Essential role for histone deacetylase 11 (HDAC11) in neutrophil biology

Short abstract

Expression of HDAC11 may be involved in myeloid cell biology.

Abstract

Epigenetic changes in chromatin structure have been recently associated with the deregulated expression of critical genes in normal and malignant processes. HDAC11, the newest member of the HDAC family of enzymes, functions as a negative regulator of IL‐10 expression in APCs, as previously described by our lab. However, at the present time, its role in other hematopoietic cells, specifically in neutrophils, has not been fully explored. In this report, for the first time, we present a novel physiologic role for HDAC11 as a multifaceted regulator of neutrophils. Thus far, we have been able to demonstrate a lineage‐restricted overexpression of HDAC11 in neutrophils and committed neutrophil precursors (promyelocytes). Additionally, we show that HDAC11 appears to associate with the transcription machinery, possibly regulating the expression of inflammatory and migratory genes in neutrophils. Given the prevalence of neutrophils in the peripheral circulation and their central role in the first line of defense, our results highlight a unique and novel role for HDAC11. With the consideration of the emergence of new, selective HDAC11 inhibitors, we believe that our findings will have significant implications in a wide range of diseases spanning malignancies, autoimmunity, and inflammation.

Abbreviations

APL

acute promyelocytic leukemia

ATRA

all‐trans retinoic acid

BM

bone marrow

CBA

cytometric bead array

CBC

complete blood count

ChIP

chromatin immunoprecipitation

DC

dendritic cell

eGFP

enhanced GFP

HDACs

histone deacetylase

HDAC11KO

histone deacetylase 11 knockout

HL60

human promyelocytic leukemia cell

MGAL

Murine Genetic Analysis Laboratory

MPO

myeloperoxidase

NET

neutrophil extracellular trap

PN

peritoneal neutrophil

qRT‐PCR

quantitative RT‐PCR

RFP

red fluorescent protein

Tg

transgenic

TV

tail vein

WT

wild‐type

Introduction

In humans, the predominant circulating leukocytes—neutrophils—are known to be produced at immense numbers through sequences of increasingly differentiated precursor myeloid progenitors in the BM before entering the bloodstream and account for 50–70% of the entire circulating population [1, 2, 3–4]. During granulopoiesis, neutrophils are produced at the rate of 1 × 10¹¹ each day, with a significant increase within hours during infections [5], and in patients with various cancer [6, 7] and therefore, more than one‐half of the BM is devoted to the production of these cells at a steady state [8]. Neutrophils play a pivotal and well‐defined role in the host defense, where they eradicate invading microorganisms [9], and although they have been labeled short lived, new in vivo deuterium labeling analysis has revealed that these cells may have a circulatory lifespan of up to 5 d [10]. Moreover, it is known that neutrophils can influence the immune response by way of communicating with DCs, macrophages [11, 12], as well as B cells [13] and T cells [14]. In fact, accumulating series of evidence also suggests that neutrophils have the potential to gain phenotypic as well as functional properties classically assigned to APCs [15, 16] and in the presence of cancer, appose tumor progression [1, 17] conversely, given the appropriate signals, and regulate tumor growth [18, 19, 20–21]. These antagonistic populations of neutrophils, referred to as N1 (or tumor inhibiting) and N2 (or tumor promoting), probably exist as a dynamic array of activation states, rather than only two distinct populations [17, 22]. Additionally, neutrophils can form structures called NETs that are involved in a process called NETosis—a contributor of innate immunity and induced or stimulated by infection, inflammation, trauma, cytokine production, activated platelets, autoantibodies, and pathogens [23, 24, 25, 26–27]. In recent years, NETosis has been identified as an additional pathway of programmed cells death [28], during which nuclear chromatin relaxes and forms fibrous web‐like structures composed of DNA and histone‐associated granular proteins [25, 29]. Therefore, it is of no surprise that these cells can also damage cells and tissues of self (host), highlighting the importance of regulating genes functionally responsible for these pathologic occurrences [30, 31]. Regulation of neutrophil function and differentiation, in particular, genes involved in the inflammatory responses and chronic diseases, is mostly regulated at the transcriptional levels [32], and subsequently, an identification of factors involved in these processes would offer significant insight into the molecular mechanisms governing functional outcome. For a number of years, regulation of normal and malignant hematopoiesis by epigenetic factors has shown to be an area of significant interest [33, 34]. Epigenetic changes in chromatin structure have been associated with the deregulation of critical genes in normal as well as malignant hematopoiesis [35, 36]. HDACs alter chromatin by deacetylation of histone tails, resulting in transcriptionally inactive chromatin [37]. HDAC inhibitors have also been identified to alter the cytokine production profile [38], ultimately influencing the fate and expansion of hematopoietic cells [35, 39]. However, presently, the exact role of specific HDACs in regulation of hematopoietic processes is yet to be elucidated. Moreover, it has been demonstrated that cytokine production by myeloid cells is regulated by changes in the acetylation status of specific gene promoters [40, 41]. HDAC11 has been described as a negative regulator of IL‐10 expression in myeloid cells [42]. Furthermore, it has been shown that lack of HDAC11 increases the suppressive capacity of myeloid‐derived suppressor cells [43]. HDAC11, the most recently identified HDAC, is the sole member of the class IV HDAC subfamily [44]. The functional role of HDAC11 remains poorly characterized. Initially, it was believed that HDAC11 had a limited tissue expression restricted to kidney, heart, brain, skeletal muscle, and testis [44], but it has recently been documented also to be expressed in hematopoietic cells, where it plays an integral role in the regulation of immune tolerance through its action in APCs. However, at the present time, its regulatory role in myeloid differentiation and specifically, neutrophil function is yet to be characterized. In this manuscript, we reveal a previously unknown role of HDAC11, which may involve the regulation of neutrophil function. Here, we demonstrate that expression of HDAC11 correlates with neutrophil maturation, migration, and phagocytic function. We also show that HDAC11 may be involved in the transcriptional machinery of IL‐6, TNF‐α, and CXCR2/CXCL2.

