【作者】 杨京；

【导师】 陈林；

【作者基本信息】 第三军医大学 ， 重症医学， 2013， 博士

【摘要】 骨质疏松（osteoporosis）是由于骨强度降低引起的骨折风险增加的慢性复杂代谢性疾病。随着社会老年化，骨质疏松已成为严重威胁我国人口健康的高发疾病。其主要危害是骨折风险增加。男性骨质疏松骨折发病率低于女性，但死亡率却明显高于女性，其发生机制尚不清楚。既往研究主要集中在女性绝经后骨质疏松，由于不同性别间骨质疏松的环境及遗传影响因素存在差异，非常有必要对男性骨质疏松相关表型进行研究。骨强度降低是骨质疏松发生的关键病理生理改变。骨强度反映骨的密度和骨质量。骨密度（bone mineral density, BMD）可反映约70%的骨强度情况，并且是目前骨质疏松诊断的金标准，所以骨密度也是最常使用的骨强度变量。髋部几何形态（hipgeometry）独立于BMD能反映骨强度。在男性和女性中都发现股骨颈剖面形态参数是独立的髋部骨折风险危险因子。骨面积与骨质疏松骨折发生率显著相关，有流行病学研究发现股骨颈骨面积是独立的髋部骨折风险因素。定量超声（Quantitativeultrasound，QUS）可无创检测骨结构及材料性状变化，可反映骨质量的变化，因此也可被认为是骨强度参数之一。上述骨强度参数均受到环境和遗传的协同调控。在峰值骨密度达到阶段，年龄及环境因素对BMD的累积影响效应相对较低，人群中的变异主要受遗传因素影响。研究峰值骨密度年龄段人群的骨强度遗传变异度有助于排除非遗传因素的影响，可能更易发现相关基因。随着人类基因组计划的完成、生物信息学、统计学等学科的发展，通过候选基因法（candidate gene approaches）、连锁分析（linkage analysis）和全基因组关联分析（genome-wide association study, GWAS)技术的应用，骨强度相关的许多候选基因得到了鉴定和明确。但这些研究在不同人群/人种中的结果存在差异或矛盾之处。这提示我们有必要在中国人群中研究/验证骨强度相关基因。既往研究提示参与调节骨骼的发育和成年期骨代谢的基因可能也是骨质疏松/骨强度表型的相关基因。成纤维生长因子受体（fibroblast growth factor receptors，FGFRs）1、2、3型在骨骼细胞表达并参与调节其生长发育和成年期骨代谢。多种FGFRs基因突变可导致人类骨骼遗传病，出现头颅、脊柱、胸廓、四肢长骨畸形。既往在欧美人群中发现FGFR1和FGFR2的基因多态性与BMD相关。这些研究集中在白人老年人群中，并且只检测了骨强度的一个变量，考虑到骨强度变量间遗传差异度、不同人种的遗传异质性、欧美与中国BMD环境影响因素的不同等，有必要在中国青年人群中检测FGFRs是否是BMD及其它骨强度变量的相关基因。我们既往发现这三种FGFRs可通过调节骨骼细胞生物学活性影响骨生物力学性能与骨密度。FGFRs也可通过调节磷代谢来影响骨密度。骨细胞合成的FGF23可在肾小管上皮细胞中特异性结合FGFR1、3、4并与Klotho形成复合物来抑制肾脏磷重吸收，调节血磷平衡，间接影响骨骼的矿化。FGF23也可依赖于维生素D受体（vitaminD receptor, VDR）调节肠道磷的吸收。但FGFRs是否参与肠道磷吸收尚不清楚。为此，本研究将首先在中国青年人群中检测FGFRs是否是BMD及其他骨强度变量的相关基因。另外利用动物模型探讨FGFR1在肠磷吸收中的作用及可能的机制，并检测其对骨强度的影响。主要实验方法：1.中国青年汉族男性骨强度环境影响因素及其与FGFRs SNPs相关性分析1.1研究对象人群资料采集共计812名某陆军汉族新兵被纳入本研究，测量身高（m）、体重（kg）、腰围（cm）、臀围（cm），同时计算BMI。问卷表记录出生日期、籍贯、既往生活习惯、家庭成员骨骼健康情况。1.2骨强度各变量测量采用Prodigy DXA测量第1到第4腰椎（L1-4）与左侧股骨颈（Femoral Neck, FN）、全髋（Total hip, TH）的BMD（g/cm2）、骨面积值（cm2）。左侧髋部形态参数由HSA软件（Prodigy，GE）基于骨密度检测结果进行自动分析获取：包括髋轴长度(hip axislength, HAL)(mm)、股骨颈角度(neck–shaft angle, Angle)(o)、骨骼强度系数(femoralstrength index, SI)、横截面积(cross-sectional area, CSA)(mm2)及横断面转动惯量(cross-sectional moment of inertia, CSMI)(mm4)。SONOST2000跟骨定量超声仪测量QUS参数，包括超声速率（Speed of sound, SOS）（m/s）、宽带超声衰减（Broadbandultrasound attenuation, BUA）（Db/MHz），并计算刚度系数（QUS-SI）。1.3ELISA法检测血清骨代谢指标（Tracp5b、BALP）。1.4基因组DNA提取后利用iMLDR法对FGFRs基因型进行检测。1.5统计分析采用SPSS11.0软件进行统计。计量资料以均数±标准差(x±SD)表示。Bartlett检验各计量资料方差是否齐性，Shapiro-Wilks检验检测各计量资料是否符合正态分布。卡方检验计算各SNP位点分布是否符合Hardy-Weinberg定律。PS v3.0.43软件计算检验效能。Haploview4.2计算LD及连锁不平衡系数r2。Stepwise回归分析各骨强度变量的协变量，年龄、身高、体重、腰围、臀围、BMI等为候选变量。ANOVA法分析各SNP位点基因型间年龄、身高、体重、腰围、臀围、BMI的差异。ANCOVA法分析各SNP位点基因型与各骨强度变量的相关性，分析各单体型与BMD的相关性。P<0.05表示差异显著。2. FGFR1对肠道发育和钙磷吸收的影响2.1FGFR1-vil条件性敲除小鼠繁殖及鉴定。2.2测量小鼠出生后体重、肠道长度、周径、肠绒毛高度、隐窝深度的变化。2.3AB-PAS染色计数杯状细胞和潘氏细胞。ALP染色检测肠细胞分化情况。BrdU检测细胞增生变化。TUNEL法检测肠绒毛上皮细胞凋亡变化。2.4免疫组化或免疫荧光检测肠绒毛上皮细胞FGFR1、NPT2b、P-ERK、P-JNK表达。2.5臧红固绿染色、TRAP染色检测小鼠骨骼病理形态及破骨细胞活性变化。micro-CT分析小鼠股骨近端松质骨和中段皮质骨变化。2.6检测不同磷含量饮食条件下小鼠粪便、尿液及血清钙磷含量。2.7定量PCR检测钙吸收相关基因（calbindin D9k、TRPV6、VDR）、磷吸收/重吸收相关基因（NPT2a、2b、2c）、FGFR1、IGF1在肠道或肾脏的表达情况。2.8WB检测肠绒毛组织P-P38及不同磷含量饮食条件下NPT2b蛋白含量变化。