Introduction

The World Health Organization’s Global Cancer Observatory (GCO) estimated an occurrence of ~ 20 million new cases of cancer and ~ 9.7 million cancer deaths during 2022 (Global Cancer Observatory, 2022). Conventional cancer treatments such as radiotherapy, surgery, and chemotherapy, alongside contemporary advancements in immunotherapy and hormonal therapy, have collectively led to substantial improvements in cancer prognosis1. However, the development and refinement of therapeutic strategies remains paramount in the effort to improve patient outcomes. Photodynamic therapy (PDT) is a minimally invasive cancer treatment that has been approved by the FDA for use against some cancers2. Fundamentally, PDT involves the administration of a photosensitive drug known as a photosensitizer (PS) that preferentially accumulates within tumors, which are subsequently irradiated with wavelengths of visible light that provide optimal absorbance and tissue penetration. Once subjected to light, PSs are excited from a ground state and achieve a triplet state through intersystem crossing, resulting in the formation of singlet oxygen and reactive oxygen species (ROS) that damage cellular components and initiate cell death3.

5-aminolevulinic acid (5-ALA) is a non-proteogenic amino acid and a key precursor in the biosynthesis of heme. The generation of 5-ALA is mediated by aminolevulinic acid synthase (ALAS), which catalyzes the condensation of glycine and succinyl-CoA at the mitochondrial membrane. Through a series of metabolic reactions in the cytosol and mitochondria, 5-ALA is converted to protoporphyrin IX (PPIX), a fluorescent PS and immediate precursor of heme. Importantly, the generation of 5-ALA by ALAS is the rate limiting step in the pathway, a key element that is exploited by 5-ALA-PDT4. Under homeostatic conditions, ALAS is regulated by negative feedback from free heme, preventing excessive heme production that would otherwise result in oxidative stress. However, the addition of exogenous water soluble 5-ALA bypasses homeostatic control of 5-ALA synthesis, leading to an accumulation of intracellular photosensitive PPIX. Many cancers display a preferential accumulation of PPIX compared to normal cells following 5-ALA administration, which has been attributed to metabolic reprogramming that upregulates components of the heme biosynthesis pathway during oncogenesis5. The preferential accumulation of PPIX in cancer cells is another exploitable element of 5-ALA-PDT that allows for the selective elimination of tumors with minimal damage to normal cells. Importantly, the subcellular accumulation of PPIX occurs primarily in mitochondria, which can become damaged by ROS generated during PDT and cause mitochondrial dysfunction that results in the initiation of programmed cell death6.

Although 5-ALA-PDT is effective at killing cancer cells, it has limitations. One limitation in vivo is the poor penetration of 630 nm light used to excite PPIX. Additionally, we have observed that 5–10% of cells survive in vitro experiments. The resistant population that remains post-treatment could lead to tumor resurgence with a phenotype that is less-responsive to subsequent 5-ALA-PDT. As such, identifying the mechanisms of 5-ALA-PDT resistance is crucial to improve treatment outcomes. Previous work in our lab demonstrated that RasV12-transformed NIH3T3 cells have upregulated PPIX biosynthesis, but also upregulate ATP-binding cassette (ABC) transporters and ferrochelatase (FECH) via MEK signalling-axes that reduce PPIX accumulation7. ABC transporters are responsible for PPIX efflux outside of the cell by exporting PPIX through the mitochondrial and plasma membranes8,9. Inhibition of MEK and downstream targets significantly enhanced the accumulation of PPIX in cancer cells, demonstrating a key mechanism of resistance to 5-ALA-PDT.

Cancer stem cells (CSCs) are a minor population of cancer cells that possess high proliferative capability and produce cancer cells with heterogenous phenotypes in a similar fashion to the differentiation of normal stem cells. Both tissue resident stem cells and differentiated cells have the potential to become CSCs because stem cells have a low genetic barrier to transform into cancer cells, and partially differentiated progenitor cells can reacquire stem cell characteristics during transformation10. CSCs are known to drive resistance to cancer treatments and express high levels of ABC transporters that mediate the export of therapeutic drugs, preventing elimination and promoting tumor recurrence11,12,13,14. As a result, development of strategies to enhance the ability of PDT to target CSCs is an active area of research, with an emphasis on drug delivery systems that overcome mechanisms of resistance15,16,17. Therefore, it is pertinent to ascertain the sensitivity of CSCs to 5-ALA-PDT, given the potential resistant phenotype that could serve as founding population for future tumors. In this study, our goal is to determine if CSCs are resistant to 5-ALA-PDT, and if CSCs establish 5-ALA-PDT resistant populations post-treatment.

