npg
© 2014 Nature America, Inc. All rights reserved.
news and views One unexpected aspect of this study is the finding that SAMHD1 does not degrade RNA in the context of RNA-DNA hybrids, which conflicts with previous work12. Therefore, how is SAMHD1’s nucleolytic activity directed to viral, rather than host, RNA? The authors speculate that SAMHD1 might bind to one or more components of the viral reverse transcription apparatus. SAMHD1’s RNase could act on either the tRNALys3 primer that initiates reverse transcription, the HIV-1 genomic RNA that serves as the template for cDNA generation or both. Northern blotting revealed no changes in the tRNALys3 levels. Using both quantitative RT-PCR and RNA-seq, Ryoo et al.5 identified HIV genomic RNA as the principal target and showed that knockdown of SAMHD1 in primary human monocyte–derived macrophages and CD4+ T cells increased the stability of HIV-1 RNA following infection. Despite the increasing evidence favoring SAMHD1’s RNase as the dominant player in HIV restriction, its dNTPase activity cannot be entirely dismissed. Recently, timecourse studies showed that HIV infection can be rescued by blocking SAMHD1 function, even though the virus was exposed to active SAMHD1 for several hours14. Such a rescue would not be predicted if SAMHD1
acts through its RNase activity alone, as this nucleolytic action would presumably be nonreversible. An intriguing possibility suggested by Ryoo et al.5 is that both enzymatic functions are important, but they operate in different cellular environments. SAMHD1’s dNTPase activity depends upon tetramerization of the enzyme, which occurs in a dGTP-dependent manner15,16. Conversely, the RNase activity of SAMHD1 does not depend on either dGTP binding or tetramer formation. It is possible that that tetramerization impairs RNase activity, but this has not been investigated to date. The levels of intracellular dGTP could therefore dictate which enzymatic function of SAMHD1 predominates. In this scenario, dGTP serves as both a regulator and a substrate of SAMHD1. The fact that dGTP levels are generally low in resting CD4+ T cells and macrophages suggests that the RNase activity may be dominant in these cells. However, slight changes in activation state could boost dGTP levels sufficiently to stimulate tetramerization and dNTPase activation. This dGTP-regulated dual-enzyme model for SAMHD1 action is attractive, but further experiments are needed to resolve whether one or both of SAMHD1’s enzymatic actions are in play during HIV
restriction. A more complete understanding of how SAMHD1 restricts HIV as well as of the counterattack leveled by the lentiviral Vpx proteins will provide new insights into viral pathogenesis and inform new strategies for preserving SAMHD1 restriction activity. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. 1. Hrecka, K. et al. Nature 474, 658–661 (2011). 2. Laguette, N. et al. Nature 474, 654–657 (2011). 3. Baldauf, H.M. et al. Nat. Med. 18, 1682–1687 (2012). 4. Descours, B. et al. Retrovirology 9, 87 (2012). 5. Ryoo, J. et al. Nat. Med. 20, 936–941 (2014). 6. Goldstone, D.C. et al. Nature 480, 379–382 (2011). 7. Powell, R.D., Holland, P.J., Hollis, T. & Perrino, F.W. J. Biol. Chem. 286, 43596–43600 (2011). 8. Lahouassa, H. et al. Nat. Immunol. 13, 223–228 (2012). 9. Cribier, A., Descours, B., Valadão, A.L., Laguette, N. & Benkirane, M. Cell Reports 3, 1036–1043 (2013). 10. White, T.E. et al. Cell Host Microbe 13, 441–451 (2013). 11. Welbourn, S., Dutta, S.M., Semmes, O.J. & Strebel, K. J. Virol. 87, 11516–11524 (2013). 12. Beloglazova, N. et al. J. Biol. Chem. 288, 8101–8110 (2013). 13. Crow, Y.J. & Livingston, J.H. Dev. Med. Child Neurol. 50, 410–416 (2008). 14. Hofmann, H. et al. J. Virol. 87, 11741–11750 (2013). 15. Ji, X. et al. Nat. Struct. Mol. Biol. 20, 1304–1309 (2013). 16. Yan, J. et al. J. Biol. Chem. 288, 10406–10417 (2013).
