Advertisement

Hepatectomy-induced apoptotic extracellular vesicles stimulate neutrophils to secrete regenerative growth factors

Open AccessPublished:August 17, 2022DOI:https://doi.org/10.1016/j.jhep.2022.07.027

      Highlights

      • Partial hepatectomy induces release of apoptotic cell debris into the circulation.
      • Neutrophils efficiently clear apoptotic extracellular vesicles by efferocytosis.
      • Efferocytosing neutrophils secrete mitogens such as hepatocyte growth factor.
      • Hepatectomy-induced apoptotic cell debris correlate with hepatocyte growth factor levels.
      • Hepatocyte growth factor-positive neutrophils rapidly increase in the blood of patients after partial hepatectomy.

      Background & Aims

      Surgical resection of the cancerous tissue represents one of the few curative treatment options for neoplastic liver disease. Such partial hepatectomy (PHx) induces hepatocyte hyperplasia, which restores liver function. PHx is associated with bacterial translocation, leading to an immediate immune response involving neutrophils and macrophages, which are indispensable for the priming phase of liver regeneration. Additionally, PHx induces longer-lasting intrahepatic apoptosis. Herein, we investigated the effect of apoptotic extracellular vesicles (aEVs) on neutrophil function and their role in this later phase of liver regeneration.

      Methods

      A total of 124 patients undergoing PHx were included in this study. Blood levels of the apoptosis marker caspase-cleaved cytokeratin-18 (M30) and circulating aEVs were analyzed preoperatively and on the first and fifth postoperative days. Additionally, the in vitro effects of aEVs on the secretome, phenotype and functions of neutrophils were investigated.

      Results

      Circulating aEVs increased at the first postoperative day and were associated with higher concentrations of M30, which was only observed in patients with complete liver recovery. Efferocytosis of aEVs by neutrophils induced an activated phenotype (CD11bhighCD16highCD66bhighCD62Llow); however, classical inflammatory responses such as NETosis, respiratory burst, degranulation, or secretion of pro-inflammatory cytokines were not observed. Instead, efferocytosing neutrophils released various growth factors including fibroblast growth factor-2 and hepatocyte growth factor (HGF). Accordingly, we observed an increase of HGF-positive neutrophils after PHx and a correlation of plasma HGF with M30 levels.

      Conclusions

      These data suggest that the clearance of PHx-induced aEVs leads to a population of non-inflammatory but regenerative neutrophils, which may support human liver regeneration.

      Lay summary

      In this study, we show that the surgical removal of a diseased part of the liver triggers a specific type of programmed cell death in the residual liver tissue. This results in the release of vesicles from dying cells into the blood, where they are cleared by circulating immune cells. These respond by secreting hepatocyte growth factors that could potentially support the regeneration of the liver remnant.

      Graphical abstract

      Keywords

      Introduction

      Partial hepatectomy (PHx) remains one of the few curative treatment options for patients with neoplastic liver disease. Within a few weeks after surgery, liver size and function are restored as a consequence of a tightly regulated sequence of events.
      • Abu Rmilah A.
      • Zhou W.
      • Nelson E.
      • Lin L.
      • Amiot B.
      • Nyberg S.L.
      Understanding the marvels behind liver regeneration.
      The first hours of liver regeneration, referred to as the priming phase, are thought to be driven by changes in portal venous flow, bacterial translocation and subsequent activation of innate immunity, e.g. via the complement system and Kupffer cells. Additionally, neutrophils and monocytes are recruited, the importance of which has been demonstrated in depletion experiments resulting in compromised liver regeneration in mice undergoing PHx.
      • Selzner N.
      • Selzner M.
      • Odermatt B.
      • Tian Y.
      • Van Rooijen N.
      • Clavien P.A.
      ICAM-1 triggers liver regeneration through leukocyte recruitment and Kupffer cell-dependent release of TNF-α/IL-6 in mice.
      This early phase is followed by the progression phase, characterized by compensatory hepatocyte hypertrophy, which is primarily driven by growth factors including hepatocyte growth factor (HGF), epidermal growth factor and fibroblast growth factor (FGF),
      • Abu Rmilah A.
      • Zhou W.
      • Nelson E.
      • Lin L.
      • Amiot B.
      • Nyberg S.L.
      Understanding the marvels behind liver regeneration.
      of which different sources have been reported. While initial HGF release is believed to be extracellular matrix-derived,
      • Lindroos P.M.
      • Zarnegar R.
      • Michalopoulos G.K.
      Hepatocyte growth factor (hepatopoietin A) rapidly increases in plasma before DNA synthesis and liver regeneration stimulated by partial hepatectomy and carbon tetrachloride administration.
      a second peak occurs hours later, when newly synthesized HGF enters the circulation. Hepatic stellate cells and endothelial cells are suggested to account for HGF production
      • Pediaditakis P.
      • Lopez-Talavera J.C.
      • Petersen B.
      • Monga S.P.S.
      • Michalopoulos G.K.
      The processing and utilization of hepatocyte growth factor/scatter factor following partial hepatectomy in the rat.
      ,
      • Schirmacher P.
      • Geerts A.
      • Jung W.
      • Pietrangelo A.
      • Rogler C.E.
      • Dienes H.P.
      The role of Ito cells in the biosynthesis of HGF-SF in the liver.
      ; however, early reports indicate that non-parenchymal cells also contribute to this second HGF peak during the progression phase of liver regeneration.
      • Schirmacher P.
      • Geerts A.
      • Jung W.
      • Pietrangelo A.
      • Rogler C.E.
      • Dienes H.P.
      The role of Ito cells in the biosynthesis of HGF-SF in the liver.
      Neutrophils are well-known for their diverse inflammatory functions; however, a growing body of evidence suggests that these cells are also actively involved in regulating immunosuppression, angiogenesis and tissue repair.
      • Scapini P.
      • Marini O.
      • Tecchio C.
      • Cassatella M.A.
      Human neutrophils in the saga of cellular heterogeneity: insights and open questions.
      ,
      • Peiseler M.
      • Kubes P.
      More friend than foe: the emerging role of neutrophils in tissue repair.
      Specifically, neutrophils store and release vascular endothelial growth factor (VEGF) and the pro-resolving molecules annexin A1 and lipoxins.
      • Jones H.R.
      • Robb C.T.
      • Perretti M.
      • Rossi A.G.
      The role of neutrophils in inflammation resolution.
      In addition, they clear injury sites from cell debris in a process termed efferocytosis and instruct inflammatory monocytes/macrophages to adopt a pro-regenerative phenotype.
      • Peiseler M.
      • Kubes P.
      More friend than foe: the emerging role of neutrophils in tissue repair.
      ,
      • Yang W.
      • Tao Y.
      • Wu Y.
      • Zhao X.
      • Ye W.
      • Zhao D.
      • et al.
      Neutrophils promote the development of reparative macrophages mediated by ROS to orchestrate liver repair.
      Accordingly, neutrophil depletion worsens tissue repair after experimental heart failure
      • Horckmans M.
      • Ring L.
      • Duchene J.
      • Santovito D.
      • Schloss M.J.
      • Drechsler M.
      • et al.
      Neutrophils orchestrate post-myocardial infarction healing by polarizing macrophages towards a reparative phenotype.
      and impedes liver regeneration after PHx
      • Selzner N.
      • Selzner M.
      • Odermatt B.
      • Tian Y.
      • Van Rooijen N.
      • Clavien P.A.
      ICAM-1 triggers liver regeneration through leukocyte recruitment and Kupffer cell-dependent release of TNF-α/IL-6 in mice.
      and acetaminophen-induced liver injury.
      • Yang W.
      • Tao Y.
      • Wu Y.
      • Zhao X.
      • Ye W.
      • Zhao D.
      • et al.
      Neutrophils promote the development of reparative macrophages mediated by ROS to orchestrate liver repair.
      Yet the role of neutrophils in tissue repair/regeneration remains incompletely understood. During the last decade, it has become evident that efferocytosis (i.e. uptake of apoptotic cells) activates an anti-inflammatory and regenerative response in phagocytes.
      • Doran A.C.
      • Yurdagul A.
      • Tabas I.
      Efferocytosis in health and disease.
      Apoptotic cells are typically cleared by neighboring phagocytes within the damaged tissue; however, massive cell death occurring as a consequence of PHx might surpass the local clearing capacity.
      • Hamazaki K.
      • Sato S.
      • Yunoki M.
      • Noda T.
      • Moreira L.F.
      • Mimura H.
      • et al.
      Kupffer cell function in chronic liver injury and after partial hepatectomy.
      As a result, apoptotic cell debris leaks into the circulation, where potential systemic responses are induced.
      • Sachet M.
      • Liang Y.Y.
      • Oehler R.
      The immune response to secondary necrotic cells.
      In this study, we could show that PHx-induced apoptosis results in a substantial release of apoptotic cell debris into the circulation. In vitro, apoptotic extracellular vesicles (aEVs) were primarily cleared by neutrophils, with aEV uptake inducing a pro-regenerative phenotype, distinct from previously described neutrophil subpopulations. Our findings suggest that initial bacterial translocation and later aEV release differentially affect liver regeneration after PHx, by eliciting inflammatory and pro-regenerative neutrophil phenotypes, respectively.

