黄 欣 论文题目：含ITIM的新型膜蛋白分子DC–LLIR及其跨膜缺失突变体的克隆与功能研究 作者简介：黄 欣，女，1971年12月出生，1997年09月师从于第二军医大学曹雪涛教授，于2000年06月获博士学位。
摘 要
树突状细胞（Dendritic cells, DCs）的功能特点是能激发初始T细胞（naive T cells）免疫反应，因此DCs是目前所发现的最重要的抗原提呈细胞 （Antigen presenting cells, APC）。其生物学功能特点表现为捕获抗原的高效性、很强的迁移能力以及高表达能激活T细胞的表面分子，但目前对其生物学功能的作用机理尚未完全阐明；已知DCs 的功能特点与其不成熟（immature）和成熟（mature）状态密切相关：不成熟的DCs具有高效的抗原摄取和处理功能，但对T细胞的激活作用较弱；而在成熟状态下，DCs的主要功能是抗原提呈，此时DCs的抗原摄取能力已大大降低。不成熟状态的DCs可以被炎性刺激因素诱导而趋于成熟，钥孔血蓝蛋白（Keyhole limpet hemocyanin，KLH）作为一种强抗原可以诱导人DCs发生从不成熟状态到成熟状态的转变，DCs同时获得了激发T细胞免疫的能力。细胞功能差异的根本原因在于细胞基因表达的差异，通过差异筛选不同成熟状态DCs差异表达的基因可以获得与DCs功能相关的基因信息，从而有助于深入认识DCs生物学功能的分子基础及相关分子本身的作用原理。
筛选差异表达基因的技术分为两大类：差示PCR技术（Differential
display polymerase chain reaction，DD-PCR）和减数杂交技术（Subtracting hybridization）。减数杂交技术在实际应用中要比差示PCR技术高效得多，并且随着相关技术的发展，其在克隆差异表达及低丰度表达的基因方面得到了更加广泛而有成效的应用。为有效地筛选KLH刺激后DCs特异表达的基因，本室首先建立了一种基于“长距离”PCR技术（Long-distance
PCR，LD-PCR）的减数杂交方法，该方法着重于克服基于PCR的减数克隆技术中的普遍缺陷，即由于PCR优先扩增短片段而造成的分子集群偏移。将此方法应用于从KLH刺激的DCs中克隆特异表达的基因，结果发现特异表达基因约占克隆基因片段的70%。其中有一个长度为1067bp的新基因片段GC13，基因分析显示包含一个开放读框（Open reading frame, ORF），读框内序列符合真核基因的编码规律，终止码的下游有加尾信号，但在同一开放读框中的基因上游没有终止码，而且5’端ATG周围的序列不符合Kozak典型的编码结构。根据这些分析结果，利用快速末端cDNA扩增（Rapid
amplification of cDNA ends, RACE）作5’上游序列的克隆，结果RACE扩增出单一的条带，将这一条带回收测序后得到了240bp的序列，证明是GC13基因片段的上游序列。拼接成的片段长1305bp，包含一个完整的开放读框，从第269位碱基到982位碱基，起始ATG的周围虽然不符合Kozak 序列，但在其上游有多个终止码出现，而且读框内的序列符合真核基因的惯用编码规律，所以认为这是一个完整的读框。预测编码区推导出的氨基酸长237个残基，该序列利用ANTHEPROT软件和ExPASy分析服务进行分析显示，从第45位氨基酸到69位氨基酸是一个完整的跨膜区段，N端没有信号序列，提示该蛋白是一II型跨膜分子；其次，其胞外序列包含一个C型lectin 结构域（C-type lectin domain）；第185位氨基酸残基是一保守的天门冬酰胺，推测为N-糖基化位点，提示该分子在体内是一糖蛋白；另外，可能也是这一分子最重要的一点，在它的N端，也就是在细胞内区段，有一ITYAEV结构，它符合V/L/IxYxxL/V这一免疫受体中酪氨酸依赖的抑制型基序（Immunoreceptor tyrosine-based inhibitory
motif, ITIM）。根据这一分子的结构特征，我们将其命名为树突状细胞来源的lectin样的免疫受体（DCs-derived，lectin-like immunoreceptor，DC-LLIR）。该基因已于1998年6月登录在基因公共数据库，序列号为AF067800。
DCs的功能有很大一部分需要通过膜受体来介导，研究得较早的与DCs 功能相关的膜表面分子有DEC-205 和MMR（Macrophage-mannose receptor），这两个分子并非只表达于DCs，但它们对于DCs摄取抗原的功能都很重要，这两个分子都是I型跨膜蛋白，胞外含有数个C型lectin 结构域。另外，也有带lectin 结构域的II型跨膜分子被发现，如CLEC-1、CLEC-2（C-type
lectin-like receptors）、Dectin-1、Dectin-2（Dendritic cell-associated C-type lectin），这些分子的胞外段都含有一个CRD，除了在DCs表达以外，也在髓系细胞和单核细胞中表达。此外还发现表达在DCs表面的抑制型受体，如ILT3和ILT4（Immunoglobulin-like transcript），其胞内段含ITIM，胞外段是免疫球蛋白样结构，这两个分子在髓系细胞和单核细胞中也有表达，能向细胞内传递抑制性信号。本实验中发现的新型分子DC-LLIR是一II型跨膜蛋白，胞外含一个lectin结构域，胞内含ITIM，与以往在DCs中发现的受体分子在结构上不完全相同。其不同的胞外段提示这一分子可能会结合不同的抗原或配体，而相同的胞内基序则预示胞内信号传导会通过相同的信号通路，引起类似的功能作用。
