Targeting macrophages and their recruitment in the oral cavity using swellable (+) alpha tocopheryl phosphate nanostructures

The phosphorylation of (+) alpha tocopherol produces adhesive nanostructures that interact with oral biofilms to restrict their growth. The aim of this work was to understand if these adhesive (+) alpha tocopheryl phosphate ( α -TP) nanostructures could also control macrophage responses to the presence of oral bacteria. The (+) α -TP planar bilayer fragments (175 nm ± 21 nm) formed in a Trizma®/ethanol vehicle swelled when exposed to the cell lines (maximum stabilized size = 29 μ m). The swelled (+) α -TP aggregates showed selective toxicity towards THP-1 macrophages (LD 50 = 304 μ M) compared to human gingival fibroblasts (HGF-1 cells; LD 50 N 5 mM), and they inhibited heat killed bacteria stimulated MCP-1 production in both macrophages (control 57.3 ± 18.1 pg/mL vs (+) α -TP 6.5 ± 3.2 pg/mL) and HGF-1 cells (control 673.5 ± 133 pg/mL vs (+) α -TP - 463.9 ± 68.9 pg/mL).

Targeting macrophages and their recruitment in the oral cavity using swellable (+) alpha tocopheryl phosphate nanostructures Macrophages perform a number of important regulatory functions in the human body, 1 but in several pathologies including, malignant tumors, 2 inflammatory disease, 3 metabolic disease, 4 infections, 5 and periodontitis, 6 their dysfunction is thought to contribute to disease progression. 7 Anti-cytokine therapies can act to counteract macrophage dysfunction, but their 'off-target' side-effects render the currently available agents inappropriate for this indication. 8 Macrophages are highly mobile and thus they are difficult to specifically target through traditional drug delivery approaches. 9 However, their ability to actively recognize and phagocytose foreign material provides a potential route to specifically deliver agents into macrophages using bespoke drug carrier systems.
One class of lipids that have the potential to be formed into materials that could target macrophages is the tocopherol lipids. 10 In recent work a novel tocopherol analogue, (+) alpha tocopheryl phosphate (α-TP) was synthesized and was shown to form oral bioretentive nanomaterials that disrupt biofilm growth. 11 In the mouth the ionic phosphate moiety of α-TP interacts with simple electrolytes and this gives it the potential to swell and change shape, 12 which could facilitate macrophage phagocytosis and release of the active from the nanomaterial structure. However, the ability of α-TP nanomaterials to selectively target macrophage responses in the mouth has yet to be tested.
In the mouth, macrophages and their cytokine products play an important role, along with enzymes, in both periodontal soft tissue and jawbone destruction. 13 As a consequence, macrophages and the cytokines that stimulate macrophage recruitment (e.g., MCP-1 14 ) have become targets in the search for new agents to improve oral health. 15 Therefore, the aim of this work was to understand if (+) α-TP nanomaterials could be used to target Nanomedicine: Nanotechnology, Biology, and Medicine 21 (2019) 102010 nanomedjournal.com Funding information: The study was financed by an Engineering and Physical Sciences Research Council (EPSRC) CASE award with Johnson and Johnson. EPSRC had no role in study design, data collection and interpretation, or the decision to submit the work for publication. Johnson and Johnson had a role in some experimental design and data interpretation. macrophage dysfunction in the mouth. It was predicted that the surface charge and size of the aggregates would influence their uptake into oral tissue; therefore, three types of tocopherol aggregates were used in this study: (+) α-T, which was predicted to display a neutral surface, and two structural isomers of α-TP, the (+) isomer and (±) isomer , both of which were predicted to be negatively charged in the mouth.

Methods
(+) α-TP synthesis (+) α-TP synthesis used phosphorus oxychloride with triethylamine in anhydrous THF for 3 h at room temperature as previously described. 11 The pure product was obtained after C18 column chromatography. (+) α-TP was considered sterile due to the presence of 70% isopropanol in the purification step.
Characterization of (+) α-T, (+) α-TP, (±) α-TP aggregates in cell culture medium The size of the aggregates was monitored in cell culture media to mimic the gum tissue environment. The volume median diameters of (+) α-T, (±) α-TP and (+) α-TP were measured using a laser diffraction technique (Mastersizer X, version 2.15, Malvern Instruments, UK). The vitamin E derivatives (3 mM) dispersed initially in 20% ethanol, 80% water vehicles with 150 mM Trizma®. Samples were diluted in cell culture medium (DMEM or RPMI without phenol red, FBS, or pen strep) to reach an obscuration ca 20%. The volume median diameters were recorded every 3 min for 15 min (N = 3).

