SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References

Photophysical, photostability, electrochemical and molecular-orbital characteristics are analyzed for a set of stable dicyanobacteriochlorins that are promising photosensitizers for photodynamic therapy (PDT). The bacteriochlorins are the parent compound (BC), dicyano derivative (NC)2BC and corresponding zinc (NC)2BC-Zn and palladium chelate (NC)2BC-Pd. The order of PDT activity against HeLa human cancer cells in vitro is (NC)2BC-Pd > (NC)2BC > (NC)2BC-Zn ≈ BC. The near-infrared absorption feature of each dicyanobacteriochlorin is bathochromically shifted 35–50 nm (748–763 nm) from that for BC (713 nm). Intersystem crossing to the PDT-active triplet excited state is essentially quantitative for (NC)2BC-Pd. Phosphorescence from (NC)2BC-Pd occurs at 1122 nm (1.1 eV). This value and the measured ground-state redox potentials fix the triplet excited-state redox properties, which underpin PDT activity via Type-1 (electron transfer) pathways. A perhaps counterintuitive (but readily explicable) result is that of the three dicyanobacteriochlorins, the photosensitizer with the shortest triplet lifetime (7 μs), (NC)2BC-Pd has the highest activity. Photostabilities of the dicyanobacteriochlorins and other bacteriochlorins studied recently are investigated and discussed in terms of four phenomena: aggregation, reduction, oxidation and chemical reaction. Collectively, the results and analysis provide fundamental insights concerning the molecular design of PDT agents.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References

Photodynamic therapy (PDT) is an emerging treatment approach that uses a nontoxic photosensitizer and harmless visible or near-infrared (NIR) light to kill diseased cells by generating reactive oxygen species, such as singlet oxygen, superoxide and hydroxyl radicals [1]. To enhance the efficacy and selectivity of response, molecular tuning of the chemical and electronic-structure characteristics of photosensitizers is employed to produce (1) favorable photophysical properties; (2) chemical and photochemical stability; (3) preferential delivery to the destination tissue or cell type, including subcellular compartments; and (4) production of reactive oxygen species that are especially lethal to the target cell.

An important consideration is to choose or tune the photosensitizer to have strong absorption in the NIR region where light penetration through tissue is maximal because both absorption and scattering of light at these wavelengths by endogenous chromophores are minimal. In this regard, bacteriochlorins and related macrocycles are ideal because of the intense NIR (720–850 nm) absorption resulting from the reduction of two pyrrolic rings in the tetrapyrrole macrocycle compared to chlorins (one reduced ring, modestly intense red absorption) and porphyrins (no reduced rings and weak or no red or NIR absorption). Naturally occurring bacteriochlorins or derivatives thereof, such as WST9 (Tookad) and WST11 (Chart 1) can have high PDT efficacy, but have drawbacks due to instability (in the dark and in the light) and limitations on molecular tailoring because nearly all sites about the perimeter of the macrocycle already bear substituents.

image

Chart 1. Representative native bacteriochlorin derivatives (WST9 = Tookad, WST11) [20, 21] and a synthetic bacteriochlorin (TDCPBS) [27, 28] that have been examined in PDT studies.

Download figure to PowerPoint

A de novo synthetic route has been developed to access bacteriochlorins that are stabilized against adventitious oxidation by the presence of a geminal dimethyl group in each reduced ring [2-4]. This route has been used to prepare bacteriochlorins containing a variety of positively charged substituents, groups to impart water solubility, and/or groups that can vary the balance between hydrophobicity and hydrophilicity. The PDT activity of these bacteriochlorins has been tested against B16 mouse melanoma cells [5], HeLa human cervical cancer cells [6] and microbial cells (Gram-positive, Gram-negative bacteria and yeast; [7]) in vitro. A number of these synthetic bacteriochlorins are highly efficacious (i.e. LD50 ≤100 nm) for cell killing against human cancer cell lines. Representative structures are shown in Chart 2. The bacteriochlorins employed are all the metal-free (free base) forms. The PDT efficacy of tetrapyrrole chromophores (e.g. porphyrins, chlorins, bacteriochlorins, phthalocyanines) often depends on the metalation state [8], as was found for a set of imidazole-substituted metalloporphyrins (Chart 3) targeted against HeLa and CT26 cancer cells [9, 10].

image

Chart 2. Imidazole-substituted porphyrins examined previously for PDT activity [9, 10].

Download figure to PowerPoint

image

Chart 3. Representative synthetic bacteriochlorins examined previously for PDT activity [5-7].

Download figure to PowerPoint

Some of the synthetic bacteriochlorins are more readily photobleached than others. Photobleaching is defined as loss of absorption can stem from a variety of causes (oxidation, reduction, aggregation, chemical reaction). Understanding the interplay of molecular structure and origin of photobleaching in principle could lead to improvements in photostability and thereby afford increased PDT activity.

Recently we synthesized and examined the PDT efficacy against HeLa cancer cells of a set of free base and metal-substituted bacteriochlorins bearing electron-withdrawing, cyano groups (Chart 4) [11]. In addition to photoactivity, analyses were performed of cellular uptake, subcellular distribution, and the propensity of the compounds to form singlet oxygen (Type-2 photochemistry) or hydroxyl radicals (Type-1 photochemistry). The dicyanobacterio-chlorins show (1) the same trend in efficacy with metal ion (e.g. Pd2+ > Zn2+) as determined for the imidazole-substituted (and other) porphyrins; (2) increased photostability and (3) efficacies comparable to or surpassing those of the best of the synthetic bacteriochlorins studied previously and other common PDT agents.

image

Chart 4. Synthetic bacteriochlorins examined recently for PDT activity [11].

Download figure to PowerPoint

In this work, the photophysical, redox and molecular-orbital characteristics of the unsubstituted and dicyanobacteriochlorins are presented along with more extensive photostability data. This information allows analysis of the recent results [11] concerning photoactivity and reactive oxygen species production in terms of fundamental electronic properties, including excited-state redox potentials. To increase the generality of the findings, the analysis is extended to include results on the imidazole-substituted porphyrins and representative synthetic bacteriochlorins studied previously, for which photostability characteristics are presented here for the first time. The collective findings provide fundamental insights into the design and electronic tuning of bacteriochlorin photosensitizers for enhanced PDT efficacy.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References
Photosensitizers

The syntheses of BC [4], (NC)2BC [3] and (NC)2BC-Zn [3] have been described. The synthesis of (NC)2BC-Pd will be reported elsewhere. Absorption and emission studies indicate that samples of (NC)2BC-Pd studied herein contain <1% of the free base (NC)2BC starting material. Photofrin and Verteporfin (liposomal benzoporphyrin derivative, BPD) were gifts from QLT, Inc. (Vancouver, Canada) and Lutex was a gift from Pharmacyclics, Inc. (Sunnyvale, CA).

Photophysical characterization

Static absorption (Varian Cary 100 or Shimadzu UV-1800) and fluorescence (Spex Fluorolog Tau 2 or PTI Quantamaster 40) measurements were performed at room temperature, as were in all other studies. Determination of the fluorescence quantum yield (Φf), singlet excited-state lifetime (τS) and triplet yield (ΦT) utilized dilute (μm) Ar-purged toluene and methanol solutions. Measurements of the triplet lifetime (τT) utilized Ar-purged 2-methyl tetrahydrofuran solutions. Samples for Φf measurements had an absorbance ≤0.1 at the excitation wavelength to minimize front-face effects and similarly low absorbance in the Qy(0,0) band to minimize inner-filter effects.

Static emission measurements employed 2–4 nm excitation- and detection-monochromator bandwidths and 0.2-nm data intervals. Emission spectra were corrected for detection-system spectral response. Some measurements employed an extended NIR sensitive detector that drives a lock-in amplifier frequency-referenced to the rate of chopping of the excitation light prior to the excitation monochromator. Fluorescence quantum yields were determined relative to several different standards. These standards are (1) chlorophyll a in deoxygenated toluene (Φf = 0.325) [12], which is the value measured in benzene [13]; (2) free base meso-tetraphenylporphyrin (FbTPP) in nondegassed toluene, for which Φf = 0.070 was established with respect to the zinc chelate ZnTPP in nondegassed toluene (Φf = 0.030) [14], a value consistent with prior results on FbTPP [15]; and [3] 8,8,18,18-tetramethylbacteriochlorin [4] in Ar-purged toluene, for which Φf = 0.14 was established with respect to chlorophyll a in benzene and FbTPP in toluene.

Singlet excited-state lifetimes (τS) for all compounds except (NC)2BC-Pd were obtained using time-correlated-single-photon-counting detection on an apparatus with an approximately Gaussian instrument response function with a full-width-at-half-maximum of ∼1 ns (Photon Technology International LaserStrobe TM-3). Samples were excited in the Soret or Qx regions using excitation pulses at 337 nm from a nitrogen laser or in the blue to green spectral regions from a dye laser pumped by the nitrogen laser. The τS value for (NC)2BC-Pd was obtained by ultrafast transient absorption spectroscopy, probing disappearance of the bleaching in the Qy ground-state absorption band and excited-state absorption features following excitation in the Qx band with an ∼130 fs excitation flash [16]. The τT values were similarly determined using transient absorption spectroscopy, probing the decay of bleaching of the Soret band and excited-state absorption features following excitation in the Qx band with ∼5 ns pulses from a Q-switched Nd:YAG laser (532 nm) or a dye laser pumped by the Nd:YAG laser [9, 16].

