- Email: [email protected]
Experimental Evolution to Study Virus Emergence
Leading Edge
Previews Experimental Evolution to Study Virus Emergence Nathan D. Grubaugh1,* and Kristian G. Andersen1,2,3,* 1Department
of Immunology and Microbial Science, The Scripps Research Institute, La Jolla, CA 92037, USA of Integrative Structural and Computational Biology, The Scripps Research Institute, La Jolla, CA 92037, USA 3Scripps Translational Science Institute, La Jolla, CA 92037, USA *Correspondence: [email protected] (N.D.G.), [email protected] (K.G.A.) http://dx.doi.org/10.1016/j.cell.2017.03.018 2Department
Understanding how viruses adapt to new environments and acquire new phenotypes is critical for developing comprehensive responses to outbreaks. By studying the emergence of vaccine-derived poliovirus outbreaks, Stern et al. describe how a combination of sequence analysis and experimental evolution can be used to reveal adaptive pathways. Poliovirus, which once afflicted millions of people a year, has been brought to near eradication by aggressive vaccine campaigns (Kew et al., 2005). Efforts to clear the last remaining poliovirus strongholds, however, are obstructed in areas with low vaccine coverage, which can facilitate outbreaks of poliomyelitis caused by the oral poliovirus vaccine (OPV). OPV is derived from an avirulent strain that is still able to replicate and has a remarkable safety record (Kew et al., 2005). Vaccination with OPV will elicit a strong immune response, which leads to protection from poliovirus infection in vaccinated individuals. Because OPV can replicate within the recipient, it may transmit to unvaccinated individuals and can revert to a more pathogenic form (Kew et al., 2005). OPV transmission can only occur when herd immunity to poliovirus is low, highlighting the critical importance of ensuring high vaccine coverage. In a new study in this issue of Cell, Stern et al. (2017) investigated how OPV type 2 Sabin strain (OPV2, one of three strains in the trivalent vaccine) can evolve elevated virulence during periods of uncontrolled circulation in low-vaccine areas. Their study gives deep insights into how an attenuated virus can regain virulence, and their experimental and analyses frameworks will be relevant for the study of emerging human pathogens (Figure 1). Stern et al. (2017) hypothesized that independent outbreaks of OPV2 evolved similar sets of mutations leading to increased transmission and virulence. Two critical aspects allowed them to track OPV2 evolution during transmission to properly test this hypothesis: (1) the
ancestral virus for each outbreak was known (i.e., the attenuated OPV2 vaccine) and (2) open sharing of sequencing data from vaccine-derived poliovirus infections that occurred during previous outbreaks in Belarus, China, Egypt, Madagascar, and Nigeria. From a dataset of 424 OPV2derived genomes that led to human disease, the authors detected 841 sites that occurred repeatedly in multiple outbreaks. They designed a probabilistic model called ParaSel to differentiate between sites under positive selection and random drift, to identify seven sites in the OPV2 genome that the authors hypothesize evolved to restore poliovirus virulence. Three substitutions, A481G, U2909C, and U398C, were found in almost all OPV2-derived sequences across all outbreaks, while the others only emerged after months of transmission. Furthermore, they found that the A481G, U2909C, and U398C mutations occurred in that order during replication in its first host—OPV2 vaccinees. These data suggest that adaptive evolution occurs in ordered waves, akin to a multistep evolutionary process (Borderı´a et al., 2015; Tsetsarkin et al., 2014) where additional fitness peaks are revealed after the initial gate-keeper mutation(s). The first adaptive wave occurs rapidly and is followed by additional waves of evolutionary fine-tuning. Thus, understanding the conditions that promote the first wave of virus adaptation is of utmost importance for responding to emerging and zoonotic pathogens. Through a series of elegant in vivo and in vitro experiments, Stern et al. (2017) demonstrated how OPV2 could restore virulence. To create a safe vaccine strain, the attenuated OPV2 is manufactured by
adapting the virus to non-human cells at sub-physiological (33 C) temperatures (Kew et al., 2005). By sequencing poliovirus directly from OPV2 vaccinees and after serial passages in human cells at physiological temperatures, the authors determined that the first wave of adaptive mutations (A481G, U2909C, and U398C) were immediately selected for. They used a reverse genetics platform to engineer each mutation and combination into OPV2 to test fitness and virulence. These experiments showed that A481G is likely the true gatekeeper mutation, as it was the only one that the authors found to individually have a significant impact on virus fitness in cell culture and virulence in mice. The other mutations were shown to provide additional benefits through epistatic interactions with A481G, but did not significantly impact fitness on their own. Therefore, the first adaptive wave during an outbreak may consist of several steps as the virus evolves to reach optimal fitness. However, future research is needed to understand the interactions between the mutations and reveal the mechanisms leading to increased virulence. Stern et al. (2017) made a significant discovery that adaptive mutations occurring in nature could, under the right circumstances, be replicated using experimental models. We cannot, however, hope to exactly mimic what occurs during outbreaks in a test tube. In addition, we can likely only make small inferences about large evolutionary events, like host jumps and new transmission pathways. For example, the experimental framework used by Stern et al. can provide general— yet critical—insights into virus-host interactions, but they cannot provide the full