MATERIALS AND METHODS

Mice and cell lines

C57BL/6 WT mice were purchased from Charles River Laboratories (Wilmington, MA, USA), Tg‐HDAC11‐eGFP reporter mice [45] were provided by the Nathaniel Heintz Institute through the Mutant Mouse Resource & Research Centers, and HDAC11KO, kindly supplied by Merck (Kenilworth, NJ, USA) and obtained from the E.S. lab, respectively. All strains of mice were housed in the same designated room at the animal facility (Vincent A. Stabile Research Building, Moffitt Cancer Center, Tampa, FL, USA), kept in a pathogen‐free condition, and handled in accordance with the requirements of the U.S. Guidelines for Animal Experiments. The HL60 APL cell line was purchased from American Type Culture Collection (Manassas, VA, USA) and cultured and maintained in RPMI (HyClone RPMI‐1640) with 10% FBS at 5% CO₂ and 37°C. Aging HDAC11KO and C57BL/6 WT mice were housed in the same room under identical conditions (mentioned above) for 18 mo (n = 5/group and strain).

qRT‐PCR

Total RNA was prepared from centrifugally pelleted and presorted cells (RNeasy mini columns and RNAse‐free DNAse; Qiagen, Germantown, MD, USA). cDNA was prepared using the iScript cDNA Synthesis Kit (Bio‐Rad Laboratories, Hercules, CA, USA), and qRT‐PCR reactions were conducted using the SYBR Green Two‐Step qRT‐PCR (Bio‐Rad Laboratories) with transcript‐specific primers (supplied upon request) and cDNA from samples as templates. qRT‐PCR amplification reactions were resolved on CFX iCycler (Bio‐Rad Laboratories), and fold changes were quantified (2^−ΔΔCt) [46]. Primers used HDAC11: forward: ACACGAGGCGCTATCTCAAC, reverse: ACGCGTTCAAACAGGAACTT; TNF‐α forward (5′‐CCGATGGGTTGTACCTTGTC‐3′), reverse (3′‐AGATAGCAAATCGGCTGACG‐5′); IL‐6 forward (5′‐CCGGAGAGGAGACTTCACAG‐3′), reverse (3′‐TCCACGTTTCCCAGAGAAC‐5′); 18s forward (5′‐GTAACCCGTTGAACCCCATT‐3′), reverse (3′‐CCGTCCAATCGGTAGTAGCG‐5′); CXCL2 forward (5′‐TCCAGAGCTTGAGCGTGACG‐3′), reverse (3′‐TTCAGGGTCAAGGCAAACTT‐5′); and CXCR2 forward (5′‐TCTGCTACGGGTTCACACTG‐3′), reverse (3′‐ACAAGGACGACAGCGAAGAT‐5′).

Flow cytometry

Flow cytometric analysis of BM aspirates was performed using fluorochrome‐labeled mAb (anti‐CD3, ‐CD4, ‐CD8, ‐CD19, ‐NK1.1, ‐CD11b, ‐Ly6G, ‐CD45, ‐CD117, ‐CD127, ‐CD11C, and ‐Ly6G; BD Biosciences, San Jose, CA, USA, and eBioscience, San Diego, CA, USA) and the vitality dye DAPI (Sigma‐Aldrich, St. Louis, MO, USA). Data were acquired on an LSR II cytometer (BD Biosciences, San Jose, CA, USA) at least 10,000 events of the smallest population of interest were analyzed using Kaluza (BD Biosciences) and FlowJo software 10.0.07r2 (TreeStar, Ashland, OR, USA). Annexin V flow cytometry was performed using BD Bioscience's LSR II flow cytometer using purified PNs from WT C57BL/6 and HDAC11KO mice. An FITC‐conjugated Annexin V antibody and the DNA dye propidium iodide were used in conjunction with FACS buffer. Gates were created for the analysis of apoptotic, viable, and dead cells from total cells. Data were collected to a limit of 10,000 events of the population of interest.

Immunoblotting

The cells were lysed in a lysis buffer containing 50 mM Tris, 280 mM NaCl, HCl, pH 8.0, 0.5% Igepal, 5 mM MgCl₂, 10% glycerol, and 10× protease inhibitor (Roche, Basel, Switzerland) and phosphatase inhibitor (Santa Cruz Biotechnology, Dallas, TX, USA). Next, lysates were sonicated for 8 min (on ice, for 2 cycles of 30 s on/30 s rest) and then mixed with 4× gel loading buffer and boiled for 10 min. Samples were then resolved on 10% gels and transferred to nitrocellulose membranes, which were blocked with 5% milk‐PBS‐Tween. Bands were detected by scanning blots with an LI‐COR Odyssey imaging system. Anti‐GAPDH (68795) was purchased from Sigma‐Aldrich. Anti‐β‐actin (sc‐47778) was purchased from Santa Cruz Biotechnology. Anti‐HDAC11 (3611P‐100) was purchased from BioVision (Milpitas, CA, USA).

Harvesting and in vitro culturing of mouse PNs

C57BL/6 WT or HDAC11KO mice (8 wk old, female) were injected i.p. with 1 ml sterile 5% thioglycollate solution (BD Biosciences). Mice were euthanized, and the peritoneal cavity was flushed immediately. The cavity was flushed with 10 ml ice‐cold RPMI‐1640 medium (HyClone Laboratories, Logan, UT, USA). Cells from the initial wash were pelleted, and neutrophils were positively selected using biotin anti‐mouse Ly‐6G antibody (127604; BioLegend, San Diego, CA, USA) and Mouse Biotin Positive Selection Kit (EasySep; Stemcell Technologies, Vancouver, BC, Canada). The PNs were cultured in 6‐well (5 × 10⁶ cell/ml) plates in complete RPMI‐1640 medium in the presence of 10% FBS (HyClone Laboratories) and 1% penicillin–streptomycin (Corning, Corning, NY, USA), stimulated with 2.5 μg/ml LPS (L2880; Sigma‐Aldrich). The purity of the PNs was ∼90% after selection.

