Expression and secretion of galectin‐12 in the context of neutrophilic differentiation of human promyeloblastic HL‐60 cells

Galectin‐12 is a tissue‐specific galectin that has been largely defined by its role in the regulation of adipocyte differentiation and lipogenesis. This study aimed to evaluate the role of galectin‐12 in the differentiation and polarization of neutrophils within a model of acute myeloid leukemia HL‐60 cells. All‐trans retinoic acid and dimethyl sulfoxide were used to induce differentiation of HL‐60 cells which led to the generation of two phenotypes of neutrophil‐like cells with opposite changes in galectin‐12 gene (LGALS12) expression and different functional responses to N‐formyl‐ l‐methionyl‐ l‐leucyl‐ l‐phenylalanine. These phenotypes showed significant differences of differentially expressed genes on a global scale based on bioinformatics analysis of available Gene Expression Omnibus (GEO) data sets. We also demonstrated that HL‐60 cells could secrete and accumulate galectin‐12 in cell culture medium under normal growth conditions. This secretion was found to be entirely inhibited upon neutrophilic differentiation and was accompanied by an increase in intracellular lipid droplet content and significant enrichment of 22 lipid gene ontology terms related to lipid metabolism in differentiated cells. These findings suggest that galectin‐12 could serve as a marker of neutrophilic plasticity or polarization into different phenotypes and that galectin‐12 secretion may be influenced by lipid droplet biogenesis.


| INTRODUCTION
Galectins are a family of soluble proteins initially characterized by their binding affinity towards β-galactoside-containing glycans (Johannes et al., 2018).Galectin-12 is a tissue-specific galectin that has previously been shown to play a role in the differentiation and/or polarization of various cell types including adipocytes, sebocytes, macrophages, and neutrophils (Harrison et al., 2007;McTague et al., 2022;Tsao et al., 2022;Wan et al., 2016;Xue et al., 2016;Yang et al., 2004).It is well-established that galectin-12 is required as a lipogenic factor for the proper differentiation of adipocytes and sebocytes engaging signaling pathways associated with lipogenesis in human, mouse, and porcine models (Wu et al., 2018;Yang et al., 2004Yang et al., , 2011)).The function of galectin-12 in neutrophil and macrophage differentiation is less studied but suggests its role as a modulator of cell plasticity.For instance, galectin-12 gene (LGALS12) knockdown in human acute promyelocytic leukemia NB4 cells led to an increase in all-trans retinoic acid (ATRA)-induced neutrophil-like differentiation (Xue et al., 2016), while genetic knockout of galectin-12 in mice was found to promote M2 polarization of macrophages (anti-inflammatory and tumorigenic phenotype) from bone marrow in response to lipopolysaccharide (Wan et al., 2016).Along these lines, we have previously reported that the expression of LGALS12 in human promyeloblastic HL-60 cells can differ depending on the inducer of neutrophilic differentiation, namely, either ATRA or dimethyl sulfoxide (DMSO); however, the cell biological significance and molecular signature of these phenotypes were not addressed (McTague et al., 2022;Vinnai et al., 2017).This aspect deserves special attention because similar to M1/M2 polarization of macrophages, a paradigm of functionally diverse phenotypes of neutrophils (N1 and N2) has been developed for tumor-associated cells (Fridlender et al., 2009) and furthermore, a recognition of the diversity of neutrophils is an emerging concept in the context of cancer therapy (Quail et al., 2022).
Remarkably, the neutrophilic differentiation of human NB4 promyelocytic leukemia cells was found to be accompanied by an increase in intracellular lipid droplet content (Xue et al., 2016).
Lipid droplets can be used in neutrophils as stores for cytokines and the number of lipid droplets increases in leukocytes upon inflammation in the body (Melo & Weller, 2016).Galectin-12 protein exhibits a highly hydrophobic profile and is specifically located intracellularly in lipid droplets (Maller et al., 2020;Yang et al., 2011) that might prevent its secretion.All galectins lack a signal sequence for export through the endoplasmic reticulum-Golgi system and are therefore secreted out of cells through unconventional mechanisms including direct translocation, release in exosomes, microvesicle shedding, and secretory autophagy among others (Popa et al., 2018).In fact, there are currently no studies specifically investigating the secretion of galectin-12 and only negligible secretion has been observed in 3T3-L1 adipocytes (Hotta et al., 2001;Yang et al., 2011), although there have been low nanogram levels of this protein detected in circulation (Galdino et al., 2021;Nowowiejska et al., 2023).
In the present study, we used a model of ATRA-and DMSOinduced differentiation of HL-60 cells (Sham et al., 1995)  were purchased from ThermoFisher Scientific.