A local circadian clock calls time on lung inflammation A A Roger Thompson, Sarah R Walmsley & Moira K B Whyte Inflammatory diseases typically display circadian variation in symptom severity. A new study in mice shows how a pulmonary epithelial cell clock controls neutrophil recruitment to the lungs and provides insight into interactions between local and systemic circadian clocks. Controlling excessive neutrophilic inflammation remains a major unmet clinical need in inflammatory lung diseases, including chronic obstructive pulmonary disease (COPD) and neutrophilic asthma, in which the essential antimicrobial functions of the neutrophil may be superseded by persistent inflammation and tissue damage. This excessive inflammation often coexists with persistence of bacteria in the lower airways, implying ineffective host defenses despite the presence of inflammation1. A.A. Roger Thompson, Sarah R. Walmsley & Moira K.B. Whyte are in the Academic Unit of Respiratory Medicine, Department of Infection and Immunity, The Medical School, University of Sheffield, Sheffield, UK. e-mail: [email protected]
Understanding the mechanisms responsible for recruiting neutrophils to the lungs and their subsequent activation and host-defense function is crucial to direct therapeutic interventions in these conditions. Inflammatory diseases are often characterized by diurnal variation in symptom severity, such as prominent early morning symptoms of wheeze in poorly controlled asthma and worsening of lung function in COPD2,3. A circadian clock, set by a central timekeeper in the brain’s suprachiasmatic nucleus, is responsive to external signals and can entrain peripheral rhythms via both endocrine and neural pathways, thus setting a circadian rhythm that affects the body as a whole. Peripheral tissues and cells also possess an intrinsic circadian cycle, employing the same genetic machinery
nature medicine volume 20 | number 8 | AUGUST 2014
as the central clock and involving feedback loop–regulated transcription and translation of clock genes including CLOCK, ARNTL (here called BMAL1), PER1–PER3 and CRY1 and CRY2 (ref. 4). In recent years, circadian rhythms in immune cells have been shown to affect systemic cytokine responses, which are largely of myeloid cell origin, with consequences for cellular trafficking and host responses to pathogens5–9. Although this clock machinery functions autonomously in isolated immune cells, there are also interactions between central and peripheral rhythms that affect the circadian regulation of inflammatory responses in individual tissues and that still remain unclear. In this issue of Nature Medicine, Gibbs et al.10 uncover a new mechanism governing circadian neutrophil 809
news and views
Tissue homeostasis
Debbie Maizels/ Nature Publishing Group
CXCL5
GR Club cell
Infectious/inflammatory stimuli (S. pneumoniae infection or LPS challenge)
Ciliated airway
CXCL5 CXCL5 CXCL5
Local clock disruption (Bmal1 deletion in bronchiolar club cells)
Increased neutrophil recruitment
Excess local inflammation; inefficient antibacterial response
npg
© 2014 Nature America, Inc. All rights reserved.
Figure 1 The pulmonary clock and inflammatory responses. Gibbs et al.10 show that inhalation of infectious (S. pneumoniae) or inflammatory (lipopolysaccharide (LPS) challenge) stimuli in mice results in release of CXCL5 from bronchiolar club cells within the airways and induction of neutrophilic inflammation, which peaks at the start of rest phase (dawn). Endogenous glucocorticoids rhythmically suppress Cxcl5 transcription in bronchiolar cells, reducing the amounts of secreted CXCL5 and subsequent neutrophilmediated inflammation. This regulation is lost upon local failure of the pulmonary clock through deletion of the clock gene Bmal1 (and potentially through alteration of clock gene function by environmental stimuli, or a naturally occurring disease mutation) in bronchiolar cells and cannot be overcome by exogenous glucocorticoid treatment.
recruitment to the lungs in response to local bacterial stimuli that is mediated by the circadian regulation of the chemokine CXCL5 in pulmonary bronchiolar cells. Disruption of this lung clock increased recruitment of neutrophils to the lung without improving bacterial clearance and, importantly, promoted resistance to anti-inflammatory glucocorticoids. Gibbs et al.10 first found in mice that local pulmonary responses to inhaled endotoxin and pneumococcal infection are under circadian regulation: peak inflammation was seen after endotoxin challenge was given at ‘dawn’ (the start of rest phase in mice). This is in marked contrast to circadian variation in systemic cytokine responses, measured after intraperitoneal endotoxin challenge, in which case maximal cytokine release was observed when mice were challenged at ‘dusk’ (start of active phase)7. Using mice with conditional deletion of the clock gene Bmal1 in either myeloid cells (LysM-Bmal1−/−) or bronchiolar club cells (Ccsp-Bmal1−/−) to specifically disrupt the circadian clock in these populations, they found that the local pulmonary inflammation was regulated by circadian clock function in bronchiolar club cells (formerly known as Clara cells) but not in alveolar macrophages or other myeloid cells10. Furthermore, the authors also observed increased pulmonary neutrophilia in mice lacking club cell Bmal1 after an intratracheal challenge with Streptococcus pneumoniae, although, interestingly, the increased inflammation did not improve the clearance of bacteria10. Gibbs et al.10 found that expression of a number of inflammatory mediators was increased in mice lacking Bmal1 in bronchiolar club cells compared with wild-type mice after pulmonary endotoxin challenge, notably the chemokine CXCL5, a known neutrophil chemoattractant. In wild-type mice, prolonged secretion of CXCL5 from bronchiolar cells into the airways in response to endotoxin coincided with neutrophil persistence in the lungs. 810