Methods

Cell culture

The human lung cancer cell line H1299, human breast cancer cell line Hs578T and human colon cancer cell lines DLD-1, were purchased from the American Type Culture Collection (ATCC). All cells were cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) (Corning, MA) supplemented with 10% fetal bovine serum (HyClone, Cytiva), 1 mM sodium pyruvate (Life Technologies) and antibiotic-antimycotic (Thermo Scientific). Cells were maintained in 10 cm culture dishes at 37 °C with 5% CO2. For generation of 5-ALA-PDT resistant cells, DLD-1 cells (5000 cells/well) were plated in 96-well plates and subjected to 5-ALA-PDT 24 h later. At 72 to 96 h following 5-ALA-PDT, live cells were trypsinized, pooled and re-plated in a 10 cm dish to generate a cell line with a PDT-conditioned population (DLD-1 PDT×1). DLD-1 PDT×1 cells were plated in 96-well plates and subjected to another round of 5-ALA-PDT to generate DLD-1 PDT×2. This cycle was repeated until we obtained cell lines with population of DLD-1 that had been conditioned by PDT to varying degrees, with a maximum of four treatments (DLD-1 PDT×4).

5-ALA-PDT

Human cancer cells plated in 96-well plates were treated with 5-aminolevulinic acid (5-ALA) (5 mM) (Sigma, A3785, USA) diluted in culture media (as described in previous section) and incubated at 37 °C for 4 h. Following the incubation period, culture media was replaced, and the cells were subjected to PDT using a Theralase TLC 3000 A modular light source (Theralase Technologies Inc., Toronto, Canada; λ = 618–630 nm, fluence rate = 150 mW/cm2, energy density (ED) = 27 J/cm2) for a duration of 3 min.

Metabolic activity assay

Metabolic activity was measured using the Colorimetric Cell Viability Kit I (WST-8) following manufacturer’s instructions (PromoCell GmbH, Germany). Ten (10) µl of WST-8 reagent was added to each well of 96-well plates containing cells and incubated in the dark at 37 °C for 35 min. Absorbance was measured at 450 nm using a Bio-Rad Model 3550 microplate reader and metabolic activity was determined based on standard curves for each cell line.

Flow cytometry

Cells were fixed with Inside Stain Kit (Miltenyl Biotec, Teterow, Germany) and subsequently stained with anti-CD133 antibody (Miltenyl Biotec) and Zombie Violet™ Fixable Viability Kit (BioLegend, CA, USA) following manufacturer’s instructions. Dead or dying cells with compromised cellular membranes are stained and represented graphically in red, whereas live cells remained unstained and are represented in green. Flow cytometry analysis was conducted using a CytoFLEX Flow Cytometer (Beckman Coulter, CA, USA). The data was analyzed using FlowJo (FlowJo LLC, OR).

Soft agar assay

UltraPure TM agarose (0.7%) (Invitrogen, CA, USA) and purified agar (1%) (Oxoid LTD, Hampshire UK) were mixed and incubated in a hot water bath at 37 °C. First, the mixtures of agarose and 2X DMEM were plated in a 6-well plate (1.5 ml/well) to generate the bottom layer. After 15 min of cooling to solidify the bottom layer, it was overlaid with mixtures of cell suspension and agarose to generate the top layer (2500 cells/well). Once the top layer solidified, the plate was incubated at 37 °C for 21 days, after which colonies were counted under microscope.

Western blot analysis

Anti-ALDH1 antibody was purchased from Abcam (Boston, MA, USA) and HRP-conjugated secondary antibody from Santa Cruz Biotechnology (Dallas, TX, USA). Protein samples were prepared, and western blot was conducted as previously described18.

