Programmed Necrosis and Disease:We interrupt your regular programming to bring you necroinflammation

Abstract

Compared to the tidy and immunologically silent death during apoptosis, necrosis seems like a chaotic and unorganized demise. However, we now recognize that there is a method to its madness, as many forms of necrotic cell death are indeed programmed and function beyond lytic cell death to support homeostasis and immunity. Inherently more immunogenic than their apoptotic counterpart, programmed necrosis, such as necroptosis, pyroptosis, ferroptosis, and NETosis, releases inflammatory cytokines and danger-associated molecular patterns (DAMPs), skewing the milieu to a pro-inflammatory state. Moreover, impaired clearance of dead cells often leads to inflammation. Importantly, these pathways have all been implicated in inflammatory and autoimmune diseases, therefore careful understanding of their molecular mechanisms can have long lasting effects on how we interpret their role in disease and how we translate these mechanisms into therapy.

Facts

• Necroinflammation is triggered by the release of inflammatory cytokines and DAMPs from cells undergoing necrotic cell death.

• Necroptosis, pyroptosis, ferroptosis, and NETosis are forms of programmed necrosis, wherein lytic cell death is mediated by an activatable genetic program.

• Defects in these programmed necrosis pathways and efferocytosis pathway have been linked to aberrant inflammation and autoimmunity.

Open Questions

• What crosstalk exists among these types of cell death within a complex organ?

• How can researchers exploit these pathways as therapeutic targets in vivo?

• What are the non-canonical roles for some of the key players in these pathways, and how do those roles impact on host defense?

Introduction: Making a Murderer

Inflammation is a complex biological response encompassing clinical signs such as heat, pain, redness, and swelling (calor, dolor, rubor, and tumor in Latin) [1], generally triggered by harmful stimuli including microbial infection and tissue damage [2]. Discovery of the pattern recognition theory by Charles Janeway shed light on the molecular mechanisms by which the innate immune system distinguishes self vs. non-self to mount a response against invasive microbes [3], specifically via innate surveillance for foreign pathological motifs termed pathogen-associated molecular patterns (PAMPs) by classes of pattern recognition receptors (PRRs) [4]. This system, however, is insufficient to explain uncontrolled inflammation, such as autoinflammatory and autoimmune diseases, and spontaneous tumor rejection. Polly Matzinger’s complementary arm to Janeway’s theory, termed the danger theory, describes sensing of endogenous “self” danger signals by the host immune system senses to initiate an appropriate immune response [5]. Accumulating evidence has revealed that various cellular stresses and modes of cell death themselves can trigger inflammation by generating danger-associated molecular patterns (DAMPs), even in the absence of microbe-derived PAMPs [6, 7]. Indeed, many DAMPs signal via PRRs, implicating the integration of signaling pathways in response to both exogenous and endogenous stimuli.

Until recently, apoptosis was considered the sole type of programmed cell death responsible for proper development, homeostasis, and host defense, while necrosis was thought of as cell death in response to physicochemical traumas [8]. We now recognize that many different types of genetically regulated cell death programs, like necroptosis, pyroptosis, ferroptosis, and NETosis, exist and are strongly associated with various inflammatory and autoimmune diseases [9]. Programmed necrosis is defined as a genetically controlled cell death with morphological features such as cellular swelling (oncosis), membrane rupture, and release of cellular contents, in contrast to the organized packaging that occurs during apoptosis. The distinct nature of these cell death processes translates into differing levels of immunogenicity - apoptosis is generally considered immunologically silent, while necrosis is reported to trigger inflammation [2], although impaired clearance of apoptotic cells (a process termed efferocytosis) can result in secondary necrosis, leading to the onset of inflammation [10, 11]. Importantly, cellular contents, normally sequestered safely within the cell to serve innocuous and valuable roles during normal physiology, are released extracellularly during necrotic cell death and can act as DAMPs [12, 13].

In this review, we describe the molecular aspects of distinct programmed cell death pathways, including their execution, their clearance, and their roles as triggers of inflammation. Finally, we explore the association of programmed necrosis and inflammatory and autoimmune diseases, as well as its potential as a therapeutic target [9].

Types of programmed cell death: How to Get Away with Murder

Apoptosis/secondary necrosis: The Walking Dead

Apoptosis (from the Greek for “falling off”) is the quintessential genetically-controlled programmed cell death pathway, designed to prevent unwanted inflammation during development, homeostasis, and infection [8]. The formation of membrane blebs and apoptotic bodies during apoptosis enables safe partitioning and enclosure of cellular contents (potential DAMPs) within the plasma membrane, and the active exposure of distinct cell death extracellular signals ensures effective recognition by phagocytes during subsequent scavenging [8, 10].