2.9共转染FGFR1siRNA及NPT2b报告基因质粒，检测其荧光素酶活性变化。2.10采用SPSS11.0软件进行统计。计量资料以均数±标准差(x±SD)表示。独立样本T检验比较组间差异。P<0.05表示差异显著。主要实验结果：1.中国汉族青年男性各骨强度变量的环境影响因素1.1BMD的环境影响因素年龄、体重、BMI、腰围与L1-4、FN及TH的BMD及BMC均显著正相关。三个部位峰值BMD均出现在22岁组。地域对BMD存在显著影响，西南地区组人群L1-4BMD显著低于北方地区组和东南地区组（P<0.05）。种族也是BMD的一个重要影响因素，本研究人群L1-4峰值BMD值均高于韩国及印度人群，髋部峰值BMD略低于印度人群，但仍高于韩国人群。军事训练后三个部位骨密度均显著高于训练前。训练后血清TRACP5b含量显著降低，BALP含量显著增加（P<0.001）。相关分析显示，TRACP5b和BALP与年龄负相关（P<0.001）。训练前后BALP变化率与L1-4BMD和全髋BMD变化率显著正相关（P<0.05）。1.2骨面积、髋部形态及跟骨QUS参数的环境影响因素年龄是影响腰椎骨面积的正性调节因素（P<0.05），但各年龄组间股骨颈及全髋骨面积均无显著差异。经过身体测量参数校正，发现年龄为CSA及CSMI的主要变异变量，21-23岁年龄组的CSA及CSMI显著高于低龄组（17-18岁）（P<0.05）。22及23岁年龄组SI显著高于18岁组（P<0.05）。HAL和Angle变异与年龄变化无关。年龄为三个QUS参数的主要变异变量（P<0.05）。SOS、BUA、QUS-SI的最大均值均出现在23岁年龄组。20-23岁年龄组的SOS、BUA显著高于低龄组（17-19岁）（P<0.05）。2.中国汉族青年男性各骨强度变量与FGFRs SNP相关性2.1BMD与FGFRs SNP相关性位于FGFR2基因exon1区域的rs1047111位点的基因多态性与腰椎（P=0.013）及股骨颈骨密度（P=0.0012）相关。FGFR1基因上rs2956724、rs6983315、rs6474354及rs4733930构成单倍体，但其4种单倍体型与腰椎、股骨颈和全髋BMD均无显著相关性。2.2骨面积与FGFRs SNP相关性位于FGFR1基因intron4的rs2288696(P=0.006)、位于intron3的rs2956724(P=0.028)及rs6474354(P=0.025)位点的基因多态性与股骨颈面积相关（表4-2）。位于FGFR15’-Flanking区域的rs10958704(P=0.043)位点的基因多态性与髋部面积相关。2.3髋部形态与FGFRs SNP相关性FGFR1基因内位于intron4区域的rs2288696基因多态性与髋部CSA和CSMI相关（P=0.049,0.035），单个G等位基因也存在显著效应，GG和GA基因型的CSA及CSMI均显著低于AA基因型。髋部SI与位于5’-Flanking的rs10958704基因多态性相关（P=0.027）。2.4跟骨QUS参数与FGFRs SNP相关性位于FGFR1基因intron2的rs4733946位点与跟骨SOS（P=0.017）及QUS-S（IP=0.025）显著相关，位于intron3的rs2956724与跟骨BUA（P=0.000）及QUS-SI（P=0.026）显著相关，位于5’-Flanking的rs10958704与跟骨SOS（P=0.023）显著相关。3. FGFR1通过MAPK-P38通路正性调节肠腔扩张及细胞凋亡3.1肠绒毛上皮细胞特异性敲除FGFR1降低肠周径定量PCR及免疫组化结果显示FGFR1在肠绒毛敲除效率在70%左右，肠绒毛上皮只检测到少量细胞残存表达FGFR1。FGFR1-vil小鼠（敲除小鼠）的体重及小肠长度无显著改变。FGFR1-vil小鼠回肠周径降低，回肠绒毛高度增加，但隐窝深度无显著差异。FGFR1-vil小鼠回肠IGF1表达降低，且与FGFR1mRNA表达水平正相关。提示FGFR1可能通过影响上皮下层IGF1表达调节肠腔扩张。3.2肠绒毛上皮细胞特异性敲除FGFR1不影响上皮细胞增生、分化但抑制凋亡BrdU掺入在两组间无显著差异，提示肠上皮敲除FGFR1不影响细胞增生。利用AB-PAS染色计数杯状细胞和潘氏细胞未见明显差异。ALP活性从近端小肠到远端小肠逐渐减弱，但两组间无明显差异。提示FGFR1不影响杯状细胞、潘氏细胞和肠细胞的分化。FGFR1-vil小鼠肠绒毛上皮细胞凋亡信号阳性细胞数量减少，提示FGFR1正性调节肠上皮细胞凋亡。3.3肠绒毛上皮细胞特异性敲除FGFR1后P-P38水平降低使用免疫组化检测到FGFR1-vil小鼠肠绒毛上皮细胞P-ERK与WT比较无显著差异，而P-JNK表达减少，但WT小鼠表达量也较低。WB结果发现FGFR1-vil小鼠肠绒毛上皮组织P-P38蛋白量显著减少。这提示FGFR1可能主要通过MAPK-P38通路影响肠道发育。4. FGFR1不依赖于食物磷含量和VDR途径，通过转录前负性调控NPT2b抑制肠道钙磷吸收4.1FGFR1-vil小鼠血磷正常，肠磷吸收增加，尿磷重吸收减少正常饮食（0.9%Pi），FGFR1-vil小鼠血清钙磷与WT小鼠比较无显著差异。利用0.5M NaH2PO4溶液灌胃小鼠制造急性高磷血症，FGFR1-vil小鼠小鼠血磷增加幅度更明显（P<0.05）。长期给予小鼠高磷饲料（1.25%Pi），FGFR1-vil小鼠小鼠血磷增高（P<0.05），而WT小鼠血磷仍维持在正常水平。正常饮食FGFR1-vil小鼠尿磷含量显著增加，同时粪便磷含量显著减少（P<0.05）。尿钙和粪便钙含量也有类似变化趋势。低磷饮食FGFR1-vil小鼠尿钙、尿磷含量增加，同时粪便钙含量降低。这提示FGFR1-vil小鼠肠磷吸收增加，同时尿磷重吸收减少以维持血磷稳定。4.2FGFR1-vil小鼠肠上皮NPT2b表达增加，肾皮质NPT2a表达降低FGFR1-vil小鼠回肠组织NPT2b mRNA表达水平和蛋白含量均增加。低磷饲料增加WT小鼠NPT2b蛋白量，而高磷饮食降低NPT2b蛋白量。FGFR1-vil小鼠NPT2b蛋白表达量对饮食P含量变化有类似反应趋势，但变化幅度较WT小鼠显著降低。转染FGFR1siRNA后CaCO2细胞中人NPT2b启动子活性增加。提示FGFR1不依赖于食物磷含量转录前负性调控NPT2b，FGFR1-vil小鼠肾皮质中，调节磷重吸收的NPT2a mRNA表达水平降低，但NPT2cmRNA表达水平无显著变化。提示肠磷吸收异常时可通过改变NPT2a表达来调节肾磷重吸收。FGFR1-vil小鼠肠上皮组织内钙吸收相关基因钙结合蛋白D9k（calbindin D9k）、VDR和V型瞬时感受器电位阳离子通道6（transient receptor potential cation channel,subfamily V, member6, TRPV6） mRNA表达水平无显著改变。这提示FGFR1不影响VDR介导的肠钙主动吸收。4.3FGFR1-vil小鼠骨骼表型未见异常臧红固绿染色显示FGFR1-vil小鼠同WT小鼠比较，病理形态无明显改变，TRAP染色显示破骨细胞活性也无明显差异。micro-CT分析股骨近端松质骨和中段皮质骨变化，两组小鼠间仍未见显著差异。全文结论：1.中国青年男性BMD受到运动、地域、种族等环境因素的影响，在22岁达到峰值BMD。2.年龄是获得峰值骨密度年龄段中国青年男性骨面积、髋部形态和QUS参数的主要影响因素。3. FGFR1是一个与骨面积、髋部形态和骨质量相关的多效应基因，FGFR2是BMD的相关基因。4.肠绒毛上皮表达的FGFR1通过激活MAPK-P38通路促进肠绒毛上皮细胞凋亡并正性调节肠腔扩张。5.肠绒毛上皮表达的FGFR1通过转录前负性调控NPT2b表达，不依赖食物磷含量和VDR途径抑制肠磷吸收。更多还原

