Elsevier

Journal of Autoimmunity

Volume 55, December 2014, Pages 33-41
Journal of Autoimmunity

Altered subcellular localization of IL-33 leads to non-resolving lethal inflammation

https://doi.org/10.1016/j.jaut.2014.02.012Get rights and content

Highlights

  • Nuclear compartmentalization of IL-33 is vital for immune homeostasis.

  • First demonstration of altered subcellular cytokine localization resulting in death.

  • NLS-deficient IL-33 promotes non-resolving lethal inflammation via ST2 signaling.

  • The IL33tm1/+ mouse could be useful in evaluating therapeutics for chronic diseases.

Abstract

Non-resolving inflammation is a major contributor to chronic disease pathogenesis, including that of cancer, chronic obstructive pulmonary disease, asthma, arthritis, inflammatory bowel disease, multiple sclerosis and obesity. Some cytokines, such as IL-1α and IL-33, may act as endogenous alarmins that contribute to non-resolving inflammation. These cytokines are constitutively expressed in the nucleus and are thought to promote inflammation only upon release during tissue damage or cell necrosis. However, the importance of their nuclear localization in immune homeostasis is not fully understood. We describe herein a novel mouse model in which the nuclear localization signal of IL-33 is abolished and demonstrate for the first time that, alone, altered subcellular localization of IL-33 dramatically affects immune homeostasis. Heterozygous IL33tm1/+ mice display elevated serum IL-33 levels, indicating that IL-33 is constitutively released when not actively targeted to the nucleus. IL33tm1/+ mice succumb to lethal inflammation characterized by eosinophil-dominated immune cell infiltration of multiple organs. The profound inflammatory phenotype is dependent on mediators downstream of ST2 as ST2-null mice are protected in spite of high serum IL-33 levels. Importantly, IL-33 transcript levels in this knock-in mouse model remain under endogenous control. We adopt the term “nuclear alarmin” to describe a danger signal that is primarily regulated by nuclear compartmentalization in this fashion.

Introduction

Non-resolving inflammation is one of the major forces contributing to disease pathogenesis, including cancer, chronic obstructive pulmonary disease, asthma, arthritis, inflammatory bowel disease, multiple sclerosis, and obesity. There is a plethora of literature describing the principles of inflammation [1]. Multiple processes and molecular signals from necrotic cells alert the immune system following tissue injury with the release of multiple “danger signals” or “alarmins” that initiate an inflammatory response [2]. Infection can activate the innate immune system directly to induce localized inflammation, or act in concert with the signals arising from tissue injury to drive inflammation [3]. Once initiated, the processes that drive and perpetuate non-resolving inflammation are less-well understood and may include constitutive activation, loss of inhibitory signals or unrecognized ongoing triggers. Thus, the debate regarding which signals trigger non-resolving inflammation during chronic disease remains ongoing [3].

Chronic or non-resolving inflammation is characterized by persistent inflammation without a known inflammatory provocation, and without evident autoimmunity [3]. It can be caused by persistent stimulation, including inflammatory stimuli of exogenous origin. A prolonged or excessive response can delay or abrogate the resolution phase of the response, thereby leading to non-resolving inflammation. Finally, a non-productive resolution phase can lead to chronic or persistent inflammation. Thus, multiple and diverse signals that, under normal circumstances, contribute to the initiation and/or resolution of inflammation can potentially drive non-resolving chronic inflammation when homeostasis is perturbed.

IL-33 is a proinflammatory cytokine originally described as “Nuclear Factor from High Endothelial Venules” and widely reported to be a potent activator of the innate immune system and of type 2 immune responses [4]. It can act as an endogenous “danger signal” or “alarmin” when released in the context of pathogenic necrotic cell death [5], [6], [7]. The IL-33 receptor (ST2) is expressed by a wide range of cell types, including almost all innate immune cells [7], [8]. The broad distribution of the ST2 receptor favors the involvement of IL-33 in the pathogenesis of a wide range of diseases, including skin inflammatory conditions [9], [10], [11].

IL-33, IL-1α and HMGB1 are cytokines normally found in the nucleus of the cells that express them [12], [13]. Such nuclear expression has contributed to the notion that these cytokines may play dual roles as transcriptional regulators in addition to their immunomodulatory functions. Nuclear IL-33 is found in nearly all epithelial barrier tissues, and is expressed by some leukocyte populations [5], [6], [14], [15], [16], [17]. The role of IL-33 in the nucleus remains unclear. For instance, it was initially proposed that IL-33 in the nucleus functions as a transcriptional-repressor and then as a pro-inflammatory cytokine once it is processed and released [18]. However, current thinking holds that IL-33 is being sequestered in the nucleus via its N-terminal nuclear localization domain (NLD) in order to limit its proinflammatory potential, whereas release into the extracellular space during tissue injury potently activates the immune system [5], [12]. That full-length IL-33 is biologically active once released from the nucleus supports this view [19]. While it has been debated that cleavage of full-length IL-33 is required for its proinflammatory activity, similar to mature IL-1β, cleavage of the full length IL-33 by caspase-1 or caspase-3 disrupts the cytokine domain, resulting in a non-functional cleavage product [6], [20], [21], [22].

Given the evidence that nuclear sequestration may limit the systemic effects of IL-33, we asked whether loss of this localization signal would result in pathologic inflammation. To this end we created a knock-in allele of IL-33 in which the nuclear localization domain was replaced by DsRed fluorescent protein. In the absence of nuclear localization, we show that IL-33 is released into the circulation. Importantly, cytokine expression from the knock-in allele remains under the control of endogenous IL-33 regulatory elements. Nonetheless, in this knock-in mouse model, systemic release of IL-33 led to widespread non-resolving inflammation, culminating in the death of the animal. To our knowledge, this is the first in vivo demonstration of how a change in the sub-cellular compartmentalization of a cytokine leads to rampant non-resolving inflammation across multiple organs. These data demonstrate the importance of sub-cellular localization and compartmentalization of IL-33 for immune homeostasis.

