Gene News

基因快讯2002年第3期   



应用抗体芯片对白血病免疫组型的分析

Larissa Belov, Odetta de la Vega, Cristobal G. dos Remedios,
Stephen P. Mulligan, and Richard I. Christopherson
Departments of Biochemistry [L. B., O. d. l. V., R. I. C.] and Anatomy and Histology, 
Institute for Biomedical Research [C. G. d. R.], University of Sydney, NSW 2006, 
and Department of Hematology, Concord Hospital, Concord, NSW 2139 [S. P. M.], Australia

(续前)

结果
    预备实验表明,多余的细胞聚合体和非特异结合的白细胞可通过以下方法降到最低:洗涤细胞以除去外在的营养物;添加螯合剂(EDTA);把温浴温度从37℃降到22℃。在这样的条件下,很少甚至没有细胞聚合体在抗体芯片上被检测到,背景很低甚至没有。

    图1 表示三个独特的细胞系与含60个抗体点的芯片杂交的模式,并用非常规的暗视野显微镜观察。杂交点的强度反映了与特定位置上抗体结合的细胞的密度(Fig. 1a)。CCRF-CEM 细胞与抗体如CD4, CD5, CD7, CD8, CD38, CD44, CD45,CD45RO, CD71, 和 CD95 上有较高的杂交信号,但与CD3 和 CD52 信号较低(Fig. 1b). Raji细胞与CD10, CD19, CD20, CD21, CD22, CD23, CD37, CD38, CD45, CD45RA, CD52, CD71, CD79b, CD80, 和 CD95有较高的杂交信号;与sIg的杂交信号适中;与CD154, k 杂交信号较低(Fig. 1c). HL-60与CD4, CD13, CD33, CD44, CD64, CD71, 和CD117杂交信号较高;与 CD11b, CD11c, CD15, CD38, CD45, CD45RO, CD95, 和 KOR (CD66c)杂交信号适中;与CD8, CD14, 和 CD16 杂交信号较低(Fig.1C)。图2说明,当Raji细胞(密度107 cell/ml)结合于HLA-DR和CD38抗体时,在样品的细胞密度与结合细胞的数量存在着线形关系。当样品的细胞在6×105 cell/ml时,很少甚至没有杂交信号被检测到。

    图3显示从正常组织与各种白血病病人的白细胞的杂交模式。正常外周血白细胞与30多个抗体有杂交信号。抗T细胞抗原的抗体(CD2, CD3, CD4, CD5,CD7)比抗B细胞抗原的抗体(CD19, CD20, CD21, CD22,CD23 和 CD24)能结合更多的细胞,反应为T淋巴细胞与B淋巴细胞之比为70:30,它们占了外周血白细胞的20-40%(17)。抗CD44与CD52的抗体与白细胞高强度结合,这在大多数白细胞中是一致。白细胞与CD13和CD33结合强度适中,反映出单核细胞/粒细胞在外周血中的浓度(>40%白细胞)。抗CD41和CD42a的抗体与血小板结合,同时抗glycophorin A(CD235a)与红细胞结合。在新鲜的白细胞准备液中都发现这两个类型细胞的存在。

    CLL病人中含有大量白细胞的样品 (>30×109cells/liter),阳性斑点较少(<25; Fig. 3c),密度均一,反映了单克隆CLL细胞相对于正常白细胞的优势(4–10×109cells/liter; Ref. 17)。对于白细胞数较少的CLL病人来说(8–14×109cells/liter; Fig. 3d),其结合模式仍然可以与正常白细胞相区别,因为对应于B细胞抗原的点密度要大于或等于T细胞的点。我们还没有检测白细胞数小于8×109cells/liter的CLL病人,但是当CLL细胞和正常白细胞以1:3相混合(相当于血液白细胞总数5×109cells/liter)时CLL的模式仍然是明显的,提示LD array可用于检测较早期的白血病或淋巴瘤。
来自于一个HCL病人的白细胞(Fig. 3e)与抗CD103和FMC-7紧密结合,而不是CD23和CD5,与典型的HCL免疫表型相一致,从而与CLL区分开来。与抗CD5结合的小量细胞用可溶性抗CD3和抗CD20荧光标记后经共聚焦显微镜检测证明是T细胞。NHL(Fig. 3f)与CLL和HCL相比,具有B细胞免疫表型(CD51/ CD79b1/CD1032/FMC-71)。AML病人的白细胞模式是双表型的(Fig. 3f),除了单细胞谱系抗原(CD4, CD11b, CD33, CD36, CD38, CD64, and HLA-DR)的表达外,还有T淋巴细胞标记CD2的表达。该免疫表型通过流式细胞仪确证。所有的白细胞都有T细胞免疫表型(Fig. 3h)。从LD微阵列获得的HCL, NHL, 和T-ALL (Fig. 3, e, f, and h)的免疫表型和由病理实验室提供的流式细胞数据非常吻合。

