Multi-sample measurement of hyperpolarized pyruvate-to-lactate flux in melanoma cells

Abstract

Hyperpolarized [1-¹³C] pyruvate can be used to examine the metabolic state of cancer cells, highlighting a key metabolic characteristic of cancer: the upregulated metabolic flux to lactate, even in the presence of oxygen (Warburg effect). Thus, the rate constant of ¹³C exchange of pyruvate to lactate, k_PL, can serve as a metabolic biomarker of cancer presence, aggressiveness and therapy response. Established in vitro hyperpolarized experiments dissolve the probe for each cell sample independently, an inefficient process that consumes excessive time and resources. Expanding on our previous development of a microcoil with greatly increased detection sensitivity (10³-fold) compared with traditional in vitro methods, we present a novel microcoil equipped with a 10-μL vertical reservoir and an experimental protocol utilizing deuterated dissolution buffer to measure metabolic flux in multiple mass-limited cell suspension samples using a single dissolution. This method increases efficiency and potentially reduces the methodological variability associated with hyperpolarized experiments. This technique was used to measure pyruvate-to-lactate flux in melanoma cells to assess BRAF-inhibition treatment response. There was a significant reduction of k_PL in BRAF^V600E cells following 24 and 48 hours of treatment with 2 μM vemurafenib (P ≤ .05). This agrees with significant changes observed in the pool sizes of extracellular lactate (P ≤ .05) and glucose (P ≤ .001) following 6 and 48 hours of treatment, respectively, and a significant reduction in cell proliferation following 72 hours of treatment (P ≤ .01). BRAF inhibition had no significant effect on the metabolic flux of BRAF^WT cells. These data demonstrate a 6–8–fold increase in efficiency for the measurement of k_PL in cell suspension samples compared with traditional hyperpolarized in vitro methods.

1 |. INTRODUCTION

Magnetic resonance spectroscopy (MRS) of ¹³C-labeled hyperpolarized (HP) probes provides an unparalleled real-time visualization of in vitro and in vivo metabolic flux.¹ Compared with MRS performed at thermal equilibrium, hyperpolarization enhances the signal of a molecule by several orders of magnitude. Dynamic nuclear polarization (DNP), the most established method for hyperpolarization, provides a potential 10 000-fold increase in signal.² This increase in sensitivity enables measurement of time-resolved conversion kinetics of a hyperpolarized substrate through enzyme-catalyzed reactions.³ [1-¹³C] pyruvate is the most widely studied probe for DNP-MRS; this is in part due to its ability to be polarized efficiently, relatively long T₁ (~1 minute at 3 T),⁴ rapid cellular uptake and fundamental role in metabolism. HP pyruvate has been examined for its potential clinical value in the detection of liver disease,⁵ cardiac ischemia^6,7 and, most extensively, oncology.^8–12 This latter application relates to the Warburg effect,¹³ a metabolic hallmark of cancer¹⁴ characterized by increased conversion of pyruvate to lactate, even in the presence of oxygen.¹⁵ Pyruvate-to-lactate flux may therefore serve as a useful biomarker of cancer presence, aggressiveness and treatment response.¹⁶

We previously developed a microcoil¹⁷ featuring a micro-reservoir in the horizontal plane for quantitative measurement of metabolic flux in live cells. The microcoil system greatly increases detection sensitivity (10³-fold) compared with traditional in vitro methods, and thus can measure HP lactate signal using fewer cells than traditional in vitro DNP-MRS experiments that also require the HP probe to be dissolved for each sample independently. The process of assembling and executing a dissolution experiment is costly in both time and resources. Additionally, each new dissolution introduces fluctuations in variables such as the polarization and concentration of the dissolved HP probe and the shim of the MR spectra acquired. These variables affect the signal-to-noise ratio (SNR) of the product peaks detected, and thus the confidence in the calculated metabolic flux for each experiment. Therefore, a protocol that allows multiple samples to be analyzed using a single HP dissolution would offer multiple benefits.

