- Email: [email protected]
Transfection of macrophages by collagen hollow spheres loaded with polyplexes: A step towards modulating inflammation
Acta Biomaterialia 8 (2012) 4208–4214
Contents lists available at SciVerse ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Transfection of macrophages by collagen hollow spheres loaded with polyplexes: A step towards modulating inflammation q Christophe Helary ⇑, Shane Browne, Asha Mathew, Wenxin Wang, Abhay Pandit ⇑ Network of Excellence for Functional Biomaterials, National University of Ireland, Galway, Ireland
a r t i c l e
i n f o
Article history: Available online 15 June 2012 Keywords: Chronic inflammation Macrophages Collagen microspheres Polyplexes Gene delivery
a b s t r a c t Macrophages are key orchestrators of inflammation as they secrete proteases and inflammatory cytokines. To date, therapies aimed at modulating macrophage phenotype have failed due to the short half-life of biomolecules in the body. Therefore, inhibition of inflammation by gene therapy constitutes a new hope. In the present study, we have assessed collagen hollow spheres as a reservoir system for polyplexes in order to transfect human macrophages while preserving cell viability. Polyplexes were formed by complexing G-Luc plasmid with a poly(2-dimethylaminoethyl methacrylate) poly(ethylene glycol) based hyperbranched polymer. Several ratios of polymer/pDNA (5:1, 8:1, 10:1 w/w) complexes in two different sphere sizes (1.24 and 4.5 lm) were tested. Collagen hollow spheres were loaded with polyplexes up to 80 lg of pDNA per mg of microspheres. The release of polyplexes from the spheres was delayed and prolonged i.e. 20% of the initial amount released in 5 days. Following incubation with polyplex-loaded microspheres, macrophages were transfected (polyplex pDNA:polymer ratio 1:10 w/w). In addition, collagen hollow spheres maintained cell viability as more than 80% of cells were viable after 4 days in culture. In contrast, when used alone, polyplexes were seen to be toxic, while there was no transfection detected. Taken together, these results show that collagen hollow spheres may be used as a reservoir for controlled gene delivery to macrophages. Unlike existing gene delivery systems, this system allows for macrophage transfection with minimal toxicity. Hence, this system has a potential for the delivery of a therapeutic gene in order to modulate inflammation. Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Macrophages are key cells in the resolution of inflammation. During the inflammatory phase following injury, macrophages adopt a classically activated phenotype characterized by the secretion of reactive oxygen species, inflammatory cytokines and proteases [1,2]. Following exposure to biological signals such as interleukin (IL)-13, IL-4 or IL-10, macrophages progressively change their phenotype to adopt an alternatively activated phenotype. This phenotype promotes wound healing by the secretion of IL-10, VEGF and TGF-b1 to suppress the inflammatory response and promote matrix formation and stabilization [2]. In the case of many chronic inflammatory conditions, macrophages are locked in a pro-inflammatory phenotype [3].
q Part of the Special Issue ‘‘Advanced Functional Polymers in Medicine (AFPM)’’, guest editors: Professors Luigi Ambrosio, Dirk W. Grijpma and Andreas Lendlein. ⇑ Corresponding authors. E-mail addresses: [email protected] (C. Helary), abhay.pandit@ nuigalway.ie (A. Pandit).
With the aim of modulating inflammation, several trials based on the injection of growth factors have been performed. Unfortunately, owing to the short half-life and the rapid diffusion of biomolecules in vivo [4,5], these attempts have thus far proved unsuccessful. Hence, gene therapy in the form of the transfection of a gene encoding for a protein capable of modulating inflammation presents a new possibility to modulate the pro-inflammatory environment. Unfortunately macrophages, which are a non-dividing cell type, are difficult to transfect. Firstly, viral transfecting reagents cannot be used as they trigger an immune response [6]. Secondly, pDNA barely penetrates into the nucleus of non-dividing cells. Lastly, regular non-viral reagents reduce macrophage viability after transfection [7,8]. In gene therapy research, the utilization of cationic polymers has proven to have several advantages when compared to cationic lipids. Cationic polymers are more stable than lipids and complex a large amount of pDNA to form positively charged polyplexes. Unfortunately, however, they are toxic for macrophages [8]. To overcome this, polymeric hollow spheres made from natural polymers have been developed as a gene depot system [9]. These systems allow for the sustained release of polyplexes, which can
1742-7061/$ - see front matter Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.actbio.2012.06.017
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
protect macrophages from the toxicity of large, bolus doses of polyplexes. Several studies have shown that hollow spheres can be fabricated from natural polymers using a template method with sizes on the nanometer to micrometer scale possible [10]. Hollow spheres made from natural polymers have the advantage that they will not release toxic degradation products and are not detrimental to cell viability [11]. This technique relies on the coating of polystyrene beads by natural polymers such as chitosan or elastin-like peptides due to an electrostatic interaction between the negatively charged polystyrene beads and the positively charged polymers in solution. After the removal of the polystyrene bead template using the solvent tetrahydrofuran (THF), hollow spheres can be obtained. These studies demonstrate that spheres can be efficiently loaded with polyplexes, which modulates the polyplex release pattern and allows for transfection of cells in vitro [11]. In addition, sustained release of polyplexes protected cells from polymer cytotoxicity. Recently, Browne and collaborators fabricated a new type of hollow sphere made from atelocollagen type I [13]. This type of sphere is easy to fabricate and can be loaded with a large amount of polyplexes. It was hypothesized in this study that a reservoir system composed of atellocollagen type 1 can load and release polyplexes in a consistent and controlled manner, with the released polyplexes capable of transfection while protecting macrophages against polymer toxicity associated with large doses. The objective of this study was to use collagen hollow spheres as a reservoir for polyplexes to transfect human macrophages. 2. Materials and methods 2.1. Collagen hollow sphere fabrication Collagen hollow spheres of defined sizes were fabricated using the template method as previously described [13]. Briefly, 1.2 or 4.5 lm sulfonated polystyrene beads were incubated with an atelocollagenÒ solution (5 mg ml1 in 0.5 M acetic acid) for 4 h (supplementary information no. 1). After cross-linking with pentaerythritol poly(ethylene glycol) ether tetrasuccinimidyl glutarate 4 arm-StarPEG for 2 h, the polystyrene core was removed with THF, leaving an atellocollagen hollow sphere of defined size.
