Hemophagocytic lymphohistiocytosis (HLH) is a rare pediatric syndrome associated with
inherited defects in the expression or extracellular secretion of perforin, a pore-forming
protein expressed by cytotoxic CD8 T cells or natural killer cells (1). Impairments
in perforin function prevent these cytolytic cells from inducing apoptosis in virally
infected (or otherwise damaged) host cells, leading to a persistent nidus of innate
immune activation. The consequent “cytokine storm” of HLH manifests as a febrile illness
with features of multiorgan dysfunction, such as hepatitis, splenomegaly, cytopenia
(especially anemia and thrombocytopenia), and central nervous system dysfunction (2).
Histology of affected tissues reveals expansion of aberrantly activated T cells and
macrophages that phagocytose host myeloid and erythroid cells. A diagnosis of HLH
is catastrophic, with a >90% mortality at a young age in the absence of chemotherapeutics
or bone marrow transplantation (3).
Although primary HLH is rare, there is increasing recognition of secondary forms of
HLH (sHLH), characterized by an acquired loss of cytolytic cell function (1, 4, 5).
One variant of sHLH is macrophage activation syndrome (MAS), a feared complication
of pediatric rheumatologic diseases such as systemic juvenile idiopathic arthritis
(sJIA) (6). Although the etiology of MAS is complex, emerging studies indicate that
the chronic inflammatory activation of autoinflammatory diseases such as sJIA suppresses
cytolytic cell function, potentially leading to an unremitting inflammatory response
to virally infected cells (1, 6, 7). Persistent induction of macrophage activation
leads to hemophagocytosis and the release of numerous proinflammatory cytokines (7,
8), mimicking the clinical findings of primary HLH.
Given that many adult patients suffer from chronic inflammatory diseases, there is
increasing concern that these patients may develop MAS-like disease states when hospitalized
for acute insults such as infection or malignancy (9). As such, MAS may be underrecognized
in adult ICUs. There is accordingly a need to mechanistically understand the proinflammatory
pathways responsible for not only the onset of sHLH/MAS but also the consequent multisystemic
organ injury responsible for disease morbidity and mortality.
In this issue of the Journal, Verweyen and colleagues (pp. 526–539) investigate the
role of IL-18 in the pathogenesis of MAS (10). IL-18 is a member of the IL-1 cytokine
family, which includes IL-1β. Similar to IL-1β, IL-18 is transcribed and translated
as a propeptide and then cleaved into an active form by caspases. As therapeutic inhibition
of IL-1β yielded only mixed benefits in the treatment of MAS, there is increasing
interest in the role of IL-18 as a potential therapeutically targetable mediator of
this disease state (1). Intriguingly, Verweyen and colleagues establish the scientific
premise for their investigations of MAS by studying IL-18 in a disease state at the
opposite end of the inflammatory spectrum: postseptic immunoparalysis. Healthy human
volunteers exposed to two sequential doses of intravenous LPS developed endotoxin
tolerance, characterized by suppressed TNFα (tumor necrosis factor α), IL-6, and IL-1β
release into the circulation after the second LPS exposure (11). Endotoxin tolerance
was similarly induced ex vivo by sequential dosing of healthy human peripheral blood
monocytes with LPS. In these in vivo human and ex vivo monocyte studies, the authors
observed that IL-18 escaped endotoxin tolerance, contrasting the suppression of tumor
necrosis factor α, IL-6, and IL-1β. The authors suggest that this escape from endotoxin
tolerance may be a consequence of a uniquely delayed induction of IL-18 transcription
after LPS. Although transcription of other cytokines peaked and resolved rapidly after
the first LPS dose, the delayed kinetics of IL-18 transcription led to ample IL-18
mRNA availability at the time of the second LPS dose, potentially providing continued
substrate for protein translation. The authors speculate that these unique IL-18 kinetics,
which corroborate a recently published study of IL-18 and IL-1β by Zhu and Kanneganti
(12), allow for persistent expression of an inflammatory cytokine that escapes LPS
tolerance, a finding potentially relevant to unremitting auto-inflammatory states
such as MAS.
