Ex-vivo Clonally Expanded B Lymphocytes Infiltrating Colorectal Carcinoma Are of Mature Immunophenotype and Produce Functional IgG

Background Tumor infiltrating B cells (TiBc) have not yet been investigated in detail. This may at least in part be due to technical difficulties. Here we describe a straightforward and reproducible method to isolate and culture TiBc from primary colorectal carcinomas (CRC). Methods/Results TiBc cultures were generated by Epstein-Barr virus (EBV) immortalization. With this method, monoclonal TiBc cultures were obtained for 14/19 CRCs. As assessed by flow cytometry and ELISA, TiBc showed an activated immunophenotype (CD23+, CD80+) and produced immunoglobulin (Ig; IgG secretion in 55% of the cultures). In functional in vitro analysis, most of the IgGs specifically bound to allogeneic CRC target cells. These data suggest that TiBc are antigen-experienced and thus may exhibit functionality in situ. Additionally, mini-cultures generated from 12 further CRCs revealed TiBc outgrowth exclusively in the presence of EBV. Conclusion In summary, this simple method provides a cellular tool and our data set the stage for analysing the bivalent role of TiBc; being antigen-presenting cells on the one hand and tumor-specific antibody producers on the other. Additionally, the generation of long-term TiBc cultures and their monoclonal Ig may serve to identify novel tumor-specific antigens.


Introduction
Lymphocytes are major components of mononuclear cells infiltrating human malignant tumors. Their positive prognostic relevance is well established [1][2][3]. Recent studies focused on T cells, for they are believed to have the greatest antitumoral potential [4]. Indeed, a dense infiltrate of CD8+ cells was shown to correlate with prolonged survival of colorectal carcinoma (CRC) patients [5,6].
However, tumor-infiltrating B cells (TiBc) may play an important role, too. They are considered to be positive regulators of immunity often collaborating with T cells to generate potent, unrelenting immune responses [7]. Beside their antibody (Ab) producing capacity, they enhance T cell responses by secreting stimulatory cytokines and chemokines, or by serving as local antigen-presenting cells (APCs).
The tumor entity in which most efforts have been undertaken in this regard is breast cancer. There, TiBc are present in about 24% of tumors and comprise up to 40% of lymphocytic infiltrates [8,9]. They have been shown to undergo antigen-driven clonal prolifer-ation and affinity maturation in situ [10]. Very recently, in a large patient cohort of different histological and biological subtypes, Mahmoud and colleagues provided evidence for a favorable outcome when high numbers of CD20 + TiBc are present. Of note, this prognostic value was independent from any clinicopathological marker [11]. So far, very few studies assessed the functional relevance of TiBc and the antibodies produced thereof. Using recombinant Ab cloning techniques, Hansen et al. reported an antigen-driven humoral immune response directed against b-actin exposed on apoptotic mammary carcinoma cells [12]. Following engraftment of lung cancer tissue in immunodeficient mice, Yasuda and co-workers identified TiBc producing tumor specific Abs against mutated p53 [13]. Thus, at least a proportion of TiBc accumulating in solid tumors can produce tumor antigen specific antibodies. TiBc could consequently be helpful for the identification of novel or at least of relevant tumor specific antigens (TSA).
Despite the obvious importance of B cells in immunological circuits, the functional role of TiBc was so far not examined in detail. This may be attributable to the difficulties in obtaining sufficient cell numbers, either by fresh isolation or by long-term culture. To overcome this obstacle, we designed a direct and reproducible method to generate immortalized B cell lines from primary tumor tissues. We used EBV-transformation by culturing B cells in the presence of virus-containing supernatant of the marmoset cell line B95/8 [14]. Clonality of outgrowing TiBc cultures, isotypes and function of secreted immunoglobulins (Ig) were analyzed.

Tumor specimens
Resection specimens of primary CRC without prior chemo-or radiotherapy (n = 19) were received fresh from surgery. Prior written informed consent was obtained from all patients, and all procedures were approved by the Ethics Committee of the University of Rostock (reference number II HV 43/2004) in accordance with generally accepted guidelines for the use of human material.
Molecular classification was done according to [15]. These data as well as staging information compiled from the clinical charts are summarized in Table 1.

Generation of TiBc cultures from primary tumors
Tumor fragments were cut into small pieces and homogenized. Cell suspensions were obtained following nylon passage. Either a total of 5610 6 cells were cultured in a cell culture flask or cells were distributed in one 48-well plate in the presence or absence of EBV-containing supernatant of B95/8 cells. To prevent T and NK cell growth, cyclosporin A (5.5610 27 M) was added. Typically, outgrowth of TiBc cultures was observed after a period of 4-8 weeks. Established TiBc were harvested, expanded and further characterized. All cell culture reagents were obtained from PAA (Cölbe, Germany), antibiotics and antifungal agents were from the pharmacy of the university hospital Rostock.