In Cxcl5-deficient mice, neutrophils were still recruited, but the diurnal variations in neutrophilic response were abolished (Fig. 1). The authors gained mechanistic insights into the rhythmic regulation of CXCL5 through investigation of the interplay between glucocorticoidmediated repression of Cxcl5 transcription and local clock function. Glucocorticoids, via activated glucocorticoid receptor (GR) binding to glucocorticoid response elements, repress transcription of proinflammatory cytokines, including CXCL5, and are frequently used to treat pulmonary inflammation. In the mice lacking Bmal1, disruption of the bronchiolar clock was associated with reduced glucocorticoid receptor occupancy at the Cxcl5 gene locus and augmented histone H3K27 acetylation at the Cxcl5 promoter (a marker of active transcription), suggesting intact local clock machinery is required for optimal glucocorticoid function. Whether clock components interact directly with GR signaling, however, remains to be determined. Importantly, exogenous steroids did not suppress endotoxininduced CXCL5 or lung neutrophil counts in these mice and so could not overcome the defect in bronchiolar cell clock function. This study provides important new insights into the regulation of neutrophilic inflammation in the lung and has broader significance for understanding the pathogenesis of human airway diseases and potential therapies. In diseases where the bronchiolar clock is intact, an immediate therapeutic impact might be had by altering the timing of treatment administration, resulting in reduced dosing or increased efficacy of anti-inflammatory treatments. In diseases where the bronchiolar clock is found to be defective, this study would predict enhanced CXCL5 release, increased neutrophil recruitment and excessive inflammation that is not suppressed by exogenous steroids. A crucial question, therefore, is whether the bronchiolar clock is disrupted in individuals with chronic
neutrophilic disorders such as COPD, where there is evidence of impaired club cell function in progressive disease11. Cigarette smoke has also been shown to disrupt circadian clock gene oscillations in mouse lung tissue, and lung tissue BMAL1 levels were reduced in samples from smokers and subjects with COPD compared to healthy control samples12. Thus, if these findings are mirrored specifically in bronchiolar club cells and with other inhaled irritants, they potentially explain the ineffectiveness of current glucocorticoid-based anti-inflammatory treatments. Direct therapeutic targeting of CXCL5 could be attractive, given that Gibbs et al.10 suggest it may reduce, rather than abolish, neutrophil recruitment and thus cause less impairment of host defenses and less susceptibility to secondary infection. A note of caution, however, is that CXCL5 has roles beyond the recruitment of neutrophils that may be beneficial, for example limiting macrophage foam cell formation in atherosclerosis13. Gibbs et al.10 also show augmentation of histone H3K27 acetylation (a marker of transcriptional activation) after deletion of Bmal1, suggesting the interaction between the circadian regulation of GR recruitment and altered chromatin architecture may also have therapeutic implications for histone modification strategies. It will be necessary to extend similar studies to disorders where inflammation is less neutrophil dominant to determine local clock involvement in recruitment of other immune cell types. Exploration of other pulmonary disease models will help determine whether the observed dependence of inflammatory responses on the bronchiolar clock also applies to stimuli that induce eosinophil recruitment, such as inhaled allergens, or to lower-dose bacterial challenges, which are normally contained by alveolar macrophages without substantial neutrophil recruitment. Interestingly, the authors’ previous studies showed that temporal gating of the systemic cytokine response required intact clock function in myeloid cells, as it was absent in LysM-Bmal1−/− mice7. It is, therefore, interesting to speculate whether the development of bacteremia following airway infection and thus the progression from a local (pulmonary) to a systemic infection could result in transitioning of circadian control from epithelial to myeloid cells. The dominance of local over systemic, and epithelial over myeloid, circadian responses in the context of pneumonia and sepsis has yet to be defined. Other inflammatory disorders, such as rheumatoid arthritis, also show marked diurnal variation in symptoms, and CXCL5 and neutrophils have been implicated in pathogenesis of this disease14. In this disease context, identifying
volume 20 | number 8 | AUGUST 2014 nature medicine
news and views whether a tissue-specific clock exists in joint tissue would help to determine whether diseaseassociated impairments in clock function also occur in other tissues and disease processes. The important findings described by Gibbs et al.10 elucidate mechanisms that underpin well-recognized clinical phenotypes in patients with inflammatory lung disease2,3,15 and provide avenues to developing new therapeutic approaches to tackle neutrophil-dominant inflammation.
COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. 1. Brusselle, G.G. et al. Lancet 378, 1015–1026 (2011). 2. Barnes, P.J. Am. J. Med. 79, 5–9 (1985). 3. Calverley, P.M. et al. Thorax 58, 855–860 (2003). 4. Hastings, M.H. et al. Nat. Rev. Neurosci. 4, 649–661 (2003). 5. Lee, J.H. & Sancar, A. Proc. Natl. Acad. Sci. USA 108, 12036–12041 (2011). 6. Keller, M. et al. Proc. Natl. Acad. Sci. USA 106, 21407–21412 (2009).