The Φisc values (triplet yields) were obtained using transient absorption spectroscopy. The extent of bleaching of the ground-state Qx bands due to the formation of the lowest singlet excited state was measured immediately following a 130 fs flash in the Qy(0,0) band and compared with that due to the formation of the lowest triplet excited state at the asymptote of the singlet excited-state decay [9, 16].

Photostability studies

Photostability measurements utilized excitation light obtained from a 300 W Xenon lamp (Model R300-3 lamp and PS300-1 power supply; ILC Technologies, Sunnyvale, CA) that passed through a 70 cm path cell containing deionized water followed by a monochromator with a 10 nm bandpass. The light intensity at the sample position was measured using a calibrated diode and optometer (Models 221 and S471; United Detector Technologies, San Diego, CA) and was typically ∼5 mW cm−2 at the wavelength (720–750 nm) of the NIR absorption maximum of the bacteriochlorin sample (A ∼0.5 in 1 cm). The solutions contained ambient (atmospheric) O2 or had the O2 removed by purging with Ar for ∼1 h or by repeated freeze-pump-thaw cycles on a high vacuum line that achieved a vacuum of <10−6 Torr. The O2 was removed from all aqueous micellar solutions via the latter method. Samples were stirred using a micro magnetic bar at the bottom of the cuvette during illumination.

Electrochemistry

Electrochemical studies were performed using previously described instrumentation [17]. The solvent was butyronitrile (Burdick and Jackson) containing 0.1 m tetrabutylammonium hexafluorophosphate (Aldrich; recrystallized three times from methanol and dried at 110°C in vacuo) as the supporting electrolyte. The electrochemical cell was housed in a glovebox. The E½ values were obtained with square wave voltammetry (frequency 10 Hz) under conditions where the ferrocene couple has a potential of +0.19 V.

Density functional theory calculations

DFT calculations were performed with Spartan ‘10 for Windows version 1.2.0 in parallel mode on a PC equipped with an Intel i7-975 cpu, 24 GB ram, and three 300 GB, 10 k rpm hard drives (except for molecular mechanics and semi-empirical models, the calculation methods used in Spartan ′08 or ′10 have been documented in Shao et al. [18]). The calculations employed the hybrid B3LYP functional and basis sets 6-31G* and LACVP (the former for atoms H to Kr and lanl2dz for atoms Kr and above). The equilibrium geometries were fully optimized using the default parameters of the Spartan program. Molecular-orbital images were plotted using an isovalue of 0.016.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References

Sensitizer activity, uptake, localization and reactive oxygen species production

In our recent study [11], the PDT activity against HeLa cancer cells in vitro was studied using the parent bacteriochlorin (BC), dicyano derivative (NC)2BC and the corresponding zinc chelate (NC)2BC-Zn and palladium chelate (NC)2BC-Pd (Chart 4). The photosensitzers were delivered either by (1) direct dilution (denoted “dd”) of a solution of bacteriochlorin in an organic solvent [5 mm in N,N-dimethylacetamide (DMA) or tetrahydrofuran (THF)] into serum-containing complete culture medium; or (2) encapsulation into aqueous Cremophore EL micelles (CrEL) followed by dilution into the same complete medium. For both delivery methods, the order of PDT activity against HeLa cells after incubation for 24 h and illumination with 10 J cm−2 of NIR light is (NC)2BC-Pd > (NC)2BC > (NC)2BC-Zn > BC. As can be seen from Table 1, the LD50 values improve using CrEL versus direct dilution only modestly for (NC)2BC-Pd (18 versus 25 nm) and (NC)2BC (25 versus 60 nm), but significantly for (NC)2BC-Zn (60 versus 1000 nm) and BC (350 versus 1800 nm).

Table 1. PDT activity, photostability and photophysical properties of the bacteriochlorins.a
CompoundLD50 dd (nm)LD50 CrEL (nm)Uptake dd (nmol mg−1)Uptake CrEL (nmol mg−1)Efficacy/uptake ddbEfficacy/uptake CrELbPhoto-stability DMAcPhoto-stability CrELcPhoto-stability CrEL no O2c Φ f τS(ns) Φ T τT(μs)
  1. a

    All measured quantities were obtained at room temperature. Data for B16-B31 are taken from refs. [5, 6], and [16]. LD50 and uptake data for the unsubstituted and dicyano bacteriochlorins are taken from ref. [11]. For the LD50, uptake, and efficacy/uptake values, “dd” reflects direct dilution of the DMA stock solution of the photosensitizer into complete medium and CrEL reflects dilution of the micellar solution into complete medium. Measurements of the fluorescence yield (Φf), singlet excited-state lifetime (τS) and triplet yield (ΦT) employed Ar-purged toluene solutions and of the triplet lifetime (τT) Ar-purged THF solutions.

  2. b

    Calculated by dividing the reciprocal of the LD50 by the uptake value and normalizing to the value for BC dd.

  3. c

    Photostability as measured by the percentage of the initial Qy absorbance remaining after 100 J cm−2 fluence of light given at that wavelength.

  4. d

    The values are reached at ≤50 J cm−2 fluence.

Unsubstituted & dicyano BCs
BC 18003505.33.71.07.30.130.110.540.144.00.62169
(NC)2BC 60254.33.337.1114.30.910.90 0.154.10.4384
(NC)2BC-Zn 1000602.12.14.576.40.520.040.670.153.90.63121
(NC)2BC-Pd 25184.82.379.6227.20.830.94 ≤0.0070.0230.997
Representative prior studied BCs
B16 100 11.7 8.1 0.020.010.560.113.60.40190
B17 80 13.5 8.8 0.090.210.710.083.00.56198
B19 15 0.001 632400    0.153.00.55159
B22 3000 7.7 0 0.600.11d0.760.114.00.4877
B29 100 8.2 11.6 0.090.11d0.780.123.80.53190
B31 800 1.8 6.6 0.020.05d0.730.093.50.5354

Values for the uptake of 5 μm bacteriochlorin at different time points in the incubation medium are also given in Table 1. The highest uptake using direct dilution is found for BC and (NC)2BC-Pd, followed by (NC)2BC and (NC)2BC-Zn. The use of CrEL substantially decreases the uptake of (NC)2BC-Pd, (NC)2BC and BC but has little effect for (NC)2BC-Zn. Table 1 also gives the PDT efficacy per unit uptake, obtained by dividing the reciprocal of the LD50 value by the uptake value and then normalizing to the value of BC using direct dilution. The tabulated values of efficacy corrected for uptake for both delivery methods show the same order as the tabulated values for activity alone: (NC)2BC-Pd (NC)2BC (NC)2BC-Zn BC. Thus, although cell uptake certainly influences the overall PDT activity, other factors appear to dominate. In this regard, the greater efficacy of (NC)2BC-Pd versus (NC)2BC-Zn parallels our prior results on imidazole-substituted metalloporphyrins (Chart 3) wherein 2-Pd is more active than 2-Zn, exhibiting LD50 values of 55 and 833 nm under the excitation conditions employed [9, 10]. The studies on the dicyanobacteriochlorins complement and extend this prior work by inclusion of a free base analogue, providing additional information for analysis of activity.

Fluorescence microscopy studies reveal that subcellular localization of the unsubstituted and dicyanobacteriochlorins is in the endoplasmic reticulum, mitochondria and lysosomes depending on the compound. The least active compound, BC, is found mainly in the endoplasmic reticulum and lysozymes, and the next compound, (NC)2BC-Zn, is also found mainly in lysozymes with little evidence for association with mitochondria. The second most active bacteriochlorin, (NC)2BC, localizes mainly in the endoplasmic reticulum and mitochondria. The most active compound, (NC)2BC-Pd, also targets mitochondria (and lysosomes) as deduced by the damage done by PDT activity monitored after illumination. The results suggest that the propensity for mitochondrial targeting contributes to the relative PDT efficacy of the four bacteriochlorins. Mitochondria are considered to be an important PDT target because of the vital cellular functions performed, including regulation of metabolism, cell-cycle control development, and induction of apopotic cell death after mitochondrial damage.

The propensities of the bacteriochlorins to produce singlet oxygen by energy transfer (Type-2 mechanism) or hydroxyl radical by electron transfer (Type-1 mechanism) were probed using fluorescent dyes sensitive to the particular reactive oxygen species. Of the four bacteriochlorins, (NC)2BC-Pd produces the most singlet oxygen and also the most hydroxyl radicals, with the balance favoring the latter (Type-1 photochemistry). The next active compound, (NC)2BC, also produces a relatively high ratio of hydroxyl radicals to singlet oxygen. The two least active photosensitizers, (NC)2BC-Zn and BC, give low ratios of hydroxyl radicals versus singlet oxygen and therefore carry out less Type-1 than Type-2 photochemistry. In summary, the most active bacteriochlorins (against HeLa cells) produce the greatest amounts of hydroxyl radicals. Similar results were found previously for the imidazole-substituted porphyrins (Chart 3), for which 2-Pd produces more hydroxyl radicals and, as noted above, have greater PDT activity than 2-Zn [10].

Initial results of photobleaching studies obtained previously [11] on the four bacteriochlorins are also given in Table 1. Listed is the fraction of the NIR Qy(0,0) absorbance remaining after 100 J cm−2 of illumination in the same optical band. Values are given for the compounds in DMA and CrEL solutions in the presence of ambient O2 and for the latter in the absence of O2. The most active compounds, (NC)2BC-Pd and (NC)2BC, are the most photostable, with typically >90% of the Qy absorbance remaining even in the presence of ambient O2. The two lesser active compounds, (NC)2BC-Zn and BC, show lower photostability in the presence of O2 and marked improvement if O2 is removed from the solution. Below, more extensive measurements of the photostability of the dicyanobacteriochlorins are presented and compared with those for representative synthetic bacteriochlorins whose PDT activity (but not photostability) have been studied previously.

