OX40 Ligand and CD30 Ligand Are Expressed on Adult but Not Neonatal CD4+CD3– Inducer Cells: Evidence That IL-7 Signals Regulate CD30 Ligand
http://www.100md.com
免疫学杂志 2005年第11期
Abstract
In this report, we have examined the expression of the T cell survival signals, OX40 ligand (OX40L) and CD30 ligand (CD30L) on CD4+CD3–CD11c–B220–IL-7R+ inducer cells from birth to adulthood in mice. We found that adult but not neonatal inducer cells expressed high levels of OX40L and CD30L, whereas their expression of TNF-related activation-induced cytokine (TRANCE) and receptor activator of NF-B (RANK) was comparable. The failure of neonatal inducer cells to express the ligands that rescue T cells helps to explain why exposure to Ag in neonatal life induces tolerance rather than immunity. The expression of OX40L and CD30L on inducer cells increased gradually in the first few weeks of life achieving essentially normal levels around the time mice were weaned. We found that IL-7 signaling through the common cytokine receptor -chain was critical for the optimal expression of both TNF-related activation-induced cytokine and CD30L but not OX40L. Furthermore, glucocorticoids, which potently suppress T effector function, did not influence the expression of OX40L and CD30L in the presence of IL-7.
Introduction
The capacity of animals to mount high affinity long-lived Ab responses is associated with the development of segregated B and T cell areas in secondary lymphoid tissue and also the development of lymph nodes. Recent work has identified a key role for a CD4+CD3–CD11c–B220–IL-7R+ cell (inducer cell) in this process (1). These cells are found in neonatal lymph nodes before they are colonized by lymphocytes (2). They are one of only a few cell types that express lymphotoxin (LT)3 12 (LT12) (2), which is the ligand for the LT receptor that plays a critical role in organizing secondary lymphoid tissues (3, 4). Mice deficient in this signaling pathway lack lymph nodes, and the spleen is not organized into B and T cell areas. Evidence that LT receptor signals from inducer cells initiate a sequence of events that leads to the development of lymph nodes comes from mice deficient in these cells. Inducer cells express IL-7 receptor chain (IL-7R), common cytokine receptor -chain (c chain, CD132), ROR (retinoic acid receptor-related orphan nuclear hormone receptor), CXCR5, TNF-related activation-induced cytokine (TRANCE), and receptor activator of NF-B (RANK). Mice deficient in ROR (5) lack lymph nodes and Peyer’s patches; mice deficient in TRANCE (6) and RANK (7) signals lack most lymph nodes; mice deficient in IL-7 (8) and c chain signals (9) lack Peyer’s patch anlagen and some lymph nodes; and CXCR5-deficient mice have impaired Peyer’s patch formation and are deficient in some lymph nodes (10).
In adult mice, we identified the adult equivalent of inducer cells, and reported that they were located in B follicles and at the interface between the B and T cell areas (11). These cells had a very similar phenotype to those described in the neonate (11), but in addition to the expression of LT12 and TRANCE, these cells were also found to express high levels of the TNF ligands for OX40 and CD30. We found evidence for direct interactions between primed CD4 T cells and these cells during the course of immune responses, and showed that the survival of primed T cells was partially dependent on OX40 signals from these cells. Because OX40 and CD30 are genetically linked in human (chromosome 1) and mouse (chromosome 4) (12) and share common signaling pathways, it seemed likely that there were redundant signaling pathways (13). Recently, we have generated and examined the phenotype of mice deficient in both OX40 and CD30 and shown that they lack T cell memory for Ab production, and this failure results from deficient OX40 and CD30 signals delivered by inducer cells (14).
In this study, we have investigated the expression of OX40 ligand (OX40L) and CD30 ligand (CD30L) during the development of inducer cells from the neonatal period to adulthood. We found that inducer cells isolated from neonatal mice did not express OX40L and CD30L, whereas their expression of TRANCE and RANK was similar to adult mice. This suggested that there were specific signals that induced expression of OX40L and CD30L on inducer cells. We found that mice deficient in either IL-7 signals or c chain had substantially reduced CD30L expression, whereas OX40L expression was comparable. Furthermore, addition of IL-7 to neonatal inducer cells specifically up-regulated CD30L expression. The IL-7-induced CD30L expression was also observed on memory phenotype CD4 and CD8 T cells as well as naive CD8 T cells, suggesting that there is a common IL-7-dependent pathway for CD30L expression. We failed to identify specific signals that up-regulated OX40L expression on neonatal inducer cells. The expression of OX40L and CD30L was not dependent on MHC class I or II signals or T cells. Furthermore, we show that although glucocorticoids down-regulated CD30L and TRANCE expression on inducer cells, OX40L expression was only modestly affected. However, in the presence of IL-7, CD30L and TRANCE expression was restored on glucocorticoid-treated cells.
