<<

HIGHLIGHTS

HAEMATOPOIESIS is the process embryos before circulation, has whereby mature cells of dis- been investigated previously by the tinct lineages are produced from authors. These experiments estab- mulitpotent haematopoietic stem lished that lymphoid potential is Stem cell origins cells (HSCs). Despite decades of restricted to precursors of intra- study, the origin of these HSCs dur- embryonic origin. In transplanta- ing mammalian embryogenesis has tion experiments using normal, remained unknown. Reporting in irradiated mice as recipients, nei- , Cumano and colleagues ther precursors from the YS or from show that, in mice, HSCs are gener- Sp were able to reconstitute an ated in an intraembryonic region, adult haematopoietic compartment. termed the splanchnopleura (Sp), As this might have been due to and not from the extraembryonic technical limitations of the system yolk sac (YS) blood islands. used, the authors developed a new During vertebrate embryogene- organ culture and cell-transfer sys- sis, haematopoietic cells first appear tem to investigate this further. The within the extraembryonic YS. As protocol involved culturing the Sp the and fetal liver (the and YS, again separated before the main haematopoietic organs in onset of blood circulation, for 4 mammals) require an input of days (to ensure a sufficient number exogenous haematopoietic precur- of HSCs were available), before sors to generate differentiated prog- transplanting the precursors into eny, it was suggested that HSCs recombinase activating gene (Rag)–/– originate in the YS and later migrate or Rag –/– crossed with common γ- to the fetal liver. However, the chain (Rag γc –/–)mice.Rag –/– mice intraembryonic Sp region has also lack T and B cells (so reducing pos- been shown to have haematopoietic sible competition between donor activity. and recipient precursors) and The differentiation potential of Ragγc–/– mice additionally lack nat- haematopoietic precursors from the ural killer cells (which might kill YS and Sp, separated from mouse haematopoietic precursors).

NATURAL KILLER CELLS But is this preferential proliferation actually driven by Ly49H recognition of MCMV? Vaccinia , which induces NK cell activation, did not cause the Natural killer selection selective proliferation of Ly49H+ cells, indicating this might indeed be an MCMV- Natural killer (NK) cells are often called murine cytomegalovirus (MCMV). This specific response. In addition, ‘innate ’ as they combine innate implies that NK cells are capable of virus- administration of Ly49H was recognition with the effector mechanisms of specific recognition. shown to inhibit bulk NK cell expansion in + T lymphocytes — secretion So, are Ly49H NK cells preferentially response to MCMV,which shows that γ γ (- ; IFN- ) and cytotoxic killing. activated by MCMV? The present study Ly49H has a direct role in MCMV-triggered However, in terms of responses to infection, assessed this in two ways — by scoring for NK cell activation. γ NK cells might be more similar to IFN- production through intracellular This study provides a new model of NK cell lymphocytes than previously thought. cytokine staining and by measuring NK cell activity in viral infection.Yet it remains to be A report from Dokun et al. in Nature proliferation using a bromodeoxyuridine seen whether this delayed, preferential indicates that, similar to T and (BrdU)-incorporation assay. In the early phase proliferation of NK cells is itself important B cells, -specific NK cells selectively of the response to MCMV there is a burst of for antiviral immunity. expand in response to infection. IFN-γ production by NK cells, which peaks at NK cells identify their targets by 36 hours. However, at 2 days post-infection, Jen Bell integrating signals from their inhibitory and there are comparable percentages of Ly49H+ References and links activation receptors. But not all NK cells are and Ly49H– NK cells producing IFN-γ, and ORIGINAL RESEARCH PAPER Dokun, A. O. et al. Specific created equal. Although their antigen there was no difference in the proliferation of and nonspecific NK cell activation during virus infection. Nature Immunol. 2, 951–956 (2001) receptors are invariant, NK cell subsets carry the two subsets. By contrast, the authors found FURTHER READING Cerwenka, A. & Lanier, L. L. Natural distinct complements of receptors. that by day 6 post-infection, there has been an killer cells, and . Nature Rev. Immunol. 1, 41–49 Previously, this group showed that, in mice, outgrowth of Ly49H+ NK cells, and these cells (2001) WEB SITE an activation of the family, proliferate to a much greater degree than their Wayne Yokoyama’s lab: http://www.hhmi.org/ Ly49H, confers specific protection against Ly49H– counterparts. research/investigators/yokoyama.htm

