A Lesson in Tolerance — Maternal Instruction to Fetal Cells William J

The new england journal of medicine

clinical implications of basic research

A Lesson in Tolerance — Maternal Instruction to Fetal Cells William J. Burlingham, Ph.D.

Most people would agree that mothers are influ- fetal thymus and head toward peripheral lymph ential in shaping their children’s outlook on life. nodes (Fig. 1). The newly arising T cells encoun- A recent study by Mold and colleagues1 discloses ter maternal cells that are most likely to be the a new facet of maternal influence: maternal im- progeny of pluripotent stem cells that have crossed mune cells “teach” those of the fetus how to bal- the placental barrier and managed to avoid elimi- ance the need for self-defense on the one hand nation by fetal natural killer cells. Since a high and the requirement for immunologic tolerance proportion of cells in a mature T-cell repertoire on the other. Balance is critical. Too much re- are capable of recognizing foreign HLA antigens, straint of immunity could lead to a lethal infec- many T cells will detect the maternal cells as tion in the newborn; too little could lead to auto- well as soluble HLA antigen encoded by nonin- immunity. herited HLA alleles and undergo activation. Acti- The study by Mold et al. offers a rare glimpse vated T cells make interleukin-2 and express the into the development of the human adaptive im- high-affinity interleukin-2 receptor, and so they mune system, and it suggests that the balance is will proliferate and differentiate into effector markedly shifted, in utero, toward tolerance by T cells. The latter could easily wipe out cells bear- means of the preferential induction of lympho- ing unfamiliar HLA and non-HLA antigens, as cytes known as regulatory T cells. The investi- happens routinely in adults who have received gators compared leukocytes from the lymphoid transfusions or transplants. tissues of fetuses of 20 weeks’ gestation with However, the authors found that the fetal leukocytes from umbilical-cord blood, the thy- lymph node did not cultivate strong immune re- muses of infants, and the lymph nodes and pe- sponses to antigens; rather, it behaved more like ripheral blood of adults. They observed no bias an immune-privileged site,2 where the regulation in regulatory T cells in the fetal thymus. Just as of immunity allowing maternal T cells to sur- in the thymuses of infants and adults, only a vive, rather than antigen-specific sensitization, small number of CD4 T cells (5 to 10%) were was favored. The fetal lymph node produced an regulatory, as indicated by expression of the abundance of transforming growth factor β regulatory T-cell marker, the forkhead box P3 (TGF-β). TGF-β, in combination with antigens (FOXP3) protein (a nuclear transcription factor). and interleukin-2, stimulates FOXP3 expression In contrast, they observed a bias toward regula- and induction of regulatory T cells that are spe- tory T cells that was as high as 20 to 25% in the cific to maternal antigens. These suppressive cells, fetal lymph nodes and spleen, where mature which might recognize a noninherited maternal T cells go to encounter antigens. This bias dis- peptide presented by the inherited HLA class II appeared at birth and was not observed in adult molecule, blocked the proliferation of effector lymph nodes. Maternally derived cells benefited T cells and spared the remaining maternal cells from this transient immune deviation; this find- from destruction. The result was maternal micro- ing has potentially far-reaching consequences for chimerism (up to 0.8% of maternal cells) in the human health and disease. fetus that was about 10 times greater than that Whence comes this remarkable pro-tolerance previously reported in adults. bias in the adaptive immune system of the fetus? Shortly after birth, the lymph node is no long Fully competent T cells emerge from the human er such a hospitable environment for maternal

n engl j med 360;13 nejm.org march 26, 2009 1355

Mature cell

Placenta Fetal cell Noninherited Inherited HLA antigen maternal HLA antigen

T cell Fetal thymus Pluripotent stem cell from mother Activation

Mature maternal cells eliminated

Proliferation Interleukin-2 and differentiation TGF-β

T cell

Effector T cell Fetal lymph Maternal node peptide

FOXP3 FOXP3

Maternal cells persist Regulatory T cell

Figure 1. Immunity and the Maternal−Fetal Axis. Mold and colleagues1 recently described a mechanism through which fetal tolerance of maternal antigens is initiated. NoninheritedCOLOR FIGURE HLA antigens (orange) and inherited maternal HLA antigens (blue) are shown. Noninherited maternal HLA antigens includeDraft 8 intact HLA3/10/09 recognized by fetal effector T cells and maternal peptides recognized by the FOXP3-positive regulatory T cells. TheAuthor differentBurlingham results of effec- Fig # 1 tor T cells versus regulatory T-cell pathways with respect to the persistence of mature maternal cells (microchimerism)Title Theare pathway indicated. to tolerance ME DE Phimister Artist Knoper AUTHOR PLEASE NOTE: Figure has been redrawn and type has been reset Please check carefully Issue date 3/26/09