| Cell culture and cell treatments
Human acute promyeloblastic leukemia HL-60 cells and mouse preadipocyte 3T3-L1 cells were obtained from ATCC and cultured in IMDM and DMEM, respectively, supplemented with 10% fetal bovine serum, 50 IU/mL penicillin, and 50 μg/mL streptomycin at 37°C and 5% CO 2 .To induce neutrophilic differentiation, HL-60 cells were treated with either 1.3% DMSO or 1 µM ATRA for 72 h starting at a concentration of 0.4 × 10 6 cells/mL (McTague et al., 2022;Vinnai et al., 2017).To induce adipocytic differentiation, 3T3-L1 cells were grown in DMEM until confluency was reached.The cells were then maintained for an additional 48 h to achieve contact-induced growth inhibition and reach a "postconfluency" state.Once postconfluent, a differentiation cocktail was added to cell medium at final concentrations of 0.5 mM 3-isobutyl-1-methylxanthine, 1 µM dexamethasone, and 10 µg/mL human insulin (Yang et al., 2004).After 48 h, the differentiation medium was replaced with postdifferentiation medium of complete DMEM supplemented with 10 µg/mL human insulin.An inhibitory analysis was used to examine the functionality of pathways associated with unconventional secretion of galectins: HL-60 cells were treated with inhibitors of exosome formation (GW 4869, 20 µM) (Hekmatirad et al., 2021), microvesicle formation (Y-27632, 10 µM) (Sapet et al., 2006), and secretory autophagy (3-MA, 2 mM) (Davuluri et al., 2021) for 48 h followed by collection of cell culture media for testing galectin-12 secretion levels.with Fourier descriptors set up as 7 (the relative proportions) and 2 (the absolute number).Before morphological analysis, apoptotic, overlapping, and poorly contrasted nuclei were removed together with any cellular debris while contacting nuclei were carefully separated using the paintbrush tool (Supporting Information S1: Figure S1).The data on circularity ( π 4 ×area perimeter 2 ) and roundness ( π 4 × area × major_axis 2 ) of nuclei were obtained using Analyze > Analyze particles ImageJ command with the size exclusion threshold set up at a minimal size unit of 0.01.Measurements were generated for each nucleus ranging from 0 to 1, where 1 indicated a perfect circularity and roundness.

| Staining and analysis of lipid droplets
Two methods were used to stain lipid droplets in cells including fluorescence staining with BODIPY 493/503 (Qiu & Simon, 2016) and histological staining with ORO (Xue et al., 2016).While these methods were directly applied to adherent 3T3-L1 cells, we needed to develop and introduce an additional step for suspension HL-60 cell culture.Namely, HL-60 cells from suspension were adhered to culture dishes coated with Con A as previously described for cell adhesion assay (Timoshenko et al., 2016).Briefly, the cell culture dishes (∅ 30 mm) were treated with 2 mL of Con A solution (50 µg/mL in Ca 2+ /Mg 2+ -free D-PBS) overnight at 4°C, then rinsed with 2 mL D-PBS, and HL-60 cells (1.0 × 10 6 in 2 mL of complete medium) were added for 1 h at 37°C, followed with another brief rinse to remove nonadherent cells.Histological staining of lipid droplets with ORO was performed only for 3T3-L1 cells, which were rinsed twice with D-PBS and fixed with 4% paraformaldehyde in D-PBS for 30 min at room temperature.
Cells were then rinsed twice with distilled water and incubated in a 60% 2-propanol solution at room temperature for 5 min, which was replaced with ORO solution (3 mg/mL in 60% 2-propanol) for 10 min (Kinkel et al., 2004).The cells were then washed with distilled water five times before imaging under a Leica DM IL LED brightfield inverted microscope using a Leica EC3 camera and LAS software.

| RNA isolation, cDNA synthesis, and polymerase chain reaction (PCR) assay
Total RNA was isolated using TRIzol™ and quantified using a Nanodrop 2000c UV-Vis spectrophotometer (ThermoFisher Scientific) with a A 260/280 of >1.8 as a quality threshold.Reverse transcription of 500-1000 ng RNA was conducted using the Advanced cDNA Synthesis Kit.Quantitative PCR (qPCR) was conducted using the CFX Connect real-time PCR system (Bio-Rad) with a two-step cycling regime (5 s at 95°C and 25 s at 62°C) as described elsewhere (McTague et al., 2022).The reaction mix (20 µL) consisted of 10 µL of SsoAdvanced Universal SYBR™ Green Supermix (Bio-Rad), 1 μL of primer supermix (10 µM forward/reverse primers), 1 µL of 2-4× diluted cDNA template, and 8 μL of nucleasefree water.PCR primer sequences were verified by BLAST and oligonucleotides were synthesized at the UWO BioCorp OligoFactory or Invitrogen Life Technologies (Table 1).Relative mRNA levels were determined by the Livak (2 −ΔΔC t) method (Livak & Schmittgen, 2001) using ACTB and Gapdh as reference genes for HL-60 and 3T3-L1 cells, respectively.The average of all untreated control samples was taken and used as the calibrator to determine the relative expression of the target gene in treatment samples.