Preparation of heat killed bacteria
Oral heat killed bacteria were used to model the inflammatory environment found in vivo. 19,20 Human saliva (1 donor) was collected, formed into bacterial pellets, re-suspended in cell culture media (1 mL) (DMEM or RPMI-1640) and heat killed using a heating block (Grant, QBA1 series), UK) at 80°C for 10 min. The heat killed bacteria were diluted to 0.18 OD 620 with cell culture medium under sterile conditions followed by a 1/100 dilution of the solution in cell culture medium to form the inflammatory stimuli.

Determination of HGF-1 and THP-1 cell line viability
The effects of (+)/(±) α-TP, (+) α-T and heat killed human saliva bacteria, individually and in combination, were tested on HGF-1 and THP-1 cells. CHX and CPC were also tested on THP-1 macrophages as controls. To perform these assessments HGF-1 cells (passages 3-8) were seeded (1 × 10 4 cells/well) 21 in 96 well microplates (100 μL/well) and were incubated for 24 h at 37°C in a 5% CO 2 atmosphere to allow for cell adhesion. The culture medium was then aspirated, and the cells were treated for 4 h with (+) α-TP (0.05, 0.5, 5, 50, 500, or 5000 μM), (+) α-T (500 or 5000 μM) or (±) α-TP (500 μM) (100 μL/well) after which the test samples were aspirated and the cells washed with HBSS (200 μL/well). Either complete media or heat killed human bacteria from human saliva in complete media (1/100 dilution) were applied (100 μL/well) and the samples were incubated for an additional 15 h at 37°C in a 5% CO 2 atmosphere. The THP-1 monocytes were seeded (1 × 10 4 cells/well) with PMA (5 ng/mL) to allow monocyte differentiation and subsequent cell adhesion in 96 well microplates. They were incubated for 48 h at 37°C in a 5% CO 2 atmosphere. 22 The culture medium was aspirated; cells were washed with HBSS and cultured in serum free media for 3 h. The cells were then treated for 2 h (the time was reduced from 4 h to reduce the time the cells were starved from FBS) 23 with (+) α-TP (0.05, 0.5, 5, 50, 500, or 5000 μM), (±) α-TP (500 μM), (+) α-T (500 or 5000 μM), CHX (50, 100, 150, 500, or 5000 μM) or CPC (0.05, 0.5, 5, 50, or 500 μM) (100 μL/well) after which the test samples were aspirated, and the cells washed with HBSS (200 μL/well). Either complete media or heat killed human bacteria from human saliva in complete media (1/100 dilution) were applied (100 μL/well) and incubated for an additional 15 h at 37°C in a 5% CO 2 atmosphere. The 15 h supernatants were removed and used for ELISA assays, fresh complete media (100 μL) were added to the wells and then the colorimetric MTS tetrazolium compound (20 μL) was added. Plates were then incubated at 37°C in a 5% CO 2 atmosphere for 4 h and then absorbances at 490 nm (iEMS Incubator/Shaker, Thermo Scientific, UK) were measured with reference subtractions at 650 nm. Untreated control cells were assigned a value of 100% viability (negative control); cells treated with 1% triton X (dispersed in cell culture media) were assigned a value of 0% viability (positive control). All the other conditions were compared to the controls using Eq.
(1) were ABS was the corrected absorbance's. Results were expressed as means ± standard deviations of triplicate assays from three different experiments. Lethal dose 50% (LD 50 ) values were calculated using the dose response model in Origin 2016 (Silverdale scientific ltd, UK).
Determination of cytokine secretion MCP-1, IL-6 and IL-8 expression was measured in the selected cell line supernatants in response to heat killed bacteria and this was repeated in the presence of the tocopherols in order to assess their anti-cytokine effects. 24 ELISA kits were used to quantify the protein concentrations of MCP-1, IL-8 and IL-6 produced in the microplate cell supernatants according to the manufacturer's protocols. Preliminary experiments showed that there was no detectable MCP-1 in the heat killed bacteria applied to the cells and hence the detected MCP-1 was solely generated from the HGF-1 cells. The test solutions at a concentration of 500 μM were not toxic against HGF-1 cells and hence this concentration was selected to assess their effect on MCP-1 release.
mRNA expression assay mRNA transcript expression assays were performed using q-PCR on HGF-1 cells to understand how MCP-1, IL-8 and IL-6 expression was being regulated in the presence and absence of the three different tocopherol aggregates. The HGF-1 cells were cultured and treated using the same method as the cell viability assay with the exception that after the 15 h incubation with the inflammatory stimuli supernatants were removed, and the cells washed with PBS (100 μL/well) and then harvested with TRI-Reagent® (100 μL/well). Cell treatments groups were in quintuplet and were combined in microcentrifuge tubes (500 μL, 50,000 cells) (N = 3). The total RNA was extracted using MagMax™-96 for Microarrays Kit. RNA was quantified using the NanoDrop (Thermo Scientific, UK). RNA integrity was analyzed using the Bioanalyzer (Agilent, UK). A 50 ng aliquot of RNA per sample was reverse transcribed to cDNA using High-Capacity RNA-to-cDNA™ Kit. Expressions of MCP-1, IL-6 and IL-8 were analyzed using probes from the Universal Probe Library (UPL, Roche). Actin beta (ACTB), Selenocysteine lyase (SCLY) and tRNA-yW synthesizing protein 1 homolog (TYW1) were used as reference genes (for primer sequence and probe selection, see supplementary material, Tables S1 and S2). Assays were designed following instructions from the Universal Probe Library Assay Design Centre. 25 Quantitative PCR (qPCR) was performed using TaqMan Universal PCR Master Mix, following manufacturer's protocol. Each 10 μL reaction contained 0.2 μM forward primer and reverse primer and 0.1 μM UPL probe. cDNA was diluted 10-fold, and 4 μL of diluted cDNA was used per reaction. qPCR was performed on Applied Biosystems 7900HT Real-Time PCR System under the following cycling conditions: 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Data were collected at the end of each cycle.

