Macrophage Migration Inhibitory Factor on Apoptotic Extracellular Vesicles Regulates Compensatory Proliferation

Apoptotic cells can signal to neighboring cells to stimulate proliferation and compensate for cell loss to maintain tissue homeostasis. While apoptotic cell-derived extracellular vesicles (AEVs) can transmit instructional cues to mediate communication with neighboring cells, the molecular mechanisms that induce cell division are not well understood. Here we show that macrophage migration inhibitory factor (MIF)-containing AEVs regulate compensatory proliferation via ERK signaling in epithelial stem cells of larval zebrafish. Time-lapse imaging showed efferocytosis of AEVs from dying epithelial stem cells by healthy neighboring stem cells. Proteomic and ultrastructure analysis of purified AEVs identified MIF localization on the AEV surface. Pharmacological inhibition or genetic mutation of MIF, or its cognate receptor CD74, decreased levels of phosphorylated ERK and compensatory proliferation in the neighboring epithelial stem cells. Disruption of MIF activity also caused decreased numbers of macrophages patrolling near AEVs, while depletion of the macrophage lineage resulted in a reduced proliferative response by the epithelial stem cells. We propose that AEVs carrying MIF directly stimulate epithelial stem cell repopulation and guide macrophages to cell non-autonomously induce localized proliferation to sustain overall cell numbers during tissue maintenance.


INTRODUCTION
The ability to maintain epithelial tissue homeostasis has important implications for the health of multi-cellular organisms. Failure to adequately replace dead or missing cells can predispose epithelia to failed tissue maintenance, loss of barrier function [1][2][3] , and increased susceptibility to infection 4 . Alternatively, unchecked cell growth with minimal removal of defective cells can result in hyperplasia 5 , a hallmark of cancer. Studies in Drosophila have shown that dying cells are able to secrete mitogenic signals to neighboring cells to stimulate proliferation 6 . Similar effects were observed in Hydra through the transfer of Wnt3 from dying cells to facilitate regeneration 7 . While the link between dying cells and proliferation has long been established, what is not entirely clear is how signals are transmitted from the dying cell to neighboring cells. Recent work has proposed apoptotic bodies or apoptotic extracellular vesicles as transporters of mitogenic signals from dying cells to neighboring cells [8][9][10] .
Cells undergoing programmed cell death fragment into membrane bound vesicles that are approximately 1-5µm in diameter [11][12][13] , called apoptotic bodies 11 or apoptotic extracellular vesicles (AEVs), to prevent the contents from spilling out into the extracellular space. Similar to other extracellular vesicles, AEVs are enriched in contents that can regulate or interact with neighboring cells. For instance, AEVs can participate in the horizontal transfer of DNA 14,15 , microRNA 16 , splicing factors 17 , and biologically active proteins 18 . While apoptosis has been dogmatically regarded as an "immunologically silent" form of cell death eliciting minimal inflammation compared to necrosis 19 , damage associated molecular patterns (DAMPs) such as Hsp70 and HMGB1 have been observed in AEVs 20 . Yet, how the contents of AEVs may regulate the local microenvironment after cell death in living tissues remains poorly understood.
The technical challenges to perturb living epithelia in the presence of intact immune system and image subsequent changes in real time has thus far prevented a detailed characterization of how apoptosis can stimulate proliferation. Zebrafish larvae possess an experimentally accessible bi-layered epidermis that is similar in structure and function to those coating organ systems in mammals 21-23 , providing a system to rapidly interrogate the coordination of apoptosis and proliferation. The keratinocytes in the basal layer serve as the resident stem cell population that contributes to all of the strata in the adult epidermis 24 , and also express defined markers found in epithelial stem cells such as TP63 [25][26][27] . Our previous work showed that these basal stem cells contribute to the clearance of AEVs that stimulate their proliferation in the tail fin epidermis of zebrafish larvae 9 .
Here we performed proteomic analysis of purified AEVs in zebrafish and identified proteins associated with tissue regeneration and modulation of the immune system. We further characterized Macrophage Migration Inhibitory Factor (MIF) as a putative regulator of AEV-mediated signaling during epithelial tissue maintenance. MIF has been characterized as a cytokine, chemokine, and molecular chaperone 28 , and despite its name, plays a role in leukocyte recruitment [29][30][31] and proliferation and migration of epithelial cells 32 . These data suggest MIF is capable of exerting both mitogenic and immunogenic effects, yet how these are regulated during apoptosis and compensatory proliferation in vivo are not well understood.
This study investigates the role of MIF in apoptosis-induced proliferation of the basal epithelial stem cells in zebrafish larvae. We show that apoptotic extracellular vesicles (AEVs) have MIF on their surface, which stimulates proliferation in surrounding epithelial stem cells via the upregulation of phosphorylated ERK.
Moreover, our findings indicate that apoptosis stimulates increased mobilization of macrophages, but their contribution to AEV engulfment and clearance is minimal. Our data suggest that AEVs carrying MIF stimulate macrophages to participate in compensatory proliferation in a cell non-autonomous manner. Together, these findings highlight the dynamic interplay between AEVs, neighboring epithelial stem cells, and macrophages during resolution of cell death and maintenance of overall cell numbers.

