Induction of T cell development and establishment of T cell competence from embryonic stem cells differentiated in vitro

Thomas M Schmitt¹, Renée F de Pooter¹, Matthew A Gronski², Sarah K Cho^{1, 3}, Pamela S Ohashi² & Juan Carlos Zúñiga-Pflücker¹

Summary:

Embryonic stem cells(ESCs) are undifferentiated cells which are pluripotent and capable of being differentiating into all cell types. Direct injection of ESCs into a recipient will lead to teratoma formation, where there is spontaneous differentiation into mixed cell types and tumor-like group; thus, in most cases, ESCs must be differentiated first before they can be used in tissue engineering therapies. In this paper, the authors used a basic co-culture technique in order to induce differentiation. T-cells start their life cycle in the bone marrow, where they differentiate from hemopoaetic stem cells, then migrate to the thymus to mature. Thinking that the conditions in the bone marrow might be important to induce differentiation, the authors co-cultured ESCs with the bone marrow stromal cell line OP9. However, just co-culturing with OP9 induces b-progenitor differentiation, thus the authors tried co-culturing with OP9-DL1(OP9 expressing Delta-like ligand 1). Delta is the transmembrane protein expressed on the cell surface that binds to the receptor Notch, which unleashed a well-studied signaling cascade leading to changes in protein expression. The authors monitored both CD25 and CD44, hallmarks of the T cell lineage, and Il7r and RAG1(among other proteins), which are shown on both B and T cell progenitors, and in the ESC-OD9-DL1 cultures, found the expression of all of these.

Figure: RT-PCR for various genes, some of them controls, and others that are proteins expressed in lymphocytes and CTLs.

The next step was to determine whether or not these cells would follow the normal differentiation pattern when injected into a thymic microenvironment. First, flow cytometry was performed to isolate the CD25+/CD44+ and CD25+/CD44- double negative T cell progenitors after 2 weeks. You see, before T cell maturation in the thymus, they are negative for both the CD8 and CD4 cell surface ligands(the ones that you see on natural killer and helper t cells, respectively), and thus are termed “double negative.” Then, these cells were injected into CD45.1-congenic fetal thymic lobes, and cultured for 5 days. These thymic lobes were then injected under the skin of sublethally irradiated Rag 2 -/- mice. These two factors ensured that the mice did not have any native T-cells, and could not recombine any native T cells (Rag 2 is necessary for VDJ recombination). In a non-grafted control, they were devoid of CD4+ and CD8+ T cells. But sure enough, in the grafted control, the T cell repertoire was reconstituted, and there were both CD4 and CD8 cells present in the spleen and lymph nodes. These mice were then challenged using lymphocytic choriomeningitis virus, and remarkably, these new T cells proved to be able to mount an effective immune response on par with the normal response from normal C57BL/6 mice.

Figure: B: FACs for various T cell markers, comparing cells taken from the splenocytes of the Rag 2 -/- 3 weeks after graft.
C: Graph quantifying Chromium 51 assay data on cell lysis. Gp33 is the LCMV while AV is a control.

Critiques:

The authors of this paper were generally very meticulous, making sure that the appropriate controls were set, and looking at a wide variety of factors to ensure that indeed T-cells were being grown in the ESC-OP9-DL1 cultures. However, there were still a few things. The authors used reconstituted thymiclobes from CD45 congenic B6 mice, and cultured them in FTOC. Their method of lymphocyte depletion was to culture thse lobes in a medium containing 1.25 mM deoxyguanosine for 5 days. However, it is possible that some native T-cells may have survived, and that these were the ones that were competent and that these cells were the ones which expanded and reconstituted the immune system of the Rag 2 -/- mice when injected. Instead of depletion using deoxyguanosine, they could have used have used thymuses from Rag 2 -/- mice instead, or at least screened for any residual T cells that may have survived. In the methods, they even said that “nearly all thymocytes were of donor origin,”meaning that some endogenous thymocytes had remained. In fact, in another experiment, the one where they challenged mice with LCMV, they used fetal thymic lobes from Rag2 -/- mice, as opposed to from the wild-type mice. One wonders why they switched the source of lobes for this last experiment, and why they did not do this before. Lastly, they only compared the LCMV responses of the Rag2 -/- mice injected with ESCs and wild-type mice, and did not include a “negative control” using normal Rag 2 -/- mice to see that they would indeed die off, or at least have different levels of cell death through the Chromium 51 release assay which they used to test for T cell activity.

Nature Immunology 5, 410 - 417 (2004)
Published online: 21 March 2004; | doi:10.1038/ni1055

http://www.nature.com/ni/journal/v5/n4/abs/ni1055.html