The human host-defence peptide (HDP) LL-37 not only displays anti-microbial activity but also immune-modulating properties that trigger intracellular signalling events in host cells. Since the cytolytic activity of high LL-37 concentrations affects cell viability, the function of LL-37 requires tight regulation. Eukaryotic cells therefore benefit from protective measures to prevent harmful effects of LL-37. p33, also known as globular C1q receptor (gC1qR), is reported to act as an LL-37 antagonist by binding the peptide, thereby reducing its cytotoxic activity. In the present report, we show that high levels of endogenous p33 correlate with an increased viability in human cells treated with LL-37. Sub-cellular localization analysis showed p33 distribution at the mitochondria, the plasma membrane and in the cytosol. Strikingly, cytosolic overexpression of p33 significantly antagonized detrimental effects of LL-37 on cell fitness, whereas the reverse effect was observed by siRNA-induced down-regulation of p33. However, modulation of p33 expression had no effect on LL-37-induced plasma membrane pore forming capacity pointing to an intracellular mechanism. A scavenging function of intracellular p33 is further supported by co-immunoprecipitation experiments, showing a direct interaction between intracellular p33 and LL-37. Thus, our findings support an important role of intracellular p33 in maintaining cell viability by counteracting LL-37-induced cytotoxicity.
Host-defence peptides (HDPs) are at the forefront of innate immune defence mechanisms against microbial infections. Although initially believed to be solely involved in disrupting bacterial cell membranes, it is now clear that their functions in the host response to infection are far more complex . For instance, HDPs have been revealed to trigger inflammatory reactions in eukaryotic cells , act as chemoattractant  or promote cellular differentiation .
The cathelicidins represent one family of HDPs, with human cationic antimicrobial protein-18 (hCAP-18) as the only currently known human member . As for all cathelicidins, hCAP-18 represents a two-domain pro-peptide form, from which the C-terminal peptide is released . hCAP-18 is produced by epithelial cells and neutrophils and upon exocytosis processed to the HDP LL-37 by proteinase 3, which is secreted from activated neutrophils [2,3]. LL-37 possesses both hydrophobic and hydrophilic properties  and exerts its anti-microbial effects through disruption of the bacterial cell wall causing cell lysis of both Gram-negative and Gram-positive bacteria and by neutralizing bacterial lipopolysaccharides [6–8]. Apart from its anti-microbial properties, LL-37 has been reported to influence inflammatory cell recruitment, modulate Toll-like receptor (TLR) function and type I interferon responses and play a role in auto-immunity .
Under pathological conditions, high levels of LL-37 have been found to reduce cell viability and promote apoptosis in osteoblasts, vascular smooth muscle cells, periodontal ligament cells, neutrophils, airway epithelial cells and T-cells [10–15]. Notably, very high concentrations, exceeding several 100 μM of LL-37, have been detected in lesions from patients suffering from the auto-immune diseases psoriasis, rosacea and ulcerative colitis, suggesting an important role of LL-37 in these disease processes [16–18]. Also under other inflammatory conditions, such as chronic periodontitis, concentrations of LL-37 can locally reach levels within the micromolar range .
In order to counteract the detrimental effects evoked by LL-37, the host relies on endogenous mechanisms. Though this area of research is not widely explored, one putative candidate in host protection is the human protein p33, also referred to as p32 or globular C1q receptor (gC1qR). p33 is expressed in different human cell types such as neutrophils, endothelial cells, lymphocytes, platelets and osteoblasts [20–22] and was originally described as a complement C1q-binding protein [23,24]. Mature p33 comprises 209 amino acid residues and derived from a precursor protein carrying a mitochondrial signal sequence. p33 is implied to be trimeric, with one of its two surfaces highly negatively charged due to a high number of glutamic and aspartic acids [28 and 20 amino acids (aa) respectively; 25,26]. Therefore, p33 was proposed to act as a scavenger for cationic HDPs such as LL-37. Indeed, exogenously applied trimeric p33 binds LL-37 and other HDPs. HDPs show high affinity for the negatively charged β-sheet (aa 115–135) of p33, thereby neutralizing their anti-microbial activity . p33 has been proposed to prevent HDP-induced cytotoxicity in endothelial cells and erythrocytes through this scavenging mechanism . Moreover, exogenous p33 (1–10 μM) antagonizes LL-37-induced apoptosis in osteoblasts in a concentration-dependent manner .