| Flow cytometry
HL-60 cells were differentiated with ATRA or DMSO for 72 h and the cell concentration and viability were assessed using the trypan blue exclusion test.Cells were washed with D-PBS and stained with 2 µM BODIPY 493/503 for 15 min at 37°C in 4 mL cell suspension (1.0 × 10 6 cells/mL) or left unstained to serve as negative controls.
The cells were washed three times with and resuspended in fluorescence-activated cell sorting buffer (Ca 2+/ Mg 2+ -free D-PBS, 1% of 0.5 M ethylenediaminetetraacetic acid, 5 µg/mL bovine serum albumin) at 1.0 × 10 6 cells/300 µL.SYTOX Blue (1 µM) was added to the final suspension to act as a dead cell indicator.Cells were analyzed using a CytoFLEX S flow cytometer (Beckman Coulter) at the London Regional Flow Cytometry Facility (Robarts Research Institute, Western University, Ontario) and the results were analyzed using FlowJo TM v10.8.1 Software (BD Life Sciences).
The scopoletin fluorescence was measured using an AMINCO-Bowman Series 2 Luminescence Spectrometer (SLM AMINCO) with an excitation wavelength of 350 nm and emission wavelength of 460 nm.The sample was maintained at 37°C and once a stable fluorescence signal was achieved, an activator of the phagocyte NADPH oxidase either PMA (1 µM) or fMLP (100 nM) was added to induce respiratory burst in the cells.The rate of hydrogen peroxide generation was calculated as the maximum slope of scopoletin oxidation traces using RStudio (version 1.4.1717)(Vinnai et al., 2017).

| Bioinformatics analysis
RNA-Seq data for ATRA/DMSO-differentiated HL-60 cells were obtained from the GEO accessions GSE93996 and GSE103706.
Reads were aligned to Genome Reference Consortium Human Build 38 release 108 genome assembly using STAR version 2.7.10a (Dobin et al., 2013) in two-pass mode with annotation, allowing only unique mapping, and up to five mismatches per read mapped.Per gene read counts were generated using htseq-count (HTSeq framework version 1.99.2) in "union" mode (Anders et al., 2015).All subsequent analysis was conducted using R. Before differential gene expression analysis data sets were merged by ComBat for RNA-Seq counts using R package sva (Johnson et al., 2007;Leek et al., 2012).Genes expressed at the level at or above one fragment count per million in at least three samples were considered for the subsequent analysis.Differential gene expression analysis and estimation of log 2 fold changes and false discovery rate (FDR) adjusted p values (Benjamini & Hochberg, 1995) were conducted using the voom method (Law et al., 2014).Changes in gene expression were deemed significant at FDR-adjusted p < 0.01 and log 2 fold change of ±1 cutoffs.Gene Ontology (GO) enrichment analysis was performed using topGO R package with the "classic" and "weight01" algorithms and a weighted p < 0.05 cut-off to generate the top GO terms for biological processes (BP), molecular function (MF), and cellular components (CC) for the list of DEGs (Alexa et al., 2006).Gene Set Enrichment T A B L E 1 Sequences and properties of polymerase chain reaction primers and amplicons.Analysis (GSEA) was performed using piano and fgsea R packages (Korotkevich et al., 2021;Väremo et al., 2013).GSEA was conducted for BP, MF, and CC terms and visualized as network plots comparing ATRA-and DMSO-differentiated cells.Only GO terms with a minimum of 20 associated genes were considered for analysis.GSEA was also conducted on a subset of all lipid-related GO terms, and normalized enrichment score was reported with significance cut-off at FDR-adjusted p < 0.05 (Liberzon et al., 2011).
One-way or Kruskal-Wallis analysis of variance followed respectively by Tukey's/Dunnett's or Dunn's multiple comparison tests were used for comparison between means considering the threshold of statistically significant differences at p < 0.05.Data were presented as means ± standard deviation for a minimum of three biological replicates for each experiment.

