Neutrophils mediate critical innate immune responses by migrating to sites of infection or inflammation, phagocytosing microorganisms, and releasing an arsenal of antimicrobial agents, including reactive oxygen species. These functions are shared by other innate immune cell types, but an interesting feature of neutrophils is their hallmark lobulated nuclei. Although why this bizarre nuclear shape forms is still being elucidated, studies of two intermediate filament proteins that associate with the nuclear envelope, lamin A and C, indicate that expression levels of these proteins govern nuclear maturation. These A-type lamins also modulate nuclear stiffness, the loss of which may be critical to the migration of not only neutrophils but also cancer cells that become prone to metastasis. We investigated whether increased expression of either lamin A or C affects neutrophil nuclear morphologic maturation, but more importantly we tested whether overexpression of either lamin also affects neutrophil functional responses, using two mouse myeloid progenitor models that can be induced toward functionally responsive neutrophil-like cells. Collectively, our results demonstrate that overexpression of either lamin A or C not only disrupts nuclear lobulation but also causes aberrant functional responses critical to innate immunity, including chemotaxis, phagocytosis, and reactive oxygen species production. Moreover, the lamin A–overexpressing cells exhibit decreased expression of a critical NADPH oxidase complex factor, gp91phox, and transcriptomic profiling demonstrated differential expression of a number of myeloid differentiation and functional pathway components. Taken together, these data demonstrate that A-type lamin expression levels modulate not only nuclear morphologic features but also gene expression changes as neutrophils mature.
As critical effectors of innate immunity, neutrophils mobilize from pools of mature cells localized to the bone marrow in response to inflammation and/or infection (1, 2). Their impressive arsenal of functions includes navigation through capillaries and transendothelial migration during chemotaxis, engulfment of microbes during phagocytosis, and pathogen killing responses, in part supported by the production of reactive oxygen species (ROS) and release of proteolytic enzymes (3, 4). These processes are made possible due to dramatic changes in the expression of function-specific proteins as neutrophils mature in the bone marrow, including those associated with 1) chemotaxis, 2) steps leading to phagocytosis (integrins, adhesion proteins and pathogen binding receptors, e.g., Fcγ receptors), 3) components of the NADPH oxidase (NOX) complex responsible for oxidant production, and 4) primary or secondary granules that contain proteolytic and bactericidal enzymes. Multiple transcription factors that tightly regulate the genes encoding these proteins have been identified, including C/EBPα, C/EBPε, PU.1, and GA-binding protein (5–11). However, these factors are not the sole regulators of neutrophil development, as epigenetic mechanisms have also been shown to play a critical role in neutrophil maturation and acquisition of functional responses (12, 13). Interestingly, the neutrophil nucleus undergoes dramatic structural changes during maturation as it forms a lobulated architecture, which may not only affect mechanical dynamics of the nucleus (e.g., deformability and stiffness) but also impact chromatin organization that lead to changes in gene expression profiles critical to neutrophil function.
Our understanding of the molecular mechanisms that promote neutrophil nuclear lobulation or its role in neutrophil function has only recently been elucidated. Much of this understanding comes from studies of the lamin B receptor (LBR), an inner nuclear membrane component that binds lamins, chromatin, and chromatin-binding proteins. Deficient expression of LBR results in neutrophil nuclear hypolobulation, a characteristic of the benign disorder Pelger-Huët anomaly (PHA) (14, 15). Complete loss of LBR leads to far more severe phenotypes, including early lethality in HEM/Greenberg skeletal dysplasia in humans and the ichthyosis (ic) phenotype in mice (16–18). Importantly, our studies of progenitors derived from ic mice demonstrated that derived neutrophils with hypolobulated nuclei also exhibited deficient chemotaxis (19). The simplest interpretation is that LBR deficiency causes a decrease in nuclear deformation coincident with increased stiffness, which impedes migration of neutrophils as they navigate tight spaces in chemotaxis chambers. However, our further studies of these ic mice–derived neutrophils revealed deficient ROS production together with decreased expression of gp91phox (also known as Nox2 or Cybb), a critical component of the NOX complex; these data indicate that Lbr expression also is important to additional neutrophil functions, including the respiratory burst. Although some of the deficient ROS production was attributed to decreased sterol reductase activities contributed by Lbr, this did not account for aberrant gp91phox gene expression (20). Taken together, these past studies suggest that deficient nuclear lobulation in the maturing neutrophil can impact gene expression changes critical to several innate immune functions.
In addition to LBR, expression profiles of proteins comprising the nuclear lamina are also important to nuclear maturation and overall structural characteristics of the mature nucleus. The nuclear lamina is located beneath the inner nuclear membrane and comprised of a complex filamentous network of type V intermediate filament proteins, primarily A- and B-type lamins (21, 22). B-type lamins (B1 and B2) are ubiquitously expressed, critical to the survival and proliferation of multiple cell types, and required throughout organogenesis (23, 24). However, they are not believed to modulate nuclear mechanical properties (25). In contrast, lamin A and its variant lamin C, the two major isoforms encoded by the LMNA gene, exhibit variable expression levels depending on the cell type and associated tissue, and these levels are thought to modulate mechanical stability of nuclei in different cells (26, 27). For example, cells of structurally firm tissues or that undergo intensive shearing stresses (e.g., bone, muscle, skin) tend to contain high amounts of lamin A and C, and thus have relatively stiff nuclei. In comparison, certain cells in less dense tissues, in particular the hematopoietic system, exhibit low to undetectable lamin A or C expression and have flexible nuclei; these include inactive T and B cells along with differentiated neutrophils (28, 29, and recently reviewed in Ref. 30). Evidence for a direct role of A-type lamins in nuclear structural features comes from studies of human acute myelocytic HL-60 cells, which express detectable levels of lamins A or C as myeloblasts, but diminished levels as they are induced with all-trans retinoic acid (ATRA) toward neutrophil-like cells (31–33). Importantly, neutrophils differentiated from HL-60 cells that were modified to continuously express the immature form of lamin A (prelamin A, which normally becomes converted into the mature form by cleavage of 18 aa from the C terminus) exhibited phenotypes similar to mouse ic cells that lack Lbr expression, that is, hypolobulated or round-shaped nuclei plus deficient passage through small (3–5 μm) constrictions (34). This same study also indicated that the respiratory burst response was modestly deficient in the lamin A–overexpressing cells. Thus, expression levels of lamin A, and perhaps even lamin C, may play an important role in supporting not only nuclear lobulation, a feature that facilitates nuclear deformation, but also changes during maturation that are required for multiple neutrophil functions, including chemotaxis and ROS production.