Migration assay analysis

Medium (25 μl; control medium) containing serum‐free medium or chemokine MIP‐2 (10 μg/ml) was loaded into the lower chambers of the Transwell system. Cell suspension (50 μl; 1 × 10⁶/ml or 2 × 10⁶/ml) in 10% FBS RPMI‐1640 medium was added to the upper compartment of the chemotaxis chamber. The two compartments were separated by a 5 μM pore‐size polycarbonate filter. Spontaneous migration was determined as the movement of cells toward the control medium.

ChIP

Mouse PNs (3 × 10⁷) were cross‐linked with 1% formaldehyde (F8775; 25 ml; Sigma‐Aldrich) for 10 min at room temperature. Then, the cross‐linking was stopped by 0.125 M glycine for 10 min at room temperature. The cells were washed twice with PBS and proceeded with sample preparation. The cell pellet was lysed by 1 ml lyse buffer (50 mM HEPES, pH 7.8, 3 mM MgCl₂, 20 mM KCl, 0.25% Triton X‐100, 0.5% Nonidet P‐40) on ice for 10 min. The cell lysate was transferred to a “dounce homogenizer,” and 25 strokes were applied using the loose pestle A. The cell pellet was washed once with 1 ml wash buffer (10 mM Tris‐HCl, pH 8.0, 1 mM EDTA, 200 mM NaCl). The pellet was resuspended in 600 μl sonication buffer (50 mM HEPES, pH 7.8, 140 mM NaCl, 1 mM EDTA, 0.5% SDS, 1% Triton X‐100). The samples were sonicated and centrifuged for 10 min. The lysate was diluted at 1:10 to dilution buffer I (50 mM HEPES, pH 7.8, 140 mM NaCl, 1 mM EDTA, 1% Triton X‐100), precleared using Protein A agarose beads (16‐156; Millipore, Billerica, MA, USA), and incubated with rabbit anti‐HDAC11 antibody cocktail (3611P, BioVision; and H4539‐200UL, Sigma‐Aldrich) on a rotator at 4°C overnight. Biotin‐conjugated normal rabbit IgG‐B (SC‐2763; Santa Cruz Biotechnology) was added at the same amount of rabbit anti‐HDAC11 antibody cocktail in parallel of each sample and incubated on a rotator at 4°C overnight. Then, 50 µl protein A agarose beads were applied to each sample and incubated at 4°C for 4 h. The beads were then washed twice each with dilution buffer I, dilution buffer II (50 mM HEPES, pH 7.8, 500 mM NaCl, 1 mM EDTA, 1% Triton X‐100), LiCl buffer (20 mM Tris‐HCl, pH 8.0, 250 mM LiCl, 1 mM EDTA, 0.5% Triton X‐100), and TE buffer (10 mM Tris‐HCl, pH 8.0, 1 mM EDTA). Then, the beads were eluted in 100 µl elution buffer (50 mM Tris‐HCl, pH 8.0, 1 mM EDTA, 1% SDS) and incubated at 65°C for 15 min, and the immunocomplexes were collected in the soluble fraction. NaCl was added to a final concentration of 200 mM, and the cross‐linking was reverted by incubating at 65°C overnight. Then, RNAse A (158922; Qiagen) was added to 20 µg/ml to each sample and incubated at 37°C for 30 min. Proteinase K (EO0491; Thermo Fisher Scientific, Waltham, MA, USA) was then added to 100 µg/ml and incubated at 42°C for 2 h. DNA was purified by using the QIAquick PCR Purification Kit (28104; Qiagen), and purified DNA was eluted in 100 µl TE buffer for analysis by qRT‐PCR. DNA (2 µl) was used for each qRT‐PCR reaction. The following PCR primers were used: TNF‐α forward (5′‐CTCGGAAAACTTCCTTGGTG‐3′), reverse (3′‐CGATGGAGAAGAAACCGAGA‐5′); IL‐6 forward (5′‐TCTGCAGAGTGAAGAAGCTGA‐3′), reverse (3′‐GATTCCAGGCTGAAAGTAGGC‐5′); CXCL2 (MIP‐2) forward (5′‐GGCTAGAACTGAGGGCTAC‐3′), reverse (3′‐ACATCCATTCTTGTCCCACTG‐5′); and CXCR2 forward (5′‐CCTCTGACTCCCACACATCT‐3′), reverse (3′‐GCTTGCTGTGCTTTTGGTTT‐5′).

Phagocytosis assay

Upon euthanasia, 1 ml blood was obtained from an experimental mouse via cardiac puncture with a heparinized syringe. Blood was centrifuged at 2000 rpm for 2 min in a microcentrifuge tube, and sera was aspirated and added to a single tube of lyophilized pHrodo Red E. coli BioParticles (Thermo Fisher Scientific). The BioParticles were vortexed and then incubated for 30 min in a 37°C water bath for optimal opsonization. BioParticles were centrifuged at 2000 rpm for 2 min and washed twice with PBS (pH 7.4, without Ca and Mg), and resuspended in 2 ml RPMI‐1640 with 10% FBS. Purified neutrophils were resuspended at 1 × 10⁶ cells/ml and plated at 100 μl/well in quadruplicate in a 96‐well, flat‐bottom, tissue culture‐treated plate per mouse. BioParticles were added to the plate immediately before being placed into IncuCyte high‐throughput live cell microscope system (Essen BioScience, Ann Arbor, MI, USA) for detection of RFP expression.

IncuCyte real‐time microscopy analysis

Cells in a 96‐well plate were imaged through a 10× objective lens at 5 min intervals for 3 h using the IncuCyte Zoom HD live cell imaging system (Essen BioScience). The system is located in a 37°C/5% CO₂ cell culture incubator to maintain proper incubation conditions. Analysis was performed using the default red fluorescence processing definition within the IncuCyte Zoom 2014B software. The number of red cells and confluency of measured red fluorescence were translated from the images and exported using the Excel format.

