【作者】 彭彦孟；

【导师】 吴军；

【作者基本信息】 第三军医大学 ， 外科学， 2013， 博士

【摘要】 天然杀伤细胞（NK细胞）是天然免疫系统中非常重要的一类细胞毒性淋巴细胞，它是免疫系统中的第一道防线。NK细胞是一群特殊的细胞，它不同于获得性免疫在发挥功能的时候是需要抗体和主要组织相容性复合体（MHC）的。在哺乳类动物受到病毒感染或者肿瘤侵袭的时候，NK细胞会非常迅速地产生反应。NK细胞的活化是由激活性受体和抑制性受体之间的平衡所决定的，如果激活信号占主导地位NK细胞就会被活化，反过来如果抑制信号占主导那么NK细胞将受到抑制。NKG2D和NKp44是与杀伤肿瘤密切相关的NK细胞的激活性受体成员。在NK细胞的胞浆里有很多溶细胞颗粒，当NK细胞活化时，它们会释放到细胞与细胞的接触面。这些颗粒包括溶细胞的蛋白，比如穿孔素（perforin）和颗粒酶（granzyme），它们在促使靶细胞的死亡的过程中发挥重要作用。CD107a又叫做LAMP-1是一种溶酶体相关的膜蛋白，属于糖蛋白类，当NK细胞脱颗粒时，原来处于管腔膜的CD107a将暴露在细胞表面的突触上。Alter等发现当NK细胞被靶细胞刺激的时候CD107a也会在其细胞表面检测到，因此作为NK发挥杀伤作用时候脱颗粒的标志物。细胞毒T淋巴细胞相关抗原4（CTLA4）是表达在激活的T细胞表面而作负调节其活性的一种抗原。而CTLA4Ig是由人CTLA4胞外区和人IgG1的重链恒定区组成的融合蛋白。它通过竞争性结合抗原递呈细胞（APC）上的CD80、CD86分子来阻碍T细胞上的CD28同CD80、CD86结合给T细胞传递共刺激信号，从而成功诱导T细胞的耐受。很多对CTLA4Ig的体外和体内的研究表明其能够治疗自身免疫性疾病和控制器官排斥反应。CTLA4Ig的商业化产品Abatacept和Balatacept已被美国FDA批准治疗诸如关节炎之类的自身免疫性疾病，同时还可以用于控制移植排斥反应。但是，在使用了CTLA4Ig治疗的病人中，他们的肿瘤发生率和感染率并没有如预期的显著升高，这一点引起了我们的关注。为什么当CTLA4Ig作用于T细胞，使其活性受到抑制时，抗肿瘤和抗感染的免疫能力依然维持而没有减弱？同时，已被大家公认，在对肿瘤的监视和抗感染中，获得性免疫和天然免疫系统同时发挥着作用，那么是否在这个过程中，当获得性免疫被抑制的同时天然免疫系统发挥了作用呢？这一点还有待验证。Grohmann等人发现CTLA4Ig能够通过APC表面的B7分子影响APC的功能。与此同时，有研究证实静息的NK细胞可以表达CD86分子，而当NK细胞被激活时，可以同时表达能够被CTL4Ig结合的CD80和CD86分子。这些前人的研究结果促使我们对探索CTLA4Ig是否能通过调节NK细胞的功能从而在机体的免疫调节中起到作用产生了兴趣。在我们的研究中，体内和体外实验数据表明CTLA4Ig能够通过激活NK细胞增强其杀伤靶细胞的能力，同时，也提出了在杀伤靶细胞的过程中可能存在的新的分子机制。我们研究的主要发现和结论总结如下：一．体内试验中CTLA4Ig减少了黑色素瘤B16F0的转移和延长了荷瘤小鼠的存活时间：为了研究CTLA4Ig在肿瘤免疫中的作用，我们采用性别和年龄匹配的B6小鼠作为研究对象。在第0天尾静脉接种B16F0细胞，然后分别在第0天、第3天和第6天尾静脉注射CTLA4Ig和对照IgG。以小鼠的生存曲线和其肺部肿瘤的转移数量为观察指标。结果表明，CTLA4Ig实验组小鼠的生存时间明显长于对照组，CTLA4Ig实验组小鼠肺部肿瘤的转移数量也明显少于对照组。二．体内实验中CTLA4Ig增强小鼠NK细胞的毒性：1. CTLA4Ig通过作用于NK细胞来减少SCID小鼠体内黑色素瘤的转移；已经被证实CTLA4Ig是通过结合B7分子从而产生对获得性免疫中T细胞的调节，文献和我们的实验表明NK细胞中也表达B7分子，我们假设CTLA4Ig可能会对NK细胞产生调节，而NK细胞又是肿瘤细胞的强有力的效应细胞，我们采用SCID小鼠和B6小鼠作为研究对象。实验结果表明，CTLA4Ig处理的SCID荷瘤小鼠能够有效抑制肿瘤的转移；根据以上实验的提示，NK细胞可能在肿瘤转移中起到了非常重要的作用。因此，为了进一步明确可能存在这种作用，我们又用PK136清除B6小鼠体内的NK细胞，结果显示CTLA4Ig处理组和对照组肿瘤的转移数量没有明显差异，这又证明了CTLA4Ig是通过增强NK细胞的毒性来抑制了肿瘤的转移。2. CTLA4Ig增强了肿瘤侵润区域NK细胞的毒性；为了检测在CTLA4Ig作用后NK细胞的细胞毒性的变化，在我们的实验研究中，肿瘤接种后第10天处死小鼠，分离小鼠肺脏的淋巴细胞，采用磁珠分选技术纯化NK细胞，测试其毒性分子的表达和杀伤靶细胞能力的强弱。结果显示，CTLA4Ig组肿瘤侵润部位的NK细胞毒性明显高于IgG对照组的NK细胞毒性，因此提示CTLA4Ig能够增强NK细胞的毒性来减少肿瘤的转移。3. CTLA4Ig能够提高肿瘤侵润部位NK细胞的CD107a和perforin的表达；因为NK细胞脱颗粒的标志物CD107a和效应分子穿孔素是NK细胞毒性的两个重要指标。因此，检测了肿瘤侵润部位NK细胞CD107a和perforin的表达，结果显示CTLA4Ig处理组其CD107a和perforin的表达明显高于对照组。这提示肿瘤侵润部位的NK细胞可以被CTLA4Ig激活从而显著提高其杀伤靶细胞的能力。三、体外实验中CTLA4Ig能够提高NK细胞毒性对靶细胞的杀伤；1.在体外CTLA4Ig能够提高小鼠脾脏NK细胞毒性对靶细胞的杀伤；为了在体外验证CTLA4Ig对NK细胞毒性的作用，我们先分离B6小鼠脾脏的NK细胞，然后用磁珠分选技术纯化NK细胞，随后分别和加入了CTLA4Ig与对照IgG的YAC-1细胞共培养，然后检测NK细胞对YAC-1细胞的杀伤。实验结果显示CTLA4Ig实验组的NK细胞杀伤YAC-1细胞的能力比对照组明显增强。2.在体外CTLA4Ig能够提高荷瘤小鼠肿瘤侵润部位NK细胞的毒性；实验中取出B6荷瘤小鼠的肺脏，随后用磁珠分选技术纯化分离肺脏肿瘤侵润的NK细胞，然后按设定比例与YAC-1肿瘤细胞共培养，检测NK细胞对YAC-1的杀伤作用，通过实验组CTLA4Ig和对照组IgG的NK细胞杀伤作用的比较，结果显示CTLA4Ig能够显著提高荷瘤小鼠肿瘤侵润部位NK细胞的细胞毒性。3.在体外CTLA4Ig能够提高人外周血NK细胞的细胞毒性：通过分离人外周血单个核细胞（PBMC），用磁珠分选技术纯化其NK细胞，然按设定比例与K562肿瘤细胞共培养，最后检测NK细胞对K562的杀伤作用。通过比较实验组CTLA4Ig和对照组IgG中NK细胞对靶细胞的杀伤，结果显示CTLA4Ig能够显著提高人外周血NK细胞的细胞毒性。4.