Antibacterial Nanoparticles with Natural Photosensitizers Extracted from Spinach Leaves.

We prepared antibacterial polystyrene nanoparticles (NPs) with natural photosensitizers from chlorophyll (Chl) extract via a simple nanoprecipitation method using the same solvent for dissolution of the polystyrene matrix and extraction of Chls from spinach leaves. A high photo-oxidation and antibacterial effect was demonstrated on Escherichia coli and was based on the photogeneration of singlet oxygen O2(1Δg), which was directly monitored by NIR luminescence measurements and indirectly verified using a chemical trap. The photoactivity of NPs was triggered by visible light, with enhanced red absorption by Chls. To reduce the quenching effect of carotenoids (β-carotene, lutein, etc.) in the Chl extract, diluted and/or preirradiated samples, in which the photo-oxidized carotenoids lose their quenching effect, were used for preparation of the NPs. For enhanced photo-oxidation and antibacterial effects, a sulfonated polystyrene matrix was used for preparation of a stable dispersion of sulfonated NPs, with the quenching effect of carotenoids being suppressed.

Introduction

Chlorophylls (Chls) are widely abundant photosensitizers in nature that enable conversion of solar energy to biochemically useful forms.[1,2] They are very cheap, commercially available, main group metal-based, and provide excellent efficiency for singlet oxygen-mediated chemistry.[3] The photosensitization process from the S1 states after excitation with red and/or blue radiation competes with a very efficient intersystem crossing to the T1 state and the formation of highly reactive singlet oxygen [O2(1Δg)]. In general, O2(1Δg) is generated during photosynthesis in the photosystem II center.[4,5] It is a signal molecule, but at high concentrations, it can oxidize and destroy target structures and can be applied for photodynamic inactivation of bacteria and viruses[6,7] and for photodynamic therapy to treat tumors.[8,9]

Chl extract from spinach (ChlE) and other plants is highly promising for applications in reactions photosensitized by O2(1Δg). A recent paper[10] reported quantum yield of singlet oxygen ΦΔ = 0.58 for a Chl-based extract, similar to the value of ΦΔ = 0.51 for Chl a. Note that ΦΔ values for Chl-based compounds in different solvents were summarized in previous comprehensive reviews published more than 20 years ago.[11,12] Application of the extract is limited by the presence of carotenoids that dramatically decrease photo-oxidation efficiency,[13] which is a harmful process for natural or artificial photosynthesis.[14,15] Both Chl triplet states and O2(1Δg) can be quenched by carotenoids via energy transfer (physical quenching). Additionally, carotenoids (carotenes and xanthophylls)[16] quench O2(1Δg) in a chemical way to form endoperoxides as a major oxidation product by cleavage of double bonds of the polyene chain, resulting in a variety of aldehydes with different chain lengths.[17] Carotenoids also have a positive role; they absorb visible light in the blue-green region, in which Chls have a low extinction coefficient, and transfer excitation energy to Chls.[18] Another undesirable property of the Chl extract from green plants is its degradation upon irradiation with light. Thus, improving the Chl photostability is necessary to facilitate its biomedical application.[19] Heavy metal-substituted Chls can be more stable than natural magnesium Chls[20] but exhibit a lower O2(1Δg) production efficiency than Mg or Zn derivatives.[21]

Isolation of pure photosensitizers from ChlE requires complicated purification methods and toxic solvents to isolate individual photosensitizers. Chl s can also be used as starting materials for the synthesis of functionalized chlorins[22] to improve photodynamic activity. Previous studies[10,23] indicate that ChlE itself can be an excellent “green” photosensitizer for environmental applications, even in the presence of carotenoids that are diluted in solutions in comparison with their level in plants.

There are plenty of organic and inorganic nanocarriers under investigation for transport of photosensitizers, some of them have good photocatalytic and/or antibacterial properties.[24] In our previous studies, we prepared different nanofiber materials[25] and nanoparticles (NPs)[26] with a porphyrin photosensitizer, which generated O2(1Δg) “on demand” after excitation with visible light and exhibited antibacterial properties. In general, NPs with a high surface/volume ratio photogenerate a high amount of O2(1Δg) and can move in close proximity to the biological targets to be destroyed, which can overcome the O2(1Δg) diffusion limitations. The encapsulated photosensitizer is well protected against external quenchers, which allows quenching of triplet states by oxygen, exclusively in the interior of NPs. These antibacterial NPs are considered an alternative to antibiotics and have strong potential to solve the problem of bacterial multidrug resistance.[27]