Broadly, apoptosis is mediated by a family of cysteine proteases called caspases, which initiate apoptosis via two different pathways [8, 14]. The intrinsic pathway (also known as the mitochondrial pathway) executes apoptosis in response to developmental signals and cellular stresses, such as DNA damage and activates by an imbalance in power between the pro-apoptotic and anti-apoptotic members of the Bcl-2 family. These protein mediate mitochondrial outer membrane permeabilization (MOMP), the release of cytochrome c from mitochondria into the cytosol, and the formation of the apoptosome (with caspase 9 and APAF-1) to activate downstream executioner caspase cascades [15]. During the extrinsic pathway, death receptor engagement by ligands such as tumor necrosis factor (TNF), Fas ligand (CD95L, FasL), and TNF-related apoptosis-inducing ligand (TRAIL) activates caspase 8 to signal to executioner caspases [16]. These caspase signaling cascades not only carry out the actual execution, they also mediate the molecular and cellular changes that contribute to apoptosis’s immunotolerance. Recent studies have demonstrated that the mitochondrial apoptotic pathway triggers potent inflammation in the absence of caspase activity via in concert with cGAS/STING and NF-κB signalings [17–19]. Furthermore, caspase signaling also contributes to externalization of “eat-me” signals to facilitate their rapid clearance (discussed below), DNA fragmentation, and membrane blebbing.

When apoptotic cells are not scavenged timely and efficiently, they undergo secondary necrosis (autolysis), characterized by damaged cell membrane and loss of cell integrity. Although secondary necrosis is suggested to be a natural outcome of apoptosis in single cell organisms, it also occurs in multicellular animals [20]. In vivo evidence of secondary necrosis can be observed under both physiological and pathological conditions, wherein apoptotic cells are shed into tissues with reduced phagocyte presence (e.g. airway lumen) or when excessive apoptosis outcompetes the capacity for efferocytosis, respectively [21, 22]. While secondary necrosis results in morphological changes that overlap with programmed necrosis, including cytoplasmic swelling, permeabilization of membranes, and the release of intracellular components, it is generally considered to be less immunogenic than primary necrosis since the DAMPs undergo danger-attenuating modifications during apoptosis [22]. Moreover, secondary necrotic cells can be cleared by micropinocytosis, thus dampening their inflammatory effects [20, 22].

Programmed necrosis: Homicide: Life on the Street

Many types of necrosis, both accidental and programmed, share common morphological features, including cellular swelling and plasma membrane permeabilization, resulting in the release of potentially harmful cellular contents and the induction of inflammation [7, 23]. Until recently, necrosis has been considered a solely accidental and non-programmed cell death modality. However, extensive genetic and pharmacological studies have clearly demonstrated that cells can and do undergo programmed necrotic processes, such as necroptosis, pyroptosis, ferroptosis, and NETosis (Fig. 1), that are induced by RIPK3/MLKL, inflammasome activation, iron metabolism, and reactive oxygen species (ROS), respectively [7]. Despite several shared morphological features, each programmed necrotic pathway utilizes distinct molecular process and generates different immunological outcomes. Since programmed necrosis is closely associated with multitude of inflammatory and autoimmune diseases, thorough understanding of these mechanisms is needed to identify effective therapeutic targets for disease [9, 24].

Fig. 1.

Molecular pathways of cell death and their roles in inflammation. Depending on death stimuli and context, live cells can undergo apoptosis or programmed necrosis. When caspase 3/7-dependent apoptotic cells are timely scavenged by efferocytosis, efferocytes like macrophages release anti-inflammatory cytokines and prevent unwanted inflammation. In the absence of effective clearance, apoptotic cells are proceeded to secondary necrosis and elicit some inflammatory responses. Upon damage signals such as infection or metabolic stress, cells trigger genetically programmed necrosis. Necroptosis and pyroptosis are RIPK3-MLKL- and inflammasome-GSDMD-mediated processes, respectively, displaying typical lytic morphology similar to primary necrosis. In contrast, ferroptosis is triggered by lipid peroxidation, and shows damaged mitochondria and reduced cellular volume. NETosis is a ROS-induced lytic cell death resulting in the extrusion of neutrophil extracellular traps (NETs), consisting of genomic DNA complexed with cellular proteins. Programmed necrotic cells generally release DAMPs and inflammatory cytokines that stimulate innate immune cells and promote necroinflammation

Necroptosis: Murder, She Wrote

Necroptosis was first reported in a seminal study wherein inhibition of caspase 8 sensitized fibroblasts to TNF-α-induced necrotic death [25]. Indeed, engagement of aforementioned death receptor pathways can trigger necroptotic cell death in the absence of caspase 8 activity, via RHIM-containing kinases like receptor-interacting protein kinase 1 (RIPK1) and RIPK3 [26, 27]. Additionally, signals that activate TIR-domain-containing adapter-inducing interferon-β (TRIF) or DNA-dependent activator of interferon-regulatory factors (DAI), both of which encode RHIM domains, by Toll like receptor 3/4 (TLR3/4) agonists including dsRNA, LPS, or viral DNA/RNA, respectively, are known to induce necroptosis [28, 29]. Furthermore, necroptosis can be triggered by other signals including T cell receptor (TCR) engagement, retinoic acid-inducible gene I (RIG-I) stimulation, and type I and type II interferon (IFN) signaling [30–32].