7. Gibbs, J.E. et al. Proc. Natl. Acad. Sci. USA 109, 582–587 (2012). 8. Nguyen, K.D. et al. Science 341, 1483–1488 (2013). 9. Bellet, M.M. et al. Proc. Natl. Acad. Sci. USA 110, 9897–9902 (2013). 10. Gibbs, J. et al. Nat. Med. 20, 919–926 (2014). 11. Park, H.Y. et al. Am. J. Respir. Crit. Care Med. 188, 1413–1419 (2013). 12. Hwang, J.W. et al. FASEB J. 28, 176–194 (2014). 13. Rousselle, A. et al. J. Clin. Invest. 123, 1343–1347 (2013). 14. Wright, H.L. et al. Nat. Rev. Rheumatol. doi:10.1038/ nrrheum.2014.80 (10 June 2014). 15. Mackay, T.W. et al. Thorax 49, 257–262 (1994).
Lightening up a notch: Notch regulation of energy metabolism
npg
© 2014 Nature America, Inc. All rights reserved.
Thomas Gridley & Shingo Kajimura Inhibiting Notch signaling induces adipose browning, improves systemic glucose tolerance and insulin sensitivity, and suppresses weight gain in mice. A fundamental concept of obesity is the lack of balance between energy intake and energy expenditure. Currently, all US Food and Drug Administration–approved antiobesity medications aim to limit energy intake, either through appetite suppression or inhibition of lipid absorption by the intestine1. Although these medications are effective in the short term, several adverse effects, such as depression or steatorrhea, are often associated with long-term treatment, and new avenues are required. All mammals harbor two types of adipose tissue: white adipose tissue (WAT) and brown adipose tissue (BAT). WAT functions mainly in the storage of excess energy, whereas BAT specializes in dissipating energy in the form of heat and functions as a defense against cold and obesity through the BAT-specific protein uncoupling protein 1 (Ucp1). Although BAT was formerly considered to be restricted to infants and small animals, the recent discovery of BAT in adult humans led to the exciting notion that increasing energy expenditure by activating BAT thermogenesis could provide a novel approach to modulating energy balance2. Recent studies indicate that adult humans and rodents3–6 have a ‘recruitable’ form of thermogenic adipocytes, termed ‘beige adipocytes’, Thomas Gridley is at the Center for Molecular Medicine, Maine Medical Center Research Institute, Scarborough, Maine, USA. Shingo Kajimura is at the University of California–San Francisco Diabetes Center, Department of Cell and Tissue Biology, University of California– San Francisco, San Francisco, California, USA. e-mail: [email protected] or [email protected]
whose development can be induced by certain environmental stimuli (termed ‘browning’ of white fat), such as chronic cold exposure. In this issue of Nature Medicine, Bi et al.7 describe a new role for the Notch signaling pathway8, a pathway known to have multiple roles in development but not previously known to play an important role in regulating adipose browning and energy homeostasis in mammals. Thus, the authors identify a potential new therapeutic avenue for treatment of obesity by modulating this important developmental pathway. The authors initially carried out gene expression analyses of white adipose tissue depots in mice and found that an inverse correlation existed between the expression levels of Notch family receptors and targets (such as the transcriptional repressor Hes1), and the expression of the BAT-specific gene Ucp1 and two transcriptional regulators crucial for brown adipocyte development, Ppargc1a and Prdm16 (ref. 2). To assess whether there was a direct relationship between levels of Notch signaling and the differentiation of BAT, the authors selectively deleted in adipocytes either the gene encoding the Notch1 receptor or the gene encoding the primary transcriptional effector of the Notch signaling pathway, Rbpj. Mice in which Notch signaling was selectively reduced in adipocytes exhibited morphological browning of subcutaneous white fat depots, upregulated expression of BATselective genes and elevated thermogenesis. These mice also exhibited an improvement in glucose homeostasis and insulin sensitivity and had higher rates of whole-body energy
nature medicine volume 20 | number 8 | AUGUST 2014
expenditure than wild-type mice. Strikingly, these mice with selective Notch signaling reduction were protected from high-fat diet–induced obesity (Fig. 1). Bi et al.7 then analyzed mice in which Notch signaling was selectively increased in adipocytes through expression of a constitutively active form of the Notch1 receptor. Compared to wild-type mice, these mice in which Notch1 was overexpressed in adipocytes exhibited impaired glucose homeostasis and insulin sensitivity, lower metabolic rate and core body temperature, and reduced expression of the Ucp1 and Pgc1-α (encoded by Ppargc1a) proteins in WAT. The authors then further investigated the molecular mechanisms of the Notch signaling-regulated adipose browning effects. They showed that the transcriptional repressor Hes1, whose transcription is induced by Notch signal reception, directly bound in cultured mouse adipocytes to consensus Hes1 binding sites that are present in the proximal promoter regions of both human and mouse Prdm16 and Ppargc1a genes. Furthermore, they showed that the Hes1 protein could suppress transcription of the Ppargc1a gene. In turn, Notch1-deficient mouse adipocytes, or wild-type adipocytes in which Notch signaling had been suppressed by an inhibitor of γ-secretase (an enzyme complex required for proteolytic processing and signal transduction by Notch family receptors), had elevated levels of BAT-selective proteins, such as Ucp1, Pgc1-α and Prdm16, and increased cellular respiration. These results suggest that the browning effects resulting from inhibition of Notch signaling are cell autonomous. 811
© 2014 Nature America, Inc. All rights reserved.
