Chimera: from bane to blessing

New data suggest that a mild preparative regimen of antibodies that block CD40 ligand and deplete host NK cells may make allogeneic hematopoietic stem cell transplants safe, establish long-term immunologic tolerance, and broaden the ap-plicability of cord blood as a source of stem cells by making engraftment more efﬁ-cient. deplete cytotoxic T cells. On day 28, a novel in vivo cytotoxicity assay was performed in which 10 million BALB/c donor splenocytes labeled with carboxyﬂuorescein succinimidyl ester (CFSE) were administered to the mice intravenously, and the elimination of the cells by the host was followed by ﬂow cytometry on peripheral blood samples 2 days later. The results demonstrate that NK cells mediate elimination of 94% of the donor cells after after anti-CD40L

New data suggest that a mild preparative regimen of antibodies that block CD40 ligand and deplete host NK cells may make allogeneic hematopoietic stem cell transplants safe, establish long-term immunologic tolerance, and broaden the applicability of cord blood as a source of stem cells by making engraftment more efficient. I n Greek mythology, "chimera," derived from the Greek word for "billy goat," referred to a fire-breathing she-demon that was part lion, part goat, and part dragon or snake. Getting rid of this killer was quite an achievement, and Bellerophon was heaped with praise and riches for his courageous and clever dispatch of the beast. The term "chimera" has come down to us through the ages with 3 definitions: the original monster, an impossible and fanciful creation of the imagination (eg, Woody Allen's malefactor with the body of a crab and the head of social worker), and an organism containing tissues from at least 2 genetically distinct parents. It is this last definition that is of compelling medical interest, based largely on the work of Ray Owen.
In 1945, Dr Owen was the first to demonstrate immunologic chimerism, when he found that the majority of dizygotic bovine twins had identical blood types. 1 This chimerism was thought to result from blood (and, by inference, hematopoietic stem cell) mixing through placental vascular anastomoses. The implied immunologic tolerance associated with the condition was formally documented in 1952 by Billingham et al, 2 who showed that dizygotic chimeric twin cattle were tolerant to skin grafts from each other but rapidly rejected third-party grafts.
Chimerism and its associated immunologic tolerance are easy to induce in fetuses and neonates but difficult to induce in adults with their own immunologic integrity. Many barriers exist to engraftment of donor lymphoid and hematopoietic tissues in adult hosts, but most of these barriers can be overcome by eliminating or greatly suppressing host immune defenses. Clinically, this elimination is usually accomplished by a cytotoxic preparative regimen that usually involves drugs but occasionally also uses radiation therapy. However, these regimens generally do not eliminate natural killer (NK) cells or NK activity. The more recent exploration of nonmyelosuppressive preparative regimens has chiefly sought to block the host T-cell response but has largely ignored NK activity. But since the work of Gustavo Cudkowicz (see, for example, Cudkowicz and Stimpfling 3 ) in the 1960s, we have known that donor marrow can be rejected even in the face of major histocompatibility complex (MHC) compatibility, and subsequent work from many groups has confirmed a role for NK cells in this process.
In this issue of Blood, Westerhuis and colleagues demonstrate that immunologic tolerance across a major histocompatibility barrier associated with immunologic chimerism is greatly facilitated by depleting host NK cells, in this case with anti-NK1.1 monoclonal antibody (see figure). In the experiment shown here, C57BL6 mice were treated with anti-CD40 ligand antibody (anti-CD154) and given 1 million allogeneic BALB/c bone marrow cells on day 0. At days 23 and 27, mice received either phosphate-buffered saline (PBS), anti-NK1.1 to deplete NK cells, or anti-CD8 to deplete cytotoxic T cells. On day 28, a novel in vivo cytotoxicity assay was performed in which 10 million BALB/c donor splenocytes labeled with carboxyfluorescein succinimidyl ester (CFSE) were administered to the mice intravenously, and the elimination of the cells by the host was followed by flow cytometry on peripheral blood samples 2 days later. The results demonstrate that NK cells mediate elimination of 94% of the donor cells after NK cells mediate the elimination of donor cells after anti-CD40L mAb treatment. anti-CD40 ligand antibody is used as a nonmyelosuppressive preparative regimen for allogeneic transplantation. The rejection of donor type cells is largely prevented by depleting NK cells. In other experiments, adding NK cell depletion to the preparative regimen enhanced allogeneic bone marrow donor cells' ability to establish stable chimerism by at least 3-fold. Without NK cell depletion, an inoculum of 30 million BALB/c marrow cells established donor chimerism in only one of 5 C57BL6 mice; with NK cell depletion, the same dose of donor cells was 100% effective at establishing long-term donor chimerism.