Downstream of these triggers are several molecular pathway options, depending on the availability of cellular inhibitor of apoptosis (cIAPs), FLICE-like inhibitory proteins (FLIP), or caspase 8, and the cell’s choice dictates its fate in the form of nuclear factor (NF)-κB-mediated survival, apoptosis, and necroptosis [7, 23]. Death receptor-induced necroptosis is currently the most deeply characterized necrotic pathway. Upon engagement of the TNF receptor (TNFR) by TNF-α, the cytosolic tail of TNFR recruits TNF receptor-associated death domain (TRADD) to promote the assembly of Complex I, in association with RIPK1, cIAP1/2, TNF receptor-associated factor 2/5 (TRAF2/5), and linear ubiquitin chain assembly complex (LUBAC). In this complex, RIPK1 is poly-ubiquitylated by cIAPs, which facilitates its interaction with TGF-β-activated kinase 1 (TAK1) and TAK1-associated binding protein (TAB). Further association with the inhibitor of NF-κB kinase (IKK) complex, promoted by LUBAC-mediated linear ubiquitination, activates canonical NF-κB signaling to promote cell survival, proliferation, and a potent pro-inflammatory response [33]. When RIPK1 is deubiquitinated by the deubiquitinase associated with cylindromatosis (CYLD), the TRADD-dependent Complex IIa is formed and executes apoptosis through caspase 8 activity. Likewise, if cIAPs, TAK1, or NF-κB essential modulator (NEMO) is depleted (by SMAC mimetics), RIPK1-dependent Complex IIb (sometimes referred to as the RIPoptosome) is assembled and dictates an alternative path of caspase 8-dependent apoptosis [33].

In the absence of caspase 8, activation of the necroptotic pathway requires assembly of the RIPK1/RIPK3 complex, called the Necrosome [34, 35]. Auto- and trans-phosphorylation of RIPK1 and RIPK3 enables recruitment and activation of the pseudokinase, mixed lineage kinase-like (MLKL), followed by oligomerization, binding to plasma membrane-associated phosphatidylinositol phosphates, and the disruption of plasma membrane integrity [26, 36, 37]. The RIPK1/RIPK3 complex also facilitates mitochondrially-derived reactive oxygen species (ROS) production, which reinforces the stability of the necrosome [38]. A recent study showed that ESCRT-III regulates the dynamics and kinetics of plasma membrane integrity during necroptosis, thus modulating the level of DAMPs release to facilitate an appropriate subsequent adaptive immune response [39]. As a consequence of plasma membrane rupture, extracellular DAMPs and cytokines affect neighboring cells and trigger inflammation via signaling through their cognate receptors [40]. Indeed, necroptotic cells release many types of DAMPs including HMGB1, nucleic acids, IL-1 family members, ATP, uric acid, and S100 protein, which contribute to the activation of innate immune cells like dendritic cells and macrophages [13, 40], although IL-33 seems to dampen inflammation through modulating regulatory T cells [13, 24]. In addition, coupling RIPK3 with RIPK1-NF-κB axis may play an important role in the subsequent stimulation of T cell response by enhancing inflammatory cytokines like IL-6 and CXCL1 [41].

In vivo evidence for necroptosis and its pathophysiological role is also rapidly accumulating. Recent studies indicate that necroptosis is very common during damage associated with physical trauma, infection, neurodegeneration, and ischemia. Despite the capacity of necroptosis to induce inflammation, emerging evidence suggests that the main pathological role of necroptosis is host defense. During specific circumstances like infection, cells may adopt this prompt and immunogenic death to eliminate the pathogen’s replicative niche [1, 50]. The phenotypic discrepancies observed between RIPK3- and MLKL-deficient mice support the idea of necroptotic machinery being a critical component of inflammation, beyond the function of necroptosis [42, 43]. RIPK3-deficient mice display enhanced protection against ischemic injury and sterile sepsis models via cell death-independent mechanisms, and RIPK3 deficiency also limits viral pathogenesis of West Nile Virus (WNV) independently of cell death [44, 45]. Moreover, RIPK3 can still promote inflammasome activation, even in the absence of MLKL [46]. Together, these results imply that RIPK3 has a role in regulating inflammation and the immune response independently of MLKL activation and necroptosis.

Pyroptosis: Burn Notice

Pyroptosis (from the Greek, pyro meaning fire), which is another form of programmed cell death mediated by excessive inflammasome activation and displays features common to both apoptosis (caspase dependence, chromatic condensation, and DNA fragmentation) and necrosis (cellular swelling, plasma membrane pore formation, and membrane rupture) [14, 47, 48]. The inflammasome is a multimeric complex responsible for the conversion of inactive pro-forms of inflammatory cytokines into their mature, active forms (IL-1β and IL-18). While the exact composition of the inflammasome can differ depending on the stimuli, it is broadly comprised of caspase 1 and a PRR that senses PAMPs or DAMPs, such as NOD-like receptors NLRP1 (anthrax lethal toxin), NLRP3 (various PAMPs and DAMPs), NLRC4 (flagellin, type III/IV secretion systems), AIM2 (dsDNA), and Pyrin (bacterial toxins) [10, 49, 50]. These PRRs contain either a caspase activation and recruitment domain (CARD) or a pyrin domain (PYD). Receptors containing CARD like NLRP1 and NLRC4 can either directly interact with pro-caspase 1, or indirectly associate with pro-caspase 1 via apoptosis-associated speck-like protein containing a CARD (ASC) to induce caspase 1 activation. PRRs without a CARD domain, such as NLRP3, AIM, and Pyrin require ASC to complex with pro-caspase 1^10,50. Caspase 11 was also reported to induce IL-1β maturation and pyroptosis in response to gram negative bacterial infections [51]. Recent studies have revealed that inflammatory caspases (human caspase 1/4/5, murine caspase 1/11) can cleave a conserved site at gasdermin D (GSDMD), and activate the pore-forming capacity of GSDMD to elicit the lytic feature of pyroptosis [51–53]. While pyroptosis plays a critical role in the containment of bacterial spread, the activation of the inflammatory cytokines, IL-1β and IL-18, by caspase 1 reveals another layer of inflammatory outcomes in addition to various DAMPs release [47, 48].