Quantification of intracellular PPIX

Parental and PDT-resistant (PDT×1–4) DLD-1 cells were plated in 24-well plates (5 × 104 cells/well) and incubated for 24 h to attain confluency. Following this, cells were treated with 5-ALA (5 mM) and incubated in the dark for 8–24 h. At each time point, samples were obtained by washing cells with PBS before lysis with supplemented RIPA buffer (RIPA, PMSF, aprotinin, and Halt™ phosphatase and protease inhibitor cocktail (Thermo Scientific)). Lysate was collected in opaque amber tubes and stored at -80 °C. In minimally lit conditions, lysate was diluted in PBS (2.5% v/v), vigorously vortexed, added to solid black 96-well plates (Greiner Bio-one), and covered with foil. Fluorescence of PPIX was measured immediately with a Biotek Synergy MX plate reader with an excitation wavelength of 405 nm and emission wavelength of 635 nm.

Statistical analyses

Statistical analyses were performed using Prism 7.0 (GraphPad). Student’s t-test was used for inter-group comparison, and one-way ANOVA with Tukey’s posthoc test was used to compare between multiple groups. P < 0.01 were considered statistically significant.

Results

CD133 + cells are not sensitive to 5-ALA-PDT

To determine if the CD133 + stem cell (CSC) population is sensitive to 5-ALA-PDT, we quantified the percentage of CD133 + cell populations in three different cell lines [human colon cancer (DLD-1), human lung cancer (H1299) and human breast cancer (Hs578T)] (Fig. 1). The average percentage of CD133 + stem cell population in DLD-1, H1299 and Hs578T were 1.0 ± 0.4%, 1.9 ± 0.7% and 1.6 ± 0.3% respectively. We next sought to determine if the CD133 + cells are sensitive to 5-ALA-PDT. The concentration of 5-ALA (5mM) we used in the study did not affect metabolic activity without light exposure (Supplementary Fig. 1). Each cell line was treated with 5-ALA for 4 h and then irradiated for 3 min using a Theralase TLC 3000 A modular light source (27 J/cm2) (5-ALA-PDT). At 24 h after 5-ALA-PDT, the cells were incubated with Zombie Violet, fixed with inside stain kit (Miltenyl Biotec), and then incubated with an anti-CD133 antibody for 10 min. Cell survival [dead cells (red) and live cells (green), see methods] and CD133 expression was evaluated by flow cytometry (Fig. 2). In the scatter plots of DLD-1 cells, 0.77% of live cells were CD133 + before 5-ALA-PDT which increased to 2.03% after 5-ALA-PDT (Fig. 2A), indicating that this population is enriched in surviving cells. In the total population of DLD-1 cells, live cells decreased from 83 ± 10% to 46 ± 21% after 5-ALA-PDT, but the reduction was not statically significant. In contrast, the live cell populations of CD133 + DLD-1 cells were 77 ± 2% and 69 ± 14% before and after 5-ALA-PDT respectively, suggesting that 5-ALA-PDT is not effective on CD133 + DLD-1 cells. For H1299, a lung cancer cell line, the percentage of CD133 + cells in the total population was increased to 3.67% from 2.32% after 5-ALA-PDT (Fig. 2B). Moreover, there was a significant difference in the percentage of live cells in the total cell population before and after 5-ALA-PDT, with an approximate 44% decrease of living cells following 5-ALA-PDT (Fig. 2B). In contrast, CD133 + H1299 cells were resistant to 5-ALA-PDT as 76 ± 5% of CD133 + cells were alive even after 5-ALA-PDT and statistically insignificant compared to the control. Similar results were found in Hs578T cells (Fig. 2C). The CD133 + population was increased after 5-ALA-PDT and cell death was effectively induced in the total cell population of Hs578T cells as the live cell population significantly reduced from 79 ± 5% to 30 ± 7%. However, the live percentages of CD133 + Hs578T cells were not significantly reduced by 5-ALA-PDT. These results demonstrate that the CD133 + populations of each cell line were not sensitive to cell death induced by 5-ALA-PDT.

Fig. 1
figure 1

CD133 + stem cell population in DLD-1, H1299 and Hs578T cells. DLD-1 (human colon cancer), H1299 (human lung cancer) and Hs578T (human breast cancer) cells were stained with anti-CD133 antibody. Flow cytometry analysis was conducted using the CytoFLEX Flow Cytometer and analyzed using FlowJo.

Full size image
Fig. 2
figure 2

CD133 + stem cells are not sensitive to 5-ALA-PDT. (A) DLD-1, (B) H1299 and (C) Hs578T cells were treated with 5-ALA (5mM) (5-ALA-PDT) or with control vehicle (control) for 4 h and then irradiated for 3 min using a Theralase TLC 3000 A modular light source (27 J/cm2). At 24 h post-irradiation, Cell survival (dead cells (red) and live cells (green)) by Zombie Violet staining (colorimetric assay) and numbers of CD133 positive cells was measured by flow cytometry analysis. **p < 0.01 by one-way ANOVA with Tukey’s post-hoc test.