In this study, we assess the functions of neutrophils with ectopic lamin A or C expression using two different mouse models of neutrophil differentiation that we transduced with retroviral vectors encoding either lamin. Each of the two models was generated from wild-type bone marrow that was engineered to ectopically express a dominant negative retinoic acid receptor α (RARα); the erythroid, myeloid, and lymphoid (EML) cell line is considered multipotent but can be derived into more mature EML-derived promyelocyte (EPRO) cells, whereas the mouse promyelocyte (MPRO) cell line is committed to the neutrophil lineage (35, 36). In each model, a homogenous population of lobulated, functionally mature cells can be generated that exhibit many of the responses displayed by wild-type, peripheral blood neutrophils (37, 38). We found that ectopic expression of either lamin A (as prelamin A) or C causes loss of nuclear lobulation during neutrophil differentiation, along with deficient chemotaxis by the mature cells in response to keratinocyte-derived chemokine (KC), the mouse homolog of human CXCL1. Moreover, mutant neutrophils overexpressing either A-type lamin exhibited abnormal phagocytosis, and in the case of mutant EPRO cells (e.g., EPRO-lamin A or EPRO-lamin C) displayed a deficient respiratory burst. Further investigation revealed deficient expression of a component of the NOX complex in either mature EPRO-lamin A or EPRO-lamin C cells. We then identified differences in global gene expression patterns in the differentiated EPRO-lamin A cells as compared with control neutrophils. Taken together, our results indicate that not only do A-type lamins play a critical role in nuclear lobulation and functional responses, but they may also influence gene expression patterns as neutrophils mature from early progenitors in the bone marrow. Our results add to a growing body of evidence that A-type lamins are far more important to cell differentiation than simply providing nuclear stiffness, by perhaps regulating gene expression patterns through their interactions with chromatin.
Materials and Methods
Cell lines and primary granulocytes
EML cells (American Type Culture Collection, Manassas, VA) were cultured in IMDM (HyClone, Logan, UT) supplemented with 20% horse serum (Life Technologies, Grand Island, NY) and 15% baby hamster kidney-MKL–conditioned medium as a source of stem cell factor (SCF). Promyelocytes were derived from EPRO cells as previously described (38) and then were maintained in IMDM with 20% horse serum and recombinant murine GM-CSF (10 ng/ml; PeproTech, Rocky Hill, NJ). EPRO cells were further differentiated into mature neutrophil-like cells when cultured for an additional 5 d in media supplemented with ATRA (10 μM; Sigma-Aldrich, St. Louis, MO). MPRO cells were maintained in AIM V media (Life Technologies), supplemented with 5% FBS (Life Technologies) plus 10% baby hamster kidney-HM5–conditioned medium and induced to neutrophil differentiation by addition of ATRA (10 μM) for 4 d. Phoenix producer cells, a subclone of HEK-293T cells, were maintained in DMEM (HyClone) with 10% FBS. Primary granulocytes were generated from ex vivo culture of bone marrow from C57BL/6 mice purchased from Charles River Laboratories (Wilmington, MA), using a previously published protocol (39). Mice were housed prior to euthanasia at the University of Massachusetts Lowell animal facilities under protocols approved by the University of Massachusetts Lowell Institutional Animal Care and Use Committee. All culture media were supplemented with penicillin (50 U/ml) and streptomycin sulfate (50 μg/ml; HyClone). The cultures were maintained in a humidified atmosphere incubator at 37°C injected with 5% CO2.
Lamin A and C expression vectors
The full-length murine lamin A (Lmna) gene was purchased as a cDNA from Open Biosystems (Lafayette, CO), excised from the provided pCMV-SPORT6 vector and then ligated into the pMSCVpuro vector (Clontech, Mountain View, CA) at XhoI and HpaI restriction sites with T4 DNA ligase (New England Biolabs, Ipswich, MA). Positive clones were sequenced (Beckman Coulter Genomics, Danvers, MA) with both strands of the inserted fragments confirmed by >2-fold read coverage using custom-designed internal PCR sequencing primers (IDT, Coralville, IA), or universal primers (M13 or T7). The resulting sequences were analyzed by aligning the reads against the Mus musculus genomic sequence database using the National Center for Biotechnology Information BLASTn alignment feature (40). The lamin C gene isoform (Lmnc) was generated using total RNA isolated from a murine macrophage cell line model, GM-CSF ER-Hoxb8, that was induced to differentiate for 5 d as previously described (41), and then reverse transcription was performed using SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). The cDNA was then amplified using 1× Taq reaction buffer with a 10:1 ratio of Taq DNA polymerase and Platinum Pfx DNA polymerase (Invitrogen), plus a forward primer from the Lmna gene, 5′-CGCTCGAGATGGAGACCCCGTCACAGC-3′ (IDT), and a reverse primer specific to the lamin C isoform, 5′-CGGAATTCTCAGCGGCGGCTGC-3′ (IDT). PCR was carried out using the following parameters: 95°C for 2 min; 95°C for 1 min, 64.4°C for 45 s, and 72°C for 2 min for 30 cycles; and a final extension cycle of 72°C for 10 min. The amplified product was cloned into pCR2.1-TOPO using the TOPO TA cloning kit (Invitrogen), and positive plasmids were sequenced (Beckman Coulter Genomics) with both strands of the inserted fragments confirmed by >2-fold read coverage using custom-designed internal PCR sequencing primers (IDT), or universal primers (M13 or T7). The resulting sequences were analyzed by aligning the reads against the M. musculus genomic sequence database (40). The Lmnc cDNA was then extracted and ligated into the pMSCVpuro vector (Clontech/Takara Bio, Mountain View, CA) at the EcoRI restriction site, and a positive clone was selected with the correct orientation using ClaI or HindIII restriction enzyme digestion and fragment analyses. Both resulting expression vectors were then sequenced from both DNA strands, and correct sequences were confirmed by comparisons to wild-type genomic sequence with the BLASTn alignment tool as described above.
Transfection and transductions with EML or MPRO cells
Phoenix producer cells were seeded at 5 × 105 total cells in 60-mm tissue culture–treated plates and cultured for 24 h. The cells then were transfected with 6 μg of either pMSCVpuro retroviral overexpression vector (containing either Lmna or Lmnc cDNA) and 1 μg of pCL-ECO retrovirus packaging vector (Imgenex, San Diego, CA) using XtremeGene transfection reagent (Roche, Brighton, MA) according to the manufacturers’ protocols. At 24 h posttransfection, the media were removed and replaced with exponentially growing EML or MPRO cell suspensions (105 cells/ml) supplemented with Polybrene (hexadimethrine bromide; 4 μg/ml; Sigma-Aldrich) and then cocultured for 24 h. The suspension cells were carefully removed from the Phoenix monolayer, centrifuged (150 × g, 5 min), and then resuspended in growth media in six-well tissue culture plates. After 24 h of recovery, puromycin (1 μg/ml; Sigma-Aldrich) was added and resistant cells were expanded for 21 d under puromycin selection to generate stable polyclonal populations of cells for further analyses.
Growth profile analyses
EML, EPRO, or MPRO cells were passaged at a 1:1 ratio with new growth media 24 h prior to assay to ensure exponentially growing and healthy cells. Cells were then harvested, washed twice with PBS, and seeded at 105 cells/ml in basal growth media in triplicate wells of a 12-well tissue culture plate for each genotype. Growth assays were performed by counting viable cells using the trypan blue dye (HyClone) exclusion method with a hemocytometer at 24, 48, 72, and 96 h time points, and cultures were passaged as needed to maintain required cell densities.
Wright–Giemsa staining and morphologic analyses
Cytospins were performed using 105 cells harvested by centrifugation (1400 × g, 5 min) and resuspending the cell pellet in 300 μl of PBS with 0.1% BSA (Fisher Scientific, Hampton, NH). Cytocentrifugations were then performed at 55 × g for 5 min, and slides were air-dried for ∼5 min and then stained in glass Coplin jars as previously performed (39). Dried slides were mounted with Permount (Fisher Scientific) and glass coverslips. The cells were visually inspected with an oil immersion ×60 objective with bright-field optics and imaged with an Olympus DP camera and Controller image analysis software (Olympus, Center Valley, PA). The distribution of cell types was evaluated by visually inspecting >150 cells for each differentiation time point and assessing morphologic features from at least six separate sets of inductions. Cells were designated into the two categories defined as either “ring/multilobed” when the nuclear structure had a distinct donut shape or at least three well-defined segments, or as “ovoid/bilobed” when no distinct segments were identified, only two connected lobes were present, or the nucleus was kidney-shaped. Data presented indicate the percentage of the total population designated in the ring/multilobed category.