【Abstract】 BackgroundOsteoporosis is a chronic complex metabolism disease characterized by decreasedbone strength that increases the risk of fracture. Osteoporosis has become an importantpublic health problem with increasing incidence rate as ageing population expanding. Themost harmful aspect of osteoporosis is the increased risk of fracture. About70%ofosteoporotic fractures occur in women, but the fracture-related mortality ratio is higher inmen. The mechanisms under this phenomenon remain unknown. Most of previousresearches focus on postmenopausal osteoporosis in women. Considering the genderdifferences of environmental and hereditary factors on osteoporosis, studying thephenotypes of osteoporosis in men is important for understanding the pathogenesis of maleosteoporosis.Decreased bone strength is the key point for the development of osteoporosis. Bonestrength reflects bone density and quality. Bone mineral density (BMD) is the gold criteriafor diagnosis of osteoporosis. BMD predicts about70%of variation of bone strength and isthe most frequently used trait in studying bone strength. Hip geometry can reflect bonestrength independent of BMD. Femoral geometry has been found being an independent riskfactor for hip fracture in both genders. Bone area is significantly associated with anincidence of osteoporotic fracture. Epidemiologic studies found that the area at femoralneck was an independent risk factor for hip fracture. Quantitative ultrasound (QUS) detectsbone architecture and material characters. QUS parameters can reflect bone quality and aredefined as bone strength traits as well.These bone strength traits are regulated by both environmental and hereditary factorswhile the main contributor is genetic variation. The accumulated effects of environmentalfactors on BMD are relatively weak in peak BMD achieved age stage. Studying geneticvariation of bone strength traits in this age stage helps eliminating the effects of*This study is supported by the National Natural Science Foundation of China (No.31000555). non-hereditary factors on BMD. Many candidate genes of bone strength have beenidentified by using candidate gene approaches, linkage analysis and genome-wideassociation studies (GWAS) along with the accomplishment of Human Genome Projectsand development of bioinformatics and statistics. But the results show more discrepancythan agreement among studies based on different populations or races. This suggests weshould study or validate the association genes of