| Bioinformatics analysis of differentially expressed genes between ATRA-and DMSO-induced phenotypes of HL-60 cells
Two established models to study neutrophil-like differentiation of HL-60 cells use treatments with ATRA or high doses of DMSO; however, the functional and genetic differences between these phenotypes are not sufficiently explored (Mar & Quackenbush, 2009).
To gain more insights into the genomic diversity of these phenotypes, PMA (a cell-permeable activator of protein kinase C).As was evident with DAPI staining, both ATRA and DMSO induced a lobular nuclear morphology versus the oval/rounded shape seen for the nuclei of the undifferentiated control cells (Figure 2a).The quantification of the nuclear shape using ImageJ revealed significant (p < 0.05) decrease in both the circularity and roundness of the nuclei within the treatment groups in comparison to control cells (Figure 2b).ATRA and DMSO also induced similar significant upregulation of NCF1 and NCF2 genes (Figure 2c) encoding essential components of the phagocyte plasma membrane NADPH oxidase complex, namely, p47phox and p67phox (Belambri et al., 2018).However, the difference between the cell phenotypes was observed when the respiratory burst was examined: while PMA stimulated efficient generation of H 2 O 2 by both phenotypes signifying their neutrophilic differentiation (Figure 2d), fMLP was only able to induce this response in DMSO-treated cells (Figure 2e).To disclose this difference, we analyzed the expression of genes encoding N-formyl peptide receptors (FPR1, FPR2, and FPR3) in HL-60 cells and observed that only FPR1 was expressed and significantly upregulated by DMSO (10.1-fold, p < 0.01) and not ATRA (p = 0.5398) as per qPCR assay (Figure 2f).neutrophil-like differentiation with DMSO led to a significant drop (3.7-fold, p < 0.01) (Figure 2g).Remarkably, the magnitude of these changes in the expression of LGALS12 was much smaller than in 3T3-L1 mouse cells undergoing adipocyte differentiation, which was used as a well-established positive control and experimental system to study galectin-12 expression and functions (Hotta et al., 2001;Yang et al., 2004Yang et al., , 2011Yang et al., , 2016)).Indeed, Lgals12 demonstrated significant 2905-fold upregulation (p < 0.001) observed at day 4 of adipocyte differentiation when comparing to subconfluent and overconfluent cells (Figure 2h), which was congruent with a significant upregulation (204-fold, p < 0.001) in the expression of adipocytic differentiation marker Fabp4 in this case (Figure 2i).Thus, the involvement of LGALS12 in regulating cellular differentiation can vary and our results with HL-60 cells suggest a different implication of LGALS12 as a marker of specific subtypes of neutrophilic lineages.

| Patterns of galectin-12 protein abundance in HL-60 and 3T3-L1 cells
Given that the expression of LGALS12 inversely changed depending on the type of HL-60 cell differentiation, next we used ELISA to measure the intracellular and secreted levels of galectin-12 protein.The intracellular concentration of galectin-12 in HL-60 cells was readily detectable and relatively stable at a level of 13.41 ± 2.97 ng/mg of total protein through all samples (Figure 3a), which was lower but comparable to the level observed in differentiated 3T3-L1 cells (31.08 ± 8.14 ng/mg of total protein) as a positive control (Figure 3b).However, while 3T3-L1 cells demonstrated a significant increase of intracellular galectin-12 at 3-5 days of differentiation, galectin-12 secretion of 3T3-L1 cells was negligible with all samples falling below the minimal detection limit (18.75 pg/mL).This pattern was different

| Cellular differentiation and accumulation of lipid droplets in HL-60 and 3T3-L1 cells
The unique property of galectin-12 to be associated with lipid droplets (Yang et al., 2011(Yang et al., , 2016) )  inspect lipid droplets in HL-60 cells, we used BODIPY 493/503 staining and fluorescence microscopy imaging (Figure 4b).Untreated HL-60 cells showed minimal staining for lipid droplets, while both ATRA and DMSO induced a significant increase in their numbers as quantified using ImageJ (Figure 4c).To further validate the observed increase in lipid droplet content, we used flow cytometry analysis of cells stained with BODIPY 493/503.Once again, we observed a significant increase in the number of stained cells following ATRA and DMSO treatments and the effect of DMSO was significantly higher than ATRA (Figure 4d).DMSO also increased the number of dead cells under these treatments (Figure 4e), which were excluded from calculations of lipid droplets staining.To supplement these findings, we conducted gene set enrichment analysis for lipid-related GO terms using RNA-Seq data sets from GEO (GSE93996 and GSE103706) covering ATRA-and DMSO-differentiated HL-60 cells (Table 2).There were seven GO terms that were significantly T A B L E 2 Significantly enriched lipid-related GO terms in differentiated HL-60 cells.

| DISCUSSION
Consistent with previous findings, ATRA-induced differentiation led to an upregulation of LGALS12 while DMSO led to a downregulation (McTague et al., 2022;Vinnai et al., 2017;Xue et al., 2016).ATRA and DMSO produced distinct phenotypes of neutrophil-like HL-60 cells based on LGALS12 and FPR1 expression, and the functional response to fMLP as measured through hydrogen peroxide generation.
Comparing the gene expression profiles of the two differentiation models using existing RNA-Seq data also found that the DMSOinduced phenotype was associated with distinctively dysregulated processes related to cell division, protein translation, and signaling based on GSEA.There was also overlap of DEGs between the two phenotypes, reflecting that both agents have differentiation mechanisms that converge at a certain point.