4209
polymers: polyethylenimine 25 kDa (PEI; Sigma, Ireland) and SuperfectÒ (Qiagen, Ireland). Linear PEI 25 kDa was used with an optimal ratio of 2:1 (polymer/G-Luc plasmid) and the partially degraded polyamido amine dendrimer (SuperfectÒ) was used with an optimal ratio of 8:1. 2.4. Analysis of polyplexes The size, zeta potential and polydispersity index of polyplexes were measured by Zetasizer (Malvern Instruments Zetasizer Nano-2590). This analysis was performed on polyplexes formed with the polymer B81 after 12 h of incubation. 2.5. Loading studies One mg of 1.2 or 4.5 lm collagen hollow spheres was resuspended in 300 ll of PBS. Polyplexes (ratio 10:1 w/w) containing a predetermined amount of pDNA (40, 80, 160 and 320 lg) were added to these spheres, followed by a volume of THF to achieve a final concentration of 50% (v/v). The mixtures were then agitated for 6 h at room temperature. After this period of incubation, the vials were opened to let the THF evaporate for 6 h. Next, the suspensions were centrifuged at 13,000g and the sphere/polyplex complexes were washed four times with ultrapure water. The supernatants collected were used to estimate the amount of free pDNA. A PicoGreenÒ assay (Invitrogen, Ireland) was performed to quantify free pDNA. To obtain free pDNA, polyplex samples collected during loading were treated with polyglutamic acid (PGA) of 10 mg ml1 concentration at 37 °C for 30 min with a modification of the protocol described in Ref. [11]. Then 100 ll of PicoGreenÒ was added to each sample of similar volume and the fluorescence was analyzed at 480 nm. A standard curve was prepared using naked pDNA. First the amount of pDNA detected in the supernatants was determined, then the loading efficiency of the hollow spheres was determined using the formula: Loading efficiency ¼
Initial pDNA added pDNA remaining in supernatant Initial pDNA added
2.6. Release pattern of polyplexes from collagen hollow spheres 2.2. Characterization by scanning electron microscopy (SEM) Prior to analysis, collagen hollow spheres were fixed with 4% paraformaldehyde. A drop of sample containing collagen hollow spheres was placed on adhesive carbon tabs mounted on SEM specimen stubs and then dried. The specimens were subsequently coated with gold using an Emitech K550 coating system. SEM images were obtained using a Hitachi S-4700 field emission microscope operating with a beam voltage of 15 kV. 2.3. Polyplex formation Polyplexes were formed by complexing the synthesized polymer (B81) with G-Luc plasmid encoding for the protein Gaussia Luciferase (New England Biolabs, Ireland). The transfecting agent B81 (14 kDa) synthesized in our group is composed of a linear poly(2dimethylaminoethyl methacrylate) block (pDMAEMA), a hyperbranched poly(ethylene glycol methyl ether methacrylate) block and an ethylene dimethacrylate block (see structure in supplementary information no. 2). The polymer B81 was synthesized via the deactivation enhanced atom transfer radical polymerization technique. Polyplexes were formed in phosphate-buffered saline (PBS) with weight ratios from 5:1 to 10:1 (Polymer B81/G-Luc plasmid) for 1 h. To assess the transfection efficiency with the polymer B81, we compared the results with those of the commercial
To characterize the release pattern of the polyplex-loaded collagen hollow spheres, 1 mg of polyplex-loaded spheres of 1.2 or 4.5 lm size was resuspended in PBS and incubated at 37 °C. The spheres used for this experiment were loaded with 40 lg of pDNA per mg of spheres. At each time point, the suspension was centrifuged and a sample of the supernatant taken. The released polyplexes were treated with polyglutamic acid (Sigma, Ireland) and quantified using the PicoGreenÒ (Invitrogen, Ireland) assay. To determine the release pattern, five time points were considered: days 1, 2, 3, 4 and 5. The results were expressed as a percentage of initial amounts of pDNA loaded in the spheres. 2.7. Macrophage differentiation and activation The human myelogenous leukemia cell line THP-1 was obtained from the ATCC. The cells were maintained in RPMI 1640 (GIBCOBRL) supplemented with 10% fetal bovine serum (FBS), penicillin (100 U ml1) and streptomycin (100 lg ml1) at 37 °C in a 5% CO2 humidified incubator. The mature macrophage-like state was induced by treating THP-1 cells for 24 h with phorbol 12-myristate 13 acetate (PMA) at 100 ng ml1 diluted in serum-free medium. Cells were seeded at 100,000 cells well1 into a 48-well plastic well plate. The following day, plastic-adherent cells were washed twice with cold, sterile Dulbecco’s PBS and incubated with fresh
4210
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
RPMI-1640 lacking PMA but containing 1% FBS, penicillin (100 U ml1) and streptomycin (100 lg ml1). The cells were then activated with 100 ng ml1 lipopolysaccharide in RPMI-1640 medium containing 1% FBS (supplementary information no. 3). 2.8. Collagen hollow sphere uptake by activated THP-1 THP-1 cells were differentiated and activated as previously described. First, 300 lg of FITC-labelled microspheres of three sizes was added to 250,000 cells and incubated for 24 h. The cells were then rinsed with Hanks’ balanced salt solution three times to remove any spheres that were not internalized. A cell scraper was used to detach the cells from the surface, and these were then fixed with 4% PFA and examined by flow cytometry (BD FACS Canto 6) for internalization. The three sizes of microspheres examined were 100 nm, 1 lm and 10 lm. 2.9. Transfection studies Activated macrophages-like cells were incubated with polyplex-loaded collagen hollow spheres for 4 days. Two sizes of spheres were tested: 1.24 and 4.5 lm. Cells without treatment were kept as controls. For the tests, 100,000 cells were incubated with 100 lg of loaded spheres. Different loadings of spheres were tested: 40 and 80 lg of pDNA per mg of spheres. Regarding the polyplexes formed with polymer B81, three different weight ratios were tested: 5:1, 8:1 and 10:1. In parallel, the same experiment was carried out with PEI and partially degraded polyamido amine dendrimer (SuperfectÒ) at their optimal ratios. All experiments were performed in 1% FBS medium. For comparison, macrophages were incubated with polyplexes alone in FBS-free medium for 4 h. Then, polyplexes were removed and replaced by fresh medium (1% FBS). After the 96 h incubation period, 50 ll of medium was collected from each of the samples for quantification of luciferase expression using a Gaussia luciferase assay kit (New England Biolabs). The luciferase assay was performed in accordance with the manufacturer’s instructions. The luciferase expression level was expressed as relative light units. 2.10. Cell metabolic activity AlamarBlueÒ assay (Sigma, Ireland) was performed as previously described to quantify the cellular metabolic activity of macrophages incubated with polyplex-loaded spheres [11]. The viability was assessed as a percentage of the metabolic activity of control cells. 2.11. Statistical analysis All experiments were replicated in triplicate and the results expressed as the mean value ± standard deviation. Statistical significance was determined using Student’s test, with p < 0.05 considered significant. Statistical analysis was carried out between control samples and macrophages incubated with collagen hollow spheres loaded with polyplexes. 3. Results and discussion 3.1. Structure of collagen hollow spheres Following fabrication using the template method, the structure of collagen hollow spheres was analysed by SEM. This analysis revealed the presence of isolated collagen hollow spheres of 1.24 and 4.5 lm sizes. The microsphere populations were monodispersed and each sphere was of the same size as the template used
Fig. 1. Structure of collagen hollow spheres. The analysis by SEM revealed the presence of 1.24 (A) or 4.5 lm collagen microspheres (B). In addition, collagen hollow spheres appeared to be monodispersed and isolated.
(Fig. 1). Collagen hollow sphere fabrication was first described by Browne et al. in order to generate a new gene depot system [13]. The collagen hollow spheres were produced using the electrostatic properties of atelocollagen on a polystyrene template. In the context of macrophages and inflammation, the utilization of collagen hollow spheres has several advantages when compared to artificial materials. Atelocollagen is collagen I without the telopeptides at the N- and C- terminals, which enables much of its antigenicity to be removed by enzymatic cleavage. As these telopeptides are responsible for the mild immunogenicity of collagen, atelocollagen does not trigger inflammation when applied in an organism either of the same or different species. Collagen hollow spheres were cross-linked with 1 mM 4 arm-StarPEG. The aim of this step was to stabilize and tune the physical properties and the porosity of collagen microspheres [12]. It has been shown that 4 arm-StarPEG is suitable for use as a cross-linker in collagen-based tissue engineering systems as it does not affect cell viability [14]. Hence, collagen hollow spheres can be considered as a desirable system to transfect macrophages and modulate inflammation because they are fabricated from a material that does not trigger a host response. 3.2. Structure and polydispersity of G-Luc pDNA/polymer B81 polyplexes The size of polyplexes formed at a 10:1 ratio (w/w) was around 50–80 nm, with a zeta potential of 4 ± 2 mV [15]. Polyplexes formed at ratios of 5:1 and 8:1 (polymer/pDNA) had weaker charges (data not shown). To interact with cells and promote their endocytosis, polyplexes need to have a positive charge. Therefore, the ratio of 10:1 is the most suitable for cell transfection in these conditions. The polydispersity index of polyplexes, measured by Zetasizer, was 0.911 after 12 h of incubation in PBS. This result shows that polyplexes do not aggregate during this period. Hence, cell transfection was performed by isolated polyplexes. It has been shown that PEGylation prevents aggregation of polyplexes [16– 18]. 3.3. Maximum loading of polyplexes within collagen hollow spheres The loading study was carried out with polyplexes formed by the G-Luc plasmid and the polymer B81 (n = 6 for each condition), at various weight ratios of polyplexes to collagen microspheres. With a ratio of 1:320 (mg of spheres per lg of pDNA), the polyplexes precipitated when they were incubated for 24 h with collagen hollow spheres. The other ratios tested did not precipitate. The maximal loading of polyplexes was about 80 lg of pDNA per mg of collagen hollow spheres when a ratio of 1:160 (mg of spheres per lg of pDNA) was used for the loading study (Fig. 2A). The efficiency of loading with this ratio was 50% whereas it was almost 100% with the lowest ratio of 1:40. The loading of collagen hollow spheres by polyplexes was similar to that obtained with 500 nm
4211
A
Loading (µg pDNA G-Luc)
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
faster, reaching 70% after 48 h [20]. The polyplex release assessed from macroporous scaffolds such as collagen and fibrin hydrogels exhibits the same features of release as artificial microspheres, i.e. a burst of polyplexes during the first 48 h and a slow, low release thereafter [21]. In contrast, collagen sponges allow delayed delivery, with only 20% of polyplexes being released after 50 days [22]. Using dendrimers (SuperfectÒ) or PEI, the release of polyplexes from collagen spheres was faster. Indeed, 40% of the initial dose was delivered after 24 h and 60% after 3 days (data not shown) [13].
200
(n = 6) 150 100 50 0
B
DNA Release (% of initial amount)
1: 40 1: 80 1: 160 Ratios (mg microspheres/ µg pDNA)
100
Using the polymer B81 with the ratio of 5:1, no transfection was detected in macrophages, with no toxic effects noted (Fig. 3E). Indeed, macrophages retained their spindle shape, and their viability corresponded to 80% of that detected in samples incubated without polyplexes (Fig. 3A and E). From the 8:1 ratio, polymer toxicity in macrophages was observed, but with no detectable levels of transfection. Cells were less numerous than in control samples.