Ferroptosis: Iron Chef

The term ‘ferroptosis’ was coined to describe a distinct form of cell death observed in RAS-transformed tumor cells induced by inhibition of system X_C^– (using small molecules, like erastin) or inhibition of the downstream component glutathione peroxidase 4 (GPX4) [54]. A cell’s ability to undergo ferroptosis seems to be linked with its state of amino acid, polyunsaturated fatty acid (PUFA), and iron metabolic activity, as well as biosynthesis of glutathione (GSH) [55]. Ferroptosis exhibits unique morphological changes, like decreased cell volume and damaged mitochondria. Apoptotic features (DNA fragmentation, membrane blebbing) or necrotic features (oncosis, necrotic membrane rupture) are largely absent in ferroptosis [54]. Despite this, ferroptosis can still potently induce inflammation via the release of HMGB1, IL-33, and other unidentified pathways [54, 55]. Mechanistically, erastin treatment inhibits the system X_C^-Cys/Glu antiporter, which facilitates the exchange of extracellular _L-Cys for intracellular _L-Glu. This imbalance results in a depletion of intracellular GSH and inhibition of GPX4, the net effect being high intracellular levels of H₂O₂ and excessive lipid peroxidation, ultimately causing cell death in a manner independent of apoptotic or necroptotic machinery [54–56].

Intracellular cysteine availability is a rate-limiting step for GSH synthesis, and inhibiting the 1:1 exchange of cystine for glutamate through the antiporter is critical for the initiation of ferroptosis. Thus, excessive level of extracellular glutamate may trigger ferroptosis under physiological conditions [57]. In some cells, cysteine can be biosynthesized via the trans-sulfuration pathway using methionine, resulting in ferroptotic resistance [58]. Importantly, PUFAs are highly susceptible to lipid peroxidation, and the accumulation of PUFA hydroperoxides proves to be the lethal step in ferroptosis, as lipid peroxides can be broken into lipid radicals that react with proteins and nucleic acids [55].

Iron availability is a critical factor in determining ferroptosis, as lipid peroxidation is mediated by the activity of lipoxygenases via a phosphorylase b kinase gamma catalytic chain, liver/testis isoform (PHKG2)-dependent iron pool [59]. Silencing of iron-responsive element-binding protein 2 (IREB2), a master regulator of iron metabolism, decreases ferroptosis [54], while selective autophagic clearance of ferritin enhances susceptibility of ferroptosis by releasing free iron [60, 61].

NETosis: Deadliest Catch

Neutrophils are a short-lived but critical component at the frontline of host defense. In addition to phagocytosis, neutrophil produce and release of antimicrobiocidal granule contents (degranulation), ROS (oxidative bursts) and neutrophil extracellular traps (NETs) [62]. NETs are web-like structures comprised of decondensed chromatin bound to cytosolic and granular proteins. The NETs are able to trap, immobilize, inactivate, and kill microorganisms, as well as stimulate the immune response [62, 63]. An extracellular release of NETs is usually accompanied by a lytic form of cell death termed NETosis, which can be induced by microbial infections (bacteria, fungi, viruses, and parasites) or endogenous ‘sterile’ stimuli (uric acid crystal, cholesterol crystal, and immune complex) [62, 63].

Once activated, ROS is produced by either the NADPH oxidase (NOX) or mitochondrial pathway, resulting in the translocation of granule-associated enzymes like myeloperoxidase (MPO) and neutrophil elastase (NE) to the nucleus, where NE and MPO synergize to disrupt and decondense chromatin [64]. Following the disintegration of the nuclear envelope, the chromatin decondensation process expands into the cytoplasm of living cells, where DNA and cytoplasmic and granular components co-mingle. Finally, the plasma membrane is permeabilized, and NETs are released into the extracellular space, typically 3-8 h after neutrophil activation [62]. While DNase I is responsible for the clearance of NETs, the mechanism governing this process is unclear, and NET-associated proteins persist after DNA degradation, indicative of possible additional mechanisms [65]. In vitro studies have demonstrated that, in addition to DNase I, macrophages play an important scavenging role in NET clearance [66].

The morphology of NETosis suggests that this type of cell death is pro-inflammatory, owing to its lytic nature and the release of histones, nucleic acids, and cytotoxic granular molecules. Indeed, there is evidence that NETosis can initiate inflammation via the release of DAMPs. Crystal-induced NETs trigger the transcription of IL-6 and IL-1β in macrophages [67]. Moreover, oxidized nucleic acids can stimulate STING-dependent type I IFN production in myeloid cells. cathelicidin-DNA complexes promote the IFN response via endosomal TLRs in plasmacytoid dendritic cells, and cathelicidin can activate NLRP3 inflammasome in LPS-primed macrophages [63]. However, the immunological consequences of NETosis remain controversial as it is also suggested that NETs can help resolve local inflammation at a high concentration [68].