Clonality analysis of TiBc
Genomic DNA was extracted from pelleted TiBc cultures with the Qiagen extraction kit (Hilden, Germany). Clonality was tested using the BIOMED-2 multiplex PCRs protocols to detect Ig gene rearrangements with Genescan detection [17]. Denatured PCR products are size separated in a denaturing capillary sequencing polymer and detected via automated scanning with a laser. Monoclonal cell samples will give rise to PCR products of identical size (single peak), whereas in polyclonal samples many different IGH PCR products will be formed, which show a characteristic Gaussian size distribution.

Histological and immunohistochemical assessment of TiBc
Peritumoral lymphocytic infiltration patterns were assessed by routine H&E staining. Additionally, representative paraffin-blocks were selected for immunohistochemistry with commercial antibodies directed against CD3, CD20, CD79a, and CD21 using standard protocols.

HIg-and Cell-ELISA
Ig isotype production of TiBc was quantified by ELISAs specific for IgG, IgA, IgM, IgD and IgE (Bethyl Laboratories, Montgomery, TX, USA). Briefly, ELISA plates were coated with antihuman Ig (50 mg/ml; 4uC), washed and blocked. TiBc supernatants (1:50; collected for a period of six month) were added, bound Ig was detected with horse-radish-peroxidase-conjugated antihuman Ig (1:50.000), and visualized by 3.3.5.5-Tetramethylbenzidin. Reactions were stopped using 2 M H 2 SO 4 . Absorbance was measured on an ELISA plate reader at 450 nm (reference 620 nm). Ig concentrations were determined by comparison with a standard curve generated from serial dilutions of Ig standards.
For Cell-ELISA, 96-well plates, seeded with fixed tumor cells, were used. All washing and blocking steps were done as described. One mg/well IgG was added and incubated at room temperature (2 h). Therapeutic Abs Cetuximab TM (anti-EGFR) and Trastuzumab TM (anti-Her2/neu) served as positive, Rituximab TM (anti-CD20) as negative controls. As a measure for IgG tumor cell binding, extinctions from the background (medium alone, ,0.20) were subtracted.
In vitro Binding assays CRC cell lines (0.5610 6 ) were fixed in 4% formalin for 30 min and incubated with IgG-supernatants from several TiBc cultures (1 mg/ml; 1 h; 4uC). Therapeutic control Abs were applied as described. After washing, cells were stained with secondary antihuman IgG-FITC conjugated Ab (1:40). Data acquisition was done as described.
Immunofluorescence was performed with fixed tumor cells followed by blocking (2% BSA; 1 h) and incubation with IgGsupernatants from selected TiBc clones (1 mg/ml, 2 h). Afterwards, Alexa 546-conjugated anti-human IgG (1:100, 1 h) was applied. Cell nuclei were counterstained with DAPI. Tumor cell binding was visualized by fluorescence microscopy (Leica DM 4000 M; Leica Wetzlar, Germany). Pictures were obtained using the provided evaluation software (Leica Image Pro Plus 6.0 image analysis software).

Statistics
Statistical differences were determined by using the two-sided Fisher's exact T-test. The tests were performed with Sigma-Stat 3.0 (Jandel Corporation, San Rafael, CA, USA). The criterion for significance was taken to be p,0.05.

Patterns of lymphocytic infiltration
TiBc were observed to concentrate in the stroma at the invasive margin, and they were never seen between the epithelia of the neoplastic glands ( Figure 1A, B). Distribution of TiBc, tumor-infiltrating CD3 + T cells, as well as mature plasma cells was quite heterogeneous, even within a single tumor (Figure 1 C, D). In addition, at the invasive margin follicular aggregates of TiBc residing in a mesh of follicular dendritic cells were seen ( Figure 1E, inset), and these varied in frequencies between cases.

Generation of EBV-transformed TiBc cultures
Establishment of TiBc bulk cultures was attempted by EBV transformation for 19 tumors (Table 1). Initially, we generated 14 TiBc bulk cultures (74% success rate) but permanent growth could not be obtained in three cases. However, the remaining 11 cultures proceeded to grow continually. In these, a characteristic in vitro pattern of follicular aggregates was observed -similar to peripheral EBV-transformed B cells. Success in obtaining cultures did not correlate with density or pattern of lymphocytic infiltration of the primaries. Moreover, it was quite independent from the amount of TiBc harvested upon initial isolation (Table 1). However, due to the technical challenge of FACS analysis from primary tumor tissue, those data on CD19 + CD20 + B cells in primary CRC need to be interpreted with caution.
To sum up these findings, this method circumvents problems implicated with low B cell numbers obtained by isolation from fresh primary tumors and the technical demands of handling freshly isolated TiBc [11].