bone strength traits in Chinese population.Previous studies indicated the bone development and metabolism related genes couldbe associated with osteoporosis or bone strength traits. Fibroblast growth factor receptors(FGFRs)1,2,3express in bone tissues and regulate bone development and metabolism.Multiple point mutations in these genes lead to human skeletal hereditary diseases whichinvolved skull, spine, thoracic and limbs. The SNPs of FGFR1and FGFR2were foundassociated with BMD in white elder people. Considering the genetic variation among bonestrength traits, genetic heterogeneity of races and differences of environmental factorsbetween European and Chinese, it’s necessary to indentify the association between FGFRsSNPs and bone strength traits in Chinese young men.Our lab found the three FGFRs affected bone mechanical properties and BMD throughregulating the biologic activity of multiple types of bone cells. FGFRs can also influenceBMD by regulating phosphate metabolism. FGF23, mainly produced by osteocytes, bindsto FGFR1,3,4coupling with Klotho in kidney and depresses renal phosphate reabsorption.FGF23also have effect on intestinal phosphate absorption in a vitamin D receptor (VDR)dependent way. But it’s not clear the role of FGFRs in intestinal phosphate absorption.This study will firstly analyze the association between FGFRs SNPs and bone strengthtraits in Chinese young men, and investigate the effect of FGFR1on intestinal phosphateabsorption and the related mechanisms as well. Furthermore, the effect of alteration ofphosphate metabolism on bone strength will be detected.Main Methods1. Analysis of environmental and hereditary factors affecting bone strength traitsin Chinese young men of Han ethnicity1.1Subjects and body measurementA total of812Chinese male army recruits of Han ethnicity were involved in this study.Height (cm), weight (kg), waist circumference (WC)(cm), hip circumference (HC)(cm), BMI was measured. A questionnaire recorded age, daily calcium intake and activity, bonediseases of family members and so on.This study received ethical approval from the Ethics Committee of Daping Hospital,Third Military Medical University, China.1.2Bone strength traits measurementBMD (g/cm2) and bone area (cm2) at lumbar (L1-4), left femoral Neck (FN) and totalhip (TH) were measured using DXA (Prodigy, GE). Left hip geometry: hip axis length(HAL)(mm), neck–shaft angle (Angle)(o), femoral strength index (SI), cross-sectional area(CSA)(mm2) and cross-sectional moment of inertia (CSMI)(mm4) were assessed using HipStructure Analysis (HSA) software (Prodigy, GE). Speed of sound (SOS)(m/s), broadbandultrasound attenuation (BUA)(Db/MHz) and stiffness index (QUS-SI) were measuredusing QUS (SONOST2000) at left calcaneus.1.3Serum bone metabolism markers (Tracp5b, BALP) were quantified using ELISAkit.1.4The genotypes of FGFRs were determined by iMLDR assays.1.5Statistic analysisAll the