Discussion
Here we report that inducer cells isolated from neonatal mice lack expression of the T cell survival molecules, OX40L and CD30L, although expression of TRANCE and RANK was normal. We have found that mice deficient in both OX40 and CD30 show grossly impaired CD4 T cell memory responses because they fail to receive survival signals from inducer cells (14). The data provided here on neonatal inducer cells suggest that T cells primed in the neonate would be rendered effectively deficient in OX40 and CD30 survival signals from inducer cells, so they would not become memory cells. This observation may help explain the phenomenon of neonatal tolerance first described 50 years ago by Medawar et al. (20). Although neonatal mice did not express OX40L and CD30L on inducer cells, by 2 wk of age both molecules were expressed, and almost normal levels were achieved by 3 wk, the time mice were weaned. These data suggest that there must be distinct signaling pathway/s that regulate expression of these molecules on inducer cells after birth.
We demonstrated that IL-7 signals to inducer cells up-regulate TRANCE and CD30L expression. On adult inducer cells, CD30L expression was augmented, and on neonatal inducer cells, CD30L expression was induced by IL-7. Furthermore, mice deficient in IL-7, IL-7R, and c chain had reduced CD30L expression. In contrast to CD30L, OX40L expression was not dependent on IL-7 or c chain, and we could identify no other cytokine signals (IL-3, IL-4, IL-9, IL-10, IL-12, or IL-15) that would induce its expression on neonatal inducer cells (data not shown).
IL-7 mediated up-regulation of CD30L expression was not exclusive to inducer cells because memory CD4 and CD8 T cells also up-regulated CD30L expression in the presence of IL-7, suggesting that there is a common signaling pathway in T cells and inducer cells. We have observed previously that IL-4, which also signals via c chain-like IL-7, does not affect the expression of either OX40L or CD30L on inducer cells (21). Its effects on CD4 T cells are also different: IL-4 down-regulated OX40L and CD30L expression on primed CD4 T cells, whereas IL-7 induced CD30L but not OX40L expression. This indicates that although the cytokines, IL-4 and IL-7, both share signaling pathways through c chain, there must be distinct signaling pathways that are c chain independent, particularly for IL-4.
Because inducer cells express CD4, we tested whether MHC class I or II molecules might be required for their development and the expression of TNF ligands. No differences in number or phenotype were observed. Similarly the expression of these molecules was normal in mice deficient for OX40 and CD30 expression suggesting that reverse signaling through OX40L and CD30L was unimportant for inducer cell development.
Finally we investigated whether glucocorticoids, which are potent suppressors of T cell-mediated effector responses, modified the expression of TNF ligands on inducer cells. After 2 days of culture with cortisol or the more potent synthetic glucocorticoid, dexamethasone, inducer cells down-regulated both CD30L and TRANCE expression compared with control. After 6 days of culture, there were significantly fewer inducer cells in the dexamethasone-treated cultures, presumably due to down-regulation of TRANCE-dependent survival signals for inducer cells, but OX40L expression was spared. The decreased survival in the presence of glucocorticoids was reversed if inducer cells were cultured in the presence of IL-7. The failure of glucocorticoids to attenuate the expression of the T cell survival signal, OX40L explain why glucocorticoids, which are effective at suppressing effector T cell responses (22, 23), fail to eliminate CD4 T cell memory (24).
In this study, we show that OX40L and CD30L expression on adult inducer cells, which we have linked with CD4 T cell survival and memory for Ab responses, is not expressed in neonatal life on inducer cells. We provide evidence that the expression of TRANCE and CD30L on inducer cells is regulated by IL-7, c chain, and glucocorticoid signals. In contrast, although OX40L was not expressed on neonatal inducer cells, regulation of its expression was not affected by the above signals. We suggest that the deficient expression of OX40L and CD30L on neonatal inducer cells could be a contributory mechanism for neonatal tolerance induction.
