Facile preparation of multifunctional superparamagnetic PHBV microspheres containing SPIONs for biomedical applications

8/19/2019 Facile preparation of multifunctional superparamagnetic PHBV microspheres containing SPIONs for biomedical appli…

1/12


2/12

www.nature.com/scientificreports/

2SCIENTIFIC REPORTS | 6:23140 | DOI: 10.1038/srep23140

and partly complementary characteristics. Among these methods, the emulsion-solvent extraction/evaporation

method possesses significant competitive advantages including good reproducibility and high degree o controlover particle characteristics such as particle size, which make it one o the most popular methods to producemicrospheres20,21.

Iron oxide nanoparticles, the only Food and Drug Administration (FDA)- and European Medicines Agency(EMA)-approved metal oxide nanoparticles, have attracted tremendous attention in targeted drug delivery, MRIand hyperthermia therapy 13. Te surace hydrophilicity o iron oxide nanoparticles enables them to be efficientlyencapsulated by hydrophilic polymers, which in most cases are natural derived polymers3,11,18. However, thehydrophilic iron oxide nanoparticles strongly tend to partition into the external aqueous phase during emulsifica-tion, which leads to great loss o iron oxide nanoparticles or even ailure in preparing hydrophobic polymer-basedmagnetic microspheres2,15. Most o the hydrophobic polymers are synthetic derived. Compared to natural pol-ymers, synthetic polymers oen offer better control o physicochemical properties22. Obviously, a good stabilityo iron oxide nanoparticles in hydrophobic polymer solutions is a prerequisite to the successul abrication andexpanding application o magnetic polymer microspheres. Surace modification with atty acids has been shownto improve the stability o iron oxide nanoparticles in dichloromethane (DCM)23, which is a typical organic sol-

vent used to dissolve hydrophobic polymers.

In the present study, superparamagnetic iron oxide nanoparticles (SPIONs) were surace modified with lauricacid, which belongs to the amily o atty acid. It was hypothesized that lauric acid could enable the SPIONs to bestable in hydrophobic polymer solution, and thereore acilitate the successul preparation o magnetic polymermicrospheres with high encapsulation efficiency/loading efficiency using emulsion-solvent extraction/evapora-tion method. Lauric acid was used in the present study also because the derived lauric acid-modified SPIONshave great potential or magnetic drug targeting, as shown in a previous study 14,24. Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), which belongs to the class o polyhydroxyalkanoates (PHAs), is a promising bio-technology derived hydrophobic polymer used or biomedical applications due to its biocompatibility, nontoxicand tailorable biodegradability 25,26. In addition, unlike poly(lactic acid) (PLA) and poly(lactic-co-glycolic acid)(PLGA), PHBV does not produce acidic degradation products which may be harmul or human tissues, and itdegrades into its monomers in vivo which are normal in the blood27.

In the first part o this investigation, SPIONs were surace modified with lauric acid, and then encapsulatedinto PHBV microspheres using solid-in-oil-in-water (S/O/W) emulsion-solvent extraction/evaporation method.Te physicochemical properties and in vitro biocompatibility o the prepared magnetic PHBV microspheres wereinvestigated. It should be pointed out that although various types o magnetic polymer microspheres have been

developed1–3,8,11,15–17,19, only ew in vitro or in vivo studies have been reported on these systems1–4. Tus, there isa strong demand to understand the biocompatibility o magnetic polymer microspheres, and the influence thatpolymer encapsulation may have on the toxicity o SPIONs. PHAs based-magnetic particles (rom nm to mm sizerange) have only been reported recently 28,29, and their biocompatibility has not been investigated yet to the besto our knowledge. Furthermore, in the present study, tetracycline hydrochloride (CH), a widely used antibiotic,was chosen as a model o water-soluble drugs to demonstrate the drug delivery capability and to investigate thedrug release behavior o the magnetic PHBV microspheres in vitro.

Results and DiscussionSurface modication on SPIONs0. Te Fourier transorm inrared spectroscopy (FIR) spectra o thetwo investigated iron oxide nanoparticles SPIONs0 (without any coating) and SPIONsLA (with lauric acid coat-ing) are shown in Fig. 1. Te characteristic band o Fe–O stretching vibration was observed at 584 cm−1 or bothSPIONs0 and SPIONsLA16,24. All the labelled bands o SPIONsLA except 584 cm−1, 1541 cm−1 and 1575 cm−1 belongto lauric acid24,30. In addition, compared to the other bands o (ree) lauric acid, the band at 1708 cm−1 whichbelongs to carboxylic acid groups was relatively weak in SPIONsLA, and new bands at 1541 cm−1 and 1575 cm−1,

Figure 1. FTIR spectra o SPIONs beore (SPIONs0) and afer (SPIONsLA) surace modification with lauricacid. (Te different relevant peaks are explained in the text).

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


3/12



which were attributed to carboxylate groups, appeared in SPIONsLA31,32. Te appearance o carboxylate groupsas well as weakened carboxylic groups indicates that chemical bonds ormed between lauric acid and iron oxidenanoparticles30,32. Tereore, it is confirmed that lauric acid has been successully coated on SPIONs0.

SPIONsLA are quite homogeneous as shown by the transmission electron microscopy (EM) image (Fig. 2(a)).Te diameter o the iron oxide cores was 11± 1 nm, which is much smaller than the superparamagnetic criticalsize (20 nm) o Fe3O4 nanoparticles

33. At a higher magnification (Fig. 2(b)), lattice planes o individual particlesare visible. Te small and homogeneous particle size o SPIONsLA is expected to enhance their stability in hydro-phobic polymer solution.