      Materials and methods

      Materials

      Materials are listed in Table S4. All general chemicals were purchased, if not indicated otherwise, from Merck-Millipore (Darmstadt, Germany). Jurkat cells and HepG2 cells have been purchased at ATCC (LGC Standards GmbH, Wesel, Germany).

      Patients and procedures

      One hundred and four patients undergoing liver surgery at Medical University of Vienna and Clinic Landstraße (Vienna, Austria) between March 2013 and March 2018 were included for perioperative monitoring. Of these, 95 patients underwent 1-step liver resection classified as minor (<3 segments) or major (≥3 segments) according to the Brisbane 2000 nomenclature.
      • Sorrentino C.
      • Miele L.
      • Porta A.
      • Pinto A.
      • Morello S.
      Activation of the A2B adenosine receptor in B16 melanomas induces CXCL12 expression in FAP-positive tumor stromal cells, enhancing tumor progression.
      The remaining 7 patients underwent 2-stage hepatectomy (associating liver partition and portal vein ligation in staged hepatectomy [ALPPS]). Briefly, ligation of the portal venous branches for the liver lobe to be resected and parenchymal transection were carried out in a first surgery, followed by the completion of the resection after sufficient regeneration of the future liver remnant. Baseline demographics were prospectively documented within our institutional data base (Table S1). Complications were classified according to the scheme given by Dindo et al.
      • Khan A.S.
      • Garcia-Aroz S.
      • Ansari M.A.
      • Atiq S.M.
      • Senter-Zapata M.
      • Fowler K.
      • et al.
      Assessment and optimization of liver volume before major hepatic resection: current guidelines and a narrative review.
      Any complication ≥ grade 1 was defined as postoperative morbidity, while complications ≥ grade 3 were defined as severe morbidity. In case of multiple complications, the most severe was used for grading. Liver dysfunction was defined according to the criteria issued by the international study group on liver surgery.
      • Khan A.S.
      • Garcia-Aroz S.
      • Ansari M.A.
      • Atiq S.M.
      • Senter-Zapata M.
      • Fowler K.
      • et al.
      Assessment and optimization of liver volume before major hepatic resection: current guidelines and a narrative review.
      Particularly, deranged values of serum bilirubin and prothrombin time on or after postoperative day (POD) 5 were defined as liver dysfunction. In case of preoperatively deranged parameters, the given criteria had to be fulfilled on 2 consecutive days after POD5. Patients were assigned to the “recovery” cohort if they did not fulfill these criteria or if they returned to normal levels of serum bilirubin and prothrombin time before POD5 and were not subjected to additional routine laboratory examination. Postoperative mortality was defined as death within 90 days after completion of liver surgery. For the analysis of neutrophilic HGF, we further included 22 patients undergoing liver resection between November 2021 and May 2022. Patient characteristics are shown in Table S3. The study was approved by the Institutional Ethics Committee (1186/2018) and all patients gave written informed consent.

      Assessment of circulating cell death markers

      Blood was drawn into Serum Separator Clot Activator tubes (Greiner Bio-One) and centrifuged at 1,000xg, 4°C for 10 minutes. Patient sera were analyzed for cytokeratin-18 isoforms using M30 Apoptosense and M65 Classic ELISA according to manufacturers’ instructions. Because cytokeratin-18 is highly expressed in hepatocytes, its caspase-cleaved isoform is a generally accepted biomarker of hepatocytic apoptosis.
      • Wimmer K.
      • Sachet M.
      • Oehler R.
      Circulating biomarkers of cell death.

      Analysis of liver-specific aEVs

      In pilot experiments, all assessed hepatocyte-specific surface markers were shed during apoptosis (data not shown), making positive staining of hepatocyte-derived aEVs impossible. Since EVs can be released from multiple cell types,
      • Shah R.
      • Patel T.
      • Freedman J.E.
      Circulating extracellular vesicles in human disease.
      we chose an exclusion approach, using CD235 (erythrocytes), CD45 (leukocytes), CD144 (endothelial cells), CD61 (thrombocytes) and CD31 (endothelial cells, thrombocytes) to identify non-hepatic EVs. Large, annexin V-positive, lineage marker-negative EVs were regarded as aEVs (for gating strategy, see Fig. S1). Blood was drawn into CTAD-tubes and centrifuged at 450xg (at 4°C) for 5 minutes. Twenty microliters of the supernatant were mixed with 80 μl of antibodies (anti-CD45-PE, -CD235-PC5, -CD144-PC7, -CD61-PE, -CD31-PE, annexin V-FITC, and Calcein Violet 450-AM), before flow cytometric analysis (CytoFlex, Beckman-Coulter).

      In vitro aEV preparation

      108 Jurkat or 8x10
      • Scapini P.
      • Marini O.
      • Tecchio C.
      • Cassatella M.A.
      Human neutrophils in the saga of cellular heterogeneity: insights and open questions.
      HepG2 cells were seeded in 10 cm2 petri dishes in DMEM/F-12 supplemented with 10% FCS. Apoptosis was induced by irradiation with 100 mJ/cm2 UV-C (CL-1000 Ultraviolet Linker, UVP, Thermo-Fisher) followed by incubation at 37°C for 24 hours. Apoptosis was confirmed as described in the supplementary methods. Cells were centrifuged at 450xg for 5 minutes and, subsequently, at 7,000xg (at 4°C) for 10 minutes to pellet apoptotic cell remnants (aCRs) and aEVs, respectively. Apoptotic cell products were quantified by flow cytometry (Gallios, Beckman-Coulter) and resistive cell counting was employed for size measurement (MOXI Z Mini Automated Cell Counter, ORFLO-Technologies).