We do not know whether inducer cells isolated from adults that express OX40L and CD30L are the direct descendants of neonatal inducer cells, although labeling studies with thymidine analogues suggest that they turnover slowly (25). We have provided evidence that CD30L expression can be induced by IL-7, but we do not know which cell provides this signal in secondary lymphoid tissue. An intriguing possibility is that it comes from follicular dendritic cells (FDCs) (26) in B follicles close to where inducer cells are found. FDCs depend on B cell LT12 for their survival (27), and we have found that inducer cells isolated from neonatal mice with B cells (CD3Tg26 mice) did have higher levels of CD30L (data not shown) than when isolated from RAG-deficient mice. However, this cannot be the sole explanation, because adult inducer cells isolated from RAG-deficient mice that lack FDCs have high levels of CD30L. In any case, we have not identified the signals that regulate OX40L expression.
Inducer cells have been reported in the blood (28), and we found that in the adult they express L-selectin (11), so precursors could potentially migrate from the blood into secondary lymphoid organs. It could be that the neonatal population that we isolate is simply replaced by an adult population that expresses OX40L and CD30L. Nevertheless, understanding the molecular mechanism that regulates this transition may provide new approaches to modifying pathogenic T cell responses that are refractory to conventional immune suppression including glucocorticoid therapy.
Acknowledgments
We thank Rose Zamoyska and Benedict Seddon for providing spleens from the IL-7R-deficient mice and common -chain-deficient RAG–/– mice.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by a Wellcome Programme Grant (to P.J.L.L.).
2 Address correspondence and reprint requests to Dr. Peter J. L. Lane, Medical Research Council Centre for Immune Regulation, Birmingham Medical School, Vincent Drive, Birmingham B15 2TT, UK. E-mail address: p.j.l.lane{at}bham.ac.uk
3 Abbreviations used in this paper: LT, lymphotoxin; c chain, common cytokine receptor -chain; TRANCE, TNF-related activation-induced cytokine; RANK, receptor activator of NF-B; OX40L, OX40 ligand; CD30L, CD30 ligand; Tg, transgenic; FDC, follicular dendritic cell.
Received for publication December 8, 2004. Accepted for publication March 15, 2005.
References
Mebius, R. E.. 2003. Organogenesis of lymphoid tissues. [Published Erratum appears in 2003 Nat. Rev. Immunol. 3: 509.]. Nat. Rev. Immunol. 3: 292-303.
Mebius, R. E., P. Rennert, I. L. Weissman. 1997. Developing lymph nodes collect CD4+CD3– LT+ cells that can differentiate to APC, NK cells, and follicular cells but not T or B cells. Immunity 7: 493-504.
Futterer, A., K. Mink, A. Luz, M. H. Kosco-Vilbois, K. Pfeffer. 1998. The lymphotoxin receptor controls organogenesis and affinity maturation in peripheral lymphoid tissues. Immunity 9: 59-70.
Fu, Y. X., D. D. Chaplin. 1999. Development and maturation of secondary lymphoid tissues. Annu. Rev. Immunol. 17: 399-433.
Sun, Z., D. Unutmaz, Y. R. Zou, M. J. Sunshine, A. Pierani, S. Brenner-Morton, R. E. Mebius, D. R. Littman. 2000. Requirement for ROR in thymocyte survival and lymphoid organ development. Science 288: 2369-2373.
Kim, D., R. E. Mebius, J. D. MacMicking, S. Jung, T. Cupedo, Y. Castellanos, J. Rho, B. R. Wong, R. Josien, N. Kim, et al 2000. Regulation of peripheral lymph node genesis by the tumor necrosis
Dougall, W. C., M. Glaccum, K. Charrier, K. Rohrbach, K. Brasel, T. De Smedt, E. Daro, J. Smith, M. E. Tometsko, C. R. Maliszewski, et al 1999. RANK is essential for osteoclast and lymph node development. Genes Dev. 13: 2412-2424.
