This is the author's final version of the contribution published as:
Malarial pigment hemozoin impairs chemotactic motility
and transendothelial migration of monocytes via
4-hydroxynonenal.
(PMID:25017964)
Skorokhod OA, Barrera V, Heller R, Carta F, Turrini F, Arese P, Schwarzer E.
Free Radic Biol Med. 2014 Oct;75 210-221.
doi:10.1016/j.freeradbiomed.2014.07.004. PMID: 25017964.
The publisher's version is available at:
https://www.sciencedirect.com/science/article/pii/S0891584914003189?via
%3Dihub
Abbreviations
RBC
red blood cell
4-HNE
4-hydroxynonenal
DC
dendritic cell
FMLP
formylmethionylleucylphenylalanine
HZ
hemozoin
MCP-1
monocyte chemoattractant protein-1
M-SFM
macrophage serum-free medium
PBMC
peripheral blood mononuclear cell
TNF
tumor necrosis factor
Chaps
3-[(3-cholamidopropyl)dimethylammonio]-1-propane sulfonate
HUVEC
human umbilical vein endothelial cell
ECGS
endothelial cell growth factor
FACS
fluorescence-activated cell sorting
SDS–PAGE
sodium dodecyl sulfate–polyacrylamide gel electrophoresis.
Keywords
Monocyte
Hemozoin
Migration
Lipoperoxidation
4-Hydroxynonenal
Actin
Coronin
Introduction
Malaria
Free radicals
During its growth in the host red blood cell (RBC) the malaria parasite Plasmodium falciparum digests hemoglobin and accumulates natural hemozoin (nHZ) in the digestive vacuole, which is expelled as a residual body (RB) upon
schizogony. nHZ consists of a biocrystallized β-hematin formed from undigested heme dimers and tightly adherent lipids and proteins from the parasite and the host [1], [2]. In vitro and in vivo, nHZ-laden RB and nHZ-containing late parasite stages (trophozoites and schizonts) are avidly phagocytosed, as confirmed by high proportions of HZ-laden monocytes and granulocytes in circulation [3],[4] or enriched in the microvasculature of spleen and liver [5], kidney [6], lung [7], brain [8], and possibly other organs and tissues. Several important functions are inhibited in nHZ-laden monocytes, such as oxidative burst [9], phagocytosis and extracellular killing [10], MHC class II-dependent antigen presentation [11], [12], and differentiation and maturation to monocyte-derived dendritic cells (DCs) [13], [14], [15]. Cell motility in response to chemotactic stimuli and cell migration are key events for monocytes to perform
phagocytosis and killing of invaders and to supply peripheral tissues with macrophages and DC precursors. Whereas the proinflammatory response to synthetic HZ (β-hematin) and nHZ is known to trigger leukocyte
recruitment [16], [17], [18], [19], the effect of nHZ phagocytosis on monocyte motility has received little attention so far.
This study was aimed first to elucidate the effects of nHZ on motility and transendothelial migration of monocytes and, second, to investigate the role of 4-hydroxynonenal (4-HNE), a lipoperoxidation product generated intracellularly by nHZ [20], in the motility impairment caused by phagocytosed nHZ. For these purposes, cell migration, morphology, and cytoskeleton organization in response to the chemoattractants MCP-1, formylmethionylleucylphenylalanine (FMLP), and TNF were assayed in in vitro cultured and nHZ-fed or HNE-treated primary human monocytes, and 4-HNE modifications of cytoskeleton proteins were identified.
Constrained cell movement and defective cytoskeleton dynamics observed in nHZ-laden monocytes were caused by 4-HNE and may contribute to imperfect adaptive immunity and secondary infections frequently observed in malaria patients.
Materials and methods
Reagents
Unless otherwise stated, reagents were obtained from Sigma–Aldrich (St Louis, MO, USA).
Primary cells
All donors were healthy volunteers and gave their informed written consent in accordance with the Declaration of Helsinki and after approval from and in accordance with the local research ethics committees of the Universities of Jena (Germany) and Torino (Italy).
Monocyte preparation
Peripheral blood mononuclear cells (PBMCs) were isolated from freshly collected blood as described [9], [11], washed twice, and suspended either in macrophage serum-free medium (M-SFM; Invitrogen, San Diego, CA, USA) at
3×107 cells/ml, for migration assays, or in RPMI 1640 medium for plating at 2×107 cells/well in six-well plates (Falcon,
Becton–Dickinson, San Jose, CA, USA) for 4-HNE–protein conjugate identification. The plates were incubated in a humidified CO2/air incubator at 37 °C for 30 min, nonadherent cells were removed by three washes, and M-SFM was
added. To confirm data obtained from PBMCs, selected experiments were performed with immunopurified monocytes. Monocytes were enriched to >85% purity as judged by flow cytometry analysis (FACS; FACSCalibur cytofluorograph, Becton–Dickinson; of CD14 and MHC class II expression) by negative selection with Dynabeads CD2 and CD19 (Dynal Biotech, Oslo, Norway) following the manufacturer’s instructions. Monocytes were suspended at 1×106 cells/ml
M-SFM.
Preparation, quantification, and opsonization of nHZ and latex beads
nHZ was isolated on a Percoll gradient from synchronous parasite culture supernatants after schizont rupture,
quantified, and opsonized with fresh human serum as described [9], [11],[21], [22]. Opsonized nHZ was tested negative for endotoxin [21] and DNase treatment was performed where indicated [21]. Quantification of nHZ was performed by luminescence assay [23]. For opsonization of latex beads, equal volumes of the manufacturer’s latex beads suspension, phosphate-buffered saline (PBS), and fresh human serum were mixed and incubated at 37 °C for 30 min.