Absorption and fluorescence spectra

Figure 5A shows the absorption spectra for BC, (NC)2BC, (NC)2BC-Zn and (NC)2BC-Pd in toluene. The wavelengths of all the major absorption features for the bacteriochlorins in toluene and methanol are given in Table 2. For each bacteriochlorin the spectral positions are similar in the two solvents, and the same is found for the compounds in DMA and in CrEL solutions. All four compounds show good solubility in the four media.

Table 2. Absorption and fluorescence properties of dicyanobacteriochlorins.a
CompoundSolventBy abs (nm)Bx abs (nm)Qx abs (nm)Qy abs (nm)Qy em (nm)inline image Φ f τS (ns)
  1. a

    All measurements were made at room temperature.

BC Tol3403654897137160.850.144.0
MeOH3353594857087120.720.0954.1
(NC) 2 BC Tol3473725157487521.30.154.1
MeOH3423675107437471.30.0954.1
(NC) 2 BC-Zn Tol3433805467617631.80.153.9
MeOH3413745537527571.90.154.1
(NC) 2 BC-Pd Tol3263745187517533.40.0070.023
MeOH3233715147447492.80.007 
image

Figure 5. Absorption spectra (solid) and fluorescence spectra (dashed) of bacteriochlorins in toluene: BC (black), (NC)2BC (green), (NC)2BC-Zn (blue), (NC)2BC-Pd (red). Panel A shows the full near-UV to NIR absorption spectrum. Panel B focuses on the NIR (Qy) region of the absorption spectra and the companion fluorescence spectra.

Download figure to PowerPoint

Figure 5B focuses on the NIR absorption feature, the Qy(0,0) band. This band for the metal-free dicyanobacteriochlorin (NC)2BC is substantially (35 nm) hypsochromically shifted versus the parent bacteriochlorin BC (748 versus 713 nm in toluene). The incorporation of the central Pd2+ ion of (NC)2BC-Pd slightly bathochromically shifts the band to 751 nm while the Zn2+ of (NC)2BC-Zn causes a larger shift to 761 nm. Figure 5B shows that the corresponding Qy(0,0) fluorescence band for each bacteriochlorin lies only 2–5 nm to longer wavelength than the Qy(0,0) absorption maximum (Table 2).

Fluorescence yields, singlet excited-state lifetimes and triplet yields

Compounds BC, (NC)2BC and (NC)2BC-Zn in toluene and methanol have similar singlet excited-state lifetimes (τS = 3.9–4.1 ns) and fluorescence quantum yields (Φf = 0.14–0.15). The results are summarized in Tables 1 and 2. The lifetime of (NC)2BC-Pd is dramatically reduced (τS = 23 ps) as is the fluorescence yield (Φf ≤ 0.007). The latter value is an upper limit because the possibility cannot be excluded that some of the weak emission derives from trace demetalated analogue (NC)2BC, which has spectral overlap with (NC)2BC-Pd. Indeed, based on the relative τS values, one would have expected a Φf value of ∼0.0008 for (NC)2BC-Pd. The Φf and τs values of BC, (NC)2BC and (NC)2BC-Zn are comparable to those that we have found recently for synthetic free base bacteriochlorins (see e.g. Table 1) [5-7, 16, 19] and zinc bacteriochlorins [19] bearing a number of substituent patterns. The results for (NC)2BC-Pd are in keeping with values for two other synthetic palladium bacteriochlorins [19], and palladium-substituted bacteriochlorophyll and the corresponding sulfonato derivative [20, 21].

The reduced τS and Φf values for (NC)2BC-Pd (like other palladium tetrapyrroles) derives from heavy-metal (and perhaps d-orbital) enhancement of singlet-to-triplet intersystem crossing, which in turn makes the latter process virtually quantitative (ΦT = 0.99). The yield of intersystem crossing (i.e. the triplet yield) for (NC)2BC-Pd can be compared with 0.63 for (NC)2BC-Zn and 0.43 for (NC)2BC (Table 1). The ΦT value for BC (0.62) is greater than that for (NC)2BC, reflecting the effect of the cyano groups to draw electron density from the macrocycle and thereby diminish spin-orbit coupling, which underlies the intersystem-crossing process. The ΦT values obtained here for the dicyanobacteriochlorins are consistent with the results of prior studies on synthetic bacteriochlorins, which give average yields of ∼0.5 for free base bacteriochlorins, ∼0.7 for zinc chelates,∼0.8 for indium chelates and ∼1 for palladium chelates [11-13, 16, 19]. Essentially quantitative intersystem crossing is also found for palladium-substituted bacteriochlorophyll and derivatives thereof [20, 21]. These values for bacteriochlorins can be compared with the typical triplet yields of 0.7–0.8, 0.8–0.9 and 0.9–1 for free base, zinc, and palladium porphyrins, respectively [9, 22].

These comparisons show that one can expect a maximal enhancement factor in triplet yield of ∼1.5 for porphyrins and ∼2 for bacteriochlorins upon changing the metalation state (i.e. palladium versus free base). The data in Table 1 indicate that PDT activity, as judged by LD50 or related measures that account for uptake, show variations in activity of one or two orders of magnitude. Although heavy-metal (e.g. palladium) substitution may not give a significant gain in PDT efficacy due to an increased triplet yield, activity could be enhanced via accompanying changes in more critical characteristics, such as ground- and excited-state redox properties, shorter T1 lifetimes, and altered axial ligation, as discussed previously [9] and below.

Phosphorescence spectra and triplet excited-state energies

Figure 6 shows an emission scan for (NC)2BC-Pd in deoxygenated THF at room temperature; the scan was collected using a detection system with extended NIR sensitivity. The spectrum shows fluorescence (747 nm) close in position to that found in toluene (753 nm; Fig. 5B) as well as weak phosphorescence at 1122 nm. The phosphorescence is verified as originating from (NC)2BC-Pd using excitation spectra and is completely quenched (as expected) if the solution contains ambient O2. Phosphorescence was not observed for (NC)2BC-Zn under the same conditions, and is expected to be much weaker.

image

Figure 6. Emission spectrum of (NC)2BC-Pd in 2-methyltetrahydrofuran. The band at 747 nm is fluorescence and the band at 1122 nm is phosphorescence. The spectra were acquired using excitation in the Soret band (374 nm). The phosphorescence was also observed using excitation in the Qy band (740 nm).

Download figure to PowerPoint

Because of the rather small number of reports of bacteriochlorin phosphorescence, this emission was also measured for two other recently synthesized palladium chelates (B9 and B87) for which the other photophysical properties have been obtained recently [19]. The structures of these palladium bacteriochlorins and their zinc and free base analogues are shown in Chart 5. The phosphorescence, verified by excitation spectra, occurs at 1114 nm for B9 and 1118 nm for B87, similar to 1122 nm for (NC)2BC-Pd. Each wavelength corresponds to a triplet excited-state (T1) energy of 1.11 eV. The latter value is lower than that of the singlet excited state (S1) for the three bacteriochlorins by 0.56, 0.52 and 0.54 eV respectively.

image

Chart 5. Two sets of synthetic bacteriochlorins prepared and studied recently [19].

Download figure to PowerPoint

The above-noted phosphorescence wavelengths, T1 energies and S1 − T1 energy gaps for the three synthetic palladium bacteriochlorins are within the span of most reported values. The compounds include bacteriochlorophyll (BChl) and its zinc- and palladium-substituted analogues (1170–1221 nm; 1.01–1.06 eV; 0.56–0.58 eV), the free base analogue bacteriopheophytin (1097 nm; 1.13 eV; 0.50 eV), zinc tetraphenylbacteriochlorin (1157 nm; 1.07 eV; 0.57 eV) and free base tetraphenylbacteriochlorin (1053 nm; 1.17 eV; 0.50 eV) [23, 24]. Similar T1 energies (1.15 and 1.22 eV) have been determined for two chlorinated tetrasulfonatophenylbacteriochlorins using photoacoustic spectroscopy [25]. The T1 energies for the latter two compounds could be slightly lower than reported because the analysis assumed that internal conversion is effectively zero (i.e., Φf + ΦT = 1) whereas recent results on a large group of free base bacteriochlorins indicate that this nonradiative pathway typically has at least twice the quantum yield as fluorescence [16]. The triplet energy of 1.30 eV for free base tetraphenylbacteriochlorin [22] is modestly higher than the above-noted phosphorescence energy of 1.17 eV [24]. Elucidation of triplet excited-state energies is important for deducing excited-state redox potentials and assessing PDT mechanisms as discussed recently [22, 25] and below.

Triplet excited-state lifetimes

The lifetime of the lowest triplet excited state (τT) in deoxygenated solution is in the range 80–170 μs for BC, (NC)2BC and (NC)2BC-Zn, but is dramatically reduced to 7 μs for (NC)2BC-Pd (Table 1). Similar triplet lifetimes (3–10 μs) have been reported for other synthetic palladium bacteriochlorins [19] and for palladium-substituted bacteriochlorophylls [20, 21]. The shorter lifetimes of the palladium versus zinc or free base bacteriochlorins can be attributed to heavy-atom enhancement of both triplet excited-state decay pathways (phosphorescence emission and nonradiative intersystem crossing). The triplet lifetime is reduced to <1 μs for each of the bacteriochlorins in the presence of ambient O2, consistent with the added contribution of the energy/electron-transfer processes that underlie the formation of the reactive oxygen species for PDT.

