Photochemistry and Photobiology, 2015, 91: 923–930
Effects of Silencing Heme Biosynthesis Enzymes on 5-Aminolevulinic Acid-mediated Protoporphyrin IX Fluorescence and Photodynamic Therapy Xue Yang1, Weihua Li2, Pratheeba Palasuberniam1, Kenneth A. Myers3, Chenguang Wang2 and Bin Chen*1 1
Department of Pharmaceutical Sciences, Philadelphia College of Pharmacy, University of the Sciences, Philadelphia, PA Key Laboratory of Tianjin Radiation and Molecular Nuclear Medicine; Institute of Radiation Medicine, Peking Union Medical College & Chinese Academy of Medical Sciences, Tianjin, China 3 Department of Biological Sciences, Misher College of Arts & Sciences, University of the Sciences, Philadelphia, PA 2
Received 2 February 2015, accepted 6 March 2015, DOI: 10.1111/php.12454
destruction (1). As the first biosynthetic product in heme biosynthesis pathway in mammalian cells, ALA is metabolically converted to protoporphyrin IX (PpIX), a heme precursor with good fluorescent and photosensitizing properties (2). Because PpIX is found preferentially accumulated in tumor tissues following the administration of exogenous ALA, ALA-PpIX has been evaluated for tumor fluorescence imaging and PDT treatments in a variety of tumors including brain (3), skin (4), bladder (5), cervix (6) and colon (7) tumors. Protoporphyrin IX production depends on the activity of heme biosynthesis pathway that is composed of eight enzymes localized in the mitochondrion and cytosol (8). Heme biosynthesis begins with the biosynthesis of ALA from succinyl-CoA and glycine by ALA synthase (ALAS) in mitochondria. ALA is then transported out of mitochondria to the cytosol where it is converted to coproporphyrinogen III through four consecutive enzymatic reactions catalyzed by four enzymes including porphobilinogen synthase (PBGS) and porphobilinogen deaminase (PBGD) (8). Coproporphyrinogen III is transported back into mitochondria to produce PpIX via oxidation by two peroxidases. Heme biosynthesis pathway ends with the chelation between PpIX and ferrous iron to form heme catalyzed by ferrochelatase (FECH) in mitochondria (8). Heme biosynthesis is tightly controlled by the rate limiting enzyme ALAS that is the first enzyme in the pathway, which is under negative feedback inhibition by pathway end product heme (9). Both ALAS and newly synthesized heme are localized in the mitochondrial matrix, ensuring close monitoring of cellular heme level and rapid adjustment of heme biosynthesis to maintain tissue homeostasis. Because of the tight regulation of heme biosynthesis pathway, cellular PpIX level normally is so low that its fluorescence is barely detectable. However, providing cells with excess exogenous ALA bypasses negative feedback inhibition on heme biosynthesis and results in more production of heme intermediates particularly PpIX, which enables fluorescence imaging and PDT (2). Successful application of ALA for tumor imaging and targeting relies on preferential accumulation of PpIX in tumor tissues after ALA administration. However, in spite of extensive exploration, why tumor tissues accumulate more ALA-mediated PpIX than surrounding normal tissues remains a fundamental question
ABSTRACT Aminolevulinic acid (ALA)-mediated protoporphyrin IX (PpIX) production is being explored for tumor fluorescence imaging and photodynamic therapy (PDT). As a prodrug, ALA is converted in heme biosynthesis pathway to PpIX with fluorescent and photosensitizing properties. To better understand the role of heme biosynthesis enzymes in ALA-mediated PpIX fluorescence and PDT efficacy, we used lentiviral shRNA to silence the expression of porphobilinogen synthase (PBGS), porphobilinogen deaminase (PBGD) and ferrochelatase (FECH) in SkBr3 human breast cancer cells. PBGS and PBGD are the first two cytosolic enzymes involved in PpIX biosynthesis, and FECH is the enzyme responsible for converting PpIX to heme. PpIX fluorescence was examined by flow cytometry and confocal fluorescence microscopy. Cytotoxicity was assessed after ALA-mediated PDT. Silencing PBGS or PBGD significantly reduced ALA-stimulated PpIX fluorescence, whereas silencing FECH elevated basal and ALA-stimulated PpIX fluorescence. However, compared with vector control cells, the ratio of ALA-stimulated fluorescence to basal fluorescence without ALA was significantly reduced in all knockdown cell lines. PBGS or PBGD knockdown cells exhibited significant resistance to ALA-PDT, while increased sensitivity to ALA-PDT was found in FECH knockdown cells. These results demonstrate the importance of PBGS, PBGD and FECH in ALA-mediated PpIX fluorescence and PDT efficacy.
INTRODUCTION Current cancer treatments combine selective removal of tumor tissues by surgery, preferably under imaging guidance, and preferential killing of tumor cells by anticancer agents, radiation and other modalities. Therapeutics with dual functions of tumor imaging and targeting are therefore highly desirable. Aminolevulinic acid (ALA) is a prodrug that is being explored as a fluorescence imaging probe for tumor detection and surgical dissection as well as a photodynamic therapy (PDT) agent for tumor *Corresponding author email: [email protected] (Bin Chen) © 2015 The American Society of Photobiology
923
924
Xue Yang et al.
to be addressed. Since PpIX is a metabolic product of heme biosynthesis, it is important to understand how modulating heme biosynthetic enzyme level affects PpIX production and PDT effects. Here, we used shRNA to knockdown three enzymes (PBGS, PBGD and FECH) involved in heme biosynthesis and examined the effects of silencing these enzymes on PpIX fluorescence and PDT outcomes in SkBr3 human breast cancer cells. Our results demonstrate that reducing the level of any of these three enzymes significantly affects PpIX fluorescence and PDT outcomes.
MATERIALS AND METHODS Chemicals and antibodies. Delta-aminolevulinic acid hydrochloride (ALA) from Frontier Scientific Inc. (Logan, UT) and rhodamine 123 from Life Technologies (Grand Island, NY) were dissolved in phosphatebuffered saline (PBS) solution, sterilized by passing through filters and stored in a 20°C freezer. They were directly added into cell culture medium for treatments. All primary antibodies against heme biosynthesis enzymes were purchased from Santa Cruz Biotechnology (Dallas, TX). Horseradish peroxidase-conjugated secondary antibody was obtained from Cell Signaling Inc. (Danvers, MA). Cell culture and transduction with lentiviral shRNA. SkBr3 human breast cancer cells (kindly provided by Dr. Chenguang Wang at Thomas Jefferson University) were routinely maintained in complete Dulbecco’s Modified Eagle Medium (DMEM) (Mediatech, Manassas, VA) supplemented with 9% fetal bovine serum (Atlanta Biologicals, Flowery Branch, GA) and 1% penicillin and streptomycin solution (Mediatech). Cells were cultured at 37°C in a humidified 5% CO2 incubator. SkBr3 cells were transduced with human pLKO.1 lentiviral shRNAs purchased from Open Biosystems (Lafayette, CO) to silence heme biosynthesis enzymes. For each enzyme, five different shRNA variants were evaluated and two best variants were chosen for experiment after knockdown effect verification by Western blot. Two PBGS shRNA sequences are shPBGS2 (50 -GATGAGCTACAGTGCCAAATT-30 ) and shPBGS5 (50 -TCCTGATGACATACAGCCTAT-30 ). Two PBGD shRNA sequences are shPBGD1 (50 -GCCAACTTGTTGCTGAGCAAA-30 ) and shPBGD4 (50 -CGGCTCAGATAGCATACAAGA-30 ). Two FECH shRNA seque nces are shFECH1 (50 -GCTTTGCAGATCATATTCTAA-30 ) and shFECH2 (50 -CCAAGGAGTGTGGAGTTGAAA-30 ). Lentiviruses were prepared by transient cotransfection of lentiviral vector encoding shRNA targeting sequence with viral packaging vectors into HEK 293T cells using calcium phosphate precipitation according to the manufacturer’s instruction. Briefly, HEK 293T cells were incubated with shRNA targeting different enzymes or GFP shRNA for vector control and viral packaging vectors in complete DMEM at 37°C for 6 h. After shRNAcontaining medium was removed, cells were incubated with fresh DMEM for another 36 h. Medium was then collected and filtered through a 0.45 lm filter to obtain virus particles. SkBr3 cells were infected at approximately 70% confluence in complete DMEM supplemented with 8 lg mL1 of polybrene. After 48-h incubation, the medium was changed to complete DMEM. Gene silence efficiency was verified by Western blot.