Proteomic analysis of epithelial stem cell derived-AEVs (esAEVs) identifies proteins associated with wound healing and regeneration
We used a zebrafish model to induce death in a subset of the basal stem cells in the bi-layered larval epidermis 21,33 . The zc1036 GAL4 enhancer trap (BASAL-GET) line was used to drive mosaic expression of the bacterial enzyme nsfB, or nitroreductase (NTR) fused to mCherry 34 in basal epithelial stem cells 21 ( Figure 1A). After the addition of Metronidazole (referred to as MTZ) to 4-day post-fertilization (dpf) larvae for 4 hours ( Figure 1B), the NTR-positive cells convert MTZ into a cytotoxic byproduct that results in DNA damage 33 and apoptosis 35 . The dying epithelial stem cells display classic markers of apoptosis, such as increased activated-caspase 3 ( Figure   1C), and the formation of apoptotic extracellular vesicles (AEVs) in vivo ( Figure 1D, Supp. Movie 1). The dynamics of epithelial stem cell-derived AEV (esAEV) biogenesis was captured using time-lapse confocal imaging. Formation of esAEVs was observed within one hour of cell shrinkage and had an average diameter of 2.41 microns ( Figure 1E). Due to the mosaic nature of our genetic system ( Figure 1F'), we can visualize the interaction of dying epithelial stem cells with the remaining healthy neighbor cells ( Figure 1F'').  hours post-treatment (hpt) with MTZ, TP-63 positive epithelial stem cells were visualized engulfing AEVs and apoptotic cell corpses ( Figure 1F''', Supp. Movie 2). The most engulfment events were observed at 8 hpt ( Figure   1G), with an average of 15 individual epithelial stem cells engulfing esAEVs, and decreased over time. By 18hpt, we observed a significant increase in the number of actively proliferating cells that incorporated 5-bromo-2'deoxyuridine (BrdU) ( Figure 1H). These findings are in line with our previous studies that demonstrated that ~63% of engulfing epithelial stem cells went on to divide 9 and support the mitogenic potential of esAEVs.
To identify proteins associated with esAEVs that regulate compensatory proliferation, we purified esAEVs using differential centrifugation and performed proteomic analysis (Figure 2A). Particle size and concentration using tunable resistive pulse sensing identified a fraction enriched in ~2um AEVs ( Figure 2B), consistent with that observed in vivo ( Figure 1D-E). Liquid chromatography with tandem mass spectroscopy proteomic analysis of the isolated esAEVs identified 421 unique proteins when compared to extracellular vesicles isolated during homeostatic conditions with no apoptosis ( Figure 2C). Gene Ontological (GO) analysis defined 14 clusters that had an enrichment score greater than or equal to 1.3, with the highest enrichment scores being pathways involved in cell metabolism ( Figure 2D). This analysis also showed that esAEVs are enriched in proteins involved in biological processes such as regeneration, cell-cell junction organization, and lipid modification. We also found several major Damage Associated Molecular Patterns (DAMPs) 36-38 such as heat-shock proteins (Q90473, and Q645R1), calreticulin (Q6DI13), protein S100 (Q6XG62), and histones H2A and H4 (Q0D272, E7FE07). We also identified proteins such as Angiosinogen (Q502R9), Pro-epidermal growth factor (EGF) (B3DH82), Low-density lipoprotein receptor-related protein (LRP1) (A0A8M2B922) and Galectin-3 (Q6TGN4) which are known to be involved in pathways pertaining to cell proliferation or cell growth (Supplemental Table 1). Finally, we assessed the list for proteins that could play a dual role in stimulating proliferation in epithelial stem cells and regulate inflammatory responses from the immune system. This led to the identification of macrophage migration inhibitory factor (MIF) (F6PCE0). In sum, these data provide new insights into the protein components of esAEVs and their potential role in facilitating compensatory proliferation.
MIF is an attractive target for investigation due to its role in stimulating proliferation, inflammation, and immune cell dynamics. Therefore, we focused our efforts on characterizing the role of MIF in esAEV-mediated signaling. To validate the presence and localization of MIF on purified esAEVs, we assayed for MIF using immunogold labeling transmission electron microscopy. Annexin V served as a control to detect externalized phosphatidylserine (PS) on the surface of esAEVs. We observed no background staining when the gold nanoparticles were administered alone (0.000 +/-0.000). Anti-MIF nanoparticle labeling was applied, it was found to be localized on the surface of esAEVs (28.90 +/-4.373 nanogold particles) at levels comparable to Annexin V (29.34 +/-4.161 nanogold particles) ( Figure 2E). In contrast, the MIF ortholog D-DT (also referred to as MIF-2) displayed little to no detectable nanoparticle localization (Supp. Figure 1). Together, these data suggest that MIF localizes on the surface of esAEVs and could initiate MIF signaling to surrounding epithelial and immune cells.

