Quantification of Protoporphyrin IX Accumulation in Glioblastoma Cells: A New Technique

Introduction. 5-Aminolevulinic Acid (5-ALA) is a precursor of heme synthesis. A metabolite, protoporphyrin IX (PpIX), selectively accumulates in neoplastic tissue including glioblastoma. Presurgical administration of 5-ALA forms the basis of fluorescence-guided resection (FGR) of glioblastoma (GBM) tumors. However, not all gliomas accumulate sufficient quantities of PpIX to fluoresce, thus limiting the utility of FGR. We therefore developed an assay to determine cellular and pharmacological factors that impact PpIX fluorescence in GBM. This assay takes advantage of a GBM cell line engineered to express yellow fluorescent protein. Methods. The human GBM cell line U87MG was transfected with a YFP expression vector. After treatment with a series of 5-ALA doses, both PpIX and YFP fluorescence were measured. The ratio of PpIX to YFP fluorescence was calculated. Results. YFP fluorescence permitted the quantification of cell numbers and did not interfere with 5-ALA metabolism. The PpIX/YFP fluorescence ratio provided accurate relative PpIX levels, allowing for the assessment of PpIX accumulation in tissue. Conclusion. Constitutive YFP expression strongly correlates with cell number and permits PpIX quantification. Absolute PpIX fluorescence alone does not provide information regarding PpIX accumulation within the cells. Our research indicates that our PpIX/YFP ratio assay may be a promising model for in vitro 5-ALA testing and its interactions with other compounds during FGR surgery.


Introduction
Glioblastoma multiforme (GBM) is the most common and aggressive primary malignant brain tumor. With the current standard of care, recurrence is inevitable and less than 5% of GBM patients will survive five years after diagnosis [1]. Subtotal surgical resection may contribute to recurrence as tumor initiating cells are left in place. A valuable tool to achieve gross total resection and thereby increase progression-free survival is the use of 5-Aminolevulinic Acid (5-ALA) for fluorescence-guided resection (FGR) [2]. 5-ALA is a substrate of the heme biosynthesis pathway and is the source of carbon for porphyrin synthesis [3]. Malignant gliomas metabolize 5-ALA and accumulate the fluorescent compound, protoporphyrin IX (PpIX) [4]. The presence of PpIX fluorescence within the tumor bed allows for discrimination between neoplastic and nonneoplastic cells [4]. Yet, not all brain tumors accumulate sufficient quantities of PpIX to make the use of 5-ALA advantageous during surgery [5,6], and it is not fully understood how 5-ALA might affect other physiological processes. At this time, it is unknown if prescribed medications of tumor patients or if drugs given to patients just prior to/during surgery interfere with 5-ALA synthesis and/or PpIX accumulation. It is difficult, however, to test for the effects of these potential drug interactions during in vitro analyses, because fluorescence of PpIX is merely an indication of presence/absence of accumulation. PpIX fluorescence detection alone does not provide understanding about the number of metabolically active cells in an assay. This lack of information coupled with the great potential of 5-ALA in FGR suggested that a quantitative analysis of the metabolism of 5-ALA to fluorescent PpIX in tumors cells would improve the methodologies used to detect neoplasms during surgery. To accomplish this, we developed an in vitro assay using a YFP-tagged glioblastoma cell line. This cell line was then used to measure the rate of 5-ALA incorporation and conversion to protoporphyrin IX.

Cell
Lines. U87MG cells (ATCC, Manassas, VA, USA) were cultured in phenol red-free Eagle's minimum essential medium (EMEM) with L-glutamine (Lonza, Portsmouth, NH, USA) and 10% fetal bovine serum (PAA, Westborough, MA, USA) at 37 ∘ C and 5% CO 2 . U87MG cells were stably transfected with the YFP expression vector (pEYFP-C1) to create U87-YFP cells using the Neon transfection system (Invitrogen, Life Technologies, Carlsbad, CA, USA) followed by selection with G418 antibiotic (Gold Biotechnology, St. Louis, MO, USA). U87MG transfection with YFP was conducted as follows: U87MG cells were harvested using Trypsin-Versene (Lonza, Portsmouth, NH, USA), counted, and resuspended in Resuspension Buffer R (Neon transfection system, Invitrogen, Life Technologies, Carlsbad, CA, USA) to a cell concentration of 5 × 10 6 cells/mL. A cell+plasmid solution (9.5 uL resuspended cells and 0.5 uL of the pEYFP-plasmid at a concentration of 1.7 ug/mL) was prepared and 10 uL was aspirated into a Neon reaction tip. The Neon reaction tip was inserted into a Neon transfection tube filled with approximately 3 mL of Electrolytic Buffer (Neon transfection system, Invitrogen, Life Technologies, Carlsbad, CA, USA). Transfection was performed using a 1,300 pulse voltage with a 30 ms pulse width, as a single pulse. The 10 uL of transfected U87s was then pipetted into a prewarmed EMEM + 10% FBS media solution lacking antibiotics in a 96-well plate. Cells were allowed to settle and incubate for 24 hours, and then the EMEM + 10% FBS solution was replaced with media containing 0.8 mg/mL G418 antibiotic. After approximately two weeks, all of the remaining cells expressed YFP. A maintenance level of G418 (0.2 mg/mL) was then used to maintain selection pressure for YFP expressing U87s. As cells became confluent they were transferred to 24well plates, then 6-well plates, and then finally T25 culture flasks.

