IN THIS ISSUE

Progenitor cells migrating to the fetal thymus from liver and bone marrow generate NK cells and dendritic cells (DCs) in addition to T cells. However, the cell lineage potential of T cell progenitors in the adult thymus has not been studied rigorously. Lu et al. (p. 5848 ) isolated early progenitors (c-kit⁺CD25⁻ and c-kit⁺CD25⁺ cells) from thymi of adult mice by flow cytometry and cultured them with a 2′-deoxyguanosine-treated fetal thymus lobe. In vitro generation of CD4⁺CD8⁺ cells was slower for adult than fetal c-kit⁺CD25⁻ cells. The level of expansion before TCRβ-chain gene rearrangement in single-cell cultures was much higher for adult c-kit⁺CD25⁻ vs adult c-kit⁺CD25⁺ cells and was significantly higher vs fetal c-kit⁺CD25⁻ cells. Adult thymus progenitors produced ∼100 times more CD4⁺CD8⁺ T cells in thymi of irradiated recipients than did fetal thymus progenitors. Using their multilineage progenitor assay on individual cells and a limiting dilution analysis, the authors demonstrated that adult thymus c-kit⁺CD25⁻ cells contained predominantly T lineage cells. Adult thymus c-kit⁺CD25⁻ progenitors with potential for T plus NK cells or T cells only generated DCs after culture of individual cells in a modified fetal thymus organ culture system. The data show that the process of early progenitor commitment for T cell, NK cell, and DC lineages is similar in the adult and fetal thymus, although T cell progenitors in the adult thymus generate more T cells.

Although numerous immunization strategies incorporating tumor-associated Ags have been attempted, patient benefits have been minimal. One explanation is that the numbers of Ag-specific CD8⁺ T cells generated are too low to be effective. Rosenberg et al. (p. 6169 ) studied the effect of peptide vaccination on 95 melanoma patients who had no clinical evidence of disease but were at high risk for recurrence. Patients were injected with a synthetic form of the melanoma peptide gp100_209–217, altered by a single amino acid substitution, in IFA using three different 48-wk injection schedules. Peptide-reactive or peptide-binding cells were detected using an ELISPOT or tetramer assay, respectively; an in vitro sensitization assay measured IFN-γ release by PBMCs in response to peptide. The first two assays showed that 53–89% of all patients were immunized to the native melanoma peptide. The third assay indicated 53% of patients had significant release of IFN-γ from peptide-stimulated PBMCs. Six of 17 patients who developed highly reactive peptide-specific CD8⁺ T cells also had PBMCs with strong reactivity against two HLA-A2⁺, but not against two HLA-A2⁻, melanoma cell lines. By the in vitro sensitization assay, PBMCs from 33 of 62 patients reacted specifically to the HLA-A2⁺ melanoma cell lines. However, recurrence rates at 3 years were 45–65% with no significant differences among the three study groups. Numbers of peptide-specific CTLs were comparable statistically between patients who recurred and those who remained disease free. The data indicate that induction of tumor Ag-specific CD8⁺ T cells is not a good measurement of successful anti-melanoma vaccination.

tudies indicate that glucocorticoid treatment lessens the severity of multiple sclerosis and its murine counterpart, experimental autoimmune encephalomyelitis (EAE). In Powell et al. (p. 5611 ), the Whitacre laboratory followed up on their previous observation that high levels of stress-induced corticosterone correlated with reduced EAE severity. They examined the role of macrophage migration inhibitory factor (MIF), an antagonist of glucocorticoid activity, in the disease process. EAE was induced in wild-type or MIF^−/− mice by immunization with a specific myelin oligodendrocyte glycoprotein peptide. Although there was no difference in clinical score or overall incidence of disease during the acute phase, the cumulative disease score was significantly lower in the MIF-deficient animals, all of which recovered around day 20 postimmunization. Serum corticosterone levels were comparable in the two groups of mice before immunization; however, the levels were higher in the immunized MIF-deficient mice before and during the acute phase. Peptide-stimulated MIF^−/− lymphoid cells had higher production of IL-4 during the acute phase and of IL-10 during the progressive phase but lower production of IFN-γ before disease onset and of TNF-α and IL-6 during acute disease compared with wild-type cells. The authors conclude that EAE disease progression, but not onset, is exacerbated by MIF-induced production of Th1 cytokines and glucocorticoid antagonism and that high levels of corticosterone can modulate disease progression.

Although it is known that JNK1 is required for Ag-stimulated expansion of CD8⁺ T cells in vitro and that more chemically induced skin tumors develop in mice lacking JNK1, the role of JNK1 in the immune surveillance function of CD8⁺ T cells is unknown. Gao et al. (p. 5783 ) found that tumors developed earlier and were larger in both JNK^−/− and IFN-γ^−/− mice injected with melanoma or lymphoma cells compared with wild-type controls. The high susceptibility of CD8^−/− mice to injected tumor cells was reduced by reconstitution with wild-type CD8⁺ T cells but not JNK^−/− or IFN-γ^−/− CD8⁺ T cells. Both IFN-γ gene transcription and IFN-γ production were lower and delayed in naive JNK^−/− CD8⁺ T cells stimulated with anti-CD3 mAb plus anti-CD28 mAb in the presence of cytokines compared with wild-type CD8⁺ T cell controls. Similar results were found for JNK^−/− CD8⁺ T cells transgenic for an OVA-specific TCR cocultured with OVA peptide-loaded dendritic cells. Expression of T-bet was lower and eomesodermin was down-regulated in activated JNK^−/− CD8⁺ T cells compared with wild-type cells; both transcription factors contribute to IFN-γ production in CD8⁺ T cells. Perforin levels were reduced significantly in OVA/dendritic cell-activated TCR transgenic JNK^−/− CD8⁺ T cells vs wild-type cells. In vivo cytolytic activity of naive JNK^−/− CD8⁺ T cells was reduced >30% over that of wild-type CD8⁺ T cells. The authors provide evidence that tumor immune surveillance by CD8⁺ T cells is regulated through JNK1 control of IFN-γ production and CTL function.