Given that
LGALS12 is differentially expressed between ATRA and DMSO-induced neutrophil-like HL-60 cells, this gene could serve as a marker of neutrophil polarization.Previously, a model characterizing tumor-associated neutrophils as N1 and N2 polarized cell populations was proposed (Fridlender et al., 2009).This paradigm paralleled the terminology used to characterize M1/M2 macrophage polarization and described antitumor (N1) and pro-tumor (N2) neutrophils in cancer.The role of galectin-12 in neutrophils could be analogous to what is observed during macrophage polarization where knockdown of LGALS12 led to M2 polarization (Lin et al., 2020;Wan et al., 2016).In our study, DMSO produced cells with low LGALS12 expression.Low LGALS12 was reported to be associated with a poor prognosis of acute myeloid leukemia (El Leithy et al., 2015) and low LGALS12 was also observed in patients with colorectal cancer (Katzenmaier et al., 2017), prostate cancer (Laderach et al., 2013) and breast cancer (Tazhitdinova & Timoshenko, 2021).
On the other hand, ATRA induced upregulation of LGALS12 and this change might support anticancer activity of ATRA, which is used for differentiation therapy of acute promyelocytic leukemia (Madan & Koeffler, 2020).Considering that galectin-12 can also induce cell cycle arrest at the G1 phase, cell growth inhibition, and apoptosis in cancer cells (Hotta et al., 2001;Yang et al., 2001), a potential role of LGALS12 as a type of tumor suppressor gene cannot be excluded, similar to other tissue-specific galectins (Kaminker et al., 2024;Kim et al., 2013;Nagy, 2002;Satelli et al., 2011;Wiersma et al., 2013;Yang et al., 2001).These findings together allow us to suggest that the ATRA-induced phenotype of HL-60 cells with a high LGALS12 represents antitumorigenic N1 cells, while DMSO-induced phenotype with a low LGALS12 represents protumorigenic N2 cells (Figure 5), that is, LGALS12 may be considered as a regulatory marker of cell plasticity or polarization of neutrophils similar to macrophages (Lin et al., 2020;Wan et al., 2016).
The other aim of our study was to evaluate galectin-12 abundance and secretion in HL-60 cells.We compared these parameters between HL-60 cells and 3T3-L1 cells, which represent a reliable model to study galectin-12 (Yang et al., 2004).The intracellular levels of galectin-12 in HL-60 were relatively high and did not change under neutrophilic differentiation conditions despite the changes in the gene expression, which may depend on specific miRNAs, mRNA stability, and other factors differentially affecting galectin-12 translation and stability under the treatments with ATRA and DMSO.This pattern was different from a dramatic timedependent increase of galectin-12 in mouse 3T3-L1 cells upon adipocyte differentiation.However, 3T3-L1 cells did not secrete galectin-12 even despite high intracellular levels, which was consistent with previous studies and resulted from the hydrophobic profile of this galectin and its association with lipid droplets in adipocytes (Yang et al., 2011(Yang et al., , 2016)).As to the HL-60 cells, the basal secretion of galectin-12 was evident and was completely blocked during neutrophilic differentiation induced by both ATRA and DMSO.This the association to lipid droplets exists beyond the adipocyte cell model (Xue et al., 2016).Therefore, galectin-12 secretion seems to be tissue-specific as it is more easily detectable in neutrophil-like cells.It is also possible that neutrophils may serve as a source of galectin-12, which is detected in circulation (Galdino et al., 2021;Nowowiejska et al., 2023).Functional implications of the secreted galectin-12 rather than its lipogenic potential require further elaboration in this context.
There is little known surrounding the presence and regulation of lipid droplets in HL-60 cells.GSEA of existing HL-60 RNA-Seq data demonstrated that several lipid-related ontology terms were positively enriched upon differentiation with ATRA or DMSO.Furthermore, both fluorescence microscopy and flow cytometry found a significant increase in lipid droplets upon differentiation with DMSO and ATRA.One study found that lipid droplets in ATRAdifferentiated HL-60 cells are not detectable until stimulation with Porphyromonas gingivalis lipopolysaccharide which increased both size and number of observable droplets (Nose et al., 2013).This suggests that differentiation with ATRA may prime machinery for lipid droplet synthesis but that there needs to be an inflammatory signal for their formation.Additionally, it is not yet established whether galectin-12 colocalizes to lipid droplets in neutrophils like in adipocytes.In adipocytes, galectin-12 colocalizes with perilipin-1 which is not expressed in neutrophils (Itabe et al., 2017;Nose et al., 2013).It was previously observed that HL-60 cells had no expression of perilipin-1, −2, and −5, while perilipin-3 expression increased upon stimulation with lipopolysaccharide (Nose et al., 2013).
Whether galectin-12 interacts with perilipin-3 in HL-60 cells in a manner similar to what is observed with perilipin-1 in adipocytes requires future studies.
Our study still leaves unanswered the question about mechanisms, which are responsible for the secretion of galectin-12 by HL-60 cells.The effects of biochemical inhibitors were highly variable and only 3-MA, a context-dependent modulator of autophagy (Klionsky et al., 2021;Wu et al., 2010), was efficient in our model increasing the secretion of galectin-12.To shed more light on unconventional secretory mechanisms, these relatively simple assays might be supplemented with not only measuring specific autophagy marker(s) but also more comprehensive studies of individual secretory pathways using an arsenal of relevant methods and techniques (Pompa & De Marchis, 2016).
In conclusion, the findings of our study suggest that LGALS12 can serve as a marker of neutrophil polarization due to its varied expression between two distinct phenotypes of neutrophil-like HL-60 cells induced by ATRA and DMSO (Figure 5).As such, LGALS12 assay might be useful for screening drugs with antitumor activity considering a diversity of neutrophils in the context of cancer therapy.We also demonstrated that galectin-12 is secreted from HL-60 cells and that this process is inhibited upon neutrophilic differentiation and accompanied by an increase in the number of lipid droplets in cells, which may account for intracellular storage of this galectin.Molecular mechanisms controlling unconventional secretion of galectin-12 are still unclear and require further elaboration as well as unraveling functional implications of this galectin in circulation.