esAEVs transporting MIF play a role in regulating epithelial stem cell proliferation
To investigate the role of MIF signaling in apoptosis-induced proliferation and esAEV signaling, we used a combination of pharmacological and genetic approaches to disrupt MIF and its cognate receptor CD74 ( Figure   3A). The zebrafish genome contains one copy of mif 39 , and two copies of the receptor genes, termed cd74a and cd74b 40 (Supp. Figure 2). Larvae expressing NTR in epithelial stem cells (NTR+) were treated with MTZ to induce apoptosis, treated with DMSO (vehicle control), ISO-1 41 , or 4-IPP 42 to block MIF signaling, and then incubated with BrdU to label the dividing cells. A compensatory proliferative response was observed with induced apoptosis (MTZ treatment alone compared to DMSO), however we observed a statistically significant decrease in proliferation when induced apoptosis was combined individually with either ISO-1 or 4-IPP treatment to inhibit MIF signaling ( Figure 3B-C). To complement the pharmacological approach, we used CRISPR/Cas-9-mediated mutation of mif, cd74a, and cd74b. In the F0 generation of mif, cd74a, and cd74b CRISPR-deleted animals (Supp. Figure 2A-L), or "crispants" 43,44 , we observed a statistically significant decrease in the number of BrdU positive epithelial stem cells ( Figure 3D-F). In contrast, F0 larvae injected with guide RNAs for tyrosinase that disrupt pigmentation showed no statistical change in proliferation after induction of apoptosis (Supp. Figure 3A-B). These data support a role for the MIF/CD74 signaling axis in esAEV-induced proliferation.
Our data supports the idea that MIF on the surface of AEVs plays a role in stimulating proliferation to replace cells lost by apoptosis. In addition to its presence on other EV populations 45,46 , MIF has been shown to be either cytosolic or secreted 47 . To test if secreted MIF is contributing to apoptosis-induced proliferation, we first overexpressed human MIF fluorescently tagged with GFP and did not observe a change in epithelial stem cell proliferation (Supp. Figure 5E). Next, we treated larvae with Brefeldin A to disrupt proteins that are secreted through the E.R and found no significant changes in esAEV formation or compensatory proliferation (Supp. Figure 6A-B). MIF can also be secreted via a non-classical secretion pathway that does not involve targeting to the ER 48,49 , and is transported via ABC transporters 50 . Therefore, we also treated larvae with glyburide, a compound that inhibits MIF secretion in THC-1 cells by targeting ABCA1 transport 51 . Glyburide is well tolerated in larval zebrafish without causing alterations to development 52 . After treatment with 25uM glyburide combined with MTZ, we found no significant difference in the number of esAEVs produced or number of BrdU positive epithelial stem cells (Supp. Figure 6C-D). Taken together, these data suggest that MIF delivery on AEVs, rather than by secretion, contributes to compensatory proliferation by the basal epithelial stem cells.