1,2 µm Spheres
80
4,5 µm Spheres
60 (n = 6)
40
3.5. Transfection of macrophages by polyplexes alone
20 0 0
1
2
3
4
5
A
B
C
D
Time (Days)
500
120
(n=6)
100
400
80 300 60 200 40 100
20 0
ct
8:
1
:1 I2 Su p
er
fe
PE
10 :
1
tio ra
tio ra
tio
5:
8:
1
l ro
1
0
ra
The release of loaded polyplexes used at a weight ratio of 10:1 (polymer B81/G-Luc pDNA) was delayed and prolonged. Moreover, the release pattern observed with 1.2 lm spheres was similar to that observed with 4.5 lm microspheres (n = 6 for each sphere size). Regardless of the microsphere size, the amount of pDNA released into the medium increased gradually over time to reach about 20% of the initial amount of pDNA after 5 days (Fig. 2B). As the maximal loading was 80 lg of pDNA per mg of sphere, 16 lg of plasmid was released after 5 days. In the transfection experiments, only 100 lg of microspheres was added to macrophages and two different loadings were tested. After 4 days at 37 °C, 0.64 and 1.28 lg of pDNA were released from microspheres loaded with 40 and 80 lg of pDNA per mg, respectively. In order to modulate inflammation, it is an advantage to have a system that allows for a slow rate and low level of release of polyplexes. The low rate of release prevents the potential toxicity of polymer [19], while the slowness permits a prolonged effect over 25 days. By comparison, the polyplex release from PLA microspheres is
Luciferase Activity (RLU)
3.4. Release pattern of polyplexes from collagen hollow spheres
E
Co nt
elastin-like polypeptide hollow spheres [11]. With this type of microsphere, the maximal loading was about 70 lg of pDNA per mg of spheres. These results showed the absence of relation between the sphere size and the loading, in that there was no significant difference between the loading efficiency of either size of microsphere.
Viability (% of Control)
Fig. 2. Loading and release profile of polyplexes at a ratio of 10:1 (polymer/pDNA) encapsulated within collagen hollow spheres. (A) Polyplex loading behaviour of collagen hollow spheres. Regardless of the sphere size, the maximal loading was around 90 lg of pDNA per mg of collagen hollow spheres. This loading was observed with the ratio of 1 mg of spheres per 160 lg of pDNA. Polyplexes were formed with polymer B81 and G-Luc plasmid using the 10:1 ratio (polymer/pDNA). (B) Release pattern of polyplexes from collagen hollow spheres. Regardless of the sphere size, the release of polyplexes was low and slow to reach about 20% of the initial pDNA amount after 5 days.
Polymer B81 Fig. 3. Transfection of macrophages by polyplexes. Using a ratio of 5:1, polyplexes formed from polymer B81 were not toxic but no transfection was detected (A and E). From the ratio of 8:1, polyplexes were toxic for cells while they did not permit their transfection (B and E). The commercial polymer PEI used with a ratio of 2:1 (C and E) and the partially degraded polyamido amine dendrimer (SuperfectÒ) used with a ratio of 8:1 (D and E) were also cytotoxic for macrophages. In addition, no transfection was observed. Histograms show luciferase activity and squares represent cell viability.
4212
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
40
30
20
10
10 um
1u m
10 0n m
0
C on tro l
Uptake of FITC-labelled Collagen microspheres (%)
Moreover, macrophages were no longer spindle shaped, but rather rounded with small extensions. As assessed by AlamarBlueÒ cell metabolic assay, cell viability represented about 50% of that observed in control macrophages (Fig. 3B and E). With the ratio of 10:1, no transfection was observed and only 40% of cells were viable. In this study, two commercial polymers were also tested in order to compare their efficacy with polymer B81. The observation by optical microscopy showed that PEI and partially degraded polyamido amine dendrimer (SuperfectÒ) were toxic for macrophages (Fig. 3C and D). As measured by AlamarBlueÒ, only 12 and 50% of cells were viable after 4 h of incubation with PEI and partially degraded polyamido amine dendrimer (SuperfectÒ), respectively (Fig. 3E). Furthermore, no transfection was detected in macrophages incubated with polyplexes formed from these two commercial polymers. PEI and dendrimers (SuperfectÒ) are known to be toxic to nondividing cells. It is thought that the presence of primary amines in their structure is responsible for this toxicity [23]. Nevertheless, this toxicity is not observed on the macrophage cell line RAW 264.7, as these macrophage-like cells can proliferate in culture. This ability to multiply makes transfection easier, as cells which divide are easier to transfect, and it also masks the toxic effect of these polymers [24]. However, macrophages do not proliferate in vivo. Hence, THP-1 cell line differentiated into macrophage-like cells is a model closer to physiological conditions as they do not multiply in culture. The utilization of THP-1 cells in these experiments displayed the toxicity of commercial polymers. In order to minimize polymer toxicity, a PEGylated polymer, B81 was tested. PEGylation shields the positive charge of the molecule and reduces toxicity [23]. In addition, the tertiary amine of the pDAEMA should improve the biocompatibility in comparison with transfecting agents with primary amines [11]. Unfortunately, these modifications were not sufficient to protect cells from polymer toxicity, as toxicity was seen with B81 despite no significant levels of transfection being observed.
Sphere Size Fig. 4. Phagocytosis of collagen hollow spheres by macrophages. The analysis by flow cytometry showed the engulfment of 100 nm collagen spheres. In contrast, 1 and 10 lm spheres were taken up less by macrophages.
3.6. Engulfment of collagen hollow spheres by macrophages The analysis by flow cytometry revealed that only 7 ± 2% of macrophages had engulfed 1 lm collagen hollow spheres after 24 h of incubation (Fig. 4). This figure dropped to 3 ± 1% when macrophages were incubated with bigger spheres (10 lm). In contrast, the engulfment of 100 nm spheres was observed in 30% of macrophages. Hence, collagen hollow spheres showed little engulfment by macrophages when their size was bigger than 1 lm. An internalization of spheres would cause high cell toxicity due to the high polyplex concentration. Therefore, this result provides evidence that collagen hollow spheres can be considered as a potential reservoir system to transfect macrophages and modulate inflammation. The mechanism of action relies on the diffusion of polyplexes from the collagen spheres which transfect macrophages after release. 3.7. Transfection of macrophages by collagen hollow spheres loaded with polyplexes No transfection was observed with the 5:1 and 8:1 ratios, regardless of the loading and the microsphere size used (Fig. 5A– C). With the 10:1 ratio, a statistically significant transfection was detected in macrophages (Fig. 5B). The luciferase activity detected in macrophages incubated with 40 and 80 lg of pDNA per mg of 4.5 lm spheres was 5 and 10 times higher than that in control samples, respectively (p < 0.05). The transfection of macrophages was not associated with polymer toxicity as more than 80% of macrophages were viable after 4 days of incubation with collagen hollow spheres. The results obtained with 1.2 lm collagen hollow spheres were similar except for the samples loaded with 40 lg of pDNA. With the ratio of 10:1, the luciferase activity was 10 times higher than the control cells (Fig. 5C). Therefore, the low, slow release of polyplexes allowed for the transfection while preventing polymer toxicity. In vitro, THP-1 cells are difficult to transfect. The classic way to transfect them was achieved by a cationic lipid, oligofectamineÒ [25]. The utilization of this transfecting reagent has several drawbacks, most notably toxicity. For example, it can inhibit the enzyme protein kinase C, thereby disturbing cell metabolism [26]. Another possibility based on nucleofection permits transfection of these cells, but only prior to them being differentiated into macrophages [8]. Polyplexes formed with PEI or partially degraded polyamido amine dendrimer (SuperfectÒ) were used in this study in order to compare with the transfection capability of polymer B81. Collagen hollow spheres were loaded with polyplexes made from these commercial polymers. Two loadings were tested: 40 and 80 lg of pDNA per mg of spheres. A transfection, similar to that obtained with B81, was observed with PEI (ratio 2:1), when loaded with 80 lg of pDNA per mg of spheres (p < 0.05). Unfortunately, this transfection was associated with high toxicity as only 40% of macrophages were viable after 4 days (Fig. 6). With the 40 lg loading, no transfection was detected while the cell toxicity remained the same as 80 lg of pDNA. In contrast, no toxicity and no transfection were detected when macrophages were incubated with collagen hollow spheres loaded with polyplexes made with partially degraded polyamido amine dendrimer (SuperfectÒ). This result adds further evidence to suggest that collagen hollow spheres loaded with polyplexes made from polymer B81 are suitable to transfect macrophages. 4. Conclusions These results show that collagen hollow spheres can be considered as a potent reservoir for controlled gene delivery. Unlike bolus
4213
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
A
*
100
1500 80 60
1000
*
40
500
(n=6)
:1
1
10
8:
1
tio
tio
ra
ra
ra
ra
tio
10
5:
:1
1 8:
tio
tio
5: ra
ra
C
tio
on
tro
1
l
0
4.5 µm spheres loaded with 80 µg pDNA/mg spheres
4.5 µm spheres loaded with 40 µg pDNA/mg spheres
Luciferase Activity (RLU)
4.51 µm Collgen Spheres
20 0
C
Viability (% of Control)
120
2000
2000
120
*
1500
100 80
*
1000
60 40
500
Viability (% of Control)
Luciferase Activity (RLU)
B
1.24 µm Collgen Spheres (n=6)
20
:1
1
10 tio
ra
tio
8:
1 ra
tio
5:
:1
4.5 µm spheres loaded with 40 µg pDNA/mg spheres
ra
tio ra
tio
10
8:
1 5: ra
tio ra
nt ro Co
1
0 l
0
4.5 µm spheres loaded with 80 µg pDNA/mg spheres
Fig. 5. Transfection of macrophages by collagen hollow spheres loaded with polyplexes. (A) Microscopic view of THP-1 like macrophages incubated for 4 days with 4.5 lm collagen hollow spheres. (B) Transfection efficacy of 4.5 lm collagen hollow spheres loaded with polyplexes (polymer B81/G-Luc plasmid) and their effect on cell viability. Using a ratio of 10:1 (polymer/pDNA), luciferase activity was detected, evidencing macrophage transfection (n = 6, p < 0.05). The transfection efficacy obtained with the loading 80 lg pDNA per mg of spheres was twice as high as that obtained with 40 lg. In addition, macrophage transfection was associated with high cell viability as more than 80% of macrophages were viable. (C) Transfection efficacy of 1.24 lm collagen hollow spheres loaded with polyplexes and their effect on cell viability. The results were similar to those obtained with 4.5 lm spheres, i.e. macrophage transfection detected with the ratio 10:1 (polymer/pDNA) associated with high viability (n = 6). Histograms show luciferase activity and lozenges represent cell viability.