在体外CTLA4Ig能够显著提高NK92MI细胞的细胞毒性；因为通过磁珠分选的方法分离出来的NK细胞并没有达到绝对纯化，其他细胞可能会干扰实验的结果，因此我们采用了人NK92MI这种NK细胞系，来进一步观察CTLA4Ig对NK细胞毒性的影响。同时，NK92MI细胞上几乎不表达CD16这种，所以可以进一步排除CTLA4Ig中Fc段可能带来的ADCC作用的干扰。实验中，将NK92MI细胞与靶细胞K562共培养，通过检测实验组CTLA4Ig和对照组IgG处理的NK92MI细胞对K562细胞的杀伤作用，结果显示CTLA4Ig能够显著提高NK92MI的细胞毒性。四、CD86分子是CTLA4Ig介导的NK细胞毒性增强的激活性受体；1. NK细胞上CD86分子的表达能够在靶细胞和细胞因子刺激的环境中显著提高；根据文献报道，CTLA4Ig能够高亲和力的结合CD80和CD86分子，我们猜想CTLA4Ig可能是通过与它们结合起到增强NK细胞的功能从而杀伤肿瘤细胞的。为了进一步明确NK细胞表面有CD80和CD86的表达，我们检测了在正常生理条件下它们的表达，流式结果提示在小鼠脾脏NK细胞中大约有6%的CD86的表达，但是没有CD80的表达；在体外与YAC-1共培养或者IL-15刺激下的NK细胞CD86的表达有显著提高。另外，我们检测了在荷瘤小鼠B16肿瘤侵润部位的NK细胞表面CD80和CD86的表达，结果提示黑色素瘤B16细胞能够显著提高CD86分子的表达,CD80的表达却没有明显提高。以上结果提示在肿瘤环境中NK细胞能够显著提高其表面CD86的表达。2. CTLA4Ig能够显著提高NK细胞激活性受体NKG2D和NKp44的表达水平；总所周知，NK细胞的杀伤功能与其表面的激活性受体有密切的联系，同时NK细胞在杀伤靶细胞的过程中的一条途径是ADCC作用。为了排除CTLA4Ig可能介导的ADCC作用，同时为了验证CD86在NK细胞中的作用，我们采用了NK92MI这种能够自身高表达CD86分子的细胞系作为效应细胞，但同时其表面无CD32、CD64的表达，而只有少量CD16这种低亲力FcR的表达。在体外实验的结果提示，CTLA4Ig能够显著提高NK92MI细胞的激活性受体NKG2D和NKp44的表达水平。3. CD86可能在增强NK细胞对靶细胞的杀伤作用里起到类似激活性受体的作用；为了排除抗CD80抗体和抗CD86抗体这两种封闭性抗体对NK细胞毒性的影响，我们将抗CD80抗体和抗CD86抗体分别加入到NK92MI培养体系中，然后检测NK细胞对肿瘤细胞的杀伤作用，发现没有显著变化。然后，我们用抗CD86抗体和CTLA4Ig进行竞争实验。结果显示，当我们把抗CD86封闭抗体和CTLA4Ig同时加入到NK92MI和K562肿瘤细胞的共培养体系中，然后检测NK92MI细胞对肿瘤细胞的杀伤作用，发现其杀伤作用与对照组有部分减弱；当我们把抗CD86封闭抗体提前CTLA4Ig2小时加入NK92MI和K562肿瘤细胞的共培养体系中，目的先让抗CD86封闭抗体先占据CD86分子结合位点，然后检测其对肿瘤细胞的杀伤时，我们发现由CTLA4Ig产生使NK细胞毒性增强的作用被逆转了。这些结果提示，CD86可能在CTLA4Ig增强NK细胞抗肿瘤毒性的过程中起到关键作用。总结：我们验证了CTLA4Ig能够显著减少B16肿瘤的转移以及延长荷瘤小鼠的生存时间；同时观察到在清除了NK细胞的小鼠注射CTLA4Ig后，并没有改善肿瘤细胞的转移；在体内CTLA4Ig作用后能够显著提高NK细胞毒性相关的CD107a和perforin的表达。另外，我们在体内外实验中验证了在肿瘤环境的刺激下，NK细胞能够提高CD86分子的表达；高表达CD86分子的人NK细胞系，体外实验在CTLA4Ig的作用下能明显增强其杀伤靶细胞的能力。最后我们用抗体封闭实验，验证了CTLA4Ig可能是通过CD86分子起到使NK细胞毒性增强的作用。综上所诉，我们发现了一条由CTLA4Ig介导的增加NK细胞杀伤能力的新途径，证明了CD86分子在NK细胞表面起到了类似激活性受体的作用，而在与CTLA4Ig结合后起到增强NK细胞杀伤能力的新机制。更多还原

【Abstract】 Natural Killer cells (NK cells) are a very important type of cytotoxic lymphocytes ininnate immunity which is the first barrier of the immune system. NK cells are a unique,different from acquired immune system which needs antibody and MHC to act. They makevery fast responses when mammality was infected by virus or invaded by tumor cells.NK cell activation is determined by the balance of activating and inhibitory receptorstimulation. If the activating signal is more prominent then NK cell will be activated,similarly if the inhibitory receptor signaling dominants then NK cell activity will be inhibited.NKG2D and NKp44, the two activating NK receptors, are involved in tumor cell lysis byactivated NK cells. NK cells have a lot of cytolytic granules in their cytoplasm. When NKcells are activated, they release these granules to the site of cell–cell contact. These granulescontain cytolytic proteins, such as perforin and granzyme, that are involved in inducingtarget cell death. Lining the membrane of the granules is the lysosomal-associatedmembrane protein, LAMP-1or CD107a. CD107a is a glycoprotein