Among the variety of nanocarriers suitable for photodynamic inactivation, polystyrene NPs are superior due to their biocompatibility, low toxicity, high oxygen permeability/diffusion coefficient, and negligible leakage of nonpolar encapsulated photosensitizes to aqueous media.[28]

This report describes a simple method for preparation of an aqueous dispersion of polystyrene NPs with encapsulated ChlE from spinach (Spinacia oleracea). In addition to Chl, the extract also contained carotenoids, which have a negative influence on photoactive NPs with respect to their photo-oxidation, antibacterial activity, and photostability. We used three simple methods to suppress the quenching effect of carotenoids: working with diluted ChlE (i), using ChlE preirradiated with light to photo-oxidize the carotenoids (ii), and/or using small sulfonated polystyrene NPs with a higher surface/volume ratio to suppress the effect of carotenoids inside the NPs (iii). The main advantages of this approach in the synthesis of photoactive NPs are the application of a “green” photosensitizer abundant in nature with absorption in both the red and blue regions of the visible spectrum and using the same solvent for extraction of Chl and preparation of NPs. Moreover, glassy polystyrene NPs prepared by the nanoprecipitation method are more stable than liposome- or polymeric micelle-based NPs[10,29,30] and should preserve favorable photophysical properties not influenced by photoelectron transfer from excited Chl molecules to the metal core typically for gold[31] and silver[32] NPs.

Results and Discussion

ChlE High-Performance Liquid Chromatography Analysis

Chl a and Chl b are highly abundant in green leaves and are accompanied by protective carotenoids.[33] HPLC analysis revealed that the retention times and mass spectrometry (MS) profiles of compounds found in ChlE correspond to Chl a, Chl b, β-carotene, and lutein standards (Table , Figure S1, panel B, and Figure S2 in Supporting Information).

Table 1

Retention Time of Standards and Concentration of Corresponding Compounds in ChlEa

pigment	Chl a	Chl b	β-carotene	lutein
tR (min)	17.6	9.8	56.7	4.3
c (mg/l)	154 (190)	54 (62)	38	28 (39)
content in w/w %	56 (58)	20 (19)	14 (11)	10 (12)

The values in parentheses include the isomers of the respective compounds.

For HPLC quantitative estimation, a mixture of standards (Chl a, Chl b, β-carotene, and lutein) at concentrations of 0.5, 1, 2.5, 5, and 7.5 mg/L, respectively, in MeOH/THF 9:1 (v/v) was applied and used to construct calibration plots for each component (peak area vs. concentration). Then, ChlE diluted 10× in MeOH was tested. The concentration of each component was estimated from calibration plots (Table ). Evidently, Chl a is the main compound in ChlE.

Note that the amount/concentration of Chl photosensitizers and carotenoids in ChlE depends on the species of green plant and changes during a season.[34]

Spectroscopic and Photophysical Properties

The UV–vis absorption spectrum of THF extract prepared from spinach (ChlE) was compared with that of Chl a, Chl b, β-carotene, and lutein standards (Figure A). The spectrum of ChlE consists of a Qy band of Chls at 664 nm corresponding predominantly to the Chl a spectrum and of a Soret band in the blue region of the spectrum partially overlaid with the absorption band of other species (Chl b, carotenoids, etc.). The absorption spectrum allows efficient harvesting of red and blue radiation in the solar spectrum.

Figure 1

(A) Normalized UV/vis spectrum of ChlE in tetrahydrofuran (THF) compared with the Chl a, Chl b, β-carotene, and lutein spectra. (B) Dependence of singlet oxygen lifetime (τΔ) on the absorbance in the Qy band after dilution of original ChlE, and after irradiation of ChlE with a 250 W Xe lamp with a long pass filter (λ ≥ 400 nm). The red line corresponds to τΔ in THF measured for standard ZnPc and Chl a photosensitizers.

We measured the weak near infrared luminescence of singlet oxygen, O2(1Δg) at ∼1270 nm (Figure S3 and S4 in Supporting Information) and estimated the quantum yield (ΦΔ) by comparing the amplitudes of singlet oxygen decay kinetics after excitation with a dye laser (664 nm and 28 ns pulse length) at the same absorbance at the excitation wavelength (Table ).