Fluorescence
microscopy and ImageJ were used to analyze morphology of DAPI-stained nuclei of HL-60 cells, which were prepared as cytospins on glass slides.Briefly, 200 μL of cell suspension (0.5 × 10 6 cells/mL in D-PBS) were loaded into Shandon Cytofunnel and centrifuged for 5 min at 500 rpm using a Shandon Cytospin 2 centrifuge.The cytospins were fixed with cold methanol, dried, and mounted with Fluoroshield mounting medium containing DAPI.The samples were viewed under a Zeiss AXIO Imager.A1 fluorescence microscope equipped with DAPI filter cube and images were captured using a high-resolution monochrome XCD-X700 CCD camera (Sony Corporation) with Northern Eclipse 8.0 software (Empix Imaging).These images were converted to binary form using ImageJ tools (Process > Binary > Make Binary) and the binary image contours were smoothened using a plugin Shape smoothing (version 1.2)

For
fluorescence staining of lipid droplets, 3T3-L1 or HL-60 cells were rinsed twice with Ca 2+ /Mg 2+ -free D-PBS and then incubated with BODIPY 493/503 solution (2 µM in D-PBS) for 15 min at 37°C.Following this incubation, cells were rinsed with D-PBS twice and fixed with 4% paraformaldehyde in D-PBS for 30 min at room temperature.These cells were rinsed again with D-PBS three times before mounting with Fluoroshield mounting medium containing DAPI.The samples were viewed and imaged using a Zeiss AXIO Imager.A1 fluorescence microscope equipped with DAPI and FITC filter cubes.The number of visible lipid droplets was counted for each cell using Cell counter plugin (version 2.2.2) for ImageJ.
we conducted a comparative analysis of two RNA-Seq data sets from the GEO representing differentiated HL-60 cells induced by ATRA (GSE93996) or DMSO (GSE103706).Principal component analysis (PCA) of gene expression demonstrated that transcriptomic effects of ATRA and DMSO treatment were distinct (Figure1a).The first two principal components explained 74.2% of total data variance and clearly separated samples by treatment.Differential gene expression analysis revealed that DMSO treatment resulted in the differential expression of over 2.5 times more genes than ATRA.In ATRA-treated cells, there were 881 and 772 significantly upregulated and downregulated genes, respectively (minimum log 2 fold change of ±1, p < 0.01), whereas DMSO-treated cells had 1885 significantly upregulated and 2459 significantly downregulated genes.There were 570 upregulated and 587 downregulated genes common to both treatments (Figure1b).In addition, 52 differentially expressed genes demonstrated opposite directionality of transcriptional change.Moreover, 21 were upregulated by DMSO treatment but downregulated by ATRA, while 31 were downregulated by DMSO but upregulated by ATRA treatment.GO terms enrichment analysis of differentially expressed gene lists indicated that both ATRA and DMSO treatment led to transcriptional changes expected for neutrophilic differentiation.Selective analysis of expression of genes encoding phagocyte NADPH oxidase subunits as markers of neutrophilic differentiation (CYBA, CYBB, NCF1, NCF2, NCF4, and RAC1) confirmed their upregulation by both ATRA and DMSO, while FPR1 gene encoding fMLP receptor was primarily upregulated in DMSO-treated cells (Figure1c).Considering the role of galectins in cellular differentiation(Tazhitdinova & Timoshenko, 2021), we also analyzed their expression profiles and noticed a variety of variations between ATRA and DMSO models, which were related to a set of six galectin genes (LGALS1, LGALS3, LGALS8, LGALS9, LGALS12, and CLC) known to be specifically expressed in HL-60 cells (Figure1c).Further Gene Set Enrichment Analysis (GSEA) based on Biological Process GO demonstrated that terms associated with cell division, DNA replication and repair, and translation and protein folding were found to be significantly downregulated with DMSO but not ATRA.Meanwhile, terms associated with cell signal reception and transduction, and cell cytoskeletal functions were upregulated to a greater extent by ATRA compared to DMSO (Figure1d, Supporting Information S1: FigureS2).Thus, ATRA-and DMSO-induced neutrophilic phenotypes of HL-60 cells are characterized by significant differences in gene expression profiles.3.2 | Opposing changes in the expression ofLGALS12 in HL-60 cells treated with ATRA and DMSO Following the bioinformatics analysis above and discrepancy of reports on the expression and functions of LGALS12 in myeloid cells(McTague et al., 2022;Vinnai et al., 2017;Xue et al., 2016), we sought to revisit this aspect in a model of neutrophilic differentiation of HL-60 cells.The cells were treated for 3 days with ATRA (1 μM) and DMSO (1.3%) and the following features of differentiated cells were verified: nuclear morphology, expression of differentiation markers, and functional responses to phagocyte membrane NADPH oxidase inducers such as fMLP (an agonist of formyl peptide receptors) and