The number of cell suspension samples that can be examined using a single HP dissolution is limited by the spin–lattice relaxation time (T₁) of the HP probe. The T₁ of a ¹³C nucleus can be extended by substitution of neighboring protium nuclei (¹H) with deuterium (²H), either via synthesis of deuterium-labeled analogues or, more easily, by dissolving the ¹³C-labeled probe in D₂O.^18–20 Our laboratory has previously characterized this latter, nonsynthetic approach to extend the T₁ and T₂ of glutamine, urea and arginine with a demonstration of improved in vivo SNR.²¹

Expanding on the previously designed microcoil, we redesigned the micro-reservoir in the vertical plane to facilitate rapid loading and removal of consecutive samples. The aim of this work was to develop an experimental protocol to enable measuring metabolic flux in multiple mass-limited cell suspension samples, thereby increasing experimental efficiency and providing greater control of the methodological variability associated with HP experiments. This novel method exploits the use of deuterated dissolution buffer in combination with a microcoil design conducive to rapid turnover of samples. Following development of the experimental protocol, we applied this technique to the measurement of pyruvate-to-lactate flux in melanoma cells for the assessment of BRAF-inhibition (BRAFi) treatment response.

2 |. EXPERIMENTAL

2.1 |. Microcoil fabrication

A microwell was designed with a diameter of 1.5 mm and height of 1.0 mm, and it was connected to a microfunnel structure (bottom/top diameter: 1.5/4.0 mm, height: 2 mm) for easy loading of samples. The design was fabricated with a 3D printer (Micro Plus HD, EnvisionTec). The magnet wire (32 American Wire Gauge) was wound four times around the printed microwell structure to form a vertical microcoil, and it was connected to a custom-designed electronic circuit board (Advanced Circuits) for MRS signal detection (Figure 1A). Variable and fixed capacitors were mounted to make a dual-tuned ¹H/¹³C resonance circuit with the microcoil, and a mechanical switch was mounted to control the resonance mode.

FIGURE 1.

A, photograph of the microcoil circuit board. Scale bar: 10 mm. B, schematic demonstrating the experimental protocol for the measurement of k_PL in multiple cell suspension samples. A ¹³C NMR pulse sequence was designed to acquire an FID every 250 ms for the entirety of the experiment. Following addition of HP [1-¹³C] pyruvate, the first cell suspension/pyruvate mixture was loaded into the microwell, and the microcoil moved to the center of the MRI scanner for ~ 10 seconds of signal acquisition. The microcoil was removed from the magnetic field and the cell suspension/pyruvate mixture removed from the microwell. Simultaneously, the subsequent cell suspension sample was mixed with HP [1-¹³C] pyruvate for loading into the microwell and the process was repeated. C, representative HP-MR spectra acquired every 250 ms for a single cell suspension sample (left); example plot of peak-areas for pyruvate, pyruvate hydrate and lactate demonstrating periods of acquisition and reloading of the microcoil with the following cell suspension sample (right)

2.2 |. Cell culture

SK-mel 28 and SK-mel 24 cells were grown in Gibco Dulbecco’s modified Eagle’s medium (Thermo Fisher Scientific, 11965–092), supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin and incubated at 37°C in 5% CO₂. Cell number was measured using a Cellometer Mini Cell Counter (Nexcelom Bioscience); the cell suspension was mixed at a 1:1 ratio with the trypan blue solution before the measurement. For all BRAFi experiments, SK-mel 28 and SK-mel 24 cells were treated with a 2 μM dose of vemurafenib, based on previous literature.²² For HP microcoil experiments, cells were plated 24 hours before treatment and incubated overnight to ensure cell adherence. SK-mel 28 and SK-mel 24 cells were treated with either vemurafenib or 0.1% dimethyl sulfoxide (DMSO, vehicle) for 24 or 48 hours. Immediately before the dissolution of HP [1-¹³C] pyruvate, cells were trypsinized, counted and divided into cell pellets of 2 × 10⁶ per sample. The order of samples loaded using the microcoil multi-shot protocol (Figure 1B) alternated between vemurafenib and vehicle treatment.

