Regulated cell death and inflammasome activation in gut injury following traumatic surgery in vitro and in vivo: implication for postoperative death due to multiorgan dysfunction

Postoperative multi-organ dysfunction (MOD) is associated with significant mortality and morbidity. Necroptosis has been implicated in different types of solid organ injury; however, the mechanisms linking necroptosis to inflammation require further elucidation. The present study examines the involvement of necroptosis and NLR family pyrin domain containing 3 (NLRP3) inflammasome in small intestine injury following traumatic surgery. Kidney transplantation in rats and renal ischaemia-reperfusion (I/R) in mice were used as traumatic and laparotomic surgery models to study necroptosis and inflammasome activation in the small intestinal post-surgery; additional groups also received receptor-interacting protein kinase 1 (RIPK1) inhibitor necrostatin-1s (Nec-1s). To investigate whether necroptosis regulates inflammasome activity in vitro, necroptosis was induced in human colonic epithelial cancer cells (Caco-2) by a combination of tumour necrosis factor-alpha (TNFα), SMAC mimetic LCL-161 and pan-caspase inhibitor Q-VD-Oph (together, TLQ), and necroptosis was blocked by Nec-1s or mixed lineage kinase-domain like (MLKL) inhibitor necrosulfonamide (NSA). Renal transplantation and renal ischaemia-reperfusion (I/R) upregulated the expression of necroptosis mediators (RIPK1; RIPK3; phosphorylated-MLKL) and inflammasome components (P2X purinoceptor subfamily 7, P2X7R; NLRP3; caspase-1) in the small intestines at 24 h, and Nec-1s suppressed the expression of inflammasome components. TLQ treatment induced NLRP3 inflammasome, promoted cleavage of caspase-1 and interleukin-1 beta (IL-1β), and stimulated extracellular ATP release from Caco-2 cells, and MLKL inhibitor NSA prevented TLQ-induced inflammasome activity and ATP release from Caco-2 cells. Our work suggested that necroptosis and inflammasome interactively promote remote postoperative small intestinal injury, at least in part, through ATP purinergic signalling. Necroptosis-inflammasome axis may be considered as novel therapeutic target for tackling postoperative MOD in the critical care settings.


INTRODUCTION
Postoperative death currently ranks as the 3 rd greatest contributor to death globally, with an estimated 4.2 million death occurring within 30 days after surgery each year [1].Increasing clinical evidence suggested that surgery per se inflicted trauma and stress to the body to cause dysfunction of multiple organs (MOD) [2].Postoperative MOD represents a formidable challenge to patient survival and quality of life, and greater effort is required to investigate underlying mechanisms and development strategies to prevent and treat MOD development following surgery to improve long-term outcome.
Notably, major noncardiac surgery is followed by dysfunction of multiple organs, encompassing the cardiovascular, pulmonary, gastrointestinal and renal systems, which is distant from the primary site of operation and is accompanied by systemic inflammation [3][4][5][6].Hypoperfusion during/immediately after surgery is considered to be the initiating event in postoperative MOD, whereby multiple organs experience low flow perfusion to cause potential ischaemia and reperfusion injury in addition to traumatic cell death.Subsequently, death cells release proinflammatory cytokines and damage-associated molecular patterns (DAMPs) into the systemic circulation with potential dissemination to a distant site to cause remote organ injuries [7,8].Indeed, it has been well documented the onset of respiratory failure in acute kidney injury (AKI) patients, and the onset of renal dysfunction in lung injury patients [8].The development of kidney, hepatic and cardiac injuries in Covid-19 pneumonia patients is another example of multi-organ cross-talk and highlights the difficulty in tackling MOD in the intensive care units [9]; post-surgical multiple organ injury may share such injurious crosstalk feature although the initiator (surgery vs virus) is totally different.
Necroptosis was reported as a key cell death mechanism underlying ischaemia-reperfusion organ injury [10,11] and chronic inflammatory disease [12][13][14].Our group previously demonstrated that necroptosis also plays a central role in remote lung injury that developed secondary to renal transplantation [15,16].The fact that inhibiting necroptosis prevents sterile (non-pathogen induced) inflammation suggests necroptosis as a form of "inflammatory" cell death, and highlights the therapeutic potential of targeting necroptosis in the prevention and treatment of multiple organ dysfunction and inflammation [17][18][19].Necroptosis machinery was shown to lead to NLR family pyrin domain containing 3 (NLRP3) inflammasome activation within macrophages/bone marrow-derived monocytes in the context of autoimmune diseases (e.g., rheumatoid arthritis, dermatitis) [14,20].
The involvement of necroptosis and inflammasome has been largely neglected in surgical setting.We consider the small intestine to be an important contributor to postoperative MOD, as the injured intestines may serve as a robust source of immunogenic mediators to precipitate and aggravate postoperative MOD [21].In the present study, we aim to investigate the development of necroptosis and associated inflammasome pathway in the small intestine following a major surgery, and to further explore how necroptosis modulates inflammasome activity to contribute to MOD in in vitro and in vivo rats and mice models.