Cell 169, March 23, 2017 ª 2017 Elsevier Inc. 1
Figure 1. Experimental Evolution Can Provide Insights about Virus Emergence Sequencing viruses directly from clinical samples during an outbreak (top panel) can reveal patterns of selection and convergent evolution (i.e., occur in more than one lineage). Stern et al. (2017) have revealed that adaptive mutations occur in waves. The first waves contains the gate-keeper mutations that are immediately selected for in a new environment (in this case, OPV2 in humans) and provide the largest fitness increases. Additional adaptive waves fine-tune virus fitness to its transmission cycle. The darker silhouettes represent the main transmission path and the lighter silhouettes represent paths that lead to extinction. Experimental evolution (bottom panel) can be used to provide insights about adaptation during virus emergence, especially during host jumps. OPV2 was attenuated in non-human cells, and here this shows that replication in human cells is sufficient to generate the first wave of adaptive mutations detected during the outbreak.
spectrum of interactions that occur during human infection. A chikungunya virus (CHIKV) epidemic provides another example of how outbreak investigation and laboratory experiments can be combined to understand virus emergence, similar to the work by Stern et al. (2017). Beginning in 2005, a large CHIKV epidemic swept throughout the Indian Ocean basin. A crucial finding was that the mosquito-borne CHIKV was being transmitted by a non-canonical vector, Aedes albopictus. This observation raised many questions about how virus adaptation might have contributed to the epidemic. Analysis of CHIKV genomes sequenced from clinical samples collected during the epidemic revealed sites potentially under positive selection. The most notable of these was a novel amino acid substitution in the envelope protein (E1-A226V) from the CHIKV lineage
2 Cell 169, March 23, 2017
that dominated the epidemic (Schuffenecker et al., 2006). Using in vivo models, it was demonstrated that the E1-A226V mutation significantly enhanced transmission efficiency by Ae. albopictus mosquitoes to humans (Tsetsarkin et al., 2007) and opened new pathways for continued fitness optimization in the vector (Tsetsarkin et al., 2014). Importantly, the E1A226V mutation could be generated de novo from the pre-endemic CHIKV lineage during experimental infection of Ae. albopictus mosquitoes (Stapleford et al., 2014). These findings, along with the research from Stern and colleagues on OPV2, represent examples of rapid virus adaptation during vector and host jumps that can have detrimental effects to human health. The work by Stern et al. creates a framework upon which future experimental evolution studies could be uti-
lized to investigate the mechanisms underpinning emerging virus outbreaks of significant public health concern. A recent example is the West African Ebola epidemic of 2013–2016, where an amino acid substitution in the viral glycoprotein (GP-A82V) appeared and eventually became dominant in the viral population (Diehl et al., 2016; Urbanowicz et al., 2016). This mutation enhanced the ability of Ebola virus to infect primate cells, but not in other species tested, suggesting adaptation to humans (Diehl et al., 2016; Urbanowicz et al., 2016). These studies, however, lack explanations for the context in which the mutation emerged, the mechanism(s) leading to altered Ebola virus fitness, and evidence to determine whether the mutation impacted transmission and scale of the Ebola outbreak. The analysis methods and experimental framework described by Stern et al. could be used to probe these important questions about Ebola virus evolution and adaptation. The experiments described by Stern et al. (2017) were used to reconstruct events that have already occurred, but these experiments can also be used to look ahead to virus emergence. For example, experimental evolution studies revealed that highly pathogenic avian influenza A/H5N1 virus could acquire an aerosolized transmission route between mammals (Herfst et al., 2012). The importance of such work is not to show that H5N1 will necessarily become aerosolized, but to know that the virus is potentially only a few steps away from reaching pandemic potential. We can also use a forward-thinking approach to design better attenuated live virus vaccines. Experimental models and sequencing during clinical trials could be used to detect attenuation losses and help the redesign process before the vaccines go onto the market. Continued studies of this nature can help us become more prepared for future shifts in virus transmission that could significantly impact public health. ACKNOWLEDGMENTS N.D.G. is supported by a NIH training grant 5T32AI007244-33. K.G.A. is a Pew Biomedical Scholar, and his work is supported by an NIH National Center for Advancing Translational Studies Clinical and Translational Science Award UL1TR001114 and NIAID contract HHSN272201400048C.
