- Email: [email protected]
Exonuclease TREX1 also Has a Sweet Tooth
Immunity
Previews creates ER stress that mimicks ‘‘terminal’’ and not ‘‘adaptive’’ UPR, activating the RIDD mode of IRE1a. Because this mode of activity arises even at low multiplicity of infection (Bronner et al., 2015), yet without activation of cell death pathways during virulent Brucella infection (Celli and Tsolis, 2015), perhaps Brucella infection changes the quality of the host cell UPR. What protein interactions underlie such quality changes still remains to be determined. The involvement of IRE1a in activating NLRP3 upstream of the caspase-2 and Bid cleavages provides a novel functional connection between the ER and mitochondria. Such a functional interaction is often thought of in terms of mitochondrial-associated-membranes (MAMs), which are areas of contact between the two organelles. In fact, Zhou et al. showed that NLRP3 is associated with the ER and moves to mitochondria (Zhou et al., 2011). Perhaps the relocation of NLRP3 is a
manifestation of the function of MAMs in either sensing or causing mitochondrial dysfunction? While Brucella abortus infection selectively activates the IRE1a pathway of the UPR, the data of Bronner et al. bring to mind the possible activation of inflammasomes via one of the other UPR sensors. The PERK pathway of the UPR has already been invoked through its transcription factor, ATF5 (Oslowski et al., 2012), which promotes inflammasome activation by upregulating TXNIP expression. Because different inflammasomes are employed in response to distinct stimuli, it is likely that infections by virulent bacteria activate multiple inflammasomes and by more than one UPR pathway.
Bronner, D., Abuaita, B., Chen, X., Fitzgerald, K., Nun˜ez, G., He, Y., Yin, X.-M., and O’Riordan, M. (2015). Immunity 43, this issue, 451–462. Celli, J., and Tsolis, R.M. (2015). Nat. Rev. Microbiol. 13, 71–82. Hollien, J., and Weissman, J.S. (2006). Science 313, 104–107. Janssens, S., Pulendran, B., and Lambrecht, B.N. (2014). Nat. Immunol. 15, 910–919. Lamkanfi, M., and Dixit, V.M. (2014). Cell 157, 1013–1022. Lerner, A.G., Upton, J.P., Praveen, P.V., Ghosh, R., Nakagawa, Y., Igbaria, A., Shen, S., Nguyen, V., Backes, B.J., Heiman, M., et al. (2012). Cell Metab. 16, 250–264. Menu, P., Mayor, A., Zhou, R., Tardivel, A., Ichijo, H., Mori, K., and Tschopp, J. (2012). Cell Death Dis. 3, e261. Oslowski, C.M., Hara, T., O’Sullivan-Murphy, B., Kanekura, K., Lu, S., Hara, M., Ishigaki, S., Zhu, L.J., Hayashi, E., Hui, S.T., et al. (2012). Cell Metab. 16, 265–273.
REFERENCES Arellano-Reynoso, B., Dı´az-Aparicio, E., Leal-Herna´ndez, M., Herna´ndez, L., and Gorvel, J.P. (2004). Vet. Microbiol. 98, 307–312.
Zhou, R., Yazdi, A.S., Menu, P., and Tschopp, J. (2011). Nature 469, 221–225.
Exonuclease TREX1 also Has a Sweet Tooth Winfried Barchet1,2,* and Thomas Zillinger1 1Institute
of Clinical Chemistry and Clinical Pharmacology Centre for Infection Research (DZIF) University Hospital, University of Bonn, 53127 Bonn, Germany *Correspondence: [email protected] http://dx.doi.org/10.1016/j.immuni.2015.08.026 2German
TREX1 regulates innate immune responses by counteracting DNA accumulation in the cytosol. In this issue of Immunity, Hasan et al. (2015) show that TREX1 also safeguards the cell against free glycan build-up in the endoplasmic reticulum, thereby preventing glycan-induced inflammation. Nucleic acid sensors of the innate immune system alert to the presence of pathogens, in particular of viruses. In response, type I interferon (IFN), the key antiviral cytokine, as well as other pro-inflammatory mediators are secreted, which in turn induce the transcription of multiple interferon inducible genes (ISGs). Due to genetic predisposition, some individuals are prone to erroneous or excessive reactivity to their own nucleic acids, which leads to autoimmune pathologies characterized by chronic ISG signatures (Crow and Manel, 2015). Recent
years have seen immense progress in unraveling nucleic acid receptors and their signaling pathways, and many of the gene variants associated with ISG-driven autoimmune syndromes have been identified (Crow and Manel, 2015). In this context, mutations in three prime repair exonuclease 1 (TREX1) were the first identified monogenetic cause of severe lupus (Lee-Kirsch et al., 2007) as well as of the rare autoimmune diseases Aicardi Goutie`res syndrome and retinal vasculopathy with cerebral leukodystrophy (RVCL) (Richards et al., 2007).