It is o great importance to increase the stability o SPIONs (Solid (S) phase) in PHBV–DCM solution (Oil(O) phase), because a stable S/O solution can reduce the loss o S phase during emulsification, which is a pre-requisite to obtain a high encapsulation efficiency. In addition, a stable S/O droplet acilitates the productiono more spherical shaped particles aer solidification. As shown in Fig. 3(a), SPIONs0 sink to the bottom o theDCM, while SPIONsLA already partly disperse in the DCM even without sonication. Aer sonication (Fig. 3(b)),

SPIONsLA

are still stably present in the DCM, however most o the SPIONs0

were located in the water/DCMinterace or inside the water. Tis result confirms that hydrophilic SPIONs0 turn hydrophobic aer surace modi-fication with lauric acid, which thereore enables the SPIONsLA to be stable in DCM, with the primary monolayero lauric acid acting as the stabilizing layer34.

Composition of magnetic PHBV microspheres. Te main composition o the pure PHBV microspheres(MS) and magnetic PHBV microspheres loaded with different content o SPIONsLA (MMSL (low content) andMMSH (high content)) was investigated by X-ray diffraction (XRD) analysis (Fig. 4(a)). All the samples showedcharacteristic peaks o PHBV at 13.5°, 16.9°, 19.9°, 21.6°, 22.8°, 25.6°, 27.2° and 30.8°35. Both MMSL and MMSHshowed the characteristic peaks o Fe3O4 at 18.2°, 30.2°, 35.6°, 43.3°, 53.7°, 57.2° and 63.1°

36, and the intensities othese peaks increased when the amount o loaded SPIONsLA was increased. Te XRD patterns confirm that Fe3O4 is successully encapsulated in the PHBV microspheres.

Te thermogravimetric behaviors o SPIONsLA and microspheres are shown in Fig. 4(b). Te weight loss oSPIONsLA was mainly due to the decomposition o lauric acid. Compared to SPIONsLA, all the microspheres

Figure 2. TEM images o SPIONsLA at different magnifications (a,b). Te inset in (a) shows the particle sizedistribution o iron oxide cores.

Figure 3. Photos o SPIONs0 and SPIONsLA in water/DCM: (a) just afer adding DCM and water, (b) 2 minafer sonication.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


4/12



presented much higher weight loss due to the decomposition o PHBV. Te thermogravimetric analysis (GA)curves o the microspheres containing more SPIONsLA shied towards higher temperature. Tis may be due tothe increased thermal stability o PHBV-based nanocomposites in the presence o more nanoparticles35.

Te loading efficiency o SPIONsLA in MMSL and MMSH was calculated to be 12.5 wt% and 20.3 wt%, and theencapsulation efficiency was 87.4% and 71.0%, respectively. Te density o Fe3O4 (~5 g/cm

3) is much higher thanthat o PHBV (~1.26 g/cm3) and DCM (~1.33 g/cm3). Tereore, more SPIONsLA loaded in the PHBV–DCM (i.e.,higher loading efficiency) may increase the instability o the emulsion which eventually leads to more SPIONsLA been lost, i.e., lower encapsulation efficiency. SPIONs have been reported to be encapsulated in various polymermicrospheres, while their encapsulation efficiency, as an important evaluation index o magnetic microsphere,

has been reported in only ew studies1,2,4,15,17,19,37. Te encapsulation efficiency in the present study is similar tothat (estimated to be ~65− 84%) o PLGA/PLA-PEG magnetic microspheres produced by O/W single emul-sion method using hydrophobic magnetite gel as the magnetic component 19, and is significantly higher thanthat (1.8− 4.5%) o hydrophilic errofluid loaded PLGA microspheres produced by conventional W/O/W doubleemulsion method2,4. In addition, the encapsulation efficiency achieved in this study is higher than that (60%)o hydrophilic F3O4 nanoparticles loaded PLA microspheres produced by a delicate W/O/W double emulsionmethod15.

Surface morphology and internal structure. Te shape and surace morphology o the microspheres, asobserved by scanning electron microscopy (SEM), are shown in Fig. 5. Both MMSL and MMSH appeared spheri-cal in shape, and the microspheres had quite homogeneous particle size. Many SPIONs LA were seen to be embed-ded into the surace o the microspheres. Tis phenomenon may be explained by the Pickering emulsion theory,specifically, iron oxide nanoparticles existed at the O/W interace to assist the suractant to orm a stable O/Wemulsion11,29. Te amount o loaded SPIONsLA seems to have no obvious influence on the surace morphology othe obtained microspheres. Furthermore, in order to reveal the internal structure, a cross-section near the center

Figure 4. (a) XRD patterns and (b) GA o MS (pure PHBV microspheres) and MMSL, MMSH (magneticPHBV microspheres). GA o SPIONsLA was perormed or comparison.

Figure 5. SEM images o magnetic PHBV microspheres at different magnifications: surace morphology o(a,b) MMSL and (c,d) MMSH, and internal microstructure o (e, ) MMSH.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


5/12



o MMSH was obtained using ocused ion beam (FIB) milling. As shown in Fig. 5(e,), SPIONsLA are observedon the cross-section as well as on the wall o the inner pores o the microspheres. Te pores inside the micro-spheres are due to the ast skin ormation during the solvent extraction, which restricts the urther shrinkage/densification o the microspheres during solidification20. Te surace morphology and internal structure provesthat SPIONsLA are distributed over the whole matrix o microspheres, which indicates that SPIONs LA are stable inPHBV–DCM solution and are efficiently encapsulated in the PHBV microspheres.

