Alternatively activated macrophages promote resolution of necrosis following acute liver injury

Background & Aim Following acetaminophen (APAP) overdose, acute liver injury (ALI) can occur in patients that present too late for N-acetylcysteine treatment, potentially leading to acute liver failure, systemic inflammation, and death. Macrophages influence the progression and resolution of ALI due to their innate immunological function and paracrine activity. Syngeneic primary bone marrow-derived macrophages (BMDMs) were tested as a cell-based therapy in a mouse model of APAP-induced ALI (APAP-ALI). Methods Several phenotypically distinct BMDM populations were delivered intravenously to APAP-ALI mice when hepatic necrosis was established, and then evaluated based on their effects on injury, inflammation, immunity, and regeneration. In vivo phagocytosis assays were used to interrogate the phenotype and function of alternatively activated BMDMs (AAMs) post-injection. Finally, primary human AAMs sourced from healthy volunteers were evaluated in immunocompetent APAP-ALI mice. Results BMDMs rapidly localised to the liver and spleen within 4 h of administration. Injection of AAMs specifically reduced hepatocellular necrosis, HMGB1 translocation, and infiltrating neutrophils following APAP-ALI. AAM delivery also stimulated proliferation in hepatocytes and endothelium, and reduced levels of several circulating proinflammatory cytokines within 24 h. AAMs displayed a high phagocytic activity both in vitro and in injured liver tissue post-injection. Crosstalk with the host innate immune system was demonstrated by reduced infiltrating host Ly6Chi macrophages in AAM-treated mice. Importantly, therapeutic efficacy was partially recapitulated using clinical-grade primary human AAMs in immunocompetent APAP-ALI mice, underscoring the translational potential of these findings. Conclusion We identify that AAMs have value as a cell-based therapy in an experimental model of APAP-ALI. Human AAMs warrant further evaluation as a potential cell-based therapy for APAP overdose patients with established liver injury. Lay summary After an overdose of acetaminophen (paracetamol), some patients present to hospital too late for the current antidote (N-acetylcysteine) to be effective. We tested whether macrophages, an injury-responsive leukocyte that can scavenge dead/dying cells, could serve as a cell-based therapy in an experimental model of acetaminophen overdose. Injection of alternatively activated macrophages rapidly reduced liver injury and reduced several mediators of inflammation. Macrophages show promise to serve as a potential cell-based therapy for acute liver injury.