representingapproximately50%of the proteins in the lysosomal membrane. During degranulation,CD107a on the luminal membrane is exposed in the immunological synapse and thenantibody could bind. Alter et al. showed that CD107a can also be detected on the surface ofNK cells following stimulation with target cells, so CD107a is considered as a marker ofdegranulation.Cytotoxic T-lymphocyte-associated antigen4(CTLA4) is expressed on the surface of Tcells hours or days after activation and functions as a negative regulator of T cell activation.CTLA4Ig is constructed by genetically fusing the external domain of human CTLA4to theheavy-chain constant region of human IgG1. CTLA4Ig has been shown to induce T celltolerance by binding to both CD80and CD86on antigen-presenting cells (APC), whichprevents the binding of CD28to CD80and CD86to deliver costimulatory signals to T cells.Many in vitro and in vivo studies demonstrated that CTLA4Ig could be used for controlling ofautoimmune diseases and allograft rejection. Both commercial products of CTLA4Ig, Abatacept and Belatecept (Bristol-Myers Squibb), have been approved by the FDA as atherapy for treating autoimmune diseases such as arthritis, and for controlling graft rejection.One interesting aspect of CTLA4Ig is that patients treated with this fusion proteinexperienced not much higher incidence of tumors and infectious episodes than controls. Weare interested in why immunocompetence against tumors remains while T cell activation issuppressed by CTLA4Ig. It is believed that surveillance for tumorigenesis and anti-infectionare mediated by both adaptive and innate immune cells. Whether the innate immunity is intactwhen the adaptive immunity is suppressed by CTLA4Ig has not been examined. Grohmannet.al. found that CTLA4Ig could influence APC function via the interaction with B7molecules on APC. Moreover, recent studies demonstrated that resting NK cells couldexpress CD86and that activated NK cells express both CD80and CD86receptors that mightbe bound by CTLA4Ig. These results prompted us to study the possibility that CTLA4Ig canregulate tumor surveillance and anti-infection immune by modulating NK cell function.In the present study, we reveal the novel function of CTLA4Ig via enhancement of NKcell cytotoxicity to target cells in vivo and in vitro and also show the underlying moleculemechanism was involve in this activity. Our main findings and conclusions of this study aresummarized as follows:1. CTLA4Ig reduces melanoma metastasis in vivo and prolonged host survival.To define the role of CTLA4Ig in tumor immunity, sex-and age-matched B6mice wereinjected with B16F0cells on Day0, followed by intravenous injection of either CTLA4Ig(n=10) or isotype control IgG (n=10) on Days0,3and6, respectively. The survival time andthe melanoma lung metastasis of each animal were monitored. The