Reverse transcription of mRNA and quantitative PCR
Total RNA was extracted from 107 cells using either TRI reagent (Molecular Research Center, Cincinnati, OH) or the RNeasy RNA mini kit (Qiagen, Germantown, MD), each according to the manufacturer’s protocol. The RNA concentrations were first determined (NanoDrop; Thermo Scientific, Waltham, MA) and then used for cDNA synthesis by reverse transcription with the SuperScript III first-strand kit (Invitrogen) per the manufacturer’s protocol. Briefly, ∼5 μg of RNA was mixed with oligo(dT) primers (1 μl of 500 μg/ml) and dNTP (1 μl of 10 mM stock concentration) supplemented with double-distilled H2O for a final reaction volume of 13 μl. The reaction mixture was heated at 65°C for 5 min and then cooled on ice for ∼2 min. To each reaction, 5× first-strand buffer (4 μl), DTT (1 μl of 0.1 M), RNase OUT (1 μl of 40 U/μl), and SuperScript III reverse transcriptase (1 μl of 200 U/μl) was added and then incubated at 50°C for 1 h followed by the reaction termination during a 15-min incubation at 70°C. For RNA removal, RNase H (1 μl of 2 U/μl) was added to the reaction mixture and incubated at 37°C for 20 min. Real-time PCRs were performed using 1 μl of the synthesized cDNA reaction mix, the recommended volume of SsoAdvanced SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA), and quantitative PCR (qPCR) primers (0.2 pmol/μl). Specific qPCR forward and reverse primers were used in the reactions for Lbr (forward, 5′-GATTCTGAGCCACGACAACAA-3′, reverse, 5′-AGCTGGAAATCGAGCCACTTT-3′), Lmna (forward, 5′-GCAGCGTCACCAAAAAGCG-3′, reverse, 5′-CCGCACGAACTTTCCCTCT-3′), Ltf (forward, 5′-TGAGGCCCTTGGACTCTGT-3′), CD11b (forward, 5′-ATGGACGCTGATGGCAATACC-3′), and mS18 as the reference transcript (forward, 5′-GGCGGAGATATGCTCATGTG-3′, reverse, 5′-GTCTGGGATCTTGTACTGTCGT-3′). Real-time PCR was performed using a CFX96 Touch real-time PCR detection system (Bio-Rad) using the following protocol: 95°C for 5 min; 95, 56, and 72°C (30 s each) for 40 cycles; and a final extension at 72°C for 3 min. Melt curve analysis started at 95°C with a temperature decrement rate of 0.5°C every 15 s until the temperature reached 45°C. Data analysis was performed using the ΔΔCt method.
Western blotting analyses
For all Western blots, 107 cells were collected, centrifuged (250 × g, 5 min) and rinsed twice in PBS containing a cOmplete protease inhibitor cocktail tablet (Sigma-Aldrich) and PMSF (1 mM, Sigma-Aldrich). The pellet was resuspended in residual buffer and then lysed by adding 300 μl of Laemmli lysis buffer (10% [w/v] glycerol, 104 mM SDS, 62.5 mM Tris [pH 6.8], 50 mM DTT) as previously described (42). The lysate samples were heated at 98°C for 5 min and passed six times through a 26½G syringe to shear DNA prior to resolving in polyacrylamide gels by SDS-PAGE. Equal amounts of protein extracts (∼10 μg) were loaded into NuPAGE Bis-Tris gradient gels (4–12%; Invitrogen) and then electroblotted onto 0.2-μm immunoblot polyvinylidene difluoride membranes (Bio-Rad). Prior to probing, membranes were activated in methanol (100%; Fisher Scientific) for ∼1 min, then incubated for 1 h at room temperature (RT) with gentle agitation in blocking buffer (5% [w/v] dried non-fat milk [Bio-Rad] in TBS with 0.1% Tween 20 [TBST; Bio-Rad]). Membranes were probed with primary Abs at dilutions and conditions recommended by the manufacturers, as follows: rabbit anti-lamin A/C (Cell Signaling Technology, Danvers, MA), 1:1000 dilution, overnight incubation at 4°C; monoclonal mouse anti–α-tubulin (Sigma-Aldrich), 1:1500 dilution, 1-h incubation at RT; mouse anti-gp91phox (BD Biosciences, San Jose, CA), 1:1000 dilution, overnight incubation at 4°C; and rabbit anti-p47phox (Invitrogen), 1:1000 dilution, overnight incubation at 4°C. Following washes in TBST, goat anti-rabbit (HRP conjugated) (1:2000 dilution, Cell Signaling Technology) or goat anti-mouse (HRP conjugated) (1:2000 dilution, Thermo Scientific) were then used against the primary Abs for 1 h at RT followed by three washes in TBST (10 min each) and detection with Clarity Western ECL substrate (Bio-Rad) to produce a chemiluminescent product. Chemiluminescence was detected and images of the blot were captured with a ChemiDoc MP Imager using the provided ImageLab software (v.5.1; Bio-Rad). The same software was also used for quantitative analysis of gp91phox expression in EPRO-lamin A and C versus vector lysates using band densitometry measurements from the developed blots. The signals were first normalized based on tubulin expression and quantified in arbitrary units. The percent of loss of gp91phox was calculated using data from three independent assays.
Cell surface marker staining and imaging flow cytometry
For each cell type, 106 cells were centrifuged (250 × g, 5 min), washed in PBS, and then resuspended in PBS with 2% FBS at 100 μl per sample. Purified rat anti-mouse CD16/CD32 (Fc Block; 50 μg/ml; BD Biosciences) was added to each sample and allowed to incubate for 15 min on ice. Fluorophore-conjugated primary Abs (4 μg/ml for each Ab; BD Biosciences) were added to the blocked cells as follows: FITC-conjugated anti–GR-1 (553126) or corresponding isotype control (553988), or PE-conjugated anti-CD11b (561689) or corresponding isotype control (553989). Cells were incubated for 45 min protected from light, washed in PBS with 2% FBS (400 μl), centrifuged (1400 × g, 5 min, 4°C), and resuspended in fresh PBS with 2% FBS (35 μl). The samples were then processed with a FlowSight imaging flow cytometer (Luminex, Austin, TX) using laser powers and gating strategies described previously (39).
Cells (105 total) were cytocentrifuged onto a glass slide, allowed to air-dry, and then fixed with 4% paraformaldehyde (Fisher Scientific) in PBS for 20 min at RT. The slides were then washed in PBS and the cells were permeabilized with 0.2% Triton X-100 (J.T. Baker, Austin, TX) in TBS (pH 6.8) plus 0.5% BSA for 20 min at RT. The slides were rinsed in TBS and blocked with TBS plus 0.5% BSA for 30 min at 37°C. The cells were then stained with anti-mouse lamin A/C Ab (Alexa Fluor 488 conjugate; Cell Signaling Technology) in TBS plus 0.1% BSA (1:50 dilution) for 1 h at RT, and washed in fresh TBS plus 0.5% BSA for 1 min. Finally, the slides were rinsed in PBS, air-dried, and mounted with ProLong Gold antifade mountant with DAPI (Molecular Probes, Eugene, OR). After 24 h, the cells were imaged under the Leica SP8 Lightning confocal microscope (Leica Microsystems, Buffalo Grove, IL) using 405 and 488-nm lasers in a sequential mode with a ×40 oil objective and ×2 zoom.