TREX1 functions as one of the three major DNA-degrading nucleases and has been shown to prevent the build-up of immune stimulatory DNA in the cytosol (Stetson et al., 2008), thus acting as a critical negative regulator of the cGAMP synthase (cGAS) signaling pathway (Gray et al., 2015). Therefore, compared to other monogenetic and complex genetic causes of autoimmune inflammation, the physiological role of TREX1 and the mechanisms leading to immune pathology are considered comparably well understood.
Immunity 43, September 15, 2015 ª2015 Elsevier Inc. 411
Immunity
Previews
Figure 1. The Dual Immune-Regulatory Role of Exonuclease TREX1 Endogenous or pathogen-derived DNA that enters the cytosol is promptly degraded by TREX1. Patients with DNase-defective TREX1 develop autoimmune disease, including familial chilblain lupus (FCL) and Aicardi Goutie`res syndrome (AGS) that are characterized by cytosolic DNA accumulation, which in turn leads to chronic activation of the cGAMP synthase (cGAS) signaling pathway, resulting in pathologically elevated type I IFN. With its C terminus, TREX1 spans the ER membrane and interacts with the oligosaccharyl transferase (OST) complex. Here, TREX1 quality controls N-linked protein glycosylation by preventing erroneous hydrolysis of the lipid-linked oligosaccharide (LLO) precursor. In patients suffering from retinal vasculopathy with cerebral leukodystrophy (RVCL), TREX1 frame-shift mutations exclusively affect the latter function. RVCL pathology is characterized by the build-up of free glycans, which leads to the activation of an unknown innate immune receptor that promotes excessive production of the inflammatory chemokine CXCL10 in a tank binding kinase 1 (TBK1)-dependent fashion. By replicating the most frequent nuclease or C-terminal mutations found in humans, two newly generated genetically targeted mouse models now allow for the differential investigation of both TREX1 pathway deficiencies.
However, in this issue of Immunity, Hasan et al. (2015) report a previously unrecognized function of TREX1 as a guarantor of N-linked protein glycosylation. The group has shown that the C terminus of TREX1 traverses the endoplasmic reticulum (ER) membrane and directly interacts with components of the oligosaccharyl transferase (OST) complex that transfers complex sugars (glycans) from a lipid-linked oligosaccharide (LLO) precursor onto nascent protein strands in the ER (Figure 1). Both in isolated cells and in various tissues in vivo, Hasan et al. (2015) have demonstrated that, in the absence of TREX1, erroneous
hydrolysis of LLO leads to a several fold increase in free glycans, a mixture of LLO hydrolysis and degradation products. Overexpression of nucleasemutated full-length TREX1 or of the C terminus alone are both able to normalize free glycan amounts, indicating that not the cytosolic exonuclease but rather the ER-localized portion of TREX1 is responsible for regulating N-glycosylation. The group then has proceeded to show that free glycans purified from TREX1-deficient cells were immune stimulatory for macrophages and upregulated the inflammatory chemokine CXCL10 and other ISGs but not type I IFN. These find-
