Here we describe that peripheral human B cells from healthy donors were capable of very extended proliferation after specific external activation, justifying these cells' description as being conditionally immortalized. By CD40L/IL-4 stimulation, B cells were maintained in a proliferative and activated state characterized by the expression of activation markers and the absence of telomere erosion. They proliferated beyond the 170–180 population doublings that have been proposed to indicate immortalization of lymphocytes 
. Their long-term proliferation remained dependent on the signals provided by exogenous CD40L and IL-4, both in early and late passage. Therefore, the state of these cells bears a much closer resemblance to physiological T-cell driven clonal expansion than to malignant transformation.
The CD40/IL-4 cultivation system for human B cells was first described many years ago 
. Similar CD40 systems have since been used for investigations in various fields like B cell differentiation, immunoglobulin production and T cell activation, documented in many studies of which we can only mention same examples 
. It seems therefore remarkable that the generation of conditionally immortalized B cell lines with a CD40 system has, to our knowledge, not been described up to now. visualizes the fact that in previous studies CD40-activated B cells were generally cultivated for about 70 days or less. Therefore, they had undergone much less proliferation doublings than required to cross the limits to proliferation that were described for human cells that are mortal in vitro, like fibroblasts 
, CD8+ T cells 
, or EBV-transformed B cells 
Culture periods of CD40/IL-4-stimulated B cells reported in different studies.
Previously, CD40/IL-4-activated B cell cultures, in the hands of different investigators, could have different fates. In their initial study 
, Banchereau et al. described two phenomena interfering with CD40/IL4-driven B cell outgrowth: the spread of EBV in the cultures in later passage, and an unexplained “dying out” of the cultures before 10 weeks of cultivation. O'Nions and Allday 
observed that CD40-activated B cells ceased to proliferate even earlier, associated with their differentiation to plasma cells 
. Schultze et al. 
cultivated CD40-activated B cells for up to 65 days without observing B cell differentiation or a decrease in proliferation, but again EBV infection was found in an unspecified proportion of the cultures. The authors of the latter study, in contrast to others, also used cyclosporin A in their culture protocol, which might have been an important factor in achieving prolonged B cell proliferation, and it appears possible that they might have achieved unlimited CD40/IL-4-dependent proliferation using their protocol, provided EBV could have been eliminated.
We suppose that the generation of conditionally immortalized B cells in this study was made possible due to (a) the use of cyclosporin A 
, reducing the in vitro reactivation of T cells, (b) the elimination of EBV by using small numbers of 100,000 PBMCs or less, containing 20,000 B cells or less, to initiate each culture, and (c) a reduction of the absolute number of T cells introduced into each culture (<90,000), again by using small numbers of unpurified PBMCs per culture. To our knowledge, none of the previous studies combined all of these three aspects. For example, Banchereau et al. 
purified B cells per culture (>97% purity) in the absence of cyclosporin A. Compared to our protocol, less T cells were introduced per culture, but the risk of reactivating them was raised, and more total B cells were used per culture, raising the risk of introducing EBV. Similar considerations apply for the other studies. It is likely that the inhibition and elimination of T cells in such a culture system is crucial because both B cell-specific and, if EBV is present, EBV-specific T cells might have a role in preventing outgrowth, in analogy to the situation in EBV-mediated B cell outgrowth 
We observed that conditionally immortalized CD40-activated B cells stabilized or re-increased their telomere length over time. Upon stimulation in vivo
in germinal centers, human B cells can elongate their telomeres even beyond their length in naive B cells 
. Telomere elongation can also be a result of in vitro
B cell stimulation 
, and the maintenance of telomerase activity in activated B cells for several weeks has been described 
. Still, on average, the telomeres of B cells, as well as those of T cells, are becoming shorter during a human lifetime 
, suggesting that endogenous telomerase cannot globally compensate for telomere loss during extended proliferation in vivo 
. Our results show, however, that in vitro
a subset of B cells may fully compensate for telomere loss during extended periods of time if a sufficiently strong exogenous stimulus is regularly provided. Specific immune responses depend on extensive clonal expansion of specific T or B lymphocytes, and telomerase is very probably critical in securing immune function by maintaining the lymphocytes' replicative potential 
. Notably, although T cell activation is coupled to telomerase activation, telomeres of T cells inevitably shorten after their primary activation 
, and to achieve immortalization of T cells in vitro
it was generally necessary and sufficient to ectopically express telomerase 
. Consistent with previous reports, our data suggest that the situation in human B cells is different. At least a subset of them appears to be equipped with an immortalization program that ensures telomere stabilization and can be accessed and maintained by applying extracellular ligands only. Such a difference in the proliferative potential of B and T cells might not be surprising, because B cells, having to undergo somatic hypermutation accompanied by heavy selection of affinity-matured cells, probably require an even greater proliferative potential than T cells.