Western blot. Western blot was performed as described previously (10). Briefly, cells were lysed in NP40 lysis buffer supplemented with protease and phosphatase inhibitors. Cell lysates were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) and then electrophoretically transferred to polyvinylidene fluoride (PVDF) membranes. Blots were first incubated with primary antibodies for different heme biosynthetic enzymes (Santa Cruz Biotechnology) and then with horseradish peroxidase-conjugated secondary antibody (Cell Signaling Inc.) following the manufacturer-recommended protocol. Immunoblots were incubated with SuperSignal West Dura extended duration substrate (Thermo Scientific) and immunoreactive bands were captured with an UVP imaging system (UVP LLC, Upland, CA). NIH ImageJ software was used to quantify the density of immunoreactive bands and normalized to the corresponding actin band intensity for data analysis. Spectrofluorometric analysis. SkBr3 vector control and heme biosynthesis enzyme knockdown cells were implanted in 100 mm-cell culture dishes in complete DMEM medium, and allowed to grow for 2 days. Cells were incubated with serum-free DMEM with or without (for control) 1 mM ALA for 4 h. After the medium was removed, cells were rinsed with PBS and lysed in 1% SDS solution. Cell lysates were collected and centrifuged to obtain supernatants. Fluorescence spectrum of cell lysate supernatant was measured with a Fluoromax-3 fluorescence spectrometer (Horiba JY, Edison, NJ). PpIX fluorescence by flow cytometry. Cellular PpIX fluorescence intensity was quantified by flow cytometry. SkBr3 parental, vector control and heme biosynthesis enzyme knockdown cells were seeded in 6well plates and allowed to grow for about 24 h in complete DMEM. After serum-containing medium was removed, cells were rinsed with PBS twice and incubated with serum-free DMEM containing 1 mM ALA for ALA-treated cells or no ALA for control cells for 4 h. At the end of 4-h incubation, cells were rinsed with PBS, trypsinized, and resuspended in PBS for flow cytometry measurement. Cellular PpIX fluorescence was measured with a FACSCalibur flow cytometer (BD Biosciences) in the FL3 channel (488 nm excitation, 650 nm long pass emission), which captures the second emission band of PpIX peaked around 700 nm. About 20 000 cells were measured and recorded for each experiment. Experiments were repeated four times. Confocal fluorescence microscopic imaging of PpIX fluorescence. Protoporphyrin IX fluorescence in the SkBr3 vector control and heme biosynthesis enzyme knockdown cells was imaged with a confocal fluorescence microscope. Cells were implanted in glass bottom cell culture dishes (MatTek, Ashland, MA) and grown in complete DMEM for 2 days. Cells were rinsed with PBS twice and incubated with serum-free DMEM with or without ALA (1 mM) for 4 h. At about 30 min before due, mitochondrial marker rhodamine 123 was added into the medium to label mitochondria. After drug-containing medium was removed, cells were washed with PBS three times and incubated in serum-free DMEM for confocal imaging. Live-cell imaging was performed on a Nikon TiE (Eclipse) confocal microscope using a 609 1.40 NA oil immersion objective equipped with a CSU-X spinning disk confocal scan head (Yokogawa), a temperaturecontrolled linear encoded x, y robotic stage (ASI Technologies, Inc.), a multi-bandpass dichromatic mirror (Semrock) and bandpass filters (Chroma Technology Corp.) in an electronic filter wheel. Microscope system was appropriately set to image the fluorescence of PpIX (405 nm excitation, 700 37.5 nm emission) and rhodamine 123 (488 nm
Figure 1. Heme biosynthesis pathway. Diagram shows four mitochondrial enzymes and four cytosolic enzymes and metabolites involved in heme biosynthesis.
Photochemistry and Photobiology, 2015, 91
excitation, 525 18 nm emission). Laser illumination was provided by a 50 mW monolithic laser combiner (MLC400; Agilent Technologies) and images were acquired using a Clara interline CCD camera (Andor Technology). The microscope system, laser launch and camera were controlled using Nikon Elements software. The exposure time for PpIX and rhodamine 123 was set at 500 ms and 200 ms, respectively. Differential interference contrast (DIC) images were acquired using exposure times in the range of 100–200 ms at the same magnification (609). Fluorescence images from different channels were pseudocolored and merged to generate composite images using NIH ImageJ software. PDT treatment and cytotoxicity assay. Cells were implanted in 96-well plates in complete DMEM and allowed to adhere overnight. Cells were rinsed with PBS once and incubated with serum-free DMEM containing ALA or no ALA (for control) for 4 h. Cells were then treated with 5 mW cm2 irradiance of 633-nm light for 5 or 10 min, which results in light fluence of 1.5 and 3.0 J cm2, respectively. Light illumination was provided by a diode laser system (High Power Devices Inc., North Brunswick, NJ) coupled to a 600-lm core diameter optical fiber fitted with a microlens at the end of fiber to achieve homogeneous irradiation. Light intensity was measured with an optical power meter (Thorlabs, Inc., North Newton, NJ). Immediately after light treatment, serum-free medium was replaced with complete DMEM. Cell viability was determined at 24 h after treatment by CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (MTS assay, Promega, Madison, WI) following the manufacturer’s instruction. Experiments were repeated four times. Statistical analysis. Two-tailed Students’ t-test was used to determine statistical difference between two groups. Statistical significance was accepted at P < 0.05.
A
Fluorescence unit
B
8.0×10 5
PpIX 100 ng/ml in DMSO
7.0×10 5 6.0×10 5 5.0×10 5 4.0×10 5 3.0×10 5 2.0×10 5 1.0×10 5 0
450
550
650
750
Wavelength (nm)
Fluor escence unit
Vector shPBGS5 shPBGD1 shFECH1
1.5×10 5 1.0×10 5 5.0×10 4 2.5×10 4 0
450
550
650
750
Wavelength (nm)
ALA (1 mM, 4 h) Fluor escence unit
1.4×10 6
Vector shPBGS5 shPBGD1 shFECH1
1.2×10 6 1.0×10 6 8.0×10 5 5.0×10 5
RESULTS Silencing PBGS, PBGD and FECH reduced enzyme protein levels
Without ALA
2.0×10 5
925
Heme biosynthesis pathway includes eight enzymatic steps that take place in the mitochondrion and cytosol (Fig. 1). The first and last three steps are catalyzed by enzymes in mitochondria, which include ALAS, coproporphyrinogen III oxidase (CPOX), protoporphyrinogen III oxidase (PPOX) and FECH. The rest four steps are catalyzed by enzymes in the cytosol including PBGS, PBGD, uroporphyrinogen III synthase (UROS) and uroporphyrinogen III decarboxylase (UROD). PBGS, PBGD and FECH were knockdown in SkBr3 cells using shRNA delivered by a lentiviral vector. Five shRNA variants targeting different sequences of each enzyme gene were used. Gene knockdown efficiency was examined by Western blot analysis. For each enzyme, two shRNA-expressing cell lines with the highest reduction in enzyme protein level were selected for the experiment. Figure 2A shows the enzyme level of parental, vector control and selected PBGS (shPBGS2 and 5), PBGD (shPBGD1 and 4) and FECH (shFECH1 and 2) knockdown cells. Compared with the parental and vector control cells, knockdown cells exhibited about 50% or higher reduction in enzyme protein level.
2.5×10 5 0
450
550
650
750
Wavelength (nm) Figure 2. (A) Silencing PBGS, PBGD and FECH reduced enzyme protein levels. Protein levels of PBGS, PBGD and FECH in parental, vector control and heme biosynthesis enzyme knockdown SkBr3 cells were examined by Western blot. Band intensity relative to actin loading control was normalized to the vector control and shown above the blot. (B) Fluorescence emission spectra of PpIX in DMSO and cell lysates without or with ALA stimulation. Fluorescence spectra were taken with excitation at 400 nm.
Silencing PBGS and PBGD reduced ALA-stimulated PpIX fluorescence, whereas silencing FECH increased PpIX fluorescence without and with ALA stimulation Fluorescence spectra of vector and knockdown cell lysates were measured with a spectrofluorometer and compared with PpIX fluorescence emission (Fig. 2B). Without ALA stimulation, vector, shPBGS5 and shPBGD1 cell lysates showed a little fluorescence, whereas shFECH1 cell lysate exhibited fluorescence emission resembling PpIX fluorescence with two emission peaks at about 632 and 692 nm. Similar PpIX-like fluorescence emission at
926
Xue Yang et al. cence, while silencing FECH (shFECH1) increased PpIX fluorescence. Fluorescence increase in shFECH2 cells was not statistically significant compared with the vector control (P > 0.05). To indicate the extent of ALA-mediated PpIX fluorescence increase, ratio of ALA-stimulated fluorescence to its basal fluorescence was calculated. For parental and vector control cells, ALA caused 50–60-folds increase in fluorescence (Fig. 3D). This ratio was significantly decreased in all knockdown cell lines. Protoporphyrin IX fluorescence in vector control and shRNAknockdown cells was imaged by a confocal microscope. To determine the relationship between PpIX localization and mitochondria, cells were labeled with mitochondrial dye rhodamine 123. Without ALA stimulation, fluorescence was not visible in vector control, shPBGS5, shPBGD1 cells (Fig. 4). In agreement with fluorescence spectrum and flow cytometry data, silencing FECH increased PpIX level, resulting in visible PpIX fluorescence in shFECH1 cells even without ALA stimulation. ALA stimulation led to visible PpIX fluorescence in all cells (Fig. 5),
much higher intensity was recorded in cell lysates after ALA stimulation. To quantify fluorescence intensity, cells were analyzed with a flow cytometer in the FL3 channel to record fluorescence emission above 650 nm. Based on fluorescence spectra of cell lysates, fluorescence detected within this range was mainly attributed to PpIX emission (Fig. 2B). Representative fluorescence histograms in parental, vector control and shRNA-knockdown SkBr3 cells with or without ALA stimulation are shown in Fig. 3A. Results of four independent experiments are summarized in Fig. 3B–D. Parental and vector control cells displayed low basal fluorescence without ALA stimulation. Knockdown of PBGS and PBGD had little effect on basal fluorescence (Fig. 3B). However, silencing FECH significantly enhanced basal fluorescence. Although incubation with ALA (1 mM) for 4 h greatly increased fluorescence in all cell lines (Fig. 3C), significant differences in fluorescence intensity were found between vector control and heme enzyme knockdown cells. Particularly, knockdown of PBGS (shPBGS2 and 5) and PBGD (shPBGD1 and 4) resulted in reduced PpIX fluores-
A
B
Basal fluorescence without ALA
*
Fluorescence (FL3)
200
150
** 100
50
H C FE sh
Fluorescence with ALA (1 mM, 4 h)
1600
*
1400
Fluorescence (FL3)
2
1 H C
D
FE
G sh
sh
P
B sh
sh
P
P
B
G
G
D
4
1
5 S
2 B
G B P sh
C
S
rl ct to r
ec V
P
ar
en ta l
0
1200 1000 800
**
600 400
**
200
**
**
H 2 C
H 1 C
sh FE
4 G D B
sh FE
1 G D B
sh P
sh P
5 B
G S
2 G S B
sh P
ct rl
sh P
D
V ec to r
P
ar en ta l
0
Ratio of ALA-stimulated to basal fluorescence 70
Fold increase
60 50 40 30 20 10
* **
**
*
**
**
P ar en ta l V ec to r ct rl sh P B G S 2 sh P B G S 5 sh P B G D 1 sh P B G D 4 sh FE C H 1 sh FE C H 2
0
Figure 3. Effects of silencing PBGS, PBGD and FECH on PpIX fluorescence. Parental, vector control and heme biosynthesis enzyme knockdown SkBr3 cells were incubated in serum-free DMEM with or without 1 mM ALA for 4 h and PpIX fluorescence was measured by a flow cytometer. (A) PpIX fluorescence histogram from a representative experiment. The number shown in each histogram is mean PpIX fluorescence intensity. (B) Fluorescence without ALA stimulation. (C) Fluorescence with ALA (1 mM, 4 h) stimulation. (D) The ratio of ALA-stimulated fluorescence to basal fluorescence without ALA stimulation. Data represent mean SE from four independent experiments. *P < 0.05, **P < 0.01, compared with vector control.