      Efferocytosis

      EDTA-anticoagulated blood was stimulated with CFSE-labeled aCRs (1:1 ratio) or aEVs (1:10 ratio) prepared from apoptotic Jurkat cells for 2 hours, at 37°C, before antibody staining and erythrocyte lysis (Versalyse). Samples were analyzed by flow cytometry (Gallios, Beckman-Coulter) or image flow cytometry (Amnis ImageStream, Luminex). Small particles were excluded and CFSE+/CD14+ and CFSE+/CD15+ events were considered as efferocytosing monocytes or granulocytes, respectively. Image flow cytometry evaluation is described in Fig. S3.

      Secretome analysis

      Secreted cytokines/growth factors were analyzed using multiplex growth factor and chemokine panels (LEGENDPlex) and a 45-plex (ProcartaPlex) bead-based immunoassay, according to manufacturers’ recommendations.
      For further details regarding the materials and methods used, please refer to the CTAT table and supplementary information.

      Results

      PHx is associated with the release of apoptotic cell debris into the circulation

      To investigate the release of cell debris into the blood of patients after PHx, 2 cell death markers (both isoforms of cytokeratin-18) were analyzed: M30, specifically reflecting apoptosis and M65, indicating total cell death (apoptosis and necrosis). Patients undergoing single-step (PHx) or 2-step (ALPPS) liver resection were included in this study (procedural details are illustrated in Fig. 1A). M30 and M65 levels markedly increased in patients at the first postoperative day (POD1) (Fig. 1B). In patients with complete liver recovery, M30 concentrations were significantly elevated from before surgery to POD1, while a smaller increase was observed in patients developing liver dysfunction. M65, in contrast, significantly increased in patients with and without postoperative liver dysfunction. M30 and M65 levels declined from POD1 to POD5 in patients undergoing conventional PHx, while a sustained increase was evident in patients undergoing ALPPS, indicating continued release of cell debris from the atrophying liver lobe that is left in situ (Fig. 1B). When analyzing tissue samples from after the first and second surgery in patients undergoing ALPPS, we observed higher numbers of cells positive for active caspase-3 in the atrophic lobe after the second surgery, confirming ongoing apoptosis in the atrophying liver lobe (Fig. 1C).
      Figure thumbnail gr1
      Fig. 1Circulating markers of hepatocytic apoptosis after PHx.
      (A) Schematic drawings comparing single-step and 2-step PHx. In 1-step PHx, the diseased liver lobe is resected within 1 surgery (indicated with X) leaving behind the healthy liver remnant for recovery. In the 2-step approach (ALPPS), the hepatic vein supporting the diseased lobe is ligated and the liver parenchyma is cut at the future resection line during the first surgery (indicated with Ø). The diseased lobe becomes atrophic and is resected in a second surgery (X). Blood was taken preOP and on POD1 and POD5. (B) Levels of M30 (apoptosis marker) and M65 (total cell death marker) in 95 patients after PHx (left) and in 7 after ALPPS (right; whiskers indicate 5-95% percentiles). The middle graph shows M30 and M65 concentrations of patients with (n = 73) and without (n = 22) complete liver function recovery. Significance was tested using Wilcoxon signed-rank tests. (C) Immunofluorescence images of tissue samples from the healthy liver remnant and atrophic lobe in patients undergoing ALPPS during the 1st and 2nd surgery (anti-active caspase-3, red; DAPI, blue). (D) Circulating aEVs. Left: example of annexin V+/calcein+/CD235-/CD45-/CD144-/CD61-/CD31- aEVs. Right: circulating aEVs in 3 PHx and 3 control patients (colorectal cancer patients undergoing colon/rectum resection). Data are presented as mean ± SD. Significance was tested using paired Student's t tests. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. aEVs, apoptotic extracellular vesicles; ALPPS, associating liver partition and portal vein ligation in staged hepatectomy; PHx, partial hepatectomy; POD, postoperative day; PreOP, 1 day before surgery; SSC, side scatter.
      To ultimately demonstrate aEV release after PHx, circulating aEVs were assessed by flow cytometry before surgery and on POD1 (Fig. 1D). Indeed, circulating aEVs significantly increased on POD1 in patients after PHx, before returning to almost baseline on POD5. In contrast, aEV levels remained unchanged in patients undergoing a comparable open surgery not involving the liver.

      aEVs are efficiently cleared by neutrophils

      Apoptosis results in the formation of 2 forms of cell debris: aEVs, which are released in high numbers from dying cells and aCRs, which are left over at the end of this process. A detailed characterization of in vitro-generated aEVs and aCRs revealed that both apoptotic end products are covered by an intact membrane (Fig. 2A, left) and expose phosphatidylserine on their surface (i.e. they bind annexin V; Fig. 2A, middle), however, they significantly differ in size. Specifically, aEVs had a much smaller mean diameter than aCRs or viable cells (2.94 ± 0.03 μm, 7.22 ± 3.64 μm, and 10.95 ± 0.03 μm, respectively), as determined by resistive pulse sensing experiments (Fig. S2). Transmission electron microscopy further showed that aEVs contained degraded cytosolic and membranous structures, while residual nuclear fragments were mainly detectable in aCRs (Fig. 2A).
      Figure thumbnail gr2
      Fig. 2Efferocytosis of aEVs by white blood cells.
      (A) aEV characterization. AEVs and aCRs were isolated from apoptotic Jurkat cells and characterized by SEM, flow cytometry (same settings as in D), and TEM. (B) aEV efferocytosis. Healthy donor-derived whole blood was stimulated with CFSE-labeled aEVs or aCRs for 2 hours, before flow cytometric analysis. Graphs present mean ± SD of CFSE+/CD14+ monocytes and CFSE+/CD15+ granulocytes (n = 6). Significance was tested using paired Student's t tests. (C) Image flow cytometric analysis of white blood cells incubated with CFSE-labeled aEVs. Upper panel: representative examples of CD14+ monocytes (left) and CD15+ granulocytes (right) with internalized or attached aEVs. Lower panel: delta-centroid (Δc) value of each cell as a measure of the distance between the 2 colors in the composite image plotted against the internalization score, which describes how much membrane surrounds the CFSE signal. Cells with positive internalization score and low Δc were considered as phagocytes with internalized aEVs.∗∗∗p <0.001. aCRs, apoptotic cell remnants; aEVs, apoptotic extracellular vesicles; SEM, scanning electron microscopy; SSC, side scatter; TEM, transmission electron microscopy.
      Next, we sought to investigate the clearance of CFSE-labeled aEVs, aCRs and viable cells (control) by the major peripheral blood phagocytes: monocytes and granulocytes. Two hours after exposure, nearly every CD14+ monocyte and CD15+ granulocyte had internalized at least 1 aEV, as indicated by positive CFSE signal (Fig. 2B). In contrast, only around 20% of CD14+ and hardly any CD15+ cells showed double positivity for CFSE-aCRs. CFSE-labeled viable cells were only observed in monocytes (around 15%), which might be explained by antibody-dependent cellular phagocytosis caused by the allogeneic setting.
      To clarify whether the association of aEVs with phagocytes reflects actual uptake or just unspecific attachment, image flow cytometry was performed. CFSE-aEVs were indeed located within CD14+ and CD15+ cell bodies (Fig. 2C, top) and could be clearly discriminated from CFSE-aEVs attached to the cell surface (Fig. 2C, middle). Quantification of internalized and attached aEVs (for further details see Fig. S3) revealed that around 80% of CD14+ and 90% of CD15+ cells contained internalized aEVs (Fig. 2C, bottom), indicating active aEV uptake by both cell types. Considering that granulocytes show a 7.7 ± 2.0-fold higher relative frequency in blood samples compared to monocytes (Fig. S4), neutrophils might play a relevant role in aEV clearance. Yet, the actual contribution of neutrophils to aEV uptake in vivo requires further investigation.