Phagocytosis of nHZ and latex beads
Monocytes suspended in M-SFM were fed with nHZ or DNase-treated nHZ at 25 fmol heme/cell or latex beads at 6400 beads/cell for 3 h before the start of transendothelial or two-dimensional (2D) migration assays. Alternatively, adherent monocytes were cultured at 2×106 cells/well in six-well plates in the presence of 25 fmol nHZ heme/cell. After 30 min
the first nHZ crystals were observed inside the monocyte and phagocytosis was completed after 3 h incubation at 37 °C. At this time cells were washed with RPMI 1640 to remove noningested nHZ. At indicated time points after nHZ supplementation, cells were collected and solubilized for 2D electrophoresis and Western blotting as described below. Phagocytosis was quantified as indicated [23]. Apoptosis was tested by FACS after annexin V–FITC staining following the manufacturer’s specifications (Apoptosis Detection Kit; Sigma–Aldrich).
Monocyte 2D chemotactic migration assay
Two-dimensional migration assay was performed in μ-slides from Ibidi (Martinsried, Germany). Monocytes were suspended in M-SFM and left unfed (control) or fed with latex beads (control meal), opsonized nHZ, or DNase-treated nHZ or treated with 10 µM 4-HNE (Biomol, Hamburg, Germany). After 1 h phagocytosis/treatment in suspension, 2×105monocytes were seeded in the central channel of the μ-slide to adhere and complete phagocytosis during a further
2 h at 37 °C. Noningested nHZ, latex beads, and 4-HNE and nonadherent cells were then removed by three washes with RPMI 1640, and adherent cells were kept in M-SFM. Cytochalasin B was added at 25 µg/ml 30 min before the
chemoattractants to block actin reorganization. The assay was started (t=0) by loading cell culture medium without or with MCP-1 (100 ng/ml, R&D Systems, Minneapolis, MN, USA), TNF (10 ng/ml (200 U/ml), Peprotech, Rocky Hill,
NJ, USA), or FMLP (200 nM) in the lateral compartment of the µ-slide to create a chemotactic gradient. Migration of cells was assessed by microscopy at 0, 30, and 120 min. The assay allowed us to detect cell movements of ≥2 µm (detection limit). Motility in the absence of external chemoattractants was determined as the percentage of monocytes that (i) migrated in random directions [24] or (ii) did not migrate. In the presence of chemoattractants, motility was recorded as the percentage of monocytes that (i) migrated toward the chemotactic attractant (positive chemotaxis), (ii) migrated away from the chemotactic attractant, or (iii) did not migrate. Additionally, the motile cell subpopulations were analyzed for linear distances covered by the cells during the indicated times. Evidence from previous experiments showed that DNase treatment did not change the nHZ effect on motility, neither did the presence of lymphocytes in the assay, as inhibition of motility was reproduced with immunopurified monocytes (not shown).
Transendothelial migration assay
Migration across a human umbilical cord vein endothelial cell (HUVEC) monolayer was analyzed. Briefly, HUVECs were isolated from anonymously acquired umbilical cords within 1–24 h after donation by collagenase perfusion and were cultured in medium 199 (M199; BioWhittaker Europe, Verviers, Belgium), supplemented with 5% (w/v) human serum albumin, 15% (v/v) fetal bovine serum, 15 µg/ml human endothelial cell growth supplement (ECGS), and 7.5 U/ml heparin until reaching confluence [25]. After the first passage, the cells were seeded on transwell inserts (8.4 mm diameter, 3-µm-diameter pores) coated with 0.2% (w/v) gelatin (Greiner Bio-One, Frickenhausen, Germany) at a density of 2.5×105cells/insert and grown in 24-well plates to confluence with addition of ECGS (22.5 µg/ml). Only
completely confluent HUVEC monolayers (as judged by hematoxylin staining) were used in the transendothelial migration assays. Just before assay, HUVEC monolayers were rinsed with M199 and 1×106 washed nHZ-fed, latex-fed,
or unfed control PBMCs were added to the transwell insert. Transmigration assay was performed in M199 supplemented with 2.5% human serum albumin and 22.5 µg/ml ECGS and chemotaxis was induced by addition of 100 ng/ml MCP-1 in the bottom well. Transendothelial migration in the presence or absence of MCP-1 was measured after 2 h at 37 °C. Monocytes that had crossed the endothelial monolayer, the gelatin coating, and the insert filter pores were counted on the underside of the insert membrane and separately in the bottom well. For the former, the insert underside was rinsed twice and attached cells were scraped off. Cells dropped from filters during transmigration were collected from the bottom well. After decoration with anti-CD14 and anti-MHC class II antibody (BD Biosciences), the number of migrated monocytes was determined by FACSort (BD Biosciences) using CytoCount beads (Dako, Glostrup, Denmark) as standard. Both monocytes still adherent to the insert and those that dropped into the bottom well medium showed similar results and are therefore presented as a sum. Chemotaxis assay was performed in triplicate and
monocytes from three different donors were tested in three independent experiments. MCP-1-dependent transmigration (chemotaxis) was calculated as the difference between numbers of transmigrated cells in the presence and absence of MCP-1. Owing to considerable differences between donors in numbers of chemotaxing control cells, normalization was performed setting the control value for each donor to 1. Parallel analysis of transmigration was performed by
microscopy. For this, the upper side of the insert membrane was swabbed to remove nonmigrated and endothelial cells. Cells adherent to the underside of the insert membrane were stained with May–Grünwald–Giemsa Quick Diff (Medion Diagnostics, Düdingen, Switzerland) and monocytes were counted in 50 to 70 independent fields by microscopy (Zeiss, Oberkochen, Germany, with a 100× oil planar apochromatic objective with 1.25 numerical aperture).
