The Peptide/Antibody-Based Surface Decoration of Calcium Phosphate Nanoparticles Carrying siRNA Influences the p65 NF-κB Protein Expression in Inflamed Cells In Vitro

Earlier studies with nanoparticles carrying siRNA were restricted to investigating the inhibition of target-specific protein expression, while almost ignoring effects related to the nanoparticle composition. Here, we demonstrate how the design and surface decoration of nanoparticles impact the p65 nuclear factor-kappa B (NF-κB) protein expression in inflamed leucocytes and endothelial cells in vitro. We prepared silica-coated calcium phosphate nanoparticles carrying encapsulated siRNA against p65 NF-κB and surface-decorated with peptides or antibodies. We show that RGD-decorated nanoparticles are efficient in down-regulating p65 NF-κB protein expression in endothelial cells as a result of an enhanced specific cellular binding and subsequent uptake of nanoparticles. In contrast, nanoparticles decorated with IgG (whether specific or not for CD69) are efficient in down-regulating p65 NF-κB protein expression in T-cells, but not in B-cells. Thus, an optimized nanoparticle decoration with xenogenic IgG may stimulate a specific cellular uptake. In summary, the composition of siRNA-loaded calcium phosphate nanoparticles can either weaken or stimulate p65 NF-κB protein expression in targeted inflamed leucocytes and endothelial cells. In general, unveiling such interactions may be very useful for the future design of anti-p65 siRNA-based nanomedicines for treatment of inflammation-associated diseases.


Introduction
It is generally known that the immune system is a complex network of diverse cell types, signaling pathways, and effector molecules, which are all necessary to provide a defense against foreign pathogens. Important cellular players are endothelial cells, phagocytic cells, such as monocytes, macrophages, dendritic cells, etc., which are able to recognize and respond to a multitude of antigens (innate immunity), as well as B-and T-lymphocytes, plasma cells, etc., which are part of adaptive immunity [1].
In particular, the nuclear factor-kappa B (NF-κB) is known to play a crucial role during immune and inflammatory responses, cell growth, survival and development [2]. Herein, NF-κB particularly regulates the expression of pro-inflammatory cytokines, matrixdegrading enzymes (matrix metalloproteinases; MMPs), adhesion molecules, and further mediators, which all determine the initiation and perpetuation of chronic inflammation [3]. For such reasons, a specific NF-κB blockade is considered to be of particular importance for therapeutic interventions in inflammatory diseases [4,5]. NF-κB plays a critical role ticle analysis by atomic absorption spectroscopy (AAS), the particles were dissolved in a 3:1 v:v mixture of H 2 O:HCl (37%; VWR, USA). For dynamic light scattering (DLS) and ζ-potential analyses, aqueous dispersions of the nanoparticles were used. Scanning electron microscopy (SEM) imaging was done on dried nanoparticles. UV/Vis spectrophotometric analyses were performed with supernatants obtained after nanoparticle centrifugation and ultracentrifugation. Functional siRNA against p65 NF-κB was obtained as functional and control (scrambled) silencing ribonucleic acid from Santa Cruz Biotechnology (USA). Functional siRNA (sc29411; M w 13.8 kDa; denoted as siRNAf in the following) was a mixture of 4 target-specific siRNA duplexes with the following sequences (from 5 →3 ): CCAUGGAGUUCCAGUACUUtt, UCAGCACCAUCAACUUUGAtt, CGAAGUGCGUA-CACAUUCUtt, GGAUUCCUGUACACCUUGAtt. A non-targeting 20-25 nt siRNA designed as negative control with proprietary base sequence (sc 37007; denoted as siRNAs in the following) was also obtained from Santa Cruz Biotechnology. For nanoparticle surface decoration, either cRGDfK-peptide [cyclo(-Arg-Gly-Asp-D-Phe-Lys)] (M w 603 Da; BACHEM AG, Bubendorf, Switzerland), IgG monoclonal anti-CD69, its k isotype control antibodies (both from Armenian hamster, BioLegend®, San Diego, CA, USA; M w 150 kDa), or a murine IgG-k-IC (Bioscience, San Diego, CA, USA; M w 150 kDa) were used.