In the present study, we show an important role of endogenous, intracellular p33 in host cell viability. We find that LL-37 differentially attenuates the viability of various cell types depending on their intracellular p33 level and demonstrate that p33 antagonizes the detrimental effects of LL-37. Together, our results suggest that the levels of intracellular p33 determine the extent of the cellular damage caused by LL-37.
Cells and cell culture
The human osteoblast-like MG63 cell line and human HeLa cells were obtained from the A.T.C.C. The human keratinocyte HaCaT cell line and neonatal dermal fibroblasts (HDFa) were purchased from CLS Cell Lines Service GmbH and Cascade Biologics respectively. The human periodontal ligament fibroblast (PDL) cells were obtained from teeth extracted for orthodontic reasons, as described previously . The patients (boys and girls 14–15 years of age) and their parents were informed both orally and in writing and a written consent was signed. This procedure was approved by the Human Ethical Committee, Lund University, Lund, Sweden. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM)/Ham's F12 (1:1) (Life Technologies) supplemented with antibiotics (penicillin 50 units/ml, streptomycin 50 μg/ml, Biochrom GmbH) and 10% FBS (Biochrom GmbH) and kept at 37°C under 5% CO2 in air in a water-jacketed cell incubator. The cells were trypsinized (0.25% trypsin/EDTA) and counted using the Luna™ Automated Cell Counter system (Logos Biosystem) in accordance with the manufacturer's instructions. Cell morphology was evaluated in photographs obtained by a Nikon TMS phase-contrast microscope equipped with a digital camera (Pixelink, Nikon). The cell lines were used up to passage 30. The primary PDL cells were used in passages 2–4. Cells were treated with LL-37 (Bachem AG) dissolved in DMSO. Controls received DMSO-vehicle as appropriate. Identical responses to LL-37 were obtained irrespective of passage number.
Transfection with p33 siRNA
The p33 siRNA (Hs_C1QBP_6) and non-coding (NC) scramble control (All Stars Negative Control), both purchased from Qiagen, were transfected into the cells using Oligofectamine transfection reagent (Life Technologies). Briefly, an RNA/oligofectamine mix was prepared in Opti MEM medium (Life Technologies) according to the manufacturer's instructions. Cells were seeded at a density of 1.2×1050 per culture dish and treated with p33 siRNA and scramble RNA at an RNA concentration of 20 nM for 24 and 96 h for MG63 and HaCaT cells respectively. The cells were allowed to equilibrate under growth-stimulated conditions for 48 h followed by a second transfection, as described above. The HaCaT cells were transfected for a longer time because they were less sensitive to the transfection. Cells were trypsinized and re-seeded at identical density for functional experiments in a 1:1 mixture of transfection medium and fresh culture medium.
Plasmid DNA for transient transfection
Human p33 lacking the mitochondrial signal sequence was used for transient transfections and overexpression. p33 was PCR amplified from the p33-pMAL-c2 vector described in Herwald et al.  using the following primers. The 5′-cgcg-gatccaccatgggcctgcacaccgacggagac-3′ forward primer introduced a BamHI restriction site as well as a Kozak consensus sequence, whereas the reverse primer 5′-ccgctcgagct-actggctcttgacaaaactcttgagg-3′ introduced a stop codon and a XhoI restriction site. PCR product and pcDNA 3.1(+) mammalian expression vector (Life Technologies) were both treated with FastDigest enzymes BamHI and XhoI (both from Fermentas) and purified by gel extraction (Qiagen) according to the manufacturer's instructions. pcDNA 3.1-p33 expression vector was then obtained by cloning the p33 fragment into pcDNA 3.1(+) using a DNA ligation kit (Novagen). Introduction of p33 was verified by sequencing using primers based on T7 promoter and BGH Reverse primer sequences (Life Technologies).