在基因公共数据库中搜索到了DC-LLIR相应的染色体序列，相应序列虽然不能完全显示外显子，但可以将DC-LLIR在染色体中的位置定为12p21，这一位置与NK细胞受体复合物（NK receptor complex, NKC）的染色体定位12p12-p13较为接近。NKC编码的大部分受体是含C型lectin 结构域的II型膜蛋白，包括CD94、NKR-P1、CD69及NKG2家族的多个分子。DC-LLIR与这些NK受体的胞外结构虽然都包含C型lectin结构域，但其中存在的差别在于NK细胞受体胞外区段的CRD已有多个保守的氨基酸残基发生了突变，目前倾向于认为这些分子不属于lectin样分子，而DC-LLIR的胞外结构是典型的lectin样结构。DC-LLIR在数据库中所作的同源分析显示与DC-LLIR胞外段在结构上最相近的分子集群是脱唾液酸糖蛋白受体（Asialoglycoprotein receptors, ASGPRs），支持DC-LLIR是一个lectin样分子。
对DC-LLIR的组织分布进行了Northern检测，结果未能检测到任何表达信号，推测可能是由于DC-LLIR基因在组织中低丰度表达所致。对胚胎组织进行的RT-PCR检测结果显示DC-LLIR在肾、肺、骨骼肌中无表达，在脾脏、肝脏中有相对较高的表达，而在心、脑、小肠中有很低的表达并且还可能存在其它剪切形式。此外，选择了部分造血细胞系进行RT-PCR检测，结果显示DC-LLIR在前髓细胞系HL-60、NB4，B细胞系Ramos、Raji和Daudi，慢性髓系白血病细胞K-562中有相对较高的表达，在T细胞系Jurkat中有很低的表达，而在T细胞系HuT78和前单核样细胞系U-937中无表达；令人感兴趣的是相同的PCR条件在HL-60细胞中还扩增出两个较小片段，提示在HL-60细胞中可能存在DC-LLIR的其它剪切形式。
对出现在HL-60细胞中的另外两个片段克隆测序，发现是DC-LLIR基因的另外两个剪切形式，与DC-LLIR对应的染色体序列比较发现其中一个转录本比DC-LLIR少一个外显子，命名为I型DC-LLIR变异体（DC-LLIRv1）；另一个由于没有完整的染色体序列，不能确定是缺失了两个、三个或是更多个外显子，命名为II型DC-LLIR 变异体（DC-LLIRv2）。这两个转录本所推导的蛋白质与DC-LLIR相比，分析软件提示缺失了跨膜区。
然后原核表达、纯化DC-LLIR蛋白用于制备抗DC-LLIR的多克隆抗体以作进一步的研究，结果DC-LLIR基因不论是直接表达还是融合表达，表达量都很低，而DC-LLIRv2在两个系统中的表达量都很高。将直接表达的DC-LLIRv2经离子交换纯化后按常规方法免疫新西兰大白兔，得到的血清用无关抗原吸附纯化；Western检测显示纯化后的抗血清稀释4000倍时能检测到0.5 mg 的DC-LLIRv2，并且与菌体蛋白没有交叉反应。由于DC-LLIR包含了DC-LLIRv2的所有表位，所以制备的抗血清可以用于检测DC-LLIR及其两个变异体。
利用抗血清对部分造血细胞系所作的Western检测提示该分子主要表达在DCs、前髓系细胞和B细胞中；并且只能检测到一个条带，分子量约为60kDa，相对于其237个氨基酸的预测分子量27458Da，说明DC-LLIR在体内是一糖基化蛋白。DC-LLIR和DC-LLIRv2基因转染成纤维细胞NIH/3T3后检测其瞬时表达的蛋白，显示两个基因在真核细胞内表达出同样大小的蛋白。对这两个基因在兔网织红细胞体系中进行体外翻译也得到了同样的结果。Western的检测结果还显示DC-LLIR以单体形式存在，提示DC-LLIR无需通过结合其它分子发挥作用。
跨膜受体的功能依赖于胞内段的结构，DC-LLIR的胞内段含有一段符合免疫受体酪氨酸依赖抑制型基序（Immunoreceptor tyrosine-based inhibitory motif, ITIM）的氨基酸序列。IMTM作为细胞膜分子向细胞内传递抑制性信号的结构，其功能基础在于当其中第三位的酪氨酸磷酸化后，能够结合细胞内含有SH-2结构域的磷酸酯酶，如SHP-1、SHP-2和/或SHIP，这一结合会使磷酸酯酶具有催化活性，催化活化通路上的信号分子去磷酸化，下调细胞的活化反应。利用这一机理，可以通过抗细胞膜分子的抗体免疫共沉淀胞膜分子和磷酸酯酶的复合物，再用抗磷酸酯酶的抗体检测磷酸酯酶是否被沉淀下来从而推测ITIM传递抑制性信号的潜在能力。SHP-1阳性的细胞HL-60以过氧钒酸钠处理后，用抗SHP-1的抗体检测多克隆抗血清免疫共沉淀的产物，结果显示DC-LLIR的ITIM能和SHP-1发生结合，说明DC-LLIR中的ITIM在磷酸化后能结合SHP-1，提示DC-LLIR可能是一种抑制型受体。
此外，在对前髓系细胞进行PMA诱导分化的研究中发现，DC-LLIRv2在HL-60和NB4细胞PMA处理后24小时表达明显增高，而DC-LLIR的表达未见明显变化，提示DC-LLIRv2可能与PMA诱导细胞分化相关。