results showed that micetreated with CTLA4Ig had significantly longer survival time and significantly lower numbersof lung metastatic tumor nodules than those treated with control IgG. These results suggestthat CTLA4Ig contribute to the anti-tumor protection.2. CTLA4Ig enhances NK cell cytotoxicity in vivo.1). CTLA4Ig reduced the melanoma metastasis in SCID mice and B6mice with NKcells depletion.It has been demonstrated that CTLA4Ig inhibits the function of the adaptive immunesystem by blocking B7/CD28interactions. We hypothesized that innate immunity, but notadaptive immunity, plays a critical role in the CTLA4Ig-mediated anti-tumor activity. To test our hypothesis, sex-and age-matched SCID mice were used, and the results showed thatCTLA4Ig could also significantly inhibit the lung metastasis of melanoma tumors in SCIDmice. Because NK cells play an important role in tumor surveillance in the body, we furtherexamined whether NK cells were involved in the CTLA4Ig-mediated anti-tumor activity.Therefore, we used the PK136depleting antibody to deplete NK cells in mice. The resultsshow that depletion of NK cells results in the abolishment of the CTLA4Ig killing target celleffect and indicate that NK cells might play some role in the CTLA4Ig-mediated killingactivity.2). CTLA4Ig enhances the tumor infiltrating NK cell cytotoxicity in vivo.The ability of NK cells to kill tumor cells are critical immune surveillance mechanismsfor eradicating developing tumors. To assess the NK cells cytotoxicity in NK-dependent-CTLA4Ig-anti-tumor activity in vivo, the B16melanoma mice treated with either CTLA4Igor control IgG were sacrificed10days after tumor inoculation; magnetic-activated cellsorting was used to isolate the infiltrating NK cells from lung tissue for the analysis ofcytolytic activity and cytokine production. The results showe that tumor-infiltrating NK cellsfrom mice treated with CTLA4Ig possess significantly higher cytolytic activity than thosetreated with control IgG. These results suggest that CTLA4Ig retards tumor metastasis byenhancing the NK cell cytotoxicity to tumor cells in vivo.3). CTLA4Ig induces higher number of perforin-producing and CD107a-positive NKcells in vivo.Because the degranulation marker CD107a and effector molecules perforin are alsoclosely associated with NK cell cytotoxicity against tumor cells, we examined the expressionof these molecules in tumor infiltrating NK cells of mice treated with either CTLA4Ig orcontrol IgG. The results showed that there were significantly higher numbers ofperforin-producing and CD107a-positive NK cells in CTLA4Ig-treated mice than controlIgG-treated mice. These results clearly show that the cytotoxicity of the infiltrating NK cellsis markedly enhanced in the process of CTLA4Ig-mediated killing activity.3. CTLA4Ig stimulats NK cell cytolytic activity in vitro.1). CTLA4Ig