This paper is important because it focuses again on the NK cell as an important barrier in establishing chimerism and immunologic tolerance. Early work demonstrated that host NK depletion facilitated engraftment and hematologic recovery in both syngeneic and allogeneic bone marrow transplants. 4 This paper joins others that have supported the idea that NK cells may be important targets for allograft engineering. It would seem that the time has come for a clinical test of the hypothesis that host NK cell depletion can enhance donor marrow cell engraftment. Given the magnitude of the effect of NK depletion reported here, if the result were verified in humans, important applications would include nonmyelosuppressive allogeneic transplantation and cord blood transplantation. In both settings, the capacity to permit complete hematopoietic engraftment with lower doses of donor cells might permit safe allogeneic transplants with a lower incidence of graft-versus-host disease. ■
B and T cells are normally segregated within lymphoid tissue by chemokine gradients that are established by local populations of lymphoid stromal cells. CXCR5 ϩ B cells are attracted to the areas forming B follicles, whereas both CCR7 ϩ T cells and antigen-presenting dendritic cells are attracted to sites that form T zones. This isolation of B cells from T-cell areas is essential for the production of high-affinity antibodies, as it excludes T cells of irrelevant specificities from the B follicle in which B-cell selection by "licensed" T FH occurs.
As others have found, 1 induction of CXCR5 on T cells occurs rapidly following priming and identifies cells destined to pro-vide help in B follicles but not inflammatory responses. Although CXCR5 is expressed on a substantial fraction of primed T cells, many such cells continue to coexpress CCR7, an expression motif that localizes them at the B-T interface where the chemokine gradients establishing the B follicle and T zone compete for influence. Since activated CXCR5expressing B cells mirror the situation in T cells by up-regulating CCR7, 2 they too localize at the B-T interface, an arrangement likely to optimize the efficiency of B-T collaboration not only for primary antibody production but also for memory responses. Only a minority of CXCR5 ϩ CCR7 ϩ -expressing T cells go on to down-regulate CCR7 and become mature T FH where they foster the development of B-cell germinal centers, the structures within which affinity maturation of the B-cell response occurs. It is not certain which signals induce T cells to do this; perhaps the signals arise from some as-yet-unidentified interaction with antigen-activated B cells.
One very interesting observation from Hardtke et al is that dual CXCR5 ϩ CCR7 ϩexpressing T cells appear in lymph nodes where there is no obvious ongoing B-cell response (see figure), suggesting that these T cells might have migrated there following priming elsewhere. This possibility is supported by the fact that these T cells are found in blood. 3 I think that these are recirculating memory cells primed to provide B-cell help and that their dual expression of CXCR5 ϩ CCR7 ϩ guides them to a CD4 ϩ CD3 Ϫ non-dendritic cell located at the B-T interface. 4 My colleagues and I have found that the development and persistence of T memory cells that provide help to B cells depend on OX40 (CD134) and CD30 survival signals from these cells. 5 The location of the T memory cells at the B-T interface puts them in pole position to provide rapid help to B cells at the earliest sign of reinfection, but the theory remains to be tested. ■ For personal use only. on July 22, 2018. by guest www.bloodjournal.org From