Transient transfection of plasmid DNA
HeLa cells were seeded in 96- or 12-well plates at a density of 1.5×1040 or 2.0×1050 per well respectively and transfected for 24 h with pcDNA3.1-p33 expression vector or pcDNA 3.1 control vector using both Lipofectamine 3000 and P3000 reagents (both from Life Technologies) according to the manufacturer's instructions. Briefly, cells were transfected in Opti MEM medium using 4 μg of DNA, 8 μl of P3000 reagent and 6 μl of Lipofectamine 3000 reagent per millilitre of medium (43 μl of transfection mixture per well in 96-well plates and 400 μl per well in 12-well plates). Cells were left in transfection mixture for 16 h before an equal volume of DMEM supplemented with 20% FBS was added, giving 2 μg/ml as the final vector concentration. No difference in cell viability was observed between cells transfected with the pcDNA 3.1-p33 and control vector. Successful transfection and overexpression of p33 was confirmed by quantitative real-time RT-PCR and Western blotting.
Cell viability was assessed using the MTT assay as described by Carmichael et al. . Briefly, cells were seeded at a density of 1.5×1040 cells/well in 96-well plates. After 24 h, cells were treated with or without LL-37 for 3 h in Phenol Red-free DMEM culture medium at 37°C in a cell incubator. Phenol Red was omitted since it may interfere with the absorbance reading. The LL-37-containing medium was removed and then the cells were incubated with MTT (0.5 mg/ml) in cell culture medium for 1 h at 37°C. After 1 h, the MTT-containing medium was removed and the blue formazan product dissolved by adding 100 μl of DMSO per well. Absorbance was monitored at 540 nm in a Multiscan GO Microplate Spectrophotometer (Thermo Scientific).
Pore formation by LL-37 was assessed by measuring the release of lactate dehydrogenase (LDH) from cells seeded in 96-well plates. Cells were incubated with or without LL-37 for 30 min in Phenol Red-free DMEM culture medium at 37°C in the cell incubator. After incubation, the plate was centrifuged at 257 g for 10 min and 30 μl of supernatant was transferred to a new plate. Total LDH release was obtained by cytolysis via sonication of untreated control cells. Fresh medium was utilized as a blank. A NADH solution in water (180 μl, 56 mM TRIS, 5.6 mM EGTA and 170 μM β-NADH) was added followed by sodium pyruvate in water (20 μl, 14 mM) and the absorbance of NADH, which had not been reduced by pyruvate-LDH, was read at 340 nm in a Multiscan GO Microplate Spectrophotometer (Thermo Scientific) for 1 h at 30°C. LDH release in response to LL-37 was normalized to total release of LDH.
For immunocytochemistry, cells were cultured on coverslips. After reaching 80% confluence, cells were washed with PBS, fixed in 4% (w/v) paraformaldehyde (PFA) and permeabilized in 0.2% (v/v) Triton X-100 for 10 min. Non-specific binding sites were blocked with 2% (w/v) BSA before polyclonal antiserum to p33 raised in rabbits (custom made antibody, Innovagen AB, 1:200 dilution) was added overnight at 4°C. The following day, cells were incubated with secondary goat anti-rabbit Alexa Fluor 555 or Alexa Fluor 488 (Invitrogen, both 1:500 dilutions) for 2 h at RT. In some experiments, cells were co-stained with the mitochondria-selective probe MitoTracker (MitoTracker Orange CM-H2TMRos, Molecular Probes). The cells were incubated with MitoTracker (500 nM) according to manufacturer's instructions. Cells were washed with PBS, rinsed with water and then the coverslips were mounted using Aqua-Poly/Mount mounting medium (Polysciences). The immunoreactive signal and MitoTracker staining were analysed using both a conventional Olympus BX60 fluorescence microscope and a laser scanning confocal microscope (LSM5 PASCAL, Carl Zeiss AG). No immunoreactivity was observed after omission of the primary p33 antibody.