Adachi, S., H. Yoshida, K. Honda, K. Maki, K. Saijo, K. Ikuta, T. Saito, S. I. Nishikawa. 1998.(Mi-Yeon Kim*, Graham Ande)
In this report, we have examined the expression of the T cell survival signals, OX40 ligand (OX40L) and CD30 ligand (CD30L) on CD4+CD3–CD11c–B220–IL-7R+ inducer cells from birth to adulthood in mice. We found that adult but not neonatal inducer cells expressed high levels of OX40L and CD30L, whereas their expression of TNF-related activation-induced cytokine (TRANCE) and receptor activator of NF-B (RANK) was comparable. The failure of neonatal inducer cells to express the ligands that rescue T cells helps to explain why exposure to Ag in neonatal life induces tolerance rather than immunity. The expression of OX40L and CD30L on inducer cells increased gradually in the first few weeks of life achieving essentially normal levels around the time mice were weaned. We found that IL-7 signaling through the common cytokine receptor -chain was critical for the optimal expression of both TNF-related activation-induced cytokine and CD30L but not OX40L. Furthermore, glucocorticoids, which potently suppress T effector function, did not influence the expression of OX40L and CD30L in the presence of IL-7.
Introduction
The capacity of animals to mount high affinity long-lived Ab responses is associated with the development of segregated B and T cell areas in secondary lymphoid tissue and also the development of lymph nodes. Recent work has identified a key role for a CD4+CD3–CD11c–B220–IL-7R+ cell (inducer cell) in this process (1). These cells are found in neonatal lymph nodes before they are colonized by lymphocytes (2). They are one of only a few cell types that express lymphotoxin (LT)3 12 (LT12) (2), which is the ligand for the LT receptor that plays a critical role in organizing secondary lymphoid tissues (3, 4). Mice deficient in this signaling pathway lack lymph nodes, and the spleen is not organized into B and T cell areas. Evidence that LT receptor signals from inducer cells initiate a sequence of events that leads to the development of lymph nodes comes from mice deficient in these cells. Inducer cells express IL-7 receptor chain (IL-7R), common cytokine receptor -chain (c chain, CD132), ROR (retinoic acid receptor-related orphan nuclear hormone receptor), CXCR5, TNF-related activation-induced cytokine (TRANCE), and receptor activator of NF-B (RANK). Mice deficient in ROR (5) lack lymph nodes and Peyer’s patches; mice deficient in TRANCE (6) and RANK (7) signals lack most lymph nodes; mice deficient in IL-7 (8) and c chain signals (9) lack Peyer’s patch anlagen and some lymph nodes; and CXCR5-deficient mice have impaired Peyer’s patch formation and are deficient in some lymph nodes (10).
In adult mice, we identified the adult equivalent of inducer cells, and reported that they were located in B follicles and at the interface between the B and T cell areas (11). These cells had a very similar phenotype to those described in the neonate (11), but in addition to the expression of LT12 and TRANCE, these cells were also found to express high levels of the TNF ligands for OX40 and CD30. We found evidence for direct interactions between primed CD4 T cells and these cells during the course of immune responses, and showed that the survival of primed T cells was partially dependent on OX40 signals from these cells. Because OX40 and CD30 are genetically linked in human (chromosome 1) and mouse (chromosome 4) (12) and share common signaling pathways, it seemed likely that there were redundant signaling pathways (13). Recently, we have generated and examined the phenotype of mice deficient in both OX40 and CD30 and shown that they lack T cell memory for Ab production, and this failure results from deficient OX40 and CD30 signals delivered by inducer cells (14).
In this study, we have investigated the expression of OX40 ligand (OX40L) and CD30 ligand (CD30L) during the development of inducer cells from the neonatal period to adulthood. We found that inducer cells isolated from neonatal mice did not express OX40L and CD30L, whereas their expression of TRANCE and RANK was similar to adult mice. This suggested that there were specific signals that induced expression of OX40L and CD30L on inducer cells. We found that mice deficient in either IL-7 signals or c chain had substantially reduced CD30L expression, whereas OX40L expression was comparable. Furthermore, addition of IL-7 to neonatal inducer cells specifically up-regulated CD30L expression. The IL-7-induced CD30L expression was also observed on memory phenotype CD4 and CD8 T cells as well as naive CD8 T cells, suggesting that there is a common IL-7-dependent pathway for CD30L expression. We failed to identify specific signals that up-regulated OX40L expression on neonatal inducer cells. The expression of OX40L and CD30L was not dependent on MHC class I or II signals or T cells. Furthermore, we show that although glucocorticoids down-regulated CD30L and TRANCE expression on inducer cells, OX40L expression was only modestly affected. However, in the presence of IL-7, CD30L and TRANCE expression was restored on glucocorticoid-treated cells.