delivery of existing polymers, this reservoir system allows for transfection of macrophages without toxicity. Hence, this system appears promising for the delivery of an inhibitory inflammatory therapeutic gene capable of modulating macrophage phenotype.
With this reservoir system, the transfection of a therapeutic gene such as IL-10, IL-12 or IL-4 could potentially trigger the polarization of inflammatory macrophages (M1) to regulatory macrophages (M2), which can promote tissue repair.
4214
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
4.5 µm Collagen Hollow Spheres (n=6) 120
*
100 1500 80 1000
60 40
500 20 0
Viability (% of Control)
Luciferase Activity (RLU)
2000
0 Control
40 µg pDNA
80 µg pDNA
PEI 2:1
40 µg pDNA
80 µg pDNA
Superfect 8:1
Fig. 6. Transfection of macrophages by 4.5 lm collagen hollow spheres loaded with polyplexes formed from commercial polymers. No transfection was detected with polyplexes formed from partially degraded polyamido amine dendrimer (SuperfectÒ) (ratio 8:1). In contrast, Luciferase activity was detected when PEI (ratio 2:1) was used and loaded into collagen hollow spheres (with 80 lg pDNA per mg of spheres). Unfortunately, the macrophage transfection was associated with cell toxicity (n = 6). Histograms show luciferase activity and lozenges represent cell viability.
Acknowledgements The authors thank Estelle Collin and Mohammad Abu-Rub for their discussion on this article. In addition, we thank Antony Sloan for editorial support while writing this article. Health Research Board (HRB RP/2008/188) for financial support of this work. Appendix A. Figures with essential colour discrimination Certain figures in this article, particularly Figs. 3 and 5, are difficult to interpret in black and white. The full colour images can be found in the on-line version, at http://dx.doi.org/10.1016/j.actbio. 2012.06.017. Appendix B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2012. 06.017. References [1] Mosser DM. The many faces of macrophage activation. J Leukoc Biol 2003; 73(2):209–12.
[2] Gordon S, Martinez FO. Alternative activation of macrophages: mechanism and functions. Immunity 2010;32(5):593–604. [3] Eming SA, Krieg T, Davidson JM. Inflammation in wound repair: molecular and cellular mechanisms. J Invest Dermatol 2007;127(3):514–25. [4] Eming SA, Kaufmann J, Lohrer R, Krieg T. Chronic wounds. Novel approaches in research and therapy. Hautarzt 2007;58(11):939–44. [5] Menke NB, Ward KR, Witten TM, Bonchev DG, Diegelmann RF. Impaired wound healing. Clin Dermatol 2007;25(1):19–25. [6] Thomas CE, Ehrhardt A, Kay MA. Progress and problems with the use of viral vectors for gene therapy. Nat Rev Genet 2003;4(5):346–58. [7] Lee SY, Cherla RP, Tesh VL. Simultaneous induction of apoptotic and survival signaling pathways in macrophage-like THP-1 cells by Shiga toxin 1. Infect Immun 2007;75(3):1291–302. [8] Schnoor M, Buers I, Sietmann A, Brodde MF, Hofnagel O, Robenek H, et al. Efficient non-viral transfection of THP-1 cells. J Immunol Methods 2009; 344(2):109–15. [9] Rethore G, Pandit A. Use of templates to fabricate nanoscale spherical structures for defined architectural control. Small 2010;6(4):488–98. [10] Dash BC, Rethore G, Monaghan M, Fitzgerald K, Gallagher W, Pandit A. The influence of size and charge of chitosan/polyglutamic acid hollow spheres on cellular internalization, viability and blood compatibility. Biomaterials 2010;31(32):8188–97. [11] Dash BC, Mahor S, Carroll O, Mathew A, Wang W, Woodhouse KA, et al. Tunable elastin-like polypeptide hollow sphere as a high payload and controlled delivery gene depot. J Control Release 2011;152(3):382–92. [12] Rethore G, Mathew A, Naik H, Pandit A. Preparation of chitosan/polyglutamic acid spheres based on the use of polystyrene template as a nonviral gene carrier. Tissue Eng Part C Methods 2009;15(4):605–13. [13] Browne S, Fontana G, Rodriguez BJ, Pandit A. A protective extracellular matrixbased gene delivery reservoir fabricated by electrostatic charge manipulation. Submitted for publication. [14] Collin EC, Grad S, Zeugolis DI, Vinatier CS, Clouet JR, Guicheux JJ, et al. An injectable vehicle for nucleus pulposus cell-based therapy. Biomaterials 2011; 32(11):2862–70. [15] Mathew A, Cao H, Collin E, Wang W, Pandit A. Hyperbranched PEGmethacrylate linear pDMAEMA block copolymer as an efficient non-viral gene delivery vector. Int J Pharmaceut 2012;434(1–2):99–105. [16] Millili PG, Selekman JA, Blocker KM, Johnson DA, Naik UP, Sullivan MO. Structural and functional consequences of poly(ethylene glycol) inclusion on DNA condensation for gene delivery. Microsc Res Tech 2010;73(9):866–77. [17] Mishra S, Webster P, Davis ME. PEGylation significantly affects cellular uptake and intracellular trafficking of non-viral gene delivery particles. Eur J Cell Biol 2004;83(3):97–111. [18] Tang GP, Zeng JM, Gao SJ, Ma YX, Shi L, Li Y, et al. Polyethylene glycol modified polyethylenimine for improved CNS gene transfer: effects of PEGylation extent. Biomaterials 2003;24(13):2351–62. [19] Storrie H, Mooney DJ. Sustained delivery of plasmid DNA from polymeric scaffolds for tissue engineering. Adv Drug Deliv Rev 2006;58(4):500–14. [20] Brunner T, Cohen S, Monsonego A. Silencing of proinflammatory genes targeted to peritoneal-residing macrophages using siRNA encapsulated in biodegradable microspheres. Biomaterials 2010;31(9):2627–36. [21] Cohen-Sacks H, Elazar V, Gao J, Golomb A, Adwan H, Korchov N, et al. Delivery and expression of pDNA embedded in collagen matrices. J Control Release 2004;95(2):309–20. [22] Scherer F, Schillinger U, Putz U, Stemberger A, Plank C. Nonviral vector loaded collagen sponges for sustained gene delivery in vitro and in vivo. J Gene Med 2002;4(6):634–43. [23] Duncan R, Izzo L. Dendrimer biocompatibility and toxicity. Adv Drug Deliv Rev 2005;57(15):2215–37. [24] Jensen LB, Griger J, Naeye B, Varkouhi AK, Raemdonck K, Schiffelers R, et al. Comparison of polymeric siRNA nanocarriers in a murine LPS-activated macrophage cell line: gene silencing, toxicity and off-target gene expression. Pharm Res 2012;29(3):669–82. [25] Shiratsuchi A, Ichiki M, Okamoto Y, Ueda N, Sugimoto N, Takuwa Y, et al. Inhibitory effect of N-palmitoylphosphatidylethanolamine on macrophage phagocytosis through inhibition of Rac1 and Cdc42. J Biochem 2009;145(1): 43–50. [26] Lv H, Zhang S, Wang B, Cui S, Yan J. Toxicity of cationic lipids and cationic polymers in gene delivery. J Control Release 2006;114(1):100–9.