news and views One unexpected aspect of this study is the finding that SAMHD1 does not degrade RNA in the context of RNA-DNA hybrids, which conflicts with previous work12. Therefore, how is SAMHD1’s nucleolytic activity directed to viral, rather than host, RNA? The authors speculate that SAMHD1 might bind to one or more components of the viral reverse transcription apparatus. SAMHD1’s RNase could act on either the tRNALys3 primer that initiates reverse transcription, the HIV-1 genomic RNA that serves as the template for cDNA generation or both. Northern blotting revealed no changes in the tRNALys3 levels. Using both quantitative RT-PCR and RNA-seq, Ryoo et al.5 identified HIV genomic RNA as the principal target and showed that knockdown of SAMHD1 in primary human monocyte–derived macrophages and CD4+ T cells increased the stability of HIV-1 RNA following infection. Despite the increasing evidence favoring SAMHD1’s RNase as the dominant player in HIV restriction, its dNTPase activity cannot be entirely dismissed. Recently, timecourse studies showed that HIV infection can be rescued by blocking SAMHD1 function, even though the virus was exposed to active SAMHD1 for several hours14. Such a rescue would not be predicted if SAMHD1
acts through its RNase activity alone, as this nucleolytic action would presumably be nonreversible. An intriguing possibility suggested by Ryoo et al.5 is that both enzymatic functions are important, but they operate in different cellular environments. SAMHD1’s dNTPase activity depends upon tetramerization of the enzyme, which occurs in a dGTP-dependent manner15,16. Conversely, the RNase activity of SAMHD1 does not depend on either dGTP binding or tetramer formation. It is possible that that tetramerization impairs RNase activity, but this has not been investigated to date. The levels of intracellular dGTP could therefore dictate which enzymatic function of SAMHD1 predominates. In this scenario, dGTP serves as both a regulator and a substrate of SAMHD1. The fact that dGTP levels are generally low in resting CD4+ T cells and macrophages suggests that the RNase activity may be dominant in these cells. However, slight changes in activation state could boost dGTP levels sufficiently to stimulate tetramerization and dNTPase activation. This dGTP-regulated dual-enzyme model for SAMHD1 action is attractive, but further experiments are needed to resolve whether one or both of SAMHD1’s enzymatic actions are in play during HIV
restriction. A more complete understanding of how SAMHD1 restricts HIV as well as of the counterattack leveled by the lentiviral Vpx proteins will provide new insights into viral pathogenesis and inform new strategies for preserving SAMHD1 restriction activity. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. 1. Hrecka, K. et al. Nature 474, 658–661 (2011). 2. Laguette, N. et al. Nature 474, 654–657 (2011). 3. Baldauf, H.M. et al. Nat. Med. 18, 1682–1687 (2012). 4. Descours, B. et al. Retrovirology 9, 87 (2012). 5. Ryoo, J. et al. Nat. Med. 20, 936–941 (2014). 6. Goldstone, D.C. et al. Nature 480, 379–382 (2011). 7. Powell, R.D., Holland, P.J., Hollis, T. & Perrino, F.W. J. Biol. Chem. 286, 43596–43600 (2011). 8. Lahouassa, H. et al. Nat. Immunol. 13, 223–228 (2012). 9. Cribier, A., Descours, B., Valadão, A.L., Laguette, N. & Benkirane, M. Cell Reports 3, 1036–1043 (2013). 10. White, T.E. et al. Cell Host Microbe 13, 441–451 (2013). 11. Welbourn, S., Dutta, S.M., Semmes, O.J. & Strebel, K. J. Virol. 87, 11516–11524 (2013). 12. Beloglazova, N. et al. J. Biol. Chem. 288, 8101–8110 (2013). 13. Crow, Y.J. & Livingston, J.H. Dev. Med. Child Neurol. 50, 410–416 (2008). 14. Hofmann, H. et al. J. Virol. 87, 11741–11750 (2013). 15. Ji, X. et al. Nat. Struct. Mol. Biol. 20, 1304–1309 (2013). 16. Yan, J. et al. J. Biol. Chem. 288, 10406–10417 (2013).