      Efferocytosis of aEVs induces neutrophil morphologic and phenotypic changes, without evoking inflammatory responses

      Having shown that both neutrophils and monocytes efficiently engulf aEVs, we aimed to investigate potential aEV-induced phenotypic alterations in these cells. aEV stimulation resulted in CD45 and CD11b increases on both cell types, which were accompanied by CD62L shedding (Fig. 3A). Elevated CD54 expression, in contrast, was only observed on monocytes. Furthermore, aEV exposure resulted in CD16 and CD66b upregulation on neutrophils. Comparable changes were observable after neutrophil stimulation with hepatocellular-derived aEVs (Fig. S6).
      Figure thumbnail gr3
      Fig. 3Impact of aEVs on neutrophil activation.
      (A) Neutrophil and monocyte surface marker expression after 2 hours of whole blood stimulation with extracellular vesicles from apoptotic (aEVs) or untreated Jurkat cells (UT), as analyzed by flow cytometry (n = 5, for gating strategy see ). (B) Scanning electron microscopy of isolated neutrophils stimulated with TNFα or aEVs, or left untreated (UT). (C) Neutrophil DNA release after stimulation with aEVs prepared from HepG2 cells (1:10 ratio), PBS (negative control) or the calcium ionophore A23187 (positive control). SytoxGreen fluorescence was measured in supernatants (n = 3). (D) Neutrophil respiratory burst after stimulation with aEVs prepared from HepG2 cells (1:10 ratio), HBSS (negative control) or fMLP (positive control). Cytochrome c reduction served as measure of NADPH oxidase activity (n = 3). (E) Elastase release of neutrophils stimulated with PBS (negative control), aEVs prepared from HepG2 cells (1:10 ratio) or fMLP (positive control) (n = 3). Graphs indicate the mean ± SD. Significance was tested using paired Student’s t tests. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001. aEVs, apoptotic extracellular vesicles; fMLP, N-formylmethionyl-leucyl-phenylalanine; HBBS, Hank’s balanced salt solution; MFI, mean fluorescence intensity; UT, untreated.
      To further assess morphological changes associated with aEV-mediated neutrophil activation, scanning electron microscopy was performed. While untreated neutrophils were perfectly round, tumor necrosis factor-α (TNFα) treatment resulted in the formation of multidirectional spike-like structures (Fig. 3B). Interestingly, aEV-stimulated neutrophils formed wide unilateral protrusions directed towards adjacent aEVs, which likely extend from the cell membrane as part of the engulfment process. Similar protrusions are observable during neutrophil extracellular trap (NET) formation
      • Muñoz L.E.
      • Bilyy R.
      • Biermann M.H.C.
      • Kienhöfer D.
      • Maueröder C.
      • Hahn J.
      • et al.
      Nanoparticles size-dependently initiate self-limiting NETosis-driven inflammation.
      ; however, aEV uptake did not induce DNA release, as one would expect from NETosing neutrophils (Fig. 3C). Moreover, aEVs failed to induce neutrophil respiratory burst (i.e. the formation of reactive oxygen species), a response strongly evoked by bacterial fMLP (N-formylmethionyl-leucyl-phenylalanine), which is abundantly present during the priming phase of liver regeneration (Fig. 3D). Additionally, aEVs did not stimulate the release of the pro-inflammatory granule protein elastase (Fig. 3E). Hence, neutrophil activation upon aEV efferocytosis differs clearly from the classical inflammatory response induced by phagocytosis of bacteria.