412 Immunity 43, September 15, 2015 ª2015 Elsevier Inc.
ings were in line with the group’s previous observation of interferon-independent activation of antiviral genes (Hasan et al., 2013). The same CXCL10-dominated ISG signature in the absence of type I IFN is also observed in patients suffering from RVCL. In RVCL patients, TREX1 lacks the ER-localized C terminus as a consequence of frame-shift mutations. The authors have shown that such mutations do not affect TREX1 nuclease activity but exclusively promote the accumulation of free glycans. Hasan et al. (2015) hypothesize that this is due to improper functioning of the OST complex and indeed find that in cells derived from RVCL patients, the OST inhibitor aclacinomycin (Acm) is able to efficiently reduce both free glycan accretion as well as the glycan-induced CXCL10 immune signature. These findings also hold true in vivo. Due to chronic inflammation in particular of the heart and liver, TREX1-deficient mice show limited viability. Intriguingly, Acm treatment of TREX1-deficient mice reduced the ISG signature, normalized liver and spleen morphology, and improved viability. In addition, a genetically targeted mouse model has recently been generated that recapitulates the most frequent human TREX1 mutation leading to the production of nucleaseinactive TREX1 protein (Grieves et al., 2015). However, in this TREX1 variant, the glycan-regulating function of TREX1 in the ER presumably remains intact, and, along with a milder course and a much later onset of autoimmune symptoms, these mice show improved viability compared to fully TREX1-deficient mice (Grieves et al., 2015). Altogether, this indicates that glycan regulation constitutes a key function of TREX1 and that deregulation of this pathway notably contributes to the severe immune pathology of TREX1-deficient mice, as well as to that of human RVCL patients. In order to study TREX1 glycan regulation independent of its nuclease function, Hasan et al. (2015) generated a new mouse model of RVCL by placing the human TREX1-V235fs mutant allele under the control of the endogenous murine promoter. Primary myeloid cells of these mice that are heterozygous carriers of this allele show elevated glycans, as well as chronic CXCL10 and other ISG
Immunity
Previews expression, with a more pronounced phenotype in TREX1-V235fs homozygotes. Although these mice showed normal viability and did not develop overt tissue inflammation, older mice developed far more autoantibodies than WT mice of the same age. Future studies with this model will reveal whether OST inhibition by Acm is able to suppress autoantibody formation. Given that TREX1 expression varies widely between cell types and activation states, the exact role of TREX-mediated regulation of N-glycosylation bears further investigation. Is TREX1 required to maintain general glycan homeostasis, or might it play a role in compensating for scenarios of cellular distress, or of infections? Although the innate immune sensing of fungal and bacterial glycans via toll-like receptors TLRs 2 and 4 or the NOD-like receptors is well understood, the data presented by Hasan et al. (2015) for the first time implicate endogenous free glycans as innate immune stimuli. However, the precise nature of the free self-glycan(s) that demonstrate immune activity remains to be established. The authors also do not identify which innate immune receptor(s) initiate the response to free glycans. This is not surprising given the sheer multitude of complex sugar-sensing immune receptors (Rabinovich et al., 2012). On the other hand, the authors’ finding that free glycan sensing is independent of cGAS and