1356 n engl j med 360;13 nejm.org march 26, 2009

Downloaded from www.nejm.org by ROBERT M. FINE MD on March 29, 2009 . Copyright © 2009 Massachusetts Medical Society. All rights reserved. clinical implications of basic research cells. Nonetheless, Mold et al. found evidence of mentally induced regulatory T cells; this idea is a lingering regulatory T-cell effect in some, but strongly supported by a recent study in mice.5 Ma- not all, healthy young adult offspring in a given ternally induced regulatory T cells may also part- family. Some FOXP3-positive regulatory T cells ly explain the reduced incidence of the formation induced in the fetus may survive as long-lived of potentially harmful antibodies in patients who memory T cells in the adult, as the authors sug- have been exposed to noninherited maternal anti- gest. Alternatively, the generation of new regu- gens through pregnancy, blood transfusions, or latory T cells may require additional antigen input transplantation; however, independent effects of from breast-feeding, ongoing maternal microchi- maternal microchimerism on B cells may also merism, or both to sustain tolerance. have a role. The expansion of antigen-specific regulatory There is still much more to be learned about T cells, both ex vivo and in vivo, is being devel- tolerogenic versus immunogenic forms of micro- oped as a clinical strategy to promote tolerance chimerism, both of which have been reported, and and thus to reduce or eliminate entirely the bur- the possibility that autoimmunity may result if den of immunosuppressive drugs on recipients of maternal antigen–specific, TGF-β–producing reg- organ transplants. The findings of Mold et al. sug- ulatory T cells become Th-17 effectors in the pres- gest that antigen-specific regulatory T cells and ence of interleukin-6. For the time being, how- microchimerism are probably complementary. ever, what a marvel are mothers and the pathway Regulatory T-cell expansion could be combined toward tolerance they map out for us before we with stem-cell transplantation to maximize the are born. induction and maintenance of tolerance. It will be No potential conflict of interest relevant to this article was re- important to determine precisely what type of ma- ported. ternal stem cell generates microchimerism and From the Department of Surgery, Division of Transplantation, tolerance in the fetus. Currently, strategies to min- University of Wisconsin Medical School, Madison. imize the side effects of immunosuppressive drugs in renal transplantation (such as corticosteroid or 1. Mold JE, Michaëlsson J, Burt TD, et al. Maternal alloantigens promote the development of tolerogenic fetal regulatory T cells cyclosporine withdrawal) are applied indiscrimi- in utero. Science 2008;322:1562-5. nately, with a significant risk of acute or chronic 2. Streilein JW. Ocular immune privilege: therapeutic opportu- rejection of the graft. Perhaps patients, particularly nities from an experiment of nature. Nat Rev Immunol 2003;3: 879-89. those with living related donors, could be evalu- 3. Burlingham WJ, Grailer AP, Heisey DM, et al. The effect of ated before transplantation to identify levels of tolerance to noninherited maternal HLA antigens on the survival microchimerism and associated regulatory T-cell of renal transplants from sibling donors. N Engl J Med 1998; 339:1657-64. activity, before they are enrolled in such a trial. 4. van Rood JJ, Loberiza FR Jr, Zhang MJ, et al. Effect of toler- Choosing donors with at least one HLA-DR and ance to noninherited maternal antigens on the occurrence of HLA-DQ match and appropriate maternal-antigen graft-versus-host disease after bone marrow transplantation from a parent or an HLA-haploidentical sibling. Blood 2002;99: mismatches could also be of value. 1572-7. The high success rate of clinical living-related 5. Molitor-Dart ML, Andrassy J, Kwun J, et al. Developmental organ and hematopoietic stem-cell transplanta- exposure to noninherited maternal antigens induces CD4+ T regulatory cells: relevance to mechanism of heart allograft tions when the donor or recipient expresses anti- tolerance. J Immunol 2007;179:6749-61. 3,4 gens that were not inherited from the mother Copyright © 2009 Massachusetts Medical Society. can now be explained by the influence of develop-

n engl j med 360;13 nejm.org march 26, 2009 1357