    用LD array确定的两个CLL病人白细胞中48个抗原的表达与流式细胞仪非常接近(表1),尤其是高水平表达的抗原。当FACS分析显示总群体的75–100%有抗原表达(CD5, CD19, CD20, CD24, CD37, CD44, CD45RA, CD52, and HLA-DR; patient 7),即表示发生了高水平的结合。35–75%的表达(CD9, CD11b, and CD21; patient 7)与中度结合相关,而15–35%的细胞表达抗原结合较弱(CD22, CD45RA, CD79a, and GPA; patient 8)。当FACS显示只有0–15%细胞阳性时结合通常是负的或+/-。有些情况下,细胞结合的FACS结果要比预期的低(CD11c, CD79a, CD95, and CD154 for patient 7),常常反映了这些抗原的低表达。然而,抗原低表达也不总是与弱结合相关的,提示有些抗体(CD22, CD23, and CD71 for patient 8)结合细胞的能力远比其它抗体强,尽管这些强结合抗体使用时的浓度(25 or 50 mg/ml)要比其它的(200 mg/ml)低。有些点显示+/-细胞结合的,在FACS结果中却是负的(CD2, CD7, CD9, CD103, CD117, and CD122 for patient 8),提示较小细胞亚群的检测用点阵列(106cells/array)而不是FACS(5000 cells counted)。但是这些结果应被谨慎对待,因为一个+/-结果代表了<50 cells/antibody dot,可能不是十分重要的。

    用LD array分别分析冻存及新鲜分离的正常白细胞或白血病细胞,得到可比的点阵模式。但是来自冻存细胞的一些抗体(CD9, CD11b, CD13, CD14, CD15, CD16, CD56, CD57, CD79a, CD95, and CD154)偶尔会有负结果,提示可能由于这些抗原表面脱落或抗原决定簇破坏导致了结合力降低甚至丧失。根据LD array分析,尽管冻存细胞丢失了一些抗原,抗体的多功能平板也允许样品间在点阵模式上有变化,而一致模式上诊断不变。

    来自20个CLL病人的白细胞样品和来自20个正常样品相比,其点阵模式显著不同,如图4。正常和CLL的白细胞间有28个抗原差异显著(P>0.0005),以星号标出。CLL白细胞抗原表达的一致模式按细胞结合的降序排列依次为:CD44, HLA-DR, CD37, CD19, CD20, CD5, CD52, CD45RA, CD22, CD24, CD45, CD23, CD21, CD11c, sIg, and CD71。κ和λ轻链表达诠释了白血病单克隆性。能够最好区别CLL和正常外周血白细胞的抗原是CD19, CD20, CD21, CD22, CD23, CD24和CD37。此外,CLL样品对特异性结合T细胞(CD2, CD3, CD4, and CD7)和各种骨髓细胞(CD4, CD11b, CD13, CD14, CD16, CD32, CD33, CD36, CD41, CD42a, CD61, and CD64)的抗体的结合力显著减弱,反映了白血病B细胞与正常白细胞的高比值。与此相一致,部分CLL样品对细胞表面标记如CD9, CD11b, CD11c, CD25, CD38, CD45RO, CD71, CD79a, CD79b, CD80, CD122, FMC-7, 和FLT3 (CD135)等是阳性的。两个对CD23阴性。正常外周血白细胞的血小板相对较高,可通过显微镜观察抗体对CD9, CD36, CD41, CD42a, CD60, and CD61 (Fig. 3b)很容易识别出来。白血病病人的样品通常血小板较低,反映了血液中这类细胞的比例较低(Fig. 3, c–h)。

 

在有些样品中,GPA点揭示了RBC(Fig. 3, e and f)污染。随后又有显示,用渗透裂解法(裂解液0.15 M ammonium chloride/10 mM potassium hydrogen carbonate/ 0.1 mM EDTA)从制备的白细胞中去除RBC以后,白细胞的结合模式没有显著改变。

Fig. 1. Binding patterns of human cell lines using the LD Array. a, key to antibody identification; b, CCRF-CEM; c, Raji; d, HL-60. The numbers in the antibody key refer to antibodies against the corresponding CD antigens; mIgG1 and mIgG2a are murine isotype control antibodies; 44 v3–10 and 44 v6 are antibodies against CD44 variants 3–10 and 6; k, l, GPA, HLA-DR, KOR, FMC7, and sIg are antibodies against human k and l light chains, GPA (CD235a), HLA-DR, KOR-SA3544 antigen (CD66c), FMC-7, and surface immunoglobulin.