stimulator of interferon genes (STING) yet dependent on tank-binding kinase 1 (TBK1) dramatically narrows down the options, because this should already exclude a number of candidates, including all currently known C-type lectin receptors. Our growing mechanistic understanding of ISG-driven autoinflammatory conditions has fueled the interest to therapeutically antagonize deregulated nucleic acid sensing pathways (Junt and Barchet, 2015). Hasan et al. (2015) now highlight a role for the innate immune sensing of endogenous complex sugars and establish that drugs that act by modulating free glycans, such as Acm, might hold therapeutic promise for RVCL and possibly also for patients suffering from other autoimmune disorders. To derive meaning from complexity, the reductionist approach to science generally serves us quite well. Yet once an immune function is paradigmatically established—as is the case for the TREX1 nuclease activity—it is tempting to assume that all observed defects are caused by variations of the same pathomechanism. This might at times lead us to behave like the proverbial blind men who try to describe an elephant by focusing exclusively on the prominent feature that is encountered first. In their seminal paper, Hasan et al. (2015) boldly approach from a different angle, discover TREX1’s sweet tooth, and thus demon-
strate that also in the case of TREX1, we are only beginning to see the whole elephant. REFERENCES Crow, Y.J., and Manel, N. (2015). Nat. Rev. Immunol. 15, 429–440. Gray, E.E., Treuting, P.M., Woodward, J.J., and Stetson, D.B. (2015). J. Immunol. 195, 1939– 1943. Grieves, J.L., Fye, J.M., Harvey, S., Grayson, J.M., Hollis, T., and Perrino, F.W. (2015). Proc. Natl. Acad. Sci. USA 112, 5117–5122. Hasan, M., Koch, J., Rakheja, D., Pattnaik, A.K., Brugarolas, J., Dozmorov, I., Levine, B., Wakeland, E.K., Lee-Kirsch, M.A., and Yan, N. (2013). Nat. Immunol. 14, 61–71. Hasan, M., Fermaintt, C.S., Gao, N., Sakai, T., Miyazaki, T., Jiang, S., Li, Q.-Z., Atkinson, J.P., Morse, H.C., III, Lehrman, M.A., and Yan, N. (2015). Immunity 43, this issue, 463–474. Junt, T., and Barchet, W. (2015). Nat. Rev. Immunol. 15, 529–544. Lee-Kirsch, M.A., Gong, M., Chowdhury, D., Senenko, L., Engel, K., Lee, Y.-A., de Silva, U., Bailey, S.L., Witte, T., Vyse, T.J., et al. (2007). Nat. Genet. 39, 1065–1067. Rabinovich, G.A., van Kooyk, Y., and Cobb, B.A. (2012). Ann. N Y Acad. Sci. 1253, 1–15. Richards, A., van den Maagdenberg, A.M., Jen, J.C., Kavanagh, D., Bertram, P., Spitzer, D., Liszewski, M.K., Barilla-Labarca, M.L., Terwindt, G.M., Kasai, Y., et al. (2007). Nat. Genet. 39, 1068–1070. Stetson, D.B., Ko, J.S., Heidmann, T., and Medzhitov, R. (2008). Cell 134, 587–598.
Immunity 43, September 15, 2015 ª2015 Elsevier Inc. 413
Previews creates ER stress that mimicks ‘‘terminal’’ and not ‘‘adaptive’’ UPR, activating the RIDD mode of IRE1a. Because this mode of activity arises even at low multiplicity of infection (Bronner et al., 2015), yet without activation of cell death pathways during virulent Brucella infection (Celli and Tsolis, 2015), perhaps Brucella infection changes the quality of the host cell UPR. What protein interactions underlie such quality changes still remains to be determined. The involvement of IRE1a in activating NLRP3 upstream of the caspase-2 and Bid cleavages provides a novel functional connection between the ER and mitochondria. Such a functional interaction is often thought of in terms of mitochondrial-associated-membranes (MAMs), which are areas of contact between the two organelles. In fact, Zhou et al. showed that NLRP3 is associated with the ER and moves to mitochondria (Zhou et al., 2011). Perhaps the relocation of NLRP3 is a
manifestation of the function of MAMs in either sensing or causing mitochondrial dysfunction? While Brucella abortus infection selectively activates the IRE1a pathway of the UPR, the data of Bronner et al. bring to mind the possible activation of inflammasomes via one of the other UPR sensors. The PERK pathway of the UPR has already been invoked through its transcription factor, ATF5 (Oslowski et al., 2012), which promotes inflammasome activation by upregulating TXNIP expression. Because different inflammasomes are employed in response to distinct stimuli, it is likely that infections by virulent bacteria activate multiple inflammasomes and by more than one UPR pathway.