REFERENCES
Bestebroer, T.M., Burke, D.F., Smith, D.J., et al. (2012). Science 336, 1534–1541.
Borderı´a, A.V., Isakov, O., Moratorio, G., Henningsson, R., Agu¨era-Gonza´lez, S., Organtini, L., Gna¨dig, N.F., Blanc, H., Alcover, A., Hafenstein, S., et al. (2015). PLoS Pathog. 11, e1004838.
Kew, O.M., Sutter, R.W., de Gourville, E.M., Dowdle, W.R., and Pallansch, M.A. (2005). Annu. Rev. Microbiol. 59, 587–635.
Diehl, W.E., Lin, A.E., Grubaugh, N.D., Carvalho, L.M., Kim, K., Kyawe, P.P., McCauley, S.M., Donnard, E., Kucukural, A., McDonel, P., et al. (2016). Cell 167, 1088–1098.e6. Herfst, S., Schrauwen, E.J.A., Linster, M., Chutinimitkul, S., de Wit, E., Munster, V.J., Sorrell, E.M.,
Stern, A., Yeh, M.T., Zinger, T., Smith, M., Wright, C., Ling, G., Nielsen, R., Macadam, A., and Andino, R. (2017). Cell 169, this issue, 35–46. Tsetsarkin, K.A., Vanlandingham, D.L., McGee, C.E., and Higgs, S. (2007). PLoS Pathog. 3, e201.
Schuffenecker, I., Iteman, I., Michault, A., Murri, S., Frangeul, L., Vaney, M.-C., Lavenir, R., Pardigon, N., Reynes, J.-M., Pettinelli, F., et al. (2006). PLoS Med. 3, e263.
Tsetsarkin, K.A., Chen, R., Yun, R., Rossi, S.L., Plante, K.S., Guerbois, M., Forrester, N., Perng, G.C., Sreekumar, E., Leal, G., et al. (2014). Nat. Commun. 5, 4084.
Stapleford, K.A., Coffey, L.L., Lay, S., Borderı´a, A.V., Duong, V., Isakov, O., Rozen-Gagnon, K., Arias-Goeta, C., Blanc, H., Beaucourt, S., et al. (2014). Cell Host Microbe 15, 706–716.
Urbanowicz, R.A., McClure, C.P., Sakuntabhai, A., Sall, A.A., Kobinger, G., Mu¨ller, M.A., Holmes, E.C., Rey, F.A., Simon-Loriere, E., and Ball, J.K. (2016). Cell 167, 1079–1087.e5.