Efferocytosis: Curb your Enthusiasm

While the role of innate immunity to sense and control invasive foreign microbes has long been appreciated, its equally important role of mediating tolerance in the face of the body’s most abundant threat, dying cells, is often overlooked. Billions of cells undergo programmed cell death in an organism daily, and their silent and proficient clearance, a process termed efferocytosis (from the Latin meaning “to take to the grave”), ensures proper development and immunological homeostasis [69, 70]. For decades, efferocytotic research focused on the clearance of apoptotic cells, considered the archetypal programmed cell death. That characterization is not without merit - apoptosis is a critical component of an organism’s proper development and arguably the most abundant programmed cell death in vivo. The apoptotic cell packages itself into immunologically inert pieces and promotes its own removal via the active advertising of its presence. Thus, the efferocytosis of apoptotic cells can be viewed as the gold standard of dying cell clearance, as the mechanisms underlying apoptosis and efferocytosis are rooted in effective immunotolerance [71].

Efferocytosis is a coordinated event, wherein dying cells recruit, prime, and present themselves to both professional phagocytes (macrophages and dendritic cells) and non-professional phagocytes (epithelial cells). The release of “find-me” signals, such as ATP, UTP, sphingosine-1-phosphate (S1P), lysophosphatidylcholine (LPC), and CX3CL1 (fractalkine), by dying cells results in the recruitment of local phagocytes to the sites of cell death, as well as licenses the phagocyte to upregulate molecules required for efferocytosis [72]. The release of these “find-me” signals is often caspase-dependent, though recent studies have demonstrated that caspase-independent deaths result in high local concentrations of these signals, undoubtedly due to their lytic nature [73]. While “find-me” signals are an important first step in efferocytosis, their function, specifically in vivo, lies beyond their chemoattractant ability [74, 75].

Phagocytes distinguish dying cells from living cells by their display of “eat-me” signals, which are actively exposed by dying cells via caspase-dependent mechanisms. While molecules such as ICAM3, oxidized LDL-like molecules, calreticulin (CRT), and C1q-bound serum proteins have been described to act as “eat-me” signals [14, 71, 73], the best characterized “eat-me” signal is phosphatidylserine (PS), a lipid confined to the inner leaflet of the plasma membrane in living cells, but translocated to the outer leaflet by caspase-3-mediated activation of the scramblase Xkr8 [76, 77]. Recent reports demonstrate that exposure of PS on dying cells precedes actual death and morphological features of apoptosis, and PS exposure is sufficient to trigger efferocytosis, though concurrent expression of don’t eat me” signals, such as CD47, can hinder this effect [78]. Thus, PS exposure acts as a first line mechanism to preemptively clear dying cells before secondary necrosis and the release of DAMPs occurs.

Phagocytes are armed with receptors that allow them to sense and migrate to areas of cell death, as well as recognize and internalize cellular corpses for degradation and processing [71]. “Find-me” signals are detected via receptors like P2Y2 (ATP, UTP), S1PRs (S1P), G2A (LPC), and CXCR3 (CX3CL1), and engagement of these receptors mediates chemoattraction as well priming the phagocyte for engulfment [79]. Phagocytes employ a variety of PS-specific receptors (PSRs) to recognize “eat-me” signals on dying cells, and engagement of these receptors mediates engulfment. Examples of these PSRs include membrane-bound receptors, such as T cell immunoglobulin mucin receptor 4 (TIM4), brain-specific angiogenesis inhibitor 1 (BAI1), and stabilin-2, or bridging molecules, such as milk fat globule-EGF factor 8 (MFG-E8) and Gas6, which further link to integrins or Tryo3-Axl-Mer (TAM) receptors [80, 81]. Different tissues preferentially express different PSRs including BAI1 (bone marrow, spleen, brain), TIM4 (kidney), and stabilin-2 (sinusoidal endothelial cells), suggesting tissue-specific PS receptor mechanisms [71, 74, 82]. While downstream molecules used by specific PSRs can differ (and are discussed in greater detail here [14]), engulfment mediated by PSRs activates the Rho family of small GTPases, converging on evolutionarily conserved Rac1 [14].

Once engulfed, phagocytes must degrade, digest, and process their cellular corpse cargo. Rab5 and Rab7, small GTPases, are recruited to the phagosome to mediate fusion to the lysosomal network, which contains acidic proteases and nucleases that digest cellular corpses into their basic factors [14]. A recently described form of non-canonical autophagy, termed LC3-associated phagocytosis (LAP), was found to be crucial for the homeostatic clearance of dying cells, as well as shaping the appropriate immune response during efferocytosis [69, 70]. While LAP can be triggered by pathogen sensing (viaTLRs or FcR) during uptake, LAP is also activated by PSR engagement during efferocytosis, resulting in the recruitment of a distinct subset of autophagic machinery to the cargo-containing, single-membraned vesicle [69, 83]. LAP requires the activity and localization of Rubicon (RUN domain protein as Beclin-1 interacting and cysteine-rich containing), whereas canonical autophagy (induced by starvation or rapamycin treatment) is Rubicon-independent [84]. LAP facilitates the rapid destruction of the engulfed cargo via fusion with the lysosomal pathway and is crucial in mediating the immunotolerant response associated with efferocytosis [69, 83, 84]. Strikingly, Rubicon^-/- mice develop systemic lupus erythematosus (SLE)-like disease with age, whereas autophagy-deficient mice do not [70].