Open circles indicate a representative mouse receiving unstained BMDMs, closed circles indicate a representative mouse receiving Vivotrack-stained BMDMs (black circles, healthy mouse; red circles, APAP-ALI mouse). (C) IF microscopy shows uniform labelling of Vivotrack-labelled GFP+ BMDMs (right-hand panels) versus unlabelled GFP+ BMDMs (left-hand panels) at 647 nm (top row). GFP+ BMDMs visualised at 488 nm (middle panels) and brightfield (bottom panels) (D) Relative ATP levels in BMDMs after Vivotrack labelling. Open circles represent three individual labelling preparations. (E) Ex vivo fluorescent imaging of individual organs 4 hours after BMDM-injection in healthy and APAP-ALI mice. (F) Organ fluorescence quantification shows BMDM localisation in liver, lung, and spleen in mice receiving Vivotrack-stained BMDMs (black circles, healthy; red circles, APAP-ALI) with unstained controls (grey circles). (G). CFSE-labelled BMDMs (black arrows) were detected in liver and spleen using IHC 20 hours after i.v. injection. Scale bars, 100 µm. APAP-ALI, acetaminophen-induced acute liver injury, ATP, adenosine triphosphate; BMDMs, bone-marrow derived macrophages; CFSE, carboxyfluorescein succinimidyl ester; GFP, green fluorescent protein; i.v., intravenous.     S5. Delivery of AAMs at 6 hours led to reduced necrosis but did not change proliferation or serum inflammation parameters by 24 hours. (A) Study design: injection of AAMs (1x10 6 , i.v.) or PBS vehicle alone six hours after APAP administration (400 mg/kg, i.p.) to fasted mice. Tissues and blood was harvested at 24 hours. (B) Representative H+E histological stains of liver tissue from APAP-ALI mice receiving PBS or AAMs. Right panel shows necrosis quantification (open circles represent individual mice, n ≥ 7 per group). (C) Representative IF images of BrdU incorporation in liver tissues from APAP-ALI mice treated with PBS or AAMs. BrdU incorporation was low at 24 hours in all APAP-ALI mice. (D) Serum chemistry parameters were not significant differences in APAP-ALI mice treated with PBS or AAMs. (E) Serum proinflammatory cytokines were not significantly different between PBS-and AAMtreated mice. Scale bars 100 μm. Mann-Whitney test (B) (C) (D, bilirubin and albumin only) and (E) for non-parametric datasets. Two-way t-test (D, except bilirubin and albumin) for parametric datasets. p-values indicated in panels.   Table S1; n≥11, Wilcoxon test). (C) Percentage weight loss and liver/body weight ratio in APAP-ALI mice treated with PBS or AAMs. Each circle represents an individual mouse (n≥7, two-way t-test).  Hepatic infiltrating macrophages were defined as viable CD45+ Ly6G− CD3− NK1.1− CD19− CD11b hi F4/80 lo cells from non-parenchymal fraction of digested livers and used to identify macrophage subsets. Quantification of absolute numbers of cells per liver was performed by expressing each subset as a proportion of NPCs, counting total number of NPCs in the digested portion of liver, calculating the total number of NPCs in the whole liver by weight differential, thereby calculating the total number of each subpopulation. Transplanted AAMs were identified as CFSE+. The percentage CFSE+ cells was calculated on the gate of total viable CD45+ Ly6G-CD3-CD19-NK1.1-cells. The negative was set on a liver from an APAP-ALI mouse receiving vehicle instead of AAMs. AAMs were further classified on their Ly6C expression, gating set using FMO (fluorescence minus one) controls. The percentage of phagocytic cells (positive and negative) was calculated on the gate of CFSE+ cells. The negative was set using the liver from an APAP-ALI mouse transplanted with AAMs but injected with the vehicle instead of PKH26PCL. PKH26PCL MFI was calculated in the same gate.  6 ) were injected intravenously into either healthy or APAP-ALI mice at 16 hours before cull at 36 hours. Liver tissue was digested to isolate myeloid cells, and exogenous AAMs FACS-sorted for downstream PCR array analysis. (B) Graph shows mean (± SD) of number of isolated viable CFSE+ AAMs following tissue digest and FACS-sorting in healthy (saline) and APAP-ALI mice (n=3, biological replicates). (C) Serum ALT activity in healthy (saline) and APAP-ALI mice. (D) Volcano plot showing differentially expressed genes in AAMs injected into APAP-ALI mice versus AAMs injected into healthy mice. The x-axis indicates fold-change (Log2 scale) of each gene, and the y-axis indicates p-values (-Log10 scale). The horizontal line represents the threshold for significance, and the vertical dashed line indicate the threshold for statistically-relevant fold-changes. Four genes in the dataset (B2m, Tlr4, C4b, and Il10) were significantly downregulated (upper left quadrant) (n=3, twoway t-test). Actb, Gusb, and Hsp90ab1 served as housekeeping genes between healthy and APAP-ALI mice using the arithmetic mean.

Fig. S12. CFSE-labelled AAMs localise throughout liver parenchyma including peri-necrotic regions in TdTomato-labelled APAP-ALI mice. (A)
AAV8.Tbg-Cre injection (5x10 11 virus particles, i.v.) was performed in R26RLSL tdTomato mice to express TdTomato specifically in hepatocytes. IF was performed to boost TdTomato and FITC fluorescence using anti-RFP and anti-FITC antibodies before confocal microscopy. Panel shows representative IF images (left, antibody treated; right, isotype control) showing injected AAMs (yellow, white arrowheads) throughout the parenchyma and TdTomato+ hepatocytes (magenta) against DAPI counterstain (cyan). Centrilobular areas of hepatocyte necrosis contained TdTomato-negative infiltrating cells with punctate areas of TdTomato debris. (B) Serum ALT activity in R26LSLTdTomato mice 36 hours after APAP administration. Circles represent individual mice.     Table S1. APAP phenotypic scoring sheet. Each mouse was scored (0-3) hourly on each parameter after APAP administration. To prevent mice exceeding severity limits in order to comply with Home Office regulations, the following pre-defined thresholds were applied: within the first six hours, if mice score 8 or above, or score 3 for either 'neurological symptoms' or 'responsiveness to touch', then proceed with humane cull. After six hours, if mice score 12 or above, or score 3 for either 'neurological symptoms' or 'responsiveness to touch', then proceed with humane cull.  Table S2. Low-density PCR array data for FACS-sorted AAMs after injection into APAP-ALI or healthy mice (n=3 biological replicates per group). Average delta Ct to the mean value of the housekeeping genes (AVG ΔCt) indicated per group, transformed average delta Ct 2^(-ΔCt) indicated per group, the fold change, p-value, and fold-change (negative value indicates a downregulation compared to PBS control group).  Table S3. Primary antibody conditions used in immunodetection assays.

Table S4
Gene Name Table S4. Primer details used in qPCR assays.