Table 2

Photophysical Parameters: Lifetime of Singlet Oxygen in THF (τΔ), the Quantum Yield of Singlet Oxygen (ΦΔ), Lifetime of the Chl Triplet States in Argon (τ0)- and Air (τair)-Saturated THF and the Fraction of the Triplet States Quenched by Oxygen in Air-Saturated Solution (FT = 1 – τ0/τAir)

	singlet oxygen	triplet states
	τΔ (μs)	τ0 (μs)b	τair (μs)b	FT	ΦΔc
ChlE	10.2	110	0.25	>0.99	0.64
ChlE*a	19.9	141	0.35	>0.99
Chl a	20.7	294	0.24	>0.99	0.66
Chl b	20.7	229	0.26	>0.99

ChlE irradiated with visible light from a 250 W Xe lamp (λ > 400 nm) for 60 min.

Estimated error less than 5%.

Estimated error less than 15%.

Taking into account ΦΔ = 0.53 for ZnPc in THF,[35] the calculated values of ΦΔ were 0.64 and 0.66 for ChlE and Chl a, respectively. This result corresponds to previous literature data for ChlE prepared from spinach.[10]

The concentration of photosensitizers (predominantly Chl a) used in experiments was proportional to the absorbance at the Qy band of Chl. The estimated concentration was 10–5 mol.L–1 (Table S1 in Supporting Information) or lower taking into account the literature value for the extinction coefficient of Chl a (∼9 × 105 L mol–1 cm–1). The singlet oxygen lifetime (τΔ) increased with the dilution of ChlE due to the lower concentration of quenchers (carotenoids, Figure B).

To remove the O2(1Δg) quenchers from the extract, purification by chromatographic methods was needed. Alternatively, we irradiated raw ChlE with visible light, under which the quenching effect of carotenoids was suppressed by their oxidation with O2(1Δg) to form photochemically inactive endoperoxides.[17] The value of τΔ also increased with increasing irradiation time up to a value of ∼20.6 μs measured for the ZnPc standard, which corresponds with the literature data for THF (Figure B).[36] This effect corresponds with degradation of the quenchers through the reaction of O2(1Δg) with the double bonds of carotenoids. Note that carotenoids exhibit complex behaviors and can be degraded by other mechanisms, for example, by free radical reaction.[37,38] The changes in the absorbance of Qy bands at 664 nm also indicate partial photodegradation of ChlE (proportional to changes in the absorbance of the Qy band).

We also measured transient absorption spectra to evaluate the kinetics of the photosensitizer (predominantly Chl a) triplet states—a precursor of O2(1Δg) (Figure S5 in Supporting Information). The kinetics of the triplet states in THF show very efficient quenching of the triplets by dissolved oxygen (Table ), leading to O2(1Δg) formation. We also found a lower lifetime of the photosensitizer triplet states for ChlE (167 μs) in argon-saturated THF in comparison with the Chl a standard (294 μs); however, more than 99% of triplets were quenched by oxygen in an air-saturated solution of ChlE, indicating that quenching by oxygen was dominant in comparison with quenching by carotenoids. Overall, the basic processes photosensitized by ChlE are summarized in Figure .

Figure 2

Simplified scheme of processes photosensitized by ChlE leading to photo-oxidation of target structures. Chl molecules were partially photodegraded during irradiation.

Photostability

A relatively photostable ChlE from the cyanobacterium Spirulina maxima was recently suggested for application in photodynamic therapy after purification.[39] We tested the photostability of our raw ChlE from spinach after irradiation with a 500 W Xe lamp with a long pass filter (λ ≥ 400 nm), and the kinetics of decomposition were followed at 664 and 459 nm (Figure ). The first wavelength (664 nm) corresponds to the absorption bands of Chl, and the second wavelength (459 nm) corresponds to the absorption band of β-carotene. The photodegradation of β-carotene dominated, and the amount of β-carotene was negligible after 20 min of irradiation both in ChlE and in a two-component model, where standards of Chl a and β-carotene were mixed. The photodegradation of both Chl and β-carotene is a complex process, including oxidation by photogenerated O2(1Δg) and energy transfer from excited Chl a to β-carotene (Figure ).[40]

Figure 3

(A) Visible spectra of ChlE (corresponding to ∼4.7 mg/L Chl a). (B) Visible spectra of a two-component model mixture of Chl a (5.3 mg/L) with β-carotene (51 mg/L) standards during irradiation. The absorption bands of chlorophyll a at 664 nm and carotene at 482 nm are assigned as (1) and (2), respectively. (C) Kinetics of photodegradation of Chl a in ChlE (1a) and in the model mixture with β-carotene (1b) and kinetics of photodegradation of β-carotene in ChlE (2a) and in the model mixture (2b). Irradiation was performed with a 500 W Xe lamp with a long pass filter (λ ≥ 400 nm).