After washing with PBS, fixation buffer was added and cells were scraped-off from the culture dish. Cells were embedded in Epon and subjected to antigen retrieval with metaperiodate . Grids were blocked with 5% (v/v) goat serum diluted in 0.2% BSA (pH 7.6, Aurion) for 15 min, followed by incubation with polyclonal rabbit anti-p33 (Innovagen AB, 1:80 dilution) overnight at 4°C. Samples were then incubated with gold-conjugated goat anti-rabbit IgG: 5 nm (BBI, 1:10 dilution) for 1 h at RT. This step was followed by fixation in 2% glutaraldehyde and then samples were post-stained with uranyl acetate and lead citrate. Rabbit IgG (BioLegend) was used as a negative control. Sections were examined with a TEM (CM100 Twin, Philips) operated at a 60 kV accelerating voltage at the Core Facility for Integrated Microscopy, Panum Institute, University of Copenhagen. The images were recorded with a side-mounted Olympus Veleta camera (Olympus).
Quantitative real-time RT-PCR
Total RNA was extracted and purified using the miRNeasy kit (Qiagen). Concentration and quality of RNA was assessed using a NanoDrop 2000C spectrophotometer (Thermo Scientific). The RNA samples were subjected to one-step quantitative real-time PCR measurements using the QuantiFast SYBR Green RT-PCR kit (Qiagen) and QuantiTect primer assays (Qiagen) on a Step One Plus real-time thermal cycler (Applied Biosystems). Gene expression was calculated using the delta CT method applying glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the housekeeping reference gene . Each sample was analysed in duplicate. The PCR primers (QuantiTect primer assays) for p33 (Hs_C1QBP_1_SG) and GAPDH (Hs_GAPDH_2_SG) were purchased from Qiagen.
Cells were washed with PBS and lysed in SDS sample buffer [Tris/HCl 62.5 mM, pH 6.8, 2% (w/v) SDS, 10% (v/v) glycerol, 1 mM PMSF]. The samples were sonicated for 2×10 s, boiled and centrifuged (16000 g). Supernatants were collected and total protein content determined by the Bio-Rad DC Protein Assay to assure equal loading in each lane. The proteins were separated by SDS/PAGE using Criterion TGX Any kD precast gels (Bio-Rad) with equal amount of protein loaded in each lane. After separation, the proteins were transferred to nitrocellulose membrane by electroblotting overnight at 4°C. The membrane was blocked in 1% casein (w/v) in TBS (Bio-Rad) and then incubated overnight with antibodies against p33 (clone 60.11, MMS-606R; Nordic BioSite AB, 1:1000 dilution), mouse monoclonal heat-shock protein 90 (Hsp90, BD Transduction Laboratories, 1:2500 dilution) and GAPDH (clone 6C5, Merck Millipore, 1:5000 dilution). The membranes were washed three times in Tris-buffered saline with Tween 20 (TBS-T) and the immunoreactive bands were visualized by chemiluminescence using a horseradish peroxidase (HRP)-conjugated secondary anti-mouse antibody followed by SuperSignal West Femto chemiluminescence reagent (Thermo Scientific). The p33 immunospecific band was analysed by photodensitometric scanning and normalized to Hsp90 or GAPDH serving as internal controls. Normalization of p33 immunoreactive signal to either Hsp90 or GAPDH gave identical results. Images were acquired using a LI-COR Odyssey Fc instrument (LI-COR Biosciences).
Co-immunoprecipitation of p33 and LL-37 was performed using a Pierce Classic IP Kit (Thermo Scientific) according to the manufacturer's instructions. Briefly, HaCaT or HeLa cells were treated with LL-37 (10 μM) for 10 min before cells were washed with PBS and lysed using the wash/lysis buffer. The lysate was divided into two equal parts, one treated with anti-p33 antibody (Innovagen AB, 6 μg) and one kept as a negative control. Lysates were incubated with agitation overnight at 4°C before antibody was retrieved with agarose beads and washed. Proteins were eluted with SDS sample buffer and LL-37 was detected by dot blots on nitrocellulose membranes, using a monoclonal antibody against LL-37 (Cathelicidin antibody OSX12, Novus Bio, 5 μg/ml), a HRP-conjugated secondary anti-mouse antibody (Cell Signaling, 1:10000 dilution) and SuperSignal West Femto chemiluminescence reagent (Thermo Scientific).