CBA

This assay involves measurement of secreted cytokines in the media by flow cytometry. In brief, 100 μl supernatant, from cells stimulated with or without LPS, was collected as specific time points, and IL‐6 and TNF‐α levels were measured following the manufacturer's recommended protocol for the CBA kit; (BD Biosciences).

Sepsis model

C57BL/6 WT and HDAC11KO mice were injected with 15 mg/kg LPS (L4391‐1MG; Sigma‐Aldrich) via TV in 100 μl vol and observed for signs of sepsis (abnormal posture/positioning; partial paralysis; failure to respond to stimuli; head tilt; circling; impaired locomotion; nonpurposeful movement; hypoactivity; hyperactivity; restlessness; self‐trauma; aggressiveness; isolation from cage mates; shallow, rapid, and/or labored breathing; cyanosis; piloerection; matted hair coat; lack of inquisitiveness; hunched posture; and/or vocalizations), 3 times daily.

RACE analysis

This experiment was out sourced to the MGAL, Mouse Biology Program, at University of California, Davis (Davis, CA, USA). Detailed protocol will be provided upon request.

Statistical analysis

Unless otherwise specified, the statistical significance between values was determined by Student's t test. Data were expressed as the means ± sd. Probability values of P ≤ 0.05 were considered significant.

RESULTS

HDAC11 is differentially expressed in various stages of myeloid differentiation

It has previously been reported that HDAC11 is a negative regulator of IL‐10 production in myeloid cells. Here, we examined the expression of HDAC11 in myelopoiesis. To follow the dynamic changes in HDAC11 transcriptional activation, we used HDAC11 promoter‐driven eGFP reporter mice (Tg‐HDAC11‐eGFP) [45], where eGFP expression is indicative of HDAC11 transcription. With the use of multiparameter flow cytometric analysis, as seen in Fig. 1A , in the BM compartment, we were able to determine that expression of HDAC11 is markedly increased in promyelocytes when compared with earlier progenitors and further amplified in the neutrophils when compared with myeloblasts. Interestingly, expression eGFP and consequently of HDAC11 in monocytes was negligible when compared with promyelocytes and neutrophils (Fig. 1A). The lymphocytes also show a moderate expression of the HDAC11 transcript. Expression of the HDAC11 protein was confirmed by immunoblotting analysis in flow‐sorted monocytes, promyelocytes, and neutrophils (Fig. 1B, inset). Quantitative mRNA analysis confirmed our findings by demonstrating an increase in the expression of the HDAC11 message (fold changes are normalized to the 18 s housekeeping gene for each sample) and further validated the Tg‐HDAC11‐eGFP reporter mouse model (Fig. 1B; graph) used in our experiments. Additionally, we further validated these results by quantifying the mRNA expression of eGFP in the same samples and were able to demonstrate that expression of HDAC11 is concomitant with the expression of the eGFP message (Fig. 1C). To ensure that our data were not an artifact of the eGFP reporter mouse model, we flow sorted monocytes, promyelocytes, neutrophils, and lymphocytes from a C57BL/6 WT mouse and examined the expression of the HDAC11 message (Fig. 1D). The results confirmed our previous findings in the Tg‐HDAC11‐eGFP reporter mouse, and consequently, the neutrophil population expressed significantly higher levels of HDAC11 when compared with the rest of the subpopulations. Next, we purchased source leukocytes (OneBlood, Florida Blood Bank, St. Petersburg, FL, USA), collected from volunteer, normal donors, and isolated (flow‐sorted cell populations) monocytes and neutrophils. Quantification of HDAC11 expression observed in this Fig. 1 shows parallel results, mirroring what we had previously seen in our murine models, where neutrophils had higher expression of the HDAC11 message (Fig. 1E, graph). Immunoblotting analysis confirmed the expression of HDAC11 in the neutrophils when compared with the monocyte population (Fig. 1E, inset).

Figure 1.

The HDAC11 message is differentially expressed in various stages of myelopoiesis. (A) Dynamic visualization of the HDAC11 message using a Tg‐HDAC11‐eGFP mouse in myeloid compartments. BM was extracted from Tg‐HDAC11‐eGFP (as well as C57BL/6 mice as control, non‐eGFP‐expressing cells; gray), and cells were labeled (as indicated above) with specific cell surface markers for identification of each population. HDAC11 expression was determined by flow cytometry analysis of eGFP reporter gene expression, where the expression of the eGFP protein corresponds to the activation of HDAC11 transcriptional machinery (representative presentation from 2 individual experiments). SSC‐H, Side‐scatter‐height. (B) Monocytes, promyelocytes, and neutrophils from Tg‐HDAC11‐eGFP mouse BM cells were sorted using the FACSAria (BD Biosciences) device with 99% purity. Cells were lysed using TRIzol reagent (Thermo Fisher Scientific), as well as radioimmunoprecipitation assay buffer (Cell Signaling Technology, Danvers, MA, USA), and RNA as well as protein was extracted, respectively. Immunoblotting [Western blot (WB)] was performed using 10% SDS gel, the image was resolved using the Dura ECL reagent (Pierce; Thermo Fisher Scientific; inset), and qRT‐PCR analysis was performed using HDAC11 primers (graph). Sorted cell populations were also morphologically confirmed (lower). (C) Coexpression of the HDAC11 message and eGFP message was confirmed by qRT‐PCR using eGFP primers (error bars = sem; data representative of 3 individual experiments; n = 3). (D) Monocytes, promyelocytes, neutrophils, B cells, and T cells were sorted from C57BL/6 WT mouse BM cells using the FACSAria (BD Biosciences) device with 98% purity. Cells were lysed using TRIzol reagent (Thermo Fisher Scientific), and RNA was extracted. qRT‐PCR analysis was performed using HDAC11 primer. This was done to confirm the HDAC11 expression profile in a non‐Tg setting (error bars = sem; data representative of 3 individual experiments; n = 3). (E) Four normal human donor peripheral blood source leukocyte samples (purchased from OneBlood, Florida Blood Bank) were sorted for monocytic and neutrophylic populations using the FACSAria (BD Biosciences) device with a 96% purity and then examined for the expression of the HDAC11 message using qRT‐PCR analysis, and the fold change between neutrophils and monocytes was normalized to the lowest expressing monocytic population (error bars = sem; n = 4). (B–E) Note: subpopulations were compared and normalized to monocytes (as they have the lowest expression of HDAC11).