A sample of ChlE after 50 min of irradiation was analyzed by HPLC. Chl a and Chl b were present in the sample, but no lutein or carotene was found (Figure S1, panel C in Supporting Information).

Sulfonated and Nonsulfonated Polystyrene Nanoparticles

Preparation and Characterization

Sulfonated and nonsulfonated polystyrene NPs were prepared by simple nanoprecipitation of THF solutions of sulfonated or nonsulfonated polystyrene enriched with THF spinach extract or standards/quenchers (see Experimental Section and Figure S6 in Supporting Information). Sulfonation of the polystyrene matrix increased NP stability, even in aqueous environments with high ionic strengths.[41]

The dispersions of sulfonated/nonsulfonated polystyrene NPs with spinach extract (ChlE@NPs) and NPs without and with the chlorophyll standard and carotene quencher were characterized by transmission electron microscopy (TEM) (Figure B,C), UV/vis absorption and fluorescence spectroscopy (Figure E), and dynamic light scattering (DLS) (Figure F).

Figure 4

(A) SEM of pristine polystyrene nanofiber material for preparation of NPs, (B) TEM of sulfonated ChlE@NP dispersion, and (C) TEM of nonsulfonated ChlE@NP dispersion. (D) Photograph of a cell containing sulfonated ChE@NP dispersion with an impinging laser beam. (E) Absorption (a) and fluorescence (λexc = 417 nm) (b) spectra of sulfonated ChlE@NP dispersion. (F) DLS size distribution of sulfonated ChlE@NP and freshly prepared nonsulfonated ChlE@NP dispersion.

The nanoprecipitation method led to mostly spherical NPs, as observed in the TEM images, where larger NPs were more visible in comparison to the bulk of small NPs revealed by DLS. The basic properties of the NPs are summarized in Table .

Table 3

Basic Properties of Sulfonated and Nonsulfonated NPs

	nonsulfonated NPs		sulfonated NPs
properties	@NPs	ChlE@NPs	@NPs	ChlE@NPs	Chla@NPs	ChlaCar@NPs
diameter (DLS) (nm)	278	227	64	61	60	60
zeta potential (mV)	–34	–31	–33	–26	–31	–34
number of NPs in 1 mL dispersion	2.60 × 1010	4.48 × 1010	1.41 × 1013	2.02 × 1013	1.98 × 1014
surface (nm2)	243 000	162 000	13 000	12 000	11 000	11 000
surface/volume (nm–1)	0.02	0.03	0.09	0.10	0.10	0.10
appearance	milky	milky green	transparent	transparent green	transparent green	milky orange

Dispersions of NPs prepared from nonsulfonated polystyrene nanofiber membranes displayed hydrodynamic diameters of approximately 230–280 nm. The sulfonated NPs had average hydrodynamic diameters of approximately 60–64 nm with a low polydispersity (PDI index of 0.2–0.3), in contrast to nonsulfonated NPs, where the freshly prepared nonsulfonated NPs were polydispersed with a tendency to precipitate. The size and zeta potential of NPs showed their high stability over time (Table S2 in Supporting Information).

According to HPLC analysis, the composition of encapsulated compounds in NPs corresponded to the composition of ChlE. Generally, chlorophylls are not stable in an acid environment and can undergo demetalation to the corresponding pheophytin but under our experimental conditions, no demetalation occurs (Figure S1, panel C in Supporting Information).

Also, no absorption or fluorescence of Chls was found in the filtrate (Figure S7 in Supporting Information), which indicates no leaching to aqueous media. The Soret absorption maximum of Chl a in the absorption spectrum of ChlE was shifted from 436 nm in THF solution to 415 nm in ChlE@NP dispersion.

Photophysical Properties

In contrast to a solution, individual components of ChlE (Chls and carotenoids) are fixed at specific places inside glassy ChlE@NPs, and only oxygen can diffuse through the polystyrene matrix[42] with a relatively high oxygen diffusion coefficient [D(O2)∼3×107 cm–1 s–1].[43] The lifetime of singlet oxygen (τΔ) measured from weak NIR luminescence is controlled by the size of the NPs,[26] by different values of τΔ in the NP interior and outside in the aqueous environment, and by additional quenching by carotenoids (Figure a). Note that only O2(1Δg) that diffuses from the NP interior to the exterior environment can be used for photo-oxidation of chemical substrates and bacterial structures (see photo-oxidation tests and photodynamic inactivation of Escherichia coli). For this reason, polystyrene with a high oxygen diffusion coefficient was selected as the starting material for NP preparation.[43]