Synthesis of Ligand-Decorated Calcium Phosphate Nanoparticles for Gene Silencing
The synthesis of bioactive silica-coated calcium phosphate nanoparticles was performed by wet-chemical precipitation as described in detail in Refs. [12,20]. Briefly, polyethyleneimine (PEI)-coated calcium phosphate nanoparticles CaP/PEI-Cy5 were prepared from aqueous solutions. These nanoparticles were fluorescent as we used Cy5-functionalized polyethyleneimine (PEI-Cy5). Next, these cationic particles were loaded with the corresponding siRNA (negatively charged) that adsorbed on the particle surface. These CaP/PEI-Cy5/siRNA nanoparticles were subsequently coated with a silica layer by a modified Stöber synthesis with TEOS to protect the siRNA from enzymatic degradation. The resulting CaP/PEI-Cy5/siRNA/SiO 2 nanoparticles were covalently functionalized by silanization with thiol groups (-SH) to enable surface decoration of the nanoparticles with ligands, i.e., peptides or antibodies ( Figure 1). In brief, 1 mL of CaP/PEI-Cy5/siRNA/SiO 2 or CaP/PEI-Cy5/SiO 2 nanoparticle dispersion was added to a stirred mixture of 4 mL absolute ethanol and 5 µL MPS (conjugation reagent) and further stirred overnight at room temperature (RT) in darkness. After this time, the nanoparticles were collected by centrifugation (1537× g, 30 min, RT), and the nanoparticle pellet was re-dispersed in 0.5 mL water, followed by vortexing and ultrasonication (cycle 0.8, amplitude 70%, 4 s). Prior to coupling with the nanoparticles, ligands (peptides/antibodies) were first activated in a reaction with the heterobifunctional crosslinker sulfo-SMCC, which contains an N-hydroxysuccinimide (NHS) ester and a maleimide functional group. This enables a covalent conjugation between amine-containing ligands and thiol-functionalized nanoparticles. In total, 0.25 mL ligand solution (0.5 mg/mL) was mixed with 0.125 mL sulfo-SMCC (4 mM) and left for activation for 4 h (RT, no stirring). Next, the activated ligand was purified by spin filtration (Amicon ® Ultra 0.5 mL; regenerated cellulose 3000 NMWL; Merck Millipore Ltd., Dublin, Ireland) to remove free sulfo-SMCC. The spin filter was first activated (14,064× g) with water. Then, the activated complex was spin-filtered by centrifugation (14,064× g), washed with 0.4 mL water and centrifuged to remove residual sulfo-SMCC. Finally, the spin filter was turned upside down, placed in a new tube, and the activated ligand was detached by centrifugation (983× g, 2 min, 4 • C). Before the activated ligand was reacted with the thiol-functionalized nanoparticles, it was analyzed by UV spectroscopy (NanoDrop) at λ max = 205 nm (E0.1%) and 280 nm (E1%) to determine the ligand concentration. In total, 0.5 mL of thiol-functionalized nanoparticles was mixed with 0.25 mL activated ligand solution and incubated for 24 h at 4 • C in darkness. The surface-functionalized nanoparticles were collected by centrifugation (21,041× g, 30 min, 4 • C), followed by re-dispersion in 0.5 mL water, followed by vortexing and gentle ultrasonication. The residual supernatant ultracentrifugation was analyzed for the presence of the free (unbound) ligand by UV spectroscopy (NanoDrop) to determine the concentration of nanoparticle-conjugated ligands. The supernatants were analyzed for the presence of free siRNA to determine the siRNA concentration in the nanoparticles during this multi-step synthesis. Finally, the nanoparticles were aliquoted, freeze-dried, and stored at −80 • C until application. and processing for their further chemical modification (silica-functionalization, thi-158 ol-functionalization, and ligand-decoration steps). Neither the silica shell nor a consistent 159 pH adjustment to 10 prevented the calcium loss, which was significant after purification 160 of silica-coated nanoparticles (up to 71% of calcium was lost during these procedures). 161 Moreover, working at higher pH of 10 did not prevent the calcium loss. A thorough 162 step-by-step investigation of this effect revealed the nature of the calcium loss, which 163 varied between 25%-60% per synthesis step (Supplementary Figure S1). After several 164 modifications of the synthetic steps, we excluded all purification steps by washing and 165 increased the concentration by a factor of 2 compared to the earlier synthesis [12]. This 166 increased the content of calcium in the final nanoparticle sample by a factor of 2 to 3. 167 Note that all synthetic steps were performed in small scale due to the low amount of 168 costly siRNA. The nanoparticle characterization was performed by DLS and zeta potential (ζ) de-179 termination in order to assess the nanoparticle size and colloidal stability (Zetasizer Nano 180 ZS; Malvern Panalytical, Germany; laser wavelength λ = 633 nm; Smoluchowski ap-181 proximation; refraction index of hydroxyapatite n = 1.65, absorption 0.01). SEM imaging 182 was performed with an ESEM Quanta 400 FEG microscope (FEI, USA) on gold/palladium 183 (80:20)-sputtered samples at an accelerating voltage of 30 kV. Calcium (Ca 2+ ) was deter-184 mined by AAS with an iCE 3000 M-Series spectrometer (Thermo Scientific, USA). The 185 efficiencies of nanoparticle loading with PEI-Cy5 and siRNA and the nanoparticle deco-186 ration with ligands were determined by UV/Vis spectrophotometry with a DS-11 FX+ 187 device (DeNovix ® , USA). The endotoxin concentration in the nanoparticles was deter-188 mined with an Endosafe ® Nexgen-PTS™ spectrophotometer (Charles River, USA), based 189 on the limulus amoebocyte lysate (LAL) chromogenic assay. For a 20 g mouse, 0.1 EU 190 (endotoxin units) is the maximum endotoxin level considered as safe [21], and we took 191 care that our particles were always below this threshold. Reglo peristaltic pumps (Is-192 matec, Germany) were used for dosing the reagent solutions during synthesis. Centrif-193 ugation of the nanoparticles was carried out with a Rotofix 32A centrifuge (Andreas 194 Hettich GmbH, Germany) and a Heraeus Fresco 21 ultracentrifuge (Thermo Scientific, 195 A thorough analysis of the calcium phosphate nanoparticles indicated a loss of calcium during the washing steps (purification) and their re-dispersion during synthesis and processing for their further chemical modification (silica-functionalization, thiolfunctionalization, and ligand-decoration steps). Neither the silica shell nor a consistent pH adjustment to 10 prevented the calcium loss, which was significant after purification of silica-coated nanoparticles (up to 71% of calcium was lost during these procedures). Moreover, working at higher pH of 10 did not prevent the calcium loss. A thorough stepby-step investigation of this effect revealed the nature of the calcium loss, which varied between 25-60% per synthesis step (Supplementary Figure S1). After several modifications of the synthetic steps, we excluded all purification steps by washing and increased the concentration by a factor of 2 compared to the earlier synthesis [12]. This increased the content of calcium in the final nanoparticle sample by a factor of 2 to 3. Note that all synthetic steps were performed in small scale due to the low amount of costly siRNA.