2.3 |. Incucyte in vitro cell growth measurements

An Incucyte HD system (Essen BioScience) was used to assess the effect of vemurafenib on SK-mel 28 and SK-mel 24 cell proliferation. Cells were plated in 96-well plates (2000 cells per well) and incubated overnight to ensure cell adherence. Twenty-four hours later, the cells were washed with phosphate-buffered saline (PBS) and the media exchanged with complete media containing 2 μM vemurafenib or 0.1% DMSO (vehicle). Frames were captured at 3-hour intervals, for a total of 72 hours, from four separate 950 × 760 μm² regions per well. Confluence was measured using the Incucyte software. Values from the four regions of each well were pooled and averaged across four replicate wells. Results are expressed graphically as fold increase of cell growth relative to the initial posttreatment confluence.

2.4 |. Western blot analysis

SK-mel 24 and SK-mel 28 cells were plated on 100-mm² dishes (~1.2 × 10⁶ cells) and incubated overnight to ensure cell adherence. The cells were washed with PBS and treated with vemurafenib or 0.1% DMSO (vehicle) for 6, 24, 48 or 72 hours (n = 3). Cells were then washed with PBS, lysed in radioimmunoprecipitation assay buffer containing protease and phosphate inhibitors (Thermo Fisher Scientific), left on ice for 30 minutes then concentrated by centrifugation at 14 000 rpm at 4°C for 10 minutes. The concentration of protein lysates was quantified using the bicinchoninic acid assay. For Western blot analysis, 25 μL of 20 μg/μL cell lysate was mixed with 25 μL 2x Laemmli loading buffer. Separated proteins in the gels were electrophoretically transferred onto polyvinylidene fluoride membranes. After washing and blocking, p-MEK1/2, p-ERK1/2, ERK1/2 and b-actin antibodies (1 mg/mL; Cell Signaling Technology) diluted in SuperBlock (TBS) blocking buffer (Thermo Fisher Scientific) were added and incubated overnight at 4°C. The bound antibodies were detected by horseradish peroxidase-conjugated antigoat Ig secondary antibody (Cell Signaling Technology) followed by ECL detection system (Thermo Scientific) according to the manufacturer’s instructions. Densitometry analysis was performed using ImageJ software; data were normalized to actin and expressed relative to the control.

2.5 |. ¹H NMR analysis of cell media

Cells were plated in six-well plates (2 × 10⁵ cells per well) and incubated overnight to ensure cell adherence. The cells were washed with PBS and the media were exchanged with complete media containing vemurafenib or 0.1% DMSO (vehicle) for 6, 24 or 48 hours prior to media collection (n = 6 replicates per group). A cell count was taken at the time of media collection (n = 3 per group) and used for normalization of the lactate and glucose concentrations. Collected media were initially stored at −80°C; on the day of ¹H nuclear magnetic resonance (NMR) analysis, 800 μL of sample was thawed and filtered through prewashed filters (Amicon Ultra 0.5 mL centrifugal filters, 3 K; Merck Millipore Ltd.) at 14 000 rpm at 4°C for 30 minutes. Media samples (560 μL) were combined with 140 μL phosphate buffer (pH 9) in D₂O. The final mixture contained 0.5 mM sodium trimethylsilylpropanesulfonate as a chemical shift reference, 3mM sodium azide as a bacteriostatic agent and 10 mM imidazole as a pH indicator. Experiments were performed on a 14.1 T NMR spectrometer equipped with an autosampler and ¹H cryoprobe (Bruker Biospin). ¹H NMR spectra were acquired with a water presaturation recycle delay of 4 seconds, acquisition time of 2.67 seconds, 90° excitation and 64 averages. Spectral data processing and extracellular lactate and glucose quantification were carried out in Chenomx NMR Suite 8.0 Professional (Chenomx Inc.).