enhances mouse spleen NK cell cytotoxicity in vitro.To assess the direct effect of CTLA4Ig on NK cells, we first examined the role ofCTLA4Ig in NK cell cytotoxicity against tumor cells in vitro. Mouse splenic NK cells were purified by MACS and were co-cultured with either CTLA4Ig or control IgG to analyze thecytolytic activity to YAC-1cells. As compared to control IgG, CTLA4Ig could significantlyenhance NK cell cytotoxicity to YAC-1cell in vitro.2). CTLA4Ig enhances tumor-infiltrating NK cells cytotoxicity ex vivo.Then, we checked the effect of CTLA4Ig on NK cell cytotoxicity to tumor cells ex vivo.The tumor-infiltrating NK cells were purified from lungs of mice bearing B6melanomatumor and co-cultured with either CTLA4Ig or control IgG to analyze the cytolytic activity toYAC-1cells. As compared to control IgG, CTLA4Ig significantly enhanced the cytotoxicityof the infiltrating NK cells ex vivo.3). CTLA4Ig enhances human NK cells cytotoxicity in vitro.Then, we also checked the effect of CTLA4Ig on human NK cell cytotoxicity againsttumor cells in vitro. The human NK cells were purified from PBMC and co-cultured witheither CTLA4Ig or control IgG to analyze the cytolytic activity to K562cells. As compared tocontrol IgG, CTLA4Ig significantly enhanced the cytotoxicity of human NK cells in vitro.4). CTLA4Ig enhances the NK92MI cell cytotoxicity in vitro.Because the NK cells isolated by using MACS were not highly purified (the purity wasless than90%), it was possible that the CTLA4Ig was acting on other cell types in the culture.To address this issue, we used a human NK cell line, NK-92MI, as the effector cells to checkthe possible direct role of CTLA4Ig in regulating NK function. The results showed thatcytotoxicity of NK-92MI cells to K562tumor cell was significantly greater increased in thepresence of CTLA4Ig than that in the presence of control IgG in vitro. The results suggestthat CTLA4Ig might directly stimulate NK cell cytotoxicity.4. CD86is the target molecule in CTLA4Ig-mediated enhancement of NK cytolyticactivity.1). NK cells could significantly increase the expression of CD86upon tumor cellstimulation.Based on the fact that CTLA4Ig binds with high affinity to CD80/CD86, wehypothesized that CD80or CD86might play in role in the ability of CTLA4Ig to enhance NKcell functions against tumor metastasis. To understand the expression of CD80and CD86onphysiological NK cells of B6mouse, the CD80/CD86expression on NK cells was examinedby flow cytometry. The results showed that prior to activation, mouse NK cells had6% expression of CD86and little CD80expression. The expression of CD86on NK cells wassignificantly increased following activation with both YAC-1tumor cells and cytokine IL-15in vitro. Furthermore, we detected the expressions of CD86and CD80on the tumor-infiltrating NK cells in vivo10days after injection of B16melanoma