可能在STK11/LKB1基因第Ⅺ功能区突变息肉发生不典型增生的病理演变过程中发挥重要作用。结论：1、规范化操作流程的建立有利于提高PJS家系生物标本收集的效率、提高标本库质量（包括生物样本和系统临床随访资料）,具有创新性和较强可操作性。2、首次报道中国PJS患者STK11/LKB1基因突变率,达67.3%,绘制了中国PJS患者STK11/LKB1基因突变图谱,且第七号外显子可能是中国PJS患者的突变热点,并发现14种新的致病突变。3、提出一种新的基因型和表型的关系,即STK11/LKB1基因第Ⅺ功能区突变与PJS患者错构瘤性息肉发生不典型增生相关；并发现p-p38MAPK蛋白可能在这一病理演变过程中发挥重要作用。4、发现Galectin-3蛋白可能成为抑制PJ息肉生长的潜在治疗靶点,其天然抑制剂低分子柑橘果胶（MCP）有望应用于PJS患者临床治疗。5、中国PJS患者因肠梗阻／肠套叠首次行剖腹手术（64%）、PJ息肉出现错构瘤伴轻／中度不典型增生（12%）以及PJS患者合并恶性肿瘤（19%）的年龄均较轻,平均年龄分别为17.29岁、30.4岁和37.4岁,为制定良好的监测方案提供重要依据,积极预防肠梗阻及恶性肿瘤的发生。更多还原

【Abstract】 BackgroundPeutz-Jeghers Syndrome (PJS, MIM#175200) is a rare autosomal dominant disorder characterized by gastrointestinal hamartomatous polyps and mucocutaneous pigmentation. Multiple gastrointestinal polyps of the PJS patients often caused intussusception, intestinal obstruction and gastrointestinal bleeding in adolescence. PJS causes not only higher cancer risk in digestive tract, but there is also an age-dependent increased risk for development of extra-intestinal malignancies compared to the general population. There is no intervention to prevent the occurrence of PJS polyps development or their malignant transformation. Therefore, exploration of PJS related genes, pathways and the mechanisms of PJS polyp development and malignant transformation has great significance.The major cause of PJS is mutations in the serine/threonine protein kinase STK11/LKB1,a family member of AMPKK and a tumor suppressor gene. It plays an important role in energy metabolism regulated by the AMPK allosteric modulator. STK11/LKB1could negatively regulate the transcription factor SPI to alter VEGF expression and thereby inhibit cell proliferation. Mutated STK11/LKB1could also regulate cell cycle arrest in G1to promote apoptosis. Nerve cells polarity is controlled by formation of the complex of STK11/LKB1and STRAD proteins leading to overexpression of KB1S431A (a serine to alanine mutation) to inhibit axis differentiation. STK11/LKB1controls cell polarity of non-small lung cancer cells by regulating JNK signaling pathways, which could regulate the transcription factor AP-1, c-Jun, JunD and ATF2. Despite evidence of the role of STK11/LKB1in energy metabolism, cell cycle and cell polarity, the mechanism of PJS polyp formation and malignant transformation is still not clear.Recently, it was found that some PJS polyps were adenomatoid polyps and may have higher risk of carcinogenesis. However, there is no conclusion whether the adenomatous polyps are evolved from hamartomas, or if they are two distinct clinical subtypes. Some scholars speculated that two mechanisms resulted in PJS malignancy, including the pathway of hamartomas-adenoma-adenocarcinoma and de novel malignant transformation. However, given the very low frequency of dysplasia, the exact role, if any, of the PJS polyp in cancer development is unclear.The attempt to correlate both the type and site of STK11/LKB1genetic mutations with the risk of malignancy has been the focus of many studies and progress in this area would represent an important research and clinical advance. Previous evidence suggested that mutations in exon3were associated with a higher cancer risk whereas in another study, statistically significant evidence correlated mutations in exon6with higher cancer risk. Schumacher et al found that missense mutations in the C terminus and regions VIB-VIII of the protein were more frequently associated with malignancies. On the other hand, in-frame deletions and splice site mutations have been found to be only rarely associated with malignancies while PJS patients with breast cancer were found to predominantly have truncating mutations. Unfortunately the analysis of additional sample sets by other groups seeking to validate these findings has not yet provided a clear genotype-phenotype association with malignancy.To date, the rate of STK11/LKB1gene mutations showed variation by geographical region (from50%to90%), and the relationship between genotype and phenotype was inconclusive. The malignant risk and pathways in PJS polyps were still controversial. Lack of concurrence between groups may be due to the different genetic backgrounds of the populations studied, but also the limited sample size. Thus, it is of great importance to establish a library of PJS pedigrees and biological specimens for