Furthermore
, we quantified the levels of LGALS12 mRNA in ATRA-and DMSO-treated HL-60 cells.The neutrophilic differentiation of HL-60 cells with ATRA resulted in a significant increase in LGALS12 expression (3.2-fold, p < 0.001), while F I G U R E 1 ATRA and DMSO activate distinct neutrophil differentiation pathways in HL-60 cells.(a) Principal component analysis of RNA-Seq data from GSE103706 and GSE93996 data sets.(b) Venn diagram of differentially expressed genes lists detected after treatment of HL-60 cells with ATRA or DMSO.(c) Heatmap of expression measures of selected galectins and neutrophil differentiation markers.Hierarchical clustering was conducted using Euclidean distance and log 2 FPKM expression measures.(d) Gene set analysis based on the gene list of all differentially expressed genes detected in RNA-Seq data sets based on the Biological Process GO terms.Nodes are GO terms.Node size reflects a number of differentially expressed genes associated with a term.Node color reflects a level of statistical significance of the up-(red) or down-(blue) regulation of expression of term-associated differentially expressed genes.Edges and edge color show magnitude of an overlap in gene identities between terms.Major thematic and semantic groupings of GO terms are outlined.ATRA, all-trans retinoic acid; DMSO, dimethyl sulfoxide; FPKM, fragments per kilobase of transcript per million mapped reads; GO, Gene Ontology.
Neutrophilic phenotypes of HL-60 cells demonstrate differential functional responses and expression of LGALS12 and FRP1 genes.HL-60 cells were grown in serum-contained IMDM suspension and nuclear morphology, gene expression, and functional responses were compared between control cells and cells treated with 1 μM ATRA or 1.3% DMSO for 72 h.(a) Typical changes in morphology of DAPI-stained nuclei and corresponding quantification of nuclear circularity and roundness using ImageJ, n = 167-568.(b) and (c) Upregulation of neutrophilic differentiation markers NCF1 and NCF2, n = 3-5.(d) and (e) Differences in rates of PMA-and fMLP-induced generation of H 2 O 2 between cellular phenotypes as normalized to the response of DMSO-treated HL-60 cells, n = 3-5.(f) and (g) Differential expression of FPR1 and LGALS12 between ATRA-and lDMSO-induced neutrophilic phenotypes of HL-60 cells, n = 3-4.(h) and (i) Gene expression of Lgals12 and Fabp4 in a positive control of mouse 3T3-L1 cells that were subconfluent (SC), overconfluent (OC) and differentiated for four days (D4), n = 3.All data are presented as means ± standard deviation and different letters between treatment groups represent significant differences between means (p < 0.05) based on one-way analysis of variance with Tukey's multiple comparisons tests.ATRA, all-trans retinoic acid; DAPI, 4′,6-diamidino-2-phenylindole; DMSO, dimethyl sulfoxide; fMLP, N-formyl-L-methionyl-L-leucyl-L-phenylalanine; IMDM, Iscove's modification of Dulbecco's modified eagle medium; PMA, phorbol 12-myristate 13-acetate.fromHL-60 cells, which were found to secrete and accumulate a low but detectable amount of galectin-12 in a time-dependent manner over 24 h of cell growth (Figure3c).Both ATRA and DMSO induced a complete inhibition of galectin-12 secretion in HL-60 cells after 48 h (Figure3d).To get insights into the potential mechanisms of galectin-12 secretion by HL-60 cells, we tested the effects of biochemical inhibitors of unconventional secretion pathways.The treatment of cells for 24 h with a modulator of secretory autophagy 3-MA (2 mM) resulted in a significant (p < 0.05) increase of galectin-12 secretion, while inhibitors of exosome (20 µM GW 4869) and microvesicle (10 µM Y-27632) formation both did not change the basal levels of galectin-12 secretion, although high variations were noticed (Figure3d).These data collectively indicate that HL-60 cells can secrete galectin-12 protein, which depends on processes of cellular differentiation and may be related to secretory autophagy.