Photochemistry and Photobiology, 2015, 91
927
Figure 4. Confocal fluorescence microscopic images of vector control and heme biosynthesis enzyme knockdown SkBr3 cells without ALA stimulation. Mitochondria were labeled by incubating cells with rhodamine 123 (250 ng mL1) for 30 min. Bars, 10 lm.
indicating enhanced PpIX production after ALA supplement. However, compared with vector control cells, fluorescence in shPBGS5 and shPBGD1 cells was much weaker, whereas fluorescence in shFECH1 cells appeared stronger. It was noted that some cells in confocal images appeared brighter than surrounding cells. This heterogeneity in fluorescence was also shown in flow cytometry histograms where cell populations with different fluorescence intensity were observed (Fig. 3A), which suggests the existence of variation in knockdown efficiency. Although being more diffused than mitochondrial marker fluorescence, PpIX fluorescence overlapped with mitochondrial marker fluorescence, indicating its localization in mitochondria (Fig. 5). Silencing PBGS and PBGD reduced cell sensitivity to ALAmediated PDT, whereas silencing FECH increased cell sensitivity to ALA-mediated PDT We next treated vector control, shPBGS5, shPBGD1 and shFECH1 SkBr3 cells with light after 4-h incubation with ALA and
evaluated cytotoxicity at 24 h after ALA-PDT (Fig. 6). Four different ALA-PDT doses were assessed. A low PDT dose (0.5 mM ALA, 1.5 J cm2) significantly reduced cell viability only in shFECH1 cells (P < 0.01), but not in vector control, shPBGS5 and shPBGD1 cells (P > 0.05), indicating that silencing FECH expression increases cell response to ALA-PDT. This is further supported by the finding that shFECH1 had significantly lower cell viability than vector control cells (P < 0.05) after PDT with a higher light dose (0.5 mM ALA, 3.0 J cm2) or a higher ALA dose (1.0 mM ALA, 1.5 J cm2). However, PDT with the highest light and ALA doses (1.0 mM ALA, 3.0 J cm2) reduced cell viability to a similarly low level (below 10%) in both vector control and shFECH1 cells. In contrast to shFECH1 which showed increased sensitivity to ALA-PDT, shPBGS5 and shPBGD1 cells exhibited significantly reduced sensitivity to PDT with varied ALA and light doses. ALA-PDT at lower doses had little effect on the viability of shPBGS5 and shPBGD1 cells and cells were still able to maintain more than 80% viability even after PDT with the highest light and ALA doses (Fig. 6).
928
Xue Yang et al.
Figure 5. Confocal fluorescence microscopic images after ALA stimulation. Vector control and heme biosynthesis enzyme knockdown SkBr3 cells were incubated in serum-free DMEM with 1 mM ALA for 4 h and imaged by a confocal microscope. Mitochondria were labeled by incubating cells with rhodamine 123 (250 ng mL1) for 30 min. Bars, 10 lm.
DISCUSSION Aminolevulinic acid-mediated tumor detection and treatment critically depend on preferential PpIX accumulation in tumor tissues following ALA administration. Through four consecutive biosynthetic steps in the cytosol and two additional reactions in mitochondria catalyzed by a total of six heme biosynthesis enzymes, ALA is converted to PpIX (8). Under the catalysis of FECH, PpIX is chelated with ferrous iron to form heme, which does not have fluorescent and photosensitizing property. Thus, enzymes involved in PpIX biosynthesis and bioconversion collectively determines ALA-mediated PpIX accumulation. We silenced the expression of the first two cytosolic enzymes in ALA-PpIX biosynthesis (PBGS, PBGD) and the mitochondrial enzyme responsible for PpIX bioconversion (FECH) in an attempt to evaluate the role of these three enzymes in ALA-PpIX accumulation and cytotoxicity to ALA-PDT. SkBr3 human breast cancer cell line was used because this cell line showed robust PpIX production
after ALA stimulation in our previous study (11). Our present data demonstrate that, although all three enzymes are important for ALA-mediated PpIX production and PDT response, silencing PBGS/PBGD and FECH have completely opposite effects. As the first cytosolic enzyme in PpIX biosynthesis, PBGS converts two molecules of endogenous and exogenous ALA to a monopyrrole metabolite porphobilinogen (PBG) (8). Because the reaction involves the removal of a water molecule, PBGS is also known as ALA dehydratase (ALAD). Four molecules of PBG are subsequently deaminated by PBGD to form a tetrapyrrole metabolite preuroporphyrinogen. Since preuroporphyrinogen is also called hydroxymethylbilane, PBGD is also known as hydroxymethylbilane synthase (HMBS). Here, we found that knockdown of PBGS or PBGD significantly decreased ALAstimulated PpIX fluorescence (Fig. 3). Moreover, comparing ALA-PpIX fluorescence between cell lines with different knockdown efficiency revealed a correlation between ALA-mediated PpIX fluorescence reduction and the extent of enzyme gene
Photochemistry and Photobiology, 2015, 91 PDT (0.5 mM ALA 4 h, 3 J/cm 2)
100
^^
80 60 40 20 0
Cell viability (% control)
Cell viability (% control)
PDT (0.5 mM ALA 4 h, 1.5 J/cm 2) 120
Vector shPBGS5 shPBGD1 shFECH1
120
100 80 60 40
**
^
*
^^
20 0
Vector shPBGS5 shPBGD1 shFECH1
**
80 60
^^^
* ^^
40 20 0
Vector shPBGS5 shPBGD1 shFECH1
PDT (1 mM ALA 4 h, 3 J/cm 2) Cell viability (% control)
Cell viability (% control)
*
***
100
PDT (1 mM ALA 4 h, 1.5 J/cm 2) 120
929
120
**
100 80
*** ^
60 40 20 0
^^^
^^^
Vector shPBGS5 shPBGD1 shFECH1
Figure 6. Effects of silencing heme biosynthesis enzyme PBGS, PBGD and FECH on ALA-mediated PDT. Vector, shPBGS5, shPBGD1 and shFECH1 cells were treated with ALA-PDT as indicated and cell viability was examined by the MTS assay at 24 h after treatment. Data represent mean SD from four independent experiments. ^P < 0.05, ^^P < 0.01, ^^^P < 0.001 compared with light only control without ALA. *P < 0.05, **P < 0.01, ***P < 0.001 compared with vector cells.