Development of small intestine injury following ischaemic kidney transplantation
On day1 after transplantation, immunofluorescence staining showed a significant increase in RIPK1 level in the small intestines in cold-ischaemia 24 h (CI24) cohorts (Fig. 1B, C; NC vs. CI24, P = 0.024), but not in the cold-ischaemia 0 h group (live transplantation).Western blot corroborated a significant upregulation of the necroptosis mediator RIPK1 in the small intestines in the CI24 group (Fig. 1D, E; NC vs. CI24, P = 0.001; Supplementary File 1).Macrophage and neutrophil infiltrations into the small intestinal mucosa were also evident after transplantation (Fig. 1F).Histology examination revealed minimal structural changes in the CI0 cohorts, whereas the CI24 Lewis recipients displayed villi blunting/deformation, mucosal oedema, and epithelial erosion/ detachment (Fig. 1G).The collective findings suggest the development of remote injury and necroptosis in the small intestines following transplantation surgery, and upregulated expression of RIPK1 supports the use of RIPK1-selective inhibitor necrostatin-1s (Nec-1s) to block necroptosis and related pathways in the small intestines.
Renal Ischaemia-reperfusion leads to necroptosis and inflammasome activation in the small intestines In a mice kidney ischemia-reperfusion model (Fig. 3A), the fluorescent intensity of necroptosis mediator RIPK3 was significantly increased in the I/R cohorts at day1 and was mostly localised to the epithelium and the underlying mucosae (arrowheads) when compared to that of naïve controls (NC) (Fig. 3B, C; NC vs. I/R, P = 0.011); Nec-1s treatment suppressed RIPK3 level to indicate prevention of necroptosis.The small intestine expression of phosphorylated MLKL (phos-MLKL) was also significantly increased in mice that underwent renal I/R but not sham surgery (Fig. 3D, E; NC vs. I/R, P = 0.045; Supplementary File 6), and Nec-1s treatment prevented CI24-induced MLKL phosphorylation.I/R also promoted inflammasome as evidenced by intestinal upregulations of NLRP3 (NC vs. I/R, P = 0.045), pro-caspase 1 (NC vs. I/R, P = 0.023; Supplementary File 7) and cleaved caspase-1 (NC vs. I/R, P = 0.046; Supplementary File 7), which were prevented in the presence of Nec-1s (Fig. 3F-H).
Small intestine samples were double-stained for phos-MLKL and NLRP3 or caspase-1 to localise necroptosis and inflammasome activity.Renal I/R significantly increased the percentage of phos-MLKL positive cells (Fig. 4A-C; NC vs. I/R, P < 0.001) that were concentrated at the intestinal villi tip (arrowheads), and to a lesser extent within the intestinal crypts.I/R also enhanced small intestinal NLRP3 (Fig. 4A, D; NC vs. I/R, P = 0.021) and caspase-1 (Fig. 4B, E; NC vs. I/R, P = 0.002) immunofluorescence intensities that were mostly concentrated in the underlying propia lamina.Nec-1s treatment ameliorated MLKL phosphorylation and attenuated I/R-induced NLRP3 and caspase-1 immunofluorescence intensities in the small intestines.

