- Email: [email protected]
Heavy Water Shedding Light on Antigen-Specific T Cell Responses
TREIMM 1453 No. of Pages 3
Spotlight
Heavy Water Shedding Light on AntigenSpecific T Cell Responses
intervention in mice, as compared to the limitations that are inherent to studies in the human system. Although CD8 T cell turnover has been studied in vivo in both mice and humans, human studies have primarily focused on bulk naïve or memory cells at steady-state or during 1,2 Liv Eidsmo and chronic viral infections [1,2]. Conse1,2, Carmen Gerlach * quently, the in vivo proliferation rates of human antigen-specific CD8 T cells CD8 T cells are crucial for long-term responding to acute antigen exposure immunity. Nevertheless, the in vivo have remained elusive.
differentiation of human naïve CD8 T cells into effector and memory populations remains ill-defined. A recent study assesses the in vivo turnover of human antigen-specific CD8 T cells and suggests that longlived memory cells arise from effector cells. CD8 T cells are key players in long-term protection against previously encountered pathogens. They are widely dispersed throughout the body and can be found in blood, lymph, and in lymphoid and non-lymphoid tissues. The precise role of individual CD8 T cells in immune defense largely depends on the antigen specificity, localization, and differentiation state of the cells.
A recent study by Akondy et al. published in Nature employed a beautiful experimental setup that allowed the authors to address fundamental questions on the turnover kinetics and differentiation of human antigen-specific CD8 T cell responses to controlled antigen exposure in vivo [3]. The authors vaccinated several groups of humans in the USA with the live attenuated yellow fever virus (YFV) vaccine that provides lifelong protection against yellow fever, and tracked the responses of YFV-specific CD8 T cells in these individuals over time using MHC–peptide tetramers. Because YFV is not endemic in the USA, and the vaccinated individuals were naïve to the YFVderived antigens at the time of vaccination, the investigators could track true primary responses to the vaccine. Furthermore, the absence of YFV in the environment of the vaccinees prevented reexposure to vaccine antigens and reactivation of YFV-specific T cells in the timeframe between vaccination and the latest point of analysis, which was up to 13 years after vaccination.
Naïve T cells initiate proliferation and differentiation when they are primed by their specific antigen. This leads to the generation of cytotoxic effector T cells that are actively involved in pathogen clearance. In addition, long-lived so-called memory T cells are formed, which provide protection from renewed encounter with the same pathogen. To determine the proliferation rates of the YFV-specific CD8 T cells, the vaccinees Currently, our most detailed understand- drank ‘heavy water’ (2H2O) during differing of CD8 T cell differentiation stems ent time periods after vaccination. Deutefrom murine in vivo studies. This is pre- rium (2H) incorporates into newly dominantly a result of the wealth of synthesized DNA, and the rate at which molecular and genetic tools that are cells incorporate this isotope provides an available for in vivo experimental estimate of the cell division rate. Likewise,
loss of deuterium labeling in cells after cessation of heavy water intake can be used to infer cell division rates. The authors showed that antigen-specific CD8 T cells proliferated most vigorously during the expansion phase of the T cell response (Figure 1). Around the peak of the T cell response, proliferation had considerably slowed down to roughly half the speed observed during the first 2 weeks. Starting with the contraction phase and throughout memory, proliferation events had become extremely rare, with an estimated doubling time of the memory population of several hundred days. While such estimates should be interpreted with some caution because different methods of cell labeling (e.g., deuterium provided in water versus glucose) and different mathematical models to interpret the data may lead to several-fold different inferred proliferation rates [4], it is remarkable how little proliferation is necessary to maintain the memory population. Both the low turnover rate, and the substantial fraction of the memory population that still contained deuterium 1–2 years after labeling, highlight how extremely long-lived memory CD8 T cells can be. Going forward, it will be interesting to investigate whether there is a link between the extent of division a cell undergoes early after antigen encounter, and the longevity of the progeny of the dividing cells. Akondy et al. found that a large fraction of the long-lived memory population still contained the deuterium label they had acquired during the first 2 weeks after vaccination. Because it takes several rounds of division before a population is maximally labeled with deuterium, one could argue that the deuterium-marked memory cells are derived from cells that have undergone extensive proliferation early during the response. On the other hand, one cell division in the presence of deuterium can be sufficient