Rejuvenation by Therapeutic Elimination of Senescent Cells Paul Krimpenfort1 and Anton Berns1,* 1Division of Molecular Genetics, the Netherlands Cancer Institute, Amsterdam, the Netherlands *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cell.2017.03.014
In this issue of Cell, Baar et al. show how FOXO4 protects senescent cell viability by keeping p53 sequestered in nuclear bodies, preventing it from inducing apoptosis. Disrupting this interaction with an all-D amino acid peptide (FOXO4-DRI) restores p53’s apoptotic role and ameliorates the consequences of senescence-associated loss of tissue homeostasis. Senescence arises as the consequence of cellular stress such as telomere erosion, unresolved DNA damage, or oncogenic signaling. For a cell, it embodies a twilight zone between regaining its normal function and death, and it serves as one of the mechanisms to stop incipient cancer cells in their tracks. Senescence is characterized by a persistent proliferative arrest in which cells display a distinct pro-inflammatory senescent-associated secretory phenotype (SASP) characterized by secretion of factors such as IL6 and IL8 (Kuilman et al., 2008). Whereas SASP exerts a supportive paracrine function during early development and wound healing (Demaria et al., 2014), the continuous secretion of these SASP factors has detrimental effects on normal tissue homeostasis and is considered to significantly contribute to aging (DiLoreto and Murphy, 2015). In this issue of Cell, Baar and colleagues have uncovered the molecular mechanisms underlying maintenance of senes-
cent cells and their effect on tissue homeostasis (Figure 1). In 2016, Van Deursen and coworkers showed that prophylactic ablation of cells that express the p16Ink4a senescent marker mitigates tissue degeneration and extends the healthspan of mice (Baker et al., 2016). Although the Baker et al. study established the causative role of senescent cells in tissue degeneration, it did not reveal the underlying mechanisms responsible for their generation. This is now addressed in the article by Baar et al. (2017). The authors compared the expression profiles of normal human IMR90 cells with their radiation-induced senescent counterparts. Whereas they expected that the senescent cells would show reduced levels of pro-apoptotic and/or an increase in anti-apoptotic proteins, they observed the opposite, suggesting lack of an upstream trigger to activate the death program. Therefore, they searched
for changes in the expression levels of transcription regulators that could underlie these changes. One immediate candidate was FOXO4, a member of a transcription factor family previously implicated in aging and longevity (Martins et al., 2016), which appeared upregulated upon senescence induction. Therefore, FOXO4 was further scrutinized for its potential role in arresting cells in a senescent state. Indeed, inhibition of FOXO4 expression using a lentiviral shRNA made cells undergo apoptosis rather than become senescent upon irradiation. More importantly, FOXO4 inhibition in already senescent cells reduced their viability. This established that FOXO4 plays a critical role in consolidating the senescent state and that its inactivation causes senescent cells to undergo apoptosis. To gain insight into the underlying mechanism, the authors went on to determine the subnuclear localization of
Cell 169, March 23, 2017 ª 2017 Elsevier Inc. 3
Previews Experimental Evolution to Study Virus Emergence Nathan D. Grubaugh1,* and Kristian G. Andersen1,2,3,* 1Department
of Immunology and Microbial Science, The Scripps Research Institute, La Jolla, CA 92037, USA of Integrative Structural and Computational Biology, The Scripps Research Institute, La Jolla, CA 92037, USA 3Scripps Translational Science Institute, La Jolla, CA 92037, USA *Correspondence: [email protected] (N.D.G.), [email protected] (K.G.A.) http://dx.doi.org/10.1016/j.cell.2017.03.018 2Department
Understanding how viruses adapt to new environments and acquire new phenotypes is critical for developing comprehensive responses to outbreaks. By studying the emergence of vaccine-derived poliovirus outbreaks, Stern et al. describe how a combination of sequence analysis and experimental evolution can be used to reveal adaptive pathways. Poliovirus, which once afflicted millions of people a year, has been brought to near eradication by aggressive vaccine campaigns (Kew et al., 2005). Efforts to clear the last remaining poliovirus strongholds, however, are obstructed in areas with low vaccine coverage, which can facilitate outbreaks of poliomyelitis caused by the oral poliovirus vaccine (OPV). OPV is derived from an avirulent strain that is still able to replicate and has a remarkable safety record (Kew et al., 2005). Vaccination with OPV will elicit a strong immune response, which leads to protection from poliovirus infection in vaccinated individuals. Because OPV can replicate within the recipient, it may transmit to unvaccinated individuals and can revert to a more pathogenic form (Kew et al., 2005). OPV transmission can only occur when herd immunity to poliovirus is low, highlighting the critical importance of ensuring high vaccine coverage. In a new study in this issue of Cell, Stern et al. (2017) investigated how OPV type 2 Sabin strain (OPV2, one of three strains in the trivalent vaccine) can evolve elevated virulence during periods of uncontrolled circulation in low-vaccine areas. Their study gives deep insights into how an attenuated virus can regain virulence, and their experimental and analyses frameworks will be relevant for the study of emerging human pathogens (Figure 1). Stern et al. (2017) hypothesized that independent outbreaks of OPV2 evolved similar sets of mutations leading to increased transmission and virulence. Two critical aspects allowed them to track OPV2 evolution during transmission to properly test this hypothesis: (1) the