综上所述，本实验首先利用减数杂交技术从KLH刺激的DCs中克隆到一个具有C型lectin结构域的新分子，得到其全长序列后在公共基因数据库登录（序列号：AF067800）；分析显示该分子是一包含C型lectin结构域的II型跨膜蛋白，胞内区段含有ITIM特征结构；染色体定位为12p21，同源分析最为相近的分子集群为ASGPRs，该分子被命名为树突状细胞来源的lectin样的免疫受体（DC-LLIR）。对于DC-LLIR在组织细胞中表达所作的Northern 、RT-PCR分析提示DC-LLIR在正常组织中表达丰度很低，主要表达在DCs、前髓系细胞、B细胞、肝、脾。在RT-PCR分析过程中克隆到了DC-LLIR的两个剪切变异体（variants），分别命名为DC-LLIRv1、DC-LLIRv2。原核表达DC-LLIRv2蛋白后制备、纯化了抗DC-LLIRv2的多克隆抗血清。Western检测显示DC-LLIR在体内是一糖基化蛋白，分子量约60kDa，以单体形式存在，主要表达在DCs、前髓系细胞、B细胞中；同时通过基因转染真核细胞瞬时表达、体外翻译DC-LLIR和DC-LLIRv2证实两个分子在体内的分子量几乎是一致的。此外，利用免疫共沉淀证实DC-LLIR胞内段的ITIM具有结合SHP-1的潜在能力，提示DC-LLIR很可能是一个新型抑制型受体。其变异体DC-LLIRv2在前髓系细胞中的表达可以被PMA诱导增高，而已知PMA能上调细胞内SHP-1的表达，这一结果提示DC-LLIRv2表达的增高可能与SHP-1的增高相关，DC-LLIRv2可能会参与SHP-1对髓系细胞分化、活化的调节过程。
关键词：C型lectin，免疫受体酪氨酸依赖抑制型基序，树突状细胞，II型膜受体，剪切变异体，减数克隆，细胞分化
Cloning and characterization of a novel
ITIM containing membrane protein DC-LLIR and its transmembrane region deletion
variants
Dendritic cells (DCs) constitute the dominant population of antigen presenting cells (APCs) by possessing potent ability to initiate T cell immunity. The indispensability of DCs as APCs in induction of primary immune response is due to their high efficiency of antigen capturing and processing, the capacity of migration, high abundance expression of specific surface molecules for activating T cells. Little is known about the mechanisms of DCs’ functions until now. The functions of DCs are varied with its mature status. DCs can uptake and process antigen efficiently but possess low T cell stimulating capacity in immature stage, and shift to high efficiency in antigen presentation and sharply reduced capacity of antigen uptake in mature stage. The immature DCs could turn into mature stage by exposure to antigen or inflammatory stimuli. Keyhole limpet hemocyanin (KLH) is a kind of potent antigen that could induce maturation of DCs and enable DCs to possess the ability of initiating T cell immunity.
The difference of cell functions is originated from the discrepancy in gene expression. Comparative studies on gene expression in different stages of DCs can promote the understanding of physiological processes of DCs and the mechanisms of correlated molecules. In this field, two main methods, the differential display polymerase chain reaction (DD-PCR) and the subtracting hybridization, have been developed. The subtracting hybridization is much powerful than DD-PCR in practice and its technical obstacles were continually