esAEVs carrying MIF activate ERK signaling in epithelial stem cells
We next sought to test if the basal epithelial stem cells express both MIF and the CD74 receptors. Hybridization chain reaction (HCR) fluorescent in situ hybridization 53 for mif in TP63:GFP transgenic animals showed high levels of expression throughout the basal epithelial cells ( Figure 4A). In contrast, we observed cd74a and cd74b were expressed at lower levels within the basal epithelial cells than mif, with cd74a showing a higher level of expression than cd74b ( Figure 4B-D). Given that MIF has also been implicated in regulating immune cell function, we also examined expression within macrophages using the mpeg1:GFP transgenic line 54 . Intriguingly, cd74a and cd74b also appear to be expressed in macrophages ( Figure 4E-H). Further, cd74a and cd74b also localize to macrophages that infiltrate to sites of injury/amputation (Supp. Figure 4A-C). Together, this indicates that esAEVs carrying MIF can signal to CD74a or b on both epithelial stem cells and macrophages.
To determine if MIF signals through CD74 to activate downstream ERK signaling 55 , we analyzed levels of phosphorylated ERK (p-ERK) with and without induced apoptosis, and after disruption of MIF/CD74 signaling.
Six hours post esAEV induction, we observed an overall increase in p-ERK fluorescence after induced apoptosis when compared to the DMSO control ( Figure Figure 7A-B). Taken together, these data indicate that esAEVs carrying MIF act through interaction with CD74a/b to upregulate p-ERK signaling to stimulate proliferation of epithelial stem cells.

MIF plays a role in macrophage surveillance activity that contributes to compensatory proliferation in a cell-non-autonomous manner
Inhibition of MIF did not completely decrease apoptosis-induced proliferation, suggesting that other contents of esAEVs play a role in affecting proliferation or there may be another microenvironmental factors contributing to esAEV-induced proliferation. As the name implies, MIF has also been implicated in regulating the migration of macrophages 56,57 . This is in line with our observed expression of cd74a and cd74b in macrophages ( Figure Figure 8A-B), with kinetics that could not support the complete clearance of AEVs from the tissue in this timeframe. These data suggest that macrophages play additional roles beyond solely the clearance of apoptotic corpses and AEVs. Treatment with either 4-IPP or ISO-1 results in decreased macrophage patrolling activity post-induction of apoptosis and AEV formation, with 4-IPP having a stronger suppressive effect than ISO-1 ( Figure 6B). The observed decrease in proliferation after ISO/4IPP treatment suggest that the macrophages may also contribute to epithelial stem cell proliferation in a cell non-autonomous manner.
To test if the presence of the immune system contributed to apoptosis-induced proliferation, we treated larvae with the anti-inflammatory dexamethasone to suppress immune cell infiltration after induced apoptosis 58 .
Dexamethasone suppressed the number of patrolling macrophages associated with epithelial stem cell apoptosis ( Figure 6C-E). Suppression of macrophage infiltration using dexamethasone was also accompanied by significantly decreased compensatory proliferation of TP63+ epithelial stem cells ( Figure 6F). To further test this idea, we used antisense morpholino oligonucleotides (MO) to target IRF8, the transcription factor that initiates the myeloid lineage and macrophage formation in early development 59 . irf8 MO larvae had significantly decreased number of macrophages (Supp. Figure 8C-E), along with reduced epithelial stem cell proliferation after induced apoptosis ( Figure 6G). Together, these data support the idea that the presence of macrophages contributes to compensatory epithelial stem cell proliferation after induced apoptosis and AEV formation. Overall, these studies suggest that AEVs play an important role in regulating stem cell proliferation and tissue regeneration. Our current studies extends these findings further by providing an in vivo assessment of AEV activity in conjunction with an intact innate immune system.

