Determination of cell volume as part of metabolomics experiments

Abstract

Cells regulate their cell volume, but cell volumes may change in response to metabolic and other perturbations. Many metabolomics experiments use cultured cells to measure changes in metabolites in response to physiological and other experimental perturbations, but the metabolomics workflow by mass spectrometry only determines total metabolite amounts in cell culture extracts. To convert metabolite amount to metabolite concentration requires knowledge of the number and volume of the cells. Measuring only metabolite amount can lead to incorrect or skewed results in cell culture experiments because cell size may change due to experimental conditions independent of change in metabolite concentration. We have developed a novel method to determine cell volume in cell culture experiments using a pair of stable isotopically labeled phenylalanine internal standards incorporated within the normal liquid chromatography-tandem mass spectrometry (LC-MS/MS) metabolomics workflow. This method relies on the flooding-dose technique where the intracellular concentration of a particular compound (in this case phenylalanine) is forced to equal its extracellular concentration. We illustrate the LC-MS/MS technique for two different mammalian cell lines. Although the method is applicable in general for determining cell volume, the major advantage of the method is its seamless incorporation within the normal metabolomics workflow.

INTRODUCTION

Cells regulate their volume by osmotic pressure gradients (1). Imbalances of intracellular and extracellular osmolarity will result in alterations of cell volume. Cells have developed a multitude of regulatory mechanisms to maintain cell volume, as changes in cell volume will modify a variety of cellular functions. Lang et al. (1) have reviewed the functional significance of cell volume regulation and have pointed out the impact that cell volume changes can have on metabolic pathways (e.g., cell swelling may increase glycogen synthesis and inhibit glycolysis) and how changes in metabolite levels (e.g., increased glycogenolysis producing glucose phosphate or increased protein degradation producing amino acids) may affect intracellular osmolarity and, thus, cell volume. For these reasons, cell volume changes should be important to consider in metabolomics.

The end point of metabolomics is to define changes in metabolite concentrations as a function of experimental or physiological perturbation. Most metabolomics experiments involve comparison of the experimental groups versus a control group. This approach is often used in both targeted metabolomics for measurement of specific metabolites and untargeted metabolomics looking for alterations in any metabolite (2). A key requirement of these approaches is normalization of amounts of sample measured in control and experimental groups (3).

Metabolomics methods using mass spectrometry, such as liquid chromatography-mass spectrometry (LC-MS), gas chromatography-mass spectrometry (GC-MS), measure total amounts of metabolites in a sample. To determine concentration, we need to know both the number of cells and their average cell volume that contributed to the sample. Cell number is a straightforward measurement, but cell volume is not as easily evaluated. Volume methods range from observation under a microscope to advanced computational tomography (4). Of these methods, the Coulter counter (5) is one of the most widely used techniques for measuring cell volume and is often regarded as the gold standard in the field. Other techniques range from the use of FACS (6) to the use of chloride ion transport (7). Different techniques provide either absolute or relative values, correlating cell volume to other parameters measured (4). These methods require expensive instrumentation, intact cells, and consume the cells used for volume measurement. Thus, separate sets of cells must be prepared for cell volume and metabolomics measurements, effectively doubling time and complexity of the experiments.

In this paper, we present a method to determine cell volume for metabolomics studies using cell culture simultaneously with the metabolomics measurements that are being made. The method is seamless with the normal metabolomics workflow and does not require separate sets of cells for the cell volume measurement. In this technique, we add a stable isotopically labeled (SIL) metabolite of known concentration to the cell culture media that will then be of equal concentration inside the cells. The technique is based on the “flooding dose” method often used for measuring protein synthesis (8). Protein synthetic rate has long been determined by measuring the rate of isotopically labeled amino incorporation into newly synthesized protein (9). The problem with the method has been accurate assessment of the intracellular enrichment (or specific activity) of the labeled amino acid. Measurement of the labeled amino acid enrichment attached to tRNA has been the best measure of the precursor enrichment but is difficult to determine (10). The concept behind the flooding dose is to raise the extracellular concentration of the labeled amino acid (cell culture media in this case) to “flood” the cells, raising the intracellular concentration to match the extracellular level (11). This approach only works for amino acids that have a low intracellular/extracellular gradient and are rapidly transported in and out of the cells, and amino acids transported by system L have been shown to meet this criterion (9). Initially, isotopically labeled leucine was used as the flooding amino acid, but later phenylalanine was found to have advantages (8). By adding an excess of SIL phenylalanine to the cell growth media, we set the extracellular concentration of phenylalanine and similarly set the intracellular concentration of phenylalanine in the cells. In a typical metabolomics cell culture experiment, cells are harvested, lysed, and the total amounts of metabolites determined. In this case, the total amount of SIL phenylalanine in the cells is also measured. By fixing the concentration of intracellular SIL phenylalanine, measuring the total amount of SIL phenylalanine and cell number, we can determine average cell volume.