Actin dynamics
Actin organization in CD14+ monocytes was analyzed 2 h after the start of the 2D chemotactic migration assay by
staining cells with Alexa Fluor 488 phalloidin, which binds to F-actin filaments, after fixation and permeabilization according to the manufacturer’s specifications (Molecular Probes/Invitrogen). Cells were washed three times with PBS for 5 min and visualized by fluorescence microscopy. Polarization as judged by the appearance of either the bright leading edge or the retraction tail [26], as well as actin foci in silent cells[27], was counted manually in at least 100 cells per experimental condition. To selectively recognize monocytes, R-phycoerythrin-conjugated anti-CD14 was applied, without revealing any interference between the two fluorochromes. According to cell treatment, monocytes from one to five donors were assayed in independent experiments.
One- and two-dimensional electrophoresis and Western blotting
Adherent cells were washed with ice-cold PBS-G (Phosphate buffered saline supplemented with 10 mM glucose), containing mannitol (5 mM final) and the Complete protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN, USA), and either lysed directly in Laemmli sample buffer for 1D electrophoresis or harvested in extraction buffer (8 M urea, 4% (w/v) Chaps, 40 mM Tris) at 75 μl/107 monocytes and 4 °C in the presence of Complete, mannitol, and
dithiothreitol for 2D electrophoresis. Samples were lysed by sonication and trituration and incubated with 150 U/ml Benzonase for 30 min at room temperature. Proteins were separated either by 1D SDS–PAGE using standard protocols or by 2D electrophoresis and stained with silver for visualization of the complete protein pattern or Coomassie G250 Brilliant Blue for spot excision or transferred to nitrocellulose membranes [28]. Membranes were probed with polyclonal rabbit anti-4-HNE–conjugate primary antibody (1:3000 dilution, Alexis Biochemicals, Enzo Life Sciences, Farmingdale, NY, USA) and horseradish peroxidase (HRP)-conjugated anti-rabbit IgG secondary antibody (1:10,000 dilution, Amersham Biosciences) to visualize and scan 4-HNE-positive proteins using ECL and Image Lab software 4.0 (Bio Rad Laboratories, Hercules, CA, USA). Human serum albumin was modified by 4-HNE (adapted from Crabb et al. [29]) and used as positive control in each Western blot to validate results.
Identification of 4-HNE–protein conjugates by FACS
Monocytes were washed in PBS–bovine serum albumin 1% (w/v) and incubated with saturating concentrations of anti-4-HNE monoclonal mouse primary antibody (1 h at room temperature, 1:50 dilution, Abcam, Cambridge, UK). After three washes, the bound antibodies were revealed by FITC-conjugated anti-mouse IgG secondary antibody (1:300 dilution). Mean fluorescence intensity was measured and data were analyzed using a FACSCalibur cytofluorograph (Becton–Dickinson) and CellQuest software (Becton–Dickinson).
Identification of 4-HNE–protein conjugates by mass spectrometry
Spots corresponding to major 4-HNE-positive signals in Western blots were excised from preparative 2D gels stained by Coomassie G250. Protein spot localization was facilitated by the silver-stained protein spots in the replicate gel. Excised spots were subjected to in-gel tryptic digestion and mass spectrometry analysis as described previously [30]. Proteins were identified by peptide mass fingerprinting against the UniProt sequence database (version 137) introducing 4-HNE as a variable modification from the default Mascot software list of modifications (Matrix Science, London, UK).
Further search parameters included taxa Homo sapiens, trypsin digest, protein molecular mass and IP, iodoacetamide modifications, monoisotopic peptide masses, two missed cleavages by trypsin, and mass deviation of 0.5 Da. Identification of protein spots was obtained from triplicate analysis.
Statistical analysis
Nonparametric Mann–Whitney U test was used to determine the significance of difference between the groups’ means (IBM SPSS Statistics 18; IBM, Chicago, IL, USA). If not otherwise indicated, p values of <0.05 were considered statistically significant. Standard error (SE) bars indicate the degree of uncertainty in the groups’ means [31].
Results
nHZ phagocytosis and 4-HNE treatment inhibit migration of monocytes toward chemoattractants
To investigate whether nHZ phagocytosis affects chemotactic migration of human monocytes, 2D migration was assessed by live-cell microscopy during the first 30 and subsequent 90 min exposure of cells to chemoattractants (see representative migrating monocytes that chemotaxed approximately 15 µm during 30 min in the assay; Fig. 1).
1.
Download high-res image (231KB)
2.
Download full-size image
Fig. 1. Representative images of monocytes moving toward chemoattractant MCP-1 in 2D chemotaxis µ-slides. Monocytes seeded in the central cross channel of 2D chemotaxis µ-slides (Ibidi) were exposed to a concentration gradient of MCP-1. Images of the same monocytes were acquired immediately (t=0, left) and 30 min (middle) after addition of 100 ng/ml MCP-1 in the lateral compartment of the µ-slide (arrow). The image on right results from the superposition of images taken at both time points and dots trace monocytes at t=0. Images are from one experiment of three performed with different donors and were captured with an inverted microscope (Leica DM IRB; Leica
Microsystems, Wetzlar, Germany) equipped with a 100× oil planar apochromatic objective with 1.32 numerical aperture, a Leica camera (DFC 420 C) and Leica DFC software, version 3.3.1.