Expression of HDAC11 is significantly increased in leukemic cells upon differentiation

With the use of the HL60 APL cell line model and upon in vitro maturation and differentiation of these cells using ATRA stimulation, we observed an increase in the expression of the HDAC11 message ( Fig. 2A , lower) and protein (Fig. 2A, inset), as demonstrated. These differentiated HL60 cells also exhibited neutrophil‐like segmentation and overexpression of maturation markers, such as CD11b and MPO (Fig. 2B).

Figure 2.

Expression of HDAC11 in APL samples and functional consequences of its manipulation in this model. (A) HL60 cells were treated with ATRA for 72 h, the expression of the HDAC11 message was measured by qRT‐PCR, and the differentiation of HL60 cells to granulocytic‐like cells was examined by morphologic analysis (Below bar graph; error bars = sem; data representative of 3 individual experiments). NT, Nontreated control. (B) The differentiation of HL60 cells to granulocytic‐like cells was examined using MPO (a marker for cell's gain of granularity; upper), as well as CD11b expression (lower; representative figure from 3 individual experiments).

Lack of HDAC11 markedly increases the expression of proinflammatory cytokines

HDAC11 is restricted mainly to the nucleus, and its expression appears to be tissue specific with higher expression in brain, heart, testis, and skeletal muscle [44]; however, its precise function in the myeloid compartment and specifically, in the neutrophils is yet to be elucidated. In the proceeding experiments, we used germline HDAC11KO (Merck; C57BL/6 background) mice that were developed using Lox/Cre technology to remove a floxed exon 3, a portion of the HDAC catalytic region. In our Supplemental Fig. 1A, we demonstrate our method of positively genotyping HDAC11KO mice. As these mice will still transcribe a truncated mRNA sequence (excluding exon 3), to validate this mouse model further, we designed a special HDAC11 primer set that was within the excised exon 3 region of the HDAC11 gene. As seen in Supplemental Fig. 1, HDAC11KO mice have no amplified HDAC11 amplicons demonstrated by qRT‐PCR analysis when compared with C57BL/6 WT counterparts (Supplemental Fig. 1B). Furthermore, immunoblotting analysis suggested that the expression of HDAC11 is absent in cells isolated from HDAC11KO mice (Supplemental Fig. 1C). Taking it a step further, we sent samples from 2 C57BL/6 WT and 2 HDACK11KO mice to MGAL, Mouse Biology Program, at University of California, Davis, for RACE‐sequencing. The data confirmed that HDAC11KO mice are devoid of exon 3 on the HDAC11 gene sequence (data will be provided upon request). Recently, it has been demonstrated that epigenetic factors, such as HDACs, have been known to affect regulation of genes involved in inflammatory responses [47]. Subsequently, we decided to investigate whether lack of HDAC11 had any functional consequences in the neutrophil population. To do this, we isolated PNs from both HDAC11KO as well as C57BL/6 WT mice, 18 h, 5% post‐thioglycollate injection. In vitro, cells were treated with 2.5 μg/ml LPS for 2 and 4 h, and the samples were analyzed for the TNF‐α and IL‐6 message and protein expression. Results revealed that in the absence of HDAC11, neutrophils became more inflammatory, as seen by the mRNA expression level of TNF‐α and IL‐6 at every time point ( Fig. 3A ). These findings were confirmed further by CBA protein analysis, where both cytokines showed a higher level of expression when compared with the C57BL/6 WT counterparts (Fig. 3B). We were also interested in considering whether the lack of functional HDAC11 changed the normal distribution of hematologic cells in the peripheral blood. In brief, we collected peripheral blood from HDAC11KO as well as their normal counterparts, C57BL/6 WT mice (submandibular blood collection), and performed a CBC analysis. Surprisingly, both cohorts—HDAC11KO as well as C57BL/6 WT control mice—had a relatively similar monocytic and granulocytic blood count profile (Fig. 3C). To our knowledge, thus far, there has been no report of HDAC11 binding directly to DNA, and consequently, there are no direct downstream genes that can be signature indicators of this interaction. In experiments for this section, we also interrogated the possibility of HDAC11 being involved in transcriptional regulation of these molecules. In brief, PNs isolated from C57BL/6 WT mice were treated with 2.5 μg/ml LPS for 3.5 h (the concentration and the time point for these experiments were determined by dose titration and time course analysis in preceding experiments), and then the cells were processed for a ChIP assay. In Fig. 3D, ChIP data suggest that HDAC11 may be recruited to the TNF‐α and IL‐6 transcriptional complex. To demonstrate that this observation was specific to HDAC11 being possibly recruited to the promoter region of TNF‐α and IL‐6, we performed a similar ChIP analysis in mice using HDAC11KO PNs that showed no recruitment of HDAC11 to the promoter regions of TNF‐α and IL‐6, interpreted by no change in the enrichment ratio when comparing LPS‐stimulated samples with nonstimulated controls (Fig. 3E).

Figure 3.