overcome with the progressing of associated strategies. To select genes differentially expressed in KLH stimulated DCs, a strategy of subtractive cloning based on “long distance” PCR was established in our laboratory. This strategy emphasized on overcoming the ubiquitous deficiency in PCR based subtractive cloning, the severe bias of fragments distribution caused by PCR amplification. By subtracting the mRNA expressed by KLH stimulated DCs with the mRNA expressed by normal DCs, about 70% fragments were identified expressed by KLH stimulated DCs specifically. One of the novel gene fragments, GC13, was predicted to contain an open reading frame (ORF). The regional variation in ORF matches the coding sequence variation of eukaryotic genes, and a polyadenylation signal sites downstream the stop codon. But the sequence upstream first ATG does not match Kozak sequences, and no stop codons site upstream the ORF. Then upstream sequence was cloned by rapid amplification of cDNA ends (RACE). A fragment with length of 240bp were amplified and revealed to be upstream sequence of GC13. The full-length sequence has a length of 1305bp and contains an ORF from nucleotide 269 to 982. Although the upstream flanking sequence does not match Kozak sequences, but because the potential start codon is the first ATG following tandem stop codons, and because the regional variation in ORF matches the coding sequence variation of eukaryotic genes, the ATG is considered to be the start codon. The deduced protein has a length of 237 amino acid and is predicted to carry a single carbohydrate recognition domain (CRD) or as C-type lectin domain at the COOH terminal and an integral transmembrane region from residues 45 to 69 according to consensus sequences analysis by ANTHEPROT software and ExPASy server. It belongs to type II membrane protein because it contains an integral transmembrane domain and without signal polypeptide. There is a consensus N-glycosylation site at residue 185 and another 6 amino acid ITYAEV that has been predicted to locate in cytoplasmic tail and matches V/L/IxYxxL/V stretch, the immunoreceptor tyrosine-based inhibitory motif (ITIM). The novel gene is designated DCs-derived lectin-like immunoreceptor (DC-LLIR) and submitted in GenBank database in June 1998 under the accession number AF067800.