Concurrent induction of necroptosis and inflammasome in small intestine epithelium-like cells
Previously reported necroptosis-inducing treatments included tumour necrosis factor-alpha (TNF-α) or lipopolysaccharides, along with cycloheximide (protein synthesis inhibitor), a SMAC (Second mitochondria-derived activator of caspase) mimetic (e.g.compound A) and/or a pan-caspase inhibitor (e.g.Z-VAD-FMK) [14,20].In the current study, we tested the necroptosis-inducing efficiency of a novel cocktail combination comprising TNF-α, the SMAC mimetic LCL-161 and the pan-caspase inhibitor Q-VD-Oph (abbreviated as TLQ).For comparison, another extra group of cells was treated with TNF-α and LCL-161 (abbreviated as TL) to activate both apoptosis and necroptosis.
As early as at 6 h after treatments (Fig. 5E), moderate phosphorylation of MLKL was seen in TLQ-treated Caco-2 cells, where phos-MLKL immunoreactive signals localised to the plasma membrane (arrowheads), in line with previous studies reporting that MLKL translocates to the plasma membrane to cause    perforation and cell death [24][25][26].NLRP3 immunofluorescence was primarily seen in cytoplasm and the intensity was increased by TL or TLQ treatment at 6 h (Fig. 5E).At 24 h, MLKL phosphorylation (Fig. 6A, B; % of pMLKL+ve cells, NC vs. TLQ, P < 0.001) and NLRP3 immunofluorescence intensity (Fig. 6A, C; NC vs. TL, P = 0.032; NC vs. TLQ, P < 0.001) were significantly elevated in TLQ-treated cells.TLQ also significantly increased the number of phos-MLKL and NLRP3 double positive cells (Fig. 6D; NC vs. TLQ, P = 0.005).Phos-MLKL and NLRP3 co-localisation suggests coactivation of necroptosis and inflammasome within the same intestinal epithelial cell, whereas singly stained cells support that necroptosis and inflammasome are activated in different, neighbouring cells.Moreover, as inflammasome activation is characterised by the nuclear-to-cytoplasm translocation of ASC and formation of cytoplasmic ASC speck [27], we have shown that TLQ and to lesser extent TL, induced cytoplasmic ASC aggregation in Caco-2 cells compared to NC (Fig. 6E, arrowheads).

DISCUSSION
The present study for the first time demonstrated concurrent activation of necroptosis and inflammasome in the small intestine following major surgery in the form of renal transplantation or renal ischaemia-reperfusion in rodents.Blocking necroptosis with the RIPK1 inhibitor Nec-1s or the MLKL inhibitor NSA effectively prevented inflammasome activation in the small intestinal cells.Taken together the current and our previously published work whereby traumatic surgery led to remote pulmonary and hepatic injury that was accompanied by increased TNF-α in the systemic circulation [15,16,28], the necroptosis-inflammasome axis may be therapeutically targeted to tackle postoperative multiorgan injury (Fig. 8).
Studies on macrophages concluded that necrosome directly activates the inflammasome complex to promote IL-1β release without inducing necroptosis or independent from MLKL.In an alternative model, it was suggested that phosphor-MLKL oligomers perforated plasma membrane to release intracellular contents as danger-associated molecular patterns (DAMPs) to target the DAMP receptors, subsequently activating the inflammasome complex [17,18].Herein we demonstrated that pMLKLpositive and NLRP3/caspase1-positive cells are in close proximity in the injured small intestines (Fig. 4), and that pMLKL and NLRP3 signals were either individually upregulated in neighbouring intestinal epithelial cells or co-localised within the same cell (Fig. 5).Moreover TLQ-induced extracellular ATP release was suppressed by MLKL inhibitor NSA (Fig. 6F).Our data suggests that postoperative small intestinal injury could encompass both paradigms of "necro-inflammation" and involves multiple cell types with heterogenous inflammasome activation pathways, potentially forming a positive amplification loop of injury and inflammation.
breakdown and gut microbiota dysbiosis predicted cognitive decline in patients with prodromal Alzheimer's disease [42].
Different therapies have been trialled to preserve gut function in surgical and critically-ill patients [45,46].Our group and others have demonstrated the alpha2-adrenergic agonist dexmedetomidine to be highly cytoprotective, organ-protective and anti-inflammatory in ischemia-reperfusion injury and sepsis [47][48][49][50].A small clinical study found that dexmedetomidine reduced intestinal injury marker diamine oxidase following liver resection [51]; future studies are necessary to ascertain whether dexmedetomidine improves postoperative intestinal function and reduces permeability with larger sample size, and to explore whether dexmedetomidine confers gut protection by inhibiting necroptosis and inflammasome activity as elucidated herein.
Our study is not without limitations.First, transgenic animals such as RIP knockout, MLKL knockout and NLRP3 knockout would be considered for further investigation as they are more specific than the inhibitors used in this study; still, NLRP3 detection in vivo and NLRP3 and p-MLKL determined in vitro pointed to the critical role of NLRP3 in regulated cell death exemplified by necroptosis.Second, we could not eliminate other detrimental effects associated with acute kidney or kidney graft injury on the gut, such as accumulation of toxic waste.Finally, the effects of alloresponses from kidney transplant on gut injuries remain unknow.However, these responses are normally initiated about a week after engraftment whilst our gut injury was determined after 24 h after surgery when allo-response has yet to occur.Despite these limitations, our work report herein encourages further studies in this area of research to enhance outcomes following major surgery in elderly and/or vulnerable patients.

CONCLUSION
In summary, our study provides novel insight into necroptosis and inflammasome activation as the pathophysiological mechanism underlying small intestinal injury following surgery, whereby necroptosis may regulate inflammasome activity through ATP release in the intestinal epithelium (Fig. 8).Our study highlights the therapeutic potential from collective targeting of necroptosis and inflammasome in the management of MOD after surgery and in critical care setting.