MATERIALS AND METHODS

Materials

Phenylalanine (abbreviated as “unlabeled Phe”) was obtained from Sigma Aldrich (St. Louis, MO). Stable isotopically labeled L-[ring-¹³C₆]phenylalanine (abbreviated as “[¹³C₆]Phe”) and [U-²H₈]phenylalanine (abbreviated as “[²H₈]Phe”) were obtained from Cambridge Isotope Laboratories Inc. (Tewksbury, MA). RMPI-1640 cell culture media was obtained from Corning Inc. (Manassas, VA). Pyruvate, folate, glucose, fetal calf serum, and glutamine used in cell culture media were also obtained from Sigma Aldrich (St. Louis, MO), 2-mercaptoethanol was obtained from Thermo Fisher Scientific, Inc. (Waltham, MA), and acetonitrile (OmniSolv LC-MS grade) was from Merck/EMD Millipore Corporation. Granulocyte-macrophage colony-stimulating factor (GM-CSF) was obtained from Peprotech (NJ).

Cell Culture and Activation

Two sets of cell culture experiments were carried out for cell volume measurements. The first set of cells was EL4 murine lymphoblast cells (RRID:MGI:2175632). The EL4 starting culture of ≈5 million cells, stored at −196°C in liquid nitrogen, was gently thawed and then allowed to grow in a T-75 flask in RPMI complete media comprising 50 mM 2-mercaptoethanol, 5% fetal calf serum (FCS), 2 mM glutamine, 1 mM pyruvate, 10 µg/mL folate, and 2.5 mg/mL glucose at 37°C in a CO₂ incubator. Media replenishment was carried out every 24 h for 4 days.

Cell culture samples were generated by replacing the standard RPMI complete media with the RPMI complete supplemented with [¹³C₆]Phe. The starter culture of 30 million EL4 cells was divided into five equal volumes into a six-well plate. [¹³C₆]Phe was supplemented (media concentration of ≈250 µM) at either 24, 4, 2, 1, or 0 h before harvesting the cell culture samples. All sets of cells were harvested at the same time. Because all subcultures had the same incubation period before harvesting, the conditions for each sample were the same for all samples, including cell number. Cells were washed with phosphate-buffered saline (PBS) twice, excess PBS was aspirated, and cell pellets were immediately frozen. Aliquots of cell media were also stored along with the cells at −70°C before lysis and metabolite extraction.

Additional sets of EL4 cells were cultured using the same protocol with increasing concentrations of [¹³C₆]Phe supplemented in the RPMI complete media. The second set of EL4 cells was cultured in triplicate with equal number of cells and harvested 2 h after supplementing with [¹³C₆]Phe in the media at 530 µM. The third set of EL4 cells was cultured in duplicate and harvested 2 and 4 h after adding [¹³C₆]Phe to the media (1,050 µM).

The second different type of cells were also measured. Bone marrow-derived dendritic cells (BMDCs) were generated as described in Lutz et al. (12). Briefly, bone marrow hematopoietic cells were differentiated in GM-CSF (20 ng/µL; Shenandoah Biotechnology, Inc.) in complete dendritic cell medium, comprising RPMI-1640, 10% FCS, 2 mM l-glutamine, 1-IU/mL Pen-Strep, and 1 mM β-mercaptoethanol, for 7 days, with media replenished at days 2 and 4 of culture. Eight sets of cells (7 × 10⁶ cells/experiment) were grown. Four sets of cells were supplemented with [¹³C₆]Phe, for 1 h and four sets of cells were supplemented for 2 h and were then harvested after washing three times in PBS. Excess PBS was aspirated, and cell pellets were frozen before analysis. Corresponding media samples were also frozen along with the cells.

Cell Lysis and Metabolite Extraction

A metabolite extraction protocol (13) was followed for cell lysis and media metabolite extraction, after adding an aliquot of the [²H₈]Phe internal standard (IS) to the tube containing the harvested cell pellets. Samples were kept on ice and extracted in cold (−80°C) lysis buffer (methanol:acetonitrile:water 4:4:2 vol/vol/vol). For media samples, an aliquot of the [²H₈]Phe internal standard was added to 100-µL aliquots of media kept on ice. The samples were mixed and then extracted in cold lysis buffer. Samples were then sonicated for 10 min with no heat. After sonicating, the centrifuge tubes were kept at 4°C for 20 min undisturbed to maximize protein precipitation. The tubes were then centrifuged at 15,000 g for 15 min at 4°C, and the supernatant containing the metabolites was collected into microcentrifuge tube and dried down in a speed-vac. The samples were then brought back up in 30 µL of 20:80 (water:acetonitrile) and transferred to LC vials for analysis by liquid chromatography-tandem mass spectrometry (LC-MS/MS).

