In This Issue

(See article on pages 669–678.)

The ER resident chaperone protein calreticulin plays a surprising variety of roles in cell regulation, some of which make it particularly appealing for antitumor vaccination. First, this protein can be cleaved in vivo to form the potent antiangiostatic factor vasostatin. In addition, when expressed in antigen presenting cells, calreticulin interacts with the peptide processing machinery to facilitate loading of associated peptides with the Class I MHC complex. Building on recent evidence that presentation of tumor-associated peptides through calreticulin can promote tumoricidal immune responses, Cheng et al. have engineered a fusion gene encoding a known viral tumor antigen, linked directly to the chaperone. They report here that mice exposed subdermally to this DNA construct enjoy seemingly complete resistance to tumor cells that express the viral antigen. Conversely, a simple mixture of the viral sequence and the calreticulin sequence has minimal effect, suggesting that calreticulin-mediated immune antigen presentation is crucial. Mice that are treated with calreticulin DNA but that lack CD8 T cells are susceptible to tumor growth, but even they appear to benefit somewhat from the effects of vasostatin, since their tumors show significantly less vascular development. It remains to be seen if this form of combined antiangiogenesis and antitumor immunity can be applied more generally by fusing calreticulin to other tumor antigens.

(See article on pages 709–715.)

The NF¹ gene encodes the tumor suppressor Neurofibromin 1, so named because of its effects on cells in the CNS, but myeloid cells with NF¹ mutations are also prone to uncontrolled growth. In one useful model of juvenile myelomonocytic leukemia (JMML), lethally irradiated mice receive hematopoietic cells from the liver of a fetal mouse lacking this growth-inhibiting protein. Although this protocol reliably yields overgrowth by mutant monocytes, Zhang and colleagues note that it is highly artificial, since leukemic cells would more typically arise as a minor population within a large number of normal cells. Testing the population dynamics in a more realistic setting, the authors show that the NF¹ mutation provides the cells only a limited growth advantage that depends on the cell's genetic background. In addition, although NF¹-/- liver cells can also yield fast-growing lymphoid cells, lymphocytic leukemias are not seen, either in recipient animals or in JMML families carrying an NF¹ mutation. In the case of the myeloid lineage, when NF¹ deficient cells are present below a threshold number, they outgrow normal cells for only a limited period, after which they represent a stable population for a period of months. Some additional event must therefore be required if the cells are to grow out of control after achieving this pattern of stable chimerism. The finding raises troubling questions about experiments modeling tumor growth using a homogeneous source of tumor-prone cells.

(See article on pages 765– 773.)

Immunosuppressive drugs generally lead to substantially improved blood cell counts in individuals with aplastic anemia (AA), a severe decline of all blood cell lineages in which the marrow is deficient in hematopoietic cells. For this reason, although AA's pathogenesis is obscure, it seems likely that autoimmune responses block hematopoiesis at an early stage but do not entirely eliminate early pluripotent stem cells. Here, Zeng and coworkers provide a first glimpse at some quirks of the immune system in AA patients. These authors have prepared T cells from within the marrow of 5 people with serious but treatable aplastic anemia. Cloning and DNA analysis of the T cell receptors from one of these patients shows that one sequence form of the complementarity determining region CDR3 is highly represented in patients' marrow. This form, dubbed JZ1.1, carries a specific heptapeptide sequence that is also found in the other patients studied but is absent in controls. Isolated T cells of this TCR type can suppress hematopoietic stem cell proliferation in co-culture experiments, diminishing the growth of myeloid and erythroid cell precursors alike. Interestingly, these JZ1.1 T cells appear to be specific for autologous stem cells derived from the patient at a time of severe disease-marrow cells derived from a healthy, MHC-matched control donor, or even from the patient after successful treatment, are not killed by exposure to JZ1.1 cells. Hence, it appears that T cells of this type are activated to respond to some still-unknown determinant that is present on early hematopoietic stem cells during periods of acute disease.