Nanoparticle Characterization
The nanoparticle characterization was performed by DLS and zeta potential (ζ) determination in order to assess the nanoparticle size and colloidal stability (Zetasizer Nano ZS; Malvern Panalytical, Germany; laser wavelength λ = 633 nm; Smoluchowski approximation; refraction index of hydroxyapatite n = 1.65, absorption 0.01). SEM imaging was performed with an ESEM Quanta 400 FEG microscope (FEI, Hillsboro, OR, USA) on gold/palladium (80:20)-sputtered samples at an accelerating voltage of 30 kV. Calcium (Ca 2+ ) was determined by AAS with an iCE 3000 M-Series spectrometer (Thermo Scientific, Waltham, MA, USA). The efficiencies of nanoparticle loading with PEI-Cy5 and siRNA and the nanoparticle decoration with ligands were determined by UV/Vis spectrophotometry with a DS-11 FX+ device (DeNovix ® , Wilmington, DE, USA). The endotoxin concentration in the nanoparticles was determined with an Endosafe ® Nexgen-PTS™ spectrophotometer (Charles River, Wilmington, MA, USA), based on the limulus amoebocyte lysate (LAL) chromogenic assay. For a 20 g mouse, 0.1 EU (endotoxin units) is the maximum endotoxin level considered as safe [21], and we took care that our particles were always below this threshold. Reglo peristaltic pumps (Ismatec, Germany) were used for dosing the reagent solutions during synthesis. Centrifugation of the nanoparticles was carried out with a Rotofix 32A centrifuge (Andreas Hettich GmbH, Tuttlingen, Germany) and a Heraeus Fresco 21 ultracentrifuge (Thermo Scientific, USA), respectively. Nanoparticle pellets were re-dispersed with an UP50H ultrasonic processor (sonotrode MS1; Hielscher Ultrasonics GmbH, Teltow, Germany). Freeze-drying of the nanoparticles was carried out with a Christ Alpha 2-4 LSC instrument (Martin Christ GmbH, Osterode am Harz, Germany). Lyophilized nanoparticles were stored at −80 • C before application. Nanoparticles were lyophilized with D-(+)-trehalose as cryoprotectant. Immediately before application, the nanoparticles were re-dispersed in the same volume of water as present before freezedrying under thorough vortexing. To calculate the concentration of nanoparticles, the Ca 2+ concentration was measured by AAS and then numerically converted to the most common calcium phosphate phase hydroxyapatite, Ca 10 (PO 4 ) 6 (OH) 2 (see [12] for details).

Nanoparticle Uptake and Cell Viability
Cells were seeded in 24-well cell culture plates and pre-incubated for 24 h. Fluorescently labeled nanoparticles were added at 0.125-1.0 mg/L Ca 2+ . After a further 24 h of incubation, the cells were enzymatically harvested, washed with PBS, and stained with annexin-V-FLUOS (Roche Diagnostics, Mannheim, Germany) according to manufacturer's recommendations for detection of dead (apoptotic) cells. Cellular nanoparticle uptake and cytotoxicity were analyzed by flow cytometry (Accuri C6 flow cytometer; BD Biosciences, Franklin Lakes, NJ, USA).