Trends in Immunology, Month Year, Vol. xx, No. yy
1
TREIMM 1453 No. of Pages 3
High
Intermediate
Extremely low
n
Expa nsio
on
Number of anƟgen-specific CD8 T cells
racƟ
t Con
studies in mice have revealed that memory cells indeed pass through an effector Prolifera on stage during their development [9]. rate Because T cell differentiation is regulated by both transcriptional and epigenetic processes, Akondy et al. characterized naïve and antigen-specific effector and long-term memory populations by RNA sequencing and epigenetic profiling. While memory T cells appeared transcriptionally more similar to naïve than Memory to effector cells, as seen before, epigenetic profiling showed that memory cells 2 weeks 4 weeks 12 years were more similar to effectors in this Time aŌer vaccinaƟon respect. For example, the loci encoding the typical effector molecules IFN-g, Effector genes Naive Effector Memory granzyme B, and perforin were accessible in long-term memory CD8 T cells, although transcription was restricted to Transcrip on effector cells (Figure 1). Their epigenetic accessibility might be a remainder of the Epigene c effector past of the memory cells, conaccessibility sistent with an effector to memory differentiation path for which the same research groups provide more definite Figure 1. [76_TD$IF]Proliferative, Transcriptional, and Epigenetic Characteristics of Human Antigen-Specific CD8 T Cell Responses. (Upper panel) The purple line represents the number of yellow fever virus (YFV)- evidence in an accompanying paper specific CD8 T cells during the different phases (expansion, contraction, memory) of the response. The blue [10]. At the same time, these epigenetic boxes represent timeframes of different proliferation rates of the YFP-specific CD8 T cells. (Lower panel) features poise the memory cells for effecTranscription was assessed by RNA sequencing. Black strings represent mRNA. Epigenetic accessibility of tor functions upon renewed antigen effector gene loci was assessed by ATAC-seq (assay for transposase-accessible chromatin using sequencing) and methylation analysis of the promoter regions. Purple circles represent methylated CpGs. An extended encounter, consistent with the raïson distance between the blue nucleosomes represents open chromatin. d’être of memory cells.
to incorporate the deuterium, and it is possible that the highly proliferative cells die during contraction. Hence, the longlived memory cells may equally well derive from cells that have undergone only limited proliferation in the past. Independent of their division history, it is still a matter of debate whether memory cells derive from effector cells or directly from naïve cells. In the latter case, one model proposes that effector and memory cells develop in parallel from the same naïve T cells, while another assumes that effector cells are the progeny of memory cells [5]. The fact that naïve T cells are phenotypically and 2
Trends in Immunology, Month Year, Vol. xx, No. yy
transcriptionally more similar to memory than to effector cells [6,7] has supported the notion that memory T cells derive directly from naïve T cells [5,8]. Notably, Akondy et al. show that the extremely long-lived memory population (8–12 years after vaccination) expresses CD45RA and CCR7, which are the classical naïve T cell markers, and this underscores the importance of employing an elaborate panel of markers to accurately delineate functionally distinct CD8 T cell subsets [8]. The alternative model of T cell differentiation proposes that naïve T cells initially differentiate into effectors, of which some become memory cells. Supporting this model, in vivo fate-mapping
The study by Akondy et al. elegantly transfers tools that are commonly used in mice to humans, and we foresee that such ‘technology transfer’ will be expanded in the future. Albeit currently far-fetched, in vivo fate-mapping during the course of an immune response may also become feasible in humans. With gene therapies entering the clinic, ex vivo genetic labeling of cells of interest and subsequent autologous cell transfer may be one possibility to allow studies in humans similar to those performed in mice. A possibly earlier achievable but less definitive alternative may be ex vivo deuterium labeling, followed by reinfusion of an albeit partially labeled cell population. Such studies would enable definitive
TREIMM 1453 No. of Pages 3
conclusions on the origin of human T cell subsets, and at the same time establish whether our mouse model-derived knowledge on T cell differentiation can be translated to humans.