Bronner, D., Abuaita, B., Chen, X., Fitzgerald, K., Nun˜ez, G., He, Y., Yin, X.-M., and O’Riordan, M. (2015). Immunity 43, this issue, 451–462. Celli, J., and Tsolis, R.M. (2015). Nat. Rev. Microbiol. 13, 71–82. Hollien, J., and Weissman, J.S. (2006). Science 313, 104–107. Janssens, S., Pulendran, B., and Lambrecht, B.N. (2014). Nat. Immunol. 15, 910–919. Lamkanfi, M., and Dixit, V.M. (2014). Cell 157, 1013–1022. Lerner, A.G., Upton, J.P., Praveen, P.V., Ghosh, R., Nakagawa, Y., Igbaria, A., Shen, S., Nguyen, V., Backes, B.J., Heiman, M., et al. (2012). Cell Metab. 16, 250–264. Menu, P., Mayor, A., Zhou, R., Tardivel, A., Ichijo, H., Mori, K., and Tschopp, J. (2012). Cell Death Dis. 3, e261. Oslowski, C.M., Hara, T., O’Sullivan-Murphy, B., Kanekura, K., Lu, S., Hara, M., Ishigaki, S., Zhu, L.J., Hayashi, E., Hui, S.T., et al. (2012). Cell Metab. 16, 265–273.
REFERENCES Arellano-Reynoso, B., Dı´az-Aparicio, E., Leal-Herna´ndez, M., Herna´ndez, L., and Gorvel, J.P. (2004). Vet. Microbiol. 98, 307–312.
Zhou, R., Yazdi, A.S., Menu, P., and Tschopp, J. (2011). Nature 469, 221–225.
Exonuclease TREX1 also Has a Sweet Tooth Winfried Barchet1,2,* and Thomas Zillinger1 1Institute
of Clinical Chemistry and Clinical Pharmacology Centre for Infection Research (DZIF) University Hospital, University of Bonn, 53127 Bonn, Germany *Correspondence: [email protected] http://dx.doi.org/10.1016/j.immuni.2015.08.026 2German
TREX1 regulates innate immune responses by counteracting DNA accumulation in the cytosol. In this issue of Immunity, Hasan et al. (2015) show that TREX1 also safeguards the cell against free glycan build-up in the endoplasmic reticulum, thereby preventing glycan-induced inflammation. Nucleic acid sensors of the innate immune system alert to the presence of pathogens, in particular of viruses. In response, type I interferon (IFN), the key antiviral cytokine, as well as other pro-inflammatory mediators are secreted, which in turn induce the transcription of multiple interferon inducible genes (ISGs). Due to genetic predisposition, some individuals are prone to erroneous or excessive reactivity to their own nucleic acids, which leads to autoimmune pathologies characterized by chronic ISG signatures (Crow and Manel, 2015). Recent
years have seen immense progress in unraveling nucleic acid receptors and their signaling pathways, and many of the gene variants associated with ISG-driven autoimmune syndromes have been identified (Crow and Manel, 2015). In this context, mutations in three prime repair exonuclease 1 (TREX1) were the first identified monogenetic cause of severe lupus (Lee-Kirsch et al., 2007) as well as of the rare autoimmune diseases Aicardi Goutie`res syndrome and retinal vasculopathy with cerebral leukodystrophy (RVCL) (Richards et al., 2007).
TREX1 functions as one of the three major DNA-degrading nucleases and has been shown to prevent the build-up of immune stimulatory DNA in the cytosol (Stetson et al., 2008), thus acting as a critical negative regulator of the cGAMP synthase (cGAS) signaling pathway (Gray et al., 2015). Therefore, compared to other monogenetic and complex genetic causes of autoimmune inflammation, the physiological role of TREX1 and the mechanisms leading to immune pathology are considered comparably well understood.
Immunity 43, September 15, 2015 ª2015 Elsevier Inc. 411
Immunity
Previews
Figure 1. The Dual Immune-Regulatory Role of Exonuclease TREX1 Endogenous or pathogen-derived DNA that enters the cytosol is promptly degraded by TREX1. Patients with DNase-defective TREX1 develop autoimmune disease, including familial chilblain lupus (FCL) and Aicardi Goutie`res syndrome (AGS) that are characterized by cytosolic DNA accumulation, which in turn leads to chronic activation of the cGAMP synthase (cGAS) signaling pathway, resulting in pathologically elevated type I IFN. With its C terminus, TREX1 spans the ER membrane and interacts with the oligosaccharyl transferase (OST) complex. Here, TREX1 quality controls N-linked protein glycosylation by preventing erroneous hydrolysis of the lipid-linked oligosaccharide (LLO) precursor. In patients suffering from retinal vasculopathy with cerebral leukodystrophy (RVCL), TREX1 frame-shift mutations exclusively affect the latter function. RVCL pathology is characterized by the build-up of free glycans, which leads to the activation of an unknown innate immune receptor that promotes excessive production of the inflammatory chemokine CXCL10 in a tank binding kinase 1 (TBK1)-dependent fashion. By replicating the most frequent nuclease or C-terminal mutations found in humans, two newly generated genetically targeted mouse models now allow for the differential investigation of both TREX1 pathway deficiencies.