Fig. 2. Relationship between cell density of the sample and number of cells binding to antibody dots on LD Arrays. Serial 2-fold dilutions of a suspension of Raji cells (107 cells/ml) were tested on LD Arrays. The number of cells bound to antibody dots for HLA-DR (f) and CD38 (F) are plotted against the density of cell suspensions. Regression analysis of data for HLA-DR (OOO) and CD38 (——) gave coefficients of correlation (R2) of 0.97 and 0.96, respectively.

Fig. 3. Binding patterns of human leukocytes using the LD Array. The leukocyte abbreviation, source, and cell density of the sample appear in brackets. a, key to antibody identification; b,normal peripheral blood leukocytes (subject 1; 4x109 cells/liter); c, CLL (patient 1; 30x109 cells/liter); d, CLL (patient 2; 9x109/liter); e, HCL (patient 3; 16x109 cells/liter); f, mantle cell lymphoma (patient 4; 19x109 cells/liter); g, AML (patient 5; 190x109 cells/liter); h, T-cell ALL (patient 6; 17x 109 cells/liter). The numbers in the antibody key refer to antibodies against the corresponding CD antigens as defined in the legend to Fig.1.

Table 1 Comparison of LD Array and flow cytometric analyses for two CLL samples Dot array results are expressed as level of cell binding to each antibody dot: 2 (none), 1/2 (very low), 1 (low), 11 (moderate), 111 (high), 1111 (very high). Flow cytometric analysis results are expressed as percentage of cells showing fluorescence for each antigen. The (1/2) to (111) results shown in brackets denote intensity of fluorescence compared to cells with control antibody (FITC- or PE-conjugated murine IgG1).

Fig. 4. Immunophenotypes of normal peripheral blood leukocytes and CLL cells. Leukocytes from peripheral blood of 20 normal individuals and 20
CLL patients (with leukocyte counts ranging from 8x109/liter to 250x109/liter) were tested using the LD Array, and the density of cell binding was scored by comparison with a set of standard dots of increasing intensity, as shown in the inset. The average values for CLL cells were calculated for each antibody and are plotted in descending order (black bars). The average values for normal peripheral blood leukocytes are also shown (gray bars). Statistical significance (P , 0.0005) of differences between CLL and normal peripheral blood leukocytes was determined by two-sample Student’s t test assuming equal variances and is shown by an asterisk (p).

Fig. 5. Confocal microscopy. Detection of CD3 (green fluorescence) and CD19 (red fluorescence) on leukocytes from (a, b, c) a normal individual
and (d, e, f) a CLL patient (patient 9; 13 3 109cells/ liter), bound to three different antibodies on an LD Array: a and d, anti-CD5; b and e, anti-CD7;c and f, anti-CD20. White scale bar, 10 mm. The binding of fluorescently labeled CD antibodies to cells captured on an LD Array is described in “Materials and Methods.”

 

    来自CLL样品的白细胞以高密度和抗CD5相结合(Fig. 3, c and d),与已知的CLL上CD5的表达相一致。这种结合的密度相对于CLL细胞与其它T细胞标记(如CD7)的结合密度相比是高的,而与B细胞标记(如CD20)的结合密度相当。如图5所示,与一个抗体点结合的细胞群体可通过标记可溶性抗-CD3-FITC(绿色)和抗-CD19-PE (红色),用共聚焦显微镜观察得到进一步验证。正常的结合到CD5和CD7点上的外周血白细胞很明显是T细胞(Fig. 5, a and b),而与CD20结合的则是B细胞(Fig. 5c)。与此相比,与CD5结合的CLL细胞则是B细胞(Fig. 5d),而在CD5和CD7上则观察到较少量的T细胞(Fig. 5, d and e)。仅有两个样品的CD20结合B细胞(Fig. 5, c and f),在CLL病人的血液中很明显是白血病B细胞。

讨论(转下页)

          目录                                

专题文献

 蛋白质组,蛋白质组学及研究技术路线

 研究蛋白质相互作用的技术平台——酵母双杂交系统的发展和应用

 应用文献
 
 应用抗体芯片对白血病免疫组型的分型

 哺乳动物神经元培养细胞中的RNAi作用分析

 线粒体DNA基因表达的缺乏导致体内细胞凋亡的升高

 实验指南

 siRNA实验成功要点

 酵母双杂交实验常见疑难解答

 特别推荐

 新一代蛋白研究工具——抗体芯片

 Liquichip液相蛋白芯片系统——新型的蛋白质研究平台

 蛋白活性研究的新工具:乙酰化系列抗体

 蛋白组研究的完整解决方案——Genomic Solution Inc.蛋白质组学研究仪器和软件

 最新动态

 最新学术讲座及活动信息