TiBc cultures are of clonal origin
Clonality analysis for IGH and IGK gene rearrangement was performed from the 11 TiBc bulk cultures. Comparison of V(D)J rearrangements and junctional mutation patterns revealed monoclonality in each of the established culture. By applying Gene Scanning fragment-length PCR, predominant clonal rearrangement (monoclonality) was identified as a single peak consisting of one type of PCR product (exemplary results are given in Figure S1A). Analysis of early passages of the cultures demonstrates, however, oligoclonal junctional mutation patterns and thus implies that monoclonality is the result of continuous in vitro selection ( Figure S1B).
Equally important, demonstration of monoclonality for all of the long-term cultures forms the basis for TiBc characterization and applying Abs produced thereof for functional analysis.
TiBc are of mature immunophenotype The most interesting result was the high expression of the B cell maturation marker CD23, suggesting growth after antigenic stimulus ( Table 2). Consistent with ongoing maturation, in each clone a heterogeneous pattern of B and plasma cell markers was observed ( Table 2). Expression of the general activation marker CD83 ranged from zero to 60%. None of the clones expressed T cell markers confirming their B lymphoid origin. Expression of MHC class I and II molecules as well as the co-stimulatory adhesion molecule CD80 was very high.

Ig repertoire produced by TiBc clones
All of the analyzed supernatants contained detectable, though varying levels of either one of the three main Ig isotypes: IgG, IgM or IgA ( Table 2). Production of IgD or IgE was not observed in any of the cultures. The pattern of Ig secretion did not alter during the observation period. Importantly, among the 11 cultures that had been included into the analysis, exclusive production of IgG was found in 55% (6/11 cultures).

IgGs specifically bind to tumor cells
Of the five TiBc-derived IgGs included into these analyses, four were observed to bind to HCT116 and three to HROC24 and HROC46 (Table 2) in initial cell ELISA screenings. Of note, the affinities of TiBc-produced IgGs were comparable to or even higher than those of the therapeutic Abs specific for EGFR and Her2/neu. These two Abs were chosen as positive controls due to their high expression on the analyzed cell lines (each 80-100%, Figure 2A).
Specific allogeneic tumor cell recognition was confirmed for three of the IgGs by flow cytometry ( Figure 2B). IgG from TiBcHROC66 and TiBcHROC90 exhibit strong avidity towards all cell lines tested and both bound to .90% of the tumor cells. Likewise, the mean fluorescence intensity, representing the number of bound Ab molecules per cell, differed only marginally between cells (data not shown). TiBcHROC57-derived IgG displayed an intermediate binding affinity with a maximum of 15% staining for HROC46 cells. When analyzing the other IgGs, weak tumor cell binding if any was observed. Less than 5% of tumor cells allowed binding of IgG from TiBcHROC59, 68, and 86 ( Figure 2A).
Finally, immunofluorescence studies were done for further confirmation of the above findings and to get an idea of the cellular antigen distribution. Tumor cell recognition was specific, and this was most pronounced for supernatants from TiB-cHROC66 and 90. Both IgGs bound to HCT116 cells with high affinity (Figure 3, upper panel). Target cell staining with TiBcHROC57-derived IgG was weaker, but still specific and mostly intracellular (Figure 3, middle panel). Again, negligible staining was shown for TiBcHROC68-derived IgG. Similar observations were made for HROC46 with highest binding for IgG from TiBcHROC66 and TiBcHROC90. TiBcHROC57derived IgG bound tumor cells to a lesser extent. Positive staining was shown for some, but not all tumor cells (Figure 3, lower panel).

No evidence for latent EBV-transformation in vivo
Growth of lymphoblastoid tumors occasionally complicates xeno-transplantation of tumors in mice [18]. This gave rise to the hypothesis that (some) TiBc could be latently infected by EBV and hide in the immunosuppressive environment of human CRCs to escape EBV-specific T cells [18]. Our in vitro system provided a perfect basis to test this hypothesis. TiBc cultures from another 12 CRC were cultured in the presence or absence of EBV (Table 3). There was no outgrowth in any of the cultures without exogenous EBV, whereas with EBV, a total of 97 TiBc clones was obtained from 10/12 cases. Additionally, this simple procedure facilitates the generation of several TiBc clones from one tumor specimen.