Supplementary methods
Serum chemistry. Serum was isolated from whole blood after centrifugation (14,000 g) after clotting. Serum chemistry was performed by measurement of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), total bilirubin, and serum albumin. ALT was measured using a described method, [1] utilising a commercial kit (Alpha Laboratories Ltd). AST and ALP were determined by a commercial kit (Randox Laboratories).
Total bilirubin was determined by the acid diazo method described by Pearlman and Lee [2] using a commercial kit (Alpha Laboratories Ltd). Mouse serum albumin measurements were determined using a commercial serum albumin kit (Alpha Laboratories Ltd). All kits were adapted for use on a Cobas Fara centrifugal analyser (Roche Diagnostics Ltd). For all assays, intra-run precision was CV < 4%. In some experiments, assays were run on plasma samples.
Plasma microRNA measurement. RNA was isolated from 10 μL once-frozen plasma using miRNeasy serum/plasma kit (Qiagen) following the manufacturer's protocol. cDNA was prepared from 5 μL purified RNA using miScript II kit (Qiagen) with HiSpec buffers according to the manufacturer's protocol. cDNA was diluted 1:10 in RNase-free water before performing PCR in duplicate on a Roche Lightcycler 480 in 384-well format using miScript SYBR Green PCR Kit (Qiagen) using primer assays for murine miR-122 and following the manufacturer's protocol. Relative expression was determined using the 2− ΔΔCT method using the mean healthy Ct value obtained from each mouse (pre-APAP), with reference to let-7d, which served as housekeeping miRNA as used previously. [3] Immunodetection assays.   Serum and liver cytokines. Cytokines were quantified in serum or liver homogenate using a V-plex Proinflammatory Panel kit (Meso Scale Discovery) following the manufacturer's instructions. Serum (25 μL) or liver homogenate (100 μg total protein, in 12.5 μL) was diluted in the plate using buffers provided. Liver tissue was homogenized in lysis buffer (150 mM NaCl, 20 mM Tris, 1 mM EDTA, 1 mM EGTA, 1 % Triton X-100, 2 x protease inhibitor cocktail, (Sigma Aldrich)). Total protein was determined using BCA protein assay kit (Thermo Fisher). Plate was read on a QuickPlex SQ 120 analyzer (Meso Scale Discovery). Standards were assayed in duplicate and samples in singlet as recommended.

Haematoxylin and eosin staining and necrosis quantification. Four micron sections were stained with haematoxylin and eosin (H+E) on a Shandon Varistain Gemini ES Automated
Slide Stainer (ThermoScientific) and mounted using the Shandon ClearVue Coverslipper (ThermoScientific). For necrosis quantification, two methods were used at different times during the project. First method: H+E slides were scanned to create a single image with Dotslide VS-ASW software (Olympus) using a motorized stage and an Olympus BX51 microscope, acquiring images using an Olympus PlanApo 2X lens and Olympus XC10 camera.
Apoptotic thymocytes were labelled using CMTMR (Invitrogen) according to the manufactuer's instructions [8] . ) to label phagocytic cells as described [9] . At 36 hours, blood was collected from sacrificed mice in EDTA-tubes and processed immediately on a Celltac α analyzer (Nihon Kohden) for hematological analysis. Plasma was harvested from the remainder of the blood via centrifugation (3800 g, 10 min, 4 °C) and frozen. Liver was digested following methods previously described [10,11]   Hepatocyte labelling and confocal microscopy. R26RtdTomatoLSL male mice (Jackson laboratory) were administered AAV8-Tbg-Cre (2.5x10 11 , i.v., Addgene) to induce TdTomato expression specifically in hepatocytes. Mice were allowed a two week washout period before APAP administration and murine CFSE-labelled AAM delivery 16 hours later.
Immunofluorescent staining for mCherry (recognises tdTomato) and FITC was performed on FFPE liver sections to improve signal detection. Confocal images were obtained on a Leica SP8 confocal microscope using the navigator function in LASX software. Z-stack images were recorded at 40X magnification and presented as a single slice or a maximum projection over 7 slices representing 2.4 μm.
In vivo cell tracking. BMDMs were stained in vitro with VivoTrack 680 (near infra-red fluorescent imaging agent, Perkin Elmer) prior to transplantation. BMDMs (5x10 6 ) were incubated with reconstituted VivoTrack (2 mL) in 4 mL DPBS and incubated for 15 min at room temperature in a dark environment and subsequently washed before injection (i.v) to control or APAP-ALI mice at 16 hours. Anaesthetized mice were maintained with gaseous oxygen/isofluorane had abdominal fur clipped before fluorescent images were taken in the dorsal position at 15 min intervals using 687 nm excitation and 722 nm emission (background correction at 487 nm) on a PhotonIMAGER™ (Biospace Lab) imaging suite. After 4 hours mice were humanely culled and organs imaged ex vivo to confirm BMDM localisation. In some experiments, BMDMs differentiated from constitutively-expressing GFP cells [12] were used to confirm staining with Vivotrack 680 in vitro.