About 50–70% of unfed control monocytes oriented and migrated along the MCP-1, TNF, or FMLP gradient (chemoattractant-directed migration in Figs. 2B–D), whereas only 14±5% (n=3) of nHZ-fed monocytes migrated toward MCP-1 (Fig. 2B), 12±2% (n=3) toward TNF (Fig. 2C), and 12% toward FMLP (Fig. 2D) in the first 30 min. Comparing the first 30-min period of chemotaxis with the subsequent 30- to 120-min period, all responsive monocytes moved immediately and no increase or recovery in motility was observed in controls and nHZ-fed cells. Hence, the functional impairment induced by nHZ persisted for at least 2 h. Additionally, nHZ inhibited migration in the absence of chemoattractant in 57±6% of cells (Fig. 2A, migration in random directions), suggesting perturbation of basic mechanisms of cell motility such as modifications in the cytoskeleton rather than chemotactic receptor modifications in nHZ-fed cells. We excluded phagocytosis-dependent actin rearrangements as the cause of the dramatic loss of mobility in nHZ-fed monocytes, as latex-fed cells were as mobile as unfed cells (Figs. 2B–D). As nHZ carries and generates substantial amounts of 4-HNE [20], we checked whether exogenous supplementation of low, biologically compatible amounts of 4-HNE was able to recapitulate those effects. Indeed, addition of 10 µM 4-HNE decreased the percentage of directed or randomly migrating monocytes similar to nHZ phagocytosis (Figs. 2A–C), pointing at 4-HNE conjugation as the possibly responsible process. Monocytes that remained mobile after nHZ uptake or 4-HNE treatment migrated shorter distances in response to MCP-1 (−54±13%; −57±17%), TNF (−46±18%; −55±15%), and FMLP (−49±21%) compared to the mobile unfed or latex-fed cells (Figs. 2F–H). In the absence of chemoattractant, 4-HNE-supplemented monocytes migrated additionally less far in comparison to unfed or latex-fed controls (−33±15% in the first 30 min and −43±22% in the subsequent 90 min; Fig. 2E).
1.
Download high-res image (783KB)
Fig. 2. nHZ phagocytosis and 4-HNE treatment inhibit migration of monocytes toward chemoattractants. Monocytes were seeded on 2D chemotaxis µ-slides, kept unfed (CTRL), fed with latex beads (LATEX) or nHZ, or treated with 4-HNE (10 µM final) or cytochalasin B (CYT B) (25 µg/ml final), before addition of chemoattractant. Latex-fed and Cyt B-treated monocytes were controls for phagocytosis-specific modification of motility and inhibition of actin-dependent motility, respectively. 2D migration was started (t=0) by adding the chemoattractant MCP-1 (100 ng/ml), TNF
(10 ng/ml), or FMLP (200 nM) or culture medium into the lateral compartment of the µ-slides. Migrated cells were recorded at 0, 30, and 120 min from addition of chemoattractant at 37 °C with an inverted microscope (Leica DM IRB; Leica Microsystems), equipped with a 20× phase-contrast objective or a 100× oil planar apochromatic objective with 1.32 numerical aperture, and a Leica camera (DFC 420 C). Leica DFC software (version 3.3.1) was applied. The imaging medium was culture medium with or without chemoattractant inside the Ibidi µ-slides during the whole cell migration period from 0 to 120 min. (A–D) Percentages of migrated and nonmigrated monocytes were reported for 0– 30 min (left) and for 30–120 min (right) in the absence (A;n=6) or presence of MCP-1 (B; n=3), TNF (C; n=3), or FMLP (D; n=1). The y-axis title on the left is valid for both plots in the row. n, number of monocyte donors assayed separately. At least 100 monocytes per donor and condition were analyzed. The plotted values are the mean values±SE of percentages of cells (A) that migrate in the absence of external chemoattractant in random directions or do not migrate or (B–D) that migrate in the presence of chemoattractant toward the chemoattractant (positive chemotaxis) or away from the chemoattractant or do not migrate. (E–H) Distances covered by migrating cells (motile subpopulations of (A–D)) were assessed for 0–30 min (left) and for 30 min–2 h (right) in the absence (E) or presence of MCP-1 (F), TNF (G), or FMLP (H). The y-axis title on the left is valid for both plots in the row. Mean values±SE of at least 100 analyzed monocytes per experimental condition from one representative donor of three analyzed are plotted. ⁎p<0.05, significant
difference vs CTRL.
The reported decrease in randomly directed motility of cells after nHZ phagocytosis but not after inert latex-bead phagocytosis (Fig. 2A) and the impaired ability to change direction of migration and to move efficiently versus chemoattractants of HZ-fed and 4-HNE-treated cells (Figs. 2B–D and F–H) suggest an HZ-specific defective reorganization of the cytoskeleton that is recapitulated by 4-HNE.
nHZ phagocytosis impairs transendothelial migration of human monocytes
The impact of motility loss on transendothelial migration (diapedesis) of nHZ-fed versus
unfed and latex-fed control monocytes was assessed with an in vitro endothelial barrier
model, a HUVEC monolayer grown to complete confluence on a gelatin-coated porous
membrane of a transwell insert. Monocytes were allowed to cross the endothelial barrier
and subsequently the pores of the membrane toward the chemoattractant MCP-1 for 2 h.
Whereas no difference from controls was detected in the basal transmigration of nHZ-fed
monocytes in the absence of chemoattractant, significantly fewer nHZ-fed compared to
unfed control monocytes crossed the endothelial barrier in the presence of MCP-1 (
Fig.
