Detection of isolated tumor cells (TC) in bone marrow (BM) from patients with breast cancer is usually accomplished by immunocytochemical (ICC) analysis of up to 2 X 10(6) mononuclear cells (MNC). However, this method is cumbersome if large numbers of BM cells (i.e. > 1 X 10(7) cells) are to be analyzed. This emphasizes the need for TC enrichment strategies. This report describes immunomagnetic separation (IMS) techniques for enrichment and detection of viable breast carcinoma cells in BM and peripheral blood (PB). The positive IMS technique was performed by incubation of MNC with 2.8 microns magnetic particles (rat antimouse IgG1 M280-Dynabeads) coated with monoclonal antibody (mAb) against epithelial surface antigens. The rosetted tumor cells were then visualized by ICC staining using alkaline phosphatase-conjugated A45-B/B3 anticytokeratin mAb (Fab). The negative IMS technique was performed by incubation of MNC with anti-CD45-coated M450-Dynabeads (4.5 microns), followed by ICC staining of the nonrosetted cells. When 1000, 100, and 10 breast carcinoma cells were mixed with 1 X 10(7) MNC, an average of 748 (n = 9), 70 (n = 10), and 7.8 TC (n = 8), respectively, were detected with the positive IMS technique. With the negative IMS technique, 648 (n = 8), 57.8 (n = 6), and 7.3 TC (n = 6), respectively, were detected. The analysis of 1 X 10(7) MNC with the IMS techniques was compared with the ICC analysis of 2 X 10(6) unseparated MNC. A mean 3.7-fold (range 1.5-6.4) to 4.2-fold (2.5-8.2) (positive IMS) and 3.1-fold (range 2.0-5.0) to 3.8-fold (2.0-6.0) (negative IMS) higher TC detection frequency was achieved after enrichment by IMS in experiments with 100 and 1000 TC/10(7) MNC. The IMS techniques were used for examination of BM samples from locally advanced breast cancer patients. A 5.3-fold mean increase (range 2.1-13.3) in the number of TC detected was obtained when the use of positive and negative IMS together was compared with the direct ICC analysis of unseparated MNC (n = 11). Enrichment of TC by IMS techniques enables us to examine large numbers of MNC from BM or PB, which can result in the detection and characterization of minimal residual disease with increased sensitivity and specificity.

Here we report that human B lymphocytes can be positively selected directly from buffy coats applying the anti-CD19 antibody AB1 coupled to magnetic beads. This isolation protocol is highly efficient and the isolated cell population is of very high purity and viability. As judged by cell cycle analysis and various parameters for cell activation, the cells are still in a resting state after isolation. Furthermore, different functional assays have shown that the isolation procedure does not interfere with either activation or proliferation/differentiation of CD19 selected cells as compared to negatively isolated cells. As a consequence of cross-linking during the isolation process, the CD19 antigen is temporarily down-regulated as measured by AB1 binding. Despite this decreased expression, monoclonal antibodies to the CD19 antigen nevertheless inhibited anti-mu plus B cell growth factor induced B cell activation as reported also for negatively isolated cells. Taken together, the presented data strongly suggest that B cells isolated through the CD19 antigen can be used in critical functional assays.

No abstract

Serological cloning of tumor-associated antigens (TAAs) using patient autoantibodies and tumor cDNA expression libraries (SEREX) has identified a wide array of tumor proteins eliciting B-cell responses in patients. However, alternative cloning strategies with the possibility of high throughput analysis of patient sera and tumor libraries may be of interest. We explored the pJuFo phage surface display system, allowing display of recombinant tumor proteins on the surface of M13 filamentous phage, for cloning of TAAs in prostate cancer (PC). Control experiments established that after a few rounds of selection on immobilized specific IgG, a high degree of enrichment of seroreactive clones was achieved. With an increasing number of selection rounds, a higher yield of positive clones was offset by an apparent loss of diversity in the repertoire of selected clones. Using autologous patient serum IgG in a combined biopanning and immunoscreening approach, we identified 13 different TAAs. Three of these (NY-ESO-1, Lage-1, and Xage-1) were known members of the cancer/testis family of TAAs, and one other protein had previously been isolated by SEREX in cancer types other than PC. Specific IgG responses against NY-ESO-1 were found in sera from 4/20 patients with hormone refractory PC, against Lage-1 in 3/20, and Xage-1 in 1/20. No reactivity against the remaining proteins was detected in other PC patients, and none of the TAAs reacted with serum from healthy subjects. The results demonstrate that phage surface display combined with postselection immunoscreening is suitable for cloning a diverse repertoire of TAAs from tumor tissue cDNA libraries. Furthermore, candidate TAAs for vaccine development of PC were identified.

Ephrin-A4 is a ligand for the erythropoietin-producing hepatocellular (Eph) receptor family of tyrosine kinases. We have identified a secreted form of ephrin-A4, denoted ephrin-A4 (s), which is encoded by an alternatively spliced mRNA and is produced by in vivo activated B cells in tonsils. Blood B cells secrete ephrin-A4 (s) upon stimulation via the B-cell antigen receptor. A subpopulation of tonsil cells in the crypts with a dendritic cell phenotype was shown to express EphA2, an Eph receptor tyrosine kinase that was found to be capable of binding an ephrin-A4 immunoglobulin chimeric protein. We conclude that ephrin-A4 (s) may play a role in the interaction between activated B lymphocytes and dendritic cells in human tonsils. (Blood. 2000;95:221-230)