The results of this study of metal-containing dicyanobacteriochlorins along with the findings in our prior studies of free base bacteriochlorins [5-7] and imidazole-substituted porphyrins [9, 10] show no clear correlation between triplet lifetime and PDT activity (Table 1). In fact, a seemingly counterintuitive result is that of the three dicyanobacteriochlorins, the photosensitizer with the shortest triplet lifetime (7 μs), (NC)2BC-Pd, has the highest activity. The same is true concerning imidazole-substituted porphyrins 2-Pd (∼10 μs) versus 2-Zn (∼6 ms) studied previously [9, 10]. The synthetic palladium tetrapyrroles, like native bacteriochlorophyll-derived WST9 (Tookad) and WST11, apparently have such favorable general characteristics (e.g. uptake/distribution, reactive oxygen species production) that a triplet lifetime of ≤10 μs allows sufficient encounters with O2 to attain high PDT activity [22]. A possible advantage to such short triplet lifetimes in reducing photobleaching is described below.

Photostability

The photostability properties of the four bacteriochlorins under study [BC, (NC)2BC, (NC)2BC-Zn, (NC)2BC-Pd] (Chart 4) were measured, as were those of representative synthetic bacteriochlorins whose PDT activity was investigated previously (B16, B17, B19, B22, B29, B31; Chart 2) [5-7]. The photostability properties of several standard PDT agents (Lutex, Photofrin, benzoporphyrin derivative BPD) [26] were also examined. Each compound was studied in multiple media (e.g. toluene, DMA, dimethylsulfoxide [DMSO], methanol [MeOH], ethanol [EtOH], acetonitrile [MeCN], CrEL) both in the presence and absence of ambient O2. The absorbance spectrum (350–900 nm) was monitored as a function of incident light fluence (up to 100 J cm−2) in the Qy(0,0) band, and in most cases the fluorescence spectrum was followed as well. In each case, the fraction of the initial peak Qy(0,0) absorbance as a function of light fluence was plotted. Table 1 gives the value at 100 J cm−2 for a number of bacteriochlorins in DMA and CrEL in the presence of ambient O2 and CrEL in the absence of O2. Representative spectra and plots are shown in Figures 8-11.

image

Figure 8. The absorption spectra of bacteriochlorins in media containing ambient O2 as a function of the fluence of the light delivered at the Qy maximum (720–750 nm). The following fluences (J cm−2) are shown in each Panel: 0 (black), 5 (red), 25 (blue), 50 (green) and 100 (dark yellow). The insets to Panels E and F show the corresponding fluorescence spectra elicited by excitation in the Qx band (480–530 nm) for the same fluences used during acquisition of the absorption spectra.

Download figure to PowerPoint

image

Figure 9. Photostability of bacteriochlorins and standard PDT agents in various media in the presence (closed symbols) or absence (open symbols) of atmospheric O2. The absorbance at the maximum of the NIR Qy(0,0) absorption band (720–750 nm) relative to that prior to illumination is plotted as a function of the fluence of incident light in that band. The panels are organized as follows: (A) unsubstituted and dicyanobacteriochlorins in CrEL and DMA, (B) B16 and B17 in CrEL, DMA and DMSO, (C) BC in various media, (D) B31 in various media, (E) B22, B29 and B31 in CrEL, (F) standard PDT agents. The data for the bacteriochlorins are colorized as follows: (NC)2BC-Pd (red), (NC)2BC-Zn (purple), (NC)2BC (blue), BC (green), B16 (magenta), B17 (dark cyan), B22 (light magenta), B29 (wine), B31 (navy blue). The data for the standard PDT agents are colorized as follows: Photofrin (orange), Lutex (violet), benzporphyrin derivative (BPD, sky blue). The symbols for the media are as follows: CrEL (circles), N,N-dimethylacetamide (DMA, squares), dimethylsulfoxide (DMSO, stars), acetonitrile (MeCN, down triangles), MeCN/H2O = 50/50 (pentagons), ethanol (ETOH, diamonds), methanol (MeOH, diamonds), toluene (up triangles), acetone (right triangles), benzonitrile (PhCN, down triangles), H2O (left triangles).

Download figure to PowerPoint

image

Figure 10. Photobleaching studies in aqueous micellar solutions (CrEL) of B16 with ambient O2 (A), B16 with no O2 (B) and BC with no O2 (C) showing the absorption spectrum as a function of light fluence (J cm−2) in the Qy band (∼720 nm). The insets to each panel show expanded views of the 400–600 nm region. The absorbance scale is 0–0.8 in each panel.

Download figure to PowerPoint

image

Figure 11. Photobleaching studies of B16 in N,N-dimethylacetamide (DMA) with ambient O2 showing the absorption spectra (A) and fluorescence spectra using excitation at the 490-nm Qx-band maximum (B) or at 376 nm on the long-wavelength side of the Soret band (C) as a function of light fluence (J cm2) in the Qy band (720 nm): 0 (black), 5 (red), 25 (blue), 50 (green), 75 (purple) and 100 (dark yellow). The insets to each panel show expanded views of certain regions of each spectrum. The absorbance scale is 0–0.6 in Panel (A).

Download figure to PowerPoint

Photobleaching is defined operationally as loss of absorption in the entire near-UV-to-NIR spectral region. Photobleaching can stem from a variety of sources, including reduction, oxidation, aggregation, or chemical reaction, which are referred to as photoreduction, photooxidation, photoaggregation and photodegradation respectively. Photoreduction and photooxidation may in part be reversed (by electron transfer with a species in the environment), as can photoaggregation (by dissociation), whereas photodegradation is typically irreversible. One photodegradation pathway for bacteriochlorins entails [2 + 2] cycloaddition with singlet oxygen (formed by energy transfer from the triplet excited-state bacteriochlorin) followed by ring opening to give a bilin species.

One of the goals with the dicyanobacteriochlorins (Chart 4) [11] was to improve photostability compared to the previously studied synthetic bacteriochlorins. The results bear out the success of that endeavor. The spectra in Fig. 8A and C, the plots in Fig. 9A, and the values in Table 1 show that (NC)2BC-Pd and (NC)2BC in the presence of ambient O2 are extremely stable. The solutions retain ≥90% of the NIR absorbance in CrEL and ≥83% in DMA after exposure to 100 J cm−2 of NIR light. In contrast, the previously studied bacteriochlorins B16, B17, B22, B29 and B31, as well as the parent unsubstituted compound BC in the same media, have ≤20% absorbance remaining after illumination with 100 J cm−2 (in some cases at ≤50 J cm−2) in the presence of ambient O2, and 50–80% after removal of O2 (Figs. 8 and 9 and Table 1). Thus, for CrEL and DMA solutions, (NC)2BC-Pd and (NC)2BC in the presence of ambient O2 are more photostable than the other bacteriochlorins even after stability has been markedly improved by the removal of O2 (Fig. 9). Bacteriochlorins (NC)2BC-Pd and (NC)2BC are also modestly more stable than common PDT agents [26] Photofrin and Lutex and considerably more stable than benzoporphyrin derivative (BPD) (Fig. 9A and F).

The zinc chelate (NC)2BC-Zn in DMA and CrEL both in the presence and absence of O2 is less photostable than (NC)2BC-Pd and (NC)2BC in these media containing O2 (Figs. 8B and 9A). (NC)2BC-Zn has comparable stability to the previously studied bacteriochlorins and BC in CrEL (with or without O2) but is typically more stable in DMA (Table 1). In effect, the incorporation of a zinc ion counteracts the photostability enhancement obtained upon incorporation of the dicyano groups; in contrast, a palladium ion can have a synergistic effect. These findings can be understood in terms of the redox properties of the chromophores, as described below. Collectively, the results show that the incorporation of (electron withdrawing) cyano groups at the 3,13-positions, along with the appropriate choice of metalation state, provide an excellent design for improving photostability of bacteriochlorin photosensitizers. Chlorine atoms, which also are electron withdrawing, have been incorporated at the ortho-positions of the aryl rings in synthetic meso-tetraarylbacteriochlorins to improve photostability (Chart 1) [27, 28].

Prior studies of the photodegradation of bacteriochlorophyll a and derivatives show that the process is highly solvent-dependent and requires O2. Typical photoproducts include chlorins such as 2-acetylchlorophyll a and other derivatives formed by two-electron oxidation of the macrocycle, bilin-like open-chain tetrapyrroles likely formed via addition of singlet oxygen and ring opening, and other species that have no featured absorption in the visible region [21, 29]. The geminal dimethyl groups in the synthetic bacteriochlorins are incorporated to avoid dehydrogenation to form chlorins and porphyrins [2, 3], which when observed are produced in such low yield that the expected features are cleanly resolved only via fluorescence. Bacteriochlorin B16 in DMA with ambient O2 after prolonged illumination gives such an example, showing a weak fluorescence feature at ∼640 nm (Figure 11C, inset) with no corresponding sharp feature in the absorption spectrum (Figure 11A). This example also reveals broad weak absorption and fluorescence-centered features around 600 nm that could represent open-chain tetrapyrrolic photoproducts.