silencing. For example, shPBGS5 cell line had a higher knockdown efficiency than shPBGS2 (Fig. 2A). It displayed lower ALA-PpIX fluorescence than shPBGS2 (Fig. 3). Similarly, shPBGD1 cell line had better knockdown effect than shPBGD4. It, therefore, showed lower fluorescence than shPBGD4 line. Because of reduced PpIX production, PBGS and PBGD knockdown cells exhibited a little or no response to ALA-PDT (Fig. 6). Thus, our present results in SkBr3 breast cancer cells together with previous studies in K562 leukemia cells (12,13) demonstrate that both PBGS and PBGD are critically important for ALA-mediated PpIX production and reducing the level/activity of any of these enzymes decreases PpIX fluorescence and PDT efficacy. Ferrochelatase catalyzes the insertion of ferrous iron into PpIX, thereby reducing PpIX level by converting it to nonfluorescent heme (8). Consistent with previous studies in human colon (14), urothelial (15) and glioma (16) cancer cells, we found in SkBr3 cells that silencing FECH expression significantly increased ALA-stimulated PpIX fluorescence. As a result, FECH knockdown cells were much more sensitive to ALA-PDT than vector control cells. Considering that a variety of human breast cancer cell lines have shown preferential ALA-PpIX accumulation (17) and ALA-PpIX fluorescence has been explored for the detection of early neoplastic and metastatic mammary tumors in transgenic mice (18) as well as primary and metastatic human breast cancers (19,20), our present finding implies that inhibiting FECH may improve ALA-based tumor detection and treatment for breast cancers by boosting more PpIX accumulation. As a matter of fact, iron-chelating agents, which abolish FECH activity by removing ferrous iron, have shown effectiveness in enhancing ALA-PpIX fluorescence and PDT effect in skin
cancers (21). However, because we found that knockdown of FECH increased basal fluorescence without ALA as well, future studies are needed to determine whether pharmacologically inhibiting FECH increases PpIX fluorescence in normal cells and, if so, how much PpIX increase in normal cells compared to tumor cells. To make it an effective approach for improving ALA-based fluorescence imaging and PDT, FECH inhibitors should possess preferential effects in tumor cells so that a greater difference in PpIX between tumor and surrounding normal cells can be achieved. It is interesting to note that, compared with parental and vector control cells, the ratio of ALA-stimulated fluorescence to basal fluorescence was significantly reduced in all lines of knockdown cells. As an indication of the extent of ALA-stimulated PpIX production, this ratio is likely dependent on the overall enzymatic capacity of six enzymes involved in PpIX biosynthesis (from PBGS to PPOX) and the activity of PpIX bioconversion enzyme FECH. As expected, knockdown of PpIX biosynthesis enzymes such as PBGS and PBGD decreased the enzyme level/activity, therefore resulting in reduced capacity of ALA-PpIX biosynthesis. Surprisingly, knockdown of FECH involved in PpIX bioconversion also decreased the ratio of ALA-stimulated PpIX to basal fluorescence. There are two possible explanations for this paradoxical finding. First, such a reduction may indicate the saturation of PpIX biosynthesis enzyme activity. Silencing FECH inhibits the conversion of PpIX to heme, which causes the accumulation of porphyrin metabolites (particularly its substrate PpIX) in heme biosynthesis pathway and activates heme biosynthesis due to the release of negative feedback inhibition on heme biosynthesis pathway (22). Imposing on this already activated PpIX biosynthesis pathway with
930
Xue Yang et al.
additional load of exogenous ALA may saturate pathway activity, thereby failing to obtain more ALA-stimulated PpIX production. Second, reduction in the ratio of ALA-PpIX over basal PpIX production in FECH knockdown cells may suggest an interaction between FECH and upstream enzymes involved in PpIX biosynthesis. In this case, PpIX biosynthesis enzyme functions may be inhibited in the absence of FECH enzyme. Studies are undergoing to determine whether such interactions exist. In summary, we evaluated the role of three heme biosynthesis enzymes PBGS, PBGD and FECH in ALA-mediated PpIX fluorescence and PDT efficacy in SkBr3 human breast cancer cells. We found that knockdown of PBGS or PBGD significantly reduced ALA-stimulated PpIX fluorescence and rendered cell resistance to ALA-PDT. In contrast, silencing FECH greatly elevated basal and ALA-stimulated PpIX fluorescence and sensitized cells to ALA-PDT. These results demonstrate that PBGS, PBGD and FECH are important for ALA-stimulated PpIX fluorescence and suggest that inhibiting FECH may enhance ALAbased tumor fluorescence detection and PDT efficacy in breast cancers. Future studies should focus on comparing the effects of FECH inhibition on ALA-PpIX in tumor versus normal cells and determining whether FECH inhibition affects PpIX biosynthesis enzymes.
7.
8. 9. 10.
11.
12.
13. 14.
Acknowledgement—This study was supported in part by Research Scholar Grant RSG-10-035-01-CCE from the American Cancer Society.
15.
REFERENCES
16.
1. Nokes, B., M. Apel, C. Jones, G. Brown and J. E. Lang (2013) Aminolevulinic acid (ALA): Photodynamic detection and potential therapeutic applications. J. Surg. Res. 181, 262–271. 2. Peng, Q., T. Warloe, K. Berg, J. Moan, M. Kongshaug, K. E. Giercksky and J. M. Nesland (1997) 5-Aminolevulinic acid-based photodynamic therapy. Clinical research and future challenges. Cancer 79, 2282–2308. 3. Stummer, W., U. Pichlmeier, T. Meinel, O. D. Wiestler, F. Zanella, H. J. Reulen and A. L.-G. S. Group (2006) Fluorescence-guided surgery with 5-aminolevulinic acid for resection of malignant glioma: A randomised controlled multicentre phase III trial. Lancet Oncol. 7, 392–401. 4. Smits, T., A. I. van Laarhoven, A. Staassen, P. C. van de Kerkhof, P. E. van Erp and M. J. Gerritsen (2009) Induction of protoporphyrin IX by aminolaevulinic acid in actinic keratosis, psoriasis and normal skin: Preferential porphyrin enrichment in differentiated cells. Br. J. Dermatol. 160, 849–857. 5. Jocham, D., F. Witjes, S. Wagner, B. Zeylemaker, J. van Moorselaar, M. O. Grimm, R. Muschter, G. Popken, F. Konig, R. Knuchel and K. H. Kurth (2005) Improved detection and treatment of bladder cancer using hexaminolevulinate imaging: A prospective, phase III multicenter study. J. Urol. 174, 862–866. 6. Barnett, A. A., J. C. Haller, F. Cairnduff, G. Lane, S. B. Brown and D. J. Roberts (2003) A randomised, double-blind, placebo-controlled trial of photodynamic therapy using 5-aminolaevulinic acid for the
17.
18.
19.
20. 21.
22.
treatment of cervical intraepithelial neoplasia. Int. J. Cancer 103, 829–832. Kato, S., J. Kawamura, K. Kawada, S. Hasegawa and Y. Sakai (2012) Fluorescence diagnosis of metastatic lymph nodes using 5aminolevulinic acid (5-ALA) in a mouse model of colon cancer. J. Surg. Res. 176, 430–436. Ponka, P. (1999) Cell biology of heme. Am. J. Med. Sci. 318, 241– 256. Heinemann, I. U., M. Jahn and D. Jahn (2008) The biochemistry of heme biosynthesis. Arch. Biochem. Biophys. 474, 238–251. Fateye, B., W. Li, C. Wang and B. Chen (2012) Combination of phosphatidylinositol 3-kinases pathway inhibitor and photodynamic therapy in endothelial and tumor cells. Photochem. Photobiol. 88, 1265–1272. Gibbs, S. L., B. Chen, J. A. O’Hara, P. J. Hoopes, T. Hasan and B. W. Pogue (2006) Protoporphyrin IX level correlates with number of mitochondria, but increase in production correlates with tumor cell size. Photochem. Photobiol. 82, 1334–1341. Feuerstein, T., A. Schauder and Z. Malik (2009) Silencing of ALA dehydratase affects ALA-photodynamic therapy efficacy in K562 erythroleukemic cells. Photochem. Photobiol. Sci. 8, 1461– 1466. Schauder, A., T. Feuerstein and Z. Malik (2011) The centrality of PBGD expression levels on ALA-PDT efficacy. Photochem. Photobiol. Sci. 10, 1310–1317. Kemmner, W., K. Wan, S. Ruttinger, B. Ebert, R. Macdonald, U. Klamm and K. T. Moesta (2008) Silencing of human ferrochelatase causes abundant protoporphyrin-IX accumulation in colon cancer. FASEB J. 22, 500–509. Miyake, M., M. Ishii, K. Kawashima, T. Kodama, K. Sugano, K. Fujimoto and Y. Hirao (2009) siRNA-mediated knockdown of the heme synthesis and degradation pathways: Modulation of treatment effect of 5-aminolevulinic acid-based photodynamic therapy in urothelial cancer cell lines. Photochem. Photobiol. 85, 1020–1027. Teng, L., M. Nakada, S. G. Zhao, Y. Endo, N. Furuyama, E. Nambu, I. V. Pyko, Y. Hayashi and J. I. Hamada (2011) Silencing of ferrochelatase enhances 5-aminolevulinic acid-based fluorescence and photodynamic therapy efficacy. Br. J. Cancer 104, 798–807. Millon, S. R., J. H. Ostrander, S. Yazdanfar, J. Q. Brown, J. E. Bender, A. Rajeha and N. Ramanujam (2010) Preferential accumulation of 5-aminolevulinic acid-induced protoporphyrin IX in breast cancer: A comprehensive study on six breast cell lines with varying phenotypes. J. Biomed. Opt. 15, 018002. Dorward, A. M., K. S. Fancher, T. M. Duffy, W. G. Beamer and H. Walt (2005) Early neoplastic and metastatic mammary tumours of transgenic mice detected by 5-aminolevulinic acid-stimulated protoporphyrin IX accumulation. Br. J. Cancer 93, 1137–1143. Frei, K. A., H. M. Bonel, H. Frick, H. Walt and R. A. Steiner (2004) Photodynamic detection of diseased axillary sentinel lymph node after oral application of aminolevulinic acid in patients with breast cancer. Br. J. Cancer 90, 805–809. Ladner, D. P., R. A. Steiner, J. Allemann, U. Haller and H. Walt (2001) Photodynamic diagnosis of breast tumours after oral application of aminolevulinic acid. Br. J. Cancer 84, 33–37. Campbell, S. M., C. A. Morton, R. Alyahya, S. Horton, A. Pye and A. Curnow (2008) Clinical investigation of the novel iron-chelating agent, CP94, to enhance topical photodynamic therapy of nodular basal cell carcinoma. Br. J. Dermatol. 159, 387–393. Sakaino, M., T. Kataoka and S. Taketani (2009) Post-transcriptional regulation of the expression of ferrochelatase by its variant mRNA. J. Biochem. 145, 733–738.