However, in this issue of Immunity, Hasan et al. (2015) report a previously unrecognized function of TREX1 as a guarantor of N-linked protein glycosylation. The group has shown that the C terminus of TREX1 traverses the endoplasmic reticulum (ER) membrane and directly interacts with components of the oligosaccharyl transferase (OST) complex that transfers complex sugars (glycans) from a lipid-linked oligosaccharide (LLO) precursor onto nascent protein strands in the ER (Figure 1). Both in isolated cells and in various tissues in vivo, Hasan et al. (2015) have demonstrated that, in the absence of TREX1, erroneous
hydrolysis of LLO leads to a several fold increase in free glycans, a mixture of LLO hydrolysis and degradation products. Overexpression of nucleasemutated full-length TREX1 or of the C terminus alone are both able to normalize free glycan amounts, indicating that not the cytosolic exonuclease but rather the ER-localized portion of TREX1 is responsible for regulating N-glycosylation. The group then has proceeded to show that free glycans purified from TREX1-deficient cells were immune stimulatory for macrophages and upregulated the inflammatory chemokine CXCL10 and other ISGs but not type I IFN. These find-
412 Immunity 43, September 15, 2015 ª2015 Elsevier Inc.
ings were in line with the group’s previous observation of interferon-independent activation of antiviral genes (Hasan et al., 2013). The same CXCL10-dominated ISG signature in the absence of type I IFN is also observed in patients suffering from RVCL. In RVCL patients, TREX1 lacks the ER-localized C terminus as a consequence of frame-shift mutations. The authors have shown that such mutations do not affect TREX1 nuclease activity but exclusively promote the accumulation of free glycans. Hasan et al. (2015) hypothesize that this is due to improper functioning of the OST complex and indeed find that in cells derived from RVCL patients, the OST inhibitor aclacinomycin (Acm) is able to efficiently reduce both free glycan accretion as well as the glycan-induced CXCL10 immune signature. These findings also hold true in vivo. Due to chronic inflammation in particular of the heart and liver, TREX1-deficient mice show limited viability. Intriguingly, Acm treatment of TREX1-deficient mice reduced the ISG signature, normalized liver and spleen morphology, and improved viability. In addition, a genetically targeted mouse model has recently been generated that recapitulates the most frequent human TREX1 mutation leading to the production of nucleaseinactive TREX1 protein (Grieves et al., 2015). However, in this TREX1 variant, the glycan-regulating function of TREX1 in the ER presumably remains intact, and, along with a milder course and a much later onset of autoimmune symptoms, these mice show improved viability compared to fully TREX1-deficient mice (Grieves et al., 2015). Altogether, this indicates that glycan regulation constitutes a key function of TREX1 and that deregulation of this pathway notably contributes to the severe immune pathology of TREX1-deficient mice, as well as to that of human RVCL patients. In order to study TREX1 glycan regulation independent of its nuclease function, Hasan et al. (2015) generated a new mouse model of RVCL by placing the human TREX1-V235fs mutant allele under the control of the endogenous murine promoter. Primary myeloid cells of these mice that are heterozygous carriers of this allele show elevated glycans, as well as chronic CXCL10 and other ISG
Immunity