Respiratory burst assays
Detection of ROS production was assayed using a chemiluminescence reaction as previously described (43). Briefly, cells (106 per replicate) were harvested, centrifuged (250 × g, 5 min), and washed once in PBS. The cell pellet was resuspended in 160 μl of HBSS (Life Technologies) with 0.1% glucose (Sigma-Aldrich) per replicate and transferred to a white 96-well plate with a clear bottom. Next, 40 μl of Diogenes reagent (National Diagnostics, Atlanta, GA) was added to each well and incubated at 37°C for 3 min. Following incubation, cells were stimulated with PMA (5 μg/ml; Sigma-Aldrich) or serum-opsonized zymosan A (OZ; 500 μg/ml; Sigma-Aldrich) (43). Luminescence was measured immediately after PMA or OZ addition and then at 1-min intervals during a period of 10 min for PMA, or at 2-min intervals for 2 h for OZ, with each detected using a Synergy HT spectrophotometer (BioTek, Winooski, VT). Unstimulated cells were also analyzed but failed to produce significant ROS (data not shown). Levels produced by unstimulated cells were subtracted from those produced by stimulated cells, although most unstimulated cells produce negligible amounts of ROS.
Particles for manual counting of phagocytosis were prepared by resuspending lyophilized FITC-conjugated zymosan beads (20 mg/ml; Molecular Probes) in HBSS plus FBS (10%) and sodium azide (2 mM; Sigma-Aldrich), and then opsonization with mouse serum for 1 h at 37°C. Cells (106 per assay) were collected by centrifugation (250 × g, 5 min), washed in HBSS, and then resuspended in 100 μl of HBSS per assay. For each assay, 700 μl of HBSS, 100 μl of freshly prepared mouse serum, 100 μl of washed cells, and 100 μl of opsonized particles were added to a polystyrene snap-cap tube (15 × 60 mm; Fisher Scientific). The tubes were transferred to the hybridization oven (45 min, 37°C) with gentle rocking whereas control samples remained on ice. After incubation, the cells were transferred to a precooled 1.5-ml microcentrifuge tube and centrifuged (500 × g, 10 min), followed by visual inspection using an Olympus BX41 fluorescent microscope and a DP camera with Controller image analysis software. To quantify phagocytic indices, numbers of fluorescence particles were determined in each positive cell and presented as either averages for positive cells or actual numbers per positive cells.
Cell migration in response to KC was measured as previously described (43). Briefly, cells (105 per replicate) were harvested by centrifugation (250 × g, 5 min), washed in PBS, and resuspended in serum-free media (80 μl per replicate). IMDM or IMDM supplemented with 1% FBS and the chemoattractant (KC, 100 ng/ml; PeproTech) were added to the bottom wells of a 3-μm-pore transwell plate at 225 μl per well (Corning, Corning, NY). The upper chamber plate was then applied, and the prepared cell suspension was added to the top chambers at 80 μl per well. The transwell plate was incubated for 2 h at 37°C and 5% CO2. After the incubation, the transwell apparatus was slowly disassembled and cells in the bottom wells were quantified either by counting live cells with the trypan blue exclusion assay or using CyQUANT reagent (Promega, Madison, WA). For measurements with CyQUANT, the dye binding solution (2×) was dispensed into each well at 225 μl per well. The microplate was covered and returned to the incubator for 60 min, after which the fluorescence intensity of each sample was measured using a microplate reader (Synergy HT; BioTek) with excitation wavelength at 508 nm and emission wavelength detection at 527 nm.
Total RNA samples from 107 control versus EPRO-lamin A cells were prepared at the initial (day 0) and final (day 5 [D5]) stages of neutrophil differentiation by cell lysis in 1 ml of TRI reagent (Sigma-Aldrich) and RNA extraction with an RNeasy kit (Qiagen). The concentrations of RNA samples were measured with a Qubit 3.0 fluorometer (Invitrogen) followed by RNA integrity analysis using an RNA 6000 Nano kit on the BioAnalyzer 2100 system (Agilent Technologies, Santa Clara, CA). All samples had a RNA integrity analysis value >9.0. The cDNA libraries for sequencing were prepared using a TruSeq stranded mRNA kit with total RNA input of 250 ng according to the manufacturer’s protocol (Illumina, San Diego, CA). The libraries were then quantified by qPCR using the KAPA library quantification universal kit (Roche), and library size distributions were analyzed with a high-sensitivity DNA assay using the BioAnalyzer 2100 system. Finally, the library concentrations were normalized and libraries were pooled. The library pool was sequenced on a NextSeq 500 sequencer at 2 × 75-cycle reads, paired end (Illumina).
Differential gene expression analysis
Raw sequencing files in fastq format were uploaded onto the Galaxy (version 20.01) server and checked for read quality using the FastQC tool. The adapter sequences were removed by applying the Trimmomatic function and the resulting reads were mapped to the mouse genome (mm10) using the BWA tool. The featureCounts function was next applied to quantify the mapped reads, and the output files were analyzed for differentially expressed genes using the edgeR package in RStudio software (version 1.2.5001). The results were plotted on a volcano graph using EnhancedVolcano function with cutoff values of 0.05 for the p value and a 2-fold change. Gene Ontology (GO) analysis of differentially expressed genes was performed using the DAVID online bioinformatics tool (version 6.8). All differentially expressed genes with a p value <0.05 were uploaded and evaluated for their effects on biological processes (BPs) based on the GO database (GO_BP). The statistically significant BP terms with annotation level 5 (GO_BP_5) were then summarized using the REVIGO tool, an online software for categorization of long lists of GO terms. The final BPs were plotted using Excel software (Microsoft, Redmond, WA) and hierarchized based on their statistical significance (−log10 p value).
Unless specified otherwise (e.g., RNA sequencing analyses), all data shown that include numerical values with error bars are averages ± SD from at least three biological replicates, calculated using Excel software. Any indicated p value for statistical significance was calculated using an unpaired two-sample Student t test assuming equal variances (Excel) or one-way ANOVA with Fisher’s post hoc comparisons (StatPlus software, Apple, Cupertino, CA).