Generally the photobleaching of the synthetic bacteriochlorins (bearing geminal dimethyl groups) is evidenced by a decrease in the entire near-UV-to-NIR absorption spectrum along with a similar disappearance of fluorescence. The diminution of the absorption profile is often accompanied by an apparent increase in the baseline, particularly at wavelengths shorter than 500 nm and increasing toward the near-UV region. This behavior is illustrated by the absorption spectra for (NC)2BC-Zn, BC and B16 in CrEL with ambient O2 obtained after 100 J cm−2 illumination (Figs. 8B, D and 10A; dark yellow). The featureless spectral profiles could reflect formation of nondescript photoproducts analogous to those for bacteriochlorophyll a [29] or light scattering due to photoaggregation. The analogous profiles for B16 and BC in CrEL in the absence of O2 (Fig. 10B and C) show Soret, Qx and Qy features that are decreased in intensity, broadened and bathochromically shifted after prolonged illumination. This behavior could reflect the formation of small aggregates (possibly as small as dimers) that are partially soluble and remain suspended in the micellar environment. The ability of CrEL to suppress aggregation or aid disaggregation underlies the motivation for its use as a delivery vehicle for the bacteriochlorins [6, 11]. On the other hand, aggregates may actually play a positive role in PDT by serving as reservoirs to provide fresh photosensitizer in place of those that have undergone irreversible photodegradation [29].

A relationship between photobleaching and the solubilization characteristics of the bacteriochlorin in a given medium is illustrated by the behavior of unsubstituted bacteriochlorin BC in several media in the presence of ambient O2 (Figs. 8D–F and 9C). The photostability is poor in CrEL (as noted above), MeCN and DMA but excellent in toluene, where 93% of the absorbance remains after 100 J cm−2 illumination. Thus, in the (nonpolar) medium, the highly nonpolar and thus highly soluble bacteriochlorin BC shows photostability that rivals that of dicyanobacteriochlorins (NC)2BC-Pd and (NC)2BC (Figs. 8A, C, E and 9). Solubility characteristics may also contribute to the marked medium dependence of photobleaching of the charged bacteriochlorin B31 (Chart 2 and Fig. 9E). Solvent polarity can also impact photooxidation contributions to photobleaching via the concentration of ambient O2 in the medium and stabilization of electron-transfer products. Collectively, these results and considerations suggest that the photostability of a photosensitizer may be much different in organic solvents, delivery media and the cellular milieu.

A contribution of a photoaggregation mechanism to the observed photobleaching could occur by several pathways. One pathway is an extension of excimer formation [30] whereby a bacteriochlorin in the T1 state associates with another bacteriochlorin in the ground state and so on to associate more chromophores. Such a pathway raises the possibility that photobleaching could be reduced for photosensitizers that have shorter T1 lifetimes (such as palladium tetrapyrroles) and thus provide less time for such diffusional encounters. A second pathway would involve electron transfer from a bacteriochlorin in the T1 excited state to O2 (or another electron acceptor) to form the bacteriochlorin π-cation radical. Interaction of the latter species with a ground-state bacteriochlorin would form a dimeric complex, analogous to formation of the monocation dimer of magnesium octaethylporphyrin [31], and so on to associate more chromophores. Dimer (and aggregate) formation would be enhanced because the monomer cation radical has a half-filled HOMO and thus the bonding orbital of the interacting chromophore pair would be filled and the higher energy antibonding orbital would be half filled (or empty), thereby giving a net bonding interaction that would not occur for two neutral species (where both orbitals are filled). The bacteriochlorin π-cation radical in the latter mechanism is presumably an intermediate in more common photooxidation pathways for degradation (e.g. to form chlorins). Thus, it is possible that multiple photobleaching pathways may be intertwined.

Electrochemical and molecular-orbital characteristics of the dicyanobacteriochlorins

The first oxidation potential (Eox) and first reduction potential (Ered) of the four bacteriochlorins [BC, (NC)2BC, (NC)2BC-Zn, (NC)2BC-Pd] were measured and are listed in Table 3. The table also gives the energy of the highest occupied molecular orbital (EHOMO) and the lowest unoccupied molecular orbital (ELUMO) obtained from DFT calculations for each photosensitizer. Electron densities (and energies) of the four frontier molecular orbitals for each of these four bacteriochlorins are shown in Fig. 12. Studies of a large number of chlorins have shown essentially linear relationships between Eox and EHOMO and between Ered and ELUMO [32]. As Eox becomes more positive, EHOMO becomes more negative, and the molecule is harder to oxidize. As Ered becomes less negative, ELUMO becomes more negative, and the molecule is easier to reduce. For comparison Table 3 also lists redox and MO data for two other sets of free base, zinc and palladium bacteriochlorins investigated recently [19]. The Table also gives data acquired here or previously [9] for imidazole-substituted porphyrins and tetraphenylporphyrins.

Table 3. Spectral, state-energy, redox and molecular-orbital properties of the bacteriochlorins.a
CompoundQy babs (nm)Qy c flu (nm)T1 phos (nm)S1d energy (eV)T1e energy (eV)S1 − T1 energy (eV)S0 Redox fPotentialsT1 Redox gPotentialsOrbital h Energy
Eox (V)Ered (V)T1 Eox (V)T1 Ered (V)HOMO (eV)LUMO (eV)
  1. a

    Properties of bacteriochlorins. All measured quantities were obtained at room temperature.

  2. b

    Peak wavelength of the Qy(0,0) absorption band in toluene.

  3. c

    Peak wavelength of the Qy(0,0) fluorescence band in toluene.

  4. d

    Energy of the S1 excited state calculated from the average energy positions of the Qy(0,0) absorption and fluorescence bands in the prior two columns.

  5. e

    Energy of the T1 excited state determined by phosphorescence for values indicated in normal font. The values in italics for the free base and zinc bacteriochlorins were obtained from the S1 energy by assuming the same S1 − T1 energy gap as in the palladium analogue.

  6. f

    Ground-state redox potentials measured in butyronitrile/ 0.1 m n-BuN4PF6 versus FeCp2/FeCp2+ = +0.19 V.

  7. g

    Redox potentials for the T1 excited state calculated from the ground state redox potentials and the T1 energy.

  8. h

    Frontier molecular orbital energies from DFT calculations obtained using the 6-31G* basis set except for (NC)2BC-Pd, for which the LACVP basis set was used. For PdTPP and ZnTPP the values listed are for the lowest-energy structure, wherein the sign of the torsional angle of a phenyl ring with respect to the porphyrin macrocycle alternates progressing around the porphyrin macrocycle.

  9. i

    The HOMO and LUMO energies listed for 2-Zn were calculated using the actual iodide counterion; for comparison, the values calculated using a chloride counterion are −5.58 and −2.67 eV respectively.

  10. j

    The HOMO and LUMO energies listed for 2-Pd were calculated using the actual iodide counterion; for comparison, the values calculated using a chloride counterion are −5.67 and −2.68 eV respectively.

  11. k

    From refs. [22] and [34].

DicyanoBC set
BC713716 1.74 1.20  +0.09−1.67−1.11−0.47−4.45−2.20
(NC)2BC748752 1.65 1.11  +0.60−1.10−0.51+0.01−5.22−3.10
(NC)2BC-Zn761763 1.63 1.09  +0.31−1.14−0.78−0.05−5.08−3.07
(NC)2BC-Pd75175311221.651.110.54+0.52−1.03−0.59+0.08−5.16−3.11
Other FbBCs
B16717722 1.72      −4.36−2.12
B17731737 1.68      −4.46−2.28
Ref BC-T set
B1 (Fb)736742 1.68 1.12  +0.21−1.49−0.91−0.37−4.40−2.22
B4 (Zn)749756 1.65 1.09  −0.04−1.60−1.13−0.51−4.26−2.20
B9 (Pd)73974511141.671.110.56+0.43−1.14−0.68−0.03−4.36−2.26
Ref BC-MME set
B88 (Fb)758765 1.63 1.11  +0.38−1.29−0.73−0.18−4.65−2.48
B84 (Zn)773780 1.60 1.08  +0.08−1.41−1.00−0.33−4.55−2.51
B87 (Pd)75876511181.631.110.52+0.29−1.29−0.82−0.18−4.63−2.54
Imidazole porphyrins
1-Zn5765797382.141.680.47    −4.97−2.35
2-Zn5765797382.141.680.46+0.36−1.39−1.32+0.29−5.72i−2.81i
2-Pd5475506722.251.840.42    −5.81j−2.83j
Ref porphyrins
FbTPP647650 1.911.47k0.44+0.83−1.45−0.64+0.02−4.90−2.20
ZnTPP5885997842.081.580.51+0.56−1.58−1.020−4.96−2.12
PdTPP5535606962.211.780.45+0.82−1.61−0.96+0.17−5.36−2.42
image

Figure 12. Electron densities and energies of frontier molecular orbitals. The DFT calculations employed basis sets 6-31G* for BC, (NC)2BC, and (NC)2BC-Zn and LACVP* for (NC)2BC-Pd. Effectively the same electron densities and the same MO energies (−5.74, −5.09, −3.09, −1.45 eV) were obtained for (NC)2BC-Zn using LACVP*. The LUMO + 2 for (NC)2BC-Pd has similar macrocycle electron density and energy (−1.41 eV) as the LUMO + 1 for the other molecules.

Download figure to PowerPoint

Incorporation of the two cyano groups of (NC)2BC makes the molecule harder to oxidize than BC (Eox = 0.60 versus 0.09 V; EHOMO = −5.22 versus −4.45 eV). Similarly, the two cyano groups of (NC)2BC make the molecule easier to reduce than BC (Ered = −1.10 versus −1.67 V; ELUMO = −3.10 versus −2.20 eV). Compared to (NC)2BC, zinc chelate (NC)2BC-Zn is significantly easier to oxidize (−290 mV) and only slightly harder to reduce (−40 mV); these redox-potential shifts are in concert with the relative magnitudes and signs (to less negative values) of the energy shifts in the HOMO (+140 meV) and LUMO (+3 meV). Compared to (NC)2BC, palladium chelate (NC)2BC-Pd is slightly easier to oxidize (−80 mV) and also slightly easier to reduce (+70 mV); the former tracks the small shift to less negative HOMO energy (+60 meV) whereas the latter is not paralleled by the virtually unchanged LUMO energy (−10 meV). In summary, the data in Table 3 show that of the three dicyanobacteriochlorins, (NC)2BC-Zn is the easiest to oxidize (although it is harder to oxidize than BC) and (NC)2BC-Pd is the easiest to reduce (and is also easier to reduce than BC).