Discussion
In various human cancers, CRC among them, the degree of lymphocytic tumor infiltration correlates with patients' clinical course. Previous studies largely focussed on T cells [5,6]. Besides, a few preliminary studies described the positive impact on systemic humoral immune responses not only for early detection but also for cancer immunotherapy [19,20]. We here began to elucidate the role of the humoral anti-tumor immune response within the tumor microenvironment. First, by EBV-transformation we were able to generate long-term monoclonal B-cell lines from TiBc. This procedure was comparatively easy to carry out and quite efficient, yielding 11 lines from 19 CRC specimens. We wish to point out, that this protocol, backed up by proof of monoclonality by the BIOMED protocols used in the diagnostics of lymphomas, offers itself as a basis to start from for any researchers working in this field. It circumvents efficiently the problems implicated in handling fresh TiBc isolates that often are very low in numbers.
The mature immunophenotype of TiBc from primary CRC observed in this study suggests activation and antigen-induced maturation, making accidental tumor tissue infiltration quite unlikely. In agreement with these observations, B cells infiltrating human breast carcinoma tissues were shown to undergo antigendriven proliferation, somatic hypermutation and affinity maturation, thus representing a tumor specific humoral immune response [11]. This interpretation is further supported by the fact that, as seen by immunohistochemistry, in CRC the great majority of TiBc resides in follicular aggregates. Likewise, peritumoral follicular aggregates of lymphocytes have long ago been reported as ''Crohn's like reaction''. This positive prognosticator for CRC was interpreted as an immune-mediated anti-tumor effect [21,22].
Most of the TiBc cultures started to produce IgG directly after cultivation, IgA secretion was found in one case. Beside IgG, IgA is known to be the other main isotype, secreted exclusively by B cells following a specific stimulus, and this provides a further argument for our hypothesis that B cell infiltration into the tumors is not accidental. Unraveling the functional capacity of this IgA will be of great interest, however in this work we focused on IgG-secreting TiBc cultures. IgG is generally considered to have the greatest potential for therapeutic purposes, because it is most efficient in mediating complement-or NK cell-mediated tumor cell lysis, and has an extended plasma half life [23][24][25][26].
Using functional assays, specific binding to tumor cells was demonstrated especially for two of the IgGs (from TiBc66 and TiBc90). Both intra-and extracellular antigens were recognized and since all binding experiments were performed in allogeneic settings, these antigens must be shared between different tumor cells. These findings are in line with and add to reports on TSA by other groups [11,12,27]. Investigations to identify the antigens bound by TiBc-derived IgGs at the molecular level are on their way, and we are confident that this will help to understand the process of CRC immunoediting [28].
High expression of MHC class I and II molecules as well as the co-stimulatory adhesion molecule CD80 is another interesting finding in this study. This suggests a role for TiBc not only as producers of tumor-specific Abs but also as local APCs, potentially modulating T cell restricted immune responses. One might speculate that tumor-antigen specific TiBc constitute highly efficient APC especially when compared with antigen-unexperienced circulating B cells. However, recruitment of TSA-specific B cells (namely regulatory B cells) by tumor cells in order to suppress antitumoral T cells in situ is also imaginable [29].   Table 3. Data of colorectal carcinomas used for generation of TiBc cultures with and without EBV, percentage of CD19 + CD20 + B cells from primary tumors (flow cytometry), and outcomes. Finally, it was reported that EBV might induce transient production of low-affinity, unstable IgM [30]. Since isotype switch during long-term culture was not observed in any of the TiBc cultures, we consider it most likely that our cases of TiBc IgM production (n = 4) are a primary antigen-driven immune response and not induced by EBV transformation. It will be of interest to analyze, if stimulation of the IgM-producing TiBc with autologous T helper cells will induce Ig class switching, possibly leading to the production of highly affine IgG. Secondly, since tumor-specific IgM antibodies have been described, specific recognition of tumor structures may be anticipated [31][32][33].
The major findings of our study may be summarized as follows: (a) human CRCs harbor B cells (i.e. TiBc) that can be long-term cultured subsequent to EBV-immortalization; (b) TiBc harvested from these are antigen-experienced and secrete Ig of the three main Ig isotypes (IgG, IgA, IgM) in amounts sufficient for functional studies; (c) IgG derived from several TiBc clones strongly bind to antigenic structures present on several allogeneic tumor cell lines; and (d) TiBc show no evidence for latent EBVinfection in vivo.
On the basis of these novel findings, Ig-producing long-term TiBc cultures may provide a convenient cellular tool for the identification of novel or relevant TSA. Figure S1 Clonality analysis using the BIOMED-2 multiplex PCRs protocols to detect Ig gene rearrangements with Genescan detection. (A) IGH and IGK gene rearrangement from selected TiBc cultures giving rise to PCR products of identical size (single peak). This confirms monoclonality of TiBc cultures. Monoclonal B-LCLs were taken as control. (B) Exemplary data for early passage TiBcHROC68 culture. Oligoclonality was detected by different PCR products indicating the presence of several subclones at early culture. (TIF)