      AEVs induce a regenerative neutrophil secretome

      During the priming phase of liver regeneration, bacterial translocation is believed to trigger an initial inflammatory response, while growth factors seem to play an important role in the subsequent progression phase.
      • Fausto N.
      • Campbell J.S.
      • Riehle K.J.
      Liver regeneration.
      Analyzing the secretome of whole blood cells stimulated with E. coli (to reflect bacterial translocation), we found elevated secretion of inflammatory chemoattractants including CCL3, CCL4, CCL11, CXCL8, CXCL9, and CXCL10 (Fig. 4A). AEVs, in contrast, elicited a profoundly different response with 7 proteins induced exclusively by aEVs (indicated with an asterisk): stem cell factor, erythropoietin, VEGF, granulocyte-macrophage colony-stimulating factor, HGF, FGF2, and transforming growth factor-α. Stimulation of isolated neutrophils with aEVs led to a comparable increase of regenerative, but not pro-inflammatory proteins (Table S2), suggesting a major contribution of neutrophil-derived factors to the secretome of whole blood cells. Most importantly, FGF2, platelet-derived growth factor subunit B and HGF were induced in neutrophils and whole blood samples to a similar extent (Fig. 4B). Intracellular HGF flow cytometry further confirmed elevated HGF expression in CD66b-positive neutrophils after exposure to aEVs (Fig. 4C). Since neutrophils contain lower amounts of chemokine and growth factor mRNAs compared to monocytes,
      • Maraux M.
      • Gaillardet A.
      • Gally A.
      • Saas P.
      • Cherrier T.
      Human primary neutrophil mRNA does not contaminate human resolving macrophage mRNA after efferocytosis.
      we sought to exclude the possibility that the aforementioned factors derive from contaminating monocytes in the neutrophil preparation. Analyzing the secretome of monocytes isolated from the same donors, we found that among all significant proteins of the neutrophil data set, only CXCL8 was expressed at a higher level in monocytes than in neutrophils (Table S2).
      Figure thumbnail gr4
      Fig. 4Neutrophil secretome after aEV efferocytosis.
      (A) Whole blood samples were incubated with PBS, aEVs prepared from apoptotic Jurkat cells (1:10 ratio), or E. coli for 18 hours. Supernatants were analyzed using a multiplex immunoassay (green, high concentrations; red, low concentrations). (B) Whole blood or neutrophils isolated from the same healthy donors were stimulated for 18 hours with PBS or aEVs prepared from HepG2 cells (1:10 ratio). Supernatants were analyzed using a 45-plex immunoassay. Selected growth factor concentrations are shown (n = 3). (C) Neutrophil intracellular HGF 4 hours after stimulation with PBS or aEVs prepared from HepG2 cells (1:10 ratio), as analyzed by flow cytometry (n = 3). Neutrophils were isolated preoperatively from patients undergoing PHx. (D) HepG2 metabolic activity and cell proliferative activity following 24 hours of stimulation with supernatants of neutrophils primed with PBS or aEVs prepared from HepG2 cells (1:10 ratio), or supernatants of only aEVs, as analyzed by Alamar blue assay (n = 5) and cell cycle analysis (n = 3). (E) Effect of aEV-primed neutrophils on adaptive immunity (T-cell proliferation): T cells were pre-activated for 1 day with anti-CD3/-CD28 and labelled with CTV, before co-culture with autologous neutrophils at a 1:5 ratio for 72 hours. Neutrophils were primed for 2 hours with aEVs prepared from HepG2 cells (1:10), TNFα (100 ng/ml) or LPS (10 ng/ml) prior to co-culture (n = 3). (F) Effect of aEV-primed neutrophils on macrophage response: macrophages were stimulated with supernatants of neutrophils primed for 6 hours with PBS or aEVs prepared from HepG2 cells (1:10 ratio) and changes in LPS-induced TNFα release were analyzed (n = 3). Data are presented as mean ± SD. Statistical significance was tested using paired Student’s t tests. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001. aEVs, apoptotic extracellular vesicles; LPS, lipopolysaccharide; PBS, phosphate buffered saline; PHx, partial hepatectomy; PMN, polymorphonuclear neutrophils.
      Next, we aimed to investigate the impact of aEV-treated neutrophils on the human hepatocyte cell line HepG2. Fig. 4D shows that supernatants of aEV-stimulated neutrophils, but not aEVs or neutrophils alone, significantly enhanced the metabolic activity of these cells and increased the number of replicating cells (i.e. in S or G2 phase). These results support the pro-regenerative potential of neutrophil-secreted factors on hepatocytes.
      To further investigate whether the observed aEV-mediated, non-inflammatory neutrophil activation is accompanied by an immunosuppressive polarization, which is primarily defined by the capacity of neutrophils to suppress T-cell responses,
      • Scapini P.
      • Marini O.
      • Tecchio C.
      • Cassatella M.A.
      Human neutrophils in the saga of cellular heterogeneity: insights and open questions.
      autologous neutrophil-T cell co-cultures were performed. Neutrophils pre-treated with aEVs did not differentially modulate T-cell proliferation when compared to neutrophils treated with TNFα or lipopolysaccharide (Fig. 4E and Fig. S7). Moreover, although neutrophil supernatants markedly impaired pro-inflammatory macrophage polarization, as defined by reduced lipopolysaccharide-induced TNFα release, aEV stimulation neither augmented nor abrogated this effect (Fig. 4F). Taken together, aEV uptake induces a pro-resolving neutrophil polarization without evident pro-inflammatory or immunosuppressive responses.

      PHx induces the formation of activated neutrophils expressing HGF

      Using transmission electron microscopy, we observed a significant increase in intrahepatic neutrophils 2 hours after induction of liver regeneration (Fig. 5A), indicating an extravasation of these cells during the early response to PHx. Twenty-two hours later (POD1), the circulating pool of neutrophils is replenished as demonstrated by the strongly elevated percentage of immature neutrophils negative for the maturation marker CD10 (Fig. 5B). The remaining mature neutrophils showed a significant upregulation of CD66b, but not CD11b on POD1. Interestingly, CD10+/CD66bhigh cells had been previously described as a non-inflammatory neutrophil subpopulation.
      • Marini O.
      • Costa S.
      • Bevilacqua D.
      • Calzetti F.
      • Tamassia N.
      • Spina C.
      • et al.
      Mature CD10+ and immature CD10- neutrophils present in G-CSF-treated donors display opposite effects on T cells.
      Intracellular HGF flow cytometry further revealed a substantial increase in the percentage of HGF-positive neutrophils on POD1 in patients undergoing major PHx (Fig. 5C), suggesting a pro-regenerative polarization of circulating neutrophils. Since growth factors have been proposed to primarily play a role in major hepatectomy, while hepatocyte hypertrophy is more relevant after minor resection,
      • Mitchell C.
      • Nivison M.
      • Jackson L.F.
      • Fox R.
      • Lee D.C.
      • Campbell J.S.
      • et al.
      Heparin-binding epidermal growth factor-like growth factor links hepatocyte priming with cell cycle progression during liver regeneration.
      we sought to investigate neutrophilic HGF in patients undergoing minor PHx. In these patients, no increase in HGF expression was evident on POD1. Notably, neutrophil numbers increased equally in patients undergoing either minor or major resections (Fig. 5C).
      Figure thumbnail gr5
      Fig. 5Neutrophils in patients undergoing PHx.
      (A) Liver tissue samples were collected before and 2 hours after induction of liver regeneration via portal vein ligation and CD66b immunofluorescence was quantified (n = 25, whiskers indicate 5-95% percentiles). Significance was calculated using Wilcoxon signed-rank test. Transmission electron microscopy was used to confirm intrahepatic neutrophil accumulation. (B) Cell surface markers on circulating neutrophils. Blood samples preOP and on POD1 were analyzed by flow cytometry (n = 5, for gating strategy, see ). (C) Neutrophil counts and HGF expression measured on preOP and POD1 using flow cytometry (for gating strategy, see ). Patients were grouped based on resection size into minor (n = 11) and major (n = 11) resection. Data are presented as mean ± SD. Statistical significance was tested using paired Student’s t tests. ∗p <0.05. PHx, partial hepatectomy; POD, postoperative day; PreOP, 1 day before surgery.

      Circulating apoptotic debris correlates with plasma HGF in patients after PHx

      Having observed that HGF-positive neutrophils increase after PHx, we aimed to investigate the correlation between neutrophils, apoptosis, and HGF in patients with PHx in more detail. Similar to the time course observed for M30 (Fig. 1B), levels of HGF as well as myeloperoxidase and elastase (as markers for inflammatory neutrophil activation) peaked on POD1 (Fig. 6A). While M30 levels correlated well with HGF, its correlation with the inflammatory neutrophil markers myeloperoxidase and elastase was notably weak (Fig. 6B). This supports the view that the efferocytosis-induced, pro-regenerative neutrophil polarization contributes to systemic HGF increase following PHx. Patients undergoing ALPPS showed prolonged secretion of M30 with the highest levels on POD5 (Fig. 1B). At this time point, M30 and HGF levels strongly correlated (r = 0.857), strengthening the link between these parameters in vivo (Fig. 6C).
      Figure thumbnail gr6
      Fig. 6M30 correlation with neutrophil activation markers after PHx.
      Blood samples were collected preOP and on POD1 and POD5 from patients undergoing 1-step (PHx) or 2-step partial hepatectomy (ALPPS). (A) Time course of plasma HGF (n = 12, 13 and 12 for preOP, POD1 and POD5), MPO (n = 48, 43 and 40 for preOP, POD1 and POD5) and elastase concentrations (n = 38, 45 and 38 for preOP, POD1 and POD5). Whiskers indicate 5-95% percentiles. Significance was tested using Wilcoxon signed-rank tests. (B) Correlation between M30 and HGF, MPO, and elastase in PHx samples (n = 58, 127 and 117, respectively). (C) Correlation between plasma HGF and M30 in ALPPS patients on preOP, POD1 and POD5 (n = 7). ∗p <0.05, ∗∗p <0.01. MPO, myeloperoxidase; PHx, partial hepatectomy; POD, postoperative day; PreOP, 1 day before surgery.