Figure 13A plots the orbital energies (Table 3) versus the cell-killing LD50 values (Table 1) for the bacteriochlorins obtained using direct dilution into the complete culture medium. A qualitatively similar plot is obtained using the LD50 values when the bacteriochlorins were first encapsulated in micelles upon treatment with CrEL before dilution into the complete culture medium. Figure 13B plots the redox potentials of the T1 excited state versus LD50 values. The excited-state redox potentials are obtained from the ground-state redox potentials and the T1 energies (Table 3). The correlations with LD50 are discussed below.

image

Figure 13. Electronic properties versus cell-killing LD50 value for bacteriochlorins using direct dilution of stock solution into complete culture medium. The data points for increasing LD50 (left to right) reflecting a decrease in photosensitizer activity are (NC)2BC-Pd (a) > (NC)2BC (b) > (NC)2BC-Zn (c) > BC (d). (A) Calculated HOMO energy (open triangles) and LUMO energy (closed triangles). (B) Redox potential of the T1 excited state (Table 3). The lines in both panels reflect the linear fits to the data.

Download figure to PowerPoint

Excited-state redox properties and reactive oxygen species production

Until recently it has been generally accepted that singlet oxygen (formed by energy transfer; i.e. Type-2 photochemistry) is the major mediator of toxicity in PDT. Studies have now shown that hydroxyl radicals, superoxide and other species formed by electron transfer (Type-1 photochemistry) also play an important role in PDT activity depending on the photosensitizer and specific disease or cell type. The palladium analogues of bacteriochlorophyll a (WST9 and WST11; Chart 1) produce superoxide and hydroxyl radicals [20, 21, 33]. Synthetic free base meso-tetraarylbacteriochlorins (particularly halogenated sulfonamide analogues; Chart 1) also produce superoxide and hydroxyl radicals as well as the Type-2 product singlet oxygen [25, 27]. It has been suggested that the latter bacteriochlorins are expected to produce both Type-1 and Type-2 photoproducts whereas porphyrins give significantly less electron-transfer (Type-1) products because porphyrins are harder to oxidize than bacteriochlorins [22, 25].

The diverse free base synthetic bacteriochlorins shown in Chart 2 and free base and metal-containing dicyanobacteriochlorins (Chart 4) that are the subject of this study also produce hydroxyl radicals as well as singlet oxygen [5-7, 11]. As noted above, the most active dicyanobacteriochlorin (NC)2BC-Pd and the second most active analogue (NC)2BC give a higher ratio of hydroxyl radicals versus singlet oxygen upon illumination, whereas the reverse is true for (NC)2BC-Zn and BC. Indeed, (NC)2BC-Pd has greater PDT efficacy and a much greater propensity to produce hydroxyl radicals than (NC)2BC-Zn. In a prior study of imidazole-substituted metalloporphyrins (Chart 3), the palladium chelate (2-Pd) was similarly found to have greater PDT activity and produce much more hydroxyl radicals than the zinc analogue (2-Zn) [10].

The parallelism in relative photoactivity and hydroxyl-radical formation for the palladium versus zinc chelates of the dicyanobacteriochlorins and imidazole-substituted porphyrins gives insights into fundamental mechanistic issues. In particular, the palladium chelates of both the dicyanobacteriochlorins and porphyrins are found (or expected) to be harder to oxidize and easier to reduce than the zinc chelates (both in the ground state and the T1 states). These findings together with general comparisons of the excited-state redox properties of bacteriochlorins versus porphyrins, and the effects of the dicyano substituents of the bacteriochlorins on photoactivity (and photostability), must be considered collectively in assessing the mechanisms of production of reactive oxygen species via Type-1 photochemistry (e.g. hydroxyl radical formation) for these tetrapyrroles.

The Type-1 mechanism is virtually always thought to involve photooxidation of the photosensitizer (P) in the T1 excited state (PT) by electron transfer to O2 via Eq. (1A).

  • display math(1A)
  • display math(1B)

A number of reaction sequences have been considered for the subsequent formation of hydrogen peroxide and hydroxyl radicals from the superoxide ion (O2−•) [10, 20-22, 25, 33]. The one-electron oxidized photosensitizer is then returned to the starting condition by electron transfer from endogenous electron donor (D) [Eq. (1B)]. The endogenous electron donor has been proposed to be human serum albumin for the palladium bacteriochlorophyll a derivatives, where complexes between the two species underlie a proposed photocatalytic role of such photosensitizers via Eqs. (1A) and (1B) [21]. Photosensitizer oxidation mechanisms involving Eq. (1A) would generally imply that photoactivity should track the T1 excited-state oxidation potential of the photosensitizer. Therefore, more potent photosensitizers should be easier to oxidize in the T1 excited state than less efficacious compounds, all other factors being equal.

An alternative mechanism is photoreduction of the photosensitizer in the T1 excited state by an endogenous electron donor (D) followed by electron transfer from the photosensitizer anion to O2, to again produce superoxide ion, via Eqs. (2A) and (2B).

  • display math(2A)
  • display math(2B)

Subsequent dark reactions of superoxide ion leading to hydrogen peroxide and hydroxyl radicals and other reactive oxygen species would proceed just as in the photooxidation mechanism described above [Eq. (1A)]. Additionally, hydroxyl radicals potentially can be formed more directly from hydrogen peroxide and the reduced photosensitizer produced in Eq. 2A [10]. The photoreduction mechanism [(Eqs. (2A) plus (2B)] would also give a photocatalytic role for the photosensitizer in an analogous manner (to that noted above via the photooxidation pathway) for human serum albumin (as the electron donor) and palladium bacteriochlorophyll a derivatives (as photosensitizer) [21]. Furthermore, ascorbate has been proposed to perform a similar role as a sacrificial electron donor to tetrapyrrole photosensitizers in the PT state [(Eq. 2A)] or P+• forms [(Eq. (1A)] [22]. Photosensitizer reduction mechanisms involving Eq. (2A) would generally imply that photoactivity should track the T1 excited-state reduction potential of the photosensitizer. Therefore, more potent photosensitizers should be easier to reduce in the T1 excited state than less efficacious compounds, all other factors being equal.

We have considered these two alternative mechanisms (photosensitizer oxidation and reduction) previously in the study of imidazole-substituted palladium versus zinc porphyrins [9, 10]. The photoreduction pathway rather than the photooxidation pathway seems to best explain the findings that porphyrin 2-Pd produces more hydroxyl radicals and is more efficacious than 2-Zn (Chart 3). This follows because based on the T1 energies and ground-state redox potentials for these compounds and PdTPP and ZnTPP (Table 3 and ref. [35]), the 2-Pd is expected to be ∼150 mV easier to reduce and ∼50 mV harder to oxidize than 2-Zn in the T1 state. Similarly, for the dicyanobacteriochlorins, (NC)2BC-Pd is 130 mV easier to reduce and 190 mV harder to oxidize than (NC)2BC-Zn in the T1 state.

The latter trend can be extended to the entire set of unsubstituted and dicyanobacteriochlorins (Chart 4 and Table 3). The potential for first T1 reduction becomes less negative or more positive (the excited bacteriochlorin becomes easier to reduce) in the order BC (−0.47 V) < (NC)2BC-Zn (−0.05 V) < (NC)2BC (0.01V) < (NC)2BC-Pd (0.08V). The potential for first T1 oxidation becomes more negative (the excited bacteriochlorin becomes easier to oxidize) in the order (NC)2BC (−0.51 V) > (NC)2BC-Pd (−0.59 V) > (NC)2BC-Zn (−0.78 V) > BC (−1.11 V). Figure 13B shows the excited-state redox potentials versus LD50 (using the direct-dilution medium). A similar plot is obtained using the LD50 values obtained using CrEL delivery. Good correlations are seen for both the T1 oxidation and reduction potentials, although the implications for activity are, at first glance, reversed. Increasing PDT activity (decreasing LD50) correlates with greater ease of reduction and greater difficulty of oxidation.

The simplest interpretation of the parallel findings for the dicyanobacteriochlorins and imidazole-substituted porphyrins is that the photoreduction of the T1 excited state contributes significantly to inherent PDT efficacy (i.e. reactive oxygen species production once the photosensitizer is excited). However, at least two additional factors must be considered as follows.

  1. The absolute (and relative) rates of electron transfer will ultimately depend on the relationship between net free energy change for the excited-state electron-transfer reaction (which depends on the T1 redox potentials) and the total reorganization energy involving the photosensitizer, O2, and the medium. In other words, depending on where electron transfer for a particular photosensitizer is on this curve (i.e. the Marcus plot), an increase in the free energy driving force for the process (via a change in redox potentials or T1 energy) could either increase or decrease the rate of the reaction. The standard considerations given above consider that a greater magnitude of the free energy change implies an increased rate and yield of electron transfer.
  2. The overall PDT activity via the Type-1 mechanism may not be dictated by the rates and yields of electron transfer (or the Type-2 contribution by the rates and yields of energy transfer) if the limiting factor is photostability. In this regard, it is normally thought that photoactivity proceeds via photooxidation [Eq. (1A)] and most electron-transfer routes to photobleaching proceed via a similar mechanism. In particular, for a bacteriochlorin the excited-state oxidation process in Eq. (1A) involves one-electron oxidation of the macrocycle to form the π-cation radical, and ultimately formation of a typical chlorin photoproduct by two-electron oxidation (21,29). Even if formation of chlorin does not proceed via the bacteriochlorin π-cation radical, the general reasoning seems to be that, within the context of a Type-1 mechanism, both PDT activity and photobleaching proceed via a photooxidation (rather than photoreduction) pathway.