Contents lists available at SciVerse ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Transfection of macrophages by collagen hollow spheres loaded with polyplexes: A step towards modulating inflammation q Christophe Helary ⇑, Shane Browne, Asha Mathew, Wenxin Wang, Abhay Pandit ⇑ Network of Excellence for Functional Biomaterials, National University of Ireland, Galway, Ireland
a r t i c l e
i n f o
Article history: Available online 15 June 2012 Keywords: Chronic inflammation Macrophages Collagen microspheres Polyplexes Gene delivery
a b s t r a c t Macrophages are key orchestrators of inflammation as they secrete proteases and inflammatory cytokines. To date, therapies aimed at modulating macrophage phenotype have failed due to the short half-life of biomolecules in the body. Therefore, inhibition of inflammation by gene therapy constitutes a new hope. In the present study, we have assessed collagen hollow spheres as a reservoir system for polyplexes in order to transfect human macrophages while preserving cell viability. Polyplexes were formed by complexing G-Luc plasmid with a poly(2-dimethylaminoethyl methacrylate) poly(ethylene glycol) based hyperbranched polymer. Several ratios of polymer/pDNA (5:1, 8:1, 10:1 w/w) complexes in two different sphere sizes (1.24 and 4.5 lm) were tested. Collagen hollow spheres were loaded with polyplexes up to 80 lg of pDNA per mg of microspheres. The release of polyplexes from the spheres was delayed and prolonged i.e. 20% of the initial amount released in 5 days. Following incubation with polyplex-loaded microspheres, macrophages were transfected (polyplex pDNA:polymer ratio 1:10 w/w). In addition, collagen hollow spheres maintained cell viability as more than 80% of cells were viable after 4 days in culture. In contrast, when used alone, polyplexes were seen to be toxic, while there was no transfection detected. Taken together, these results show that collagen hollow spheres may be used as a reservoir for controlled gene delivery to macrophages. Unlike existing gene delivery systems, this system allows for macrophage transfection with minimal toxicity. Hence, this system has a potential for the delivery of a therapeutic gene in order to modulate inflammation. Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Macrophages are key cells in the resolution of inflammation. During the inflammatory phase following injury, macrophages adopt a classically activated phenotype characterized by the secretion of reactive oxygen species, inflammatory cytokines and proteases [1,2]. Following exposure to biological signals such as interleukin (IL)-13, IL-4 or IL-10, macrophages progressively change their phenotype to adopt an alternatively activated phenotype. This phenotype promotes wound healing by the secretion of IL-10, VEGF and TGF-b1 to suppress the inflammatory response and promote matrix formation and stabilization [2]. In the case of many chronic inflammatory conditions, macrophages are locked in a pro-inflammatory phenotype [3].
q Part of the Special Issue ‘‘Advanced Functional Polymers in Medicine (AFPM)’’, guest editors: Professors Luigi Ambrosio, Dirk W. Grijpma and Andreas Lendlein. ⇑ Corresponding authors. E-mail addresses: [email protected] (C. Helary), abhay.pandit@ nuigalway.ie (A. Pandit).
With the aim of modulating inflammation, several trials based on the injection of growth factors have been performed. Unfortunately, owing to the short half-life and the rapid diffusion of biomolecules in vivo [4,5], these attempts have thus far proved unsuccessful. Hence, gene therapy in the form of the transfection of a gene encoding for a protein capable of modulating inflammation presents a new possibility to modulate the pro-inflammatory environment. Unfortunately macrophages, which are a non-dividing cell type, are difficult to transfect. Firstly, viral transfecting reagents cannot be used as they trigger an immune response [6]. Secondly, pDNA barely penetrates into the nucleus of non-dividing cells. Lastly, regular non-viral reagents reduce macrophage viability after transfection [7,8]. In gene therapy research, the utilization of cationic polymers has proven to have several advantages when compared to cationic lipids. Cationic polymers are more stable than lipids and complex a large amount of pDNA to form positively charged polyplexes. Unfortunately, however, they are toxic for macrophages [8]. To overcome this, polymeric hollow spheres made from natural polymers have been developed as a gene depot system [9]. These systems allow for the sustained release of polyplexes, which can
1742-7061/$ - see front matter Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.actbio.2012.06.017
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
protect macrophages from the toxicity of large, bolus doses of polyplexes. Several studies have shown that hollow spheres can be fabricated from natural polymers using a template method with sizes on the nanometer to micrometer scale possible [10]. Hollow spheres made from natural polymers have the advantage that they will not release toxic degradation products and are not detrimental to cell viability [11]. This technique relies on the coating of polystyrene beads by natural polymers such as chitosan or elastin-like peptides due to an electrostatic interaction between the negatively charged polystyrene beads and the positively charged polymers in solution. After the removal of the polystyrene bead template using the solvent tetrahydrofuran (THF), hollow spheres can be obtained. These studies demonstrate that spheres can be efficiently loaded with polyplexes, which modulates the polyplex release pattern and allows for transfection of cells in vitro [11]. In addition, sustained release of polyplexes protected cells from polymer cytotoxicity. Recently, Browne and collaborators fabricated a new type of hollow sphere made from atelocollagen type I [13]. This type of sphere is easy to fabricate and can be loaded with a large amount of polyplexes. It was hypothesized in this study that a reservoir system composed of atellocollagen type 1 can load and release polyplexes in a consistent and controlled manner, with the released polyplexes capable of transfection while protecting macrophages against polymer toxicity associated with large doses. The objective of this study was to use collagen hollow spheres as a reservoir for polyplexes to transfect human macrophages. 2. Materials and methods 2.1. Collagen hollow sphere fabrication Collagen hollow spheres of defined sizes were fabricated using the template method as previously described [13]. Briefly, 1.2 or 4.5 lm sulfonated polystyrene beads were incubated with an atelocollagenÒ solution (5 mg ml1 in 0.5 M acetic acid) for 4 h (supplementary information no. 1). After cross-linking with pentaerythritol poly(ethylene glycol) ether tetrasuccinimidyl glutarate 4 arm-StarPEG for 2 h, the polystyrene core was removed with THF, leaving an atellocollagen hollow sphere of defined size.
4209
polymers: polyethylenimine 25 kDa (PEI; Sigma, Ireland) and SuperfectÒ (Qiagen, Ireland). Linear PEI 25 kDa was used with an optimal ratio of 2:1 (polymer/G-Luc plasmid) and the partially degraded polyamido amine dendrimer (SuperfectÒ) was used with an optimal ratio of 8:1. 2.4. Analysis of polyplexes The size, zeta potential and polydispersity index of polyplexes were measured by Zetasizer (Malvern Instruments Zetasizer Nano-2590). This analysis was performed on polyplexes formed with the polymer B81 after 12 h of incubation. 2.5. Loading studies One mg of 1.2 or 4.5 lm collagen hollow spheres was resuspended in 300 ll of PBS. Polyplexes (ratio 10:1 w/w) containing a predetermined amount of pDNA (40, 80, 160 and 320 lg) were added to these spheres, followed by a volume of THF to achieve a final concentration of 50% (v/v). The mixtures were then agitated for 6 h at room temperature. After this period of incubation, the vials were opened to let the THF evaporate for 6 h. Next, the suspensions were centrifuged at 13,000g and the sphere/polyplex complexes were washed four times with ultrapure water. The supernatants collected were used to estimate the amount of free pDNA. A PicoGreenÒ assay (Invitrogen, Ireland) was performed to quantify free pDNA. To obtain free pDNA, polyplex samples collected during loading were treated with polyglutamic acid (PGA) of 10 mg ml1 concentration at 37 °C for 30 min with a modification of the protocol described in Ref. [11]. Then 100 ll of PicoGreenÒ was added to each sample of similar volume and the fluorescence was analyzed at 480 nm. A standard curve was prepared using naked pDNA. First the amount of pDNA detected in the supernatants was determined, then the loading efficiency of the hollow spheres was determined using the formula: Loading efficiency ¼
Initial pDNA added pDNA remaining in supernatant Initial pDNA added
2.6. Release pattern of polyplexes from collagen hollow spheres 2.2. Characterization by scanning electron microscopy (SEM) Prior to analysis, collagen hollow spheres were fixed with 4% paraformaldehyde. A drop of sample containing collagen hollow spheres was placed on adhesive carbon tabs mounted on SEM specimen stubs and then dried. The specimens were subsequently coated with gold using an Emitech K550 coating system. SEM images were obtained using a Hitachi S-4700 field emission microscope operating with a beam voltage of 15 kV. 2.3. Polyplex formation Polyplexes were formed by complexing the synthesized polymer (B81) with G-Luc plasmid encoding for the protein Gaussia Luciferase (New England Biolabs, Ireland). The transfecting agent B81 (14 kDa) synthesized in our group is composed of a linear poly(2dimethylaminoethyl methacrylate) block (pDMAEMA), a hyperbranched poly(ethylene glycol methyl ether methacrylate) block and an ethylene dimethacrylate block (see structure in supplementary information no. 2). The polymer B81 was synthesized via the deactivation enhanced atom transfer radical polymerization technique. Polyplexes were formed in phosphate-buffered saline (PBS) with weight ratios from 5:1 to 10:1 (Polymer B81/G-Luc plasmid) for 1 h. To assess the transfection efficiency with the polymer B81, we compared the results with those of the commercial