      Discussion

      In this study, we demonstrated that: 1) liver-specific aEVs are shed into the circulation during liver regeneration, 2) efferocytosis of aEVs by neutrophils induces a unique pro-regenerative polarization, which can also be detected in circulation after hepatic resection, 3) the atrophying liver lobe generated via the ALPPS procedure continuously sheds apoptotic cell debris into the circulation, generating an enormous source for neutrophils to be primed towards a pro-regenerative phenotype and possibly contributing to the accelerated hepatic regeneration seen in these patients. This adds to a growing body of evidence that neutrophils, besides their classical pro-inflammatory properties, are critically involved in non-inflammatory/pro-regenerative processes.
      In the hepatic microenvironment, Kupffer cells are primarily responsible for dead cell clearance.
      • De Oliveira T.H.C.
      • Marques P.E.
      • Proost P.
      • Teixeira M.M.M.
      Neutrophils: a cornerstone of liver ischemia and reperfusion injury.
      After PHx, however, their phagocytic capacity declines,
      • Hamazaki K.
      • Sato S.
      • Yunoki M.
      • Noda T.
      • Moreira L.F.
      • Mimura H.
      • et al.
      Kupffer cell function in chronic liver injury and after partial hepatectomy.
      which together with the sudden increase of cell death may cause the local clearance capacity to be surpassed. Consequently, elevated amounts of aEVs might be released into the circulation, where they are available for interaction with blood cells. Indeed, we found that both neutrophils and monocytes efficiently engulf aEVs; however, while the effects of efferocytosis on monocyte/macrophage polarization are well understood,
      • Doran A.C.
      • Yurdagul A.
      • Tabas I.
      Efferocytosis in health and disease.
      limited data are available for neutrophils. Apoptotic cell uptake by neutrophils has been demonstrated to prevent respiratory burst, NETosis and secretion of pro-inflammatory cytokines,
      • Esmann L.
      • Idel C.
      • Sarkar A.
      • Hellberg L.
      • Behnen M.
      • Möller S.
      • et al.
      Phagocytosis of apoptotic cells by neutrophil granulocytes: diminished proinflammatory neutrophil functions in the presence of apoptotic cells.
      ,
      • Manfredi A.A.
      • Ramirez G.A.
      • Rovere-Querini P.
      • Maugeri N.
      The neutrophil’s choice: phagocytose vs make neutrophil extracellular traps.
      which is in accordance with our findings. Yet, we observed markers of neutrophil activation following aEV exposure, which was accompanied by the release of pro-regenerative growth factors, such as HGF. While the absent inflammatory neutrophil responses would point towards a non-inflammatory or even immunosuppressive activation, aEV-treated neutrophils did not suppress T-cell proliferation, as one would expect from classical immunosuppressive neutrophils like granulocytic myeloid-derived suppressor cells.
      • Pillay J.
      • Tak T.
      • Kamp V.M.
      • Koenderman L.
      Immune suppression by neutrophils and granulocytic myeloid-derived suppressor cells: similarities and differences.
      aEV efferocytosis thus appears to prime neutrophils towards a non-inflammatory/pro-regenerative phenotype, which, to our knowledge, has not been described before.
      The dramatic increase of circulating aEVs after PHx and their impact on neutrophil polarization led us to investigate the potential relevance of this mechanism during liver regeneration. During the priming phase of liver regeneration, inflammatory neutrophil responses appear critical
      • Selzner N.
      • Selzner M.
      • Odermatt B.
      • Tian Y.
      • Van Rooijen N.
      • Clavien P.A.
      ICAM-1 triggers liver regeneration through leukocyte recruitment and Kupffer cell-dependent release of TNF-α/IL-6 in mice.
      ; however, the role of neutrophils in the later progression phase remains poorly understood. Apoptosis is essential in liver regeneration, which is limited in caspase-3 knockout mice after PHx.
      • Li F.
      • Huang Q.
      • Chen J.
      • Peng Y.
      • Roop D.R.
      • Bedford J.S.
      • et al.
      Apoptotic cells activate the “phoenix rising” pathway to promote wound healing and tissue regeneration.
      Since apoptosis is a tightly regulated and time consuming process that usually takes several hours,
      • Liang Y.Y.
      • Rainprecht D.
      • Eichmair E.
      • Messner B.
      • Oehler R.
      Serum-dependent processing of late apoptotic cells and their immunogenicity.
      we hypothesized that aEVs might elicit a distinct phenotype in neutrophils after the priming phase of liver regeneration – namely the progression phase. Indeed, our data indicate that significant numbers of neutrophils are mobilized from the bone marrow after liver resection, which can further be primed towards a pro-regenerative phenotype by aEV engulfment. The observed increase in HGF-positive circulating neutrophils in patients after PHx strengthens this assumption. This neutrophil phenotypic shift might allow for the continued production of pro-regenerative growth factors required during the progression phase of liver regeneration.
      While acute surgical trauma and the sudden increase of portal venous pressure after liver resection are thought to represent the main inducers of intrahepatic apoptosis, in patients undergoing ALPPS, a chronically malnourished liver lobe remains in place for 7-10 days. During this time, venous drainage from the remaining lobe is preserved, generating a potential source of liver-specific aEVs. Indeed, we observed a prolonged elevation of M30 in patients who underwent ALPPS, the levels of which strongly correlated with plasma HGF, one of the pro-regenerative proteins we found to be produced by efferocytosing neutrophils. This mechanism may partly account for the improved regeneration observed in patients who underwent ALPPS. In support of this hypothesis, various studies indicate a critical pro-regenerative role for circulating factors following ALPPS. Total blood exchange after PHx, for instance, has been shown to impair liver regeneration.
      • Eguchi S.
      • Sugiyama N.
      • Kawazoe Y.
      • Kawashita Y.
      • Fujioka H.
      • Furui J.
      • et al.
      Total blood exchange suppresses the early stage of liver regeneration following partial hepatectomy in rats.
      Additionally, plasma from ALPPS mice induced an accelerated regenerative response in regular PHx mice.
      • Schlegel A.
      • Lesurtel M.
      • Melloul E.
      • Limani P.
      • Tschuor C.
      • Humar B.
      • et al.
      ALPPS ;From human to mice highlighting accelerated and novel mechanisms of liver regeneration.
      In this study, we specifically focused on HGF as it serves as a major mitogenic stimulus driving hepatocyte proliferation. HGF levels follow a 2-phased time course during liver regeneration. The initial release phase after PHx depends on HGF conversion to its active form, causing a 17-fold increase in systemic HGF as early as 2 hours after PHx.
      • Zarnegar R.
      • DeFrances M.C.
      • Kost D.P.
      • Lindroos P.
      • Michalopoulos G.K.
      Expression of Hepatocyte Growth Factor mRNA in regenerating rat liver after partial hepatectomy.
      Pre-existing HGF storages are rapidly depleted, leading to the second phase – the production phase. Until now, hepatic stellate cells and endothelial cells were believed to primarily account for HGF production
      • Selzner N.
      • Selzner M.
      • Odermatt B.
      • Tian Y.
      • Van Rooijen N.
      • Clavien P.A.
      ICAM-1 triggers liver regeneration through leukocyte recruitment and Kupffer cell-dependent release of TNF-α/IL-6 in mice.
      ; however, a contribution of non-parenchymal cells has been suggested.
      • Schirmacher P.
      • Geerts A.
      • Jung W.
      • Pietrangelo A.
      • Rogler C.E.
      • Dienes H.P.
      The role of Ito cells in the biosynthesis of HGF-SF in the liver.
      When monitoring HGF in patients undergoing ALPPS, it was interesting to find that the initial HGF peak on POD1 did not correlate with the apoptosis marker M30, suggesting that HGF levels at this time point largely represent HGF released from tissue storages. Contrarily, a strong correlation between HGF and M30 was notable on POD5, a time point during the production phase. These results may suggest that during the second phase of HGF production, continued release of aEVs induces a neutrophil phenotypical switch that supports pro-regenerative growth factor production and thus might explain the accelerated liver regeneration observed in patients undergoing ALPPS. Importantly, postoperative HGF levels have been reported in humans to correlate with liver growth after living donor liver transplantation and ALPPS
      • Sparrelid E.
      • Johansson H.
      • Gilg S.
      • Nowak G.
      • Ellis E.
      • Isaksson B.
      Serial assessment of growth factors associated with liver regeneration in patients operated with associating liver partition and portal vein ligation for staged hepatectomy.
      ; however, further studies are required to address the role of neutrophil-derived HGF in ALPPS.
      In conclusion, our data suggest that efferocytosis of aEVs, which leak into the circulation after PHx, induces a non-inflammatory, pro-regenerative neutrophil phenotype, which might contribute to increased systemic HGF after liver resection. Our findings might not only be relevant for liver regeneration but may also represent a general mechanism of tissue repair.