3A). nHZ loading impaired the MCP-1-dependent transmigration (chemotaxis) by
45.3±5.5% (mean±SE; n=3; Fig. 3B). Phagocytosis was completed before the assay was
started and had no influence on cell motility as concluded from latex-fed cells that
chemotaxed equally well as unfed controls (not shown). The impairment of the
transmigratory ability of nHZ-fed monocytes measured by FACS was confirmed by
microscopic inspection of the bottom insert membrane on which 23.9% (donor 1) and
19.6% (donor 2) fewer nHZ-fed monocytes were counted compared to unfed controls at
2 h. Apoptosis was below 3% in both HZ-fed and unfed monocytes and excluded as a
cause of the low motility of nHZ-fed monocytes (not shown). The HUVEC monolayer
remained intact during the 2-h assay duration and neither nHZ-fed nor unfed monocytes
seemed to cause leakiness as judged by microscopic inspection.
1.
Download high-res image (181KB)
2.
Download full-size image
Fig. 3. nHZ phagocytosis impairs transendothelial migration of human monocytes. Monocytes were fed nHZ or kept unfed as controls (CTRL; see Materials and methods for details). Transendothelial migration was started by adding 1×106 PBMCs on top of a confluent HUVEC monolayer in a gelatin-coated transwell insert and culture medium with or
without MCP-1 (100 ng/ml) in the bottom well. After 2 h at 37 °C in a cell culture CO2 incubator, transmigrated
CD14+ and MHC class II+ monocytes were collected and counted by FACS. Results are presented (A) as the number of
transmigrated monocytes per transwell in the presence and absence of MCP-1 and (B) as the difference of both (MCP-1-elicited transmigration) referred to the respective control value (normalized MCP-1-dependent transmigration or chemotaxis). Data are means±SE from three transwells per condition assessed for monocytes from three different donors in three independent experiments. Significant differences are indicated by ⁎ and respective p-values.
Altered polarization and actin reorganization in nHZ-fed monocytes: role of 4-HNE
Cell motility goes along with the formation of membrane protrusions pushed forward by polymerizing F-actin. Formation of a dense filamentous actin network in the leading edge indicates cellular polarization toward chemoattractants [27], [32]. To verify the potential impact of nHZ and 4-HNE on cell polarization, F-actin was visualized by fluorescence-labeled phalloidin in unstimulated and MCP-1-stimulated nHZ-fed and 4-HNE-treated monocytes and compared to the actin pattern predominant in unfed and latex-fed control monocytes. nHZ phagocytosis as well as 4-HNE-treatment impaired the actin reorganization of monocytes in response to chemoattractants (Fig. 4A). After MCP-1 addition, three subsets of cells were distinguished with respect to their F-actin organization. First, highly responsive polarized cells with a highly dense F-actin network at the leading edge (L in micrographs,Fig. 4A) and/or actin-retraction fibers on the opposite side of the cell (uropod U, Fig. 4A) were frequent in control and latex-fed cells. These structures accounted for 40.5±4.4% of control cells, but were rare after nHZ phagocytosis and 4-HNE treatment (4.0±2.0 and 2.5±1.5% of whole cell population, respectively; Fig. 4C). Polarization became visible already 30 min after MCP-1 addition and more evident until 2 h in responsive control monocytes, whereas nHZ-fed and 4-HNE- and cytochalasin B-treated cells failed to become asymmetric during the observation period. A very low basic polarization of adherent cells (<3%) in the absence of a chemoattractant was observed irrespective of previous phagocytosis, whereupon nHZ and cytochalasin B caused an additional, significant decrease in polarization (Fig. 4B). Second, we observed cells with exclusively central focal staining. This actin pattern is characteristic of actin-rich focal adhesion protrusions or podosomes for the anchorage of cells that undergo actin reorganization (P in Fig. 4A). This was the predominant actin organization of adhering cells before MCP-1 stimulation (approximately 60%) with the exception of 4-HNE-treated and cytochalasin B-blocked cells (Fig. 4B). After MCP-1 exposure 40–50% of all cells showed this pattern irrespective of precedent treatment (Fig. 4C). Density and brightness of these foci were clearly increased by MCP-1 treatment in control and latex-fed cells but remained faint and sporadic in nHZ-fed and 4-HNE-treated cells (Fig. 4A). Third, silent cells negative for any actin reorganization due to MCP-1 were characteristic of nHZ-fed and 4-HNE-treated cells, accounting for 55 and 63% of cells, respectively, but rare in controls (11% of cells) (Figs. 4A and C). The similarity between nHZ-fed and 4-HNE-treated cells and cells with actin polymerization inhibited by cytochalasin B treatment (Fig. 4C) suggests impaired actin reorganization elicited by 4-HNE.
1.
Download high-res image (946KB)
2.