A local circadian clock calls time on lung inflammation A A Roger Thompson, Sarah R Walmsley & Moira K B Whyte Inflammatory diseases typically display circadian variation in symptom severity. A new study in mice shows how a pulmonary epithelial cell clock controls neutrophil recruitment to the lungs and provides insight into interactions between local and systemic circadian clocks. Controlling excessive neutrophilic inflammation remains a major unmet clinical need in inflammatory lung diseases, including chronic obstructive pulmonary disease (COPD) and neutrophilic asthma, in which the essential antimicrobial functions of the neutrophil may be superseded by persistent inflammation and tissue damage. This excessive inflammation often coexists with persistence of bacteria in the lower airways, implying ineffective host defenses despite the presence of inflammation1. A.A. Roger Thompson, Sarah R. Walmsley & Moira K.B. Whyte are in the Academic Unit of Respiratory Medicine, Department of Infection and Immunity, The Medical School, University of Sheffield, Sheffield, UK. e-mail: [email protected]
Understanding the mechanisms responsible for recruiting neutrophils to the lungs and their subsequent activation and host-defense function is crucial to direct therapeutic interventions in these conditions. Inflammatory diseases are often characterized by diurnal variation in symptom severity, such as prominent early morning symptoms of wheeze in poorly controlled asthma and worsening of lung function in COPD2,3. A circadian clock, set by a central timekeeper in the brain’s suprachiasmatic nucleus, is responsive to external signals and can entrain peripheral rhythms via both endocrine and neural pathways, thus setting a circadian rhythm that affects the body as a whole. Peripheral tissues and cells also possess an intrinsic circadian cycle, employing the same genetic machinery
nature medicine volume 20 | number 8 | AUGUST 2014
as the central clock and involving feedback loop–regulated transcription and translation of clock genes including CLOCK, ARNTL (here called BMAL1), PER1–PER3 and CRY1 and CRY2 (ref. 4). In recent years, circadian rhythms in immune cells have been shown to affect systemic cytokine responses, which are largely of myeloid cell origin, with consequences for cellular trafficking and host responses to pathogens5–9. Although this clock machinery functions autonomously in isolated immune cells, there are also interactions between central and peripheral rhythms that affect the circadian regulation of inflammatory responses in individual tissues and that still remain unclear. In this issue of Nature Medicine, Gibbs et al.10 uncover a new mechanism governing circadian neutrophil 809
news and views
Tissue homeostasis
Debbie Maizels/ Nature Publishing Group
CXCL5
GR Club cell
Infectious/inflammatory stimuli (S. pneumoniae infection or LPS challenge)
Ciliated airway
CXCL5 CXCL5 CXCL5
Local clock disruption (Bmal1 deletion in bronchiolar club cells)
Increased neutrophil recruitment
Excess local inflammation; inefficient antibacterial response
npg
© 2014 Nature America, Inc. All rights reserved.
Figure 1 The pulmonary clock and inflammatory responses. Gibbs et al.10 show that inhalation of infectious (S. pneumoniae) or inflammatory (lipopolysaccharide (LPS) challenge) stimuli in mice results in release of CXCL5 from bronchiolar club cells within the airways and induction of neutrophilic inflammation, which peaks at the start of rest phase (dawn). Endogenous glucocorticoids rhythmically suppress Cxcl5 transcription in bronchiolar cells, reducing the amounts of secreted CXCL5 and subsequent neutrophilmediated inflammation. This regulation is lost upon local failure of the pulmonary clock through deletion of the clock gene Bmal1 (and potentially through alteration of clock gene function by environmental stimuli, or a naturally occurring disease mutation) in bronchiolar cells and cannot be overcome by exogenous glucocorticoid treatment.
recruitment to the lungs in response to local bacterial stimuli that is mediated by the circadian regulation of the chemokine CXCL5 in pulmonary bronchiolar cells. Disruption of this lung clock increased recruitment of neutrophils to the lung without improving bacterial clearance and, importantly, promoted resistance to anti-inflammatory glucocorticoids. Gibbs et al.10 first found in mice that local pulmonary responses to inhaled endotoxin and pneumococcal infection are under circadian regulation: peak inflammation was seen after endotoxin challenge was given at ‘dawn’ (the start of rest phase in mice). This is in marked contrast to circadian variation in systemic cytokine responses, measured after intraperitoneal endotoxin challenge, in which case maximal cytokine release was observed when mice were challenged at ‘dusk’ (start of active phase)7. Using mice with conditional deletion of the clock gene Bmal1 in either myeloid cells (LysM-Bmal1−/−) or bronchiolar club cells (Ccsp-Bmal1−/−) to specifically disrupt the circadian clock in these populations, they found that the local pulmonary inflammation was regulated by circadian clock function in bronchiolar club cells (formerly known as Clara cells) but not in alveolar macrophages or other myeloid cells10. Furthermore, the authors also observed increased pulmonary neutrophilia in mice lacking club cell Bmal1 after an intratracheal challenge with Streptococcus pneumoniae, although, interestingly, the increased inflammation did not improve the clearance of bacteria10. Gibbs et al.10 found that expression of a number of inflammatory mediators was increased in mice lacking Bmal1 in bronchiolar club cells compared with wild-type mice after pulmonary endotoxin challenge, notably the chemokine CXCL5, a known neutrophil chemoattractant. In wild-type mice, prolonged secretion of CXCL5 from bronchiolar cells into the airways in response to endotoxin coincided with neutrophil persistence in the lungs. 810