Figure 5

Simplified behavior of O2(1Δg) in the NP interior and outside NPs in an aqueous environment (a). Normalized singlet oxygen kinetics after excitation of ChlE in sulfonated ChlE@NPs (b) and preirradiated sulfonated ChlE@NPs (c). The SODF signal for preirradiated sulfonated (d) and nonsulfonated ChlE@NPs (e). Experiments were carried out in an oxygen-saturated water dispersion to increase the amount of O2(1Δg) generated by the photosensitization process. The red lines are single exponential fits to the experimental data. Note that the value of τΔ is influenced by slow decay kinetics of Chls triplets in nanoparticles (Figure S8 in Supporting Information).

Prolongation of O2(1Δg) luminescence kinetics for sulfonated ChlE@NPs prepared from extract preirradiated with visible light from a Xe lamp, Figure c, τΔ = 13 μs) in comparison with NPs prepared from raw extract (Figure b, τΔ = 8 μs), indicated photodegradation of carotenoids and loss of both physical and chemical quenching properties.

Nonsulfonated ChlE@NPs did not provide any measurable O2(1Δg) luminescence signal. We used singlet oxygen-sensitized delayed fluorescence (SODF), a more sensitive method for detection of singlet oxygen inside NPs, that arises from the interaction of O2(1Δg) with the triplet states of photosensitizers.[44] The SODF signal at high excitation energy is a few orders of magnitude stronger than the weak luminescence of O2(1Δg) at 1270 nm and therefore can be used for detection of lower concentrations of O2(1Δg). Comparison of the SODF signals confirmed a dramatic decrease in the O2(1Δg) concentration for nonsulfonated ChlE@NPs (Figure e) compared with preirradiated sulfonated ChlE@NPs (Figure d). Even nonirradiated ChlE@NPs showed a much higher SODF signal (Figure S9 in Supporting Information). Recently, we found that the SODF signal increased with the average size of NPs loaded with a photosensitizer (porphyrin) as a consequence of an increased amount of communicating/interacting molecules of photosensitizers via photogenerated O2(1Δg).[41,44] In contrast to previous studies, the higher SODF signal for smaller sulfonated ChlE@NPs than for larger nonsulfonated ChlE@NPs is likely due to inhibition of SODF by carotenoid quenchers that reduce the concentration of O2(1Δg) inside ChlE@NPs.

Photostability

ChlE and both sulfonated and nonsulfonated ChlE@NPs in aqueous dispersion (6.9 × 1012 NPs/ml) were irradiated with visible light from a 500 W Xe lamp with a long pass filter (λ ≥ 400 nm) to compare their photostability. The kinetics of Chl decomposition was followed at a wavelength of 664 nm, which corresponded to the absorption band of Chl a. The sulfonated ChlE@NPs exhibited similar photostability as ChlE itself that is in contrast to more stable nonsulfonated ChlE@NPs with smaller size and higher photo-oxidation efficiency (Figure S10, panel C in Supporting Information).

Photo-oxidation of Chemical Substrates

The dispersions of NPs were irradiated with a 36 W red light-emitting diode (LED) grow light bulb (λ = 662 nm) in iodide detection solution to observe O2(1Δg) photogeneration (see Experimental Section). A linear increase in the I3– concentration (following the UV/vis absorbance change at 287 or 351 nm) proportional to the generation of O2(1Δg)[45] was found (Figure S11 in Supporting Information). Alternatively, O2(1Δg) photogeneration was confirmed by photodegradation of uric acid at 293 nm[46] (Figure S12 in Supporting Information). The experimental data indicate that oxidation of both I– and uric acid substrates is due to oxidation by O2(1Δg) for several reasons: a) formation of O2(1Δg) was confirmed from luminescence measurement (see photophysical measurements above), (b) the oxidation did not take place in inert gas-saturated samples and/or in aerated samples in the dark, and (c) a strong physical quencher of singlet oxygen (sodium azide)[47] inhibited oxidation of both substrates.