Effects of Silencing Heme Biosynthesis Enzymes on 5-Aminolevulinic Acid-mediated Protoporphyrin IX Fluorescence and Photodynamic Therapy Xue Yang1, Weihua Li2, Pratheeba Palasuberniam1, Kenneth A. Myers3, Chenguang Wang2 and Bin Chen*1 1
Department of Pharmaceutical Sciences, Philadelphia College of Pharmacy, University of the Sciences, Philadelphia, PA Key Laboratory of Tianjin Radiation and Molecular Nuclear Medicine; Institute of Radiation Medicine, Peking Union Medical College & Chinese Academy of Medical Sciences, Tianjin, China 3 Department of Biological Sciences, Misher College of Arts & Sciences, University of the Sciences, Philadelphia, PA 2
Received 2 February 2015, accepted 6 March 2015, DOI: 10.1111/php.12454
destruction (1). As the first biosynthetic product in heme biosynthesis pathway in mammalian cells, ALA is metabolically converted to protoporphyrin IX (PpIX), a heme precursor with good fluorescent and photosensitizing properties (2). Because PpIX is found preferentially accumulated in tumor tissues following the administration of exogenous ALA, ALA-PpIX has been evaluated for tumor fluorescence imaging and PDT treatments in a variety of tumors including brain (3), skin (4), bladder (5), cervix (6) and colon (7) tumors. Protoporphyrin IX production depends on the activity of heme biosynthesis pathway that is composed of eight enzymes localized in the mitochondrion and cytosol (8). Heme biosynthesis begins with the biosynthesis of ALA from succinyl-CoA and glycine by ALA synthase (ALAS) in mitochondria. ALA is then transported out of mitochondria to the cytosol where it is converted to coproporphyrinogen III through four consecutive enzymatic reactions catalyzed by four enzymes including porphobilinogen synthase (PBGS) and porphobilinogen deaminase (PBGD) (8). Coproporphyrinogen III is transported back into mitochondria to produce PpIX via oxidation by two peroxidases. Heme biosynthesis pathway ends with the chelation between PpIX and ferrous iron to form heme catalyzed by ferrochelatase (FECH) in mitochondria (8). Heme biosynthesis is tightly controlled by the rate limiting enzyme ALAS that is the first enzyme in the pathway, which is under negative feedback inhibition by pathway end product heme (9). Both ALAS and newly synthesized heme are localized in the mitochondrial matrix, ensuring close monitoring of cellular heme level and rapid adjustment of heme biosynthesis to maintain tissue homeostasis. Because of the tight regulation of heme biosynthesis pathway, cellular PpIX level normally is so low that its fluorescence is barely detectable. However, providing cells with excess exogenous ALA bypasses negative feedback inhibition on heme biosynthesis and results in more production of heme intermediates particularly PpIX, which enables fluorescence imaging and PDT (2). Successful application of ALA for tumor imaging and targeting relies on preferential accumulation of PpIX in tumor tissues after ALA administration. However, in spite of extensive exploration, why tumor tissues accumulate more ALA-mediated PpIX than surrounding normal tissues remains a fundamental question
ABSTRACT Aminolevulinic acid (ALA)-mediated protoporphyrin IX (PpIX) production is being explored for tumor fluorescence imaging and photodynamic therapy (PDT). As a prodrug, ALA is converted in heme biosynthesis pathway to PpIX with fluorescent and photosensitizing properties. To better understand the role of heme biosynthesis enzymes in ALA-mediated PpIX fluorescence and PDT efficacy, we used lentiviral shRNA to silence the expression of porphobilinogen synthase (PBGS), porphobilinogen deaminase (PBGD) and ferrochelatase (FECH) in SkBr3 human breast cancer cells. PBGS and PBGD are the first two cytosolic enzymes involved in PpIX biosynthesis, and FECH is the enzyme responsible for converting PpIX to heme. PpIX fluorescence was examined by flow cytometry and confocal fluorescence microscopy. Cytotoxicity was assessed after ALA-mediated PDT. Silencing PBGS or PBGD significantly reduced ALA-stimulated PpIX fluorescence, whereas silencing FECH elevated basal and ALA-stimulated PpIX fluorescence. However, compared with vector control cells, the ratio of ALA-stimulated fluorescence to basal fluorescence without ALA was significantly reduced in all knockdown cell lines. PBGS or PBGD knockdown cells exhibited significant resistance to ALA-PDT, while increased sensitivity to ALA-PDT was found in FECH knockdown cells. These results demonstrate the importance of PBGS, PBGD and FECH in ALA-mediated PpIX fluorescence and PDT efficacy.
INTRODUCTION Current cancer treatments combine selective removal of tumor tissues by surgery, preferably under imaging guidance, and preferential killing of tumor cells by anticancer agents, radiation and other modalities. Therapeutics with dual functions of tumor imaging and targeting are therefore highly desirable. Aminolevulinic acid (ALA) is a prodrug that is being explored as a fluorescence imaging probe for tumor detection and surgical dissection as well as a photodynamic therapy (PDT) agent for tumor *Corresponding author email: [email protected] (Bin Chen) © 2015 The American Society of Photobiology
923
924
Xue Yang et al.
to be addressed. Since PpIX is a metabolic product of heme biosynthesis, it is important to understand how modulating heme biosynthetic enzyme level affects PpIX production and PDT effects. Here, we used shRNA to knockdown three enzymes (PBGS, PBGD and FECH) involved in heme biosynthesis and examined the effects of silencing these enzymes on PpIX fluorescence and PDT outcomes in SkBr3 human breast cancer cells. Our results demonstrate that reducing the level of any of these three enzymes significantly affects PpIX fluorescence and PDT outcomes.
MATERIALS AND METHODS Chemicals and antibodies. Delta-aminolevulinic acid hydrochloride (ALA) from Frontier Scientific Inc. (Logan, UT) and rhodamine 123 from Life Technologies (Grand Island, NY) were dissolved in phosphatebuffered saline (PBS) solution, sterilized by passing through filters and stored in a 20°C freezer. They were directly added into cell culture medium for treatments. All primary antibodies against heme biosynthesis enzymes were purchased from Santa Cruz Biotechnology (Dallas, TX). Horseradish peroxidase-conjugated secondary antibody was obtained from Cell Signaling Inc. (Danvers, MA). Cell culture and transduction with lentiviral shRNA. SkBr3 human breast cancer cells (kindly provided by Dr. Chenguang Wang at Thomas Jefferson University) were routinely maintained in complete Dulbecco’s Modified Eagle Medium (DMEM) (Mediatech, Manassas, VA) supplemented with 9% fetal bovine serum (Atlanta Biologicals, Flowery Branch, GA) and 1% penicillin and streptomycin solution (Mediatech). Cells were cultured at 37°C in a humidified 5% CO2 incubator. SkBr3 cells were transduced with human pLKO.1 lentiviral shRNAs purchased from Open Biosystems (Lafayette, CO) to silence heme biosynthesis enzymes. For each enzyme, five different shRNA variants were evaluated and two best variants were chosen for experiment after knockdown effect verification by Western blot. Two PBGS shRNA sequences are shPBGS2 (50 -GATGAGCTACAGTGCCAAATT-30 ) and shPBGS5 (50 -TCCTGATGACATACAGCCTAT-30 ). Two PBGD shRNA sequences are shPBGD1 (50 -GCCAACTTGTTGCTGAGCAAA-30 ) and shPBGD4 (50 -CGGCTCAGATAGCATACAAGA-30 ). Two FECH shRNA seque nces are shFECH1 (50 -GCTTTGCAGATCATATTCTAA-30 ) and shFECH2 (50 -CCAAGGAGTGTGGAGTTGAAA-30 ). Lentiviruses were prepared by transient cotransfection of lentiviral vector encoding shRNA targeting sequence with viral packaging vectors into HEK 293T cells using calcium phosphate precipitation according to the manufacturer’s instruction. Briefly, HEK 293T cells were incubated with shRNA targeting different enzymes or GFP shRNA for vector control and viral packaging vectors in complete DMEM at 37°C for 6 h. After shRNAcontaining medium was removed, cells were incubated with fresh DMEM for another 36 h. Medium was then collected and filtered through a 0.45 lm filter to obtain virus particles. SkBr3 cells were infected at approximately 70% confluence in complete DMEM supplemented with 8 lg mL1 of polybrene. After 48-h incubation, the medium was changed to complete DMEM. Gene silence efficiency was verified by Western blot.