90 | NOVEMBER 2001 | VOLUME 1 www.nature.com/reviews/immunol © 2001 Macmillan Magazines Ltd Ragγc–/– mice injected with YS cells generated donor-derived myeloid cells, but this reconstitution was tran- sient, and after 3 months donor- derived progeny were no longer observed. No lymphocytes were ever detected in these mice. Therefore,YS cells cannot generate lymphocytes, but can provide short-term myeloid reconstitution. By contrast, after injec- tion of Sp-derived cells into Ragγc–/– recipients, donor-derived myeloid cells, as well as B and T cells were gen- erated. These myeloid and lymphoid cells were still present 8 months after injection, indicating that Sp-derived percursors can provide long-term NATURAL KILLER CELLS rejected. Tumour cells were rejected in wild-type reconstitution of recipient mice. mice depleted of CD8+ T cells and in Rag –/– mice The authors conclude that the (which lack T and B cells), but grew in Rag –/– mice only cells capable of adult long-term depleted of NK cells, indicating that conventional haematopoiesis are from the Sp Targeting NK cells are responsible for the rejection. Rae1β- region. So, during mouse embryo- or H60-transduced RMA cells were also rejected genesis, HSCs are generated intra- tumour cells in wild-type mice, but rejection required both embryonically and do not derive CD8+ T cells and NK cells. from the YS. The ‘missing-self’ hypothesis, formulated by The Lanier group also used the RMA cells to Jenny Buckland Klaus Kärre and colleagues in 1986, proposed investigate NK cell responses. RMA cells stably γ δ References and links that natural killer (NK) cells seek out and transfected with Rae1 or Rae1 were injected ORIGINAL RESEARCH PAPER Cumano, A. et al. destroy cells that have lost expression of major intraperitoneally into recipient mice. Mice Intraembryonic, but not yolk sac hematopoietic histocompatibility complex (MHC) class I injected with mock-transfected cells developed precursors, isolated before circulation, provide long-term multilineage reconstitution. Immunity . Two recent papers, published in Nature tumours and died, whereas mice challenged with 15, 477–485 (2001) and in Proceedings of the National Academy of transfected cells rejected tumour cells in an Sciences, now show that NK cells can reject NK-cell dependent manner. tumour cells that express ligands for the activating So, NK cells can reject tumour cells expressing NK receptor NKG2D, despite the expression of NKG2D ligands, despite MHC class I expression. MHC class I molecules by the tumour cells. But do mice primed with NKG2D ligand- The formulation of the missing-self hypothesis expressing tumour cells develop an adaptive predicted the existence of receptors on NK cells T-cell response against subsequent challenge by that inhibit their activity and which recognize non-transduced parental tumour cells? This is MHC class I molecules. Many of these receptors where the results from the two groups differ. have now been identified. Recent work has also Raulet and colleagues found that NKG2D ligand- identified several activating NK receptors, negative tumour cells were rejected by mice including the -like molecule NKG2D, whose previously challenged with transduced tumour engagement provides dominant activating signals cells, and that this was a CD8+ T-cell-dependent to the NK cell. Previous work by Tom Spies’ group process. By contrast, Lanier and colleagues found showed that NK cells can kill NKG2D ligand- that mice that had rejected Rae1γ-transfected expressing cells in vitro. The mouse ligands for RMA cells were unable to reject parental tumours NKG2D are retinoic acid early inducible-1 (Rae-1) on re-challenge. and H60, which are expressed by some tumour These results show that NK cells can, and do, cells, but not by normal adult cells. participate in rejection of MHC class-I bearing Both groups looked for direct evidence to tumour cells and, although there are some support the idea that tumour cells ectopically discrepancies in the results, this approach might expressing ligands for NKG2D could stimulate be effective for tumour vaccine development. antitumour responses by NK cells. The Raulet Elaine Bell group used a retroviral expression system to References and links β ORIGINAL RESEARCH PAPERS Diefenbach, A., Jensen, E. R., express Rae1 and H60 in three mouse tumour Jamieson, A. M. & Raulet, D. H. Rae1 and H60 ligands of the NKG2D cell lines that express MHC class I molecules — receptor stimulate tumour immunity. Nature 413, 165–171 (2001) | EL4 thymoma cells, RMA T-cell cells Cerwenka, A., Baron, J. L. & Lanier L. L. Ectopic expression of retinoic acid early inducible-1 gene (RAE-1) permits natural killer cell-mediated (which were used in the original Kärre study) and rejection of a MHC class I-bearing tumor in vivo. Proc. Natl Acad. Sci. B16-BL6 melanoma cells. Transduced EL4 and USA 98, 11521–11526 (2001) B16-BL6 cells that were injected subcutaneously WEB SITES Lewis Lanier’s lab: http://cc.ucsf.edu/people/lanier_lewis.html into recipient mice were rapidly and completely David Raulet’s lab: http://mcb.berkeley.edu/labs/raulet/

NATURE REVIEWS | IMMUNOLOGY VOLUME 1 | NOVEMBER 2001 | 91 © 2001 Macmillan Magazines Ltd