2.6 |. Hyperpolarization using DNP and MRS acquisition

Hyperpolarization of a 20-μL mixture of [1-¹³C] pyruvate (14.2 M, GE Healthcare) and GE trityl radical (15mM, GE Healthcare) was conducted by the DNP method using a SPINlab polarizer (5 T, 0.80 K, GE Healthcare). Following polarization, the frozen sample was rapidly dissolved in 10 mL of TRIS buffer consisting of 100 mM TRIS, 1 mM EDTA (Fisher Scientific, USA), pH 7.4 in D₂O (Cambridge Isotope Laboratories), as described previously.²¹ A buffer was also prepared using H₂O to compare the T₁ of HP [1-¹³C] pyruvate dissolved in deuterated with nondeuterated buffer and the effect of deuterated buffer on measured k_PL in SK-mel 28 cells. Before each experiment, the microcoil was calibrated with first-order shimming in the ¹H resonance mode. The resonance mode was then changed to ¹³C mode prior to ¹³C MRS acquisition.

A summary of the developed experimental protocol for measuring the k_PL of multiple cell suspension samples is shown in Figure 1B. The dissolution was neutralized in a receiving vial containing sodium hydroxide such that the final solution was pH 7.4. A vertical micro-reservoir was designed to accommodate a reloadable well for measurement. HP [1-¹³C] pyruvate was mixed with the first sample (cell suspension) with a ratio of 1:5 and the mixture was loaded into the well of the microcoil, producing a final concentration of ~ 10 mM pyruvate (10 μL total volume). The exact concentration of pyruvate is calculated for each dissolution and used as a parameter for the pyruvate-lactate conversion rate calculation. The microcoil was moved to the center of the MRI scanner, and the ¹³C NMR pulse sequence acquisition was initiated. Following 10 seconds of signal acquisition, the microcoil was removed from the magnetic field and the cell suspension/HP pyruvate mixture was removed from the microcoil well. Concurrently, the next cell suspension sample was mixed with HP [1-¹³C] pyruvate in preparation for loading into the microcoil and then the process of signal acquisition was repeated.

Time-resolved DNP-MRS data were acquired using a 3 T MRI system (Bruker); a ¹³C free induction decay (FID) was acquired every 250 ms using a 15° excitation. The T₁ of HP [1-¹³C] pyruvate was measured using a 1 T Spinsolve spectrometer (Magritek, San Diego, CA, USA) with a 5-mm ¹H/¹³C coil.

2.7 |. HP spectral data analysis

Spectroscopic data were processed using custom software developed in-house (MATLAB R2018a, MathWorks). The acquired data were transformed without zero filling or line broadening. The complex HP spectra were fit to a Lorentzian model that included baseline offset, first-order phase, center frequency, T₂*, and peak area for [1-¹³C] pyruvate, [1-¹³C] pyruvate hydrate and [1-¹³C] lactate. Metabolite dynamics were calculated by averaging the peak areas of four consecutive time points to generate data with 1-second temporal resolution. The fraction of HP [1-¹³C] pyruvate converted to [1-¹³C] lactate was calculated using the following formula for the lactate ¹³C signal ratio: lactate/(pyruvate + pyruvate hydrate + lactate). We have previously defined this conversion rate multiplied by the concentration of pyruvate as a metabolic flux metric ξ.¹⁷

2.8 |. Statistical analysis

Statistical analyses and data visualization were performed using R (http://www.R-project.org/; packages: ggplot2, dplyr) and GraphPad Prism. The significance of comparisons of the T₁ of HP [1-¹³C] pyruvate in deuterated vs nondeuterated buffer, ξ (pmols/s) measurements, in vitro cell growth and ¹H NMR extracellular concentrations were determined by Welch’s t-test (two-tailed), with treatment and vehicle compared at each time point independently. Densitometry data were compared using a one-way ANOVA with Dunnett’s multiple comparison test.