A dead cell is not like other foreign particles, however, and its engulfment essentially doubles the phagocyte’s levels of cellular components, including fats, sterols, peptides, and nucleotides. To handle this metabolic stress, efferocytosis leads to the activation of peroxisome proliferator-activated receptor γ/δ (PPARγ/δ) and liver x receptor (LXR) families, both critical modulators of cellular lipid homeostasis and immunity [85, 86]. Activation of these factors mediates the upregulation of other phagocytic machinery, such as the TAM family, and basal cholesterol efflux machinery, such as ABCA1 (ATP-binding cassette subfamily A, member 1), to accommodate this increase in cholesterol levels [87]. Indeed, cholesterol homeostasis is important for immunotolerance during the clearance of apoptotic cells, as it is closely linked to the production of anti-inflammatory cytokines, like TGF-β and IL-10, and active suppression of pro-inflammatory cytokines, such as TNF-α, IL-1, and IL-12 [88, 89]. Despite the multi-step process by which efferocytosis occurs, and the functional redundancy of some of the key players, defects in even a single component of this pathway can have deleterious effects, often resulting in autoinflammation or autoimmunity [11], highlighting the importance of this pathway in homeostasis.

DAMPs and inflammation: Jeopardy!

The efficiency with which efferocytosis operates underscores the danger of exposure to DAMPs [90, 91]. DAMPs run the gamut of cellular molecules, including nuclear or cytosolic proteins (high mobility group box 1 [HMGB1], S100 family of calcium-binding proteins, interleukin 1 [IL-1] family members, transcription factor A mitochondrial [TFAM], and histones) or non-protein molecules (ATP, uric acid, DNA, RNA, and mitochondrial DNA), which are both immunologically inert and biologically beneficially within the confines of a cell [92]. Phagocytes employ the same PRRs used to detect PAMPs to sense DAMPs, thus despite their “self” source, DAMPs are viewed as threatening as invading microbes (Table 1) [93].

Table 1.

DAMPs and the PRRs that recognize them

DAMP	PRR
Serum amyloid A	TLR1/2
Fatty acids	TLR4
Hyaluroinc acid	TLR2/4
Uric acid	NOD1/2, NLRP3
ATP	NLRP3, P2XR, P2YR, NOD1/2
snRNPs	RIG-I, MDA5, TLR7/8
dsDNA	TLR9, DAI, AIM2
Histone	TLR2/4/9, NLRP3
HMGB1	RAGE, TLR2/4
IC	FcR, TLR9
HSPs	TLR2/4
Surfactant Protein A/D	TLR2/4
Oxidized LDL	TLR4
Defensins	TLR2/4
RNA	TLR7
LL37	RAGE, TLR7/9
S100 proteins	RAGE, NOD1/2, TLR4
Reg111a	TLR4
Lactoferrin	TLR4

Abbreviations: TLR: Toll-like receptor; NOD: Nucleotide-binding oligomerization domain-containing protein; NLRP: NLR family, pyrin domain containing; ATP: adenosine triphosphate; snRNAP: small nuclear ribonucleoproteins; RIG-I: retinoic acid-inducible gene I; MDA: melanoma differentiation-associated protein; DAI (ZBP1): dna-dependent activator of interferon-regulatory factors; AIM2: absent in melanoma 2; HMGB1: high mobility group box 1; RAGE (AGER): receptor for advanced glycation endproducts; IC: immune complex; FcR: Fc receptor; HSP: heat shock protein; LDL: low-density lipoprotein; LL37 (CAMP): cathelicidin antimicrobial peptide;

The immunotolerant beauty of apoptosis lies in its sequestration of cellular contents into membrane-bound blebs which isolates DAMPs from their cognate PRRs. The developmentally and functionally homeostatic process of apoptosis is therefore akin to the nuclear compartmentalization of nucleic acids, permitting the evolution of viral nucleic acid sensors, which are critical for host defense [94]. Programmed necrosis, however, unleashes intracellular and immunostimulatory DAMPs on to phagocyte PRRs, eliciting inflammation and if left unchecked, autoimmune and autoinflammatory disorders (Fig. 2) [82]. The lytic nature of programmed necrosis means that its debris is inadvertently PS-positive, and necroptotic and pyroptotic cells can and do engage PSRs [10]. Studies have demonstrated that both necrotic and necroptotic cells engages TIM4 on phagocytes to mediate LAP, and engagement of PSRs is sufficient to engage the PPAR/LXR immunotolerant pathway [10]. Despite providing the excess cholesterol and fatty acids that tend to suppress inflammation by apoptotic cells, efferocytosis of necrotic cells does not mediate the same immunotolerant response [69]. The key to this quandary lies in the release of DAMPs – the immunostimulatory effects of DAMPs outweighs the tolerance induced by PSR engagement, and the net effect is inflammation [69,95].

Fig. 2.