Data analysis
All data were expressed as their mean ± standard deviation (SD) at N = 3. The N numbers were independent experiments; each N number had three data points (three wells). Statistical analysis of the data was performed using Levine's homogeneity test before statistical significance between the sample groups was assessed by one way analysis of variance (ANOVA) tests with post-hoc Tukey analysis in Origin 2016 (Silverdale scientific ltd, UK). Statistically significant differences were identified when P ≤ 0.05.

Aggregate characterization in cell culture media
Dynamic light scattering size analysis, zeta potential, atomic force microscopy (AFM) imaging and fluorescence spectroscopy measurements of the aggregates in 20% ethanol, 80% water, 150 mM Trizma®, pH 7.4 vehicles are reported in the supplementary material (See supplementary material, Figures  S1-S4). 11,12 The characterization data showed (+) α-T produced spherical liposomes (563 ± 1 nm, −10.5 ± 0.2 mV), (+) α-TP produced planar bilayer fragments (175 nm ± 21 nm, −14.9 ± 3.5 mV) and (±) α-TP produced spherical liposomes (104 nm ± 1.3 nm, −38.7 ± 7.0 mV) in an 20% ethanol 80% Trizma® buffer vehicle. The (+) α-TP nanomaterials size was shown to increase into the micron range when aliquoted into both the DMEM and RPMI cell culture media, with no statistically significant difference in the aggregate sizes between the two media (P N 0.05) (24.0 ± 4.9 μm and 28.52 ± 6.46 μm at 15 min). The (+) α-T aggregates swelled in both DMEM and RPMI over a period of 15 min to sizes of 3.7 ± 0.7 μm and 3.8 ± 0.6 μm respectively, but unlike (+) α-TP this size did not change beyond the 15 min time point. The (±) α-TP isomer immediately swelled to a size of 35 ± 10 μm in DMEM and 16 ± 9 μm in RPMI and then the size started to reduce upon further incubation (Figure 1, B). It was not possible to gain clear images of the swelled aggregates in the cell culture media.

HGF-1 cellular response to the tocopherol aggregates
(+) α-TP was well tolerated by HGF-1 cells until a concentration of 5 mM (0.25% w/v) at which point the cell viability significantly dropped to 75.5% ± 7.9% ( Figure 2). Both the (±) and the (+) α-TP isomers (500 μM) were found to have the same effects on cell viability (P N 0.05), which suggested that the stereochemistry was not a factor in the cell toxicity. At 5 mM (+) α-T did not reduce the cell viability (96.7% ± 6.1%), which showed that the addition of the phosphate group significantly increased the agents' cell toxicity.
The pre-treatment of the HGF-1 cells with (+) α-TP before inflammatory stimulation did not significantly reduce the mRNA transcription of MCP-1, IL-6 or IL-8 compared to the heat killed bacteria treated cells (P N 0.05, see supplementary material, Figure S6).