2.9 |. Protocol development

The use of deuterated buffer for dissolution significantly extended the T₁ of HP [1-¹³C] pyruvate from 73.8 ± 2.0 to 147.4 ± 6.3 seconds (Figure S1; P ≤ .0001). This extended T₁ relaxation rate facilitated measuring metabolic flux in successive samples following a single dissolution. In the current study, dynamic spectra could be acquired for 6–8 samples, thus allowing for n = 3–4 per group using a two-group comparison experimental design.

The signal decays for the first three samples of an experiment using a single dissolution are shown in Figure 1C, with representative ¹³C NMR spectra for a single cell suspension sample showing the resonances for HP [1-¹³C] pyruvate (171.1 ppm, dark blue shading), HP [1-¹³C] pyruvate hydrate (179.6 ppm, light blue shading) and HP [1-¹³C] lactate (183.4 ppm, red shading) acquired every 250 ms for a total acquisition time of ~10–12 seconds, which provides enough data points to reproducibly calculate metabolic flux (Figure 2B).

FIGURE 2.

A, HP-MR spectral data acquired every 250 ms from eight independent cell suspension samples (left) are averaged to provide one peak area per molecule per second (right). Box represents the four sample measurements represented in B. B, the first 5 seconds of acquisition were used to quantify metabolic flux in SK-mel 28 BRAF^V600E cells. C, the lactate ratio scales linearly with cell number (SK-mel 28 cells). D, no difference was observed in the metabolic flux of SK-mel 28 cells, ξ, between D₂O and H₂O solvation for the dissolution buffer (data expressed as median, first and third quartiles). E, pyruvate-to-lactate flux (ξ) in SK-mel 28 and SK-mel 24 cells, following 24- and 48-hour treatment with 2 μM vemurafenib or vehicle control (data are presented as mean ± SEM); *P ≤ .05

Figure 2A shows representative peak-area decays for HP [1-¹³C] pyruvate hydrate and HP [1-¹³C] lactate as acquired (left), and after removal of “reload” spectra acquired while there was no sample inside the scanner then averaging to achieve 1-second temporal resolution (right). For SK-mel 28 cells, the lactate ratio was calculated for the first 5 seconds and the slope of a linear model fit was used to approximate the initial conversion rate of pyruvate to lactate, k_PL (Figure 2B). This conversion rate scales linearly according to cell number (R² = 0.866, Figure 2C); to compare samples, this calculated metric is normalized to the number of cells loaded to give ξ in pmols/s per 10⁵ cells. Solvation of the dissolution buffer with D₂O was not found to significantly affect the measured pyruvate-lactate flux in SK-mel 28 cells compared with H₂O solvation (P = .45; Figure 2D).

The R² of the linear model was used to determine confidence in the calculated rate; an R² cut-off of 0.4 was used to discard low-SNR sample acquisitions from statistical comparison. [1-¹³C] lactate generation is assumed to be linear at the beginning of the reaction while there is a surplus of substrate. To abide by first-order rate kinetics, we used data for the duration that the conversion was linear when fitting the model, as described previously.¹⁷ The rate kinetics were linear for 5 seconds on average for SK-mel 28 cells and 10–12 seconds for SK-mel 24 cells. No comparisons were made between cell types; comparisons were only made within a given cell type.

3 |. RESULTS

3.1 |. Rapid quantitative assessment of BRAFi treatment response in BRAF^V600E and BRAF^WT melanoma cells

Treatment with 2 μM vemurafenib for both 24 and 48 hours led to a significant reduction in the flux, ξ, of HP [1-¹³C] pyruvate to [1-¹³C] lactate in BRAF^V600E SK-mel 28 cells (P ≤ .05). No significant difference (P = .80) was observed in metabolic flux, ξ, in BRAF^WT SK-mel 24 cells following 24 hours of BRAFi treatment; however, a 48-hour treatment resulted in a trend of increased HP [1-¹³C] pyruvate to [1-¹³C] lactate flux (P = .21; Figure 2E).