Systemic spread of necroinflammation. Programmed necrosis or defective efferocytosis of dying cells can initiate local inflammation. DAMPs and proinflammatory cytokines derived from necrotic cells and subsequently activated immune cells provide feed-forward signals reinforcing programmed necrosis in more cells. Continuing the vicious death cycle allows the damage of barrier function, the spread of necroinflammation to systemic level, which may ultimately cause devastating multi-organ failure

Programmed necrosis, however, is not solely a losing game. Programmed necrosis can be triggered by an immunologically calamitous event, such as infection. In situations that require immune attention, these programmed necrotic pathways have likely evolved to modulate the immune response. Necrotic cell death achieves immunoregulatory effects beyond mere death – it can destroy the pathogen’s replicative niche, it can result in the increased release of “find-me” signals (which can double as DAMPs) [73], it can engage downstream immune signaling pathways [96], and the increased DAMPs release can heighten the inflammatory cascade. During pyroptosis, pore-induced intracellular traps (PITs) can trap intracellular bacteria within structures and complement and scavenger receptor stimulation can promote the PIT clearance by neutrophils. Therefore, pyroptosis and necroptosis may represent two unique biological pathways in which the host counteracts pathogen evasion mechanisms [97].

Autoimmune and inflammatory diseases: House, MD

Programmed necrosis has a crucial role in defending against invading microorganisms and regulating innate immunity. However, its activity may be a double-edged sword in human pathophysiology, as it is involved in the development of a variety of chronic inflammatory and autoimmune diseases via different mechanisms (Table 2).

Table 2.

Programmed necrosis and molecules in diseases summarized from in vivo and clinical studies

Programmed Necrosis	Molecule	Disease(s)	Reference
Programmed necrosis and associated molecules that protect against disease(s)
Necroptosis	RIPK3	Viral and bacterial infections, breast cancer	[35,45,99–101,137]
Pyroptosis	NLRP3	Crohn’s disease, EC, bacterial, viral, fungal and parasitic infections	[138–141]
	NLRP1	Bacterial and parasitic infections	[97,141]
	AIM2	Bacterial and viral infections	[142]
	NLRC4	Bacterial, fungal infection	[143–145]
	IFI16	Viral infection	[146]
	Pyrin	Bacterial infection	[147]
	ASC	EC, bacterial, viral, parasitic and chlamydia infections	[139,141,148–150]
	Caspase-1	EC, bacterial, fungal, viral and parasitic infections	[139,149,151]
	Caspase-4/5/11	Bacterial and parasitic infection	[141,151,152]
Ferroptosis	GPX4	Neuron degeneration, acute kidney injury	[126,127,130]
	Cystathioneγ-lyse	Huntington’s disease	[129]
	Ferritin H	Renal ischemia	[153]
	Hepcidin	Acute kidney injury	[154]
NETosis	MPO	Candida infection	[131]
	DNase1	SLE, DM	[136,155]
	α1-antitrypsin	DM	[156]
	NE	Bacterial infection	[157]
Programmed necrosis and associated molecules that contributes the development of disease(s)
Necroptosis	RIPK1	ALS, AD, brain and renal ischemia, Huntington’s disease, MS, traumatic brain injury, SIRS	[102,103,108–111,158,159]
	RIPK3	ALS, AD, myocardial infarction, renal ischemia and allograft inflammation, ethanol-induced liver disease, retinitis pigmentosa, atherosclerosis, SIRS, IBD, allergic colitis	[102,104–107,110,111,160–162]
	MLKL	IBD, allergic colitis, kidney ischemia, AD, acute pancreatitis	[111,162–164]
Pyroptosis	Caspase 1	EAE/MS, AD, DM, obesity	[118,120,122,165]
	IL18	Atherosclerosis, EAE/MS	[166,167]
	IL-1	Atherosclerosis, EAE/MS,PD, DM	[117,119,123,124]
	ASC	Atherosclerosis, EAE/MS, DM,obesity	[117,120,121,168]
	NLRP3	Autoinflammatory diseases, atherosclerosis, EAE/MS, AD, PD,DM,obesity	[116,117,120–122,168–174]
	NLRC4	Autoinflammatory diseases	[115,175]
	NLRP1	Autoinflammatory diseases	[114,176]
	Pyrin	Autoinflammatory diseases, EAE/MS	[177,178]
	Cnr1	DM	[179]
Ferroptosis	Hepcidin	Ischemic brain	[24,128]
	Heme oxygenase1	Acute kidney injury	[180]
	NE	RA, DM, atherosclerosis	[67,156,181]
NETosis	MPO	RA, AAV	[181–183]
	PAD4/PADs	SLE, RA, DM, atherosclerosis	[136,181,184–186]
	H3Cit/Histone	RA, DM, thrombus formation	[136,181,187]
	PR3	DM, AAV, atherosclerosis	[67,156,183]

Disease abbreviations: EC: experimental colitis; SLE: systemic lupus erythematosus; DM: diabetes mellitus; ALS: amyotrophic lateral sclerosis; AD: Alzheimer’s disease; IBD: inflammatory bowel disease; EAE: experimental autoimmune encephalomyelitis; MS: multiple sclerosis; SIRS: TNF-induced systemic inflammatory response syndrome; PD: Parkinson’s disease; RA: rheumatoid arthritis; AAV: anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis

Necroptosis is triggered by a vast number of pathogens that express TLR3/4 ligands, resulting in the rupture of infected cells and the suppression of microbial growth [98]. While RIPK1 global knockout mice are neonatally lethal, RIPK3- and MLKL-deficient mice undergo normal development and can be utilized to examine the role of necroptosis in vivo. Mice lacking RIPK3 are highly susceptible to bacterial (Yersinia pestis [99]) and viral pathogens [100], often accompanied by defective pathogen-induced tissue necrosis [35]. Interestingly, although RIPK3-deficient animals were more susceptible to influenza A virus and WNV infections, MLKL-deficient animals were not [45, 101]. These results suggest that control of these viruses in vivo may depend on RIPK3-directed transcriptional responses instead of necroptosis. RIPK1/3 have also been shown to contribute to the progression of ischemic and degenerative diseases. Treatment with the RIPK1 kinase inhibitor, Necrostatin-1, or genetic ablation of RIPK3 effectively protects against brain and kidney ischemia [102, 103], renal allograft inflammation [104], and myocardial infarction [105, 106]. Moreover, RIPK1 deficiency ameliorates retinal degeneration (retinitis pigmentosa) [107] and neuronal degeneration (Huntington’s disease, multiple sclerosis (MS), amyotrophic lateral sclerosis (ALS) and Alzheimer’s disease (AD)) [108–111] in animal disease models.

Pyroptosis acts as a protective host mechanism against microbial infections, via destruction of its replication niche. Inflammasome-defective mice are more susceptible to bacterial, viral, or fungal challenge [97]. Recent studies indicate that parasites can also trigger pyroptosis via inflammasome activation [112]. While inflammasome complexes are essential for pathogen clearance, their uncontrolled and excessive activation is detrimental and pathogenic. Over-activation of pyroptosis by pathogens can lead to widespread cell death, acute organ failure, sepsis, and septic shock [113]. Furthermore, patients carrying a gain-of-function mutation in inflammasome genes (NLRP1, NLRP3, and NLRC4) present with a wide spectrum of multi-organ inflammatory symptoms due to spontaneously increased inflammasome activation and IL-1β production [114–116]. Finally, inflammasome and caspase 1 activation have been implicated in myriad chronic inflammatory or autoimmune diseases, including atherosclerosis [117], MS [118, 119], diabetes [120], obesity [121], AD [122], and Parkinson’s disease [123], suggesting that either pharmacological or genetic inhibition of inflammasomes could significantly alleviate symptoms and reduce disease severity in experimental animals [117–123] or in patients [124].

Ablation of the key enzyme, GPX4, in the ferroptosis pathway results in embryonic lethality in mice, implying an important role of ferroptosis in development [125]. In addition, conditional deletion of GPX4 in murine neuronal cells caused rapid motor neuron degeneration, and inducible GPX4 ablation in adult mice led to neuronal loss and inflammation in the hippocampus [126, 127]. Increased levels of hepcidin, a regulator of iron release, was involved in ischemic brain injury, and its silencing ameliorated damage [128]. Importantly, impaired cysteine biosynthesis resulted in necrotic cell death in a mouse model, and substantial depletion of cystathione γ-lyse (an enzyme for cysteine biosynthesis) was observed in Huntington’s disease patients [129]. Furthermore, renal tubular cells are particularly sensitive to ferroptosis in GPX4-deficient mice [130]. Consistently, highly reactive catalytic iron is considered as a risk factor for acute kidney injury (AKI), and the treatment with iron chelators, like desferrioxamine, displayed some protective effect despite inconsistency [13]. In addition, ferritin, hepcidin, and heme oxygenase 1 are all involved in the pathogenesis of AKI [24].

Human patients deficient for MPO, a key neutrophilic enzyme, have an increased risk for Candida albicans infection [131]. Similarly, MPO deficient mice also failed to clear C. albicans and other large pathogens, but not small bacteria [132]. Despite its role in controlling infection, NETs are now considered a source of self-antigen and a potential initiator in many autoimmune diseases, including SLE [133], rheumatoid arthritis (RA) [134], anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) [135] and type 1 diabetes (T1D). Accordingly, SLE patients have a unique low-density granulocyte (LDG) population with an increased NET formation capacity [133], and have elevated serum anti-dsDNA and anti-ribonucleoprotein proteins that are directed against the NETs components [135]. Likewise, neutrophils from RA and T1D patients produce more NETs in vitro and display increased ROS production [134, 136]. Furthermore, NETs are implicated in vascular diseases, as evidenced by NE^-/- RP3^-/- mice which are protected against the development of atherosclerosis [67].

Conclusion: Life Goes On

Paradoxical as it may seem, life requires death – for the proper development of an organism, for the regeneration of tissue and cells over time, and for the appropriate immune response during challenge. We now recognize that apoptosis is not the only show in town and that other programmed cell death pathways, such as necroptosis, pyroptosis, ferroptosis, and NETosis, have their own role to play in an organism’s homeostasis and defense. These newly characterized pathways represent possible targets for therapies against inflammatory diseases and infections that have often been refractory to current treatments. Future studies on how these pathways intersect with other key biological processes and their tissue specificities are needed to fully harness their potential.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.