Particle size and zeta potential. he median particle size o both MMSL and MMSH was 2.46 μm(Fig. 6(a)). Te span (width o particle size distribution) o MMSL and MMSH was 1.37 and 1.62, respectively.Different amounts o encapsulated SPIONsLA seem to have no obvious influence on the mean particle size o

prepared microspheres, however higher amount o SPIONsLA slightly increased the microsphere size distributionwidth. Te wider particle size distribution o MMSH can be attributed to the increased instability o the emulsioncaused by loading o more SPIONsLA. Fig. 6(b) presents the zeta potential o SPIONsLA (− 19.9 mV) and micro-spheres in water. Te zeta potential o MS was− 33.6 mV, and it increased to − 28.1 mV or MMSL and urtherincreased to− 22.8 mV or MMSH when more SPIONsLA were encapsulated. Tis phenomenon is due to the actthat the incorporated SPIONsLA possess a higher zeta potential than MS. Te relatively high negative zeta poten-tial o these microspheres is beneficial or their dispersion and stability in aqueous medium.

Magnetic properties. Fig. 7(a) shows the magnetization curves measured at room temperature orSPIONsLA, MMSL and MMSH. Te saturation magnetization data obtained at high external field strength (2 )are an approximation o the theoretical value o saturation magnetization. Te magnetization curves, speciallyor MMSL and MMSH are only slightly increasing at 2 , which means that the approximation is quite close tothe theoretical value. Te saturation magnetization values o MMSL and MMSH were 5.6 emu/g and 11.0 emu/g,respectively, and it was 56.8 emu/g or SPIONsLA. Tis behavior is more consistent with the saturation magneti-zation o bulk maghemite (around 60 emu/g) than with the saturation magnetization o bulk magnetite (around

Figure 6. (a) particle size distribution o MMSL and MMSH, (b) zeta potential o SPIONsLA, MS, MMSL andMMSH in water.

Figure 7. (a) magnetization curves measured at room temperature or SPIONsLA, MMSL and MMSH. Te inset

shows the response o MMSL and MMSH to a magnetic field. (b) T 2-weighted MR images o H2O, 0.1 wt% agarsolution and 0.25–2.0 mg/mL o MS, MMSL and MMSH dispersed in 0.1 wt% agar solution.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


6/12



90 emu/g)38. However, it has been shown earlier that dispersions o SPIONs are prone to oxidation during storage,which reduces the saturation magnetization significantly over time, e.g. by up to 28.9%39. Although the decreasein saturation magnetization could also derive rom the smaller magnetic domain size o SPIONs compared tobulk material, it has been shown earlier that this is not the case or SPIONsLA24. Tereore it is likely that theparticles had oxidized to maghemite already. Te inset in Fig. 7(a) demonstrates the dispersion and separationprocess o the magnetic microspheres. Both MMSL and MMSH were well attractable with a magnet. Tereore,the prepared magnetic PHBV microspheres possessed magnet-induced sensitivity, and this ‘proo o concept’experiment indicates that they could be utilized as magnetic targeting carriers in biomedical applications.

Te prepared magnetic microspheres were qualitatively evaluated or their magnetic contrast capability underMRI scanning. T 2-weighted MR images (Fig. 7(b)) show that MMSL and MMSH produced significant nega-tive contrast effects (signal reductions) in comparison to the control groups, i.e., H2O, agar solution and MS in0.25–2.0 mg/mL. With increasing concentration o microspheres in agar solution, there was a correspondingincrease in contrast or MMSL and MMSH. Furthermore, the contrast or MMSH was higher than MMSL at asame concentration due to the higher loading efficiency o SPIONsLA in MMSH. Tese results show that, althoughSPIONsLA are encapsulated in PHBV microspheres, they can stil l be used or magnetic imaging applications.

Drug release and antibacterial property. One o the purposes o this study was to determine i themagnetic PHBV microspheres could serve as potential vehicles to successully deliver drugs. Fig. 8(a,b) shows the

drug release profiles o SPIONsLA-CH

(CH loaded SPIONsLA

) loaded PHBV microspheres in PBS. Both MMSLand MMSH exhibited a sustained release o CH or up to 14 days. Cumulative amounts o 101 μg and 168 μgCH were released rom MMSL and MMSH, respectively. Te higher amount o CH loading in MMSH couldexplain why it showed a slightly higher initial burst release ollowed by a relatively aster release at the initial stagecompared to MMSL. Te mass ratio o CH in MMSH over that in MMSL is 1.66, which was quite close to themass ratio (1.62) o the CH carrier, i.e., SPIONsLA in MMSH over that in MMSL as determined by GA. Teadsorptive affinities o metal oxide nanoparticles or CH are likely due to several mechanisms, including elec-trostatic attraction, surace complexation and cation exchange40, while the exact mechanism in the present studyis yet to be investigated. Phosphate ions exhibit high binding efficiency to the surace o iron oxides 41, and canthus replace other substances such as drugs which are located on the surace o the SPIONs by the aorementionedmechanism. Te act that some, but not all o the drug is released via burst release could indicate a barrier effecto the polymer towards the drug loaded on iron oxide particles.

PHBV is hydrophobic42, and generally degrades quite slowly in vitro27. Tereore, the drug release rom PHBVmicrospheres is likely to be more dependent on diffusion rather than on polymer degradation26,43,44. Te drugrelease data was fitted to Higuchi equation45, and the results are shown in Fig. 8(c). Te fitting or both MMSL and

Figure 8. (a) Cumulative amount and (b) cumulative percentage o CH release rom MMSL and MMSH,(c) linear fitting o the cumulative percentage o CH release versus the square root o time (Higuchi equation),and (d) antibacterial activity against Escherichia coli (E. coli) o diluted CH solution collected at 1 h, 1 d, 3 dand 5 d.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


7/12



MMSH shows good linearity. Tus, CH release rom the magnetic PHBV microspheres can be well described byHiguchi equation, indicating a diffusion controlled release mechanism. In addition, MMSH had a higher kineticconstant (k) than MMSL, which confirms that the CH release rom MMSH was aster than rom MMSL in thefirst 24 hours (as shown in Fig. 8(b)).