MATERIALS & METHODS Animal surgery and treatment
In the present study, rat allograft kidney transplantation surgery and mouse renal ischaemia-reperfusion injury were used as two different models of laparotomic major surgery.Inbred rats and mice were purchased from Harlan, Bicester, UK and were kept in temperatureand humidity-controlled cages in a specific pathogen-free facility at Chelsea-Westminster Campus, Imperial College London.All animal procedures were carried out in accordance with the United Kingdom Animals (Scientific Procedures) Act of 1986.
Adult male rats aged 12 to 16 weeks and weighing 225-250 grs were used in the allograft kidney transplantation model, where kidney from Brown-Norway (BN, RT 1n ) rat was transplanted orthotopically into Lewis (LEW, RT1 1 ) rat using conventional microvascular techniques under isoflurane anaesthesia (4% induction, 2% maintenance) as described before [15,16,28].In the Brown-Norway donor rats, the left kidney, aorta and inferior vena cava were exposed to extract the graft to be flushed.Lewis recipient rats were randomly allocated to naïve control (no transplantation, n = 4), cold-ischaemia 0 h (CI0, n = 4) or coldischaemia 24 h (CI24, n = 4).The CI0 group was used to emulate live transplantation, whereby the recipient's left kidney was removed and flushed, followed by anastomosis of the renal arteries, veins and ureters between the donor and the recipient, with total surgical ischaemia time limited to < 45 min.Recipient's contralateral naïve kidney was removed immediately after surgery.Successful engraftment and perfusion are confirmed by an immediate colour change of the kidney graft and expansion of the renal arteries/veins.For the CI 24 h group that mimics delayed transplantation, donor's left kidney was stored in 4 °C heparinised Soltran's preservation solution for 24 h, prior to implantation into the recipient as described above.Additional CI24 groups were given a single dose of necrostatin-1s (1.65 mg/kg in 200 µL PBS, intraperitoneal, n = 4; Nec-1s, 17802, Cell Signaling Technology, MA, USA) or vehicle PBS (200 µL, intraperitoneal, n = 4) immediately after transplantation.
In view of our previous studies on renal transplantation and remote organ injury [15,16], a sample size of n = 4-5 was considered to be sufficient for the present study.For both renal transplantation and renal ischaemia/reperfusion procedures, randomisation was achieved using a computer-based random number generator.For each animal, two different investigators were involved, whereby the first investigator performed the surgical procedures +/− treatment based on the randomised numbers and was aware of the group allocations; the second investigator was responsible for harvesting samples for subsequent experiments and data analysis.As the NC animals did not receive laparotomic surgery, blinding was not feasible for the NC group.
All animals were given post-operative analgesics buprenorphine (0.1 mg/kg, subcutaneous) and carprofen (5 mg/kg, subcutaneous) daily for three days after the procedure.No animals died during the period of study.

Haematoxylin and eosin staining
On day 4, PFA-fixed small intestine specimens were dehydrated in ethanol and xylene before paraffin embedment, and were cut into 5 um thick sections for haematoxylin and eosin staining.Slides were viewed under Olympus BX4 microscope and representative images for each group were taken.

Fig. 8
Fig. 8 Putative mechanisms underlying postoperative small intestinal injury.Surgical trauma is associated with systemic increase in proinflammatory cytokines (TNF-α) to lead to necroptosis of small intestinal epithelial cells.Stimulation of TNF receptor and de-ubiquitination are required for the activation of RIPK1 and the subsequent phosphor-activation of RIPK3 and MLKL.Phosphorylated MLKL oligomers translocate to and perforate the plasma membrane, causing the extracellular release of danger-associated molecular patterns (DAMPs) including ATP.In a neighbouring cell (epithelial or immune cell), ATP activates purinergic receptor P2X7R to induce the assembly of NLRP3 inflammasome complex to facilitate the processing of pro-inflammatory cytokine IL-1β.ASC Apoptosis-associated speck-like protein containing a CARD, casp1 caspase-1, DAMPs Danger-associated molecular patterns, IL-1β Interleukin-1 beta, MLKL, mixed lineage kinase domain-like, p-MLKL Phosphorylated mixed lineage kinase-domain like, NLRP3 NLR family pyrin domain containing 3, P2X7R P2X purinoceptor subfamily 7, RIPK1 Receptor-interacting protein kinase 1, RIPK3 Receptor-interacting protein kinase 3, TNF-α Tumour necrosis factor-alpha, TNFR Tumour necrosis factor receptor.Created with Biorender.