Incubation of Cells with siRNA-Loaded Nanoparticles
To study the impact of nanoparticles carrying functional p65 siRNA on the different cell types, the following experimental routines were followed: Prior to the investigations, cells were stimulated to induce an inflammation status: (I) SVEC4-10 cells were incubated with 10 µg/mL lipopolysaccharide (LPS) for 4 h; (II) TK-1 cells were stimulated with 5 µg/mL CD3ε-Biotin antibody and 2 µg/ mL CD28 antibody for 6 h prior to investigation, (III) MOPC-315 cells were incubated with 0.5 µM ODN-2006 and 75 ng/mL IL-4 for 3 h. Stimulated cells were subsequently exposed either to: (I) the decorated nanoparticles carrying functional siRNA (siRNAf, final concentration: 1 µg/mL, 72 h) to assess their impact on p65 protein expression; (II) decorated nanoparticles carrying non-functional, scrambled siRNA (siRNAs, final concentration: 1 µg/mL, 72 h) to control p65-specific functionality of encapsulated siRNAf; III) decorated nanoparticles containing no siRNA to control their impact of the unloaded nanoparticles on the p65 protein expression; (IV) SHfunctionalized nanoparticles containing no siRNA, to control the effect of the free antibody coupling linker on target T-cells; (V) functional or non-functional p65 siRNA dissolved in the transfection agent Lipofectamine TM (1 µg/mL, Thermo Fisher, Germany) to control the impact of free (non-encapsulated) siRNA with good availability in the cytoplasm; and (VI) Lipofectamine TM (1 µg/ mL) without siRNA to control the impact of this substance per se on p65 protein expression. The nanoparticle concentration in the experimental arms (I) to (IV) was estimated by the Ca 2+ component concentration, the corresponding values were 0.8-1.0 mg/L Ca 2+ . After nanoparticle treatment, the cells were lysed for protein isolation with RIPA buffer. The total protein concentration was measured with the Bradford assay. Cell lysates were used to measure protein levels.

p65 Protein Expression after Incubation with siRNA-Loaded Nanoparticles
After treatment, the cells were lysed for protein isolation with a peqGOLD TriFastTM reagent (VWR, Germany). The total protein concentration was measured via the Bradford assay. After electrophoretic separation (10% (w/v) SDS-Page, 20 µg total protein per lane) and Western blotting, proteins were probed with antibodies against p65 NF-κB (Santa Cruz Biotechnology, Heidelberg, Germany) and β-actin (Abcam, Cambridge, UK). The p65 expression of the different treatment groups was analyzed densitometrically by the ImageJsoftware (NIH, Bethesda, MD, USA). Furthermore, the regulation of protein expression (up-or down-regulation) was calculated as the ratio of the p65 protein expression with respect to the non-treated inflammatory condition set to zero.

Expression of Affinity Molecules on the Surface of Target T-Cells
The expression of integrin αVβ3 on SVEC4-10 cells and CD69 on MOPC-315 and TK-1 cells was investigated as control for the presence of the specific anchor of our decorated nanoparticles on the cell surface. For this, SVEC4-10 cells were enzymatically harvested and stained with an anti-mouse CD51 antibody (from rat, PE-labeled, 0.01 µg/µL, BioLegend ® , USA) or the IgG isotype (isotype control, IgG1, κ Isotype, from rat, PE-labeled, 0.01 µg/µL, BioLegend ® , USA) on ice in the dark for 30 min. MOPC-315 and TK-1 were harvested, centrifuged (200× g, 5 min, 4 • C), and washed with 1 (w/v) % BSA in PBS. Cells were stained with anti-mouse CD69 (from Armenian hamster, PE-labeled, 2.5 ng/µL, BioLegend ® , USA) or the IgG isotype (isotype control, from Armenian hamster, PE-labeled, 2.5 ng/µL, Biolegend, USA) for 30 min on ice in the dark. Expression of activation markers on the cell surface was analyzed by flow cytometry, as described above.

Statistics
Data requiring statistical analyses were evaluated with the Prism 9 program (GraphPad software, San Diego, CA, USA). All data were considered to be normally distributed based on literature reports on normality distribution of the same variables. A Student's t-test or an ANOVA (analysis of variance) with Tukey's post hoc test was used to compare groups. Differences with p-values of 0.05 or less were considered as statistically significant.

Short Denomination RGD-F-NP RGD-S-NP
The efficiency of siRNA encapsulation into the nanoparticles was up to 90%. The efficiency of the nanoparticle surface decoration with RGD peptides (cyclic, cRGDfK, see Section 2) and with IgG antibodies was ∼95% and ∼50%, respectively.

The Impact of RGD-Peptide Decorated and Functional (p65 siRNAf) Nanoparticles on Endothelial Cells
In endothelial cells, the RGD-decorated and functional (p65 siRNAf) nanoparticles down-regulated the p65 NF-κB protein expression with respect to non-nanoparticle exposed cells (LPS-primed cells only, set to zero), albeit to a lesser extent when cells were transfected with p65 functional siRNA in the presence of Lipofectamine TM at equivalent concentrations (1 µg/mL, Figure 3A). The nanoparticle decoration with RGD per se stimulated p65 NF-κB expression, but the additional presence of non-functional siRNA in the nanoparticle configuration attenuated this effect. The nanoparticle binding was, at least in parts, specific for α5β3-integrin, as seen by competition assays and the analyses for the presence of integrin αV by flow cytometry (Figure 3B,C). The mentioned impact of the nanoparticles on the p65 NF-κB expression was the result of RGD-mediated binding and uptake by endothelial cells, since a larger number of nanoparticle-positive cells was observed compared to the exposure of cells to non-RGD-decorated nanoparticles ( Figure 3D). Furthermore, a bias on the mentioned effects due to nanoparticle cytotoxicity can be excluded ( Figure 3E).