data was represented as mean±SD. Bartlett tests were used to evaluate thehomogeneity of variance. Shapiro-Wilks tests were applied to analyze the distribution of allthe quantitative parameters. Chi-square tests were performed to evaluate theHardy-Weinberg equilibrium of genotypes. PS v3.0.43software was used to calculatedpower. Haploview4.2software was performed to conduct haplotypes of FGFR1. Pearson’scorrelation tests were performed to analyze the relationship between bone metabolismmarkers and BMD. Stepwise regression analyses were applied to bone strength traits usingage, height, weight, WC, HC, BMI as covariates. ANOVA analyses were used to test thedifferences of age, height, weight, WC, HC and BMI among genotypes. ANCOVA analyseswere performed to evaluate the association between genotypes of FGFRs and bone strengthtraits as well as haplotypes of FGFR1and BMD. P<0.05was considered significant.2. The role of FGFR1in intestinal development and absorption of calcium andphosphate2.1Animals: Conditional knockout of FGFR1in intestinal villus epithelial cells(FGFR1-vil) and the littermate controls (WT). 2.2Measurements of weight, intestinal length and circumference, villus height andcrypt depth after birth.2.3BrdU labeling was used to determine the proliferation of epithelial cells. Thenumbers of goblet cells and Paneth’s cells were counted after AB-PAS staining. ALPstaining was used to evaluate the differentiation of enterocytes. TUNEL assay wasperformed to detect the apoptosis of intestinal villus epithelial cells.2.4Immunohistochemistry or immunofluorescence was used to observe theexpressions of FGFR1, NPT2b, P-ERK and P-JNK in intestine.2.5Safranine O/fast green (SO/FG) staining was used to evaluate the changes of boneremodeling of tibia. The differentiation and activation of osteoclasts were detected withtartrate resistant acid phosphatase (TRAP) staining. micro-CT scanning was performed todetermine the bone content at proximal trabecular bone and midshaft cortical bone offemurs.2.6Serum, urine and fecal phosphate and calcium of mice were measured after beingfed with diet contained different contents of phosphate.2.7Real-time PCR was performed to detect the mRNA expression levels of calciumabsorption related genes (calbindin D9k, TRPV6, VDR), phosphate absorption/reabsorption related genes (NPT2a,2b,2c), FGFR1and IGF1in intestinal villus or kidneycortex tissues.2.8Western bolt was used to quantify the protein contents of P-P38and NPT2b inintestine after being fed with diet contained different contents of phosphate.2.9Luciferase activity was determined in CaCO2cells after cotransfection of FGFR1siRNA and NPT2b promotor plasmids.2.10Statistic analysis: Independent t-tests were performed to analyze the differencebetween FGFR1-vil and WT mice. P<0.05was considered significant.Main results1. Analysis of environmental and hereditary factors affecting bone strength traitsin