In order to better differentiate the antibacterial property o CH solution collected at 1 h, 1 d, 3 d and 5 dduring drug release, the solutions were diluted in different ratios. Repetition o the antibacterial test is needed orperorming statistical analysis in urther study, while the bacterial growth or both MMSL and MMSH (Fig. 8(d))exhibits an overall decreasing trend when the drug release time increases rom 1 h to 5 d, indicating increasingCH concentration. Moreover, at a given time point, MMSH in general showed less bacterial growth than MMSL,which confirms MMSH released more CH than MMSL, as shown in Fig. 8(a). Hence, the results o the prelimi-nary antibacterial tests are in good agreement with the results o drug release study and confirm the drug deliverycapability o the prepared magnetic PHBV microspheres.

In vitro cytotoxicity. Jurkat cells (suspension cells) and H-29 colon carcinoma cells (adherent cells)were used to investigate the biocompatibility o PHBV microspheres (without CH) in various concentrations.Aer 24 h and 48 h o incubation, cells were stained and analyzed in flow cytometry. Staining with the DNAdye Hoechst and differences in size enabled the clear discrimination between cells and microspheres. Usingthe cyanine dye DiIC1(5) (1,1′ ,3,3,3′ ,3′ -hexamethylindodicarbocyanine iodide), we stained cells with an active

mitochondrial membrane potential (MMP) as a measure or intact cell metabolism. Decrease o DiIC1(5) stainintensity indicates a disrupted mitochondrial membrane potential. Data were confirmed by AnnexinA5-FIC /propidium iodide staining (not shown), which detects exposure o phosphatidylserine on the outer side o theplasma membrane o apoptotic cells and disrupted plasma membranes o necrotic cells, respectively. Investigatingthe side scatter (SSC) o viable cells as mean o cellular granularity indicates a concentration dependent cellularuptake o the microspheres aer 24 h, whereas aer 48h, the SSC values have equaled or all tested concentrationso each microsphere indicating a saturation effect (Fig. 9(a)). Te cellular prolieration was dose dependentlydecreased by the presence o microspheres in Jurkat cells (Fig. 9(b)). In contrast, cell viability, reflected by intactMMP (Fig. 9(c)) and DNA integrity (Fig. 9(d)), were only slightly influenced in Jurkat cells by the microspheres,indicating their biocompatibility even in high concentrations and prolonged incubation times. AnnexinA5-FIC/propidium iodide staining confirmed these results (data not shown). In addition, due to their relatively largeparticle size, microspheres gradually sedimented, thus at the bottom o the cell culture wells locally even highermicrosphere concentrations can be assumed. Tereore, microsphere concentrations are given in both μg/mLand μg/cm2.

Figure 9. Biocompatibility o MS, MMSL and MMSH on Jurkat T cells. Experiments were perormed intriplicates. Results are given as mean± standard deviations (Te asterisks indicate significant differences inStudent’s t -test compared to the controls *P


8/12



H-29 colon carcinoma cells (adherent cells) were incubated with 0.5 mg/mL (56 μg/cm2) or 2 mg/mL (224 μg/cm2)microspheres (Fig. 10). Aer 24 h o incubation, transmission microscope images (Fig. 10(a,b)) showed thatadherent cells were still tightly attached to the plastic suraces o the wells and looked phenotypically healthy evenin the presence o high microsphere concentrations (Fig. 10(b)). For more detailed investigation, flow cytometryo H-29 cells was perormed aer trypsin detachment and staining with DiIC1(5) (Fig. 10(c)). As shown by mor-phological analysis and investigation o mitochondrial membrane potential, incubation with 0.5 mg/mL (56 μg/cm2) microspheres induced only a slight loss in cell viability (Fig. 10(d)).

Altogether, cell viability in the presence o MMSL, MMSH as well as MS was better than ree SPIONsLA at acomparable concentration as shown in the related previous study 24, thereore we conclude that encapsulation oSPIONsLA in PHBV microspheres improved their biocompatibility. Te biocompatibility o these magnetic PHBVmicrospheres ensures their potential or use in biomedical applications such as drug delivery and MRI.

ConclusionsSPIONs aer surace modification with lauric acid became much more stable in DCM solution, which thereoreallowed the preparation o magnetic SPION-containing PHBV microspheres with high encapsulation efficiency(71.0–87.4%) using emulsion-solvent extraction/evaporation method. Te obtained microspheres appearedspherical in shape, and had a median particle size o 2.46 μm. SPIONsLA were observed throughout the matrixo microspheres. Loading efficiencies o the microspheres were determined to be 12.5–20.3 wt%, and the cor-responding magnetization was 5.6–11.0 emu/g. Te zeta potential o the microspheres increased but it was stillnegative with increasing content o encapsulated SPIONsLA. Te developed magnetic PHBV microspheres werereadily detectable by MRI, and increasing contrast was observed or the microspheres loaded with higher contento SPIONsLA. Te magnetic PHBV microspheres were shown to be biocompatible towards both Jurkat cells (sus-pension cells) and H-29 cells (adherent cells) even at high concentrations. CH, as a model drug, was success-ully loaded into the microspheres, and released in a diffusion controlled manner. Te developed magnetic PHBVmicrospheres have potential applications in targeted drug delivery and MRI.