The Impact of Antibody-Decorated and Functional (p65 siRNAf) Nanoparticles on T-and B-Cells
In T-cells, all IgG-decorated nanoparticles resulted in a down-regulation of p65 NF-κB protein expression with respect to non-nanoparticle exposed cells (stimulated cells only, set to zero) independently of whether or not the siRNA was specific for p65 (siRNAf or siRNAs). Obviously, the combination of IgG and siRNA was favorable to down-regulate p65 NF-κB protein expression in those nanoparticle-exposed cells ( Figure 4A). In contrast, p65 functional siRNA transfected with Lipofectamine TM (Thermo Fisher Scientific) rather up-regulated p65 in T-cells, whereas non-functional siRNA did not. Moreover, the further nanoparticle controls had ambiguous effects on p65 NF-κB protein expression, depending on the fact of being native nanoparticles, free p65 functional siRNA transfected with Lipofectamine™ or Lipofectamine™ per se. The nanoparticle accumulation in T-cells was, at least in parts, specific for CD69 and for other antibody binding sites at the cell surface (e.g. FcγR), as detected via competition experiments (presence of excess of free IgG targeting or not CD69) together with the corroboration of the expression of CD69 on the T-cell surface via flow cytometry ( Figure 4B,C). There was an increased accumulation of IgG-decorated nanoparticles in T-cells (with or without specificity for CD69) in comparison to native nanoparticles (SH or non-decorated, Figure 4D). Moreover, a clear concentration dependency up to 0.5 mg/L Ca 2+ and a good biocompatibility (no cytotoxicity, Figure 4E) were found. In T-cells, all IgG-decorated nanoparticles resulted in a down-regulation of p65 353 NF-κB protein expression with respect to non-nanoparticle exposed cells (stimulated 354 cells only, set to zero) independently of whether or not the siRNA was specific for p65 355 (siRNAf or siRNAs). Obviously, the combination of IgG and siRNA was favorable to 356 down-regulate p65 NF-κB protein expression in those nanoparticle-exposed cells (Figure 357  4A). In contrast, p65 functional siRNA transfected with Lipofectamine TM rather 358 up-regulated p65 in T-cells, whereas non-functional siRNA did not. Moreover, the fur-359 ther nanoparticle controls had ambiguous effects on p65 NF-κB protein expression, de-360 on the T-cell surface via flow cytometry ( Figure 4B,C). There was an increased accumu-366 lation of IgG-decorated nanoparticles in T-cells (with or without specificity for CD69) in 367 comparison to native nanoparticles (SH or non-decorated, Figure 4D). Moreover, a clear 368 concentration dependency up to 0.5 mg/L Ca 2+ and a good biocompatibility (no cytotoxi-369 city, Figure 4E) were found. 370 371 Figure 4. The potential of CD69-decorated CaP/PEI-Cy5/SiO2 nanoparticles to down-regulate 372 NF-kappa B p65 protein expression in murine T-cells (TK-1). A) Potential of nanoparticles to 373 down-regulate p65 together with each of their components as determined via immunoblotting, 374 nanoparticle concentration: 1 µ g/mL siRNA. For detailed specification of nanoparticle formulations 375 used see Table 1 and 2. F = functional siRNA, S = scrambled siRNA, Lp = Lipofectamine TM . The 376 regulation is the ratio of the p65 expression with respect to the inflammatory condition 377 (CD3/CD28-stimulated cells, absence of nanoparticles) set to zero. "+ regulation" = up-regulation; "-378 regulation" = down-regulation. The dashed line depicts the non-inflammatory condition. B) Speci-379 ficity of CD69 or FcR-specific-binding of the decorated nanoparticles as determined via flow cy-380 tometry (change of nanoparticle-positive cells under competition condition (non-competition con-381 dition = 0)). C) Control for the expression of CD69 on T-cells surfaces as target molecule of deco-382 rated nanoparticles determined via flow cytometry. D) Nanoparticle uptake. E) Nanoparticle bio-383 compatibility (annexin-negative cells). Statistical differences between control or indicated groups 384 with * p < 0.05, ** p < 0.01, **** p < 0.0001. n.d.: not determined. 