In some experiments, sub-cellular fractionation was performed prior to co-immunoprecipitation . Briefly, cells were homogenized with an Ultra-Turrax homogenizer (IKA) and nuclei pelleted by centrifugation at 800 g for 10 min. The supernatant was then subjected to a second centrifugation at 30000 g for 20 min in an Optima™ L-100 XP ultra-centrifuge (Beckman Coulter) to pellet the membranes. All procedures were done at 4°C.
Summarized data are presented as means ± S.E.M. Each culture well represents one biological replicate, i.e., one observation (n=1). Number of n-values is indicated in the figure legends. Statistical significance was calculated using ANOVA and Student's two-tailed t test for unpaired comparisons with Bonferroni correction for post hoc analysis as appropriate. P-values <0.05 were considered significant.
p33 is located at the mitochondria, the plasma membrane and in the cytosol
In the first series of experiments, we aimed to localize endogenous p33 in human osteoblasts and keratinocytes. To this end, osteoblast-like MG63 cells and HaCaT keratinocytes were analysed by immunocytochemistry. As shown in Figure 1, cytoplasmic p33 immunoreactivity was detected in MG63 cells by laser scanning confocal microscopy (Figures 1A and 1B). The nuclear p33 staining was weak compared with the cytoplasmic signal, indicating cytoplasmic rather than nuclear location of p33 (Figures 1A and 1B). HaCaT cells showed similar distribution of p33 immunoreactivity to MG63 cells (Figure 1C). Two further human cell types investigated, human PDL cells and HDFa cells, also displayed a comparable expression pattern for p33 (result not shown). By employing double-staining for p33 immunoreactivity and MitoTracker for mitochondria in HaCaT cells, we found that p33 is localized to the mitochondria (Figures 1C–1E). The sub-cellular distribution of p33 was then further analysed by immunogold labelling EM. p33 immunoreactivity in MG63 cells was found at the mitochondrial periphery, at the plasma membrane and in some parts within the cytosol (Figure 2A). A similar distribution pattern for p33 was recorded in HaCaT cells (Figure 2C). Upon treatment with p33 siRNA, only a weak p33 immunoreactive signal was observed for all sub-cellular compartments in both HaCaT and MG63 cells (Figures 2B and 2D). Furthermore, a similar sub-cellular distribution pattern for p33 as that observed in MG63 and HaCaT cells was also seen in the HDFa and PDL fibroblasts (Figure 3).
p33 is localized to the mitochondria in human HaCaT keratinocytes
Sub-cellular distribution of p33 and down-regulation by p33 siRNA in MG63 and HaCaT cells
Sub-cellular distribution of p33 in HDFa and PDL cells
The extent of LL-37-induced cytotoxicity differs in various cell lines
Next, we tested whether the extent of cytotoxic effects caused by LL-37 varies in different human cell lines. For this purpose, MG63, HaCaT, PDL and HDFa cells were treated with increasing concentrations of LL-37 and cell viability assessed by the MTT and LDH assay (Figures 4A–4D). The results depicted in Figures 4(A) and 4(B) reveal that a 3-h treatment with LL-37 causes a concentration-dependent attenuation of cell viability as assessed by the MTT assay with EC50 values that are in the micromolar range (HaCaT=9 μM, MG63=5 μM, HDFa=4 μM and PDL cells=1.7 μM respectively). These data thus show that HaCaT cells are least sensitive to LL-37-treatment, whereas viability of PDL cells is most affected. The cell type specific LL-37-induced attenuation of viability was not solely dependent on LL-37 ability to perforate the plasma membrane, since there was a prominent LDH release in all four cell types (Figures 4C and 4D), which may point to an intracellular mechanism that is responsible for the LL-37-evoked attenuation of cell viability. We further noted that PDL cells display a higher LL-37-induced LDH release than all other cells investigated. The strong LL-37-evoked decrease in cell viability was also reflected in reduced cell number and morphological changes, such as cell shrinkage and membrane blebbing, which were more pronounced in PDL cells compared with HDFa cells (Figure 5).