Figure 10. Biocompatibility o MS, MMSL and MMSH on adherent HT-29 cells. Experiments wereperormed in triplicates, shown are the mean values with standard deviations. H-29 cells were seeded intocell culture wells (6 well plates; growth area 8.9 cm2, 1 mL medium), grown or 24 hours and incubated withmicrospheres or 24 h. (a,b) ransmission microscope images (magnification 10× ) o H-29 cells incubatedwith (a) 0.5 mg/mL microspheres or (b) 2 mg/mL microspheres or 24 h. (c) Flow cytometry raw data file(Forward scatter versus side scatter) depicts possibility to discriminate between microspheres (red), viable(green) and dying/dead cells (blue) by morphological parameters; (d) analysis o cell viability by morphologyand mitochondrial membrane potential. Viability o control cells was set to 100%; experiment was perormed intriplicates; shown are the mean values with standard deviations.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


9/12



Materials and MethodsMaterials. PHBV with a PHV content o 12 wt% was purchased rom Goodellow (Huntingdon, UK).Polyvinyl alcohol (PVA) (MW ~30,000), Ringer´s solution and Iron (II) chloride tetrahydrate were rom Merck(Darmstadt, Germany). DCM, propidium iodide (PI), sodium citrate, triton X-100, acetone and lauric acid wereobtained rom Sigma-Aldrich (St. Louis, USA). Ammonia solution 25% and Iron (III) chloride hexahydrate weresupplied by Roth (Karlsruhe, Germany). CH was purchased rom AppliChem (Darmstadt, Germany). RPMI1640 medium, etal cal serum (FCS), Annexin A5-FIC (Ax), Hoechst 33342 (Hoechst) and DiIC 1 (5) wereacquired rom Termo Fisher Scientific (Waltham, USA).

Preparation of SPIONs and lauric acid-coated SPIONs. Uncoated SPIONs (SPIONs0) and lauricacid-coated SPIONs (SPIONsLA) were synthesized as described in detail elsewhere25. In brie, 1.998 g o iron(II) chloride tetrahydrate and 5.406 g o iron (III) chloride hexahydrate were dissolved in 20 mL o ultrapurewater under Ar atmosphere and heated to 80 °C. 16.5 mL ammonia solution (25%) were added to orm a blackishprecipitate o SPIONs0. o obtain SPIONsLA, 1.28 g o lauric acid dissolved in acetone were added aer that andthe suspension was heated to 90 °C or another 15 minutes. Both particle suspensions were purified by dialysisusing a molecular weight cut off o 8000 Da and the pH was adjusted to 9.2 using ammonia solution. Tey werelyophilized and stored at 4 °C until urther use.

Characterization of SPIONsLA. FTIR measurement . Te chemical composition o the lyophilized nan-oparticles was determined using FIR (Nicolet 6700, Termo Scientific, USA). Te sample was mixed with KBr(spectroscopy grade, Merck, Germany) and pressed into pel lets. Spectra were recorded in absorbance modebetween 4000 and 400 cm−1 with a resolution o 4 cm−1.

Stability of SPIONsLA. 5 mg SPIONs0 and SPIONsLA were placed in a glass vial, respectively. 2 mL DCM wereadded into the glass vial ollowed by adding 4 mL water. Photos were taken immediately aer adding the water.Aerwards, the mixture was homogenized by a probe sonicator (Branson Sonifier® S-250D, Emerson, USA) at10% power output or 40 s. Photos were taken 2 min aer the sonication. 2 min was chosen here because this is theemulsification time or preparing PHBV microspheres in the ollowing study.

TEM. EM images were taken with a CM 300 UltraWIN (Philips, Eindhoven, Netherlands) operated at anacceleration voltage o 300 kV. Samples were prepared by drying 10 μL o diluted nanoparticle suspension on acarbon-coated Athene S147-2 copper grid (Plano, Wetzlar, Germany). Particle size distribution o the iron oxidecores was determined using Image J (NIH, USA), and more than 100 particles were analyzed.

Preparation of magnetic PHBV microspheres. A solid-in-oil-in-water (S/O/W) emulsion-solventextraction/evaporation method was used to prepare the magnetic PHBV microspheres by modiying the methodused in our previous study 43. Briefly, SPIONsLA (S phase) were added into 3 mL o 3% w/v PHBV–DCM solution(O phase). wo mass ratios o SPIONsLA to PHBV were chosen, i.e., 1:6 (MMSL) and 1:2.5 (MMSH). Te mix-ture was homogenized by a probe sonicator at 10% power output or 40 s. Ten, the homogeneous mixture wasadded into 75 mL o 2% w/v PVA solution (W phase) and emulsified using a homogenizer (18, IKA, Germany)at 10500 rpm or 2 min. Aerwards, the obtained emulsion was added into 225 mL o 1% w/v PVA solution. Tefinal solution was stirred at 600 rpm or 2 h by an over-head stirrer (Eurostar 20, IKA, Germany) to evaporate theorganic solvent. Te microspheres were collected by centriugation (Centriuge 5430R, Eppendor, Germany) at5000 rpm or 4 min, washed in deionized water, and lyophilized (Alpha 1-2 LDplus, Martin Christ, Germany).Te microspheres were stored in a desiccator until urther use. Pure PHBV microspheres were prepared ollowingthe procedure described above, except that no SPIONsLA were added into the PHBV solution.