385 In B-cells, the exposure to nanoparticles with 1 µ g/mL p65 functional siRNA was 386 not sufficient to down-regulate of the expression of p65 NF-κB. Instead, there was a 387 slight up-regulation of p65, independent from the nanoparticle decoration with IgG 388 (with or without CD69 antigen specificity), or the encapsulation of siRNA (p65 function-389 al or non-functional) into them. The presence of native nanoparticles up-regulated the 390 p65 protein expression above the level of non-nanoparticle-exposed but stimulated 391 B-cells. Only the Lipofectamine TM -mediated transfection of p65 functional siRNA was 392 very efficient in those cells (control experimental arm) ( Figure 5A). Furthermore, there 393 In B-cells, the exposure to nanoparticles with 1 µg/mL p65 functional siRNA was not sufficient to down-regulate of the expression of p65 NF-κB. Instead, there was a slight up-regulation of p65, independent from the nanoparticle decoration with IgG (with or without CD69 antigen specificity), or the encapsulation of siRNA (p65 functional or nonfunctional) into them. The presence of native nanoparticles up-regulated the p65 protein expression above the level of non-nanoparticle-exposed but stimulated B-cells. Only the Lipofectamine TM -mediated transfection of p65 functional siRNA was very efficient in those cells (control experimental arm) ( Figure 5A). Furthermore, there was no specificity in B-cell uptake of such decorated nanoparticles, neither with CD69 nor with IgG-isotypedecoration, although the investigated B-cells expressed CD69, as shown via flow cytometry ( Figure 5B,C). Beyond this, the nanoparticle decoration with IgG (whether specific or not for CD69) generally increased their binding and uptake in B-cells, and there was no concentration dependency of nanoparticle binding from 0.05 mg/L Ca 2+ and above ( Figure 5D). The nanoparticles showed a good biocompatibility and only a slight cytotoxic effect on B-cells at concentrations higher than 0.1 mg/L Ca 2+ ( Figure 5E). was no specificity in B-cell uptake of such decorated nanoparticles, neither with CD69 394 nor with IgG-isotype-decoration, although the investigated B-cells expressed CD69, as 395 shown via flow cytometry ( Figure 5B and C). Beyond this, the nanoparticle decoration 396 with IgG (whether specific or not for CD69) generally increased their binding and up-397 take in B-cells, and there was no concentration dependency of nanoparticle binding from 398 0.05 mg/L Ca 2+ and above ( Figure 5D). The nanoparticles showed a good biocompatibil-399 ity and only a slight cytotoxic effect on B-cells at concentrations higher than 0.1 mg/L 400 Ca 2+ ( Figure 5E). 401 402 Figure 5. Uptake behavior of CD69-decorated CaP/PEI-Cy5/SiO2 nanoparticles and potential to 403 down-regulate NF-kappa B p65 protein expression in murine B-cells (MOPC-315). A) Potential of 404 nanoparticles to down-regulate p65 together with each of their components as determined via 405 immunoblotting, nanoparticle concentration: 1 µ g/mL siRNA. For detailed specification of nano-406 particles formulations used see Table 1 and 2. F = functional siRNA, S = scrambled siRNA, Lp = 407 Lipofectamine TM . The regulation is the ratio of the p65 expression with respect to the inflammatory 408 condition (ODN-2006/IL-4-stimulated cells, absence of nanoparticles) set to zero. "+ regulation" = 409 up-regulation; "-regulation" = down-regulation. The dashed line depicts the non-inflammatory 410 condition. B) Specificity of CD69 or FcR-specific-binding of the decorated nanoparticles as de-411 termined via flow cytometry (change of nanoparticle-positive cells under competition condition 412 (non-competition condition = 0)). C) Control for the expression of CD69 on B-cells surface as target 413 molecule of decorated nanoparticles determined via flow cytometry. D) Nanoparticle uptake. E) 414 Nanoparticle biocompatibility (annexin-negative cells). n.d.: not determined. Statistical differences 415 between control or indicated groups with ** p < 0.01, *** p < 0.001, **** p < 0.0001. 416 The comparison of the nanoparticle behavior in cellular players of inflammation 417 showed that the cellular accumulation can be increased when they are decorated with 418 xenogenic IgG vs. allogenic ones ( Figure 6). Namely, there is a very strong accumulation 419 in B-cells, endothelial cells, and monocytes, but a comparatively lower one in T-cells. The 420 The comparison of the nanoparticle behavior in cellular players of inflammation showed that the cellular accumulation can be increased when they are decorated with xenogenic IgG vs. allogenic ones ( Figure 6). Namely, there is a very strong accumulation in B-cells, endothelial cells, and monocytes, but a comparatively lower one in T-cells. The effect is particularly prominent at low nanoparticle concentrations (0.01 mg/mL Ca 2+ ) and weaker at higher nanoparticle concentrations (0.1 mg/mL Ca 2+ ). In contrast, the uptake is suppressed in the presence of allogenic IgG, particularly in B-cells and in T-cells, but to a lesser extent in endothelial cells and monocytes. effect is particularly prominent at low nanoparticle concentrations (0.01 mg/mL Ca 2+ ) and 421 weaker at higher nanoparticle concentrations (0.1 mg/mL Ca 2+ ). In contrast, the uptake is 422 suppressed in the presence of allogenic IgG, particularly in B-cells and in T-cells, but to a 423 lesser extent in endothelial cells and monocytes.  Table 1 and 2. Statistical differences 432 between indicated groups with * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. 433