Lamin A and C expression in mouse models of differentiating neutrophils
As our previous studies of the wild-type EML model demonstrated that Lbr expression increases as cells differentiate into mature neutrophils (19), in this study, we began by comparing Lbr expression to that of either lamin A or C at each of three distinct stages in this same model: multipotent progenitors (EML), promyelocytes (EPRO), and mature neutrophils (EPRO cells induced with ATRA for 3 d). As shown in (Fig. 1A (upper panel), multipotent EML cells exhibit low levels of Lbr expression, but as the cells were induced into EPRO cells and then into mature neutrophils with ATRA (e.g., day 3), Lbr expression exhibited >3- and 5-fold increases, respectively. In contrast, appreciable Lmna gene transcription was identified in the EML cells, but levels significantly decreased in the EPRO cells and derived mature cells that resemble neutrophils. Similar results for Lmna gene expression profiles were observed in the MPRO cells, that is, detectable levels in undifferentiated cells but diminished levels in ATRA-induced cells (data not shown). To compare these results with protein expression, Western blot analyses were performed that revealed decreased lamin A or C protein expression in the uninduced or ATRA-induced EPRO cells as compared with EML cells (Fig. 1A, lower panel). We then analyzed gene expression patterns in ex vivo–cultured mouse bone marrow at three similar stages of neutrophil differentiation. Consistent with the EML cell model, Lbr expression was detectable at the common myeloid progenitor stage (after 3 d in SCF plus IL-3), and there was a slight increase as the cells differentiated into early promyelocytes (day 5, after culture in SCF, IL-3, and G-CSF), but levels dramatically increased upon terminal neutrophil maturation (day 7, after culture in G-CSF) (Fig. 1B, upper panel). In contrast, Lmna gene expression declined throughout the differentiation process. Results of a Western blot assay confirmed that lamin A or C protein expression diminished in the mature neutrophils, but they also revealed that lamin C consistently remains higher in expression as compared with lamin A (Fig. 1B, lower panel). Taken together, these data demonstrate that throughout myeloid cell maturation there is an inverse correlation of Lbr versus Lmna expression, suggesting that a critical balance of these nuclear envelope proteins may be important to the neutrophil maturation process.
Ectopic lamin A or C expression disrupts neutrophil nuclear lobulation
Our studies of the ic mouse model and those of others using genetically modified HL-60 cells demonstrated that loss of Lbr expression in either mouse or human myeloid progenitors results in neutrophils that exhibit features of PHA (i.e., hallmark hypolobulated or “pince-nez” nuclei) (19, 44). Studies of HL-60 cells manipulated to overexpress lamin A also caused nuclear hypolobulation along with defective migration through narrow (3–5 μm) constrictions (34). However, whether similar effects are caused by overexpression of the shorter lamin C isoform, or whether ectopic expression of either A-type lamin impacts neutrophil functional responses other than migration was not investigated. We therefore cloned each of the mouse lamin A or C genes into a retroviral expression system and transduced each into EML or MPRO cells to generate lines that overexpress either lamin isotype. We chose the pMSCVpuro retroviral vector based on our success in driving expression of wild-type and mutant forms of Lbr in our EML-ic cells that lack this protein (20). We then assessed protein expression levels using Western blotting with total protein lysates derived from each cell type at four distinct stages of neutrophil differentiation. As expected, detectable levels of lamin A or C were identified in uninduced control cells (referred to as EML-vector), but levels diminished during neutrophil differentiation (Fig. 2A, upper panel). In contrast, those cells transduced with either the lamin A or C expression vectors exhibited abundant levels of either protein throughout neutrophil maturation. Importantly, blind assessments of nuclear features identified in Wright–Giemsa-stained cells revealed that most differentiated EPRO-vector cells exhibited either ring-shaped or multilobed nuclei, demonstrating that they resemble fully mature neutrophils (Fig. 2A, lower panels, open versus filled block arrows indicate each nuclear feature in stained cells, respectively). In contrast, most of the differentiated, EPRO-derived cells that ectopically express lamin A or C (indicated as EPRO-lamin A or EPRO-lamin C neutrophils) showed ovoid, bilobed, or kidney-shaped nuclei (ovoid cells are indicated with thin arrows; asterisks in graphed data indicate significant differences with p < 0.0001). Similar results were observed in the MPRO cell model; control MPRO-vector neutrophils exhibited low levels of lamin A or C expression and abundant numbers of lobulated cells, whereas MPRO-lamin A or MPRO-lamin C neutrophils displayed severe hypolobulation (Fig. 2B). Taken together, these data confirm that elevated expression of either A-type lamin during neutrophil differentiation disrupts nuclear lobulation.
To further assess ectopic expression of each lamin in our mutant models, imaging flow cytometry analyses were performed on the control and mutant cells using an anti–A-type lamin Ab, or Abs against the Mac-1 component CD11b (depicted as Mac-1) or Gr-1 as indicators of neutrophil maturation. As shown in (Fig. 3A for the MPRO line, the population of differentiated control cells exhibits a modest decrease in A-type lamin expression as compared with undifferentiated cells, which coincides with a significant increase in Mac-1 and Gr-1 expression as would be expected for mature neutrophils. In contrast, both the undifferentiated and mature populations of MPRO-lamin A or C cells exhibit abundant expression of the corresponding lamin. Increased Mac-1 and Gr-1 expression in either mutant line at the mature state further indicates terminal differentiation. An inspection of individual cells revealed expression patterns of both types of proteins (lamin versus Mac-1 or Gr-1), and nuclear staining demonstrated the lobulated phenotype of the control nuclei as compared with ovoid nuclei in the mutant cells (Fig. 3B). To determine whether overexpression of either lamin might affect transcription of the Lbr gene, we also assessed Lbr expression as compared with genes encoding either lactoferrin or CD11b in the control versus mutant cells. As depicted in Supplemental Fig. 1, all three MPRO cell lines exhibit dramatically increased expression of the neutrophil-expressed genes along with a modest increase in Lbr expression. These results further demonstrate maturation of the mutant cells but also indicate that overexpression of either A-type lamin does not affect expression of a nuclear envelope–specific gene known to regulate nuclear maturation, specifically Lbr. Finally, immunocytochemistry was used on the MPRO control versus mutant cells to detect either lamin A or C protein localization along with nuclear features revealed with DAPI staining (Fig. 3C). As expected, the lamins are highly expressed and localized to the nuclear envelope of each mutant cell type, and most of the nuclei in the derived mutant neutrophils are either ovoid or kidney-shaped. Similar results were generated from the EML/EPRO-vector versus EML/EPRO-lamin A/C cells, all of which also exhibited a dramatically increased level of Gr-1 and Mac-1 upon differentiation, plus localized expression of ectopic lamin proteins to the nuclear envelope in the mutant cells (Supplemental Fig. 2A and 2B, respectively).
Growth responses of progenitors overexpressing either lamin A or C
Proliferation assays were performed on control (vector) versus mutant EML, EPRO, or MPRO cells to evaluate whether there were any changes in growth responses due to lamin overexpression. Cells were cultured in basal growth media for 4 d, during which cell counts were completed every 24 h for each genotype, using trypan blue staining to differentiate live versus dead cells and assessment with a hemocytometer. As shown in (Fig. 4, no appreciable differences were noted in any of the growth profiles. Thus, the increased and maintained expression of either lamin A or C does not affect survival or proliferation of either myoblast-like or promyelocyte-like cell types.
Disruption of chemotaxis and phagocytosis by ectopic lamin A or lamin C expression
Aberrant expression of the lamin B receptor in either the EPRO-ic/ic cells or the genetically modified HL-60 cells caused defective chemotaxis or cell movements through tight constrictions, as did overexpression of human lamin A in HL-60 cells (20, 34). Based on these results, we tested the mutant cells overexpressing either lamin A or C for chemotaxis using Boyden-like transwells with 3-μm pore size membranes and either KC or IMDM in the bottom chemotaxis chambers; responses were compared with those of the control cells. Migrated cells were then quantified either manually or by adding CyQuant into the bottom well and measuring levels of fluorescence. As indicated in (Fig. 5A, differentiated EPRO-vector cells exhibited impressive numbers of migrating cells in response to KC as compared with IMDM alone (∼19,000 cells from manual counts or 55,000 relative fluorescence units RFU with CyQuant, left versus right panels, respectively). In contrast, significantly less EPRO-lamin A neutrophils migrated into the bottom chamber when counted manually (∼6,700 RFU) or by CyQuant measurements (∼18,000 RFU). Mature EPRO-lamin C neutrophils also exhibited less migration, with levels slightly above 50% of that exhibited by the control (vector) cells (∼11,000 cells or 30,000 RFU, manual counts versus CyQuant, respectively). Similar results were observed with differentiated MPRO-lamin A or MPRO-lamin C cells, each with ∼50% reductions in chemotaxis as compared with the control cells (Fig. 5B). These results are consistent with the notion that loss of nuclear lobulation caused by increased lamin A or C expression increases nuclear stiffness, thereby hindering the movement of cells through tight constrictions provided by the 3-μm pores in the Boyden-like chamber membrane.