1
Dermatology and Venereology Unit, Department of Medicine Solna, Karolinska Institutet, 171 76 Stockholm, Sweden
4. Borghans, J.A.M. and de Boer, R.J. (2007) Quantification of T-cell dynamics: from telomeres to DNA labeling. Immunol. Rev. 216, 35–47
2
5. Farber, D.L. et al. (2014) Human memory T cells: generation, compartmentalization and homeostasis. Nat. Rev. Immunol. 14, 24–35
Department of Dermatology, Karolinska University Hospital Solna, 171 76 Stockholm, Sweden *Correspondence: [email protected] (C. Gerlach). https://doi.org/10.1016/j.it.2018.01.003
Acknowledgments Research in the laboratory of C.G. is supported by the Karolinska Institute, Swedish Research Council, Ragnar Söderbergs [78_TD$IF]stiftelse, Knut and Alice Wallenberg Foundation, Cancerfonden, and Jeanssons Stiftelser. Research in the laboratory of L.E. is supported by the Karolinska Institute, Swedish Research Council, Swedish City Council (ALF), Ragnar Söderbergs [78_TD$IF]4stiftelse, Marcus and Marianne Wallenberg Founda-
References 1. Hellerstein, M.K. et al. (2003) Subpopulations of long-lived and short-lived T cells in advanced HIV-1 infection. J. Clin. Invest. 112, 956–966 2. Vrisekoop, N. et al. (2015) Quantification of naive and memory T-cell turnover during HIV-1 infection. AIDS 29, 2071–2080 3. Akondy, R.S. et al. (2017) Origin and differentiation of human memory CD8 T cells after vaccination. Nature 552, 362–367
6. Best, J.A. et al. (2013) Transcriptional insights into the CD8+ T cell response to infection and memory T cell formation. Nat. Immunol. 14, 404–412 7. Kakaradov, B. et al. (2017) Early transcriptional and epigenetic regulation of CD8+ T cell differentiation revealed by single-cell RNA sequencing. Nat. Immunol. 18, 422–432 8. Mahnke, Y.D. et al. (2013) The who’s who of T-cell differentiation: human memory T-cell subsets. Eur. J. Immunol. 43, 2797–2809 9. Gerlach, C. et al. (2011) The descent of memory T cells. Ann. N. Y. Acad. Sci. 1217, 139–153 10. Youngblood, B. et al. (2017) Effector CD8 T cells dedifferentiate into long-lived memory cells. Nature 552, 404–409
tion, Hudfonden, and Psoriasisfonden.
Trends in Immunology, Month Year, Vol. xx, No. yy
3
Spotlight
Heavy Water Shedding Light on AntigenSpecific T Cell Responses
intervention in mice, as compared to the limitations that are inherent to studies in the human system. Although CD8 T cell turnover has been studied in vivo in both mice and humans, human studies have primarily focused on bulk naïve or memory cells at steady-state or during 1,2 Liv Eidsmo and chronic viral infections [1,2]. Conse1,2, Carmen Gerlach * quently, the in vivo proliferation rates of human antigen-specific CD8 T cells CD8 T cells are crucial for long-term responding to acute antigen exposure immunity. Nevertheless, the in vivo have remained elusive.
differentiation of human naïve CD8 T cells into effector and memory populations remains ill-defined. A recent study assesses the in vivo turnover of human antigen-specific CD8 T cells and suggests that longlived memory cells arise from effector cells. CD8 T cells are key players in long-term protection against previously encountered pathogens. They are widely dispersed throughout the body and can be found in blood, lymph, and in lymphoid and non-lymphoid tissues. The precise role of individual CD8 T cells in immune defense largely depends on the antigen specificity, localization, and differentiation state of the cells.
A recent study by Akondy et al. published in Nature employed a beautiful experimental setup that allowed the authors to address fundamental questions on the turnover kinetics and differentiation of human antigen-specific CD8 T cell responses to controlled antigen exposure in vivo [3]. The authors vaccinated several groups of humans in the USA with the live attenuated yellow fever virus (YFV) vaccine that provides lifelong protection against yellow fever, and tracked the responses of YFV-specific CD8 T cells in these individuals over time using MHC–peptide tetramers. Because YFV is not endemic in the USA, and the vaccinated individuals were naïve to the YFVderived antigens at the time of vaccination, the investigators could track true primary responses to the vaccine. Furthermore, the absence of YFV in the environment of the vaccinees prevented reexposure to vaccine antigens and reactivation of YFV-specific T cells in the timeframe between vaccination and the latest point of analysis, which was up to 13 years after vaccination.