Characterization of magnetic PHBV microspheres. Surface morphology and internal microstructure.Surace morphology was observed using SEM (LEO 435 VP, Cambridge, UK and Ultra Plus, Zeiss, Germany).Dried microspheres were mounted onto stubs using carbon tapes. No metallic coating was applied on themicrospheres or SEM observation. Te cross-sections o single microsphere were obtained using a FIB (HeliosNanoLab™ 660, FEI, Hillsboro, USA). Te microspheres were sputter coated first with 200 nm thick carbon,then with 2 μm thick platinum. A Ga+ beam o 30 kV between 40 and 90 pA was used to cut the microspheres.Once the samples had been polished to the desired cross-sections, SEM images were taken in secondary electronimaging mode.

Particle size and zeta potential. Particle size and zeta potential o the microspheres were analyzed by Mastersizerparticle size analyzer 2000 and Malvern Zetasizer Nano ZS (Malvern, Worcestershire, UK), respectively. For thesemeasurements, microspheres were suspended in deionized water. In addition to the median particle size (D50),the width o particle size distribution (span) was also calculated. Te span is calculated by (D90− D10)/D50, whereDi means i% o the particle distribution lies below the size Di.

Encapsulation efficiency and loading efficiency. Te encapsulation efficiency and loading efficiency o SPIONsLA in magnetic PHBV microspheres were determined by GA (Q5000, A Instruments, USA). Te samples wereheated rom 25 °C to 550 °C under nitrogen flow. Te heating rate was 10 °C/min. Te loading efficiency andencapsulation efficiency were calculated by equations (1) and (2), respectively:

= ×Loading efficiency Mass of SPIONs in microspheres/Mass of microspheres 100% (1)LA

http://-/?-http://-/?-http://-/?-http://-/?-


10/12



=

×

Encapsulation efficiency Mass of SPIONs in microspheres/Mass of SPIONs

used for preparing microspheres 100% (2)

LA LA

Magnetic properties (SQUID and MRI). Te magnetic properties o the samples in solid orm (lyophilizedpowder) were measured using a superconducting quantum intererence device (SQUID)-based susceptometer(QD-MPMS-XL-5, Quantum Design Inc, San Diego, USA) in an applied magnetic field o± 2 at room temper-ature. Te sample holder used or the measurement was gelatin capsule. Diamagnetic contribution o the sampleholder was subtracted. Te obtained magnetization is given in emu per gram (o total material). MRI studies wereconducted on a 3 Siemens rio-im scanner (Siemens, Erlangen, Germany). Imaging phantoms were preparedby dispersing microspheres in 0.1 wt% agar solution at 0.25–2.0 mg/mL.

XRD analysis. Te main composition o the microspheres was characterized using XRD (Bruker D8 ADVANCEDiffractometer, Cu K α) at room temperature. Data were collected over the 2θ range rom 10° to 70° using a stepsize o 0.014° with 1 second per step.

In vitro drug release study. Drug loading . SPIONsLA were soaked in 3 mg/mL CH–deionized watersolution or 24 h. Ten, the CH loaded SPIONsLA (SPIONsLA-CH) were collected by centriugation, washed indeionized water, and lyophilized. Te loading degree o CH on SPIONsLA was 8.8 wt%. SPIONsLA-CH loadedPHBV microspheres were prepared ollowing the same procedure as SPIONsLA loaded PHBV microspheres. Tedrug loading and drug release process were kept off light.

Drug release. 10 mg MMSL and MMSH were dispersed in 1 mL PBS solution (pH 7.4) and placed in a dialysis

bag (Spectra/Por® 1, molecular weight cut-off 6000 to 8000, Carl Roth, Germany), and then immersed in tubescontaining 4 mL PBS solution (pH 7.4). Te samples were placed in a shaking incubator (90 rpm, 37 °C). Aerpredetermined time points, 1 mL medium was collected and replaced with 1 mL resh PBS solution. CH concen-tration was measured using a UV–Vis spectrophotometer (Specord® 40, Analytik Jena, Germany) at a wavelengtho 362 nm. A calibration curve was made with CH concentration in the range 2.5–100 μg/mL, and the obtainedcalibration curve ollows Beer’s law: A = 0.0294 × C , where A is the absorbance and C is the concentration oCH. Te cumulative release (%) o CH was calculated. Te drug release study was carried out or 408 h andperormed in triplicate.

Drug release kinetics. Higuchi model was used to analyze the CH release kinetics rom the magnetic PHBVmicrospheres, which is represented by equation: Qt = kt

1/2, where Qt is the cumulative percentage o drug release,k is the kinetic constant and t is the release time45. Tis equation is generally valid or the first 60% o total amounto drug release.

Antibacterial test. 4 mg MMSL and MMSH were placed in a dialysis bag, respectively, and then immersed in2 mL PBS solution. 100 μL released medium was collected at 1 h, 1 d, 3 d and 5 d, and diluted in 10 mM sterilizedsodium phosphate buffer or antibacterial test. CH concentration in the released medium was measured using aUV–Vis spectrophotometer. Antibacterial activities o the diluted medium were evaluated using the 96-well plateassay 46. E. coli ACC 25922 was purchased rom Leibniz Institute DSMZ-German Collection o Microorganismsand Cell Cultures (Braunschweig, Germany). E. coli were inoculated into resh nutrient broth (Roth, Karlsruhe,Germany) and incubated at 37 °C or 8 h with shaking to reach the logarithmic phase o growth at a density o1–5× 107 cu/mL. Ten, these bacterial suspensions were diluted to a density o 1–2× 105 cu/mL. Individualwells o a sterilized 96-well microtitre plate were inoculated with 50 μL o samples, i.e., diluted medium and 50 μLo bacterial suspension. Control wells contained all the components except the samples, which were replaced by50 μL ampicillin solution o 10 μg/mL (positive control) or 50 μL sterilized sodium phosphate buffer o 10 mM(negative control). Te 96-well plates were incubated overnight at 37 °C with shaking. Te experiment was con-ducted in duplicates and the optical density at 600 nm was spectrophotometrically recorded aer the incubationby using a temperature-controlled automatic plate reader (ecan, Männedor, Switzerland). Percentage o bacte-rial growth (P bacterial growth) was determined by