DISCUSSION
A key question from these studies is how putative signals are transferred from apoptotic cells to neighboring healthy cells to initiate compensatory proliferation. Our studies suggest a critical role for surfacelocalized MIF on AEVs in mediating intercellular communication from dying epithelial stem cells to healthy neighboring epithelial stem cells and macrophages. Extracellular vesicles, including both various EV subtypes and exosomes, can transfer DNA, mRNA, and proteins to facilitate short-range communication between cells 63 .
Mesenchymal stem cell exosomes carrying MIF have also been shown to enhance myocardial repair, promote angiogenesis and reduce fibrosis in the heart 64 . Similarly, exosomal MIF from nasopharyngeal carcinoma promotes metastasis by enhancing macrophage survival 46 . These findings suggest that AEVs carrying MIF also play important roles in regulating diverse biological processes, including immune responses, tissue repair, and cancer progression.
The interaction between the cytokine MIF and its receptor CD74 can trigger downstream signaling cascades such as ERK1/2 to drive a change in cellular behaviors including proliferation, migration, and survival.
In our study, we showed that AEVs promote an increase in p-ERK signaling in neighboring epithelial stem cells to drive proliferation. MIF has also been shown to drive proliferation and migration of airway muscle cells 65 , spermatogonial cells 66 , dendritic cells 67 in an ERK1/2 dependent fashion. While we did not observe a change in p-ERK in macrophages, other studies have shown that EVs carrying MIF can stimulate p-ERK in macrophages 68 , promoting their activation and modulating their immune responses. A possible explanation for this is that CD74 can recruit additional co-receptors, such as CXCR2 29 or CXCR4 69 , instead of CD44 70 in a MIF dependent manner and activate other downstream signaling events such as PI3K/Akt, and calcium dependent integrin activity 29 .
What mediates the particular co-receptors that CD74 recruits is not well understood, and future characterization of these interactions may inform which downstream pathways are taking place in macrophages to facilitate compensatory proliferation.
Apoptosis has traditionally been regarded as an immunologically silent form of cell death that resolves with minimal induction of inflammation 71 . The rapid externalization of PS during apoptosis serves as an 'eat-me' signal and exhibits anti-inflammatory properties 72 . DAMPs are released during apoptosis and cellular stress events 73 , calling this previously held dogma into question 38 . Intriguingly, AEVs can also transport DAMPs such as HMGB1 20 , a molecule involved in tissue repair 74 . Other extracellular vesicle populations such as small EVs and microvesicles can carry a variety of DAMPs 36 , with HSP70 being localized on the surface of exosomes 75 . MIF can be secreted by a variety of cell types in response to tissue damage, infection, and other forms of stress, and has been shown to have pro-inflammatory and immune-regulatory effects 47,76 . We did not observe an impact on proliferation with pharmacological inhibition of MIF secretion, or with genetic overexpression of human MIF, suggesting that apoptosis is critical for the ability of MIF to promote a compensatory proliferation response. Therefore, MIF may represent a novel DAMP with unique properties and functions.
Our studies indicate that macrophages contribute to AEV-mediated compensatory proliferation. We observed expression of cd74a and cd74b in macrophages, and established a role for AEV delivered MIF in promoting macrophage surveillance activity after induction of apoptosis. Importantly, dampening of inflammation via treatment with dexamethasone or depletion of the macrophage population decreased proliferation, suggesting a cell-non autonomous mechanism for the stimulation of proliferation during the re-establishment of epithelial tissue homeostasis. This is echoed by studies showing that macrophages play a key role in facilitating tissue repair 77 , and that attenuation of macrophages impairs wound healing and tissue regeneration [78][79][80][81] .
Apoptotic cells can prime macrophages toward a more "pro-regenerative" M2 phenotype 82 , and M2 macrophages are known to secrete a variety of signals following injury 78 . Recent work has highlighted that MIF can induce macrophages to secrete mmp-9 downstream of ERK1/2 signaling 83 . The types of signals macrophages secrete in response to AEV's carrying MIF remains unclear. Identification of the signals produced by macrophages will be an interesting topic for future studies of compensatory proliferation.
In summary, our studies define a role for MIF carried by AEVs in the reestablishment of tissue homeostasis in a dynamic process that engages both epithelial stem cells and macrophages. We propose that AEVs carrying MIF play a dual role in sustaining homeostatic cell numbers, to directly stimulate epithelial stem cell repopulation and guide macrophage behavior to cell non-autonomously contribute to localized proliferation.