Electronic tuning of tetrapyrrole photosensitizers for photostability and photoactivity

Ideally, the photoactivity and photostability of a photosensitizer could be independently controlled, at least to some degree. The data in Fig. 13 and the reasoning given above suggest that such a situation could in fact contribute to the PDT efficacy of tetrapyrroles. In particular, photoactivity via the Type-1 mechanism could proceed at least in part via transient reduction of the photosensitizer [Eq. (2A)] and thus would be enhanced if the photosensitizer is made easier to reduce in the excited state. On the other hand, photobleaching likely proceeds to a significant degree by photooxidation (rather than photoreduction) and photoaggregation (perhaps via the π-cation radical), in addition to photodegradation; thus photostability would be enhanced by making the photosensitizer harder to oxidize. Rendering a tetrapyrrole chromophore harder to oxidize and easier to reduce is, in fact, what generally occurs upon the incorporation of electron-withdrawing substituents or the incorporation of a central metal ion with a high electronegativity (e.g. palladium). Such trends can be seen in the redox properties and the energies of the frontier molecular orbitals (Table 3) [22, 34, 35]. Furthermore, the strategic choice and placement of substituents allow some control over the relative oxidation and reduction potentials (via HOMO and LUMO energies), as has been shown for synthetic chlorins, bacteriochlorins and related oxophorbines and bacteriooxophorbines that incorporate the keto-bearing five-membered ring of the native photosynthetic chromophores [16, 19, 32, 12, 36, 37].

The increased photostability engendered by halogenation of the aryl rings of meso-tetraarylbacteriochlorins (e.g. TDCPBS in Chart 1) has been noted previously [22, 25]. The more difficult oxidation (and increased photostability) of palladium bacteriochlorophyll a derivatives (e.g. WST9 and WST11 in Chart 1) than bacteriochlorophyll a also has been noted [20-22, 33], and ascribed in part to the effect of the central palladium ion compared to the native magnesium ion. Additionally, the native-like bacteriochlorins bear a 3-acetyl group and the related keto group in the fused five-membered ring spanning the 12- to 15-positions (Chart 1). These groups are basically electron withdrawing in character in that they lower the HOMO and LUMO energies and make the chromophore harder to oxidize and easier to reduce than bacteriochlorins that do not contain such substituents [16, 19, 37], and the same is true of chlorin analogues [9, 36]. Thus, the photostability of WST9 and WST11 is likely provided by effects of both substituents and the electronegative palladium ion to reduce photooxidation.

Again, the effectiveness of these previously studied photosensitizers to produce hydroxyl radicals (or other Type-1 photoproducts) would be diminished compared to compounds that lacked the electron-withdrawing substituents or highly electronegative central metal if the mechanism of that process involves photooxidation [Eq. (1A)]; however, the activity would be enhanced if there is a contribution from the photoreduction pathway [Eq. (2A)]. As described above, the same is true of the bacteriochlorins bearing 3,13 substituents that are the focus of this work and our recent investigation of the PDT efficacy of these photosensitizers [11]. On the basis of shifts in MO energy and redox properties obtained upon incorporation of cyano versus acetyl groups at similar positions (Table 3) [32, 12, 36, 37], one would expect the combined effects on oxidation potentials [photostability and photoactivity via Eq. (1A)] and reduction potentials [photoactivity via Eq. (2A)] to be greater for (NC)2BC-Pd versus the palladium bacteriochlorophyll a derivatives WST9 and WST11, all other things being equal. The advantages of (NC)2BC-Pd and related synthetic bacteriochorins are [1] further enhanced photostability by the use of the geminal dimethyl groups that inhibit formation of chlorin photodegradation products and [2] the availability of macrocycle sites for additional electronic and chemical tuning.

In summary, a combination of favorable electronic effects makes the palladium bacteriochlorin (NC)2BC-Pd a photosensitizer with PDT activity (LD50 ∼25 nm) and photostability that rival or surpass those of synthetic bacteriochlorins that we have studied previously as well as many common PDT agents. The photostability of this compound is enhanced due to diminished photobleaching (via photooxidation, photoaggregation, and photodegradation), which stem from the effects of the peripheral cyano groups and central palladium ion. The photoactivity of (NC)2BC-Pd, (NC)2BC and other tetrapyrrole photosensitizers may derive in part from a contribution of a photoreduction pathway for production of lethal reactive oxygen species such as hydroxyl radicals. Collectively, the fundamental insights gained from the studies described herein may allow the design of bacteriochlorin photosensitizers that have even greater PDT efficacy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References

This work was supported by grants from the NIH (R01AI050875 to M.R.H.), and the JimmyV NCSU Cancer Therapeutics Training Program. Y.-Y.H. was supported by a grant (R41AI072854) from the National Institute of Allergy and Infectious Diseases to NIRvana Pharmaceuticals, Inc. Characterization of the photophysical, redox and molecular-orbital properties of the bacteriochlorins described herein were initially motivated by solar energy studies and supported by grants from the Division of Chemical Sciences, Geosciences and Biosciences Division, Office of Basic Energy Sciences of the U.S. Department of Energy to D.F.B. (DE-FG02-05ER15660) and D.H. (DE-FG02-05ER15661).

Conflict of Interest Statement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References