Full size image

Cancer stem cells are enriched in 5-ALA-PDT resistant populations

We hypothesize that cancer stem cells play a critical role in forming populations with resistance to 5-ALA-PDT. If this is the case, cancer stem cells could be enriched in 5-ALA-PDT resistant populations. To this end, we first established 5-ALA-PDT resistant cell lines from DLD-1 cells by subjecting them to multiple rounds of treatment (Fig. 3). We plated parental DLD-1 cells in a 96-well plate and subjected them to 5-ALA-PDT (5 mM 5-ALA, 27 J/cm2). At 3–4 days after 5-ALA-PDT, the surviving cells were pooled and further subjected to 5-ALA-PDT. This process was repeated up to 4 times (PDT×1: one cycle, PDT×2: two cycles, PDT×3: three cycles and PDT×4: four cycles) (Fig. 3A). To confirm if the established cell lines were resistant to 5-ALA-PDT, the metabolic activity of parental DLD-1 and 5-ALA-PDT resistant cells (PDT×1, PDT×2, PDT×3 and PDT×4) was measured 24 h after 5-ALA-PDT by a WST-8 assay. The metabolic activities of parental DLD-1 and PDT×1 cell lines were significantly reduced after 5-ALA-PDT, suggesting that they are sensitive to 5-ALA-PDT (Fig. 3B). In contrast, most of the PDT×2, PDT×3 or PDT×4 cells survived after 5-ALA-PDT, indicating that they are 5-ALA-PDT resistant cell lines.

Fig. 3
figure 3

Generation of 5-ALA-PDT resistant DLD-1 cells. (A) Illustration of the workflow of generation of 5-ALA-PDT resistant cells. DLD-1 cells plated in 96-well plates were subjected to 5-ALA-PDT. At 72 to 96 h after 5-ALA-PDT, live cells were pooled and re-plated in a 10 cm dish (DLD-1 PDT×1). The cycle was repeated until we obtained DLD-1 PDT×4. (B) Parental DLD-1, PDT×1, PDT×2, PDT×3 and PDT×4 cells were treated with 5-ALA (5mM) (5-ALA-PDT) or with vehicle (control) for 4 h and then irradiated for 3 min using a Theralase TLC 3000 A modular light source (27 J/cm2). At 24 h post-irradiation, metabolic activity of cells was measured by WST-8 assay. **p < 0.01 by one-way ANOVA with Tukey’s post-hoc test.

Full size image

To determine if CSCs were enriched in the DLD-1 PDT×2, PDT×3 and PDT×4 resistant cell lines, we first performed a soft agar analysis of anchorage-independent growth. The parental DLD-1, PDT×2, PDT×3 and PDT×4 resistant cells were plated in soft agar. After 21 days, colonies were counted under the microscope. PDT×3 and PDT×4 resistant cells showed a significant increase in the average number of colonies compared to the parental DLD-1 cell line, whereas a significant change was not observed in PDT×2 resistant cells (Fig. 4A). To further confirm that CSCs are enriched in the 5-ALA-PDT resistant populations, we conducted western blot analysis of ALDH1 on cell lysates prepared from parental DLD-1 cells and 5-ALA-PDT resistant DLD-1 cells (PDT×2, PDT×3 and PDT×4) (Fig. 4B and Supplementary Fig. 2). ALDH1 is another marker for CSCs, which was expressed on DLD-1 CSCs, but not in other cell lines. While we did not observe changes in ALDH1 expression between PDT×2 cells and parental DLD-1 cells, ALDH1 expression was increased in PDT×3 and PDT×4 cells. Finally, we determined the expression levels of the CSC marker CD133 + expression in the 5-ALA-PDT resistant cells (PDT×4) by flow cytometry. The mean fluorescent intensity of the CD133 + population of cells was increased in the PDT×4 cell line compared to the parental DLD-1 cell line (Fig. 4C). We also found that the percentage of CD133 + cells was significantly increased in the PDT×4 cell line compared to DLD-1 parent cells. These results further support that CSCs are resistant to 5-ALA-PDT and are enriched in 5-ALA-PDT resistant cell populations. We further determined if decreased PPIX accumulation could be one of causes of 5-ALA-PDT resistance (Fig. 4D). We found that 5-ALA-PDT resistant DLD-1 cells (PDT×1, PDT×2, PDT×3 and PDT×4) accumulated significantly lower amounts of PPIX than parental control DLD-1 cells when stimulated with 5-ALA. The results suggest that low PPIX accumulation is one of the underlying mechanisms for 5-ALA-PDT resistance.