After identifying delayed transcription of IL-18 after LPS, the authors sought to
determine the factors responsible for these unique transcriptional kinetics. Using
human peripheral blood monocytes, the authors observed that IL-18 induction was maximal
after TLR4 (Toll-like receptor 4) activation, with TLR5 agonists inducing only a blunted
activation of IL-18. In addition to TLR agonism, induction of IL-18 transcription
required type I IFN (IFN α/β) activation of JAK/STAT signaling. Conversely, type II
IFN (IFN γ) had no effect on IL-18 transcription. Type I IFN not only induced IL-18
but also controlled the kinetics of translation: pretreatment of monocytes with IFN
α/β accelerated the onset of LPS-induced IL-18 transcription. These findings were
confirmed using peripheral blood monocytes collected from a patient with a STAT1 gain-of-function
mutation. Notably, the authors did not test whether this acceleration of IL-18 transcription
reversed the previously observed ability of IL-18 to escape endotoxin tolerance. Interestingly,
type I IFN/JAK/STAT signaling had an opposite, inhibitory effect on IL-1β expression
in normal human peripheral blood monocytes, again demonstrating divergent mechanisms
of transcriptional control of these related cytokines (12).
After using models of endotoxin tolerance to identify the unique transcriptional kinetics
of IL-18, Verweyen and colleagues shifted their focus to investigate the effect of
type I IFN/JAK/STAT/IL-18 signaling on auto-inflammatory diseases such as sJIA and
MAS. In patients with sJIA or other autoinflammatory states (e.g., familial Mediterranean
fever [FMF]), peripheral blood monocyte expression of IL-18 was highly correlated
with expression of IFN-related genes, suggesting a mechanistic association. Furthermore,
microtubule destabilizing agents such as colchicine or nocodazole, commonly used to
treat autoinflammatory diseases, suppressed IL-18 and IFN β expression in LPS-treated
peripheral blood monocytes. Colchicine- or nocodazole-induced suppression of IL-18
transcription could be reversed by the administration of exogenous IFN α/β. The translational
relevance of these findings was supported by an observed suppression of circulating
IL-18 in colchicine-treated patients with FMF.
Finally, the authors confirmed the importance of JAK/STAT signaling to IL-18 expression
in vivo by analyzing samples collected from previously published studies of mouse
models of MAS (13, 14). These models, in which a MAS-like phenotype is induced by
repeated dosing with the TLR9 agonist CpG (with concurrent hemophagocytosis, if IL
10 is additionally inhibited [1]), revealed that treatment with the JAK1/2 inhibitor
ruxolitinib suppressed IL-18 expression. Furthermore, treatment of a MAS human patient,
who experienced a partial response to anti-IL-18 therapy (15), with the JAK1/3 inhibitor
tofacitinib similarly suppressed circulating IL-18, coincident with improved clinical
outcomes.
Taken together, this comprehensive work by Verweyen and colleagues elegantly used
pathologic extremes of human inflammation, ranging from postseptic immunoparalysis
to fulminant autoinflammatory disorders such as MAS and FMF, to glean new insights
into the transcriptional control of IL-18. Similar to most important studies, there
remain numerous unanswered questions. In contrast to IFN α/β signaling, the authors
found that IFN γ, a cytokine with known importance to MAS pathogenesis (1), exerted
minimal impact on IL-18 signaling. These findings demonstrate that the complex pathophysiology
of these autoinflammatory conditions likely cannot be explained by IL-18 alone. Furthermore,
it is uncertain if (and how) the authors’ work, derived largely from ex vivo studies
of LPS-treated peripheral blood monocytes, can be extrapolated to inform the in vivo
behavior of CD8 T cells and/or hemophagocytic tissue-resident macrophages pathognomonic
of MAS. Nevertheless, this work provides exciting insights into the mechanisms responsible
for control of IL-18 expression while identifying therapeutic targets (e.g., type
I IFN, JAK/STAT signaling) that may potentially help patients with autoinflammatory
disease.