cells. The data showedthat injection of B16melanoma cells could significantly increase the expression of CD86, butnot CD80, on infiltrating NK cells. These results indicated that NK cells could significantlyincrease the expression of CD86upon stimulation by tumor cells and that CTLA4Ig activatesNK cells possibly via ligation of CD86on the activated NK cells.2). CTLA4Ig induces higher level of expression of both NKG2D and NKp44.To test the role of CD86in regulating NK cell function, we used NK92-MI thatspontaneously expressed high levels of CD86, but not CD80, as effector cells for ligationwith CTLA4Ig in vitro. The results showed that the ligation of NK-92MI cells withCTLA4Ig significantly induced higher expression level of NK effector molecules NKG2Dand NKp44.3). CD86rather than CD80on NK cells might be involved in the enhancement of NKcell cytotoxicity in competition experiments.To exclude that the anti-CD80and anti-CD86antibody may affect the cytotoxicity ofNK cells, we added anti-CD80or anti-CD86antibody into NK-92MI cell culture system, asexpected, no effect of anti-CD80and anti-CD86antibody on NK-92MI cell cytotoxicity wasfound. Then, an anti-CD86antibody was used to compete for CD86molecules on NK-92MIwith CTLA4Ig. The results showed that when anti-CD86antibody, together with CTLA4Ig,was added into the NK-92MI cell culture system, the CTLA4Ig-mediated NK-92MI cellactivation was partly decreased. If anti-CD86antibody was added2hours earlier thanCTLA4Ig, which meant that the anti-CD86antibody had pre-occupied the CD86moleculeson NK cells, CTLA4Ig-mediated NK-92MI cell cytotoxicity was completely abolished.These data from the competition experiments clearly demonstrates that CD86rather thanCD80on NK cells is involved in the enhancement of NK cell cytotoxicity to target cells.Conclusion: we showed that brief administration of CTLA4Ig significantly reducedtumor metastasis and prolonged the survival of host mice bearing B16melanoma. Depletionof NK cells prior to CTLA4Ig administration eliminated the CTLA4Ig-mediated killingactivity. CTLA4Ig can significantly enhance NK cell cytotoxicity to target cells via up-regulation of NK cell effecter molecules CD107a and perforin in vivo. In addition, wedemonstrated that, upon activation, NK cells could significantly increase the expression ofCD86both in vitro and in vivo, and ligation of CD86with CTLA4Ig significantly increasedthe ability of NK cells to kill target cells. Furthermore, a human NK cell line that expressedhigh level of CD86, but not CD80, was directly activated by CTLA4Ig so that killing oftargets was enhanced; this enhanced killing could be inhibited by blocking CD86. Ourfindings uncover a novel function of CTLA4Ig in innate immunity and suggest that CD86onNK cells is an activating receptor and closely involved in the CTLA4Ig-mediatedcytotoxicity activity.更多还原