The capacity of the lamin-overexpressing cells to exhibit phagocytosis was next tested using opsonized FITC-conjugated zymosan particles and enumerating cells that engulfed particles after 45 min of incubation. As shown in (Fig. 6A, most (∼80%) of EPRO-vector neutrophils contained opsonized particles. In comparison, significantly fewer mutant cells overexpressing each lamin isoform were positive for engulfed particles (51% for lamin A versus 69% for lamin C). Moreover, the phagocytic index of each cell type, calculated as the average numbers of particles engulfed per cell, was less in the mutant versus control cells. A distribution profile of the number of internalized particles on a per-cell basis also was generated; as depicted in (Fig. 6B, the results revealed that many more mutant lamin A–overexpressing cells completely lacked internalized particles, whereas there was a fairly even distribution of control cells that contained between 1 and 4 particles, with some containing over >10 identifiable particles. MPRO-lamin A or MPRO-lamin C neutrophils also exhibited less percentages of cells with engulfed particles but their phagocytic indices were similar to those of the control cells (Fig. 6C).
Deficient respiratory burst and gp91phox expression
Destruction of pathogens by neutrophils is dependent on their capacity to generate ROS during the respiratory burst, a process mediated by NOX complexes localized to plasma and phagosome membranes (45). Assembly and activation of NOX component proteins, which include the membrane-bound factor gp91phox and the cytosolic protein p47phox, can be stimulated with either PMA or OZ. We compared the capacity of EPRO and MPRO cells, as either control or mutant forms, to produce a respiratory burst that was detected using an enhanced luminol reagent and quantitatively measuring emitted light units from stimulated cells. As depicted in (Fig. 7A (top panels, with average area under the curve indicating calculations from ROS produced by three independent inductions and analyses), control EPRO neutrophils exhibit robust and rapid production of ROS when stimulated with PMA, a molecule that easily penetrates the cell membrane and bypasses an engulfment step by directly activating protein kinase C (46). In stark contrast, EPRO-lamin A or C neutrophils exhibited dramatically reduced production of ROS. Similar results were observed when the control versus mutant neutrophils phagocytosed opsonized yeast particles—robust ROS were produced by control cells but severely diminished levels were produced by the mutant cells (Fig. 7A, lower panels). Surprisingly, the mutant MPRO neutrophils produced ROS at levels consistent with those produced by the control cells with the empty vector, whether stimulated with PMA or OZ (Fig. 7B, top and bottom panels, respectively).
We next determined whether deficient ROS production in the EPRO-lamin A or EPRO-lamin C neutrophils was due to decreased expression of a NOX component, as was previously observed with the EPRO-ic cells that produced aberrant respiratory burst along with deficient gp91phox expression (19). As shown in (Fig. 8A, expression of both gp91phox and p47phox dramatically increased in the control EPRO-vector cells after 5 d of ATRA treatment. In comparison, although expression of gp91phox increased upon induction of the EPRO-lamin A or EPRO-lamin C cells, levels were diminished compared with those expressed by the control cells. The effects on NOX appeared to only impact the gp91phox component, as levels of p47phox were comparable between mutant versus control cells. We also scanned the gp91phox expression signals depicted in (Fig. 8A to quantitate protein amounts, averaging levels from three independent inductions and corresponding Western blotting assays. These combined analyses revealed ∼80% less gp91phox protein expression in the EPRO-lamin A cells and 35% less expression in the EPRO-lamin C cells, each as compared with the control cells (Fig. 8B). Not surprisingly, mutant MPRO cells expressed normal levels of either protein, consistent with their abundant ROS production (Fig. 8C).
Differential gene expression associated with ectopic lamin A expression
Multiple studies have indicated that lamins can epigenetically regulate gene expression through their capacity to modulate chromatin organization and/or interact with transcription factors (21, 22, 47–49). However, there is little information regarding involvement of A-type lamins in regulating global gene expression, in particular during cell differentiation. Therefore, we decided to perform RNA sequencing analyses on transcripts extracted from EPRO-lamin A cells as uninduced promyelocytes or differentiated neutrophils, and then compared gene expression patterns to those expressed by the vector control cells at each stage of maturation. As depicted in (Fig. 9A through volcano plots, we found that in addition to the Lmna gene (gene ID 16905), which exhibited ∼32- and 120-fold overexpression in promyelocytes versus mature neutrophils (EPRO versus D5 ATRA-induced EPRO, respectively), there are multiple genes aberrantly expressed in the mutant cells at either stage of differentiation. Specifically, we identified 69 upregulated and 124 downregulated genes with ≥2-fold changes in the uninduced EPRO-lamin A cells, versus 59 upregulated and 246 downregulated genes in the mature EPRO-lamin A cells, each by comparison with the vector-only transduced cells. Among the upregulated genes with the highest fold change differences in D5 mutants were those encoding the olfactory receptor 811 (Olfr811; gene ID 258545), neuronal regeneration related protein (Nrep; gene ID 27528), F-box and WD-40 domain protein 17 (Fbxw; gene ID 109082), and ARP3 actin-related protein 3B (Actr3b; gene ID 242894), with 93-, 29-, 26-, and 25-fold increases, respectively. In comparison, downregulated genes included those encoding glycoprotein (transmembrane) nmb (Gpnmb; gene ID 93695), junction adhesion molecule 3 (Jam3; gene ID 83964), guanylate-binding protein 8 (Gbp8; gene ID 76074), and matrix metallopeptidase 12 (Mmp12; gene ID 17381), with 128-, 42-, 30-, and 26-fold decreases, respectively. Even the uninduced mutant cells exhibited impressive differential gene expression, including >100-fold upregulation of high mobility group nucleosomal binding domain 1 (Hmgn1; gene ID 15312) and PGE receptor 3 (Ptger3; gene ID 19218), a 45-fold increase in Nrep (gene ID 27528), and a 26-fold upregulation of translin-associated factor X (Tsnax; gene ID 53424). For downregulated genes in the uninduced promyelocytes, we observed an almost 26-fold decrease in melanoregulin (Mreg; gene ID 381269), an 18-fold decrease in aldehyde dehydrogenase family 3, subfamily A1 (Aldh3a1; gene ID 11670), a 14-fold decline in Jam3 (gene ID 83964), and an almost 13-fold decrease in arginase (Arg1; gene ID 11846) levels. We note that differential expression of the Nox2 gene was not revealed in our analyses of either developmental stage, despite loss of protein expression identified by Western blotting assays. Whether this is due to detection limits of the assay (i.e., transcription is defective but there is less than a 2-fold difference compared with wild-type cells) is unknown, but it may reflect overall differences in the transcriptome versus proteome of neutrophils, as has been recently reported along with other immune cell types (see Discussion) (50, 51).