Phenotypic consequences of HDAC11‐deficient PNs. (A) PNs from C57BL/6 WT and HDAC11KO mice were collected post‐thyoglycolate injection (5% at 18 h). The cells were treated with 2.5 μg/ml LPS for 2 and 4 h. Expression of inflammatory genes TNF‐α and IL‐6 were measured by qRT‐PCR. NT, nontreated control. (B) Protein concentrations for TNF‐α and IL‐6 were measured by CBA analysis. (C) CBC from C57BL/6 WT and HDAC11KO (n = 6/group). LYM, lymphocyte; MONO, monocyte; GRAN, granulocyte. (D) The recruitment of HDAC11 protein to chromatin fragments of TNF‐α and IL‐6 promoters was analyzed using the Pfaffl method [46] and is presented relative to input before immunoprecipitation, and the enrichment ratio was normalized to the IgG control (PNs isolated from C57BL/6 WT). (E) PNs isolated from HDAC11KO mice were used in another ChIP assay as negative control (error bars = sem; data presented for each graph are representative of 2 individual experiments).

Neutrophils isolated from mice lacking HDAC11 demonstrate much higher migratory and phagocytic capacity

The hallmark of neutrophils in the innate immune system is their capacity to migrate to the site of tissue injury and/or infection. To interrogate the migratory capacity of neutrophils lacking HDAC11, we isolated PNs (as mentioned in Fig. 3) and performed a migration assay. In brief, in this assay, isolated neutrophils were loaded in the upper compartment of the chemotaxis chamber, the serum‐free media or chemokine CXCL2 (MIP‐2–IL‐8 homolog; 10 μg/ml) were loaded into the lower chambers of a Transwell system, and the assay was concluded by detection and calculation of migratory neutrophils. The data suggest that neutrophils isolated from HDAC11KO mice have a significantly higher migratory capacity toward a bait chemokine MIP‐2 when compared with the WT counterpart ( Fig. 4A ). Next, we compared the expression of migratory receptor/chemokines, such as CXCR2 (homolog of IL‐8Rβ)/CXCL2, in HDAC11KO and C57BL/6 WT mice. In Fig. 4B, we demonstrate a higher endogenous expression of mRNA for these molecules in neutrophils lacking HDAC11. Flow cytometry analysis confirmed a significant increase in the protein level of CXCR2 on HDAC11KO neutrophils compared with C57BL/6 WT mice (Fig. 4C); however, the changes in the CXCL2 protein levels were not significant (data not shown). Furthermore, we assessed the phagocytic ability of C57BL/6 WT versus HDAC11KO mouse PNs. In brief, an equal number of PNs from C57BL/6 WT and HDAC11KO mice were cocultured with pHrodo Red E. coli BioParticles loaded with a pH‐sensitive fluorescent dye. With the use of live microscopy, we were able (in real‐time) to monitor the phagocytic capacity of PNs from each mouse. Data ascertained from these experiments demonstrated that PNs from HDAC11KO mice were functionally more potent (Fig. 4D). Given our findings discussed in the above experiments, it was deemed necessary to determine if HDAC11 was being recruited to the promoters of CXCR2 and CXCL2. Given our presented data so far, we predict that HDAC11 may be a negative regulator of genes involved in migratory response. To examine this hypothesis, in our concluding experiments, we performed a ChIP analysis. In brief, PNs from C57BL/6 WT mice were treated with 2.5 μg/ml LPS for 3.5 h, and then the cells were processed for a ChIP assay. The data indicate that there is recruitment of HDAC11 to the promoter regions of CXCR2 and CXCL2 (Fig. 4E). Additionally, ChIP assay using PNs from HDAC11KO mice demonstrates no enrichment suggesting an absence of HDAC11 recruitment to the same promoter regions (Fig. 4F). To demonstrate that this observation was specific to HDAC11, we performed a similar ChIP analysis using HDAC11KO PNs that showed no change in the enrichment ratio when comparing LPS‐stimulated samples with nonstimulated controls. The observed recruitment of HDAC11 to the promoter regions is not discriminated to a direct biding of HDAC11 to the promoter regions; however, the interrogation of this possibility requires additional and in‐depth promoter‐binding analysis, which was not the focus of our manuscript.

Figure 4.

Enhanced migratory and phagocytic capacity of HDAC11KO neutrophils. (A) Migration assay analysis between C57BL/6 WT and HDAC11KO PNs using the Transwell system and 2 × 10⁶/sample. Expression of migratory genes was analyzed in PNs isolated from WT and HDAC11KO mice using qRT‐PCR analysis (data generated from 4 individual experiments); ∗∗P < 0.01. (B) mRNA expression of CXCL2 and CXCR2 was analyzed in C57BL/6 WT and HDAC11KO mice at steady state (data generated from 2 individual experiments). (C) Expression of surface CXCR2 protein on neutrophils from HDAC11KO versus C47BL/6 WT mice. gMFI, geometric mean fluorescence intensity; ∗P < 0.05. (D) Phagocytic ability of PNs isolated from C57BL/6 WT and HDAC11KO mice was analyzed in the presence of pHrodo Red E. coli BioParticles loaded with a pH‐sensitive fluorescent dye and analyzed as a measure of cell engulfment and lysis in real time (RFP measurement by microscopy; error bars = sem; statistical analysis was done using 2‐way ANOVA; data presented are representative of 2 individual experiments). (E) The recruitment of HDAC11 protein‐to‐chromatin fragments of CXCL2 and CXCR2 promoters was analyzed in the presence and absence of LPS stimulation. (F) PNs from HDAC11KO mice were used in another ChIP assay as a negative control. The values obtained for these ChIP experiments were analyzed using the Pfaffl method [46] and are presented relative to input before immunoprecipitation; the enrichment ratio was normalized to the IgG control (error bars = sem; representative figure from 2 individual experiments).