ancestral virus for each outbreak was known (i.e., the attenuated OPV2 vaccine) and (2) open sharing of sequencing data from vaccine-derived poliovirus infections that occurred during previous outbreaks in Belarus, China, Egypt, Madagascar, and Nigeria. From a dataset of 424 OPV2derived genomes that led to human disease, the authors detected 841 sites that occurred repeatedly in multiple outbreaks. They designed a probabilistic model called ParaSel to differentiate between sites under positive selection and random drift, to identify seven sites in the OPV2 genome that the authors hypothesize evolved to restore poliovirus virulence. Three substitutions, A481G, U2909C, and U398C, were found in almost all OPV2-derived sequences across all outbreaks, while the others only emerged after months of transmission. Furthermore, they found that the A481G, U2909C, and U398C mutations occurred in that order during replication in its first host—OPV2 vaccinees. These data suggest that adaptive evolution occurs in ordered waves, akin to a multistep evolutionary process (Borderı´a et al., 2015; Tsetsarkin et al., 2014) where additional fitness peaks are revealed after the initial gate-keeper mutation(s). The first adaptive wave occurs rapidly and is followed by additional waves of evolutionary fine-tuning. Thus, understanding the conditions that promote the first wave of virus adaptation is of utmost importance for responding to emerging and zoonotic pathogens. Through a series of elegant in vivo and in vitro experiments, Stern et al. (2017) demonstrated how OPV2 could restore virulence. To create a safe vaccine strain, the attenuated OPV2 is manufactured by
adapting the virus to non-human cells at sub-physiological (33 C) temperatures (Kew et al., 2005). By sequencing poliovirus directly from OPV2 vaccinees and after serial passages in human cells at physiological temperatures, the authors determined that the first wave of adaptive mutations (A481G, U2909C, and U398C) were immediately selected for. They used a reverse genetics platform to engineer each mutation and combination into OPV2 to test fitness and virulence. These experiments showed that A481G is likely the true gatekeeper mutation, as it was the only one that the authors found to individually have a significant impact on virus fitness in cell culture and virulence in mice. The other mutations were shown to provide additional benefits through epistatic interactions with A481G, but did not significantly impact fitness on their own. Therefore, the first adaptive wave during an outbreak may consist of several steps as the virus evolves to reach optimal fitness. However, future research is needed to understand the interactions between the mutations and reveal the mechanisms leading to increased virulence. Stern et al. (2017) made a significant discovery that adaptive mutations occurring in nature could, under the right circumstances, be replicated using experimental models. We cannot, however, hope to exactly mimic what occurs during outbreaks in a test tube. In addition, we can likely only make small inferences about large evolutionary events, like host jumps and new transmission pathways. For example, the experimental framework used by Stern et al. can provide general— yet critical—insights into virus-host interactions, but they cannot provide the full
Cell 169, March 23, 2017 ª 2017 Elsevier Inc. 1
Figure 1. Experimental Evolution Can Provide Insights about Virus Emergence Sequencing viruses directly from clinical samples during an outbreak (top panel) can reveal patterns of selection and convergent evolution (i.e., occur in more than one lineage). Stern et al. (2017) have revealed that adaptive mutations occur in waves. The first waves contains the gate-keeper mutations that are immediately selected for in a new environment (in this case, OPV2 in humans) and provide the largest fitness increases. Additional adaptive waves fine-tune virus fitness to its transmission cycle. The darker silhouettes represent the main transmission path and the lighter silhouettes represent paths that lead to extinction. Experimental evolution (bottom panel) can be used to provide insights about adaptation during virus emergence, especially during host jumps. OPV2 was attenuated in non-human cells, and here this shows that replication in human cells is sufficient to generate the first wave of adaptive mutations detected during the outbreak.