To characterize the release pattern of the polyplex-loaded collagen hollow spheres, 1 mg of polyplex-loaded spheres of 1.2 or 4.5 lm size was resuspended in PBS and incubated at 37 °C. The spheres used for this experiment were loaded with 40 lg of pDNA per mg of spheres. At each time point, the suspension was centrifuged and a sample of the supernatant taken. The released polyplexes were treated with polyglutamic acid (Sigma, Ireland) and quantified using the PicoGreenÒ (Invitrogen, Ireland) assay. To determine the release pattern, five time points were considered: days 1, 2, 3, 4 and 5. The results were expressed as a percentage of initial amounts of pDNA loaded in the spheres. 2.7. Macrophage differentiation and activation The human myelogenous leukemia cell line THP-1 was obtained from the ATCC. The cells were maintained in RPMI 1640 (GIBCOBRL) supplemented with 10% fetal bovine serum (FBS), penicillin (100 U ml1) and streptomycin (100 lg ml1) at 37 °C in a 5% CO2 humidified incubator. The mature macrophage-like state was induced by treating THP-1 cells for 24 h with phorbol 12-myristate 13 acetate (PMA) at 100 ng ml1 diluted in serum-free medium. Cells were seeded at 100,000 cells well1 into a 48-well plastic well plate. The following day, plastic-adherent cells were washed twice with cold, sterile Dulbecco’s PBS and incubated with fresh
4210
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
RPMI-1640 lacking PMA but containing 1% FBS, penicillin (100 U ml1) and streptomycin (100 lg ml1). The cells were then activated with 100 ng ml1 lipopolysaccharide in RPMI-1640 medium containing 1% FBS (supplementary information no. 3). 2.8. Collagen hollow sphere uptake by activated THP-1 THP-1 cells were differentiated and activated as previously described. First, 300 lg of FITC-labelled microspheres of three sizes was added to 250,000 cells and incubated for 24 h. The cells were then rinsed with Hanks’ balanced salt solution three times to remove any spheres that were not internalized. A cell scraper was used to detach the cells from the surface, and these were then fixed with 4% PFA and examined by flow cytometry (BD FACS Canto 6) for internalization. The three sizes of microspheres examined were 100 nm, 1 lm and 10 lm. 2.9. Transfection studies Activated macrophages-like cells were incubated with polyplex-loaded collagen hollow spheres for 4 days. Two sizes of spheres were tested: 1.24 and 4.5 lm. Cells without treatment were kept as controls. For the tests, 100,000 cells were incubated with 100 lg of loaded spheres. Different loadings of spheres were tested: 40 and 80 lg of pDNA per mg of spheres. Regarding the polyplexes formed with polymer B81, three different weight ratios were tested: 5:1, 8:1 and 10:1. In parallel, the same experiment was carried out with PEI and partially degraded polyamido amine dendrimer (SuperfectÒ) at their optimal ratios. All experiments were performed in 1% FBS medium. For comparison, macrophages were incubated with polyplexes alone in FBS-free medium for 4 h. Then, polyplexes were removed and replaced by fresh medium (1% FBS). After the 96 h incubation period, 50 ll of medium was collected from each of the samples for quantification of luciferase expression using a Gaussia luciferase assay kit (New England Biolabs). The luciferase assay was performed in accordance with the manufacturer’s instructions. The luciferase expression level was expressed as relative light units. 2.10. Cell metabolic activity AlamarBlueÒ assay (Sigma, Ireland) was performed as previously described to quantify the cellular metabolic activity of macrophages incubated with polyplex-loaded spheres [11]. The viability was assessed as a percentage of the metabolic activity of control cells. 2.11. Statistical analysis All experiments were replicated in triplicate and the results expressed as the mean value ± standard deviation. Statistical significance was determined using Student’s test, with p < 0.05 considered significant. Statistical analysis was carried out between control samples and macrophages incubated with collagen hollow spheres loaded with polyplexes. 3. Results and discussion 3.1. Structure of collagen hollow spheres Following fabrication using the template method, the structure of collagen hollow spheres was analysed by SEM. This analysis revealed the presence of isolated collagen hollow spheres of 1.24 and 4.5 lm sizes. The microsphere populations were monodispersed and each sphere was of the same size as the template used
Fig. 1. Structure of collagen hollow spheres. The analysis by SEM revealed the presence of 1.24 (A) or 4.5 lm collagen microspheres (B). In addition, collagen hollow spheres appeared to be monodispersed and isolated.
(Fig. 1). Collagen hollow sphere fabrication was first described by Browne et al. in order to generate a new gene depot system [13]. The collagen hollow spheres were produced using the electrostatic properties of atelocollagen on a polystyrene template. In the context of macrophages and inflammation, the utilization of collagen hollow spheres has several advantages when compared to artificial materials. Atelocollagen is collagen I without the telopeptides at the N- and C- terminals, which enables much of its antigenicity to be removed by enzymatic cleavage. As these telopeptides are responsible for the mild immunogenicity of collagen, atelocollagen does not trigger inflammation when applied in an organism either of the same or different species. Collagen hollow spheres were cross-linked with 1 mM 4 arm-StarPEG. The aim of this step was to stabilize and tune the physical properties and the porosity of collagen microspheres [12]. It has been shown that 4 arm-StarPEG is suitable for use as a cross-linker in collagen-based tissue engineering systems as it does not affect cell viability [14]. Hence, collagen hollow spheres can be considered as a desirable system to transfect macrophages and modulate inflammation because they are fabricated from a material that does not trigger a host response. 3.2. Structure and polydispersity of G-Luc pDNA/polymer B81 polyplexes The size of polyplexes formed at a 10:1 ratio (w/w) was around 50–80 nm, with a zeta potential of 4 ± 2 mV [15]. Polyplexes formed at ratios of 5:1 and 8:1 (polymer/pDNA) had weaker charges (data not shown). To interact with cells and promote their endocytosis, polyplexes need to have a positive charge. Therefore, the ratio of 10:1 is the most suitable for cell transfection in these conditions. The polydispersity index of polyplexes, measured by Zetasizer, was 0.911 after 12 h of incubation in PBS. This result shows that polyplexes do not aggregate during this period. Hence, cell transfection was performed by isolated polyplexes. It has been shown that PEGylation prevents aggregation of polyplexes [16– 18]. 3.3. Maximum loading of polyplexes within collagen hollow spheres The loading study was carried out with polyplexes formed by the G-Luc plasmid and the polymer B81 (n = 6 for each condition), at various weight ratios of polyplexes to collagen microspheres. With a ratio of 1:320 (mg of spheres per lg of pDNA), the polyplexes precipitated when they were incubated for 24 h with collagen hollow spheres. The other ratios tested did not precipitate. The maximal loading of polyplexes was about 80 lg of pDNA per mg of collagen hollow spheres when a ratio of 1:160 (mg of spheres per lg of pDNA) was used for the loading study (Fig. 2A). The efficiency of loading with this ratio was 50% whereas it was almost 100% with the lowest ratio of 1:40. The loading of collagen hollow spheres by polyplexes was similar to that obtained with 500 nm
4211
A
Loading (µg pDNA G-Luc)
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
faster, reaching 70% after 48 h [20]. The polyplex release assessed from macroporous scaffolds such as collagen and fibrin hydrogels exhibits the same features of release as artificial microspheres, i.e. a burst of polyplexes during the first 48 h and a slow, low release thereafter [21]. In contrast, collagen sponges allow delayed delivery, with only 20% of polyplexes being released after 50 days [22]. Using dendrimers (SuperfectÒ) or PEI, the release of polyplexes from collagen spheres was faster. Indeed, 40% of the initial dose was delivered after 24 h and 60% after 3 days (data not shown) [13].
200
(n = 6) 150 100 50 0
B
DNA Release (% of initial amount)
1: 40 1: 80 1: 160 Ratios (mg microspheres/ µg pDNA)
100
Using the polymer B81 with the ratio of 5:1, no transfection was detected in macrophages, with no toxic effects noted (Fig. 3E). Indeed, macrophages retained their spindle shape, and their viability corresponded to 80% of that detected in samples incubated without polyplexes (Fig. 3A and E). From the 8:1 ratio, polymer toxicity in macrophages was observed, but with no detectable levels of transfection. Cells were less numerous than in control samples.