Statistical Analysis.
A Pearson correlation coefficient was calculated to determine a relationship between cell number and YFP fluorescence. For PpIX florescence, individual wells from each assay were analyzed as individual values. A repeated measures analysis of variance (ANOVA) was used to analyze the differences between 5-ALA treatment concentrations and time points over 72 hours (SPSS, Version 18, IBM, Armonk, NY, USA). Post hoc least significant difference (LSD) analyses were used to determine statistical difference between concentrations and time points. All values are expressed as a mean ± standard error, and significant differences were determined at the 0.05 level.

Relationship between YFP and Cell Number.
To determine if yellow fluorescence emission at 580-640 nm would be a reliable mechanism to determine cell number in a well, a standard curve of U87-YFP cells was prepared. A Pearson correlation coefficient was calculated to demonstrate the relationship between YFP fluorescence and number of cells in an assay plate well. A strong positive correlation was found ( = 0.801 ± 0.053, < 0.01), indicating a significant linear relationship between YFP RLU and the number of cells plated into a microplate well.

Discussion
There are several novel findings in this study: (1) constitutive YFP fluorescence strongly correlates with cell number, (2) in vitro PpIX fluorescence increases over time reflecting increases in cell number, (3) utilizing constitutive YFP expression in cells allows for attainable PpIX quantification in vitro, and (4) 5-ALA treatments ≥ 25 mM generated highest amounts of PpIX/cell while decreasing the number of viable cells. The utilization of YFP allowed for the detection of changes in cell numbers without using traditional techniques which typically damage or kill the cells-now longitudinal designs are feasible with the same assay plate. Our lab previously determined that YFP does not interfere with 5-ALA metabolism and little to no cross-talk is detectible when exciting these compounds. Thus, by making use of the PpIX/YFP fluorescence ratio, relative PpIX levels per cell are more accurately estimated and allow for the assessment of PpIX accumulation or lack thereof.
Because YFP was demonstrated to be a reliable indicator of cell number, a YFP expressing cell line provides a new model for quantifying PpIX accumulation in vitro. A method to quantify PpIX accumulation would allow researchers to measure of the amount of PpIX per cell. Without this quantification, PpIX fluorescence is merely an indication of absence or presence. During a time course study, for example, the number of cells in a well of an assay plate changes (increases due to cell division or decreases due to cell death). If PpIX fluorescence is recorded without accounting for the change in cell number, there is no way to determine if PpIX is accumulating within these cells. Our data suggests that higher concentrations of 5-ALA (Figure 2) may have a toxic effect in cells, yet great PpIX accumulation within living cells of the assay. These toxic effects may warrant future investigations of 5-ALA administration as a targeted therapeutic agent.
While the 5-ALA FGR method is viewed as a more efficient means for tumor resection, there are limitations. It is not fully understood how 5-ALA interacts with other physiological processes and other pharmacological agents when given systemically. 5-ALA use has been questioned because PpIX fluorescence is weak or even absent in some patients. This low PpIX fluorescence is possibly due to interactions of 5-ALA with drugs and/or other compounds circulating within tumor patients at the time of FGR surgery. Low PpIX fluorescence may also be due to the lack of individualized (patient by patient) standards regarding the presurgical administration of 5-ALA to ensure peak PpIX fluorescence. Currently, 20 mg/kg of 5-ALA per body weight is administered about 3 hours prior to surgery [7]. To our knowledge, alternative dosing strategies have not been tested for efficacy. The PpIX/YFP method may allow for dosage focused studies to be conducted. Our data indicates that new studies designed to monitor PpIX fluorescence over time should address cell growth/death during the testing process-PpIX fluorescence in vitro is dependent on the number of viable cells (Figure 3). Without understanding about the number of metabolically active cells in an assay, the fluorescence of PpIX is merely an indication of presence or absence. Here, PpIX fluoresce values increased over time (Figure 3(a)); however, YFP fluorescence values also increased over time (Figure 3(b)). Further, when comparing relative PpIX fluorescence (Figure 3(c)), the amount of PpIX per cell did not change. This knowledge allows researchers to now investigate novel approaches to optimize fluorescence. Optimizing PpIX fluorescence may improve surgical outcomes and significantly increase the number of brain tumors that accumulate sufficient quantity of PpIX for FGR.

Conclusion
In summary, we showed that constitutive YFP expression strongly correlates with cell number and can be used to normalize PpIX fluorescence in vitro. We also showed that recording PpIX fluorescence alone does not provide information regarding accumulation of PpIX within cells, and we discovered that high levels of 5-ALA treatment reduce cell numbers during the time course warranting future investigations of 5-ALA administration as a therapeutic agent. Our PpIX/YFP ratio method is a feasible approach for in vitro 5-ALA analysis and may provide a new model for testing interactions of 5-ALA with other compounds circulating within tumor patients at the time of FGR surgery. These interactions may determine why some tumors fluoresce and others do not.