HDAC11KO mice are more susceptible to LPS‐induced sepsis when compared with the C57BL/6 WT counterpart

Sepsis in humans, defined as severe inflammation in the presences of infection, is a common syndrome that kills thousands of patients each year [48]. To study this physiologic response, murine models have been established and validated. For instance, in intoxication models, mice are inoculated with proinflammatory compounds (noninfections), such as LPS [49]. Of note, mice are extremely resilient to most types of inflammation when compared with humans, but at high doses of LPS, within the range of 1–25 mg/kg (1000–10,000 times the dose that will cause septic shock in humans), mice experience severe inflammatory response and eventually succumb to sepsis [50, 51]. Given our observations thus far, it appears that HDAC11KO PNs, at steady state, have a higher innate inflammatory nature when compared with C57BL/6 WT mice. Therefore, we sought to investigate whether there was a difference between the time of inflammatory onset leading to eventual sepsis and death (post‐LPS inoculation) when comparing HDAC11KO with C57BL/6 WT mice. In brief, a cohort of HDAC11KO and age‐matched C57BL/6 WT mice (n = 8 in each group) was injected with LPS at 15 mg/kg via TV. Mice were monitored 3 times in 24 h. As seen in Fig. 5 , although all mice post‐LPS inoculation showed signs of fatigue and sluggish movements, HDAC11KO mice expired within the first 48 h of the experiment, whereas the remaining C57BL/6 WT mice fully recovered and were ultimately euthanized at the 72 h time point to mark the termination of the experiment.

Figure 5.

Septic shock experiment comparing HDAC11KO with C56BL/6 mice. A cohort (n = 8) of HDAC11KO and C57BL/6 mice was injected with 15 mg/kg LPS via TV. Survival graph representing both groups in timeline up to 72 h (data presented are representative of 2 individual experiments).

HDAC11KO mice show BM hypercellularity with granulocytic expansion and splenomegaly as a result of increased extramedullary hematopoiesis

Emergency granulopoiesis is defined as a well‐coordinated de novo production of neutrophils as a result of increased myeloid progenitor cells proliferating in the BM, signaled by and during severe infection. The ultimate goal of this physiologic occurrence is to increase the neutrophil output during an innate immune response. To examine whether HDAC11KO mice had a more robust production of neutrophils, we injected (i.p.) a cohort of HDAC11KO and C57BL/6 WT mice with CFA (Thermo Fisher Scientific) and assessed the expression of granular cells by flow cytometry. Flow cytometery analysis revealed a moderate increase of granular cells in the spleen of HDAC11KO mice when compared with C57BL/6 WT mice; however, this difference was not significant. Immunohistochemistry and H&E staining of BM cells from the same mice demonstrated extensive hypercelullarity, which did not allow decisive identification and quantification of granular cells (data not shown). However, when we compared spleen sections from aging (18 mo old) HDAC11KO and C57BL/6 WT mice, we observed that the HDAC11KO spleens had increased extramedullary hematopoiesis, resulting in an expanded red pulp (Fig. 5). In addition, sections from femoral bones revealed a marked BM hypercellularity on aging HDAC11KO mice, mostly as a result of an expansion of maturing neutrophils (Fig. 5).

DISCUSSION

In myelopoiesis, lineage‐specific transcription factors C/EBPα, C/EBPɛ, PU.1, and acute myeloid leukemia 1 cooperatively interact with specific DNA sequence response elements to initiate transcription of genes involved in differentiation [8, 52, 53–54]. More recently, epigenetic mechanisms, such as chromatin modification by acetylation/deacetylation of histone tails, have been shown to contribute to the regulation of gene expression and determination of cell population specificity [55, 56]. Conversely, it has been demonstrated that the deregulation of epigenetic mechanisms causes genetic alterations that consequently lead to manifestation of diseases, such as cancer and autoimmunity [57, 58]. More specifically, epigenetic modulations have been reported in a number of cellular processes involved in neutrophil development and functions (reviewed in Ostuni et al. [34]), including NETosis [59].

In this report, the interrogation of the transcriptional activity of HDAC11 in myeloid and lymphoid cells, using an HDAC11 promoter‐driven eGFP reporter mouse model, combined with use of lineage‐specific markers with multiparametric flow cytometry analysis, demonstrated a significant overexpression of HDAC11 in neutrophils when compared with monocytes (Fig. 1A). Moreover, analysis of hematopoietic cells isolated from the BM of Tg‐HDAC11‐eGFP mice revealed an overexpression of HDAC11 at the promyelocyte stage of neutrophil differentiation and with a significant increase in the neutrophils. Monocytes showed low‐to‐undetectable expression of HDAC11 (Fig. 1B and C). Additionally, our results using flow‐sorted leukocyte subpopulations from C57BL/6 WT mouse BM further confirmed and demonstrated a higher level of HDAC11 expression in promyelocytes and neutrophils compared with monocytes and lymphoid subsets (Fig. 1D). Equally, flow‐sorted monocytes and neutrophils from source leukocytes of normal human donors showed higher levels of HDAC11 expression in neutrophils compared with the monocyte population (Fig. 1E). These findings are of interest, as they suggest that HDAC11 may be a factor during myelopoiesis and consequently, play a role in the fate of neutrophils. Epigenetic mechanisms, such as histone modifications, play a key role in the control of multiple normal biologic processes, including hematopoiesis, as well as alterations leading to many diseases, such as autoimmunity and cancer [60, 61]. The understanding of the role of each specific HDAC in the context of the immune system in different malignancies will facilitate selective cancer immunotherapeutic modalities. Therefore, we further examined HDAC11 in a cancer model to determine whether this lineage‐specific overexpression also applied to malignancies using the HL60 APL human cell line. Our findings supported what we had seen in normal processes. Upon in vitro ATRA‐induced differentiation and maturation of APL cells, expression of HDAC11 is increased exponentially (Fig. 2A). Concurrently, the expression of MPO and CD11b (markers of granularization) was increased (Fig. 2B) during maturation of APL cells. The exploration into the physiologic role of HDAC11 overexpression in neutrophils, using a model of germline HDAC11KO mice, demonstrated that purified neutrophils lacking HDAC11 displayed an apparent overproduction of TNF‐α and IL‐6 upon stimulation with LPS at message and protein levels compared with their C57BL/6 WT counterparts (Fig. 3A and B). Our assessment of these collective data suggests that HDAC11 may be playing the role of a check‐point molecule, where it possibly controls the activation of neutrophils. Moreover, subsequent data suggest that HDAC11 may be associated with the transcriptional machinery of inflammatory molecules, TNF‐α and IL‐6, as seen by HDAC11 chromatin binding, which was demonstrated by ChIP analysis (Fig. 3D and E). Recently, Stammler et al. [62] reported that HDAC11 inhibition increases IL‐1β in DCs and macrophages, through promoting a noncanonical caspase‐8‐dependent pathway. In our experiments, ChIP analysis revealed that HDAC11 chromatin binding is distinctive to TNF‐α and IL‐6, and no binding was observed for IL‐1β (IL‐1β, data not shown). Markedly, migration assay analysis also demonstrated that neutrophils isolated from HDAC11KO mice exhibit a significantly higher migratory rate, as well as increased phagocytic activity, compared with C57BL/6 WT mice (Fig. 4A). This observation is highlighted by qRT‐PCR analysis that showed that HDAC11KO mice demonstrate an endogenous overexpression of CXCR2 and CXCL2 genes (Fig. 4B and C). Moreover, ChIP analysis results suggest that HDAC11 may also be recruited to the transcription complex, regulating the transcription machinery of CXCR2 and CXCL2 (Fig. 4E and F). Recently published data [63] suggest that HDAC11 has functional network protein association with a number of biologic processes, including gene expression, therefore, highlighting the possibility of its involvement in regulating transcription, which is yet to be determined.

