TEXT

Top Abstract Text References

The human immune system is characterized by expansion of the B-cell repertoire following recombination events that take place at both the genetic and the somatic levels. This expansion leads to the generation of an extremely diverse set of B cells. Diversity among B cells makes it difficult to devise suitable strategies for study of the B-cell clonal response (1). Like B cells, T cells are also very diverse. Suitable assays for analysis of clonal T cells have only recently become available (8, 11, 12). T-cell assays are based on the known heterogeneity of the T-cell receptor variable (V) regions and use techniques like single-stranded conformation polymorphism analysis (8) and the heteroduplex tracking assay (HTA) (11). T-cell studies have recently become more selective and narrow because of the associations that have been seen between specific T-cell gene subfamilies in health and disease (8, 11, 12). Such information about B-cell gene subfamilies is not available, and no attempt has so far been made to develop assays that can be used to study B-cell clonality. Most of the studies that have been done to analyze human clonal B cells have used strategies that involve sequencing of the variable regions of heavy-chain (V_H) genes and light-chain genes and comparisons of the lengths of the complementarity-defining regions (2, 4, 6, 9, 10).

In this study, I have applied the principle of HTA to analyze B-cell clonality. The basis of HTA was the heteromobility of DNA duplexes on the gels. The heteroduplexes were formed due to annealing of diverse immunoglobulin (Ig) genetic sequences to a single-stranded (ss) gene family-specific probe. HTA has previously been shown to be better a technique for the study of genetic heterogeneity, mainly because of its ease of operation and unbiased approach. It does not involve sequencing of every Ig clone, which can be labor-intensive and time-consuming. One other advantage of HTA is that the observed bands on the gel can be directly correlated to the number of different clones in a given population (5, 11).

The assay was performed with peripheral blood mononuclear cells (PBMCs; 5 × 10⁶). PBMCs were obtained from leukocyte-enriched blood collected at the New York University Blood Center from three healthy donors and from blood samples of four vaccines enrolled in an AIDS study (3). PBMCs were purified with Ficoll-Hypaque and were subjected to RNA extraction with the RNA Easy kit (Qiagen, Santa Clarita, Calif.). The extracted RNA (4 µg) was then reverse transcribed with the Superscript II RT kit (Gibco BRL, Gaithersburg, Md.) and random hexamers according to the manufacturer's recommendations. A nested PCR (nPCR) was initiated with the cDNA product and consensus external and internal sets of primers for the Ig heavy chain under the PCR conditions described previously (2). The final amplified nPCR products obtained from PBMCs of healthy donors were subsequently analyzed by ligating, cloning, and sequencing (Invitrogen Inc., Carlsbad, Calif.). The plasmid clones generated from the PBMCs of a healthy donor served as templates for the generation of probes for HTA. In this study, only two of the seven major V_H gene families were studied. Among the different V_H human gene families, the frequency of use of these two gene families seems to be the highest (2).

The nPCR products were generated in a 50-µl reaction volume. The amplified products were first run on a gel to compare the intensities of the DNA bands; bands that exhibited equal intensities were analyzed by HTA. The comparison of bands not only confirmed successful cDNA synthesis and a successful amplification reaction but it also ensured that equal amounts of DNA were used in the assay. HTA was carried out essentially as described previously (11) by adding 1 µl of radiolabeled ss probe to 5 µl of the nPCR products in a total volume of 10 µl of annealing buffer (100 mM NaCl, 10 mM Tris [pH 7.4], 2 mM EDTA). Heteroduplexes were formed by melting the nPCR products at 94°C for 3 min in the presence of the probe and by cooling the mixture to 4°C in a thermocycler. The duplexes thus formed were then separated on a 5% polyacrylamide gel (acrylamide and bisacrylamide [30:1]) at 250 V for 2 h. Finally, the gels were dried under vacuum and were subjected to autoradiography.

ssDNA probes were generated by PCR with a 5' biotin-tagged variable-region heavy-chain primer and a 5' ³²P-labeled JH primer. The J_H primers were radiolabeled by use of the T4 kinase kit (Amersham International plc). The amplified product was purified with magnetic M280-streptavidin Dynal beads (Dynal Inc., Oslo, Norway). The bound product was eluted with 0.1 N NaOH, collected, and neutralized with 0.2 N HCl-1 M Tris-HCl (pH 8.0) in a total volume of 50 µl.

For the study of assay performance, nPCR products were amplified from the V_H plasmid clones and were then analyzed by HTA with an ss probe. An ss V_H 3-J_H 1/4/5 probe was able to identify different duplexes generated from plasmids, and it could also distinguish the presence of two different clones in a reaction mixture. The two clones were identified as a homoduplex and a heteroduplex, respectively, on the basis of their mobilities (Fig. 1). When PBMCs were subjected to HTA is with the same (V_H 3-J_H 1/4/5) probe, a smear was observed on the autoradiographs. The smear was an indication of polyclonality for V_H 3 B cells, and it resulted from migration of Ig sequences to a separate location on the gel after they reannealed to a radiolabeled Ig ss probe. An occasional oligoclonality among the PBMC B cells was sometimes also identified as a prominent dark band among the polyclonal population (e.g., Fig. 1, lane 5).