Fig. 4
figure 4

CSCs are enriched in 5-ALA-PDT resistant DLD-1 cells. (A) Parental DLD-1 cells and 5-ALA-PDT resistant DLD-1 cells (PDT×2, PDT×3 and PDT×4) were plated for assaying anchorage-independent growth assay in soft agar. Mean ± standard error of mean (SEM) in average colonies numbers per microscopic field are presented. **p < 0.01 by one-way ANOVA with Tukey’s post-hoc test. (B) The expression levels of ALDH1 and β-actin in parental DLD-1 cells and the 5-ALA-PDT resistant DLD-1 cells was determined by western blot analysis. (C) The expression levels of CD133 on parental DLD-1 cells and 5-ALA-PDT resistant DLD-1 cells (PDT×4) was measured by flow cytometry. Representative histogram of fluorescent intensity of CD133 on parental DLD-1 and PDT×4 cells (left) and quantitative analysis of CD133 expression based on 3 independent experiments (right). **p < 0.01 by one-way ANOVA with Turkey’s post-hoc test. (D) Intracellular accumulation of PPIX was measured by fluorometric analysis of total lysate of DLD-1 cells (Parental, PDT×1, PDT×2, PDT×3, and PDT×4) treated with 5 mM of 5-ALA for 8 h and 24 h in the dark. Excitation λ = 405 nm, emission λ = 635 nm. **p < 0.01 by one-way ANOVA with Tukey’s post-hoc test.

Full size image

Discussion

Although 5-ALA-PDT has been approved for treating different types of cancers, cancer recurrence remains as a problem19,20,21. One limiting factor is the poor tissue penetration of 630 nm light used to excite PPIX. As such, the use of photosensitzers with longer wavelengths may overcome this barrier. Another key factor of cancer recurrence is resistance to treatment, leading to survival of a small number of cancer cells that potentially reform tumors. Therefore, addressing the mechanisms of cancer resistance against 5-ALA-PDT is essential to improve its efficacy.

In this study, we investigated the possible role of CSCs as a driver of 5-ALA-PDT resistance because CSCs have been reported to resist chemotherapy partially due to upregulation of ABC transporters22,23,24. ABC transporters efflux chemotherapeutics from cancer cells, which decreases its intracellular accumulation and thus prevents the cells from dying25,26. Our previous studies also demonstrated that PPIX efflux via ABCB1 transporter reduces cancer cell sensitivity to 5-ALA-PDT7,9. Therefore, we hypothesized that CSCs are less sensitive to 5-ALA-PDT and subsequently contribute to the establishment of resistant populations to 5-ALA-PDT due to their ability to efflux 5-ALA via ABC transporters. Through our efforts, we confirmed that 5-ALA-PDT is not effective in killing CSCs and that CSC populations are increased in the 5-ALA-PDT resistant cell lines. These results clearly indicate that CSCs play an essential role in establishing cancer resistance to 5-ALA-PDT in the cell lines evaluated (Fig. 5).

Fig. 5
figure 5

CSCs are resistant to 5-ALA-PDT which play critical roles in tumor recurrence. 5-ALA-PDT does not kill CSCs efficiently due to (1) low accumulation of PPIX caused by high expression levels of the ABC transporter, (2) promoted cell survivability caused by expression of stress response genes and constitutively active autophagy and (3) resistance to cell death due to low ROS levels. Surviving CSCs serve as a founding population for future tumors, resulting in cancer recurrence.

Full size image

Interestingly, our results contrast previous reports of 5-ALA-PDT efficacy against CSCs27,28,29. CSCs from oral squamous cell carcinoma cell lines were effectively differentiated by 5-ALA-PDT and susceptibility to treatment was due to increased PPIX accumulation27. Similarly, glioma CSC lines accumulate more PPIX compared to their differentiated counterparts and therefore more sensitive to treatment28. When evaluating 5-ALA-PDT against head and neck cancer CSC lines, 5-ALA-PDT reduced the expression of CSC markers and sensitized cells to chemotherapy, despite modest cell killing29. In comparison, our study showed that CSCs became enriched in PDT-resistant populations, which accumulated significantly less PPIX (Fig. 4D). These results highlight the importance of PPIX accumulation on efficacy of treatment and the consideration of CSC heterogeneity between cancers.