1,2 µm Spheres
80
4,5 µm Spheres
60 (n = 6)
40
3.5. Transfection of macrophages by polyplexes alone
20 0 0
1
2
3
4
5
A
B
C
D
Time (Days)
500
120
(n=6)
100
400
80 300 60 200 40 100
20 0
ct
8:
1
:1 I2 Su p
er
fe
PE
10 :
1
tio ra
tio ra
tio
5:
8:
1
l ro
1
0
ra
The release of loaded polyplexes used at a weight ratio of 10:1 (polymer B81/G-Luc pDNA) was delayed and prolonged. Moreover, the release pattern observed with 1.2 lm spheres was similar to that observed with 4.5 lm microspheres (n = 6 for each sphere size). Regardless of the microsphere size, the amount of pDNA released into the medium increased gradually over time to reach about 20% of the initial amount of pDNA after 5 days (Fig. 2B). As the maximal loading was 80 lg of pDNA per mg of sphere, 16 lg of plasmid was released after 5 days. In the transfection experiments, only 100 lg of microspheres was added to macrophages and two different loadings were tested. After 4 days at 37 °C, 0.64 and 1.28 lg of pDNA were released from microspheres loaded with 40 and 80 lg of pDNA per mg, respectively. In order to modulate inflammation, it is an advantage to have a system that allows for a slow rate and low level of release of polyplexes. The low rate of release prevents the potential toxicity of polymer [19], while the slowness permits a prolonged effect over 25 days. By comparison, the polyplex release from PLA microspheres is
Luciferase Activity (RLU)
3.4. Release pattern of polyplexes from collagen hollow spheres
E
Co nt
elastin-like polypeptide hollow spheres [11]. With this type of microsphere, the maximal loading was about 70 lg of pDNA per mg of spheres. These results showed the absence of relation between the sphere size and the loading, in that there was no significant difference between the loading efficiency of either size of microsphere.
Viability (% of Control)
Fig. 2. Loading and release profile of polyplexes at a ratio of 10:1 (polymer/pDNA) encapsulated within collagen hollow spheres. (A) Polyplex loading behaviour of collagen hollow spheres. Regardless of the sphere size, the maximal loading was around 90 lg of pDNA per mg of collagen hollow spheres. This loading was observed with the ratio of 1 mg of spheres per 160 lg of pDNA. Polyplexes were formed with polymer B81 and G-Luc plasmid using the 10:1 ratio (polymer/pDNA). (B) Release pattern of polyplexes from collagen hollow spheres. Regardless of the sphere size, the release of polyplexes was low and slow to reach about 20% of the initial pDNA amount after 5 days.
Polymer B81 Fig. 3. Transfection of macrophages by polyplexes. Using a ratio of 5:1, polyplexes formed from polymer B81 were not toxic but no transfection was detected (A and E). From the ratio of 8:1, polyplexes were toxic for cells while they did not permit their transfection (B and E). The commercial polymer PEI used with a ratio of 2:1 (C and E) and the partially degraded polyamido amine dendrimer (SuperfectÒ) used with a ratio of 8:1 (D and E) were also cytotoxic for macrophages. In addition, no transfection was observed. Histograms show luciferase activity and squares represent cell viability.
4212
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
40
30
20
10
10 um
1u m
10 0n m
0
C on tro l
Uptake of FITC-labelled Collagen microspheres (%)
Moreover, macrophages were no longer spindle shaped, but rather rounded with small extensions. As assessed by AlamarBlueÒ cell metabolic assay, cell viability represented about 50% of that observed in control macrophages (Fig. 3B and E). With the ratio of 10:1, no transfection was observed and only 40% of cells were viable. In this study, two commercial polymers were also tested in order to compare their efficacy with polymer B81. The observation by optical microscopy showed that PEI and partially degraded polyamido amine dendrimer (SuperfectÒ) were toxic for macrophages (Fig. 3C and D). As measured by AlamarBlueÒ, only 12 and 50% of cells were viable after 4 h of incubation with PEI and partially degraded polyamido amine dendrimer (SuperfectÒ), respectively (Fig. 3E). Furthermore, no transfection was detected in macrophages incubated with polyplexes formed from these two commercial polymers. PEI and dendrimers (SuperfectÒ) are known to be toxic to nondividing cells. It is thought that the presence of primary amines in their structure is responsible for this toxicity [23]. Nevertheless, this toxicity is not observed on the macrophage cell line RAW 264.7, as these macrophage-like cells can proliferate in culture. This ability to multiply makes transfection easier, as cells which divide are easier to transfect, and it also masks the toxic effect of these polymers [24]. However, macrophages do not proliferate in vivo. Hence, THP-1 cell line differentiated into macrophage-like cells is a model closer to physiological conditions as they do not multiply in culture. The utilization of THP-1 cells in these experiments displayed the toxicity of commercial polymers. In order to minimize polymer toxicity, a PEGylated polymer, B81 was tested. PEGylation shields the positive charge of the molecule and reduces toxicity [23]. In addition, the tertiary amine of the pDAEMA should improve the biocompatibility in comparison with transfecting agents with primary amines [11]. Unfortunately, these modifications were not sufficient to protect cells from polymer toxicity, as toxicity was seen with B81 despite no significant levels of transfection being observed.
Sphere Size Fig. 4. Phagocytosis of collagen hollow spheres by macrophages. The analysis by flow cytometry showed the engulfment of 100 nm collagen spheres. In contrast, 1 and 10 lm spheres were taken up less by macrophages.
3.6. Engulfment of collagen hollow spheres by macrophages The analysis by flow cytometry revealed that only 7 ± 2% of macrophages had engulfed 1 lm collagen hollow spheres after 24 h of incubation (Fig. 4). This figure dropped to 3 ± 1% when macrophages were incubated with bigger spheres (10 lm). In contrast, the engulfment of 100 nm spheres was observed in 30% of macrophages. Hence, collagen hollow spheres showed little engulfment by macrophages when their size was bigger than 1 lm. An internalization of spheres would cause high cell toxicity due to the high polyplex concentration. Therefore, this result provides evidence that collagen hollow spheres can be considered as a potential reservoir system to transfect macrophages and modulate inflammation. The mechanism of action relies on the diffusion of polyplexes from the collagen spheres which transfect macrophages after release. 3.7. Transfection of macrophages by collagen hollow spheres loaded with polyplexes No transfection was observed with the 5:1 and 8:1 ratios, regardless of the loading and the microsphere size used (Fig. 5A– C). With the 10:1 ratio, a statistically significant transfection was detected in macrophages (Fig. 5B). The luciferase activity detected in macrophages incubated with 40 and 80 lg of pDNA per mg of 4.5 lm spheres was 5 and 10 times higher than that in control samples, respectively (p < 0.05). The transfection of macrophages was not associated with polymer toxicity as more than 80% of macrophages were viable after 4 days of incubation with collagen hollow spheres. The results obtained with 1.2 lm collagen hollow spheres were similar except for the samples loaded with 40 lg of pDNA. With the ratio of 10:1, the luciferase activity was 10 times higher than the control cells (Fig. 5C). Therefore, the low, slow release of polyplexes allowed for the transfection while preventing polymer toxicity. In vitro, THP-1 cells are difficult to transfect. The classic way to transfect them was achieved by a cationic lipid, oligofectamineÒ [25]. The utilization of this transfecting reagent has several drawbacks, most notably toxicity. For example, it can inhibit the enzyme protein kinase C, thereby disturbing cell metabolism [26]. Another possibility based on nucleofection permits transfection of these cells, but only prior to them being differentiated into macrophages [8]. Polyplexes formed with PEI or partially degraded polyamido amine dendrimer (SuperfectÒ) were used in this study in order to compare with the transfection capability of polymer B81. Collagen hollow spheres were loaded with polyplexes made from these commercial polymers. Two loadings were tested: 40 and 80 lg of pDNA per mg of spheres. A transfection, similar to that obtained with B81, was observed with PEI (ratio 2:1), when loaded with 80 lg of pDNA per mg of spheres (p < 0.05). Unfortunately, this transfection was associated with high toxicity as only 40% of macrophages were viable after 4 days (Fig. 6). With the 40 lg loading, no transfection was detected while the cell toxicity remained the same as 80 lg of pDNA. In contrast, no toxicity and no transfection were detected when macrophages were incubated with collagen hollow spheres loaded with polyplexes made with partially degraded polyamido amine dendrimer (SuperfectÒ). This result adds further evidence to suggest that collagen hollow spheres loaded with polyplexes made from polymer B81 are suitable to transfect macrophages. 4. Conclusions These results show that collagen hollow spheres can be considered as a potent reservoir for controlled gene delivery. Unlike bolus
4213
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
A
*
100
1500 80 60
1000
*
40
500
(n=6)
:1
1
10
8:
1
tio
tio
ra
ra
ra
ra
tio
10
5:
:1
1 8:
tio
tio
5: ra
ra
C
tio
on
tro
1
l
0
4.5 µm spheres loaded with 80 µg pDNA/mg spheres
4.5 µm spheres loaded with 40 µg pDNA/mg spheres
Luciferase Activity (RLU)
4.51 µm Collgen Spheres
20 0
C
Viability (% of Control)
120
2000
2000
120
*
1500
100 80
*
1000
60 40
500
Viability (% of Control)
Luciferase Activity (RLU)
B
1.24 µm Collgen Spheres (n=6)
20
:1
1
10 tio
ra
tio
8:
1 ra
tio
5:
:1
4.5 µm spheres loaded with 40 µg pDNA/mg spheres
ra
tio ra
tio
10
8:
1 5: ra
tio ra
nt ro Co
1
0 l
0
4.5 µm spheres loaded with 80 µg pDNA/mg spheres
Fig. 5. Transfection of macrophages by collagen hollow spheres loaded with polyplexes. (A) Microscopic view of THP-1 like macrophages incubated for 4 days with 4.5 lm collagen hollow spheres. (B) Transfection efficacy of 4.5 lm collagen hollow spheres loaded with polyplexes (polymer B81/G-Luc plasmid) and their effect on cell viability. Using a ratio of 10:1 (polymer/pDNA), luciferase activity was detected, evidencing macrophage transfection (n = 6, p < 0.05). The transfection efficacy obtained with the loading 80 lg pDNA per mg of spheres was twice as high as that obtained with 40 lg. In addition, macrophage transfection was associated with high cell viability as more than 80% of macrophages were viable. (C) Transfection efficacy of 1.24 lm collagen hollow spheres loaded with polyplexes and their effect on cell viability. The results were similar to those obtained with 4.5 lm spheres, i.e. macrophage transfection detected with the ratio 10:1 (polymer/pDNA) associated with high viability (n = 6). Histograms show luciferase activity and lozenges represent cell viability.