Previews expression, with a more pronounced phenotype in TREX1-V235fs homozygotes. Although these mice showed normal viability and did not develop overt tissue inflammation, older mice developed far more autoantibodies than WT mice of the same age. Future studies with this model will reveal whether OST inhibition by Acm is able to suppress autoantibody formation. Given that TREX1 expression varies widely between cell types and activation states, the exact role of TREX-mediated regulation of N-glycosylation bears further investigation. Is TREX1 required to maintain general glycan homeostasis, or might it play a role in compensating for scenarios of cellular distress, or of infections? Although the innate immune sensing of fungal and bacterial glycans via toll-like receptors TLRs 2 and 4 or the NOD-like receptors is well understood, the data presented by Hasan et al. (2015) for the first time implicate endogenous free glycans as innate immune stimuli. However, the precise nature of the free self-glycan(s) that demonstrate immune activity remains to be established. The authors also do not identify which innate immune receptor(s) initiate the response to free glycans. This is not surprising given the sheer multitude of complex sugar-sensing immune receptors (Rabinovich et al., 2012). On the other hand, the authors’ finding that free glycan sensing is independent of cGAS and
stimulator of interferon genes (STING) yet dependent on tank-binding kinase 1 (TBK1) dramatically narrows down the options, because this should already exclude a number of candidates, including all currently known C-type lectin receptors. Our growing mechanistic understanding of ISG-driven autoinflammatory conditions has fueled the interest to therapeutically antagonize deregulated nucleic acid sensing pathways (Junt and Barchet, 2015). Hasan et al. (2015) now highlight a role for the innate immune sensing of endogenous complex sugars and establish that drugs that act by modulating free glycans, such as Acm, might hold therapeutic promise for RVCL and possibly also for patients suffering from other autoimmune disorders. To derive meaning from complexity, the reductionist approach to science generally serves us quite well. Yet once an immune function is paradigmatically established—as is the case for the TREX1 nuclease activity—it is tempting to assume that all observed defects are caused by variations of the same pathomechanism. This might at times lead us to behave like the proverbial blind men who try to describe an elephant by focusing exclusively on the prominent feature that is encountered first. In their seminal paper, Hasan et al. (2015) boldly approach from a different angle, discover TREX1’s sweet tooth, and thus demon-
strate that also in the case of TREX1, we are only beginning to see the whole elephant. REFERENCES Crow, Y.J., and Manel, N. (2015). Nat. Rev. Immunol. 15, 429–440. Gray, E.E., Treuting, P.M., Woodward, J.J., and Stetson, D.B. (2015). J. Immunol. 195, 1939– 1943. Grieves, J.L., Fye, J.M., Harvey, S., Grayson, J.M., Hollis, T., and Perrino, F.W. (2015). Proc. Natl. Acad. Sci. USA 112, 5117–5122. Hasan, M., Koch, J., Rakheja, D., Pattnaik, A.K., Brugarolas, J., Dozmorov, I., Levine, B., Wakeland, E.K., Lee-Kirsch, M.A., and Yan, N. (2013). Nat. Immunol. 14, 61–71. Hasan, M., Fermaintt, C.S., Gao, N., Sakai, T., Miyazaki, T., Jiang, S., Li, Q.-Z., Atkinson, J.P., Morse, H.C., III, Lehrman, M.A., and Yan, N. (2015). Immunity 43, this issue, 463–474. Junt, T., and Barchet, W. (2015). Nat. Rev. Immunol. 15, 529–544. Lee-Kirsch, M.A., Gong, M., Chowdhury, D., Senenko, L., Engel, K., Lee, Y.-A., de Silva, U., Bailey, S.L., Witte, T., Vyse, T.J., et al. (2007). Nat. Genet. 39, 1065–1067. Rabinovich, G.A., van Kooyk, Y., and Cobb, B.A. (2012). Ann. N Y Acad. Sci. 1253, 1–15. Richards, A., van den Maagdenberg, A.M., Jen, J.C., Kavanagh, D., Bertram, P., Spitzer, D., Liszewski, M.K., Barilla-Labarca, M.L., Terwindt, G.M., Kasai, Y., et al. (2007). Nat. Genet. 39, 1068–1070. Stetson, D.B., Ko, J.S., Heidmann, T., and Medzhitov, R. (2008). Cell 134, 587–598.
Immunity 43, September 15, 2015 ª2015 Elsevier Inc. 413