434
Independently from the type of nanoparticle surface decoration (with either RGD 435 peptide or IgG antibodies), the calcium phosphate nanoparticles had similar physico-436 chemical properties. Additionally, our data showed that 1) our RGD-decorated nanopar-437 ticles are efficient in down-regulating p65 NF-κB protein expression in endothelial cells 438 as a result of increased specific binding uptake of nanoparticles in those cells and 2) that 439 nanoparticles decorated with IgG with specificity for CD69 are efficient in 440 down-regulating p65 NF-κB protein expression in T-cells but not in B-cells, whereas the 441 nanoparticle uptake in those cells was mediated, at least in parts, by IgG-based (but 442 non-CD69-specific) nanoparticle binding. Low-dose nanoparticle decoration with xeno-443 genic IgG stimulated their uptake in leucocytes and endothelial cells. The knowledge of 444 mentioned biological interactions of the nature and surface functionalities of therapeutic 445 nanoparticles should be helpful when addressing immune cells with NF-κB-specific 446 siRNAs in the future.  Tables 1 and 2. Statistical differences between indicated groups with * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Discussion
Independently from the type of nanoparticle surface decoration (with either RGD peptide or IgG antibodies), the calcium phosphate nanoparticles had similar physicochemical properties. Additionally, our data showed that (1) our RGD-decorated nanoparticles are efficient in down-regulating p65 NF-κB protein expression in endothelial cells as a result of increased specific binding uptake of nanoparticles in those cells and (2) that nanoparticles decorated with IgG with specificity for CD69 are efficient in down-regulating p65 NF-κB protein expression in T-cells but not in B-cells, whereas the nanoparticle uptake in those cells was mediated, at least in parts, by IgG-based (but non-CD69-specific) nanoparticle binding. Low-dose nanoparticle decoration with xenogenic IgG stimulated their uptake in leucocytes and endothelial cells. The knowledge of mentioned biological interactions of the nature and surface functionalities of therapeutic nanoparticles should be helpful when addressing immune cells with NF-κB-specific siRNAs in the future.
The combination of the targeting moiety RGD-peptide with the p65 functional siRNA led to a down-regulation of p65 NF-κB in endothelial cells. Interestingly, the presence of RGD in unloaded nanoparticles (or in nanoparticles with encapsulated non-functional siRNA) was rather stimulating in terms of p65 expression in those cells. Although the underlying reasons are not yet clear, we tentatively postulate that the binding of the nanoparticles to RGD per se exerts a stimulatory effect on p65 NF-κB expression, and that p65 plays a particular role in this process. In this view, it has been demonstrated in a number of studies that integrin binding rapidly enhances the NF-κB activity (e.g., [9,22]). The nanoparticles are internalized into the endothelial cells via membrane invagination [20,21]. The internalized calcium phosphate nanoparticles are expected to be degraded in endolysosomes and the siRNA cargo released into the cytoplasm [12], where it can exert its gene-silencing effect.
The RGD peptide-decorated nanoparticles carrying p65 siRNA are effective in attenuating inflammation and inflammation-associated diseases, where p65 NF-κB is involved in cell growth, mediator secretion, and many other processes [9], especially if there is a favorable balance between the RGD-mediated cellular uptake and the intracellular availability of anti-p65 siRNA. Beyond the few studies showing the impact of injecting non-encapsulated RGD-p65-siRNA nanoparticle conjugates in mice with rheumatoid arthritis (1 mg/mL i.v. tail vein injection [23]), the particular advantage of using multi-shell calcium phosphate nanoparticles is the fact that the siRNA can be protected from degradation in the blood. We postulate that the therapeutic concentration of RGD-decorated calcium phosphate siRNA nanoparticles should be determined thoroughly also from the viewpoint that very high silica nanoparticle concentrations (25, 50, 100, and 200 µg/mL) were reported to activate NF-κB pathways by induction of oxidative stress [24].