In order to understand the potential physiological impact of the identified genes dysregulated in lamin A–overexpressing neutrophils, we performed an analysis against the GO database. All differentially expressed genes with a p value <0.05 were evaluated in BP ontology and the top, statistically significant terms were summarized using the REVIGO tool. Our results reveal that genes downregulated in day 0 mutants are associated with multiple processes, that is, cell migration, intracellular signal transduction, actin cytoskeleton organization, endocytosis, and myeloid leukocyte activation (Fig. 9B, upper panel). The terms for the upregulated genes include cell surface receptor signaling pathway, cell secretion, leukocyte cell–cell adhesion, and steroid biosynthetic process (Fig. 9B, lower panel). In D5 mutants the downregulated genes can be mostly linked to chemotaxis (e.g., cell migration, cellular response to cytokine stimulus, actin cytoskeleton organization, regulation of cell shape); however, other important processes were also listed, including regulation of programmed cell death, endocytosis, protein transport or IκB kinase/NF-κB signaling (Fig. 9C, upper panel). Finally, the upregulated genes in D5 mutants are associated with mitotic cell cycle process, microtubule cytoskeleton organization, and regulation of protein metabolic processes, among other terms (Fig. 9C, lower panel).
Studies of lamin A functions have predominantly focused on characterizing mutations in the LMNA gene that cause laminopathies, and severe dystrophies that affect development of muscle (skeletal and cardiac), cartilage, bone, and skin, among other tissues (recently reviewed in Ref. 52). Aberrant processing of prelamin A is also well characterized, as it causes premature aging known as Hutchinson-Gilford progeria syndrome, typically linked to point mutations that disrupt cleavage of C-terminal tail residues of prelamin A (53). However, there is growing interest in the roles that A-type lamins might play in both stem and cancer cell properties, in particular those that modulate tissue stiffness and thereby facilitate or impede cell migration either during embryogenesis or during metastasis. For example, LMNA expression is upregulated during osteogenesis but downregulated during adipogenesis, consistent with the relative stiffness of bone tissues versus fat cells, respectively (54). Cells of the CNS also show changes during differentiation that parallel those of the hematopoietic system, with significant lamin expression at the early stages (e.g., neuroblasts) but loss of expression in mature neurons. This may be important to the capacity of either cell type to migrate during organogenesis, either as hematopoietic progenitors migrate between the yolk sac, liver, and bone marrow, or as neural progenitors navigate through different tissues as neural networks mature. Interestingly, the lamin A-to-lamin C ratio is similar between these two disparate cells types, with higher levels of lamin C expression in the mature tissues, including cerebellum and cortex of the brain (55). Our results showing that overexpression of lamin C alone can cause aberrant nuclear morphology and migration behavior in mature cells support the notion that both lamins can play a role in cell and/or nuclear maturation.
Regarding cancer cell properties, low levels of lamin A in ovarian cancer tissues have been associated with poor prognosis, whereas higher levels correlate with better survival (56). Furthermore, knockdown of lamin A expression in HO8910 ovarian cancer cells increased their migration in either 3- or 8-μm pore size membranes, whereas overexpression impeded migration (57). Similar results have been documented in cells derived from neuroblastomas, gastric carcinoma, as well as colon, breast, and lung cancer (58–62). Our results showing abnormal nuclear structure plus deficient chemotaxis in lamin A– or lamin C–overexpressing cells (Figs. 2 and (5, respectively) are consistent with these results, suggesting that loss of either lamin isotype in cancer cells may facilitate their metastatic potential. Interestingly, knockdown of lamin A not only decreased prostate cancer cell growth but also affected their epithelial-to-mesenchymal transition states (63, 64); although we observed no changes in the proliferation of the multipotent EML or promyelocyte-like EPRO or MPRO cells that overexpress either lamin (Fig. 4), the observed aberrant functional properties of the derived mature cells may indicate that the progenitors have unidentified abnormalities. Thus, there is a significant body of evidence that lamin A or C expression levels influence not only nuclear mechanical properties but may be a key factor that determines mechanophenotypes of different cell types; in the case of cancer cells, detecting changes in LMNA gene expression may also help predict whether cancer cells are more prone to metastasis.
Beyond supporting nuclear stiffness and/or differentiation properties, A-type lamin expression must also play a significant role in nuclear lobulation in mature neutrophils, as was discovered for LBR and its role in PHA (14–16). Moreover, the importance of lamin A or C expression levels to this unique morphologic feature must be independent of Lbr, as its expression remained normal in the lamin A– or lamin C–overexpressing EPRO and MPRO cells (based on quantitative RT-PCR analyses; see Supplemental Fig. 1). Our results also add further evidence that a lobulated nucleus facilitates neutrophil chemotaxis, a process that requires navigation through tight spaces as the cells escape capillary beds and migrate through tissues to sites of infection or inflammation. The notion that ratios of lamins-to-nuclear envelope components is critically important to nuclear lobulation is also supported; that is, as Lbr expression increases lamin expression declines during neutrophil maturation (34). Interestingly, this opposing expression pattern is conserved in adipogenesis, leading to a soft tissue that exhibits low lamin A/C but high LBR expression (65). We note that studies of lamin A–overexpressing HL-60 cells failed to show the additional phenotypic changes observed with the EPRO and MPRO cells, in particular phagocytosis or changes in NOX component expression (66). However, HL-60 cells exhibit undetectable expression of secondary granules and limited chemotaxis toward chemokines known to stimulate human neutrophils, for example, IL-8 or GROα (CXCL1), as compared with robust chemotaxis with the mouse cell line models (38). The HL-60 cells studied by Yadav et al. (66) also were modified to overexpress CXCR2, the receptor for CXCL1, and thus exhibit genetically enhanced chemotaxis.
In addition to effects on neutrophil chemotaxis and phagocytosis, and consistent with responses of EPRO-ic/ic cells lacking Lbr (19), EPRO-lamin A or C cells also exhibited deficient gp91phox expression together with a defective respiratory burst as mature neutrophils (Figs. 7A, 8). However, it is perplexing that neutrophils derived from MPRO-lamin A or C progenitors failed to exhibit either phenotype. We had previously attributed at least part of the deficient ROS in ic cells to the reduced sterol reductase activities provided by Lbr, but the similar defect in gp91phox expression shown in the present study adds further evidence that lobulation itself, or perhaps transcriptional regulatory mechanisms mediated by nuclear lobulation, must contribute to upregulation of this critical membrane-bound component of cytochrome b558 (20). Interestingly, maximal Cybb (gp91phox) gene expression occurs between the metamyelocyte and band neutrophil stages, whereas Ncf1 (p47phox) is upregulated at the promyelocyte-to-myelocyte transition (67). EPRO-lamin A or EPRO-lamin C cells both began at the early EML stage (considered even earlier than the myeloblast stage), whereas MPRO cells begin at the promyelocyte stage. Based on this consideration, it is possible that activation of Cybb may be more dependent on changes in nuclear composition as compared with Ncf1 or other genes encoding NOX proteins, or this dependency occurs at an earlier stage of neutrophil maturation when nuclear envelope-lamina composition begins to change. The MPRO and EPRO cells have already become committed to neutrophil maturation and express abundant amounts of multiple neutrophil-expressed genes, including the neutrophils gelatinase and collagenase, none of which is expressed by EML cells (20, 37). Thus, disruption of critical ratios between lamins and Lbr prior to the promyelocyte stage may have a more profound effect on certain genes critical to neutrophil functions, including Cybb. Alternatively, there are critical differences in the transcriptome of mature neutrophils as compared with their proteome, as has been recently indicated by studies of neutrophils from both healthy individuals and those with neutrophil-specific diseases (e.g., severe congenital neutropenia, chronic granulomatous disease, or leukocyte adhesion deficiency) (50). These differences may explain why we observed changes in gp91phox protein expression but not changes in Cybb transcripts from our RNA sequencing results. There is also the possibility of posttranscriptional or posttranslational modifications, neither of which would affect total mRNA levels.