      Abbreviations

      aCRs, apoptotic cell remnants; aEVs, apoptotic extracellular vesicles; ALPPS, associating liver partition and portal vein ligation in staged hepatectomy; FGF, fibroblast growth factor; HGF, hepatocyte growth factor; PHx, partial hepatectomy; POD, postoperative day; TNF, tumor necrosis factor-α; VEGF, vascular endothelial growth factor.

      Financial support

      This project has been funded by the Austrian Research Promotion Agency (FFG; grant no. 8668039 ).

      Authors’ contributions

      Conceptualization: VB, VS, AA, RO, PS. Patient recruitment: TG, PS. Patient sample collection: VB, VS, GO, DP, JS, MR. Laboratory analysis: VB, VS, SG, TH, NW, LB, TW, MM, WCS, CR, VG, CR. Electron microscopy and image flow cytometry: VB, VS, YYL, SR, RK, BM, AS. Funding acquisition: RO, AA, PS. Project administration: RO, AA, PS. Supervision: MS, AA, RO, PS. Data Analysis and Visualization: VB, VS
      Writing: RO, PS, VS.

      Data availability statement

      The data that support the findings of this study are available on request from the corresponding author.

      Conflict of interests

      Thomas Hammond is employed by and is a shareholder in AstraZeneca. All other authors declare no conflicts.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Acknowledgements

      Following people supported us with their excellent technical and intellectual expertise (in alphabetical order): Jennifer Fuxsteiner, Sofia Gabbasova, Fritz Garo, Stefanie Hägele, Hubert Hayden, Günther Hofbauer, Felix Huber, Markus Kelemen, Anna Emilia Kern, Brigitte Langer, Sarang Kim, Christoph Köditz, Johannes Längle, Sina Najarnia, Viola Nori-Cucchiari, Julia Pointner, Sabine Rauscher, Benedikt Rumpf, Manuel Salzmann and Katharina Seif. EM of Jurkat cells was performed at the Core Facility Cell Imaging and Ultrastructure Research, University of Vienna-member of the Vienna Life-Science Instruments.

      Supplementary data

      The following are the supplementary data to this article:

      References

        • Abu Rmilah A.
        • Zhou W.
        • Nelson E.
        • Lin L.
        • Amiot B.
        • Nyberg S.L.
        Understanding the marvels behind liver regeneration.
        Wiley Interdiscip Rev Dev Biol. 2019; 8: e340https://doi.org/10.1002/wdev.340
        • Selzner N.
        • Selzner M.
        • Odermatt B.
        • Tian Y.
        • Van Rooijen N.
        • Clavien P.A.
        ICAM-1 triggers liver regeneration through leukocyte recruitment and Kupffer cell-dependent release of TNF-α/IL-6 in mice.
        Gastroenterology. 2003; 124: 692-700https://doi.org/10.1053/gast.2003.50098
        • Lindroos P.M.
        • Zarnegar R.
        • Michalopoulos G.K.
        Hepatocyte growth factor (hepatopoietin A) rapidly increases in plasma before DNA synthesis and liver regeneration stimulated by partial hepatectomy and carbon tetrachloride administration.
        Hepatology. 1991; 13: 743-750https://doi.org/10.1002/hep.1840130422
        • Pediaditakis P.
        • Lopez-Talavera J.C.
        • Petersen B.
        • Monga S.P.S.
        • Michalopoulos G.K.
        The processing and utilization of hepatocyte growth factor/scatter factor following partial hepatectomy in the rat.
        Hepatology. 2001; 34: 688-693https://doi.org/10.1053/jhep.2001.27811
        • Schirmacher P.
        • Geerts A.
        • Jung W.
        • Pietrangelo A.
        • Rogler C.E.
        • Dienes H.P.
        The role of Ito cells in the biosynthesis of HGF-SF in the liver.
        EXS. 1993; 65: 285-299
        • Scapini P.
        • Marini O.
        • Tecchio C.
        • Cassatella M.A.
        Human neutrophils in the saga of cellular heterogeneity: insights and open questions.
        Immunol Rev. 2016; 273: 48-60https://doi.org/10.1111/imr.12448
        • Peiseler M.
        • Kubes P.
        More friend than foe: the emerging role of neutrophils in tissue repair.
        J Clin Invest. 2019; 129: 2629-2639https://doi.org/10.1172/JCI124616
        • Jones H.R.
        • Robb C.T.
        • Perretti M.
        • Rossi A.G.
        The role of neutrophils in inflammation resolution.
        Semin Immunol. 2016; 28: 137-145https://doi.org/10.1016/j.smim.2016.03.007
        • Yang W.
        • Tao Y.
        • Wu Y.
        • Zhao X.
        • Ye W.
        • Zhao D.
        • et al.
        Neutrophils promote the development of reparative macrophages mediated by ROS to orchestrate liver repair.
        Nat Commun. 2019; 10: 1076https://doi.org/10.1038/s41467-019-09046-8
        • Horckmans M.
        • Ring L.
        • Duchene J.
        • Santovito D.
        • Schloss M.J.
        • Drechsler M.
        • et al.
        Neutrophils orchestrate post-myocardial infarction healing by polarizing macrophages towards a reparative phenotype.
        Eur Heart J. 2016; 38ehw002https://doi.org/10.1093/eurheartj/ehw002
        • Doran A.C.
        • Yurdagul A.
        • Tabas I.
        Efferocytosis in health and disease.
        Nat Rev Immunol. 2020; 20: 254-267https://doi.org/10.1038/s41577-019-0240-6
        • Hamazaki K.
        • Sato S.
        • Yunoki M.
        • Noda T.
        • Moreira L.F.
        • Mimura H.
        • et al.
        Kupffer cell function in chronic liver injury and after partial hepatectomy.
        Res Exp Med. 1994; 194: 237-246https://doi.org/10.1007/BF02576385
        • Sachet M.
        • Liang Y.Y.
        • Oehler R.
        The immune response to secondary necrotic cells.
        Apoptosis. 2017; 22: 1189-1204https://doi.org/10.1007/s10495-017-1413-z
        • Sorrentino C.
        • Miele L.
        • Porta A.
        • Pinto A.
        • Morello S.
        Activation of the A2B adenosine receptor in B16 melanomas induces CXCL12 expression in FAP-positive tumor stromal cells, enhancing tumor progression.
        Oncotarget. 2016; 7: 64274-64288https://doi.org/10.18632/oncotarget.11729
        • Khan A.S.
        • Garcia-Aroz S.
        • Ansari M.A.
        • Atiq S.M.
        • Senter-Zapata M.
        • Fowler K.
        • et al.
        Assessment and optimization of liver volume before major hepatic resection: current guidelines and a narrative review.
        Int J Surg. 2018; 52: 74-81https://doi.org/10.1016/j.ijsu.2018.01.042
        • Wimmer K.
        • Sachet M.
        • Oehler R.
        Circulating biomarkers of cell death.
        Clin Chim Acta. 2020; 500: 87-97https://doi.org/10.1016/j.cca.2019.10.003
        • Shah R.
        • Patel T.
        • Freedman J.E.
        Circulating extracellular vesicles in human disease.
        N Engl J Med. 2018; 379: 958-966https://doi.org/10.1056/NEJMra1704286
        • Muñoz L.E.
        • Bilyy R.
        • Biermann M.H.C.
        • Kienhöfer D.
        • Maueröder C.
        • Hahn J.
        • et al.
        Nanoparticles size-dependently initiate self-limiting NETosis-driven inflammation.
        Proc Natl Acad Sci U S A. 2016; 113: E5856-E5865https://doi.org/10.1073/pnas.1602230113
        • Fausto N.
        • Campbell J.S.
        • Riehle K.J.
        Liver regeneration.
        J Hepatol. 2006; 43: S45-S53https://doi.org/10.1016/S0168-8278(00)80412-2
        • Maraux M.
        • Gaillardet A.
        • Gally A.
        • Saas P.
        • Cherrier T.
        Human primary neutrophil mRNA does not contaminate human resolving macrophage mRNA after efferocytosis.
        J Immunol Methods. 2020; 483112810https://doi.org/10.1016/j.jim.2020.112810
        • Marini O.
        • Costa S.
        • Bevilacqua D.
        • Calzetti F.
        • Tamassia N.
        • Spina C.
        • et al.
        Mature CD10+ and immature CD10- neutrophils present in G-CSF-treated donors display opposite effects on T cells.
        Blood. 2017; 129: 1343-1356https://doi.org/10.1182/blood-2016-04-713206
        • Mitchell C.
        • Nivison M.
        • Jackson L.F.
        • Fox R.
        • Lee D.C.
        • Campbell J.S.
        • et al.
        Heparin-binding epidermal growth factor-like growth factor links hepatocyte priming with cell cycle progression during liver regeneration.
        J Biol Chem. 2005; https://doi.org/10.1074/jbc.M412372200
        • De Oliveira T.H.C.
        • Marques P.E.
        • Proost P.
        • Teixeira M.M.M.
        Neutrophils: a cornerstone of liver ischemia and reperfusion injury.
        Lab Investig. 2018; 98: 51-62https://doi.org/10.1038/labinvest.2017.90
        • Esmann L.
        • Idel C.
        • Sarkar A.
        • Hellberg L.
        • Behnen M.
        • Möller S.
        • et al.
        Phagocytosis of apoptotic cells by neutrophil granulocytes: diminished proinflammatory neutrophil functions in the presence of apoptotic cells.
        J Immunol. 2010; 184: 391-400https://doi.org/10.4049/jimmunol.0900564
        • Manfredi A.A.
        • Ramirez G.A.
        • Rovere-Querini P.
        • Maugeri N.
        The neutrophil’s choice: phagocytose vs make neutrophil extracellular traps.
        Front Immunol. 2018; 9: 288https://doi.org/10.3389/fimmu.2018.00288
        • Pillay J.
        • Tak T.
        • Kamp V.M.
        • Koenderman L.
        Immune suppression by neutrophils and granulocytic myeloid-derived suppressor cells: similarities and differences.
        Cell Mol Life Sci. 2013; 70: 3813-3827https://doi.org/10.1007/s00018-013-1286-4
        • Li F.
        • Huang Q.
        • Chen J.
        • Peng Y.
        • Roop D.R.
        • Bedford J.S.
        • et al.
        Apoptotic cells activate the “phoenix rising” pathway to promote wound healing and tissue regeneration.
        Sci Signal. 2010; 3: ra13https://doi.org/10.1126/scisignal.2000634
        • Liang Y.Y.
        • Rainprecht D.
        • Eichmair E.
        • Messner B.
        • Oehler R.
        Serum-dependent processing of late apoptotic cells and their immunogenicity.
        Apoptosis. 2015; 20: 1444-1456https://doi.org/10.1007/s10495-015-1163-8
        • Eguchi S.
        • Sugiyama N.
        • Kawazoe Y.
        • Kawashita Y.
        • Fujioka H.
        • Furui J.
        • et al.
        Total blood exchange suppresses the early stage of liver regeneration following partial hepatectomy in rats.
        Artif Organs. 1998; 22: 847-853https://doi.org/10.1046/j.1525-1594.1998.06166.x
        • Schlegel A.
        • Lesurtel M.
        • Melloul E.
        • Limani P.
        • Tschuor C.
        • Humar B.
        • et al.
        ALPPS ;From human to mice highlighting accelerated and novel mechanisms of liver regeneration.
        Ann Surg. 2014; 260: 839-847https://doi.org/10.1097/SLA.0000000000000949
        • Zarnegar R.
        • DeFrances M.C.
        • Kost D.P.
        • Lindroos P.
        • Michalopoulos G.K.
        Expression of Hepatocyte Growth Factor mRNA in regenerating rat liver after partial hepatectomy.
        Biochem Biophys Res Commun. 1991; 177: 559-565https://doi.org/10.1016/0006-291X(91)92020-K
        • Sparrelid E.
        • Johansson H.
        • Gilg S.
        • Nowak G.
        • Ellis E.
        • Isaksson B.
        Serial assessment of growth factors associated with liver regeneration in patients operated with associating liver partition and portal vein ligation for staged hepatectomy.
        Eur Surg Res. 2018; 59: 72-82https://doi.org/10.1159/000488078