Correlation between reduced cell viability for LL-37-treated human cell lines and cellular p33 level
Morphological analysis of LL-37-treated PDL and HDFa cells
LL-37-induced cytotoxicity depends on the cellular expression level of p33
Based on these findings, we hypothesized that the differences in cell viability are due to the endogenous p33 expression levels in the cell lines tested. Therefore, we performed Western blotting for comparative analysis of the respective cellular p33 content in these cells (Figures 4E and 4F). HaCaT cells showed the highest p33 level, whereas lower levels were found in HDFa and MG63 cells. The lowest p33 content was detected in PDL cells. Additionally, we determined p33 mRNA levels in both MG63 and HaCaT cells revealing a 30% higher p33 mRNA level in HaCaT compared with MG63 cells (result not shown). Thus, HaCaT cells show higher expression of p33 than MG63 cells on both transcript and protein level.
Down-regulation of p33 by siRNA increases LL-37-induced cytotoxic effects
In order to demonstrate that p33 expression levels determine the sensitivity of cells to LL-37 treatment, we employed a siRNA approach to down-regulate p33 mRNA and protein. p33 siRNA transfection led to a decrease in p33 mRNA and protein levels both in MG63 and HaCaT cells. As shown in Figures 6(A) and 6(C), treatment with p33 siRNA reduced p33 mRNA and protein level in MG63 cells by 80% and 90% respectively, compared with cells transfected with NC scramble and similar results were obtained when HaCaT cells were used (Figures 6B and 6D).
p33 siRNA reduces p33 mRNA and protein expression levels
To compare the cytotoxic damage in p33 siRNA and NC construct transfected cells, we chose LL-37 concentrations which were below the EC50 values for each cell type (5 μM for MG63 and 9 μM HaCaT cells). Figure 7(A) shows that treatment with 1 and 4 μM LL-37 reduced MG63 cell viability, assessed by the MTT assay, by 30% and 50% respectively, regardless whether p33 was present or knocked down in these cells (Figure 7A). In contrast, treatment of HaCaT cells with 6 and 8 μM LL-37 reduced viability significantly more strongly in p33 siRNA-treated cells compared with cells transfected with a NC construct (Figure 7B). The p33 siRNA increased sensitivity to LL-37 in the HaCaT cells but not in MG63 cells probably because HaCaT cells contain approximately six times higher level of p33 than MG63 cells (Figure 4E) and therefore the absolute decrease in p33 evoked by p33 siRNA is greater in HaCaT compared with MG63 cells. Interestingly, the p33 knockdown in MG63 and HaCaT cells did not further affect the integrity of the plasma membrane, as LL-37-evoked LDH release was not increased in cells lacking p33 (Figures 7C and 7D). Our previous work has shown that treatment with p33 siRNA has no effect on LL-37-stimulated LDH release in MG63 cells . Thus, these data suggest that the intracellular level of p33 regulates cell viability.
Treatment with p33 siRNA enhances sensitivity to LL-37 in HaCaT cells
Overexpression of p33 antagonizes LL-37-evoked cytotoxicity
Having shown that removal of p33 enhances the cytotoxic effects of LL-37, we next wished to analyse whether p33 up-regulation renders cells more resistant to LL-37. To this end, HeLa cells were employed because their transfection efficiency was significantly higher than that of MG63 and HaCaT cells. Notably, HeLa cells show similar p33 expression levels for both mRNA and protein compared with MG63 cells (result not shown). In cells transiently transfected with pcDNA 3.1-p33 vector, p33 mRNA level was increased by approximately 200 times compared with cells transfected only with pcDNA 3.1 vector as control (Figure 8A). Additional Western blot analysis revealed that the p33 protein level was increased approximately 2-fold after transfection (Figure 8B). The increase in p33 rendered the cells completely resistant against an LL-37 attack, as depicted in Figure 8(C). The data show that cell viability in HeLa cells overexpressing p33 was not affected by 4 μM LL-37, whereas the same concentration of LL-37 (4 μM) reduced cell viability by approximately 30% in control cells (Figure 8C). As before, this effect seems to be only evoked by the intracellular p33, since LL-37 (4 μM) triggered a similar LDH release in p33 overexpressing and control cells (Figure 8D). Thus, overexpression of p33 completely abolished the LL-37-induced attenuation of cell viability but showed no effect on LDH release.