Discussion
Here we report that inducer cells isolated from neonatal mice lack expression of the T cell survival molecules, OX40L and CD30L, although expression of TRANCE and RANK was normal. We have found that mice deficient in both OX40 and CD30 show grossly impaired CD4 T cell memory responses because they fail to receive survival signals from inducer cells (14). The data provided here on neonatal inducer cells suggest that T cells primed in the neonate would be rendered effectively deficient in OX40 and CD30 survival signals from inducer cells, so they would not become memory cells. This observation may help explain the phenomenon of neonatal tolerance first described 50 years ago by Medawar et al. (20). Although neonatal mice did not express OX40L and CD30L on inducer cells, by 2 wk of age both molecules were expressed, and almost normal levels were achieved by 3 wk, the time mice were weaned. These data suggest that there must be distinct signaling pathway/s that regulate expression of these molecules on inducer cells after birth.
We demonstrated that IL-7 signals to inducer cells up-regulate TRANCE and CD30L expression. On adult inducer cells, CD30L expression was augmented, and on neonatal inducer cells, CD30L expression was induced by IL-7. Furthermore, mice deficient in IL-7, IL-7R, and c chain had reduced CD30L expression. In contrast to CD30L, OX40L expression was not dependent on IL-7 or c chain, and we could identify no other cytokine signals (IL-3, IL-4, IL-9, IL-10, IL-12, or IL-15) that would induce its expression on neonatal inducer cells (data not shown).
IL-7 mediated up-regulation of CD30L expression was not exclusive to inducer cells because memory CD4 and CD8 T cells also up-regulated CD30L expression in the presence of IL-7, suggesting that there is a common signaling pathway in T cells and inducer cells. We have observed previously that IL-4, which also signals via c chain-like IL-7, does not affect the expression of either OX40L or CD30L on inducer cells (21). Its effects on CD4 T cells are also different: IL-4 down-regulated OX40L and CD30L expression on primed CD4 T cells, whereas IL-7 induced CD30L but not OX40L expression. This indicates that although the cytokines, IL-4 and IL-7, both share signaling pathways through c chain, there must be distinct signaling pathways that are c chain independent, particularly for IL-4.
Because inducer cells express CD4, we tested whether MHC class I or II molecules might be required for their development and the expression of TNF ligands. No differences in number or phenotype were observed. Similarly the expression of these molecules was normal in mice deficient for OX40 and CD30 expression suggesting that reverse signaling through OX40L and CD30L was unimportant for inducer cell development.
Finally we investigated whether glucocorticoids, which are potent suppressors of T cell-mediated effector responses, modified the expression of TNF ligands on inducer cells. After 2 days of culture with cortisol or the more potent synthetic glucocorticoid, dexamethasone, inducer cells down-regulated both CD30L and TRANCE expression compared with control. After 6 days of culture, there were significantly fewer inducer cells in the dexamethasone-treated cultures, presumably due to down-regulation of TRANCE-dependent survival signals for inducer cells, but OX40L expression was spared. The decreased survival in the presence of glucocorticoids was reversed if inducer cells were cultured in the presence of IL-7. The failure of glucocorticoids to attenuate the expression of the T cell survival signal, OX40L explain why glucocorticoids, which are effective at suppressing effector T cell responses (22, 23), fail to eliminate CD4 T cell memory (24).
In this study, we show that OX40L and CD30L expression on adult inducer cells, which we have linked with CD4 T cell survival and memory for Ab responses, is not expressed in neonatal life on inducer cells. We provide evidence that the expression of TRANCE and CD30L on inducer cells is regulated by IL-7, c chain, and glucocorticoid signals. In contrast, although OX40L was not expressed on neonatal inducer cells, regulation of its expression was not affected by the above signals. We suggest that the deficient expression of OX40L and CD30L on neonatal inducer cells could be a contributory mechanism for neonatal tolerance induction.