In Cxcl5-deficient mice, neutrophils were still recruited, but the diurnal variations in neutrophilic response were abolished (Fig. 1). The authors gained mechanistic insights into the rhythmic regulation of CXCL5 through investigation of the interplay between glucocorticoidmediated repression of Cxcl5 transcription and local clock function. Glucocorticoids, via activated glucocorticoid receptor (GR) binding to glucocorticoid response elements, repress transcription of proinflammatory cytokines, including CXCL5, and are frequently used to treat pulmonary inflammation. In the mice lacking Bmal1, disruption of the bronchiolar clock was associated with reduced glucocorticoid receptor occupancy at the Cxcl5 gene locus and augmented histone H3K27 acetylation at the Cxcl5 promoter (a marker of active transcription), suggesting intact local clock machinery is required for optimal glucocorticoid function. Whether clock components interact directly with GR signaling, however, remains to be determined. Importantly, exogenous steroids did not suppress endotoxininduced CXCL5 or lung neutrophil counts in these mice and so could not overcome the defect in bronchiolar cell clock function. This study provides important new insights into the regulation of neutrophilic inflammation in the lung and has broader significance for understanding the pathogenesis of human airway diseases and potential therapies. In diseases where the bronchiolar clock is intact, an immediate therapeutic impact might be had by altering the timing of treatment administration, resulting in reduced dosing or increased efficacy of anti-inflammatory treatments. In diseases where the bronchiolar clock is found to be defective, this study would predict enhanced CXCL5 release, increased neutrophil recruitment and excessive inflammation that is not suppressed by exogenous steroids. A crucial question, therefore, is whether the bronchiolar clock is disrupted in individuals with chronic
neutrophilic disorders such as COPD, where there is evidence of impaired club cell function in progressive disease11. Cigarette smoke has also been shown to disrupt circadian clock gene oscillations in mouse lung tissue, and lung tissue BMAL1 levels were reduced in samples from smokers and subjects with COPD compared to healthy control samples12. Thus, if these findings are mirrored specifically in bronchiolar club cells and with other inhaled irritants, they potentially explain the ineffectiveness of current glucocorticoid-based anti-inflammatory treatments. Direct therapeutic targeting of CXCL5 could be attractive, given that Gibbs et al.10 suggest it may reduce, rather than abolish, neutrophil recruitment and thus cause less impairment of host defenses and less susceptibility to secondary infection. A note of caution, however, is that CXCL5 has roles beyond the recruitment of neutrophils that may be beneficial, for example limiting macrophage foam cell formation in atherosclerosis13. Gibbs et al.10 also show augmentation of histone H3K27 acetylation (a marker of transcriptional activation) after deletion of Bmal1, suggesting the interaction between the circadian regulation of GR recruitment and altered chromatin architecture may also have therapeutic implications for histone modification strategies. It will be necessary to extend similar studies to disorders where inflammation is less neutrophil dominant to determine local clock involvement in recruitment of other immune cell types. Exploration of other pulmonary disease models will help determine whether the observed dependence of inflammatory responses on the bronchiolar clock also applies to stimuli that induce eosinophil recruitment, such as inhaled allergens, or to lower-dose bacterial challenges, which are normally contained by alveolar macrophages without substantial neutrophil recruitment. Interestingly, the authors’ previous studies showed that temporal gating of the systemic cytokine response required intact clock function in myeloid cells, as it was absent in LysM-Bmal1−/− mice7. It is, therefore, interesting to speculate whether the development of bacteremia following airway infection and thus the progression from a local (pulmonary) to a systemic infection could result in transitioning of circadian control from epithelial to myeloid cells. The dominance of local over systemic, and epithelial over myeloid, circadian responses in the context of pneumonia and sepsis has yet to be defined. Other inflammatory disorders, such as rheumatoid arthritis, also show marked diurnal variation in symptoms, and CXCL5 and neutrophils have been implicated in pathogenesis of this disease14. In this disease context, identifying
volume 20 | number 8 | AUGUST 2014 nature medicine
news and views whether a tissue-specific clock exists in joint tissue would help to determine whether diseaseassociated impairments in clock function also occur in other tissues and disease processes. The important findings described by Gibbs et al.10 elucidate mechanisms that underpin well-recognized clinical phenotypes in patients with inflammatory lung disease2,3,15 and provide avenues to developing new therapeutic approaches to tackle neutrophil-dominant inflammation.
COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. 1. Brusselle, G.G. et al. Lancet 378, 1015–1026 (2011). 2. Barnes, P.J. Am. J. Med. 79, 5–9 (1985). 3. Calverley, P.M. et al. Thorax 58, 855–860 (2003). 4. Hastings, M.H. et al. Nat. Rev. Neurosci. 4, 649–661 (2003). 5. Lee, J.H. & Sancar, A. Proc. Natl. Acad. Sci. USA 108, 12036–12041 (2011). 6. Keller, M. et al. Proc. Natl. Acad. Sci. USA 106, 21407–21412 (2009).