Our IgG-decorated nanoparticles carrying siRNA were efficient in down-regulating p65 NF-κB in T-cells, particularly when carrying antibodies against CD69 and non-functional siRNA. This result could be attributed to the multi-faceted role of NF-κB in inflamed T-cells, involving not only the canonical, i.e., p65-mediated pathway, but also the non-canonical one with additional NF-κB players, such as RANK (receptor activator of NF-κB), p100 (a member of the NF-κB protein family), NIK (NF-κB-binding kinase), or JNK (c-Jun NH 2terminal kinase) [25][26][27][28]. In T-cells, the nanoparticle decoration with IgG distinctly favored its binding and accumulation, compared to the non-decorated nanoparticles (plain, or SH-terminated). Nevertheless, the nanoparticle accumulation in those cells was lower compared to the other immune cells investigated in this study. Moreover, the IgG-mediated nanoparticle binding to T-cells was, at least in parts, specific for CD69. From competition experiments and derived from the fact that the receptor CD69 is internalized upon binding and degraded afterwards [17], we deduce that CD69-targeting promotes nanoparticle internalization into T-cells. Further nanoparticle internalization into T-cells could have happened with FcγR, where the Fc-portion of the IgG of the nanoparticle surface acted as ligand, since competition experiments have shown a weak but discernible specificity for IgG-mediated accumulation. Moreover, it is well known that T-cells express FcγR only during a narrow window following cellular activation [29]. The internalization of IgG-decorated nanoparticles into T-cells occurred non-specifically as our competition experiments showed. Potential targets are non-specific membrane invaginations after interactions between the IgG molecules and the glycocalyx of cells. Given the curvature and size of the nanoparticles, we postulate a receptor-mediated endocytosis followed by degradation in endolysosomes, and the delivery of free siRNA into the cytoplasm [12].
In B-cells, the down-regulation of p65 NF-κB was not visible despite a comparatively strong non-IgG-based nanoparticle uptake by these cells. In this case, the nanoparticle accumulation was neither CD69-nor FcγR-specific, as our competition experiments showed. The reasons for such a relatively high non-specific B-cell accumulation of IgG-decorated nanoparticles are unknown, and they may well be associated with unspecific endocytosis (see above) or MHCII recognition processes. The rather up-regulation of p65 NF-κB protein expression in B-cells could be associated with the fact that non-specific nanoparticle bindings per se activated NF-κB pathways. For example, there is a known cross-talk between the NF-κB and Fc-receptor signaling [30], and furthermore, a cross-talk between CD69-related JAK/STAT and the NF-κB signaling pathway [17,31]. In this view, it is conceivable that the specific binding of CD69 receptors on B-cells via anti-CD69 IgGmolecules on NPs may have well triggered NF-κB expression. Furthermore, it is generally known that IgG molecules contain Fc-regions and that B-cells express Fc-receptors opposed to T-cells [32]. Besides the mentioned specific CD69 receptor-mediated NP-binding, an unspecific binding could have occurred between the Fc-regions of the IgG-based NP surface decoration and Fc surface receptors of B-cells. Such interactions could have activated NF-κB signaling, and consequently, led to p65-NF-κB up-regulation. Since the process of NP uptake occurs earlier in time than the NP cargo release into endolysosomes and cytoplasma [33], the intracellularly delivered anti-p65 siRNA was not able to reverse the preceded p65-NF-κB up-regulation upon NP-binding to B-cells. This view is in agreement with our observation that the Lipofectamine TM -mediated transfection of free p65 functional siRNA into B-cells was effective in reducing their p65 NF-κB protein expression (absence of NP-encapsulated siRNA) as opposed to the IgGsurface decorated NPs. This means that upon administration in vivo, IgG-decorated nanoparticles will be effective in downregulating p65 NF-κB protein expression in endothelial cells and T-cells but not in B-cells.
In the present study, we generally refer to the total amount of p65 NF-κB protein, and not directly to the activated p65 NF-κB protein, which is known to be translocated to the cell nucleus. This means that the siRNA-mediated inhibition of the p65 NF-κB protein will not necessarily correlate with the functional inhibition of NF-κB activity. To verify such correlations, further in vitro and in vivo animal experiments should be conducted in the future. Another point to be addressed in the long run is the number of nanoparticles internalized by inflamed cells of the blood compartment and their therapeutic efficacy in the in vivo situation.
The fact that nanoparticles decorated with xenogenic IgG were better accumulated by all investigated cell types may potentially open up a strategy to regulate the intracellular availability of siRNA in the cytoplasm of specific immune cells. We expect that nanoparticle accumulation occurred mainly via non-specific processes, such as MHCII recognition in the investigated cells. Nevertheless, special attention should be paid to the fact that some unwanted activation of NF-κB-mediated pro-inflammatory pathways may occur in association with its recognition as foreign antigen [34]. A potential strategy may be accomplished by decorating calcium phosphate nanoparticles with very small amounts of xenogenic IgGs or technologically-modified IgGs, which could act as "subliminal baits" for immune cells, particularly for those which have a low intrinsic propensity to accumulate calcium phosphate nanoparticles, such as T-cells.

Conclusions
In summary, our data reveal that the composition of the calcium phosphate nanoparticles and the presence of a peptide-or antibody-based targeting moiety on their surface have an impact on the p65 NF-κB protein expression in different immune cells, and such effects can either stimulate or even weaken the functionality of the encapsulated functional siRNA against p65. Finally, a smart nanoparticle decoration with xenogenic IgG may stimulate their uptake in certain immune cells, which are otherwise difficult to address. Such interrelations have seldom been taken into account so far. It is vitally important to understand said interactions; this will allow us optimizing the functionalities of therapeutic nanoparticle-based biomedicines more specifically when addressing immune cells with anti-NF-κB siRNAs.