Chinese young men of Han ethnicity1.1Environment factors affecting BMDAge, weight, WC and HC were significantly correlated with BMD and BMCrespectively (P<0.05). All of the peak BMD at three sites was found in the22years age group. Geographic factor contributed to the variation of BMD. The mean L1-4BMD insouthwestern group was significantly lower than that in northern group and southeasterngroup (P<0.05). We also found ethnic difference in Asia countries. The mean peak BMD atL1-4in this group was higher than that in both Korean and Indian men. But the mean peakBMD of Chinese men at hip was lower than of Indian men.Military training increased BMD at three sites. Comparing with baseline, serum levelsof TRACP5b (P<0.05) were significantly decreased while elevated levels of BALP(P<0.001) were detected after training. Serum concentration of TRACP5b and BALP werenegatively related with age (P<0.001). The variation ratio of BALP was positively relatedwith that of BMD at L1-4and TH (P<0.05).1.2Environment factors affecting bone area, hip geometry and QUS parametersAge was the positive regulator of bone area at L1-4(P<0.05). There was no significantdifference of bone area at HN and TH among age groups.Age was the major factor contributed to variations of CSA and CSMI withanthropometric traits as covariants. All the mean CSA and CSMI in21-23years age groupsare significantly higher than that in17or18years age group (P<0.05). Both mean SI in22and23years age group were significantly higher than that in18years age group (P<0.05).There was no significant relationship between age and HAL or Angle.Age was significantly related with variation of three QUS parameters (P<0.05). Thepeak values of SOS, BUA and QUS-SI were found in23years age group. All the mean SOSand BUA in20-23years age groups are significantly higher than that in17-19years agegroups (P<0.05).2. Association between FGFRs SNPs and BMDrs1047111located at exon1of FGFR2was associated with BMD at L1-4(P=0.013)and FN (P=0.0012). A four loci haplotype in FGFR1constructed by rs2956724, rs6983315,rs6474354and rs4733930had no significant association with BMD.2.1Association between FGFRs SNPs and bone areaIn FGFR1gene, rs2288696located at intron4, rs2956724and rs6474354located atintron3were associated with bone area of FN while rs10958704in5’-Flanking wasassociated with that of TH (P=0.006,0.028,0.025,0.043respectively).2.2Association between FGFRs SNPs and hip geometry Two SNPs of FGFR1was associated with hip geometry parameters: rs2288696inintron4with CSA and CSMI while rs10958704in5’-Flanking with SI (P=0.049,0.035,0.027respectively).2.3Association between FGFRs SNPs and QUS traitsAll the associated SNPs of QUS parameters were found in FGFR1. rs4733946inintron2was associated with SOS (P=0.017) and QUS-SI (P=0.025). rs2956724in intron3was associated with BUA (P=0.000) and QUS-SI (P=0.026) while rs10958704in5’-Flanking with SOS (P=0.023).3. The