Naïve T cells initiate proliferation and differentiation when they are primed by their specific antigen. This leads to the generation of cytotoxic effector T cells that are actively involved in pathogen clearance. In addition, long-lived so-called memory T cells are formed, which provide protection from renewed encounter with the same pathogen. To determine the proliferation rates of the YFV-specific CD8 T cells, the vaccinees Currently, our most detailed understand- drank ‘heavy water’ (2H2O) during differing of CD8 T cell differentiation stems ent time periods after vaccination. Deutefrom murine in vivo studies. This is pre- rium (2H) incorporates into newly dominantly a result of the wealth of synthesized DNA, and the rate at which molecular and genetic tools that are cells incorporate this isotope provides an available for in vivo experimental estimate of the cell division rate. Likewise,
loss of deuterium labeling in cells after cessation of heavy water intake can be used to infer cell division rates. The authors showed that antigen-specific CD8 T cells proliferated most vigorously during the expansion phase of the T cell response (Figure 1). Around the peak of the T cell response, proliferation had considerably slowed down to roughly half the speed observed during the first 2 weeks. Starting with the contraction phase and throughout memory, proliferation events had become extremely rare, with an estimated doubling time of the memory population of several hundred days. While such estimates should be interpreted with some caution because different methods of cell labeling (e.g., deuterium provided in water versus glucose) and different mathematical models to interpret the data may lead to several-fold different inferred proliferation rates [4], it is remarkable how little proliferation is necessary to maintain the memory population. Both the low turnover rate, and the substantial fraction of the memory population that still contained deuterium 1–2 years after labeling, highlight how extremely long-lived memory CD8 T cells can be. Going forward, it will be interesting to investigate whether there is a link between the extent of division a cell undergoes early after antigen encounter, and the longevity of the progeny of the dividing cells. Akondy et al. found that a large fraction of the long-lived memory population still contained the deuterium label they had acquired during the first 2 weeks after vaccination. Because it takes several rounds of division before a population is maximally labeled with deuterium, one could argue that the deuterium-marked memory cells are derived from cells that have undergone extensive proliferation early during the response. On the other hand, one cell division in the presence of deuterium can be sufficient
Trends in Immunology, Month Year, Vol. xx, No. yy
1
TREIMM 1453 No. of Pages 3
High
Intermediate
Extremely low
n
Expa nsio
on
Number of anƟgen-specific CD8 T cells
racƟ
t Con
studies in mice have revealed that memory cells indeed pass through an effector Prolifera on stage during their development [9]. rate Because T cell differentiation is regulated by both transcriptional and epigenetic processes, Akondy et al. characterized naïve and antigen-specific effector and long-term memory populations by RNA sequencing and epigenetic profiling. While memory T cells appeared transcriptionally more similar to naïve than Memory to effector cells, as seen before, epigenetic profiling showed that memory cells 2 weeks 4 weeks 12 years were more similar to effectors in this Time aŌer vaccinaƟon respect. For example, the loci encoding the typical effector molecules IFN-g, Effector genes Naive Effector Memory granzyme B, and perforin were accessible in long-term memory CD8 T cells, although transcription was restricted to Transcrip on effector cells (Figure 1). Their epigenetic accessibility might be a remainder of the Epigene c effector past of the memory cells, conaccessibility sistent with an effector to memory differentiation path for which the same research groups provide more definite Figure 1. [76_TD$IF]Proliferative, Transcriptional, and Epigenetic Characteristics of Human Antigen-Specific CD8 T Cell Responses. (Upper panel) The purple line represents the number of yellow fever virus (YFV)- evidence in an accompanying paper specific CD8 T cells during the different phases (expansion, contraction, memory) of the response. The blue [10]. At the same time, these epigenetic boxes represent timeframes of different proliferation rates of the YFP-specific CD8 T cells. (Lower panel) features poise the memory cells for effecTranscription was assessed by RNA sequencing. Black strings represent mRNA. Epigenetic accessibility of tor functions upon renewed antigen effector gene loci was assessed by ATAC-seq (assay for transposase-accessible chromatin using sequencing) and methylation analysis of the promoter regions. Purple circles represent methylated CpGs. An extended encounter, consistent with the raïson distance between the blue nucleosomes represents open chromatin. d’être of memory cells.