Human memory T cells are defined by CCR7 and CD45RA expression. Naive T cells (T_N) express both markers, central memory cells (T_CM) express only CCR7, effector memory (T_EM) cells lack expression of both, and a third subset of memory cells (T_EMRA) express CD45RA but not CCR7. There is no consensus regarding the developmental relationship among the memory cell subsets. Willinger et al. (p. 5895 ) used oligonucleotide arrays to detect two main clusters (T_N/T_CM and T_EM/T_EMRA) of gene expression in highly purified CD8⁺ T cell subsets separated from PBMCs of four healthy donors. Clusters of differentially expressed genes fell into two major patterns (effector memory signature with high expression on T_EM/T_EMRA or naive signature with high expression on T_CM) plus two minor patterns (high expression in T_CM and T_EM or T_CM specific). Effector memory signature genes represented a variety of immune response functions that correlated with protein expression studies. FACS sorting of cytokine-stimulated cells with phosphoprotein-specific intracellular staining of STAT and NF-κB molecules indicated differential activation of cytokine signaling pathways among the four CD8⁺ T cell subsets. The cytokine signaling signatures and the transcription profiling correlated in that T_CM was intermediate between T_N and T_EM/T_EMRA. The data indicate that T_CM have features of both T_N and T_EM and are consistent with a model by which low-dose Ag promotes differentiation of T_N into noneffector T_CM, whereas high-dose Ag promotes differentiation of T_N into effector T_EM. The data do not support a linear differentiation model by which T_N differentiate into effector T_EM and then into T_CM.

Transfer of APC determinants to T cells during interaction of the TCR with MHC/peptide complexes, through a process termed trogocytosis, occurs during T cell development. Romagnoli et al. (p. 5751 ) found that immature CD4⁺CD8⁺ thymocytes (DP) had barely detectable levels of I-A, CD4⁺CD8^int and mature CD4⁺CD8⁻ thymocytes (CD4SP) had high levels of I-A, and CD4⁻CD8⁺ (CD8SP) had less I-A than CD4SP. I-A was higher on activated CD69⁺ than on unactivated CD69⁻ DP. DP from TCR-α^−/− mice had negligible amounts of I-A. I-A levels were higher on heat stable Ag (HSA)^high than on fully differentiated HSA^low CD4SP in wild-type mice, but the levels on the two cell groups were the same in TCR transgenic mice incapable of negative selection. Among CD4⁺ CD8^int cells, those with CD25^high had higher I-A and TCR levels than CD25⁻ cells; among CD4SP, TCR levels were higher on CD25⁻ than on CD25^high cells. Maintenance of the higher levels of I-A captured by CD4⁺CD25^high thymocytes required continuous interaction with stromal thymic cells. I-A levels on DP from lethally irradiated C57BL/6 mice reconstituted with MHC class II-deficient bone marrow were comparable to those on DP from control wild-type chimeras, whereas I-A levels were lower on CD4⁺CD8^int and HSA^low CD4SP from the MHC class II-deficient chimeras. I-A levels on CD4⁺CD8^int and CD25^high CD4SP were comparable in the MHC class II-deficient chimeras, but I-A levels were lower on CD25^high CD4SP from the wild-type chimeras. A high level of I-A was detected on peripheral CD4⁺CD25^high lymphocytes, but the level was lower than on CD4⁺CD25^high thymocytes. The results suggest that I-A acquisition by developing regulatory T cells is a TCR-mediated, activation-dependent process that serves as a molecular signature for thymic selection events.

Overexpression and immunogenicity of p53 in human tumors render the molecule an attractive target for immunotherapy. However, the inability to isolate human CTLs with high avidity for p53 limits this approach. Cohen et al. (p. 5799 ) transduced primary human PBLs with a single retroviral vector expressing both α- and β-chain cDNAs derived from an antihuman p53 peptide TCR of a highly avid CTL from an HLA-A2 transgenic mouse. IFN-γ release was detected in cocultures of transduced lymphocytes plus p53 peptide-loaded T2 lymphoblastoid cells or HLA-A2-transfected cells carrying a vector expressing wild-type or mutant p53. HLA-A2⁺p53⁺ tumor cell lines also stimulated the TCR-transduced lymphocytes to proliferate and to secrete IFN-γ, IL-2, and other cytokines. The incubated transduced lymphocytes had elevated CD107a expression and lysed established tumor cell lines and fresh HLA-A2⁺ p53⁺ melanoma samples in a ⁵¹Cr release assay. Normal cells or mock-transduced PBLs did not exhibit those activities, and tumor samples lacking HLA-A2 did not elicit cytokine production from the transduced lymphocytes. Both CD4⁺ and CD8⁺ T cell fractions of the TCR-transduced PBLs lysed HLA-A2⁺p53⁺ tumor cell lines but not control cell lines. The authors demonstrate the ability to transduce a single vector expressing both chains of a murine p53-specific TCR into human PBLs and to generate CD4⁺ and CD8⁺ T cells that recognize mutant p53 in tumor cell lines and in fresh melanoma samples.

Summaries written by Dorothy L. Buchhagen, Ph.D.