Treatment with 2 μM vemurafenib for 72 hours significantly inhibited cell growth in BRAF^V600E SK-mel 28 cells (P = .003), but did not inhibit cell growth in BRAF^WT SK-mel 24 cells (Figure 3A). Following 6 hours of treatment with 2 μM vemurafenib, Western blot analysis of p-MEK1/2 and p-ERK1/2 showed target inhibition in BRAF^V600E SK-mel 28 cells and upregulation in BRAF^WT SK-mel 24 cells (Figure 3B).

FIGURE 3.

Analyses of SK-mel 28 (BRAF^V600E) and SK-mel 24 (BRAF^WT) cells following treatment with 2 μM vemurafenib. A, cell proliferation analysis (data are presented as mean ± SD). Cell proliferation decreased significantly for SK-mel 28 cells following 72 hours of treatment with vemurafenib; SK-mel 24 cells showed no change following 72 hours of treatment. B, vemurafenib-mediated inhibition of the MAPK/ERK pathway in BRAF^V600E cells was confirmed via a reduction in protein expression level detected with Western blot (upper) and densitometry analysis (lower, data are presented as mean ± SEM). C, ¹H NMR spectroscopy measurements of media concentrations of lactate and glucose (data are presented as mean ± SEM); *P ≤ .05, **P ≤ .01, ***P ≤ .001, ****P ≤ .0001

Following BRAFi treatment, ¹H NMR analysis of BRAF^V600E SK-mel 28 cell media samples showed a significant reduction in extracellular lactate after 6 hours (P ≤ .05) and a significant reduction of glucose consumption after 48 hours (P ≤ .001). In BRAF^WT cells, both extracellular lactate and glucose consumption significantly increased after 48 hours of treatment with 2 μM vemurafenib (P ≤ .0001; Figure 3C).

4 |. DISCUSSION

In the experimental protocol described here, deuterated solvation of HP [1-¹³C] pyruvate combined with a highly sensitive microcoil system with a micro-reservoir in the vertical plane, provided a means of increasing the throughput of in vitro DNP-MRS experiments. Traditional HP experiments using cell suspensions require more than 10⁷ cells,^22–25 which utilizes excessive resources and incubator space. The microcoil system allows for reproducible measurements of k_PL with ~10⁵ cells. The increased sensitivity of the microcoil relative to traditional HP cell suspension methodology may be particularly useful for slow-growing patient-derived primary cells or organoid cultures, for which biological material is limited. In this respect, this method could be a useful tool for assessing multiple treatment regimens using patient-derived material for advanced personalized medicine.²⁶

The use of deuterated dissolution buffer significantly extended the spin–lattice relaxation time of [1–¹³C] pyruvate, without affecting the calculated k_PL in BRAF^V600E cells compared with H₂O solvation. This increase in T₁ is in agreement with previous literature^21,27–29 and is due to the reduction in dipolar relaxation resulting from deuteration of exchangeable protons. This effect arises because the gyromagnetic ratio of deuterium is ~7-fold weaker than that of protium, leading to weaker dipolar interactions.³⁰ Previous methods of multiple-sample spectral acquisition using a single dissolution have been limited to two³¹ or three³² independent samples. The method we present here, combining a microcoil design allowing for rapid cell suspension sample load/unload/reload capability and the extended hyperpolarized lifetime of [1-¹³C] pyruvate dissolved in D₂O, enabled up to eight reproducible high-throughput measurements of pyruvate-to-lactate metabolic flux in multiple independent samples using a single HP dissolution. The benefits of this approach include a reduction in costs and expedited experimental progress, with the potential for a two-group comparative in vitro study (n = 4) to be completed with one dissolution.