View larger version (100K): [in this window] [in a new window]   FIG. 1.   Clonal analysis of VH 3 Ig sequences by HTA. VH 3 ss radiolabeled probes and nPCR products were reacted and run on a 5% gel followed by autoradiography. Lane 1, binding of ss probe to a product amplified from a homologous plasmid; lanes 2 and 3, binding of probe to products amplified from heterologous plasmids that had some base pair difference between them; lane 4, ability of ss probe to distinguish two clonal populations in a mixture; the reaction mixture was obtained by combining the products amplified from plasmids from lanes 1 and 3 in equal proportions; lane 5, reaction of probe to product amplified from PBMCs; lane 6, probe alone. ssDNA, HTDX, and HMDX indicate the locations of ssDNA, heteroduplexes, and homoduplexes, respectively.

Next, the specificity of the assay was evaluated by mixing a fixed number (5 × 10⁵, 5 × 10⁴, 5 × 10³, 5 × 10², 5 × 10¹, or 0) of monoclonal HL-6 B cells (with human V_H 3-J_H 4 genes; kindly provided by Marshall Posner) with PBMCs (5 × 10⁶). The cells were analyzed by HTA with an ss V_H 3-J_H 1/4/5 probe. The analysis was able to detect clonality for HL-6 cells at a ratio of 1:10⁴ cells (Fig. 2). Analysis of the PBMCs from healthy individuals with the same probe yielded a smear pattern that was indicative of polyclonality (Fig. 2, lane 1). Bands above the ss probe in most cases resulted from binding of the ss probe to larger Ig fragments from the first reaction of the nPCR and were considered not significant for clonal analysis. To observe oligoclonality among B cells, I chose frozen PBMC samples from human immunodeficiency virus type 1 (HIV-1)-positive patients that had strong antibody responses to HIV-1. These samples were available from a previous study that was undertaken to understand correlates of HIV-1 protection (3). Figure 3 shows oligoclonality for two of the seven V_H gene families, V_H 1 and V_H 3, in one such individual. The V_H 1 repertoire remained more or less constant, and it did not show any noticeable change in banding pattern at various points of analysis. The V_H 3 repertoire, on the other hand, changed over the same time period in the same patient. The appearance and disappearance of new bands corresponded to the presence or disappearance of new V_H 3 B-cell clones. A smear in lane 1 of Fig. 3, for both V_H groups was probably due to the polyclonal nature of B cells. Early in HIV-1 infection, antibody responses tend to be polyclonal, and later, these responses become extremely oligoclonal (3, 7). This observation is under further investigation to understand the complexity of B-cell immunity to HIV-1. Application of the analysis to other patient samples yielded in each case banding patterns that showed unique B-cell clonal patterns (data not shown).

View larger version (62K): [in this window] [in a new window]   FIG. 2.   Detection of spiked clonal B cells among PBMCs. Different numbers of HL-6 clonal B cells were added to 5 × 106 PBMCs and the cells were analyzed by HTA. The autoradiograph depicts the results of HTA run with spiked samples with a VH 3-JH 1/4/5 probe. The arrow indicates the location of the clonal heteroduplex product for HL-6 cells within the PBMC population. The ratios at which HL-6 cells were mixed with PBMCs are indicated at the top.

View larger version (64K): [in this window] [in a new window]   FIG. 3.   VH gene usage in an HIV-1-infected (C12) individual. Lanes A to E, samples analyzed 231, 239, 251, 392, and 555 days after infection, respectively; The autoradiograph shows the results of an analysis of both the VH 1 and VH 3 gene family PCR products on a 5% gel. Analysis of the VH 1 gene family showed the presence of two dominant bands in the heteroduplex area. The open circle indicates the appearance of new faint bands among the VH 3 group. A prominent dark band present in all lanes indicated the presence of unused ss probe. Bands above the ss probe were not considered significant for the analysis.

HTA results that showed distinct banding patterns were more likely due to an overrepresentation of certain cell populations (Fig. 2). Cells that appeared to be overrepresented by HTA were often the cells that had identical Ig sequences. These clonal populations can arise either following antigen exposure or due to tumorigenicity. A distinction between the two can be made on the basis of the nature of cell types along with some supporting clinical observations. Tumor B cells generally have a tendency to persist and expand; on the contrary, immune B cells tend to disappear and are more transient. Immune B cells are both multiplied and limited by somatic mutations and the death of cells. Any mutation in the Ig sequence of antigen-specific B cells within the complementarity-determining region can alter the B-cell affinity, even though cells may retain their specificity. B cells that have same specificity but different affinities when they are subjected to HTA analysis will migrate and band at a distinct location on the gel. A similar observation was made for the two V_H 3 Ig clones that had a few base pair differences (Fig. 1). Detection by HTA of cells that arise from mutation will also depend on the sensitivity of the analysis.

In summary, I have described here for the first time a method that can be used to track clonal development among B cells in humans by HTA. The assay can find applications for the detection of B-cell lymphomas and plasmacytomas by recognizing them as predominant clonotypes or, in many cases, as the only clonotypes. It can also be used to study the immune response by observing changes in clonal types over time. The present assay design is limited in that it cannot distinguish between antigen-specific and nonspecific B cells if duplexes generated from these cells were to band at an identical location. A modification of this technique with panned antigen-specific B cells may help in resolving this limitation of HTA.

Nucleotide sequence accession number. The sequences of plasmid clones (one clone for each V_H gene family) generated from the PBMCs of a healthy donor were submitted to the GenBank-National Center for Biotechnology Information database (accession no. AF129749-54).