Download full-size image
Fig. 4. nHZ phagocytosis and 4-HNE treatment inhibit polarization and actin reorganization in monocytes. Monocytes were seeded on 2D chemotaxis µ-slides and were unfed (CTRL), fed with latex beads (LATEX) or nHZ, or treated with 4-HNE (10 µM final) or cytochalasin B (CYT B) (25 mg/L final), before addition of MCP-1. Latex-fed and Cyt B-treated monocytes were controls for phagocytosis-specific actin rearrangement and inhibition of actin polymerization, respectively. Assay was started (t=0) by adding MCP-1 (100 ng/ml) or culture medium into the lateral compartment of µ-slides (at the top of the micrograph). After 2 h exposure to MCP-1 or culture medium, the cells were fixed,
permeabilized, stained with Alexa Flour-488 phalloidin probe, and visualized by fluorescence microscopy at room temperature to check cell polarization. (A) Representative images of fluorescent phalloidin-stained monocytes exposed or not to a MCP-1 concentration gradient. Parameters used to recognize the state of actin polymerization and cellular polarization in different subcellular regions were leading edge (L), uropod (U), and central punctate staining sites (P) as explained under Results. (B, C) Percentages of monocytes with distinct actin polymerization pattern being L and/or U positive (polarized) or P positive (actin foci only) or without defined actin polymerization (nonpolarized) were recorded in the (B) absence and (C) presence of MCP-1. Mean values±SE of cell percentages are shown. (B) CTRL (n=5), LATEX (n=3), nHZ (n=3), 4-HNE (n=2), CYT B (n=3). (C) CTRL (n=6), LATEX (n=2), nHZ (n=4), 4-HNE (n=2), CYT B (n=3). n, number of donors whose monocytes were separately assayed. At least 100 monocytes were counted per donor and experimental condition. Significance of difference vs CTRL is indicated: ⁎p<0.05 and ⁎⁎p=0.06. Images
were captured with a Leica DM IRB microscope equipped with a 100× oil planar apochromatic objective with 1.32 numerical aperture and a Leica camera (DFC 420 C) and Leica DFC software, version 3.3.1.
Identification of 4-HNE–protein conjugates in nHZ-fed monocytes
The enhanced generation of 4-HNE in nHZ-fed monocytes covalently modified a number of proteins. Significantly elevated levels of 4-HNE–protein conjugates were detected by FACS already 3 h after the start of phagocytosis. Conjugates increased further until 12 h and persisted until 24 h (Fig. 5A) in nHZ-fed monocytes, whereas latex-fed or unfed control monocytes maintained low conjugate levels during the same period. To identify proteins causally involved in monocyte motility impairment, 4-HNE-conjugated proteins were analyzed in whole-cell lysate of nHZ-fed monocytes by a combined Western blotting and proteomic approach. Lysate proteins separated by 1D electrophoresis showed an early increase in 4-HNE-modifications and a stable 4-HNE–protein conjugate pattern, which persisted in each replicate for at least 24 h after the start of phagocytosis (Fig. 5B). Western blotting revealed a mildly oxidized state in control cells, too, identified by a donor-dependent and transiently appearing pattern of 4-HNE–protein conjugates. Western blots of proteins separated by 2D electrophoresis confirmed these data (not shown). Protein spots
corresponding to major positive signals in Western blot were assayed by MALDI-TOF. A total of six 4-HNE-modified monocyte proteins were identified (Fig. 5C, Table 1). Three proteins were cyto- and nucleoskeleton members (β-actin, coronin-1A, and lamin A/C; P60709, P31146, and P02545, respectively, UniProt protein database). Two proteins were glycolytic enzymes (enolase and triosephosphate isomerase, P06733 and P60174, UniProt), and one was a component of the splicing-enhancer complex (heterogeneous nuclear ribonucleoprotein H, P31943, UniProt).
1.
Download high-res image (702KB)
2.
Download full-size image
Fig. 5. Detection of 4-HNE–protein conjugates in nHZ-fed monocytes. (A) Kinetics of 4-HNE conjugation of monocyte membrane proteins. Human monocytes were analyzed for 4-HNE–protein conjugates at indicated times after
phagocytosis of nHZ and compared to unfed (CTRL), latex beads-fed (LATEX), or repeatedly 4-HNE-treated (10 µM at 0, 6, and 12.5 h) monocytes. Modified proteins were recognized by specific anti-4-HNE–protein conjugate antibodies in intact cells by FACSCalibur and measured as mean fluorescence intensity (MFI). Means±SE (n=3 different donors) are plotted. Significance of difference vs corresponding condition at time 0 is indicated: ⁎p<0.05. (B) Kinetics of 4-HNE
conjugation of proteins extracted from monocytes. 20 µg of protein from adherent nHZ-fed and unfed monocytes was separated by 10% SDS–PAGE, Western blotted, and probed with specific anti-4-HNE–protein conjugate antibody (see Materials and methods for details). 4-HNE-modified protein bands were visualized by HRP-conjugated anti-mouse secondary antibody and ECL luminol-enhanced detection and assayed by densitometry scan (Image Lab software 4.0, Bio-Rad Laboratories). (C) Identification of 4-HNE-modified proteins by mass spectrometry. 150 µg of proteins extracted 24 h after the start of nHZ phagocytosis from monocyte lysate was separated by 2D electrophoresis in duplicate gels. IEF and SDS PAGE indicate the first and second direction of protein separation by isoelectric focusing (see also pH scale at the lower edge) and PAGE, respectively. The pattern of separated and Coomassie G250-stained monocyte proteins is shown. Spots that corresponded to 4-HNE-positive spots in the Western blot of the duplicate gel (indicated by lines) were excised (see Materials and methods for details). In-gel-digested proteins were identified by MALDI-TOF and proteins with mass spectrometrically confirmed 4-HNE conjugates are listed.