may explain the inhibition of galectin-12 secretion, which was observed with differentiated HL-60 and 3T3-L1 cells.To provide support to this hypothesis, we analyzed the formation of lipid droplets in HL-60 cells with a reference to adipocyte differentiation of 3T3-L1 cells as a positive control for this process.Indeed, mouse 3T3-L1 cells readily accumulated lipid droplets over 5 days of adipocytic differentiation, which were observed using brightfield (ORO staining) and fluorescence (BODIPY 493/503 staining) microscopy (Figure4a).Notably, fluorescence microscopy was more sensitive and showed that lipid droplets surrounded the perimeter of DAPI-stained nuclei in differentiated cells, while only background diffused staining with BODIPY 493/503 was observed in the cytoplasm of untreated cells.Therefore, to F I G U R E 3 Intracellular levels and secretion of galectin-12 in two cellular differentiation models of HL-60 and 3T3-L1 cells.The measurements were performed using ELISA kits for human and mouse galectin-12.(a) Levels of galectin-12 in HL-60 cell lysates obtained from control cells (CTRL) and differentiated cells (ATRA and DMSO).(b) Levels of galectin-12 in cell lysates of 3T3-L1 undergoing adipocyte differentiation as described in Section 2: original subconfluent cell culture (SC), overconfluent cells (OC), and Days 1-7 of differentiation (D1, D3, D5, and D7).(c) Time-dependent secretion of galectin-12 from HL-60 cells over a 24 h period of cell growth in a serum-contained IMDM.(d) Galectin-12 secretion from HL-60 cells treated for 48 h with differentiation agents (ATRA and DMSO) and inhibitors of unconventional secretion pathways (20 µM GW4689, 10 µM Y-27632, 2 mM 3-MA).Data are presented as mean ± standard deviation, n = 3-4.One-way analysis of variance with Dunnett's multiple comparisons test was conducted where each treatment group was compared against the relevant controls (*p < 0.05, **p < 0.01, ***p < 0.001).ATRA, all-trans retinoic acid; DMSO, dimethyl sulfoxide; ELISA, enzyme-linked immunosorbent assay; IMDM, Iscove's modification of Dulbecco's modified eagle medium.F I G U R E 4 Comparative accumulation of lipid droplets in 3T3-L1 cells and HL-60 cells undergoing adipocytic and neutrophilic differentiation, respectively.(a) Adipocytic differentiation of 3T3-L1 cells over 5 days resulted in accumulation of lipid droplets, which were readily observed using brightfield integrated modulation contrast microscopy (red staining with Oil Red O) or fluorescence microscopy (green staining with BODIPY 493/503, supplemented with blue staining of nuclei with DAPI), 40× objective.(b) Fluorescence staining of lipid droplets in HL-60 cells with BODIPY 493/503 and cell nuclei with DAPI: typical images with magnified inserts of untreated cells (CONTROL) and cells treated for 72 h with ATRA or DMSO under 40× objective.(c) Quantification of lipid droplets in HL-60 cells based on their counts using ImageJ.(d) Flow cytometry analysis of HL-60 cells stained with BODIPY 493/503: histograms and quantification of mean fluorescence intensity between undifferentiated (CONTROL) and differentiated (ATRA and DMSO) cells.(e) Flow cytometry analysis of HL-60 cells stained with SYTOX Blue: forward scatter versus SYTOX blue plots and their quantification to assess cell viability.Data are presented as mean ± standard deviation, n = 4-5.Significant differences (p < 0.05) are represented as different letters for treatments and were determined using Kruskal-Wallis analysis of variance test followed by Dunn's test for multiple comparisons.ATRA, all-trans retinoic acid; DAPI, 4′,6-diamidino-2-phenylindole; DMSO, dimethyl sulfoxide.
(p < 0.05) enriched with both ATRA and DMSO, while an additional six and nine terms were enriched with ATRA or DMSO only, respectively.Together, these findings suggest a link between suppression of galectin-12 secretion and elevated number of lipid droplets in differentiated HL-60 cells.
is the only galectin tested in HL-60 cells thus far, that had inhibition of secretion upon neutrophilic differentiation(McTague et al., 2022).F I G U R E 5 A proposed role and regulation of galectin-12 in neutrophilic differentiation of HL-60 cells.LGALS12 is proposed to serve as a marker of N1 and N2 neutrophilic polarization due to its varied expression between two distinct phenotypes of neutrophil-like HL-60 cells induced by ATRA and DMSO.Both phenotypes accumulate lipid droplets (LD), which inhibit the secretion of galectin-12.ATRA, all-trans retinoic acid; DMSO, dimethyl sulfoxide.This finding supports the notion of positive relations between the suppression of galectin-12 secretion and an elevated number of lipid droplets in cells.Leukocytes including neutrophils, were shown to accumulate lipid droplets in response to inflammatory conditions and stimuli(Melo & Weller, 2016).Xue et al. previously observed that knockdown of galectin-12 reduced lipid droplet formation in ATRAdifferentiated NB4 promyelocytic leukemia cells, which suggests that