7. Gibbs, J.E. et al. Proc. Natl. Acad. Sci. USA 109, 582–587 (2012). 8. Nguyen, K.D. et al. Science 341, 1483–1488 (2013). 9. Bellet, M.M. et al. Proc. Natl. Acad. Sci. USA 110, 9897–9902 (2013). 10. Gibbs, J. et al. Nat. Med. 20, 919–926 (2014). 11. Park, H.Y. et al. Am. J. Respir. Crit. Care Med. 188, 1413–1419 (2013). 12. Hwang, J.W. et al. FASEB J. 28, 176–194 (2014). 13. Rousselle, A. et al. J. Clin. Invest. 123, 1343–1347 (2013). 14. Wright, H.L. et al. Nat. Rev. Rheumatol. doi:10.1038/ nrrheum.2014.80 (10 June 2014). 15. Mackay, T.W. et al. Thorax 49, 257–262 (1994).
Lightening up a notch: Notch regulation of energy metabolism
npg
© 2014 Nature America, Inc. All rights reserved.
Thomas Gridley & Shingo Kajimura Inhibiting Notch signaling induces adipose browning, improves systemic glucose tolerance and insulin sensitivity, and suppresses weight gain in mice. A fundamental concept of obesity is the lack of balance between energy intake and energy expenditure. Currently, all US Food and Drug Administration–approved antiobesity medications aim to limit energy intake, either through appetite suppression or inhibition of lipid absorption by the intestine1. Although these medications are effective in the short term, several adverse effects, such as depression or steatorrhea, are often associated with long-term treatment, and new avenues are required. All mammals harbor two types of adipose tissue: white adipose tissue (WAT) and brown adipose tissue (BAT). WAT functions mainly in the storage of excess energy, whereas BAT specializes in dissipating energy in the form of heat and functions as a defense against cold and obesity through the BAT-specific protein uncoupling protein 1 (Ucp1). Although BAT was formerly considered to be restricted to infants and small animals, the recent discovery of BAT in adult humans led to the exciting notion that increasing energy expenditure by activating BAT thermogenesis could provide a novel approach to modulating energy balance2. Recent studies indicate that adult humans and rodents3–6 have a ‘recruitable’ form of thermogenic adipocytes, termed ‘beige adipocytes’, Thomas Gridley is at the Center for Molecular Medicine, Maine Medical Center Research Institute, Scarborough, Maine, USA. Shingo Kajimura is at the University of California–San Francisco Diabetes Center, Department of Cell and Tissue Biology, University of California– San Francisco, San Francisco, California, USA. e-mail: [email protected] or [email protected]
whose development can be induced by certain environmental stimuli (termed ‘browning’ of white fat), such as chronic cold exposure. In this issue of Nature Medicine, Bi et al.7 describe a new role for the Notch signaling pathway8, a pathway known to have multiple roles in development but not previously known to play an important role in regulating adipose browning and energy homeostasis in mammals. Thus, the authors identify a potential new therapeutic avenue for treatment of obesity by modulating this important developmental pathway. The authors initially carried out gene expression analyses of white adipose tissue depots in mice and found that an inverse correlation existed between the expression levels of Notch family receptors and targets (such as the transcriptional repressor Hes1), and the expression of the BAT-specific gene Ucp1 and two transcriptional regulators crucial for brown adipocyte development, Ppargc1a and Prdm16 (ref. 2). To assess whether there was a direct relationship between levels of Notch signaling and the differentiation of BAT, the authors selectively deleted in adipocytes either the gene encoding the Notch1 receptor or the gene encoding the primary transcriptional effector of the Notch signaling pathway, Rbpj. Mice in which Notch signaling was selectively reduced in adipocytes exhibited morphological browning of subcutaneous white fat depots, upregulated expression of BATselective genes and elevated thermogenesis. These mice also exhibited an improvement in glucose homeostasis and insulin sensitivity and had higher rates of whole-body energy
nature medicine volume 20 | number 8 | AUGUST 2014
expenditure than wild-type mice. Strikingly, these mice with selective Notch signaling reduction were protected from high-fat diet–induced obesity (Fig. 1). Bi et al.7 then analyzed mice in which Notch signaling was selectively increased in adipocytes through expression of a constitutively active form of the Notch1 receptor. Compared to wild-type mice, these mice in which Notch1 was overexpressed in adipocytes exhibited impaired glucose homeostasis and insulin sensitivity, lower metabolic rate and core body temperature, and reduced expression of the Ucp1 and Pgc1-α (encoded by Ppargc1a) proteins in WAT. The authors then further investigated the molecular mechanisms of the Notch signaling-regulated adipose browning effects. They showed that the transcriptional repressor Hes1, whose transcription is induced by Notch signal reception, directly bound in cultured mouse adipocytes to consensus Hes1 binding sites that are present in the proximal promoter regions of both human and mouse Prdm16 and Ppargc1a genes. Furthermore, they showed that the Hes1 protein could suppress transcription of the Ppargc1a gene. In turn, Notch1-deficient mouse adipocytes, or wild-type adipocytes in which Notch signaling had been suppressed by an inhibitor of γ-secretase (an enzyme complex required for proteolytic processing and signal transduction by Notch family receptors), had elevated levels of BAT-selective proteins, such as Ucp1, Pgc1-α and Prdm16, and increased cellular respiration. These results suggest that the browning effects resulting from inhibition of Notch signaling are cell autonomous. 811