CD133 was used as a CSC marker in this study as we found that the levels of CD133 + cells in all three cell lines were sufficient (1-1.9%, Fig. 1) to conduct our proposed experiments. Other CSC markers such as CD24, CD32, CD44 and ALDH1A1 were not commonly expressed in the cell lines. As CD133 has been widely used as a CSC marker for different types of cancer, we utilized CD133 as a CSC marker of DLD-1 cells to determine CSC functions throughout the study. However, our research objective was not to specifically characterize the roles of CD133 and its downstream signalling pathways in 5-ALA-PDT resistance. The CSC populations (CD133 + cells) were low in the cancer cell lines (1–3%) (Fig. 1) and even in PDT×4 DLD-1 cells (8%) in Fig. 4C. Generally, the CSC population in tumors range from less than 0.02 to 25% depending on the tumor types30.

Our current study falls short to identify the precise mechanisms underlying 5-ALA-PDT resistance in CSCs. There are several possible mechanisms that could lead to CSCs resistant against 5-ALA-PDT. First, as mentioned above, CSCs could be resistant to 5-ALA-PDT due to high expression of the ABC transporters as they are commonly overexpressed by CSCs11,31. High expression of the ABC transporters may decrease PPIX accumulation in CSCs induced by 5-ALA treatment and subsequently the sensitivity of CSCs to 5-ALA-PDT. In a study examining the efficacy of photodiagnosis of cancers using 5-ALA (5-ALA PDD), Kawai et al. reported that the CSC populations of PANC-1 cells expressing high levels of ABCG2 were responsible for decreased PPIX accumulation32. We observed that 5-ALA-PDT resistant cells accumulated lower amounts of PPIX (Fig. 4D), suggesting that this mechanism underlies CSC resistance to 5-ALA-PDT. Whether the reduced accumulation of PPIX in CSCs in our results can be attributed to efflux by transporters or reduced PPIX biosynthesis remains unknown. An inherent reduction in the production of PPIX by CSCs would justify the use of photosensitizers that do not require metabolic processing. Secondly, high expression of stress response genes in CSCs could be one of the resistant mechanisms33,34. CSCs have increased genetic diversity to adapt and survive under stress conditions such as lack of oxygen, starvation, and exposure to DNA-damaging compounds35,36,37,38. Moreover, CSCs have increased activity of autophagy, which promotes cellular survival by enabling the ability to overcome stress conditions through nutrient recycling and preventing the accumulation of damaged cellular components39,40,41,42. These survival mechanisms of CSCs may reduce cell death caused by 5-ALA-PDT. Fourth, CSCs have lower levels of reactive oxygen species (ROS) compared to their differentiated counterparts, which is the essential effector for cancer cell death induced by 5-ALA-PDT43. Although our current study did not identify the cellular mechanisms that underlie CSC resistance to 5-ALA-PDT, it is essential to improve the efficacy of 5-ALA-PDT and reduce cancer recurrence after 5-ALA-PDT. Combined treatment with 5-ALA-PDT and inhibitors targeting ABC transporters has been shown to increase anticancer efficacy in vitro and in vivo, which may be an effective way to eliminate CSCs by 5-ALA-PDT44,45,46. However, this approach has shown limited success clinically due to off-target cytotoxicity and related complications with increased sensitivity to chemotherapeutics47. As the development of novel ABC transporter inhibitors for clinical trials is an active area of research, it would be a feasible idea to evaluate inhibitors in combination with 5-ALA-PDT in clinical settings in the near future if these barriers are overcome48.

To clarify the mechanism of CSC resistance to 5-ALA-PDT, this study warrants further investigation. It would be essential to determine cellular localization and quantification of PPIX, activities of enzymes involved in the heme pathway, cellular ROS levels, activation of cell death pathways, expression of ABC transporters and expression of stress response genes. Furthermore, to translate our findings to a clinical setting, it would be important to determine if CSCs are responsible for 5-ALA-PDT resistance and tumor recurrences in animal models of cancer.