Table 1. Identification of 4-HNE-modified proteins and 4-HNE-modified functional sites by MALDI-TOF and peptide mass fingerprinting
Synonymous protein
names
UniProt protein
identification
MW
SDS–
PAGE
Expected
MW
pIin
IEF
Expected
pI
Sequence
coverage
(%;P
m/P
tot)
Identified
4-HNE amino
acid
modification
Known molecular
functions of
domains/regions
hosting 4-HNE
modifications
•
β-Actin•
Actin, cytoplasmic 1P60709
44
41.737
5.7 5.29
31; 10/21 His371
SD1 interaction
ACTB_human
Lys373 (in
subdomain
(SD) 1)
with coronin-1A
triggers the
binding of
coronin-1A to a
second binding
site in SD4 of actin
•
Coronin-1A•
Coronin-like protein A (Clipin-A)P31146
53
51.026
6.4 6.25
16; 10/26 His130
Binding to -actin
β
COR1A_human
Lys132 (in
central
Ile111–
Glu204
region)
•
Coronin-like protein p57
•
Lys331
Binding to -actin
β
Cys332 (in
C-terminal
Ile297–
Lys461
region)
•
α-Enolase•
2-Phospho-D -glycerate hydrolyaseP06733 and
P06733-2
50
47.36
7.3 6.99
36; 11/21 Lys325
Not described
ENOA_human
7.7 7.55
29; 11/21 Lys342 or
357
Cys336, 338,
or 356
•
Triosephosphate isomeraseP60174
26
26.8
7
6.51
50; 12/21 Cys255 or
Lys256
Not described
TPIS_human
•
Prelamin A/C,
cleaved into Lamin A/C (=70-kDa lamin; =renal carcinoma antigen NY-REN-32)
P02545
70
74.4
7.1 6.44
29; 17/21 Lys180
Sumoylation site,
important for
nuclear
localization, with
functional
consequences in
myocytes
LMNA_human
Lys201 or
Lys208
•
Heterogeneous nuclear
ribonucleoprotein H
P31943
60
49.3
6.4 5.89
40; 10/21 Lys185,
His194
Not described
HNRH1_human
Lys200
UniProt database accession numbers and protein names are reported. Molecular weights (MW) determined experimentally by SDS–PAGE, predicted MW (expected MW), experimental IEF point (pI), and expected pI are reported. The ratio between the number of matched peptides and the total peptide number is expressed asPm/Ptot next to
sequence coverage (%). 4-HNE-modified proteins with a score above 57 (recognized rate for significant identification) were identified by mass spectrometry as described under Materials and methods with the free search program
MASCOT 2.3.02 (http://www.matrixscience.com). 4-HNE modification determination is reported as sequential number of amino acids involved in the modification.
Sites of β-actin and coronin-1A modification by 4-HNE
β-Actin and coronin-1A, the actin-associated protein that participates in the regulation of F-actin
assembly/disassembly [34], are important cytoskeleton components involved in monocyte motility such as migration and phagocytosis. The mass spectrometric results for both proteins are summarized in Table 1. In β-actin, 4-HNE modifications of His371 and Lys373 in subdomain 1 (SD1; P60709, UniProt) were detected in nHZ-fed but not in
control monocytes. In coronin-1A, 4-HNE conjugates with His130, Lys132, Lys331, and Cys332 were detectable in the central Ile111–Glu204 and the C-terminal Ile297–Lys461 region (P31146, UniProt). The modified amino acids are localized in interaction-regulating domains (for details see Table 1) and may thus explain defects in the actin reorganization process required for cell motility. Of the remaining four identified 4-HNE-conjugated proteins, only lamin A/C carried 4-HNE in a known functionally important site that regulates the protein import into the nucleus of myocytes as detailed in Table 1. There is no immediate functional association with the herein-described motility impairment and further studies will be necessary to prove the altered lamin A/C localization in nHZ-fed monocytes.
Discussion
nHZ and nHZ-containing late-stage parasites are avidly phagocytosed by monocytes in vivo[3], [4] and in vitro[9], and the undigested crystalline β-hematin core of nHZ persists in the phagocyte lysosome for long periods of time [35]. Several functions of monocytes and monocyte-derived macrophages and DCs are impaired after phagocytosis of nHZ [14], [17],[21], [36], [37]. The contact between the pro-oxidant β-hematin core of nHZ and lipids of the phago- and lysosome membrane enhances the nonenzymatic heme-catalyzed peroxidation of polyenoic fatty
acids [9], [36] generating several bioactive lipoperoxide derivatives [20],[38]. Many toxic effects observed in nHZ-laden monocytes were elicited by those derivatives, notably HETEs, HODEs, and the terminal aldehyde
4-HNE [14], [15], [20], [36], [38]. 4-HNE is a highly reactive final lipoperoxidation product, which stably binds to proteins [39], [40]. In vitro, high levels of 4-HNE and 4-HNE–protein conjugates were found in nHZ-fed
monocytes[14], [20], [41] and 4-HNE added exogenously in plausible concentrations to monocytes recapitulated a number of toxic nHZ effects [14], [36], [42]. In vivo, elevated levels of 4-HNE conjugates were found in the membrane of RBCs of malaria patients [43], [44].
Many of the monocyte functions impaired by nHZ, such as repeated phagocytosis and differentiation to functional DC, depend on efficient cell motility in vivo. In malaria patients suppressed monocyte and neutrophil motility and
chemotaxis have been reported [45], [46]. Another study showed increased migration ability in response to lymphoid chemokines, after partial maturation of human DCs induced by the synthetic HZ analog β-hematin [47]. However the role of nHZ, and its pro-oxidant derivative 4-HNE, in such suppression either in vivo or in vitro has received no attention so far. Here we show that motility and migratory response of nHZ-fed monocytes to chemotactic MCP-1, TNF, and FMLP stimuli was bluntedin vitro, as evidenced by decreased transendothelial crossing and reduced 2D chemoattractant-independent migration and chemotaxis. In nHZ-fed monocytes acquisition of a polarized morphology was impaired and the ability to form two opposing poles, leading edge and uropod, and to reorganize F-actin filaments in the lamellipodium at the front was incapacitated (Fig. 4). Already 30 min after addition of MCP-1 stimulus only approximately 10% of nHZ-fed monocytes started the polarization process vs 40–50% in controls (not shown), which is the usual rate for intact monocytes in in vitro polarization assays [24].