spectrum of interactions that occur during human infection. A chikungunya virus (CHIKV) epidemic provides another example of how outbreak investigation and laboratory experiments can be combined to understand virus emergence, similar to the work by Stern et al. (2017). Beginning in 2005, a large CHIKV epidemic swept throughout the Indian Ocean basin. A crucial finding was that the mosquito-borne CHIKV was being transmitted by a non-canonical vector, Aedes albopictus. This observation raised many questions about how virus adaptation might have contributed to the epidemic. Analysis of CHIKV genomes sequenced from clinical samples collected during the epidemic revealed sites potentially under positive selection. The most notable of these was a novel amino acid substitution in the envelope protein (E1-A226V) from the CHIKV lineage
2 Cell 169, March 23, 2017
that dominated the epidemic (Schuffenecker et al., 2006). Using in vivo models, it was demonstrated that the E1-A226V mutation significantly enhanced transmission efficiency by Ae. albopictus mosquitoes to humans (Tsetsarkin et al., 2007) and opened new pathways for continued fitness optimization in the vector (Tsetsarkin et al., 2014). Importantly, the E1A226V mutation could be generated de novo from the pre-endemic CHIKV lineage during experimental infection of Ae. albopictus mosquitoes (Stapleford et al., 2014). These findings, along with the research from Stern and colleagues on OPV2, represent examples of rapid virus adaptation during vector and host jumps that can have detrimental effects to human health. The work by Stern et al. creates a framework upon which future experimental evolution studies could be uti-
lized to investigate the mechanisms underpinning emerging virus outbreaks of significant public health concern. A recent example is the West African Ebola epidemic of 2013–2016, where an amino acid substitution in the viral glycoprotein (GP-A82V) appeared and eventually became dominant in the viral population (Diehl et al., 2016; Urbanowicz et al., 2016). This mutation enhanced the ability of Ebola virus to infect primate cells, but not in other species tested, suggesting adaptation to humans (Diehl et al., 2016; Urbanowicz et al., 2016). These studies, however, lack explanations for the context in which the mutation emerged, the mechanism(s) leading to altered Ebola virus fitness, and evidence to determine whether the mutation impacted transmission and scale of the Ebola outbreak. The analysis methods and experimental framework described by Stern et al. could be used to probe these important questions about Ebola virus evolution and adaptation. The experiments described by Stern et al. (2017) were used to reconstruct events that have already occurred, but these experiments can also be used to look ahead to virus emergence. For example, experimental evolution studies revealed that highly pathogenic avian influenza A/H5N1 virus could acquire an aerosolized transmission route between mammals (Herfst et al., 2012). The importance of such work is not to show that H5N1 will necessarily become aerosolized, but to know that the virus is potentially only a few steps away from reaching pandemic potential. We can also use a forward-thinking approach to design better attenuated live virus vaccines. Experimental models and sequencing during clinical trials could be used to detect attenuation losses and help the redesign process before the vaccines go onto the market. Continued studies of this nature can help us become more prepared for future shifts in virus transmission that could significantly impact public health. ACKNOWLEDGMENTS N.D.G. is supported by a NIH training grant 5T32AI007244-33. K.G.A. is a Pew Biomedical Scholar, and his work is supported by an NIH National Center for Advancing Translational Studies Clinical and Translational Science Award UL1TR001114 and NIAID contract HHSN272201400048C.
REFERENCES
Bestebroer, T.M., Burke, D.F., Smith, D.J., et al. (2012). Science 336, 1534–1541.
Borderı´a, A.V., Isakov, O., Moratorio, G., Henningsson, R., Agu¨era-Gonza´lez, S., Organtini, L., Gna¨dig, N.F., Blanc, H., Alcover, A., Hafenstein, S., et al. (2015). PLoS Pathog. 11, e1004838.
Kew, O.M., Sutter, R.W., de Gourville, E.M., Dowdle, W.R., and Pallansch, M.A. (2005). Annu. Rev. Microbiol. 59, 587–635.
Diehl, W.E., Lin, A.E., Grubaugh, N.D., Carvalho, L.M., Kim, K., Kyawe, P.P., McCauley, S.M., Donnard, E., Kucukural, A., McDonel, P., et al. (2016). Cell 167, 1088–1098.e6. Herfst, S., Schrauwen, E.J.A., Linster, M., Chutinimitkul, S., de Wit, E., Munster, V.J., Sorrell, E.M.,
Stern, A., Yeh, M.T., Zinger, T., Smith, M., Wright, C., Ling, G., Nielsen, R., Macadam, A., and Andino, R. (2017). Cell 169, this issue, 35–46. Tsetsarkin, K.A., Vanlandingham, D.L., McGee, C.E., and Higgs, S. (2007). PLoS Pathog. 3, e201.