Mutations of the BRAF kinase are the most common genetic alterations in melanoma, with BRAF V600E the most frequent,³³ leading to sustained activation of the mitogen-activated protein kinase (MAPK/ERK) pathway that promotes cell growth and proliferation when combined with additional genetic events.³⁴ Vemurafenib inhibits the MAPK/ERK pathway by binding with high affinity to the mutated form of BRAF, thus preventing phosphorylation of MEK1 and MEK2 and leading to cell cycle arrest and apoptosis.^35–37 A 24-hour treatment with 2 μM vemurafenib led to a significant reduction in the rate of HP pyruvate-to-lactate conversion in BRAF^V600E SK-mel 28 cells. This is consistent with the observed decrease in the pool size of extracellular lactate following 6 hours of treatment, decrease in glucose consumption following 48 hours of treatment, and decrease in cell proliferation following 72 hours of treatment. Altered cell metabolism is a key component of the downstream consequence of activating BRAF mutations in melanoma,³⁸ with evidence of reduced oxidative phosphorylation activity and increased glycolysis.³⁹ The data presented here are in agreement with the current hypotheses that indicate treatment with vemurafenib leads to a reduction of in vitro aerobic glycolysis and glucose consumption in BRAF-mutant melanoma cells.^22,23

It has been argued that transport of pyruvate into the cell via monocarboxylate transporter (MCT) 1 is the rate-determining step for conversion of extracellular pyruvate to lactate.⁴⁰ Indeed, in a previous analysis of vemurafenib-treated WM266.4 (BRAF^V600D) cells, reduced hyperpolarized pyruvate-to-lactate exchange was attributed to reduced MCT1 expression.²² Thus, the BRAFi-induced reduction in pyruvate-to-lactate flux in BRAF^V600E cells likely reflects either a change in expression of lactate dehydrogenase, the transporters MCT1 and MCT4, or a combination of these factors.²⁵

Following a 24-hour treatment with vemurafenib, BRAFi-insensitive BRAF^WT cells showed no significant change in HP pyruvate-to-lactate flux, cell proliferation and ¹H NMR media lactate and glucose pool sizes. However, in BRAF^WT cells, a 48-hour treatment resulted in a significant increase in lactate efflux and glucose consumption and a trend of increased HP pyruvate-to-lactate flux. Furthermore, vemurafenib treatment caused an increase in ERK activation in SK-mel 24 cells, which is consistent with previous analyses of BRAF^WT cell lines.⁴¹ RAF inhibitors have been shown to paradoxically increase MAPK/ERK activation in cells with wild-type BRAF through transactivation of RAF dimers.^35,42,43 Our results support increased glycolysis as a downstream metabolic consequence of increased BRAFi-driven signaling in BRAF^WT cells.

5 |. CONCLUSIONS

The combination of reloadable microcoil and D₂O solvation presented here increased the efficiency of in vitro HP metabolic flux measurements in independent samples, with an 8-fold increase in samples from a single dissolution. A potential benefit of this experimental setup is the reduction in nonbiological methodological variability associated with HP experiments, thereby increasing the power to detect biological differences between samples. The multiple-sample protocol allowed for a two-group comparative in vitro study (n = 4) to be completed with one dissolution. Application of this protocol to the measurement of pyruvate-to-lactate flux in melanoma cells for the assessment of BRAFi treatment demonstrated a significant reduction of k_PL following 24 and 48 hours of treatment in BRAF^V600E cells and no significant effect on BRAF^WT cells.

Abbreviations used:

BRAFi

BRAF-inhibition

DMSO

dimethyl sulfoxide

DNP

dynamic nuclear polarization

EDTA

ethylenediaminetetraacetic acid

ERK

extracellular signal-regulated kinase

HP

hyperpolarized

LDH

lactate dehydrogenase

MAPK

mitogen-activated protein kinase

MCT

monocarboxylate transporter

MEK

mitogen-activated protein kinase kinase

MRS

magnetic resonance spectroscopy

NMR

nuclear magnetic resonance

PBS

phosphate-buffered saline

SNR

signal-to-noise ratio