D.F.B., D.H. and J.S.L. are cofounders of NIRvana Sciences, Inc., successor to NIRvana Pharmaceuticals, which develops bacteriochlorins for clinical diagnostics. The company did not provide research support for the studies described herein.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Acknowledgements
  7. Conflict of Interest Statement
  8. References
  • 1
    Agostinis, P., K. Berg, K. A. Cengel, T. H. Foster, A. W. Girotti, S. O. Gollnick, S. M. Hahn, M. R. Hamblin, A. Juzeniene, D. Kessel, M. Korbelik, J. Moan, P. Mroz, D. Nowis, J. Piette, B. C. Wilson and J. Golab (2011) Photodynamic therapy of cancer: an update. CA Cancer J. Clin. 61, 250281.
  • 2
    Kim, H.-J. and J. S. Lindsey (2005) De novo synthesis of stable tetrahydroporphyrinic macrocycles: bacteriochlorins and a tetradehydrocorrin. J. Org. Chem. 70, 54755486.
  • 3
    Ruzié, C., M. Krayer, T. Balasubramanian and J. S. Lindsey (2008) Tailoring a bacteriochlorin building block with cationic, amphipathic, or lipophilic substituents. J. Org. Chem. 73, 58065820.
  • 4
    Taniguchi, M., D. L. Cramer, A. D. Bhise, H. L. Kee, D. F. Bocian, D. Holten and J. S. Lindsey (2008) Accessing the near-infrared spectral region with stable, synthetic, wavelength-tunable bacteriochlorins. New J. Chem. 32, 947958.
  • 5
    Mroz, P., Y.-Y. Huang, A. Szokalska, T. Zhiyentayev, S. Janjua, A.-P. Nifli, M. E. Sherwood, C. Ruzié, K. E. Borbas, D. Fan, M. Krayer, T. Balasubramanian, E. Yang, H. L. Kee, C. Kirmaier, J. R. Diers, D. F. Bocian, D. Holten, J. S. Lindsey and M. R. Hamblin (2010) Stable synthetic bacteriochlorins overcome the resistance of melanoma to photodynamic therapy. FASEB J. 24, 31603170.
  • 6
    Huang, Y.-Y., P. Mroz, T. Zhiyentayev, S. K. Sharma, T. Balasubramanian, C. Ruzié, M. Krayer, D. Fan, K. E. Borbas, E. Yang, H. L. Kee, C. Kirmaier, J. R. Diers, D. F. Bocian, D. Holten, J. S. Lindsey and M. R. Hamblin (2010) In vitro photodynamic therapy and quantitative structure-activity relationship studies with stable synthetic near-infrared-absorbing bacteriochlorin photosensitizers. J. Med. Chem. 53, 40184027.
  • 7
    Huang, L., Y.-Y. Huang, P. Mroz, G. P. Tegos, T. Zhiyentayev, S. K. Sharma, Z. Lu, T. Balasubramanian, M. Krayer, C. Ruzié, E. Yang, H. L. Kee, C. Kirmaier, J. R. Diers, D. F. Bocian, D. Holten, J. S. Lindsey and M. R. Hamblin (2010) Stable synthetic cationic bacteriochlorins as selective antimicrobial photosensitizers. Antimicrob. Agents Chemother. 54, 38343841.
  • 8
    Nyman, E. S. and P. H. Hynninen (2004) Research advances in the use of tetrapyrrolic photosensitizers for photodynamic therapy. J. Photochem. Photobiol. B: Biol. 73, 128.
  • 9
    Kee, H. L., J. Bhaumik, J. R. Diers, P. Mroz, M. R. Hamblin, D. F. Bocian, J. S. Lindsey and D. Holten (2008) Photophysical characterization of imidazolium-substituted Pd(II), In(III), and Zn(II) porphyrins as photosensitizers for photodynamic therapy. J. Photochem. Photobiol. A: Chem. 200, 346355.
  • 10
    Mroz, P., J. Bhaumik, D. K. Dogutan, Z. Aly, Z. Kamal, L. Khalid, H. L. Kee, D. F. Bocian, D. Holten, J. S. Lindsey and M. R. Hamblin (2009) Imidazole metalloporphyrins as photosensitizers for photodynamic therapy: role of molecular charge, central metal and hydroxyl radical production. Cancer Lett. 282, 6376.
  • 11
    Huang, Y.-Y., T. Balasubramanian, E. Yang, J. R. Diers, D. F. Bocian, J. S. Lindsey, D. Holten and M. R. Hamblin (2012) Stable synthetic bacteriochlorins for photodynamic therapy: role of dicyano peripheral groups, central metal substitution (2H, Zn, Pd), and Cremophor EL delivery. ChemMedChem, 7, 21552167.
  • 12
    Mass, O., M. Taniguchi, M. Ptaszek, J. W. Springer, K. M. Faries, J. R. Diers, D. F. Bocian, D. Holten and J. S. Lindsey (2011) Structural characteristics that make chlorophylls green: interplay of hydrocarbon skeleton and substituents. New J. Chem. 35, 7688.
  • 13
    Weber, G. and F. W. J. Teale (1957) Determination of the absolute quantum yield of fluorescent solutions. Trans. Faraday Soc. 53, 646655.
  • 14
    Seybold, P. G. and M. Gouterman (1969) Porphyrins XIII: fluorescence spectra and quantum yields. J. Mol. Spectrosc. 31, 113.
  • 15
    Gradyushko, A. T., A. N. Sevchenko, K. N. Solovyov and M. P. Tsvirko (1970) Energetics of photophysical processes in chlorophyll-like molecules. Photochem. Photobiol. 11, 387400.
  • 16
    Yang, E., C. Kirmaier, M. Krayer, M. Taniguchi, H.-J. Kim, J. R. Diers, D. F. Bocian, J. S. Lindsey and D. Holten (2011) Photophysical properties and electronic structure of stable, tunable synthetic bacteriochlorins: extending the features of native photosynthetic pigments. J. Phys. Chem. B 115, 1080110816.
  • 17
    Strachan, J. P., S. Gentemann, J. Seth, W. A. Kalsbeck, J. S. Lindsey, D. Holten and D. F. Bocian (1997) Effects of orbital ordering on electronic communication in multiporphyrin arrays. J. Am. Chem. Soc. 119, 1119111201.
  • 18
    Shao, Y., L. F. Molnar, Y. Jung, J. Kussmann, C. Ochsenfeld, S. T. Brown, A. T. B. Gilbert, L. V. Slipchenko, S. V. Levchenko, D. P. O'Neill, R. A. DiStasio Jr, R. C. Lochan, T. Wang, G. J. O. Beran, N. A. Besley, J. M. Herbert, C. Y. Lin, T. V. Voorhis, S. H. Chien, A. Sodt, R. P. Steele, V. A. Rassolov, P. E. Maslen, P. P. Korambath, R. D. Adamson, B. Austin, J. Baker, E. F. C. Byrd, H. Dachsel, R. J. Doerksen, A. Dreuw, B. D. Dunietz, A. D. Dutoi, T. R. Furlani, S. R. Gwaltney, A. Heyden, S. Hirata, C.-P. Hsu, G. Kedziora, R. Z. Khalliulin, P. A. Klunzinger, M. Lee, M. S. Lee, W. Liang, I. Lotan, N. Nair, B. Peters, E. I. Proynov, P. A. Pieniazek, Y. M. Rhee, J. Ritchie, E. Rosta, C. D. Sherrill, A. C. Simmonett, J. E. Subotnik, H. L. Woodcock III, W. Zhang, A. T. Bell, A. K. Chakraborty, D. M. Chipman, F. J. Keil, A. Warshel, W. J. Hehre, H. F. Schaefer III, J. Kong, A. I. Krylov, P. M. W. Gill and M. Head-Gordon (2006) Advances in methods and algorithms in a modern quantum chemistry program package. Phys. Chem. Chem. Phys. 8, 31723191.
  • 19
    Chen, C.-Y., E. Sun, D. Fan, M. Taniguchi, B. E. McDowell, E. Yang, J. R. Diers, D. F. Bocian, D. Holten and J. S. Lindsey (2012) Synthesis and physicochemical properties of metallobacteriochlorins. Inorg. Chem. 51, 94439464.
  • 20
    Vakrat-Haglili, Y., L. Weiner, V. Brumfeld, A. Brandis, Y. Salomon, B. Mcllroy, B. C. Wilson, A. Pawlak, M. Rozanowska, T. Sarna and A. Scherz (2005) The microenvironment effect on the generation of reactive oxygen species by Pd-bacteriopheophorbide. J. Am. Chem. Soc. 127, 64876497.
  • 21
    Ashur, I., R. Goldschmidt, I. Pinkas, Y. Salomon, G. Szewczyk, T. Sarna and A. Scherz (2009) Photocatalytic generation of oxygen radicals by the water-soluble bacteriochlorophyll derivative WST11, noncovalently bound to serum albumin. J. Phys. Chem. A 113, 80278037.
  • 22
    Arnaut, L. G. (2011) Design of porphyrin-based photosensitizers for photodynamic therapy. Inorg. Photochem. 63, 187233.
  • 23
    Hartwich, G., L. Fiedor, I. Simonin, E. Cmiel, W. Schäfer, D. Noy, A. Scherz and H. Scheer (1988) Metal-substituted bacteriochlorophylls. 1. Preparation and influence of metal and coordination on spectra. J. Am. Chem. Soc. 120, 36753683.
  • 24
    Takiff, L. and S. G. Boxer (1988) Phosphorescence from the primary donor in Rhodobacter sphaeroides and Rhodopseudomonas viridis reaction centers. Biochim. Biophys. Acta 932, 325334.
  • 25
    Silva, E. F. F., C. Serpa, J. M. Dabrowski, C. J. P. Monteiro, S. J. Formosinho, G. Stochel, K. Urbanska, S. Simões, M. M. Pereira and L. G. Arnaut (2010) Mechanisms of singlet-oxygen and superoxide-ion generation by porphyrins and bacteriochlorins and their implications in photodynamic therapy. Chem. Eur. J. 16, 92739286.
  • 26
    MacDonald, I. J. and T. J. Dougherty (2001) Basic principles of photodynamic therapy. J. Porphyrins Phthalocyanines 5, 105129.
  • 27
    Dabrowski, J. M., K. Urbanska, L. G. Arnaut, M. M. Pereira, A. R. Abreu, S. Simões and G. Stochel (2011) Biodistribution and photodynamic efficacy of a water-soluble, stable, halogenated bacteriochlorin against melanoma. ChemMedChem 6, 465475.
  • 28
    Dabrowski, J. M., L. G. Arnaut, M. M. Pereira, C. J. Monteiro, K. Urbanska, S. Simões and G. Stochel (2010) New halogenated water-soluble chlorin and bacteriochlorin as photostable PDT sensitizers: synthesis, spectroscopy, photophysics, and in vitro photosensitizing efficacy. ChemMedChem 5, 17701780.
  • 29
    Limantara, L., P. Koehler, B. Wilhelm, R. J. Porra and H. Scheer (2006) Photostability of bacteriochlorophyll a and derivatives: potential sensitizers for photodynamic tumor therapy. Photochem. Photobiol. 82, 770780.
  • 30
    Birks, J. B. (1970) Photophysics of Aromatic Molecules, pp 492517. Wiley-Interscience, London.
  • 31
    Fajer, J. and M. S. Davis (1978) Electron spin resonance of porphyrin π cations and anions. In The Porphyrins, Vol. IV (Edited by D. Dolphin), pp 197256. Academic Press, New York.
  • 32
    Kee, H. L., C. Kirmaier, Q. Tang, J. R. Diers, C. Muthiah, M. Taniguchi, J. K. Laha, M. Ptaszek, J. S. Lindsey, D. F. Bocian and D. Holten (2007) Effects of substituents on synthetic analogs of chlorophylls. Part 2: redox properties, optical spectra and electronic structure. Photochem. Photobiol. 83, 11251143.
  • 33
    Posen, Y., V. Kalchenko, R. Seger, A. Brandis, A. Scherz and Y. Salomon (2005) Manipulation of redox signaling in mammalian cells enabled by controlled photogeneration of reactive oxygen species. J. Cell Sci. 118, 19571969.
  • 34
    Gouterman, M. (1978) Optical spectra and electronic structure of porphyrins and related rings. In The Porphyrins, Vol. III (Edited by D. Dolphin), pp. 1165. Academic Press, New York.
  • 35
    Felton, R. H. (1978) Primary redox reactions of metalloporphyrins. In The Porphyrins, Vol. V (Edited by D. Dolphin), pp. 53125. Academic Press, New York.
  • 36
    Krayer, M., E. Yang, J. R. Diers, D. F. Bocian, D. Holten and J. S. Lindsey (2011) De novo synthesis and photophysical characterization of annulated bacteriochlorins. Mimicking and extending the properties of bacteriochlorophylls. New J. Chem. 35, 587601.
  • 37
    Springer, J. W., K. M. Faries, J. R. Diers, C. Muthia, O. Mass, H. L. Kee, C. Kirmaier, J. S. Lindsey, D. F. Bocian and D. Holten (2012) The effects of substituents on synthetic analogs of chlorophylls. Part 3: the distinctive impact of auxochromes at the 7- versus 3-positions. Photochem. Photobiol. 88, 651674.