Our cytogenetic analyses showed that a majority of long-term CD40-stimulated B cell lines contained cells with an intact karyotype, either exclusively or to a significant proportion. This situation is in marked contrast to that found in EBV-transformed B-lymphoblastoid cell lines (LCLs), where long-term growth and immortalization is necessarily associated with the acquisition of chromosomal aberrations, usually in combinations more complex than observed in any of the CD40-stimulated B cell lines analyzed by us 
. Moreover, the necessity in LCLs to select for rare mutated cells capable of further proliferation often manifests itself in the form of a proliferative crisis, a phenomenon we did not observe in CD40-stimulated B cell lines. In this context, it is noteworthy that three of our B cell lines contained cells with an additional copy of chromosome 2, an aberration that has been associated both with in vitro senescence of human T cells from old-age donors 
and with human malignancies 
. One of our B cell lines homogeneously showed trisomy 2, and this line later underwent senescence. It is an interesting question whether the occurrence of trisomy 2 generally predicts senescence in this culture system. Taken together, although it was not possible to generate chromosomally intact long-term CD40-stimulated B cells from each normal donor, they could be obtained from several donors, and the level of genetic stability of this type of B cell culture appeared to be generally higher than that observed in LCLs and no lower than that of other cells that are cultured in bulk for extended periods, for example human embryonic stem cells 
. Thus, our data seem to indicate that the acquisition of genetic abnormalities is not required for the conditional immortalization of human B cells in this system.
We provide two examples of possible applications of long-term CD40-stimulated B cells. In our first example, we showed that they strongly and specifically react to stimulatory oligodeoxynucleotides with the secretion of cytokines. Studies on the activation of human B cells by innate immune receptor ligands usually rely on the use of primary B cells, which need to be purified in complicated procedures and are only available in limited numbers and in purities well below 100%. However, a small contaminating cell population can significantly disturb such experiments. The availability of normal CD40-activated B cell lines in unlimited numbers and essentially 100% purity could greatly facilitate such studies in the future.
In our second application of conditionally immortalized CD40-activated B cells, we showed that they can be used to specifically (re-)activate and expand antigen-specific CD8+ T cells in vitro. Dominant as well as subdominant memory T cell populations, and T cells that were likely a part of the naive repertoire, could equally well be expanded in vitro using peptide-loaded long-term CD40-activated B cells. Previous studies have shown in detail that shorter-term CD40-activated B cells can efficiently activate and expand populations of frequent or rare memory T cells and naive T cells 
. Here we provide evidence that conditionally immortalized B cells can be used to the same purpose.
To our knowledge, CD40-stimulated B cells are the first example of a differentiated cell type from healthy human donors that undergoes conditional immortalization in vitro, in the absence of genetic manipulation. Their conditional immortalization was achieved by external stimulation with analogs of physiological ligands. Thus, human B cells display what was previously thought to be a unique feature of human stem cells and neoplastic cells. The future will show whether non-invasive mechanisms will be identified to break the proliferation limit of other human differentiated non-neoplastic cell types, or whether it will emerge that B cells have singular proliferation properties that are rooted in their unique biological function. It will also be important to investigate if and how individual B cell subsets differ in their capacity to undergo immortalization. However that may be, conditionally immortalized CD40-stimulated B cells present new opportunities to study B cell function and differentiation, and they offer themselves as tools for individualized cell or gene banking and for use in cellular immunotherapy.