role of FGFR1in intestinal development and absorption of calcium andphosphate3.1Conditional knockout FGFR1in intestinal epithelial cells lead to reducedcircumference of intestineThe efficiency of gene knockout was confirmed using real-time PCR and IFC.Expression levels of FGFR1in intestinal villus of FGFR1-vil mice were decreased about70%compared with that of WT mice. The body weight and intestine length had no statisticdifference between groups.Compared with WT mice, FGFR1-vil mice exhibited decreased intestinecircumference and increased villus height, but no significant change of crypts depth.Depressed mRNA expression levels of IGF1, known as a regulator of intestinal diameter,were detected in intestines of FGFR1-vil mice. The mRNA expression levels of IGF1andFGFR1were positively related which implied that FGFR1may mediate intestinal diameterexpansion through regulating expression of IGF1in subepithelial myofibroblasts.3.2FGFR1enhanced apoptosis but not proliferation or differentiation ofintestinal epithelial cellsNo remarkable difference in location and numbers of BrdU positive cells was observedbetween FGFR1-vil and WT mice which suggested FGFR1inactivation didn’t seem toaffect epithelial cells proliferation. The numbers of goblet cells and Paneth’s cells inFGFR1-vil mice were not significantly different from that in WT mice. ALP activity whichreflected numbers of enterocytes in FGFR1-vil mice was similar to that in WT mice. Thissuggested that intestine-specific FGFR1inactivation may have no effect on goblet cell,Paneth’s cell and enterocyte differentiation. TUNEL staining found a significant reduction of apoptosis in FGFR1-vil mice which implied FGFR1positively regulated epitheliaapoptosis.3.4Conditional knockout of FGFR1in intestine epithelial cells depressedactivation of MAPK-P38pathwayIFC staining found similar activated signals of P-ERK between two genotypes. Almostno P-JNK positive cells were observed in FGFR1-vil intestine while few ones were foundin intestine of WT mice. WB detected decreased P-P38protein levels in intestinal epitheliaof FGFR1-vil mice than that in WT mice. This suggested MAPK-P38may be the mainpathway contributing to the effects of FGFR1on intestinal development.4. FGFR1inhibited intestinal phosphate absorption through pre-transcriptionalregulating NPT2b expression in the diet phosphate-and VDR-independent way4.1Conditional knockout of FGFR1in intestinal epithelial cells maintainednormal serum phosphate by increasing intestinal absorption and reducing renalreabsorptionCompared with WT mice, serum calcium and phosphate in FGFR1-vil mice had nosignificant change in normal diet condition (0.9%Pi).Higher increased ratio of serumphosphate was detected in FGFR1-vil mice than that in WT mice in a hyperphosphatemiamodel induced by intragastric administration of0.5M NaH2PO4(P<0.05). In long-term highphosphate diet condition (1.25%Pi), elevated