Zebrafish handling and husbandry
Adult Zebrafish were maintained at the MD Anderson Cancer Center fish facility in accordance with the institutional guidelines and best practices for animal care. Zebrafish embryos were maintained at 28°C in E3 medium.

Transgenic lines
The To assess proliferation or recovery phenotypes, larvae were subjected to an 18-hour recovery time prior to bromodeoxyuridine (BrdU) incorporation. After the recovery period, BrdU was administered to label dividing cells, allowing us to analyze proliferation and recovery phenotypes in response to stem cell ablation.

Isolation of esAEVs and quantitative analysis
After inducing apoptosis in epithelial stem cells in 4dpf larvae, a combination of mechanical dissociation, trypsonization, and centrifugation were used to isolate esAEVs. After a 4-5 hour treatment with MTZ, 150-200 Basal-GET NTR+ larvae were transferred to a 1.5mL eppendorf tube, and washed once with tissue-culture grade 1X PBS. Once the larvae settled in the tube, 1mL of pre-warmed 37°C Trypsin was added to the tube. A scalpel was used to chop the larvae for roughly 60-90 seconds. The larvae were then placed on a nutator to rock gently for 10 minutes. Chopping and placement on the nutator was repeated twice more. The larvae were then placed in a 4°C centrifuge for 10 minutes at 650 x g. The supernatant was transferred to a new tube and centrifuged for 2 minutes at 14,500 x g, 4°C to pellet large cells. The supernatant was transferred to a new tube and centrifuged for 1 hour at 14,500 g, 4°C . In this study, tunable-resisitive pulse sensing (qNano, Izon) was used to quantify the size and concentration of esAEVs. For the qNano, CPC 2000 calibration particles were used with an NP 2000 pore. esAEVs were diluted 1:50 in measurement electrolyte prior to measuring the size and concentration.
Look to Figure 2A  Gene ontology analysis of proteins unique to esAEVs GO enrichment analysis was performed using DAVID 84 . The 421 unique proteins to esAEVs were uploaded to DAVID, where the following criteria was applied: Using the highest stringency for biological processes, a total of 72 clusters with different enrichment scores were returned. Applying an exclusion criterion of no less than 1.3 for enrichment scores, 14 clusters were appropriate for analysis. Within each cluster, terms with fold enrichments greater than 1.5 and the lowest false discovery rate (FDR) were further selected as biologically relevant.

Transmission electron microscopy and Immunogold labeling
With the assistance of the M.D. Anderson Cancer Center Electron Microscopy Core Facility, isolated AEVs were submitted for immunogold labeling. AEVs were fixed in 2% Glutaraldehyde in 0.1M PBS, pH 7.5. In order to visualize the membrane, esAEVs were whole-mounted. During the whole mount, AEVs were mildly permeabilized with saponin to expose surface antigens. Primary antibodies were administered at the following Human Anti-CD74 antibody (BioLegend, 326802) was stored at 4°C, and added to E3 at a 1:400 dilution.

Microscopy
A Zeiss LSM800 Laser Scanning confocal microscope was used for movie and image acquisition. Images were acquired as z-stacks using 20x for tail tips. 10x objective was used to acquire tiled region images of entire larvae and stitched together to present a full image. A Zeiss Axiozoom fluorescent microscope was used for acquisition of images for the BrdU immunohistochemical analyses. All microscopy images were processed using Zen software. Any adjustments made to brightness and contrast were applied consistently across images. The resulting images were then used for further analysis and quantification.