The functions of DCs are mediated in part by the expression of surface receptors. Macrophage mannose receptor (MMR) and DEC-205 are two well-studied surface receptors and considered to be involved in antigen processing of DCs although they are not only expressed on DCs. Both of them are type I membrane proteins and both the extracellular sections contain several tandem repeats of CRD. Several type II lectin-like molecules were identified recently, which include C-type lectin-like receptors (CLEC) 1, CLEC-2, dendritic cell-associated C-type lectin (Dectin) 1, Dectin 2. All of these molecules contain a single CRD in extracellular region and all of them are expressed in myeloid cells, monocytes besides DCs. Another two type II inhibitory receptors, immunoglobulin-like transcript (ILT) 3 and ILT4, are expressed on surface of monocytes, macrophages besides DCs, which belong to immunoglobulin superfamily and transmit inhibitory signals into cells. The discrepancy of extracellular sections implies DC-LLIR may bind other types of antigens or ligands, and the similarity of cytoplasmic sequence infers similar signaling pathway would be triggered upon specific binding.
A corresponding chromosomal sequence of DC-LLIR was found in GenBank database, although uncomplete, which indicates the chromosomal localization of DC-LLIR is 12p21. Such a chromosomal localization is close to that of NK cell receptors cluster (NKC) 12p12-p13, which consists of C-type lectin domain containing type II membrane proteins that include NKG2 molecules, CD94, NKR-P1A/CD161 and CD69. The C-type lectin domains in these molecules are distantly related to other C-type lectin domains because several consensus amino acids in CRDs have been changed and whether they are likely to bind carbohydrates is under suspicion. DC-LLIR would be classified as a typical C-type lectin protein according to its conserved extracellular domain that shows highest homologous to asialoglycoprotein receptors (ASGPRs).
The distribution analysis of DC-LLIR in human tissues by Northern blot showed no visible signature and this result suggests DC-LLIR may be expressed in lower abundant in the tissues detected. RT-PCR analysis of DC-LLIR in fetal tissues showed DC-LLIR was not expressed in kidney, lung, skeleton muscle, expressed with relatively low abundant in heart, brain, small intestine and expressed in relatively high abundant in liver and spleen. This result is in agreement with its high homology to ASGPRs. It is also indicated that alternative splice variants may exist in heart, brain and small intestine. mRNA expression analysis in hematopoietic cell lines showed DC-LLIR was expressed with relatively high levels in DCs, promyelocytic cells HL-60 and NB4, B cells Ramos, Raji and Daudi, with relatively low levels in chronic leukemia cells K-562, T lymphoma line Jurkat; and not express in T lymphocyte cells HuT78, histiocytic lymphoma cells U-937. It is interesting that alternative splice variants may exist in HL-60 cells.
Two short fragments emerged from HL-60 cells in RT-PCR analysis were cloned, sequenced and revealed to be two alternative splice variants of DC-LLIR. One transcript lacks an exon of DC-LLIR according to comparison with the corresponding chromosomal sequence, which was named type I variant of DC-LLIR (DC-LLIRv1). Another transcript lacks two, three or much more exons of DC-LLIR with uncertainty because the corresponding chromosomal sequence is restricted, which was named type II variant of DC-LLIR (DC-LLIRv2). It is predicted by analysis software that the transmembrane domain of DC-LLIR are absent in these two transcripts.
Then the DC-LLIR protein was expressed by prokaryotic system and purified for immunization. But both the expression levels of DC-LLIR by direct expression or fusion expression are too low to satisfy purification and immunization. In contrary, the expression levels of DC-LLIRv2 in both of the two systems are very high. The DC-LLIRv2 protein expressed by prokaryotic direct expression system was purified by ion-exchange chromatography and used to immunize New Zealand white rabbit with routine methods. The antiserum was purified by absorption with non-reacting antigen. Western blot analysis showed this purified antiserum could detect 0.5 mg of DC-LLIRv2 protein at the titration of 1:4000, and cross-reactivity with bacterial protein was not detected. Because DC-LLIR protein contains all epitopes of DC-LLIRv2, the prepared antiserum could be used to detect all three transcripts.