Western blot. Western blot was performed as described previously (10). Briefly, cells were lysed in NP40 lysis buffer supplemented with protease and phosphatase inhibitors. Cell lysates were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) and then electrophoretically transferred to polyvinylidene fluoride (PVDF) membranes. Blots were first incubated with primary antibodies for different heme biosynthetic enzymes (Santa Cruz Biotechnology) and then with horseradish peroxidase-conjugated secondary antibody (Cell Signaling Inc.) following the manufacturer-recommended protocol. Immunoblots were incubated with SuperSignal West Dura extended duration substrate (Thermo Scientific) and immunoreactive bands were captured with an UVP imaging system (UVP LLC, Upland, CA). NIH ImageJ software was used to quantify the density of immunoreactive bands and normalized to the corresponding actin band intensity for data analysis. Spectrofluorometric analysis. SkBr3 vector control and heme biosynthesis enzyme knockdown cells were implanted in 100 mm-cell culture dishes in complete DMEM medium, and allowed to grow for 2 days. Cells were incubated with serum-free DMEM with or without (for control) 1 mM ALA for 4 h. After the medium was removed, cells were rinsed with PBS and lysed in 1% SDS solution. Cell lysates were collected and centrifuged to obtain supernatants. Fluorescence spectrum of cell lysate supernatant was measured with a Fluoromax-3 fluorescence spectrometer (Horiba JY, Edison, NJ). PpIX fluorescence by flow cytometry. Cellular PpIX fluorescence intensity was quantified by flow cytometry. SkBr3 parental, vector control and heme biosynthesis enzyme knockdown cells were seeded in 6well plates and allowed to grow for about 24 h in complete DMEM. After serum-containing medium was removed, cells were rinsed with PBS twice and incubated with serum-free DMEM containing 1 mM ALA for ALA-treated cells or no ALA for control cells for 4 h. At the end of 4-h incubation, cells were rinsed with PBS, trypsinized, and resuspended in PBS for flow cytometry measurement. Cellular PpIX fluorescence was measured with a FACSCalibur flow cytometer (BD Biosciences) in the FL3 channel (488 nm excitation, 650 nm long pass emission), which captures the second emission band of PpIX peaked around 700 nm. About 20 000 cells were measured and recorded for each experiment. Experiments were repeated four times. Confocal fluorescence microscopic imaging of PpIX fluorescence. Protoporphyrin IX fluorescence in the SkBr3 vector control and heme biosynthesis enzyme knockdown cells was imaged with a confocal fluorescence microscope. Cells were implanted in glass bottom cell culture dishes (MatTek, Ashland, MA) and grown in complete DMEM for 2 days. Cells were rinsed with PBS twice and incubated with serum-free DMEM with or without ALA (1 mM) for 4 h. At about 30 min before due, mitochondrial marker rhodamine 123 was added into the medium to label mitochondria. After drug-containing medium was removed, cells were washed with PBS three times and incubated in serum-free DMEM for confocal imaging. Live-cell imaging was performed on a Nikon TiE (Eclipse) confocal microscope using a 609 1.40 NA oil immersion objective equipped with a CSU-X spinning disk confocal scan head (Yokogawa), a temperaturecontrolled linear encoded x, y robotic stage (ASI Technologies, Inc.), a multi-bandpass dichromatic mirror (Semrock) and bandpass filters (Chroma Technology Corp.) in an electronic filter wheel. Microscope system was appropriately set to image the fluorescence of PpIX (405 nm excitation, 700 37.5 nm emission) and rhodamine 123 (488 nm
Figure 1. Heme biosynthesis pathway. Diagram shows four mitochondrial enzymes and four cytosolic enzymes and metabolites involved in heme biosynthesis.
Photochemistry and Photobiology, 2015, 91
excitation, 525 18 nm emission). Laser illumination was provided by a 50 mW monolithic laser combiner (MLC400; Agilent Technologies) and images were acquired using a Clara interline CCD camera (Andor Technology). The microscope system, laser launch and camera were controlled using Nikon Elements software. The exposure time for PpIX and rhodamine 123 was set at 500 ms and 200 ms, respectively. Differential interference contrast (DIC) images were acquired using exposure times in the range of 100–200 ms at the same magnification (609). Fluorescence images from different channels were pseudocolored and merged to generate composite images using NIH ImageJ software. PDT treatment and cytotoxicity assay. Cells were implanted in 96-well plates in complete DMEM and allowed to adhere overnight. Cells were rinsed with PBS once and incubated with serum-free DMEM containing ALA or no ALA (for control) for 4 h. Cells were then treated with 5 mW cm2 irradiance of 633-nm light for 5 or 10 min, which results in light fluence of 1.5 and 3.0 J cm2, respectively. Light illumination was provided by a diode laser system (High Power Devices Inc., North Brunswick, NJ) coupled to a 600-lm core diameter optical fiber fitted with a microlens at the end of fiber to achieve homogeneous irradiation. Light intensity was measured with an optical power meter (Thorlabs, Inc., North Newton, NJ). Immediately after light treatment, serum-free medium was replaced with complete DMEM. Cell viability was determined at 24 h after treatment by CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (MTS assay, Promega, Madison, WI) following the manufacturer’s instruction. Experiments were repeated four times. Statistical analysis. Two-tailed Students’ t-test was used to determine statistical difference between two groups. Statistical significance was accepted at P < 0.05.
A
Fluorescence unit
B
8.0×10 5
PpIX 100 ng/ml in DMSO
7.0×10 5 6.0×10 5 5.0×10 5 4.0×10 5 3.0×10 5 2.0×10 5 1.0×10 5 0
450
550
650
750
Wavelength (nm)
Fluor escence unit
Vector shPBGS5 shPBGD1 shFECH1
1.5×10 5 1.0×10 5 5.0×10 4 2.5×10 4 0
450
550
650
750
Wavelength (nm)
ALA (1 mM, 4 h) Fluor escence unit
1.4×10 6
Vector shPBGS5 shPBGD1 shFECH1
1.2×10 6 1.0×10 6 8.0×10 5 5.0×10 5
RESULTS Silencing PBGS, PBGD and FECH reduced enzyme protein levels
Without ALA
2.0×10 5
925
Heme biosynthesis pathway includes eight enzymatic steps that take place in the mitochondrion and cytosol (Fig. 1). The first and last three steps are catalyzed by enzymes in mitochondria, which include ALAS, coproporphyrinogen III oxidase (CPOX), protoporphyrinogen III oxidase (PPOX) and FECH. The rest four steps are catalyzed by enzymes in the cytosol including PBGS, PBGD, uroporphyrinogen III synthase (UROS) and uroporphyrinogen III decarboxylase (UROD). PBGS, PBGD and FECH were knockdown in SkBr3 cells using shRNA delivered by a lentiviral vector. Five shRNA variants targeting different sequences of each enzyme gene were used. Gene knockdown efficiency was examined by Western blot analysis. For each enzyme, two shRNA-expressing cell lines with the highest reduction in enzyme protein level were selected for the experiment. Figure 2A shows the enzyme level of parental, vector control and selected PBGS (shPBGS2 and 5), PBGD (shPBGD1 and 4) and FECH (shFECH1 and 2) knockdown cells. Compared with the parental and vector control cells, knockdown cells exhibited about 50% or higher reduction in enzyme protein level.
2.5×10 5 0
450
550
650
750
Wavelength (nm) Figure 2. (A) Silencing PBGS, PBGD and FECH reduced enzyme protein levels. Protein levels of PBGS, PBGD and FECH in parental, vector control and heme biosynthesis enzyme knockdown SkBr3 cells were examined by Western blot. Band intensity relative to actin loading control was normalized to the vector control and shown above the blot. (B) Fluorescence emission spectra of PpIX in DMSO and cell lysates without or with ALA stimulation. Fluorescence spectra were taken with excitation at 400 nm.
Silencing PBGS and PBGD reduced ALA-stimulated PpIX fluorescence, whereas silencing FECH increased PpIX fluorescence without and with ALA stimulation Fluorescence spectra of vector and knockdown cell lysates were measured with a spectrofluorometer and compared with PpIX fluorescence emission (Fig. 2B). Without ALA stimulation, vector, shPBGS5 and shPBGD1 cell lysates showed a little fluorescence, whereas shFECH1 cell lysate exhibited fluorescence emission resembling PpIX fluorescence with two emission peaks at about 632 and 692 nm. Similar PpIX-like fluorescence emission at
926
Xue Yang et al. cence, while silencing FECH (shFECH1) increased PpIX fluorescence. Fluorescence increase in shFECH2 cells was not statistically significant compared with the vector control (P > 0.05). To indicate the extent of ALA-mediated PpIX fluorescence increase, ratio of ALA-stimulated fluorescence to its basal fluorescence was calculated. For parental and vector control cells, ALA caused 50–60-folds increase in fluorescence (Fig. 3D). This ratio was significantly decreased in all knockdown cell lines. Protoporphyrin IX fluorescence in vector control and shRNAknockdown cells was imaged by a confocal microscope. To determine the relationship between PpIX localization and mitochondria, cells were labeled with mitochondrial dye rhodamine 123. Without ALA stimulation, fluorescence was not visible in vector control, shPBGS5, shPBGD1 cells (Fig. 4). In agreement with fluorescence spectrum and flow cytometry data, silencing FECH increased PpIX level, resulting in visible PpIX fluorescence in shFECH1 cells even without ALA stimulation. ALA stimulation led to visible PpIX fluorescence in all cells (Fig. 5),
much higher intensity was recorded in cell lysates after ALA stimulation. To quantify fluorescence intensity, cells were analyzed with a flow cytometer in the FL3 channel to record fluorescence emission above 650 nm. Based on fluorescence spectra of cell lysates, fluorescence detected within this range was mainly attributed to PpIX emission (Fig. 2B). Representative fluorescence histograms in parental, vector control and shRNA-knockdown SkBr3 cells with or without ALA stimulation are shown in Fig. 3A. Results of four independent experiments are summarized in Fig. 3B–D. Parental and vector control cells displayed low basal fluorescence without ALA stimulation. Knockdown of PBGS and PBGD had little effect on basal fluorescence (Fig. 3B). However, silencing FECH significantly enhanced basal fluorescence. Although incubation with ALA (1 mM) for 4 h greatly increased fluorescence in all cell lines (Fig. 3C), significant differences in fluorescence intensity were found between vector control and heme enzyme knockdown cells. Particularly, knockdown of PBGS (shPBGS2 and 5) and PBGD (shPBGD1 and 4) resulted in reduced PpIX fluores-
A
B
Basal fluorescence without ALA
*
Fluorescence (FL3)
200
150
** 100
50
H C FE sh
Fluorescence with ALA (1 mM, 4 h)
1600
*
1400
Fluorescence (FL3)
2
1 H C
D
FE
G sh
sh
P
B sh
sh
P
P
B
G
G
D
4
1
5 S
2 B
G B P sh
C
S
rl ct to r
ec V
P
ar
en ta l
0
1200 1000 800
**
600 400
**
200
**
**
H 2 C
H 1 C
sh FE
4 G D B
sh FE
1 G D B
sh P
sh P
5 B
G S
2 G S B
sh P
ct rl
sh P
D
V ec to r
P
ar en ta l
0
Ratio of ALA-stimulated to basal fluorescence 70
Fold increase
60 50 40 30 20 10
* **
**
*
**
**
P ar en ta l V ec to r ct rl sh P B G S 2 sh P B G S 5 sh P B G D 1 sh P B G D 4 sh FE C H 1 sh FE C H 2
0
Figure 3. Effects of silencing PBGS, PBGD and FECH on PpIX fluorescence. Parental, vector control and heme biosynthesis enzyme knockdown SkBr3 cells were incubated in serum-free DMEM with or without 1 mM ALA for 4 h and PpIX fluorescence was measured by a flow cytometer. (A) PpIX fluorescence histogram from a representative experiment. The number shown in each histogram is mean PpIX fluorescence intensity. (B) Fluorescence without ALA stimulation. (C) Fluorescence with ALA (1 mM, 4 h) stimulation. (D) The ratio of ALA-stimulated fluorescence to basal fluorescence without ALA stimulation. Data represent mean SE from four independent experiments. *P < 0.05, **P < 0.01, compared with vector control.