Our RNA sequencing analyses (Fig. 9) reveal that enhanced lamin A expression influences transcription of multiple gene families at two distinct stages of neutrophil maturation, that is, promyelocytes (EPRO cells) versus functionally mature cells. These changes may be due to altered interactions of the lamina with Lbr, which binds to the heterochromatin binding protein HP1 and chromatin itself, proposed to anchor heterochromatin to the lamina and thereby epigenetically regulate chromatin activation (68). Lamin A/C also can form complexes with the lamina-associated polypeptide 2a (LAP2α), which has been shown to regulate the actions of pRb/E2f and thereby cell proliferation versus differentiation states (69–71). However, lamina composition may also directly regulate gene expression by forming scaffolds in the nucleus that can either serve as a platform in support of transcription factor activities and nuclear signaling proteins, or attenuate their functions by forming a “nuclear trap” (21). Altered lamina composition may also influence the linker of the nucleoskeleton to the cytoskeleton (LINC) complex, which mediates assembly of the nuclear membrane and nuclear pore complexes during mitosis, but also supports extracellular cues as they are relayed between the actomyosin cytoskeleton, the LINC complex, and chromatin rearrangements in maturing and/or terminally differentiated cells (72, 73). Whichever mechanism is disrupted by overexpressed A-type lamins remains unclear, but it is interesting to note the types of pathways that are affected. For example, lamin A overexpression causes changes in pathways involved in cell migration, indicating that loss of lamin expression facilitates pathways critical to movements of neutrophils during their recruitment to sites of infection, and potentially also their clearance during healing stages and wound repair (74–76). Processes involved in microtubule and actin cytoskeleton organization are also affected, suggesting that lamins may facilitate the control of actin polymerization during cell movements, again either during neutrophil recruitment or removal during healing responses. Cell adhesion and aggregation are additional mechanisms altered, so loss of lamins might prevent aberrant adhesion, perhaps important to recruitment into uninfected tissues in response to inflammation. Interestingly, signaling pathways associated with NF-κB expression were downregulated, providing a possible link to the observed loss of gp91phox expression (77). Lamin A levels may also have effects on chemokine or cytokine-mediated signal transduction or associated signaling pathways (e.g., those affecting metabolic processes or leukocyte activation; see (Fig. 9B, 9C), which could support chemotaxis and possible migration toward sites of inflammation not associated with infections. Our RNA sequencing results also reveal a connection between lamin A expression and endocytosis indicating that this lamin isotype plays a role in pathogen internalization during phagocytosis. Clearly there is more to be elucidated regarding how the nuclear lamina regulates gene expression in highly migratory cells, perhaps in connection to important roles of different LINC complex partners and how they are influenced by mechanotransduction and/or chromatin structural changes. Whether altered expression of either A-type lamin also affects components involved in NETosis, a process involving dramatic rearrangements of the neutrophil nucleus as chromatin is expelled and known to contribute to multiple human diseases, also requires investigation. Interestingly, recent studies have demonstrated that overexpression of lamin B attenuated NET formation, whereas reduced lamin B via treatment of a farnesyl transferase inhibitor enhanced NETosis (78). Comparisons of our transcriptome analysis to the proteome should also be explored, with specific attention to other LINC members that might interact with, or be affected by, A-type lamin expression, including Emerin, Sun2, and the nesprins (recently reviewed in Ref. 79).
Finally, our studies of lamin C alone add to an increasing body of evidence that this lamin isoform independently can modulate nuclear structural features and perhaps even regulate gene expression changes during cell differentiation. Interestingly, specific disruption of lamin C expression can result in either laminopathy or lipodystrophy (80, 81). Moreover, a transgenic mouse designed to only express the lamin C isoform showed no severe laminopathies characteristic of Lmna knockout, suggesting that lamin C can substitute for lamin A and therefore may support or augment those functions provided by lamin A—it was even suggested that lamin A is dispensable when replaced by the lamin C isoform (82). Lamin C may also be a potent player in mechanotransduction, perhaps even more important to this process as compared with lamin A depending on the cell type (e.g., stem versus cancer cells). Nonetheless, several studies have indicated that lamin A targets the C-isoform to the nuclear envelope and is therefore critical for any functions of lamin C in either supporting nuclear architecture, regulating mechanotransduction, or modulating chromatin activities, at least in the context of muscle cells and fibroblasts (83, 84). Our results contradict this notion. First, we observed normal localization of lamin C when overexpressed in either EPRO or MPRO cells, despite undetectable levels of lamin A. Second, the ectopically expressed lamin C alone caused hypolobulation, defects in chemotaxis and phagocytosis, and perhaps most importantly a disruption to Cybb (gp91phox) gene expression. Importantly, lamin A is generated from prelamin A by posttranslational modifications, including farnesylation plus carboxymethylation. These modifications promote actions of the metalloproteinase Zympste24 to cleave the last 18 aa, thereby creating mature lamin A. Both lamin B1 and B2 also undergo farnesylation at the C-terminal tail, and previous studies have indicated that farnesylation inhibitors, including FTase inhibitors, can influence nuclear structures by reducing farnesylated lamins (85). In contrast, lamin C is not known to be further processed and is most likely produced in its mature form. Lamin C also is not affected by the tail mutations observed in multiple forms of laminopathies or by aberrant functions of farnesylases. Also of note is that lamin C is still detectable in fully differentiated neutrophils, as shown in our studies and those of differentiated HL-60 myeloblasts (33), indicating that at least some lamin C expression may be needed in these cells to maintain nuclear integrity despite its highly flexible structure. Future studies of A-type lamins must include identifying functions of lamin C, as this isoform continues to emerge as a critical component of the lamina. Such studies should expand into additional cell types, not just in the context of hematopoiesis but also in additional tissues in which A-type lamins and the lamin B receptor are differentially expressed.
We thank Jack Lepine in the Core Research Facility at UMass Lowell for invaluable help performing imaging flow cytometry and RNA sequencing analyses.
This work was supported by the National Heart, Lung, and Blood Institute (Academic Research Enhancement Award Grant 1R15HL104593-2 to P.C.W.G).
K.S., M.G.H.P., K.M., and P.C.W.G. conceived the work and designed the experiments. K.S., M.G.H.P., K.M., A.M.B., R.M.G., and S.C.B. performed the experiments and analyzed the data. K.S., M.G.H.P., and P.C.W.G. wrote the manuscript and prepared the figures. All of the authors revised and edited the manuscript.
The online version of this article contains supplemental material.
Abbreviations used in this article
all-trans retinoic acid
erythroid, myeloid, and lymphoid
lamin B receptor
linker of the nucleoskeleton to the cytoskeleton
serum-opsonized zymosan A
relative fluorescence unit
reactive oxygen species
stem cell factor
TBS with 0.1% Tween 20
The authors have no financial conflicts of interest.