− − × .( ) ( )OD OD OD OD/ 100%s ample pos it ive cont rol nega ti ve cont rol pos it ive con trol

In vitro biocompatibility assays. Te non-adherent human -cell leukemia cell line Jurkat (ACC 282,DSMZ, Braunschweig, Germany) was cultured in RPMI 1640 medium supplemented with 10% FCS. Adherentcolon carcinoma H-29 cells (ACC/LGC GmbH, Wesel, Germany) were cultured in McCoy’s 5A medium(Gibco®, Lie echnologies GmbH, Darmstadt, Germany) supplemented with 10% FCS. All cells were cultivatedunder standard cell culture conditions in a humidified incubator (INCOmed, Schwabach, Memmert, Germany)at 37 °C and 5% CO2.

Prior to these experiments, viability and count o cells were determined using a Muse cell analyzer (MerckMillipore, Darmstadt, Germany). Only i the viability exceeded 90% the experiments were continued. Te micro-spheres were sterilized under UV light or 2 h.

Jurkat cells were adjusted to 2× 105 cells/mL in medium. 1 mL o this suspension, containing the respectiveconcentration o microspheres (without CH), was seeded into 48 well plates (Greiner Bioone, Frickenhausen,

http://-/?-http://-/?-http://-/?-http://-/?-


11/12



Germany). Aer 24 h and 48 h, 50 μL aliquots o the cells and microspheres containing suspension were taken and

stained or cell viability analysis.3.4× 105 H-29 cells were seeded into each well o a 6 well plate (PP, Schafausen, Switzerland) in medium,grown or 24 h and incubated in the presence o microspheres or urther 24 h. Ten, transmission optical micros-copy was perormed using an Axiovert 40 CFL Microscope (Zeiss, Jena, Germany) with a 10x objective withphase contrast. Ten, cells were detached with 0.25% trypsin/0.02% EDA in PBS (PAN Biotech, Aidenbach,Germany) rom the wells and resuspended to orm single-cell suspensions.

Since microspheres gradually sediment to the bottom o the cell culture wells, indication o microsphere con-centration in μg/mL and μg/cm2 is required (see able 1).

Analyses or cell viability were then perormed according to a protocol rom Munoz et al.47. For staining,250 μL o the staining solution (containing 20 μg/mL PI, 0.5 μg/mL Annexin A5-FIC, 10 nM DiI and 1 μg/mLHoechst 33342 in Ringer´s solution) were added to cell aliquots and incubated or 20 min at 4 °C. Te suspensionswere then analyzed with a Gallios flow cytometer (Beckman Coulter, Fullerton, USA) or 1 min. Bleed throughfluorescence was eliminated by electronic compensation. Forward scatter (FSC) and side scatter (SSC) wereconsidered or morphological analysis o cells. Cell cycle and DNA degradation were examined by propidiumiodide-triton (PI) staining according to the protocol o Nicoletti et al.48. Briefly, 400 μL o a solution containing

0.1% sodium citrate, 0.1% riton X-100, and 50 μg/mL PI was added to 50 μL aliquot o cells, incubated overnightat 4 °C in the dark, and nuclear fluorescence was measured in flow cytometry. Data analysis was perormed withthe soware Kaluza 1.2.

Statistical analysis. Statistical significance analysis was perormed using Student’s t -test in Microso Excel2010 (Microso, Redmond, USA). A value o P


12/12


23. Liu, X. et al. Preparation and characterization o hydrophobic superparamagnetic magnetite gel. J. Magn. Magn. Mater. 306, 248–253(2006).

24. Zaloga, J. et al. Development o a lauric acid/albumin hybrid iron oxide nanoparticle system with improved biocompatibility. Int. J.Nanomedicine 9, 4847–4866 (2014).

25. Hazer, D. B., ılıçay, E. & Hazer, B. Poly(3-hydroxyalanoate)s: Diversification and biomedical applications: A state o the art review. Mater. Sci. Eng. C Mater. Biol. Appl . 32, 637–647 (2012).

26. Li, W., Ding, Y., Yu, S., Yao, Q. & Boccaccini, A. . Multiunctional Chitosan-45S5 Bioactive Glass-Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) Microsphere C omposite Membranes or Guided issue/Bone egeneration. ACS Appl. Mater. Interfaces 7,20845–20854 (2015).

27. öse, G. ., enar, H., Hasırcı, N. & Hasırcı, V. Macroporous poly(3-hydroxybutyrate-co-3-hydroxyvalerate) matrices or bonetissue engineering. Biomaterials 24, 1949–1958 (2003).