Schuffenecker, I., Iteman, I., Michault, A., Murri, S., Frangeul, L., Vaney, M.-C., Lavenir, R., Pardigon, N., Reynes, J.-M., Pettinelli, F., et al. (2006). PLoS Med. 3, e263.
Tsetsarkin, K.A., Chen, R., Yun, R., Rossi, S.L., Plante, K.S., Guerbois, M., Forrester, N., Perng, G.C., Sreekumar, E., Leal, G., et al. (2014). Nat. Commun. 5, 4084.
Stapleford, K.A., Coffey, L.L., Lay, S., Borderı´a, A.V., Duong, V., Isakov, O., Rozen-Gagnon, K., Arias-Goeta, C., Blanc, H., Beaucourt, S., et al. (2014). Cell Host Microbe 15, 706–716.
Urbanowicz, R.A., McClure, C.P., Sakuntabhai, A., Sall, A.A., Kobinger, G., Mu¨ller, M.A., Holmes, E.C., Rey, F.A., Simon-Loriere, E., and Ball, J.K. (2016). Cell 167, 1079–1087.e5.
Rejuvenation by Therapeutic Elimination of Senescent Cells Paul Krimpenfort1 and Anton Berns1,* 1Division of Molecular Genetics, the Netherlands Cancer Institute, Amsterdam, the Netherlands *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cell.2017.03.014
In this issue of Cell, Baar et al. show how FOXO4 protects senescent cell viability by keeping p53 sequestered in nuclear bodies, preventing it from inducing apoptosis. Disrupting this interaction with an all-D amino acid peptide (FOXO4-DRI) restores p53’s apoptotic role and ameliorates the consequences of senescence-associated loss of tissue homeostasis. Senescence arises as the consequence of cellular stress such as telomere erosion, unresolved DNA damage, or oncogenic signaling. For a cell, it embodies a twilight zone between regaining its normal function and death, and it serves as one of the mechanisms to stop incipient cancer cells in their tracks. Senescence is characterized by a persistent proliferative arrest in which cells display a distinct pro-inflammatory senescent-associated secretory phenotype (SASP) characterized by secretion of factors such as IL6 and IL8 (Kuilman et al., 2008). Whereas SASP exerts a supportive paracrine function during early development and wound healing (Demaria et al., 2014), the continuous secretion of these SASP factors has detrimental effects on normal tissue homeostasis and is considered to significantly contribute to aging (DiLoreto and Murphy, 2015). In this issue of Cell, Baar and colleagues have uncovered the molecular mechanisms underlying maintenance of senes-
cent cells and their effect on tissue homeostasis (Figure 1). In 2016, Van Deursen and coworkers showed that prophylactic ablation of cells that express the p16Ink4a senescent marker mitigates tissue degeneration and extends the healthspan of mice (Baker et al., 2016). Although the Baker et al. study established the causative role of senescent cells in tissue degeneration, it did not reveal the underlying mechanisms responsible for their generation. This is now addressed in the article by Baar et al. (2017). The authors compared the expression profiles of normal human IMR90 cells with their radiation-induced senescent counterparts. Whereas they expected that the senescent cells would show reduced levels of pro-apoptotic and/or an increase in anti-apoptotic proteins, they observed the opposite, suggesting lack of an upstream trigger to activate the death program. Therefore, they searched
for changes in the expression levels of transcription regulators that could underlie these changes. One immediate candidate was FOXO4, a member of a transcription factor family previously implicated in aging and longevity (Martins et al., 2016), which appeared upregulated upon senescence induction. Therefore, FOXO4 was further scrutinized for its potential role in arresting cells in a senescent state. Indeed, inhibition of FOXO4 expression using a lentiviral shRNA made cells undergo apoptosis rather than become senescent upon irradiation. More importantly, FOXO4 inhibition in already senescent cells reduced their viability. This established that FOXO4 plays a critical role in consolidating the senescent state and that its inactivation causes senescent cells to undergo apoptosis. To gain insight into the underlying mechanism, the authors went on to determine the subnuclear localization of
Cell 169, March 23, 2017 ª 2017 Elsevier Inc. 3