LL-37 has no effect on cell viability in HeLa cells overexpressing p33
p33 interacts with LL-37
As our results imply that the toxic effects of intracellular LL-37 are neutralized by p33, we next wished to demonstrate that the two proteins bind to each other upon internalization of LL-37. These experiments were performed with HaCaT cells since they show high expression of endogenous p33 (Figure 4E). The cells were first incubated with LL-37 (10 μM) for 10 min and, after a lysis step, the interaction between p33 and LL-37 was investigated by co-immunoprecipitation. Immunoblot (dot blot) of the precipitates demonstrates interaction of the two proteins (Figure 9A) suggesting that LL-37 is targeted by p33 and this leads to a down-regulation of the deleterious effects of LL-37. In order to provide further evidence that intracellular p33 binds LL-37, we treated HaCaT and HeLa cells with LL-37 (10 μM) for 10 min, then performed sub-cellular fractionation and determined the interaction between intracellular p33 and LL-37 by co-immunoprecipitation in cytosolic fractions of HaCaT and HeLa cells. Dot blot showed interaction of p33 with LL-37 in a cytosolic fraction of HaCaT and HeLa cells (Figures 9A and 9B). Moreover, we demonstrate stronger interaction between p33 and LL-37 in a cytosolic fraction of HeLa cells transfected with the pcDNA 3.1-p33 vector compared with cells transfected with the control vector (Figure 9B). The interaction between p33 and LL-37 was approximately two times higher in p33 overexpressing HeLa cells compared with control cells assessed by densitometry (205±10% in p33 overexpressing cells compared with 100±6% in control cells, P<0.001, n=3 in both groups). These data support our working hypothesis that intracellular p33 is responsible for down-regulating the deleterious activity of LL-37.
Co-immunoprecipitation shows interaction between p33 and LL-37
LL-37 up-regulates p33 expression in MG63 cells
In the last series of experiments, we were wondering whether p33 up-regulation constitutes a host defence mechanism that makes cells less susceptible to an LL-37 attack under pathological conditions. Figure 10 shows that treatment with 4 μM LL-37 for 48 h increased MG63 p33 protein expression by approximately 50%, whereas LL-37 (8 μM) had no effect on HaCaT p33 protein expression. The LL-37-induced up-regulation of MG63 p33 protein expression was confirmed on the mRNA level (result not shown). Treatment with a higher concentration of LL-37 (8 μM) increased MG63 p33 protein level by 55% (100±11% in control cells compared with 155±13% in LL-37-treated cells, P<0.01, n=5 for both groups). In contrast, incubation with the well-known pro-apoptotic compound staurosporine (0.5, 1 and 1.5 μM) for 48 h had no effect on MG63 p33 protein expression, suggesting that LL-37-evoked up-regulation of p33 is specific for LL-37 and not secondary to the pro-apoptotic process (result not shown).
LL-37 triggers p33 protein expression in MG63 cells
Modulation of inflammatory reactions in infectious diseases is tightly regulated and their disturbance often decides the fate of disease progression . Notably, pattern recognition receptors are key players in these processes since they are among the first sentinels that sense the invading pathogen . Apart from microorganism-derived ligands, pattern recognition receptors can be activated by danger signals. These are often host factors mobilized under disease conditions that are up-regulated or released upon an inflammatory signal .
In 2010, Into et al.  reported for the first time that LL-37 is a danger-associated molecular pattern (DAMP) peptide, as it modulates the activity of TLRs and triggers the mobilization of inflammatory and chemotactic mediators, such as interleukin-6 (IL-6), IL-8 and C-X-C motif chemokine 10 (CXCL10) respectively. Interestingly, due to its amphiphilic nature caused by the hydrophobic and hydrophilic residues of LL-37, the peptide is able to penetrate eukaryotic plasma membranes and subsequently enter the cytosol. Intracellularly, LL-37 not only exerts its anti-microbial activity by killing invading microorganisms , but it can also block inflammasome formation by binding to cytosolic DNA . However, not much is known of how the activity of cytosolic LL-37 is regulated.