further study of this rare genetic disease. In the present study, we established a biological specimen bank of Chinese PJS pedigrees according to the standardized procedures. We analyzed the clinical and pathological characteristics and assessed the malignant risk in this cohort of PJS patients. Then, we directly interrogated STK11/LKB1gene mutations in a subset of (116) these patients representing52index cases and analyzed the genotype and phenotype corelation. In addition, Protein Pathway Array analysis was used to identify proteins with altered expression in PJ polyps.Methods1. Reference standardized procedures for establishing a PJS biological specimen bank and use these to develop and standardize an operational process in PJS bio-sample collection. Collection PJS polyps tissue and blood samples nationwide began under implemented protocol. Using of the Collect-Feedback-Modify mode to review and improve the specimens collection operational processes and strengthen management and quality control measures.2. Twenty-eight PJS families and33sporadic cases were identified from a regional Chinese population for a total of133patients for clinical characteristic and malignant risk evaluation.3. Peripheral blood genomic DNA samples from116Chinese PJS "patients from52unrelated families (25PJS families and27sporadic cases) were investigated for STK11mutations using a combination of conventional direct DNA sequencing and the multiplex ligation-dependent probe amplification (MLPA) assay. Phenotypic correlations were investigated.4. To identify the PJS polyps protein expression profiles,28fresh PJS polyps (with clearly STK11/LKB1mutations) and35normal tissues (without STK11/LKB1gene mutation) were used for Protein Pathway Array screening.5. Statistical methods:Differences between groups were determined using chi-squared (x2) test or Fisher’s exact test. The statistical analyses were assessed using SPSS17.0software (SPSS Inc, Chicago, IL). The differentially expressed proteins were analyzed using the Student’s t-test and Significant Analysis of Microarray (SAM) screening. A value of P<0.05or Q<5was considered statistically significant.Results1. Establishment of the PJS pedigrees biological specimen bank.After the development of standardized specimen collection process, we collected61Chinese PJS families, including133PJS patients and105normal family members in the biological specimen bank. Of which,28PJS pedigrees have family history and33cases were sporadic.2.Clinical characteristics of patients with Peutz-Jeghers syndrome.In all,85patients (85/133,64%) had received at least one laparotomy (range,1-5) and the average age at first laparotomy was17.29years (range,3-68years). Polyp-induced complications, primarily intussusception of the small bowl, were the major referring cause for laparotomies. The overall cancer frequency in our study population was25/133(18.8%) and the mean age at cancer onset was37.4years (n=27). Gastrointestinal cancer was noted in18cases (18/27,66.7%). Breast and gynecological cancers were noted in25.9%(7/27). Of particular interest, gastrointestinal dysplastic hamartomas were identified in16patients (16/133,12%) and these were detected at a mean age of30.4years.3. STK11/LKB1gene point mutations.We identified germline point mutations in27of52(51.9%) of the index cases. Of these,15were found in familial (15/25,60%) and12in sporadic (12/27.44.4%) cases. We detected8missense mutations and17different truncating mutations. Ten mutations (10/25,40%) were associated with cancer in the index patient and/or in relatives with PJS. To our knowledge,14(14/25,56%) of these mutations are novel. Intriguingly, nearly one-third (8/27,29.6%) of all point mutations clustered in exon7, the shortest of the9exons.4. STK11/LKB1gene Large Genomic DeletionsWe next tested for the presence of exonic rearrangements by using MLPA in the25PJS probands in whom no mutation was identified by Sanger sequencing. The overall frequency of large deletions was8/52(15.4%). Three (3/8,37.5%) deletions were associated with cancer in the index patient and/or in relatives.5. Genotype-phenotype correlations.Analysis of the genotype-phenotype correlations showed that a higher frequency of mutations was identified in our patients with a family history of PJS (20/25,80%) when compared to our sporadic cases (15/27,55.6%); however, the difference was not statistically significant (p=0.06, x2=3.525). Eighteen (18/83,21.7%) patients with an STK11mutation developed a malignancy, which was not statistically different than the cancer frequency in our patients without STK11mutations (5/33,15.2%)(p=0.426, x2=0.634). None of the five splice site mutations or four small in-frame deletions was associated with malignancies in the index patient or their affected relatives. In contrast, all five insertion mutations and three out of four nonsense mutations were associated with malignancies.6. A novel correlation between kinase domain XI mutations with the development of dysplastic GI polyps.Ten of the27germline mutations (37%) that we identified were present in kinase domain XI (amino acids277-309). Strikingly, nine of the ten mutations (90%) were associated with gastrointestinal tract hamartomatous polyp dysplasia.Conversely, only two of the remaining17(11.8%) mutations were associated with dysplastic changes in polyps (p=0.0001).7. p-p38MAPK protein expression significantly increased in XI domain mutation-polyps compared to I-X domain mutation-polyps.Analysis of the differently expressed proteins between13PJS polyps with mutations in STK11/LKB1I-X functional domain and four polyps carrying XI domain mutations, we found that p-p38MAPK protein was upregulated1.83times in I-X domain mutation-polyps than the normal tissue. However, in XI domain mutation carrying polyps this upregulation was more pronounced at3.23times higher than the normal tissue (p=0.003, I-X vs. XI). The results were verified by Western Blot.8. Galectin-3protein a potential molecular therapeutic target for inhibition of PJ polyps. In addition to the p-p38MAPK protein, we found39differentially expressed proteins between PJS polyps and normal tissues by Protein Pathway Array screening. Of these,13proteins had increased expression in PJS polyps and26proteins lower expression compared to normal tissues. The expressions of proteins Galectin-3, GSTP1, NQO1, COX-2and ICAM-1were verified by Western Blot and IHC.Most notably, Galectin-3protein expression in PJ polyps was7times higher than the normal mucosa. IHC showed PJS hamartoma and the dysplasia polyps had strong cytoplasmic expression, but expression was negative in normal tissue. The inhibitor of Galectin-3protein, low molecular weight citrus pectin (MCP), could be a molecular therapeutic target for inhibition of PJ polyps given that it has minimal side effects.9. Prediction of the mechanism of PJ polyps development by IPA (Ingenuity Pathway Analysis).After IPA analysis, we speculated that STK11/LKB1mutations could activate p38MAPK, which would cause the formation of PJS polyps through the inflammatory response (COX-2, ICAM-1) and metabolic pathways (GSTP1, Galectin-3, NQO1).Conclusion1. The establishment of standardized operation process is conducive to improve the efficiency of collecting PJS pedigrees biological specimens and to improve the quality of specimen bank (including biological samples and clinical follow-up data).2. Our study is the first to report the mutation discovery rate of67.3%in Chinese PJS patients. Exon7could be the mutation hot-spot in this cohort. And we also identified14novel STK11/LKB1mutations.3. The present study is the first to identify a novel correlation between STK11kinase domain Ⅺ mutations with the development of PJS dysplastic GI polyps. The extent of the association between dysplasia and the development of G1-related cancers is currently unknown but our results highlight a novel STK11genotype-phenotype association as the basis for future studies. In addition, we found p-p38MAPK protein expression was significantly higher in XI domain mutation-polyps than I-X domains.4. Galectin-3protein could be a new molecular therapeutic target for inhibition of PJ polyps. The natural inhibitor of Galectin-3,low molecular weight citrus pectin (MCP) has potential to be used in the clinical treatment of PJS patients.5. In our cohort, the average ages of the first laparotomy (64%), the gastrointestinal dysplastic hamartomas (12%) and malignancies were17.29,30.4and37.4years separately. These data provide an important theoretical basis for the development of monitoring programs to prevent the occurrence of intestinal obstruction and malignancies.更多还原