We do not know whether inducer cells isolated from adults that express OX40L and CD30L are the direct descendants of neonatal inducer cells, although labeling studies with thymidine analogues suggest that they turnover slowly (25). We have provided evidence that CD30L expression can be induced by IL-7, but we do not know which cell provides this signal in secondary lymphoid tissue. An intriguing possibility is that it comes from follicular dendritic cells (FDCs) (26) in B follicles close to where inducer cells are found. FDCs depend on B cell LT12 for their survival (27), and we have found that inducer cells isolated from neonatal mice with B cells (CD3Tg26 mice) did have higher levels of CD30L (data not shown) than when isolated from RAG-deficient mice. However, this cannot be the sole explanation, because adult inducer cells isolated from RAG-deficient mice that lack FDCs have high levels of CD30L. In any case, we have not identified the signals that regulate OX40L expression.
Inducer cells have been reported in the blood (28), and we found that in the adult they express L-selectin (11), so precursors could potentially migrate from the blood into secondary lymphoid organs. It could be that the neonatal population that we isolate is simply replaced by an adult population that expresses OX40L and CD30L. Nevertheless, understanding the molecular mechanism that regulates this transition may provide new approaches to modifying pathogenic T cell responses that are refractory to conventional immune suppression including glucocorticoid therapy.
Acknowledgments
We thank Rose Zamoyska and Benedict Seddon for providing spleens from the IL-7R-deficient mice and common -chain-deficient RAG–/– mice.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by a Wellcome Programme Grant (to P.J.L.L.).
2 Address correspondence and reprint requests to Dr. Peter J. L. Lane, Medical Research Council Centre for Immune Regulation, Birmingham Medical School, Vincent Drive, Birmingham B15 2TT, UK. E-mail address: p.j.l.lane{at}bham.ac.uk
3 Abbreviations used in this paper: LT, lymphotoxin; c chain, common cytokine receptor -chain; TRANCE, TNF-related activation-induced cytokine; RANK, receptor activator of NF-B; OX40L, OX40 ligand; CD30L, CD30 ligand; Tg, transgenic; FDC, follicular dendritic cell.
Received for publication December 8, 2004. Accepted for publication March 15, 2005.
References
Mebius, R. E.. 2003. Organogenesis of lymphoid tissues. [Published Erratum appears in 2003 Nat. Rev. Immunol. 3: 509.]. Nat. Rev. Immunol. 3: 292-303.
Mebius, R. E., P. Rennert, I. L. Weissman. 1997. Developing lymph nodes collect CD4+CD3– LT+ cells that can differentiate to APC, NK cells, and follicular cells but not T or B cells. Immunity 7: 493-504.
Futterer, A., K. Mink, A. Luz, M. H. Kosco-Vilbois, K. Pfeffer. 1998. The lymphotoxin receptor controls organogenesis and affinity maturation in peripheral lymphoid tissues. Immunity 9: 59-70.
Fu, Y. X., D. D. Chaplin. 1999. Development and maturation of secondary lymphoid tissues. Annu. Rev. Immunol. 17: 399-433.
Sun, Z., D. Unutmaz, Y. R. Zou, M. J. Sunshine, A. Pierani, S. Brenner-Morton, R. E. Mebius, D. R. Littman. 2000. Requirement for ROR in thymocyte survival and lymphoid organ development. Science 288: 2369-2373.
Kim, D., R. E. Mebius, J. D. MacMicking, S. Jung, T. Cupedo, Y. Castellanos, J. Rho, B. R. Wong, R. Josien, N. Kim, et al 2000. Regulation of peripheral lymph node genesis by the tumor necrosis
Dougall, W. C., M. Glaccum, K. Charrier, K. Rohrbach, K. Brasel, T. De Smedt, E. Daro, J. Smith, M. E. Tometsko, C. R. Maliszewski, et al 1999. RANK is essential for osteoclast and lymph node development. Genes Dev. 13: 2412-2424.
Adachi, S., H. Yoshida, K. Honda, K. Maki, K. Saijo, K. Ikuta, T. Saito, S. I. Nishikawa. 1998.(Mi-Yeon Kim*, Graham Ande)