Immunoblotting analysis of some hematopoietic cells revealed one specific binding band with size of about 60 kDa in DCs, promyelocytic cells and B cells. The results indicate DC-LLIR is a glycoprotein in vivo because the predicated molecular weight of amino acids backbone is 27458 Da. The ancillary evidences were obtained from post transfection expression and in vitro translation. Fibroblast NIH/3T3 produced same size proteins after transfected with DC-LLIR or DC-LLIRv2 genes, in agreement with the results from translation of DC-LLIR or DC-LLIRv2 gene in rabbit reticulocyte lysate. It is also indicated that DC-LLIR exists as monomer in vivo, which suggests DC-LLIR may function independently.
The cytoplasmic tail of receptor is the structural basis of its signaling transduction. An ITIM motif was identified in the intracellular region of DC-LLIR, which is a kind of motif able to deliver inhibitory signal after the extracellular part of receptor was cross-linked properly. Phosphorylation of the tyrosine residues within ITIM creates a docking site for the Src homology-2 domain (SH-2) of the intracellular phosphatases and results in the recruitment and activation of at least one of the SH-2 domain containing protein tyrosine phosphatases, SHP-1 and SHP-2, and/or the inositol phosphatase SHIP. The activation of these phosphatases results in dephosphorylation of signaling molecules, leading to down modulation of cell activity. This mechanism enables the assessment of potential ability of ITIM like sequence by immunoprecipitation. The result that the ITIM in DC-LLIR could recruit SHP-1 upon pervanadate treatment indicates DC-LLIR has the potential to function as an inhibitory receptor.
Furthermore, the mRNA of DC-LLIRv2 was increased in promyelocytic cells until 24 hr later after PMA treatment, while the mRNA expression of DC-LLIR remained unchanged, which indicates DC-LLIRv2 may associate with PMA induced differentiation.
Taken together, a novel gene, which contains a C-type lectin domain, was cloned from KLH stimulated DCs by subtractive cloning. The full length sequence of this gene was obtained by RACE and submitted into GenBank database (Accession: AF067800). The deduced protein contains an ITIM in intracellular region proximal to N terminal, which belongs to type II lectin-like membrane protein for carrying a single CRD at the COOH terminal, an integral transmembrane domain but without signal peptide. This gene was designated DC-LLIR for DCs-derived lectin-like immunoreceptor. The gene of DC-LLIR locates in a region of chromosomal 12p21 and shows highest homologous with ASGPRs. Then the distribution of DC-LLIR was analyzed by Northern blot and RT-PCR and revealed that DC-LLIR was expressed with low abundant in tissues and cells and mainly expressed in DCs, promyelocytic cells, B cells, liver and spleen. Two other fragments emerged from RT-PCR analysis were cloned and identified to be two alternative splice variants of DC-LLIR, which were named DC-LLIRv1 and DC-LLIRv2. And then DC-LLIRv2 protein was expressed in E.Coli and purified for preparing polyclonal antiserum. Immunoblotting analysis of some hematopoietic cells were performed by using this polyclonal antiserum and revealed DC-LLIR was expressed in DCs, promyelocytic cells and B cells. The mature protein is glycosylated with molecular weight of 60 kDa, exists as a monomer and displays same size with the DC-LLIRv2 protein, which is in agreement with the results from transit expression after transfected into mammalian cells and in vitro translation. Furthermore, the ITIM in DC-LLIR was demonstrated to bind SHP-1 after the tyrosine had been phosphorylated, which indicates DC-LLIR has the potential to function as an inhibitory receptor. In addition, the mRNA expression level of DC-LLIRv2, one alternative splice variant, was raised during differentiation induced by PMA. With the fact that the intracellular SHP-1 would increase upon PMA stimulation, the result indicates that the increase in DC-LLIRv2 expression may be associated with the increase of SHP-1, and DC-LLIRv2 may participate the regulatory influence of SHP-1 on activation and differentiation process in myelocytic cells.
Keywords: C-type lectin, immunoreceptor tyrosine-based inhibitory motif, dendritic cells,
type II membrane receptor, alternative splice variant, subtractive cloning, cell differentiation