Typically, occurrence of severe sepsis, also known as septic shock, is a fatal complication of infection, usually caused by dysregulated inflammatory and immune responses. The onset of sepsis is generally a result of a robust innate immune response through enhanced granulopoiesis in the BM and generation of an exuberant number of neutrophils and consequently, a massive production of proinflammatory cytokines [64]. Sepsis remains to be a serious health issue, and therefore, identification of factors contributing to it is of great interest. Thus far, numerous studies have labeled pan‐HDAC inhibitors fundamentally as negative regulators of gene expression for acute immune receptors and pathways in innate immune cells [65, 66]. However, the oversimplification of this statement cannot be generally correct, as HDAC inhibitors, depending on tissue type and time of treatment, can have alternative effects on gene expression. Consequently, numerous investigators are now focusing on selective HDAC inhibitors to narrow the target of interest. In regards to HDAC11, our observations suggest that this HDAC may be a gatekeeper of inflammatory response in neutrophils. In fact, in a murine sepsis model, we were able to show that HDAC11KO mice succumb to sepsis much faster than the C57BL/6 WT counterparts (Fig. 5), suggesting the possible presence of preprimed neutrophils in HDAC11KO mice. Additionally, the CellTiter‐Blue viability assay demonstrated a modest decrease in the number of neutrophils (isolated from the BM) extracted from HDAC11KO mice in the absence of stimulation when compared with the C57BL/6 WT counterpart; however, a significant decrease in the viability of neutrophils isolated from HDAC11KO mice was observed when these cells were stimulated with GM‐CSF compared with C57BL/6 WT counterparts (Supplemental Fig. 2). Moreover, in an aging HDAC11KO mouse experiment, we observed a noticeable increase in the BM cellularity when compared with age‐matched C57BL/6 WT control mice, which is mostly a result of an expansion of maturing neutrophils ( Fig. 6 ). This is indicative of what we have demonstrated so far in this manuscript: signifying the likelihood of an essential role for HDAC11 in neutrophil biology.

Figure 6.

Observation of splenomegaly as well as hypercellularity with granulocytic expansion in BM of HDAC11KO mice. (A) Representative sections of spleens (H&E stain) harvested from aged C57BL/6 WT (upper) and HDAC11 KO (lower) mice showing a marked expansion of the red pulp (left; original magnification, ×40), as a result of replacement of the mostly lymphocytic red pulp cellularity observed on C57BL/6 WT mice with mostly trilineage extramedullary hematopoiesis on HDAC11KO mice (right; original magnification, ×400). (B) Representative sections of femurs (H&E stain) harvested from aged C57BL/6 WT (upper) and HDAC11 KO (lower) mice showing a marked increase in the BM cellularity on HDAC11KO mice compared with WT mice (left; original magnification, ×20), mostly as a result of an expansion on maturing neutrophils (right; original magnification, ×200; n = 5/group).

In recent years, the role of neutrophils has expanded significantly from frontline combatants of infection to endowment of anti‐tumor immunity [67, 68]. In contrast, recent evidence suggests a novel protumor function for these cells [11, 69], suggesting complex and opposing roles of neutrophils in a tumor setting. Such observations continue to highlight the importance of identifying factors that may play essential roles in neutrophil activation and function. In conclusion, our data suggest that HDAC11 appears to have a dual function in neutrophils biology: 1) HDAC11 is increased as neutrophils differentiate and mature, and 2) a decrease in HDAC11 correlates with functional activity of neutrophils. In this report, our studies reveal that HDAC11 plays a role in neutrophil chemokine and cytokine biology and function. Given the multifaceted function of neutrophils, the key findings described in this report will potentially lead to targeted epigenetic therapies to influence diseases involving neutrophils.

AUTHORSHIP

P.H. and J.P.‐I. are co‐principal investigators on this manuscript. E.S. and J.C. planned, organized, and performed the research and wrote the manuscript. J.J.P., X.C., K.M., S.L.D., A.D., M.L., H.W.W., L.X., F.C., A.L.S., and P.P.‐V. performed experiments. S.W., A.V., E.S., E.M.S., P.H., and J.P.‐I. supervised the projects for this research, reviewed the manuscript, and provided funding.

DISCLOSURES

The authors declare no conflicts of interest.

Supporting information

Supplementary Material Files

Supplementary Material Files

Supplementary Material Files