The data demonstrated in the present study suggest a critical role for internalized LL-37, as the intracellular pool of the peptide is also crucial for impairing cell viability. We further find that its cytotoxic activity is regulated by p33, a protein also known as gC1qR or p32. Importantly, p33 and LL-37 show direct intracellular interaction demonstrated by co-immunoprecipitation. p33 was originally described as a receptor for the globular part of complement factor C1q , but subsequent analysis revealed that p33 possesses a canonical N-terminal specific localization sequence that targets the protein to the mitochondria . However, apart from the cell membrane and mitochondria, also other cellular locations such as the nucleus have been described for p33 . A multi-sub-cellular distribution of p33 was also observed in our experiments. Though a big p33 pool was localized to the mitochondria, the protein was also found at the plasma membrane and within the cytosol. The relative importance of mitochondrial, plasma membrane and cytosolic p33 for the binding of LL-37 may indeed vary between different cell types. We cannot completely rule out additional protective mechanisms, besides p33, explaining the cell type-dependent differences in LL-37 sensitivity. The high expression level of p33 in HaCaT cells compared with other human cell types observed in the present study may reflect that skin keratinocytes represent an outer barrier exposed to high LL-37 concentrations.
It is worth mentioning that recent work has pointed to a role of p33 in regulating cellular fitness. For instance, Xiao et al.  published a study in 2014 showing that mitochondrial p33 is pro-apoptotic and regulates mitochondrial calcium up-take, whereas Watthanasurorot et al.  employ an arthropod model to demonstrate that binding of cytosolic p33 to calreticulin protects cells from dying. Whereas the molecular mechanisms are not completely understood, it appears that shuttling of p33 between the cytosol and the mitochondria or the nucleus is essential in these processes [42,43]. Dual or multi-fold intracellular targeting of proteins with specific localization sequences is an emerging field of research [44,45] and several mechanisms have been proposed to explain this phenomenon. These include, for instance, different or several translation products, multiple mRNAs or membrane permeabilization. However, so far the subcellular distribution of p33 remains unexplained. Interestingly, in our experiments cytosolic overexpression revealed that p33 counteracts LL-37 induced cell death, but not cell perforation. This suggests that intracellular p33, either the mitochondrial or the cytosolic fraction or both, binds and inactivates LL-37.
Considering the capability of LL-37 to modulate Toll-like receptor activity and evoke inflammatory reactions, intracellular damage caused by the peptide would be counterproductive and impair the cellular response. Therefore, cells have to rely on intracellular defence mechanisms that prevent cellular damage evoked by cytosolic LL-37. Bearing in mind that mitochondria are evolutionarily of prokaryotic origin, it seems plausible that localization of p33 to this organelle prevents further damage by internalized HDPs with anti-microbial activities. Our data therefore point to an important role of p33 in preventing intracellular cytotoxic effects of LL-37 or other HDPs, without influencing the anti-microbial and inflammatory activities of these peptides.
Daniel Svensson designed, planned and preformed the experiments with the exception of immunogold EM and preparation of the p33 plasmid and wrote the paper. Laura Wilk prepared the plasmid DNA, preformed experiments on p33 overexpression and wrote the paper. Matthias Mörgelin preformed immunogold EM. Heiko Herwald designed the study and wrote the paper. Bengt-Olof Nilsson designed the study and wrote the paper.
The authors wish to thank Dr Daniel Nebel for providing the human PDL, Dr Uta Rabenhorst for providing the GFP-vector and Dr Ole Sørensen for advice on transfection protocols.
This work was supported by the Swedish Research Council [grant numbers 2003-7605 (to B.-O.N.) and SB12–0019 (to H.H.)]; the Swedish Dental Society (to B.-O.N.); the Southern Region within the Swedish Dental Association (to B.-O.N.); the Greta and Johan Kock Foundation (to B.-O.N.); the Swedish Foundation for Strategic Research [grant number K2014–56X-13413–15–3 (to H.H.)]; the Ragnar Söderberg Foundation (to H.H.); the Knut and Alice Wallenberg Foundation [grant number KAW 2011.0037 (to H.H.)]; the Alfred Österlund Foundation (to H.H.); and the Medical Faculty, Lund University.
C-X-C motif chemokine 10
Dulbecco's modified Eagle's medium
globular C1q receptor
human cationic antimicrobial protein-18
neonatal dermal fibroblasts
heat-shock protein 90
periodontal ligament fibroblasts
Tris-buffered saline with Tween 20