A fluorescence assay using terephthalic acid[48] did not reveal the generation of any other ROS (H2O2, O2–, and OH.). Also, no post-irradiation effect, for example, dark oxidation of iodide (in iodide test), typical for accumulation of H2O2 due to presence of OH and O2– was found. However, we cannot exclude the minor formation of these species; the quantitative evaluation of data for such complex systems may be limited.[49]

To estimate the relative efficiency of I– photo-oxidation to I3– by O2(1Δg), absorbance-matched dispersions of different NPs were irradiated with a red LED bulb (λ = 662 nm). Figure shows that smaller sulfonated ChlE@NPs were much more effective producers of O2(1Δg) than larger, nonsulfonated ChlE@NPs and exhibited approximately the same photo-oxidation efficiency as Chla@NPs (using a Chl a analytical standard) of a similar size.

Figure 6

Absorption spectra of Chla@NPs (a), sulfonated ChlE@NPs (b), ChlaCar@NPs (c), and nonsulfonated ChlE@NPs (d). Relative photo-oxidation efficiency estimated as absorbance changes at 351 nm that corresponds to the I3– absorption band in the photo-oxidation of I- by O2(1Δg) using sulfonated Chla@NPs (e), ChlE@NPs (f), ChlaCar@NPs (g), and nonsulfonated ChlE@NPs (h). The measurements were carried out with absorbance-matched samples at an excitation wavelength of 662 nm; samples were irradiated with a red LED bulb (λ = 662 nm).

The presence of β-carotene reduces the photo-oxidation response. Surprisingly, we found that sulfonated ChlaCar@NPs (originating from 4 mg of β-carotene) have the same photo-oxidation kinetics as a sample with a half concentration of β-carotene (2 mg, not shown). This observation may correspond with the small size of NPs: O2(1Δg) photogenerated near the surface of NPs is directly released into aqueous media and does not meet any β-carotene (carotenoid) quencher located inside the NPs. In contrast, O2(1Δg) photogenerated inside larger nonsulfonated NPs can be quenched even by a low concentration of β-carotene. This efficient quenching reduces the number of O2(1Δg) molecules that diffuse into the environment of the NPs toward target structures and inhibits total photo-oxidation (Figure ).

To evaluate the preirradiation effect on the relative efficiency of photo-oxidation of chemical/biological targets by O2(1Δg), aqueous dispersions of sulfonated ChlE@NPs and nonsulfonated ChlE@NPs were irradiated with a 36 W red LED bulb before addition of iodide detection solution and photo-oxidation tests. Preirradiation of sulfonated ChlE@NPs led only to a decrease in photo-oxidation efficiency, attributed to the common photodegradation of Chl photosensitizers (Figure S13 in Supporting Information). The large nonsulfonated ChlE@NPs exhibited a more complex profile with limited reproducibility.

In our previous study,[41] we found that polystyrene NPs can be easily removed by filtration through a 0.03 mm thin hydrophilic electrospun polyurethane (Tecophilic) nanofiber membrane. The removal of sulfonated ChlE@NPs was followed by decreased red fluorescence in the supernatant (Figure S7 in Supporting Information). After one filtration, approximately 90% of fluorescent ChlE@NPs were removed. Repeated filtration yielded complete removal of all NPs. No fluorescence or absorption was observed, and only a background DLS signal was observed. This test also verified that the hydrophobic compounds in ChlE are fixed in the glassy polystyrene matrix, which efficiently prevents release of the encapsulated compounds into the aqueous medium. In short, only components of ChlE encapsulated in ChlE@NPs exhibited an efficient photoantibacterial effect due to the formation of cytotoxic O2(1Δg).

Photodynamic Inactivation of E. coli

Previous photo-oxidation experiments revealed that O2(1Δg)-generating NPs can easily oxidize external substrates in aqueous media. To test photodynamic inactivation (PDI), a suspension of Gram-negative E. coli was mixed with a dispersion of @NPs or sulfonated ChlE@NPs and irradiated with a 36 W red LED bulb (λ = 662 nm) for 0, 10, and 20 min. No significant photodegradation of Chls during irradiation (monitored by changes in the absorbance at 664 nm) was observed (Figure S10 in Supporting Information). In contrast to controls (suspension of E. coli alone or with sulfonated @NPs), a substantial decrease in the colony forming unit (CFU) ratio was observed when sulfonated ChlE@NPs were used (Figure ). ChlE@NPs kept in the dark also exhibited no antibacterial effect (Table S3 in Supporting Information). In contrast to the frequently observed antibacterial effect of visible light itself (especially in the blue region), we found a slight increase in the CFU ratio. This stimulation effect can be attributed to the temperature increase in the cuvette chamber (0.5–1.0 °C) during irradiation.