to incorporate the deuterium, and it is possible that the highly proliferative cells die during contraction. Hence, the longlived memory cells may equally well derive from cells that have undergone only limited proliferation in the past. Independent of their division history, it is still a matter of debate whether memory cells derive from effector cells or directly from naïve cells. In the latter case, one model proposes that effector and memory cells develop in parallel from the same naïve T cells, while another assumes that effector cells are the progeny of memory cells [5]. The fact that naïve T cells are phenotypically and 2
Trends in Immunology, Month Year, Vol. xx, No. yy
transcriptionally more similar to memory than to effector cells [6,7] has supported the notion that memory T cells derive directly from naïve T cells [5,8]. Notably, Akondy et al. show that the extremely long-lived memory population (8–12 years after vaccination) expresses CD45RA and CCR7, which are the classical naïve T cell markers, and this underscores the importance of employing an elaborate panel of markers to accurately delineate functionally distinct CD8 T cell subsets [8]. The alternative model of T cell differentiation proposes that naïve T cells initially differentiate into effectors, of which some become memory cells. Supporting this model, in vivo fate-mapping
The study by Akondy et al. elegantly transfers tools that are commonly used in mice to humans, and we foresee that such ‘technology transfer’ will be expanded in the future. Albeit currently far-fetched, in vivo fate-mapping during the course of an immune response may also become feasible in humans. With gene therapies entering the clinic, ex vivo genetic labeling of cells of interest and subsequent autologous cell transfer may be one possibility to allow studies in humans similar to those performed in mice. A possibly earlier achievable but less definitive alternative may be ex vivo deuterium labeling, followed by reinfusion of an albeit partially labeled cell population. Such studies would enable definitive
TREIMM 1453 No. of Pages 3
conclusions on the origin of human T cell subsets, and at the same time establish whether our mouse model-derived knowledge on T cell differentiation can be translated to humans.
1
Dermatology and Venereology Unit, Department of Medicine Solna, Karolinska Institutet, 171 76 Stockholm, Sweden
4. Borghans, J.A.M. and de Boer, R.J. (2007) Quantification of T-cell dynamics: from telomeres to DNA labeling. Immunol. Rev. 216, 35–47
2
5. Farber, D.L. et al. (2014) Human memory T cells: generation, compartmentalization and homeostasis. Nat. Rev. Immunol. 14, 24–35
Department of Dermatology, Karolinska University Hospital Solna, 171 76 Stockholm, Sweden *Correspondence: [email protected] (C. Gerlach). https://doi.org/10.1016/j.it.2018.01.003
Acknowledgments Research in the laboratory of C.G. is supported by the Karolinska Institute, Swedish Research Council, Ragnar Söderbergs [78_TD$IF]stiftelse, Knut and Alice Wallenberg Foundation, Cancerfonden, and Jeanssons Stiftelser. Research in the laboratory of L.E. is supported by the Karolinska Institute, Swedish Research Council, Swedish City Council (ALF), Ragnar Söderbergs [78_TD$IF]4stiftelse, Marcus and Marianne Wallenberg Founda-
References 1. Hellerstein, M.K. et al. (2003) Subpopulations of long-lived and short-lived T cells in advanced HIV-1 infection. J. Clin. Invest. 112, 956–966 2. Vrisekoop, N. et al. (2015) Quantification of naive and memory T-cell turnover during HIV-1 infection. AIDS 29, 2071–2080 3. Akondy, R.S. et al. (2017) Origin and differentiation of human memory CD8 T cells after vaccination. Nature 552, 362–367
6. Best, J.A. et al. (2013) Transcriptional insights into the CD8+ T cell response to infection and memory T cell formation. Nat. Immunol. 14, 404–412 7. Kakaradov, B. et al. (2017) Early transcriptional and epigenetic regulation of CD8+ T cell differentiation revealed by single-cell RNA sequencing. Nat. Immunol. 18, 422–432 8. Mahnke, Y.D. et al. (2013) The who’s who of T-cell differentiation: human memory T-cell subsets. Eur. J. Immunol. 43, 2797–2809 9. Gerlach, C. et al. (2011) The descent of memory T cells. Ann. N. Y. Acad. Sci. 1217, 139–153 10. Youngblood, B. et al. (2017) Effector CD8 T cells dedifferentiate into long-lived memory cells. Nature 552, 404–409
tion, Hudfonden, and Psoriasisfonden.
Trends in Immunology, Month Year, Vol. xx, No. yy
3