serum phosphate was found in FGFR1-vilmice while normal serum phosphate was detected in WT mice (P<0.05). Significantlyincreased urine phosphate and decreased fecal phosphate were detected in FGFR1-vil witheither normal diet or low phosphate diet (P<0.05). This suggested that increased intestinalphosphate absorption in absence of FGFR1lead to reduced renal phosphate resorptionwhich helped normal serum phosphate maintenance.4.2Deletion of FGFR1in intestine upregulated NPT2b expression in intestinalepithelia and downregualted NPT2a expression in renal cortexBoth increased mRNA and protein expression levels of NPT2b were detected inFGFR1-vil mice. Low phosphate diet upregulated protein expression levels of NPT2b inWT mice. High phosphate diet had opposite effects on that in WT mice. The similarresponses were found in FGFR1-vil mice. But the variation ratio of NPT2b in FGFR1-vilmice was significantly lower than that in WT mice. The promoter activity of NPT2b was reduced after transfection of FGFR1siRNA in CaCO2cells. This suggested FGFR1negatively pre-transcriptionally regulatd NPT2b expression in a diet phosphate-independentway.Increased mRNA expression level of NPT2a, known as a phosphate transporter locatedat proximal tubules, was detected in kidney cortical tissues of FGFR1-vil mice. The mRNAexpression level of NPT2c was not significantly different from that in WT mice. Thissuggests abnormal intestinal phosphate absorption may regulate renal phosphatereabsorption by altering NPT2a mRNA expression.No significant difference of mRNA expression levels of calcium absorption relatedgenes (calbindin D9k, VDR and TRPV6) was found between FGFR1-vil and WT mice.This suggested FGFR1may not contribute to intestinal active absorption of calcium.4.3Conditional knockout of FGFR1in intestinal epithelial cells didn’t affect bonephenotypesSO/FG staining found no abnormal pathologic change in FGFR1-vil tibias. TRAPstaining indicated the differentiation and activation of osteoclasts were similar betweenFGFR1-vil and WT mice. micro-CT traits at proximal trabecular bone and midshaft corticalbone of femurs had no significant changes in FGFR1-vil compared with that in WT mice.Conclusions1. BMD of Chinese young men is affected by exercise, geography and race. PeakBMD at three sites (L1-4, FN and TH) are observed in22years age group.2. Age is the major factor influencing bone area, hip geometry and QUS parameters ofChinese men in the period of peak BMD achieved.3. FGFR1is a multiple effect gene associated with bone area, bone geometry and bonequality. FGFR2is the association gene of BMD in Chinese young men.4. FGFR1increases apoptosis of intestinal epithelial cells and positively regulatesintestinal diameter expansion by activating MAPK-P38pathway.5. FGFR1inhibits intestinal phosphate absorption through pre-transcriptionallydepressing NPT2b expression in the diet phosphate-and VDR-independent way.更多还原