Figure 7

(Photo)antibacterial activity estimated as the average number of CFUs of E. coli observed on agar plates after irradiation with visible light. The average number of CFUs corresponding to treatment of E. coli alone (dashed), E. coli with @NPs (gray), and E. coli with sulfonated ChlE@NPs (green) after 0, 10, and 20 min of irradiation versus samples stored in the dark based on three independent tests, and photographs of agar plates. Irradiation source: 36 W red LED bulb (λ = 662 nm).

Conclusions

Polystyrene nanoparticles with Chl photosensitizers from green plants prepared by simple nanoprecipitation methods can be inexpensive and green alternatives to photoactive antimicrobial NPs with many applications, for example, for cleaning contaminated water or for antibacterial treatment of biofilms. The high quantum yield of singlet oxygen generation for ChlE (close to that of pure Chl a) allows direct application in photodynamic treatment without a complicated purification process to prepare individual photosensitizers. The quenching effect of carotenoids in ChlE can be suppressed by preirradiation with visible light, under which carotenoids decompose more efficiently than Chl photosensitizers and lose their quenching ability due to the reaction of O2(1Δg) with the conjugated double bonds of carotenoids. Encapsulation of Chl photosensitizers from ChlE in polystyrene NPs led to higher photostability. NPs prepared from sulfonated polystyrene with small sizes (diameter of ∼60 nm) exhibited more efficient photo-oxidation and antibacterial properties than those prepared from nonsulfonated polystyrene. Moreover, efficient removal of the NPs together with the inactivated bacteria after their use by simple filtration through an electrospun membrane (a precursor of NPs) is also an important benefit.

Experimental Section

Materials

Chl (Chl a and Chl b), lutein, and β-carotene standards; uric acid sodium salt; tetraethylammonium bromide (TEAB); hydrogen peroxide; ampicillin; potassium iodide and other inorganic salts (all Sigma-Aldrich); phosphate-buffered saline (PBS); agar and LB medium (Lennox) (all Carl Roth GmbH); cyclohexanone; and sulfuric acid (both Lach-Ner, Czech Republic) were used as delivered. THF (Sigma-Aldrich) was dried with a PureSolv MD5 solvent purification system (Innovative Technology). PS GP 137 polystyrene was purchased from Synthos Kralupy (Czech Republic). ChlE was prepared from spinach. Spinach leaves [100 g, Italy: grown in a greenhouse, baby spinach (Spinacia), summer] were dried at room temperature between filter papers in the dark for 2 weeks. ChlE was prepared from 17.5 g of dried leaves extracted with 200 mL THF, filtered and used for experiments without any other purification.

Preparation of Nanofiber Material and NPs

A mixture of 0.07 wt % TEAB and 99.93 wt % polystyrene was dissolved in cyclohexanone to prepare a 17% solution for fabrication of electrospun polystyrene nanofiber material. The conductivity of the solution was enhanced by TEAB (0.12 g/kg). The electrospinning process was described in detail in our previous papers.[25,43,50]

Sulfonated NPs were prepared from electrospun polystyrene nanofiber material fixed on quartz substrates treated with immersion in 96% sulfuric acid[41] at room temperature for 48 h. The materials were washed with distilled water until a neutral pH was reached and were then stored between two pieces of polypropylene fabric cover. Typically, a gently wet, sulfonated nanofiber membrane (16 mg) was immersed in 4 mL of dry THF with ChlE (ca 0.2 mg) or Chl a standard (0.2 mg) or β-carotene standards (2.0 mg) for 60 s with stirring; then, distilled water (20 mL) was added. The concentration of standards was checked using extinction coefficients (Table S1 in Supporting Information). THF was removed by evaporation under a slight vacuum (60 °C, to a final volume of approximately 15 mL). The resulting dispersion of NPs in water was centrifuged for 10 min at 3070 g to remove microparticles and was dialyzed using Float-A-Lyzer G2 with a molecular weight cutoff of 50 kDa for 15 h in distilled water at room temperature to remove traces of sulfuric acid and THF. Nonsulfonated NPs were prepared under the same protocol (see above) without sulfonation.