Section snippets

Generation of IL-33 tm1 and IL-33 tm2 knock in mice

The targeting vectors were generated using recombineering technology as supplied by Gene Bridges GmbH, Heidelberg from the BAC RP23 380N12 (Invitrogen, Carlsbad, CA). The targeting vectors were used for homologous recombination in BALB/c ES cells. Following electroporation of a CRE-expressing plasmid, site specific recombination led to the removal of the neomycin selection cassette in vitro (Fig. S1A, C). After blastocyst injection of ES cell clones, chimeric animals were bred and DNA

Generation of IL-33 knock-in mice

The IL-33 gene is conserved throughout evolution and encodes a 30-kDa protein containing an N-terminal homeodomain, which includes the chromatin binding domains and a helix turn helix motif (mouse amino acids 1–68), and the 18-kDa C terminal cytokine domain (mouse amino acids 112–266) (Fig. 1, upper panel). The C terminal cytokine domain is sufficient for binding to and activating the ST2 receptor [9], [24]. To investigate the importance of nuclear localization to IL-33 cytokine function and

Discussion

In this study, we demonstrate that in the absence of nuclear sequestration, IL-33 is released in vivo into the systemic circulation, where it activates innate immunity and drives lethal system-wide non-resolving inflammation. Importantly, the elevated circulating level of IL-33 in this mouse model arises from basal gene transcription, rather than enforced protein overexpression (i.e., from a transgene) or administration of exogenous recombinant protein. Heterozygous knock-in IL33tm1/+ mice

Disclosures

The authors are employees of F. Hoffmann-La Roche Ltd., and Stiefel, a GSK Company.

Author contributions (order is indicative of effort contributed)

JCS conceived of the project.

JCS and AI oversaw the project.

JB, AI, CAM, JCS and MCVM planned experiments.

JB, CAM, SS, MCVM and AI performed experiments and generated data.

MCVM performed and interpreted histopathological results.

CAM and AI planned knock in strategy.

CAM generated knock in mice with assistance from SS.

JB, CAM, JCS, SHS, AI and SS analyzed and interpreted data.

JCS, SHS, JB and MCVM wrote the manuscript with input from AI and CAM.

Acknowledgments

We would like to thank Juerg H. Marty and Michael Hennig for their support, Martine Kapps, Natascha Santacroce, Liz Peixoto and Laetitia Petersen for their technical assistance, and Christelle Zundel for providing excellent technical assistance with necropsy, histology and immunohistochemistry. We would also like to thank Dr. Andrew N.J. McKenzie, from the Medical Research Council in Cambridge, UK, for providing the ST2 KO mice and Dr. Ronald Germain for critical reading of the manuscript.

References (51)

  • T. Pecaric-Petkovic et al.

    Human basophils and eosinophils are the direct target leukocytes of the novel IL-1 family member IL-33

    Blood

    (2009)
  • Z. Su et al.

    Potential autocrine regulation of interleukin-33/ST2 signaling of dendritic cells in allergic inflammation

    Mucosal Immunol

    (2013)
  • J.B. Seidelin et al.

    IL-33 is upregulated in colonocytes of ulcerative colitis

    Immunol Lett

    (2010)
  • G.P. Anderson

    Endotyping asthma: new insights into key pathogenic mechanisms in a complex, heterogeneous disease

    Lancet

    (2008)
  • C. Nathan

    Points of control in inflammation

    Nature

    (2002)
  • H. Kono et al.

    How dying cells alert the immune system to danger

    Nat Rev Immunol

    (2008)
  • C. Moussion et al.

    The IL-1-like cytokine IL-33 is constitutively expressed in the nucleus of endothelial cells and epithelial cells in vivo: a novel ‘alarmin’?

    PLoS One

    (2008)
  • T. Ohno et al.

    Interleukin-33 in allergy

    Allergy

    (2012)
  • A.E. Price et al.

    Systemically dispersed innate IL-13-expressing cells in type 2 immunity

    Proc Natl Acad Sci U S A

    (2010)
  • F.Y. Liew et al.

    Disease-associated functions of IL-33: the new kid in the IL-1 family

    Nat Rev Immunol

    (2010)
  • A. Balato et al.

    IL-33 is secreted by psoriatic keratinocytes and induces pro-inflammatory cytokines via keratinocyte and mast cell activation

    Exp Dermatol

    (2012)
  • E. Lefrancais et al.

    Mechanisms of IL-33 processing and secretion: differences and similarities between IL-1 family members

    Eur Cytokine Netw

    (2012)
  • S.L. Yu et al.

    The alarmin functions of high-mobility group box-1 and IL-33 in the pathogenesis of systemic lupus erythematosus

    Expert Rev Clin Immunol

    (2013)
  • D.E. Smith

    IL-33: a tissue derived cytokine pathway involved in allergic inflammation and asthma

    Clin Exp Allergy

    (2010)
  • C.L. Hsu et al.

    IL-33 is produced by mast cells and regulates IgE-dependent inflammation

    PLoS One

    (2010)
  • Cited by (132)

    • TSLP, IL-33, and IL-25: Not just for allergy and helminth infection

      2022, Journal of Allergy and Clinical Immunology
    View all citing articles on Scopus
    1

    J.B. and C.A.M contributed equally to this work.

    View full text