The observed alterations in cell polarization and actin reorganization appeared to be causally related to HNE, as 4-HNE supplementation recapitulated the inhibition of cell nondirected motility and chemotaxis, and 4-4-HNE conjugates with the cytoskeleton proteins β-actin and coronin-1A were observed in nHZ-fed monocytes (Fig. 2, Fig. 4, Fig. 5). 4-HNE–protein conjugates were detected 30 min after nHZ phagocytosis, suggesting rapid oxidative events in nHZ-fed monocytes in accordance with the time course of the motility decline. Early studies have shown that
submicromolar/low-micromolar concentrations of 4-HNE promote chemotactic activity on human and rat
neutrophils [48], [49], [50], [51], [52]. It is plausible that low levels of 4-HNE, a bioactive component of nHZ, may contribute to the initial stimulation of monocyte chemotaxis. By contrast, very high 4-HNE levels (up to 50 µM peak levels) were generated inside the nHZ-fed phagocyte [20]. Those high 4-HNE levels may be responsible for the long-term cell paralysis observed here. The susceptibility of actin to 4-HNE is known from a previous in vivo study [53], in which the intraperitoneal administration of iron chelate to rats elicited the formation of 4-HNE–actin conjugates in the kidneys. 4-HNE added to purified actin monomers modified the amino acid Cys374 but did not change the steady-state polymerization levels in vitro[54]. Actin Cys374 was suggested to play a scavenger role against oxidative
stress [55] but seemed to be less accessible to 4-HNE in vital cells than its neighboring amino acids, which are probably involved in the interaction of actin with other cytoskeleton components. For example, in a cellular model of lung injury 4-HNE supplemented at 1–25 μM was shown to alter the remodeling of actin fibers [56], [57]. We show here how multiple 4-HNE modifications of β-actin and coronin-1A in protein interaction sites [58], [59], [60] affected actin rearrangement (Fig. 4) in nHZ-fed monocytes. The decrease in cellular mobility of nHZ-fed monocytes in the absence of chemoattractants (Fig. 2A) highlights the importance of cytoskeleton protein modifications by 4-HNE to elicit motility loss. β-Actin and coronin-1A are essential for cytoskeleton organization [58], [59], [60]. Actin, the main component of the cytoskeleton, regulates cell morphology, phagocytosis, vesicle trafficking, adherence, polarity, and diapedesis [61]. Coronin-1A, proposed to stabilize F-actin filaments [62], is enriched in the leading edge of
chemotaxing cells [63] and is essential for granulocyte motility [64]. In monocytes and macrophages,
coronin-dependent processes have been described during phagosome formation [65]. Both migration and phagocytosis strongly depend on actin-filament turnover, and nHZ-fed monocytes were unable to repeat the phagocytic cycle [9].
Circulating blood monocytes contribute substantially to immunological surveillance [66], [67]by patrolling as sentinels the intravascular space for signs of infection or damage, with long-range lateral migration [66]. Patrolling monocytes extravasate and secrete chemoattractants to recruit further immune cells. Phagocytosis of nHZ by monocytes may lead to inefficient patrolling and transendothelial chemotaxis. Defects in monocyte motility and migratory ability may contribute to the incomplete clearing of sequestered parasites, facilitate superinfections, and explain the poor adaptive immune response to the malaria parasite, all hallmarks of severe falciparum malaria [13], [68], [69], [70]. Further studies are necessary to show whether paralysis of nHZ-laden monocytes enhances local inflammatory reaction. Immobile nHZ-fed monocytes may accumulate in the microvasculature, as shown in cerebral microvessels from Malawian children dead of cerebral malaria [8]. Cytoadherent-paralyzed HZ-fed monocytes in microvasculature may continue to produce proinflammatory molecules with deleterious consequences for the endothelium.
Blood monocytes contribute directly to the innate immune defense against the parasite and link innate and adaptive immunity to blood-stage malaria supplying peripheral tissues with macrophage and DC precursors [36]. The primary site of immune responses againstPlasmodium parasites is the spleen [71]. Antigen-presenting cells (APCs) of the marginal zone of the white pulp, such as monocytes/macrophages and DCs, phagocytose parasitized RBCs and nHZ in the red pulp and may initiate adaptive immune responses by migrating toward T cell zones deeper in the white pulp for T cell activation. The histology of the spleen of adults dying from malaria [68] revealed that HZ-laden macrophages and myeloid DCs accumulate in the red pulp/marginal zone and fail to migrate into white pulp, thereby missing T cell activation. It seems possible that the inability of APCs to move toward the white pulp is caused by nHZ and that the connected defects in actin polymerization contribute basically to it in addition to changes in receptor, cytokine, or chemokine profiles [13], [47], [72]. Therefore, the partial paralysis of APCs after parasite uptake would explain the imperfect induction and/or maintenance of immune responses in Plasmodium infections.
In summary, this study has shown that nHZ-laden monocytes displayed impaired chemotactic and random migration and decreased transendothelial crossing, dependent on cytoskeleton alterations causally related to 4-HNE generated by nHZ-elicited lipoperoxidation. HNE supplementation recapitulated inhibition of cell motility and chemotaxis, and 4-HNE-conjugates with β-actin and coronin-1A were observed in both 4-HNE-treated and nHZ-fed monocytes. nHZ/4-HNE-linked impaired motility of monocytes may offer an overarching mechanistic explanation connecting impaired phagocytosis, defective induction, and insufficient maintenance of immune responses in Plasmodiuminfections.