HPLC Analysis

HPLC analyses were performed on an Agilent Infinity 1290 liquid chromatographic system coupled with a Triple Quad 6460 tandem mass spectrometric detection system (Agilent Technologies, Germany). A SunFire C18 column (150 mm × 4.6 mm, 5 μm particle size), (Waters, Ireland), thermostatted at 20 °C, was used for separation. The mobile phase consisted of methanol (MeOH) with 0.5% formic acid at a flow rate of 0.8 mL/min. The injection volume was 20 μL, and samples were kept at 20 °C. The MS/MS measurements were performed in the multiple reaction-monitoring (MRM) mode using positive electrospray ionization (ESI). The gas flow was 10 L/min, gas temperature was 350 °C, nebulizer pressure was 55 psi, and capillary voltage was 5500 V. The optimized MS/MS conditions are shown in Table S4 in the Supporting Information. Total-ion current (TIC) chromatograms with MRM chromatograms of the individual analytes obtained for a mixture of standards are shown in Figure S1, panel A (Supporting Information), and the same TIC chromatograms recorded for ChlE are shown in Figure S1, panel B (Supporting Information).

Characterization of NPs and Electrospun Polymeric Membranes

Nanofiber and NP morphology was studied with a scanning electron Quanta 200 FEG microscope (FEI, Czech Republic). The nanofiber diameters were measured using NIS Elements 4.0 image analysis software (Laboratory Imaging, Czech Republic). The NP size and size distributions in water were determined by dynamic light scattering on a Zetasizer Nano ZS particle size analyzer from Malvern (United Kingdom).

Photophysical Properties

UV/vis absorption spectra were recorded using Unicam 340 and Varian 4000 spectrometers. The steady-state fluorescence spectra were monitored using an FLS 980 (Edinburgh Instruments, UK) spectrofluorimeter. For time-resolved measurements, the samples were excited with a Lambda Physik FL3002 laser (λexc = 664 nm and pulse length ∼28 ns) that matched the Qy band of Chls[51] in the extract and both Chl a and zinc phthalocyanine (ZnPc) standards. Time-resolved near-infrared luminescence of O2(1Δg) at 1270 nm was observed using a homemade detector unit (interference filters, Ge diode). Temporal profiles of O2(1Δg) luminescence were averaged and calculated as the difference between signals in oxygen- and argon-saturated H2O. They were fitted to a single exponential function: I = I0 exp(−t/τΔ), in which τΔ is the lifetime of singlet oxygen. The fitting procedure excluded the initial part of the plot influenced by light scattering and fluorescence (usually 1–2 μs after excitation). The details can be found in our previous paper.[26] Transient absorption spectra of photosensitizers (corresponding to triplet–triplet transitions) and decay kinetics of the photosensitizer triplet states were measured on an LKS 20 laser kinetic spectrometer (Applied Photophysics, United Kingdom) in oxygen-, air-, and argon-saturated solvents/dispersions.

Photo-oxidation Properties

A dispersion of NPs with or without encapsulated ChlE, Chl a, or β-carotene (∼1.3 × 1012 NPs/mL for sulfonated NPs and ∼1.5 × 109 NPs/mL for nonsulfonated NPs) was placed in a thermostatted 10 mm quartz cell (22 °C) that contained 0.1 M iodide detection solution or 2 × 10–4 M uric acid in 0.02 M phosphate buffer (pH = 7.0). The cell was irradiated with visible light using a stabilized xenon lamp (500 W, Newport) with a long pass filter (λ ≥ 400 nm, Newport) or using a 36 W LED bulb with an emission at 662 nm (for spectral irradiance see Figure S14 in Supporting Information). The UV/vis absorbance changes at 287 or 351 nm (attributed to the formation of I3– in the iodide test)[45] or 291 nm (attributed to photodegradation of uric acid)[52,53] were recorded at regular intervals and compared to a blank solution of the same composition that was stored in the dark. Spectral and total irradiance of the used light sources was evaluated with an ILT960 spectroradiometer SpectriLight (International Light Technologies, USA).

Antibacterial Assays

A culture of E. coli DH5α (Invitrogen, CA, USA) with the plasmid pGEM11Z (Promega, WI, USA) was incubated at 37 °C while stirring in LB medium after addition of ampicillin. Incubation was terminated when the absorbance at 560 nm reached approximately 1.2. The prepared culture was diluted 5000× to the desired concentration in PBS. NPs from stock suspensions of sulfonated NPs (∼2.0 × 1013 NPs/mL) with or without encapsulated ChlE were mixed with diluted bacterial culture at a ratio of 1:1. Two milliliters of this suspension was placed in a thermostatted 10 mm quartz cell (25 °C). While stirring, the cell was irradiated with red light produced by a stabilized deep red LED grow light bulb (36 W and 662 nm). At regular time intervals, 150 μL of the irradiated suspension was placed on an agar plate. The plates were incubated for 20 h in darkness at 37 °C to allow the individual bacteria to grow and form colonies.[41]