Photochemistry and Photobiology, 2015, 91
927
Figure 4. Confocal fluorescence microscopic images of vector control and heme biosynthesis enzyme knockdown SkBr3 cells without ALA stimulation. Mitochondria were labeled by incubating cells with rhodamine 123 (250 ng mL1) for 30 min. Bars, 10 lm.
indicating enhanced PpIX production after ALA supplement. However, compared with vector control cells, fluorescence in shPBGS5 and shPBGD1 cells was much weaker, whereas fluorescence in shFECH1 cells appeared stronger. It was noted that some cells in confocal images appeared brighter than surrounding cells. This heterogeneity in fluorescence was also shown in flow cytometry histograms where cell populations with different fluorescence intensity were observed (Fig. 3A), which suggests the existence of variation in knockdown efficiency. Although being more diffused than mitochondrial marker fluorescence, PpIX fluorescence overlapped with mitochondrial marker fluorescence, indicating its localization in mitochondria (Fig. 5). Silencing PBGS and PBGD reduced cell sensitivity to ALAmediated PDT, whereas silencing FECH increased cell sensitivity to ALA-mediated PDT We next treated vector control, shPBGS5, shPBGD1 and shFECH1 SkBr3 cells with light after 4-h incubation with ALA and
evaluated cytotoxicity at 24 h after ALA-PDT (Fig. 6). Four different ALA-PDT doses were assessed. A low PDT dose (0.5 mM ALA, 1.5 J cm2) significantly reduced cell viability only in shFECH1 cells (P < 0.01), but not in vector control, shPBGS5 and shPBGD1 cells (P > 0.05), indicating that silencing FECH expression increases cell response to ALA-PDT. This is further supported by the finding that shFECH1 had significantly lower cell viability than vector control cells (P < 0.05) after PDT with a higher light dose (0.5 mM ALA, 3.0 J cm2) or a higher ALA dose (1.0 mM ALA, 1.5 J cm2). However, PDT with the highest light and ALA doses (1.0 mM ALA, 3.0 J cm2) reduced cell viability to a similarly low level (below 10%) in both vector control and shFECH1 cells. In contrast to shFECH1 which showed increased sensitivity to ALA-PDT, shPBGS5 and shPBGD1 cells exhibited significantly reduced sensitivity to PDT with varied ALA and light doses. ALA-PDT at lower doses had little effect on the viability of shPBGS5 and shPBGD1 cells and cells were still able to maintain more than 80% viability even after PDT with the highest light and ALA doses (Fig. 6).
928
Xue Yang et al.
Figure 5. Confocal fluorescence microscopic images after ALA stimulation. Vector control and heme biosynthesis enzyme knockdown SkBr3 cells were incubated in serum-free DMEM with 1 mM ALA for 4 h and imaged by a confocal microscope. Mitochondria were labeled by incubating cells with rhodamine 123 (250 ng mL1) for 30 min. Bars, 10 lm.
DISCUSSION Aminolevulinic acid-mediated tumor detection and treatment critically depend on preferential PpIX accumulation in tumor tissues following ALA administration. Through four consecutive biosynthetic steps in the cytosol and two additional reactions in mitochondria catalyzed by a total of six heme biosynthesis enzymes, ALA is converted to PpIX (8). Under the catalysis of FECH, PpIX is chelated with ferrous iron to form heme, which does not have fluorescent and photosensitizing property. Thus, enzymes involved in PpIX biosynthesis and bioconversion collectively determines ALA-mediated PpIX accumulation. We silenced the expression of the first two cytosolic enzymes in ALA-PpIX biosynthesis (PBGS, PBGD) and the mitochondrial enzyme responsible for PpIX bioconversion (FECH) in an attempt to evaluate the role of these three enzymes in ALA-PpIX accumulation and cytotoxicity to ALA-PDT. SkBr3 human breast cancer cell line was used because this cell line showed robust PpIX production
after ALA stimulation in our previous study (11). Our present data demonstrate that, although all three enzymes are important for ALA-mediated PpIX production and PDT response, silencing PBGS/PBGD and FECH have completely opposite effects. As the first cytosolic enzyme in PpIX biosynthesis, PBGS converts two molecules of endogenous and exogenous ALA to a monopyrrole metabolite porphobilinogen (PBG) (8). Because the reaction involves the removal of a water molecule, PBGS is also known as ALA dehydratase (ALAD). Four molecules of PBG are subsequently deaminated by PBGD to form a tetrapyrrole metabolite preuroporphyrinogen. Since preuroporphyrinogen is also called hydroxymethylbilane, PBGD is also known as hydroxymethylbilane synthase (HMBS). Here, we found that knockdown of PBGS or PBGD significantly decreased ALAstimulated PpIX fluorescence (Fig. 3). Moreover, comparing ALA-PpIX fluorescence between cell lines with different knockdown efficiency revealed a correlation between ALA-mediated PpIX fluorescence reduction and the extent of enzyme gene
Photochemistry and Photobiology, 2015, 91 PDT (0.5 mM ALA 4 h, 3 J/cm 2)
100
^^
80 60 40 20 0
Cell viability (% control)
Cell viability (% control)
PDT (0.5 mM ALA 4 h, 1.5 J/cm 2) 120
Vector shPBGS5 shPBGD1 shFECH1
120
100 80 60 40
**
^
*
^^
20 0
Vector shPBGS5 shPBGD1 shFECH1
**
80 60
^^^
* ^^
40 20 0
Vector shPBGS5 shPBGD1 shFECH1
PDT (1 mM ALA 4 h, 3 J/cm 2) Cell viability (% control)
Cell viability (% control)
*
***
100
PDT (1 mM ALA 4 h, 1.5 J/cm 2) 120
929
120
**
100 80
*** ^
60 40 20 0
^^^
^^^
Vector shPBGS5 shPBGD1 shFECH1
Figure 6. Effects of silencing heme biosynthesis enzyme PBGS, PBGD and FECH on ALA-mediated PDT. Vector, shPBGS5, shPBGD1 and shFECH1 cells were treated with ALA-PDT as indicated and cell viability was examined by the MTS assay at 24 h after treatment. Data represent mean SD from four independent experiments. ^P < 0.05, ^^P < 0.01, ^^^P < 0.001 compared with light only control without ALA. *P < 0.05, **P < 0.01, ***P < 0.001 compared with vector cells.