Larvae fixation and fluorescent immunostaining
After treatment or recovery, larvae were fixed in 4% formaldehyde (Sigma) in 0.05% PBST. Larvae were left on gentle rocking overnight at 4 degrees. The fixative was washed out using 6 x 5 minute washes in PBST 0.5%. BrdU detection was done by adding 2N Hydrochloric acid in ddH2O for 45 minutes under gentle rocking at room temperature. The HCL was washed out with 6 x 5 minute washes in PBST 0.5%. Blocking was performed using 10% goat serum in blocking buffer for 1-2 hours. Antibodies were diluted in blocking solution, and used to stain larvae overnight in 4°C. The primary antibody was washed out with 6 x 20 minute PBST 0.5% washes. Larvae were blocked 1-2 hours before the secondary antibody was added. Larvae were left overnight in the dark at 4 degrees. The secondary was washed out using 6 x 20 minute PBST 0.5% washes. Body parts just above the cloaca were removed and discarded during mounting. 80% glycerol in PBS was used to preserve the fluorescence of larvae.

BrdU incorporation and detection
BrdU incorporation was performed using 10mM BrdU, and 5% DMSO in E3 at 18 hpt for 45 minutes. After incorporation, animals were washed 3x with E3, left to recover for 45 minutes in E3 alone, and then fixed in paraformaldehyde overnight at 4°C. We detected BrdU positive cells after cell permeabilization using 2N HCL for 45 minutes. The staining protocol proceeded with applying monoclonal rat anti-BrdU primary antibody followed by the appropriate secondary antibody.
Immunostaining and Quantification of p-ERK signal 4dpf Basal-GET larvae were drug treated with MTZ, and different agents to perturb MIF activity for 4 hours.
Larvae were fixed 6 hours post treatment, and stained for p-ERK (1:200 dilution) (Cell Signaling, Antibody #9101). All p-ERK signal was visualized in the far-red channel as this has the least autofluorescence in the presence of apoptotic cells. The tail epithelium of the larvae were imaged with the exact same imaging parameters. 100um x 100 um regions of interest were drawn around NTR-mCherry cells in control and damage conditions. Tails had a maximum of three ROIs. The mean fluorescent intensity was recorded across a z-stack within each ROI.

Fluorescent in situ hybridization
Probes for the in situ hybridization were purchased from Molecular Instruments, inc. The genes and the accession numbers provided to Molecular Instruments are as follows: MIF was detected using a B1 amplifier while CD74a and CD74b were detected with a B3 amplifier. Larvae were stained according to manufacturer detection and staining procedures 53 , with modifications outlined by Ruiz et al. 85 All larvae were imaged in the exact same conditions to allow for comparisons between probes, and all imaging was performed using the far-red channel to minimize background fluorescence.

Generation of CRISPR-edited larvae
The program CHOP-CHOP 86,87 was used to design highly specific guides directed against MIF, CD74a, and CD74b. 1-cell stage embryos were injected with two guides for each gene with Cas9 protein (NEB, M0646T).
Tyrosinase was an injection control to validate the efficacy of the Cas9 protein. Embryos were raised to 4dpf for further experimentation. We validated the efficiency and accuracy of Cas9 by using Sanger sequencing, followed by TIDE 88

Morpholino Oligonucleotides
The morpholinos used in this study were obtained from GeneTools, and the sequences were obtained from ZFIN (https://zfin.org/). The injection scheme for Irf8 was modeled after Madigan et al. 90 A GeneTools 25 nucleotide scrambled (Random) control oligo was used as an injection and morpholino control. Tracking and quantification of macrophage movement Using a Basal-GET line crossed to Tg(mpeg:GFP), macrophages were imaged every five minutes using a Zeiss LSM 800 confocal microscope. The number of macrophages outside the notochord and present in the epithelial tissue were manually counted every hour, and graphed over time using GraphPad Prism. Imaris was used to track the overall movement of macrophages over an 8-10 hour time period.

Software
Zeiss Zen Blue was used to analyze the mean fluorescent intensity of p-ERK signal, and fluorescent in situ hybridization. Zen blue was also used to quantify the number of BrdU+ nuclei. GraphPad Prism (version 9) was used for statistical analysis and graph generation. DAVID was used for protein Gene Ontology analysis. Imaris (version 9) was used to track the movement of macrophages in the larval zebrafish tail epithelium.

Statistical analysis
Statistical analysis was performed using GraphPad Prism version 9. A t-test was used to assess significance between two groups, and an ANOVA was used to determine the significance between three or more groups with a Tukey test as the ad hoc analysis. Results are reported with the standard error of the mean unless specified.
A p-value less than 0.05 was considered statistically significant.