delivery of existing polymers, this reservoir system allows for transfection of macrophages without toxicity. Hence, this system appears promising for the delivery of an inhibitory inflammatory therapeutic gene capable of modulating macrophage phenotype.
With this reservoir system, the transfection of a therapeutic gene such as IL-10, IL-12 or IL-4 could potentially trigger the polarization of inflammatory macrophages (M1) to regulatory macrophages (M2), which can promote tissue repair.
4214
C. Helary et al. / Acta Biomaterialia 8 (2012) 4208–4214
4.5 µm Collagen Hollow Spheres (n=6) 120
*
100 1500 80 1000
60 40
500 20 0
Viability (% of Control)
Luciferase Activity (RLU)
2000
0 Control
40 µg pDNA
80 µg pDNA
PEI 2:1
40 µg pDNA
80 µg pDNA
Superfect 8:1
Fig. 6. Transfection of macrophages by 4.5 lm collagen hollow spheres loaded with polyplexes formed from commercial polymers. No transfection was detected with polyplexes formed from partially degraded polyamido amine dendrimer (SuperfectÒ) (ratio 8:1). In contrast, Luciferase activity was detected when PEI (ratio 2:1) was used and loaded into collagen hollow spheres (with 80 lg pDNA per mg of spheres). Unfortunately, the macrophage transfection was associated with cell toxicity (n = 6). Histograms show luciferase activity and lozenges represent cell viability.
Acknowledgements The authors thank Estelle Collin and Mohammad Abu-Rub for their discussion on this article. In addition, we thank Antony Sloan for editorial support while writing this article. Health Research Board (HRB RP/2008/188) for financial support of this work. Appendix A. Figures with essential colour discrimination Certain figures in this article, particularly Figs. 3 and 5, are difficult to interpret in black and white. The full colour images can be found in the on-line version, at http://dx.doi.org/10.1016/j.actbio. 2012.06.017. Appendix B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2012. 06.017. References [1] Mosser DM. The many faces of macrophage activation. J Leukoc Biol 2003; 73(2):209–12.
[2] Gordon S, Martinez FO. Alternative activation of macrophages: mechanism and functions. Immunity 2010;32(5):593–604. [3] Eming SA, Krieg T, Davidson JM. Inflammation in wound repair: molecular and cellular mechanisms. J Invest Dermatol 2007;127(3):514–25. [4] Eming SA, Kaufmann J, Lohrer R, Krieg T. Chronic wounds. Novel approaches in research and therapy. Hautarzt 2007;58(11):939–44. [5] Menke NB, Ward KR, Witten TM, Bonchev DG, Diegelmann RF. Impaired wound healing. Clin Dermatol 2007;25(1):19–25. [6] Thomas CE, Ehrhardt A, Kay MA. Progress and problems with the use of viral vectors for gene therapy. Nat Rev Genet 2003;4(5):346–58. [7] Lee SY, Cherla RP, Tesh VL. Simultaneous induction of apoptotic and survival signaling pathways in macrophage-like THP-1 cells by Shiga toxin 1. Infect Immun 2007;75(3):1291–302. [8] Schnoor M, Buers I, Sietmann A, Brodde MF, Hofnagel O, Robenek H, et al. Efficient non-viral transfection of THP-1 cells. J Immunol Methods 2009; 344(2):109–15. [9] Rethore G, Pandit A. Use of templates to fabricate nanoscale spherical structures for defined architectural control. Small 2010;6(4):488–98. [10] Dash BC, Rethore G, Monaghan M, Fitzgerald K, Gallagher W, Pandit A. The influence of size and charge of chitosan/polyglutamic acid hollow spheres on cellular internalization, viability and blood compatibility. Biomaterials 2010;31(32):8188–97. [11] Dash BC, Mahor S, Carroll O, Mathew A, Wang W, Woodhouse KA, et al. Tunable elastin-like polypeptide hollow sphere as a high payload and controlled delivery gene depot. J Control Release 2011;152(3):382–92. [12] Rethore G, Mathew A, Naik H, Pandit A. Preparation of chitosan/polyglutamic acid spheres based on the use of polystyrene template as a nonviral gene carrier. Tissue Eng Part C Methods 2009;15(4):605–13. [13] Browne S, Fontana G, Rodriguez BJ, Pandit A. A protective extracellular matrixbased gene delivery reservoir fabricated by electrostatic charge manipulation. Submitted for publication. [14] Collin EC, Grad S, Zeugolis DI, Vinatier CS, Clouet JR, Guicheux JJ, et al. An injectable vehicle for nucleus pulposus cell-based therapy. Biomaterials 2011; 32(11):2862–70. [15] Mathew A, Cao H, Collin E, Wang W, Pandit A. Hyperbranched PEGmethacrylate linear pDMAEMA block copolymer as an efficient non-viral gene delivery vector. Int J Pharmaceut 2012;434(1–2):99–105. [16] Millili PG, Selekman JA, Blocker KM, Johnson DA, Naik UP, Sullivan MO. Structural and functional consequences of poly(ethylene glycol) inclusion on DNA condensation for gene delivery. Microsc Res Tech 2010;73(9):866–77. [17] Mishra S, Webster P, Davis ME. PEGylation significantly affects cellular uptake and intracellular trafficking of non-viral gene delivery particles. Eur J Cell Biol 2004;83(3):97–111. [18] Tang GP, Zeng JM, Gao SJ, Ma YX, Shi L, Li Y, et al. Polyethylene glycol modified polyethylenimine for improved CNS gene transfer: effects of PEGylation extent. Biomaterials 2003;24(13):2351–62. [19] Storrie H, Mooney DJ. Sustained delivery of plasmid DNA from polymeric scaffolds for tissue engineering. Adv Drug Deliv Rev 2006;58(4):500–14. [20] Brunner T, Cohen S, Monsonego A. Silencing of proinflammatory genes targeted to peritoneal-residing macrophages using siRNA encapsulated in biodegradable microspheres. Biomaterials 2010;31(9):2627–36. [21] Cohen-Sacks H, Elazar V, Gao J, Golomb A, Adwan H, Korchov N, et al. Delivery and expression of pDNA embedded in collagen matrices. J Control Release 2004;95(2):309–20. [22] Scherer F, Schillinger U, Putz U, Stemberger A, Plank C. Nonviral vector loaded collagen sponges for sustained gene delivery in vitro and in vivo. J Gene Med 2002;4(6):634–43. [23] Duncan R, Izzo L. Dendrimer biocompatibility and toxicity. Adv Drug Deliv Rev 2005;57(15):2215–37. [24] Jensen LB, Griger J, Naeye B, Varkouhi AK, Raemdonck K, Schiffelers R, et al. Comparison of polymeric siRNA nanocarriers in a murine LPS-activated macrophage cell line: gene silencing, toxicity and off-target gene expression. Pharm Res 2012;29(3):669–82. [25] Shiratsuchi A, Ichiki M, Okamoto Y, Ueda N, Sugimoto N, Takuwa Y, et al. Inhibitory effect of N-palmitoylphosphatidylethanolamine on macrophage phagocytosis through inhibition of Rac1 and Cdc42. J Biochem 2009;145(1): 43–50. [26] Lv H, Zhang S, Wang B, Cui S, Yan J. Toxicity of cationic lipids and cationic polymers in gene delivery. J Control Release 2006;114(1):100–9.