Discussion
The tocopherol nanomaterials tested in this work swelled when spiked into cell culture media. The different swelling profiles observed for the nanomaterials in the DMEM and RPMI fluids were thought to be a consequence of the media composition. The iron, which was only present in the DMEM, was identified as the most likely component to explain the differential behavior, as it is known to interact with phosphorylated compounds. 26 In both media the swelling of the nanomaterials was attributed to a transition of the lipid aggregates to multi-lamella liposomes due to the change of ionic strength in the dispersion media. 11 An alternative explanation to the observed size changes was that the nanomaterials were aggregating. However, this was thought to be unlikely as the dispersions changed size with a high degree of reproducibility, which is not common in aggregated systems. 27 Unlike macrophages, non-phagocytic cells are unlikely to internalize the micron sized carriers; hence, the changes in nanomaterial characteristics in physiological fluids were thought to be beneficial for macrophage targeting. This hypothesis was supported by the cell cytotoxicity and cytokine suppression data discussed further below. The (+) α-TP reduced MCP-1 production by HGF-1 cells exposed to heat killed bacteria by approximately 32%. In the literature there does not seem to be a report that has previously shown that α-TP can suppress MCP-1 release. The antiinflammatory agent, Bindarit, which has some structural similarities to α-TP, has previously been shown to selectively inhibit MCP-1 through the inhibition of MCP-1 mRNA, but in this work (+) α-TP was found not to inhibit mRNA transcript expression. From the lack of MCP-1 mRNA inhibition it could be deduced that (+) α-TP influenced post mRNA transcription activity such as protein synthesis or cellular protein secretion inhibition. However, because the cell signaling pathways for the heat treated saliva induced cytokine release from HGF-1 cells have not been well established the (+) alpha-TP mechanism of action was not investigated further 28 It was surprising that all the tocopherols used in this work did not inhibit cytokine release from the cell lines because (+) α-T has previously been shown to have anti-inflammatory effects. [29][30][31] In addition, (+) α-T has been observed to inhibit IL-6 from HGF-1 cells in the literature. 32 However, in previous work LPS from P. gingivalis was used to induce the cytokine and in this study heat killed bacteria from human saliva were used. It is possible that the tocopherols inhibit cytokine production on the LPS stimulation pathway and not other salivary inflammatory stimuli pathways.
The selective α-TP toxicity and subsequent reduction in MCP-1 secretion from THP-1 macrophages were thought to be a function of the nanomaterial swelling. The most selective of the aggregates in terms of α-TP toxicity and MCP-1 suppression was (+) α-TP followed by (±) α-TP then (+) α-T, which was not toxic and did not suppress MCP-1. This rank order aligned to the sizes of the aggregates when presented to the cells. This could, at least in part, be a consequence of greater internalization of the larger aggregates due to more material volume being phagocytized by individual macrophages. 33 Selective macrophage toxicity of α-TP was thought to be desirable for the treatment of the chronic inflammatory phase of periodontal disease as it could control the macrophage burden and shorten the immune response. 34 However, it is not desirable to kill all macrophages as they perform a beneficial role of engulfing microbes and preventing the spread of systemic infection. In this aspect it was promising that even though (+) α-TP showed selective toxicity towards the macrophages it was still found to be less toxic (lower LD 50 values) than two commercially used antimicrobial agents CHX and CPC; suggesting that it was capable of regulating macrophage accumulation without complete depletion.

Conclusion
Mild and moderate periodontitis affects the majority of the adult global population with 10.5%-12% of the population affected by severe periodontitis, making it the sixth most prevalent condition in 2010. 35 However, at present, subantimicrobial dose doxycycline is the only agent that has been approved for human use that inhibits the gingival inflammatory process, the cause of the destructive elements of this disease. 36 (+) α-TP has previously been shown to be tooth adherent, substantive, and capable of reducing the oral microbial burden. 11,12 In this study the swelling of (+) α-TP nano-sized aggregates was shown to have an anti-inflammatory effect on macrophages by selective macrophage toxicity thereby reducing MCP-1 generation. This effect would then dampen excessive macrophage burden and therefore reduce gingival destruction making this compound an attractive prospect for development as a multifunctional agent to improve oral health.