silencing. For example, shPBGS5 cell line had a higher knockdown efficiency than shPBGS2 (Fig. 2A). It displayed lower ALA-PpIX fluorescence than shPBGS2 (Fig. 3). Similarly, shPBGD1 cell line had better knockdown effect than shPBGD4. It, therefore, showed lower fluorescence than shPBGD4 line. Because of reduced PpIX production, PBGS and PBGD knockdown cells exhibited a little or no response to ALA-PDT (Fig. 6). Thus, our present results in SkBr3 breast cancer cells together with previous studies in K562 leukemia cells (12,13) demonstrate that both PBGS and PBGD are critically important for ALA-mediated PpIX production and reducing the level/activity of any of these enzymes decreases PpIX fluorescence and PDT efficacy. Ferrochelatase catalyzes the insertion of ferrous iron into PpIX, thereby reducing PpIX level by converting it to nonfluorescent heme (8). Consistent with previous studies in human colon (14), urothelial (15) and glioma (16) cancer cells, we found in SkBr3 cells that silencing FECH expression significantly increased ALA-stimulated PpIX fluorescence. As a result, FECH knockdown cells were much more sensitive to ALA-PDT than vector control cells. Considering that a variety of human breast cancer cell lines have shown preferential ALA-PpIX accumulation (17) and ALA-PpIX fluorescence has been explored for the detection of early neoplastic and metastatic mammary tumors in transgenic mice (18) as well as primary and metastatic human breast cancers (19,20), our present finding implies that inhibiting FECH may improve ALA-based tumor detection and treatment for breast cancers by boosting more PpIX accumulation. As a matter of fact, iron-chelating agents, which abolish FECH activity by removing ferrous iron, have shown effectiveness in enhancing ALA-PpIX fluorescence and PDT effect in skin
cancers (21). However, because we found that knockdown of FECH increased basal fluorescence without ALA as well, future studies are needed to determine whether pharmacologically inhibiting FECH increases PpIX fluorescence in normal cells and, if so, how much PpIX increase in normal cells compared to tumor cells. To make it an effective approach for improving ALA-based fluorescence imaging and PDT, FECH inhibitors should possess preferential effects in tumor cells so that a greater difference in PpIX between tumor and surrounding normal cells can be achieved. It is interesting to note that, compared with parental and vector control cells, the ratio of ALA-stimulated fluorescence to basal fluorescence was significantly reduced in all lines of knockdown cells. As an indication of the extent of ALA-stimulated PpIX production, this ratio is likely dependent on the overall enzymatic capacity of six enzymes involved in PpIX biosynthesis (from PBGS to PPOX) and the activity of PpIX bioconversion enzyme FECH. As expected, knockdown of PpIX biosynthesis enzymes such as PBGS and PBGD decreased the enzyme level/activity, therefore resulting in reduced capacity of ALA-PpIX biosynthesis. Surprisingly, knockdown of FECH involved in PpIX bioconversion also decreased the ratio of ALA-stimulated PpIX to basal fluorescence. There are two possible explanations for this paradoxical finding. First, such a reduction may indicate the saturation of PpIX biosynthesis enzyme activity. Silencing FECH inhibits the conversion of PpIX to heme, which causes the accumulation of porphyrin metabolites (particularly its substrate PpIX) in heme biosynthesis pathway and activates heme biosynthesis due to the release of negative feedback inhibition on heme biosynthesis pathway (22). Imposing on this already activated PpIX biosynthesis pathway with
930
Xue Yang et al.
additional load of exogenous ALA may saturate pathway activity, thereby failing to obtain more ALA-stimulated PpIX production. Second, reduction in the ratio of ALA-PpIX over basal PpIX production in FECH knockdown cells may suggest an interaction between FECH and upstream enzymes involved in PpIX biosynthesis. In this case, PpIX biosynthesis enzyme functions may be inhibited in the absence of FECH enzyme. Studies are undergoing to determine whether such interactions exist. In summary, we evaluated the role of three heme biosynthesis enzymes PBGS, PBGD and FECH in ALA-mediated PpIX fluorescence and PDT efficacy in SkBr3 human breast cancer cells. We found that knockdown of PBGS or PBGD significantly reduced ALA-stimulated PpIX fluorescence and rendered cell resistance to ALA-PDT. In contrast, silencing FECH greatly elevated basal and ALA-stimulated PpIX fluorescence and sensitized cells to ALA-PDT. These results demonstrate that PBGS, PBGD and FECH are important for ALA-stimulated PpIX fluorescence and suggest that inhibiting FECH may enhance ALAbased tumor fluorescence detection and PDT efficacy in breast cancers. Future studies should focus on comparing the effects of FECH inhibition on ALA-PpIX in tumor versus normal cells and determining whether FECH inhibition affects PpIX biosynthesis enzymes.
7.
8. 9. 10.
11.
12.
13. 14.
Acknowledgement—This study was supported in part by Research Scholar Grant RSG-10-035-01-CCE from the American Cancer Society.
15.
REFERENCES
16.
1. Nokes, B., M. Apel, C. Jones, G. Brown and J. E. Lang (2013) Aminolevulinic acid (ALA): Photodynamic detection and potential therapeutic applications. J. Surg. Res. 181, 262–271. 2. Peng, Q., T. Warloe, K. Berg, J. Moan, M. Kongshaug, K. E. Giercksky and J. M. Nesland (1997) 5-Aminolevulinic acid-based photodynamic therapy. Clinical research and future challenges. Cancer 79, 2282–2308. 3. Stummer, W., U. Pichlmeier, T. Meinel, O. D. Wiestler, F. Zanella, H. J. Reulen and A. L.-G. S. Group (2006) Fluorescence-guided surgery with 5-aminolevulinic acid for resection of malignant glioma: A randomised controlled multicentre phase III trial. Lancet Oncol. 7, 392–401. 4. Smits, T., A. I. van Laarhoven, A. Staassen, P. C. van de Kerkhof, P. E. van Erp and M. J. Gerritsen (2009) Induction of protoporphyrin IX by aminolaevulinic acid in actinic keratosis, psoriasis and normal skin: Preferential porphyrin enrichment in differentiated cells. Br. J. Dermatol. 160, 849–857. 5. Jocham, D., F. Witjes, S. Wagner, B. Zeylemaker, J. van Moorselaar, M. O. Grimm, R. Muschter, G. Popken, F. Konig, R. Knuchel and K. H. Kurth (2005) Improved detection and treatment of bladder cancer using hexaminolevulinate imaging: A prospective, phase III multicenter study. J. Urol. 174, 862–866. 6. Barnett, A. A., J. C. Haller, F. Cairnduff, G. Lane, S. B. Brown and D. J. Roberts (2003) A randomised, double-blind, placebo-controlled trial of photodynamic therapy using 5-aminolaevulinic acid for the
17.
18.
19.
20. 21.
22.
treatment of cervical intraepithelial neoplasia. Int. J. Cancer 103, 829–832. Kato, S., J. Kawamura, K. Kawada, S. Hasegawa and Y. Sakai (2012) Fluorescence diagnosis of metastatic lymph nodes using 5aminolevulinic acid (5-ALA) in a mouse model of colon cancer. J. Surg. Res. 176, 430–436. Ponka, P. (1999) Cell biology of heme. Am. J. Med. Sci. 318, 241– 256. Heinemann, I. U., M. Jahn and D. Jahn (2008) The biochemistry of heme biosynthesis. Arch. Biochem. Biophys. 474, 238–251. Fateye, B., W. Li, C. Wang and B. Chen (2012) Combination of phosphatidylinositol 3-kinases pathway inhibitor and photodynamic therapy in endothelial and tumor cells. Photochem. Photobiol. 88, 1265–1272. Gibbs, S. L., B. Chen, J. A. O’Hara, P. J. Hoopes, T. Hasan and B. W. Pogue (2006) Protoporphyrin IX level correlates with number of mitochondria, but increase in production correlates with tumor cell size. Photochem. Photobiol. 82, 1334–1341. Feuerstein, T., A. Schauder and Z. Malik (2009) Silencing of ALA dehydratase affects ALA-photodynamic therapy efficacy in K562 erythroleukemic cells. Photochem. Photobiol. Sci. 8, 1461– 1466. Schauder, A., T. Feuerstein and Z. Malik (2011) The centrality of PBGD expression levels on ALA-PDT efficacy. Photochem. Photobiol. Sci. 10, 1310–1317. Kemmner, W., K. Wan, S. Ruttinger, B. Ebert, R. Macdonald, U. Klamm and K. T. Moesta (2008) Silencing of human ferrochelatase causes abundant protoporphyrin-IX accumulation in colon cancer. FASEB J. 22, 500–509. Miyake, M., M. Ishii, K. Kawashima, T. Kodama, K. Sugano, K. Fujimoto and Y. Hirao (2009) siRNA-mediated knockdown of the heme synthesis and degradation pathways: Modulation of treatment effect of 5-aminolevulinic acid-based photodynamic therapy in urothelial cancer cell lines. Photochem. Photobiol. 85, 1020–1027. Teng, L., M. Nakada, S. G. Zhao, Y. Endo, N. Furuyama, E. Nambu, I. V. Pyko, Y. Hayashi and J. I. Hamada (2011) Silencing of ferrochelatase enhances 5-aminolevulinic acid-based fluorescence and photodynamic therapy efficacy. Br. J. Cancer 104, 798–807. Millon, S. R., J. H. Ostrander, S. Yazdanfar, J. Q. Brown, J. E. Bender, A. Rajeha and N. Ramanujam (2010) Preferential accumulation of 5-aminolevulinic acid-induced protoporphyrin IX in breast cancer: A comprehensive study on six breast cell lines with varying phenotypes. J. Biomed. Opt. 15, 018002. Dorward, A. M., K. S. Fancher, T. M. Duffy, W. G. Beamer and H. Walt (2005) Early neoplastic and metastatic mammary tumours of transgenic mice detected by 5-aminolevulinic acid-stimulated protoporphyrin IX accumulation. Br. J. Cancer 93, 1137–1143. Frei, K. A., H. M. Bonel, H. Frick, H. Walt and R. A. Steiner (2004) Photodynamic detection of diseased axillary sentinel lymph node after oral application of aminolevulinic acid in patients with breast cancer. Br. J. Cancer 90, 805–809. Ladner, D. P., R. A. Steiner, J. Allemann, U. Haller and H. Walt (2001) Photodynamic diagnosis of breast tumours after oral application of aminolevulinic acid. Br. J. Cancer 84, 33–37. Campbell, S. M., C. A. Morton, R. Alyahya, S. Horton, A. Pye and A. Curnow (2008) Clinical investigation of the novel iron-chelating agent, CP94, to enhance topical photodynamic therapy of nodular basal cell carcinoma. Br. J. Dermatol. 159, 387–393. Sakaino, M., T. Kataoka and S. Taketani (2009) Post-transcriptional regulation of the expression of ferrochelatase by its variant mRNA. J. Biochem. 145, 733–738.