28. Vilos, C. et al. Superparamagnetic Poly (3-hydroxybutyrate-co-3 hydroxyvalerate) (PHBV) nanoparticles or biomedicalapplications. Electron. J. Biotechnol. 16(5), doi: 10.2225/vol16-issue5-ulltext-8 (2013).29. Oa, C., Ushimaru, ., Horiishi, N., suge, . & itamoto, Y. Core–shell composite particles composed o biodegradable polymer

particles and magnetic iron oxide nanoparticles or targeted drug delivery. J. Magn. Magn. Mater. 381, 278–284 (2015).30. Wang, P.-C., Chiu, W.-Y., Lee, C.-F. & Young, .-H. Synthesis o iron oxide/poly(methyl methacrylate) composite latex particles:

Nucleation mechanism and morphology. J. Polym. Sci., Part A: Polym. Chem. 42, 5695–5705 (2004).31. Lu, Y. & Miller, J. D. Carboxyl Stretching Vibrations o Spontaneously Adsorbed and LB-ranserred Calcium Carboxylates as

Determined by FI Internal eflection Spectroscopy. J. Colloid. Interface. Sci. 256, 41–52 (2002).32. ataby, G., Cojocaru, M., Prozorov, . & Gedanen, A. Coating Carboxylic Acids on Amorphous Iron Nanoparticles.Langmuir 15,

1703–1708 (1999).33. Dunlop, D. J. Superparamagnetic and single-domain threshold sizes in magnetite. J. Geophys. es. 78, 1780–1793 (1973).34. Shen, L., Laibinis, P. E. & Hatton, . A. Bilayer Suractant Stabilized Magnetic Fluids: Synthesis and Interactions at Interaces.

Langmuir 15, 447–453 (1999).35. Yu, H. et al. einorcement o biodegradable poly(3-hydroxybutyrate-co-3-hydroxyvalerate) with cellulose nanocrystal/silver

nanohybrids as biunctional nanofillers. J. Mater. Chem. B 2, 8479–8489 (2014).36. Unterweger, H. et al. Development and characterization o magnetic iron oxide nanoparticles with a cisplatin-bearing polymer

coating or targeted drug delivery. Int. J. Nanomedicine 9, 3659–3676 (2014).37. Wamocha, H. L. et al. Cytotoxicity o release products rom magnetic nanocomposites in targeted drug delivery. J. Biomater. Appl.

27, 661–667 (2013).38. Cornell, . & Schwertmann, U. Te Iron Oxides: Structure, Properties, eactions, Occurrences and Uses, p. 117 (VCH, 1996).39. Zaloga, J.et al. Different storage conditions influence biocompatibility and physicochemical properties o iron oxide nanoparticles.

Int. J Mol. Sci. 16, 9368 (2015).40. Duan, L., Li, L., Xu, Z. & Chen, W. Adsorption o tetracycline to nano-NiO: the effect o co-existing Cu(ii) ions and environmental

implications. Environ. Sci. Process Impacts 16, 1462–1468 (2014).41. Weng, L., Van iemsdij, W. H. & Hiemstra, . Factors controlling phosphate interaction with iron oxides. J. Environ. Qual. 41,

628–635 (2012).42. Li, W., Nooeaid, P., oether, J. A., Schubert, D. W. & Boccaccini, A. . Preparation and characterization o vancomycin releasing

PHBV coated 45S5 Bioglass®-based glass–ceramic scaffolds or bone tissue engineering. J. Eur. Ceram. Soc. 34, 505–514 (2014).43. Li, W. et al. Preparation and characterization o PHBV microsphere/45S5 bioactive glass composite scaffolds with vancomycin

releasing unction. Mater. Sci. Eng. C Mater. Biol. Appl . 41, 320–328 (2014).44. Scheithauer, E. C. et al. Preparation and characterization o electrosprayed daidzein-loaded PHBV microspheres. Mater. Lett. 158,

66–69 (2015).45. Higuchi, . Mechanism o sustained-action medication. Teoretical analysis o rate o release o solid drugs dispersed in solid

matrices. J. Pharm. Sci. 52, 1145–1149 (1963).46. Pritchard, S. ., Phillips, M. & ailasapathy, . Identification o bioactive peptides in commercial Cheddar cheese. Food es. Int. 43,

1545–1548 (2010).47. Munoz, L. E. et al. Colourul death: Six-parameter classification o cell death by flow cytometry—Dead cells tell tales. Autoimmunity 46, 336–341 (2013).

48. Nicoletti, I., Migliorati, G., Pagliacci, M., Grignani, F. & iccardi, C. A rapid and simple method or measuring thymocyte apoptosisby propidium iodide staining and flow cytometry. J. Immunol. Methods 139, 271–279 (1991).

AcknowledgementsW.L. would like to acknowledge the China Scholarship Council. J.Z., C.J. and C.A. are grateul or financialsupport rom the Bavarian State Ministry o the Environment and Consumer Protection, the DFG (SPP1681)and the EU project FP7-NMP-2012-LARGE-6-309820 “NanoArthero”. he authors thank Dr. Klaus Gieb(Department o Physics, FAU), Christel Dieker, Dr. Qingqing Yao, Qiang Chen, Dr. Sandra Cabanas-Polo, Dr.Julia Will, Alina Gruenewald and Kai Zheng or experimental support.

Author ContributionsW.L. conceived the idea. W.L., J.Z., Y.L. and C.J. designed the experiments. W.L., J.Z., Y.D., Y.L. and C.J. perormedthe experiments. M.P., C.A. and A.R.B. supervised the research. W.L., J.Z., Y.L. and C.J. wrote the manuscript. All

authors discussed the results and reviewed the manuscript.

Additional InformationCompeting financial interests: Te authors declare no competing financial interests.

How to cite this article: Li, W. et al. Facile preparation o multiunctional superparamagnetic PHBVmicrospheres containing SPIONs or biomedical applications. Sci. Rep. 6, 23140; doi: 10.1038/srep23140 (2016).

Tis work is licensed under a Creative Commons Attribution 4.0 International License. Te imagesor other third party material in this article are included in the article’s Creative Commons license,

unless indicated otherwise in the credit line; i the material is not included under the Creative Commons license,users will need to obtain permission rom the license holder to reproduce the material. o view a copy o thislicense, visit http://creativecommons.org/licenses/by/4.0/
http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/

Documents

Facile preparation of multifunctional superparamagnetic PHBV microspheres containing SPIONs for biomedical applications