ESI-LC-MS/MS

A high-resolution Waters Xevo G2-XS QTOF mass spectrometer and Waters Acquity M-Class UPLC LC were used in the positive ion mode to acquire data with electrospray ionization (ESI) source. Phenylalanine was quantified using parallel reaction monitoring (PRM), where each isotopologue’s parent ion was scanned in the MS channel and further fragmented to obtain corresponding MS/MS fragments, which were measured for quantification. The precursor and product ions selected were m/z = 166.0 and 120.081 for unlabeled Phe (abbreviated as M0-Phe), m/z =172.0 and 126.102 for [¹³C₆]Phe (abbreviated as M6-Phe), and m/z = 174.0 and 128.132 for [²H₈]Phe (abbreviated as M8-Phe). These scans generate time versus intensity traces for M0-Phe, M6-Phe, and M8-Phe. Although all three isotopologues elute at the same time, they produce uniquely different traces due to their differences in mass (as noted by the different precursor/product ion selections). The areas under the elution traces were then integrated and used to quantify the amounts of M0-Phe, M6-Phe, and M8-Phe.

Following LC conditions were used for running the samples. Two microliters of each sample was injected onto a Waters Acquity UPLC BEH Amide column (1 mm × 150 mm, 1.7 µm particles) with the system flow set at 50 µL/min. Two mobile phases used for the LC were A: water with 10 mM of ammonium acetate and B: acetonitrile:water (9:1 vol:vol in 10 mM of ammonium acetate). The LC gradient was 99% B for 3 min that was then decreased to 88% B by 5 min and then to 81% B by 12 min.

The Waters TargetLynx software suite was used to integrate chromatographic-PRM peak areas with a mass detection window of ±5 ppm.

Standard Curves

M0-, M6-, and M8-Phe were weighed and subsequently diluted to make stock solutions. These stock solutions were then used to make standard curve dilutions. Serial dilutions of unlabeled Phe and [¹³C₆]Phe versus [²H₈]Phe were analyzed using LC-MS/MS.

Calculations

The PRM peak areas for M0-, M6-, and M8-Phe were integrated. These areas correspond to the amount of each Phe isotopologue in each sample. The linear relationship was established by injecting and measuring the peak areas of standards of Phe for M0-, M6-, and M8-Phe.

$$I_i={\alpha }_i{n}_i$$ (1)

where I is the area of the peak, α is the ionization constant, and n is the number of moles of each species i (M0-, M6-, or M8-Phe). Peak area ratio measurements were formed for M0-Phe and M6-Phe against the M8-Phe internal standard:

$$\frac{I_i}{I_8}={(\frac{I_i}{I_8})}_8+k_i(\frac{n_i}{n_8})$$ (2)

where I_i/I₈ is the area ratio of either M0- or M6-Phe versus M8-Phe plotted against the mole ratio of n_i/n₈, where n_i is the amount (in moles) of either M0- or M6-Phe and n₈ is the amount of M8-Phe in moles in the standard sample. The slope of the line is k_i, and the intercept, ${(\frac{I_i}{I_8})}_8$, reflects any residual M0- or M6-Phe signal present in the M8-Phe.

The amount of either unlabeled Phe or [¹³C₆]Phe is then determined in both cell media and cell extracts by knowing the amount (n₈) of [²H₈]Phe added.

$$n_i=\frac{n_8}{k_i}\left[(\frac{I_i}{I_8})-{(\frac{I_i}{I_8})}_8\right]$$ (3)

Note: for the calculation of cell volume, measurement of unlabeled Phe is not needed. For completeness, we are including these measurements to demonstrate and help validate the method.

We assume that after a “flooding dose” of [¹³C₆]Phe to the media, the concentration of [¹³C₆]Phe in the cells will rise equal to the media concentration of [¹³C₆]Phe. Assuming that the intracellular concentration of [¹³C₆]Phe is that of the media (C₆), and measuring the total amount of [¹³C₆]Phe in the cells (n₆), we can determine total cell volume (V_c):

$$V_{tl}=\frac{n_{c6}}{C_6}.$$ (4)

The average cell volume of individual cells is the total cell volume divided by the total number of cells (N):

$$V_{avg}=\frac{V_{tl}}{N}.$$ (5)

The concentration of both unlabeled Phe (C₀) and [¹³C₆]Phe (C₆) in the media are known from the amount of unlabeled Phe in the initial media and from the amount of [¹³C₆]Phe added. For completeness, we also measured these media concentrations by taking an aliquot of media (V), adding a known amount of [²H₈]Phe, and measuring the amount of M0-Phe (n₀) and M6-Phe (n₆) using Eq. 3 earlier. The media concentrations were then C₀ = n₀/V and C₆ = n₆/V for unlabeled Phe and [¹³C₆]Phe, respectively.

RESULTS

Phenylalanine Detection and Standard Curve

Using PRM, we obtained very precise ratio measurements of phenylalanine and its isotopologues by LC-MS/MS for standards, cell culture media, and cell extracts. Figure 1 shows the PRM chromatograms for the three isotopologues of phenylalanine in standards and in EL4 cells. Phenylalanine elutes around 8.8 min. There was effectively no noise or interference in any of the phenylalanine peaks, either for standards or cell extracts. The peak areas were then integrated and used in the calculations. For quantification, both M0-Phe and M6-Phe were measured against M8-Phe as the internal standard to generate standard curves.

Figure 1.

PRM-extracted ion chromatograms of phenylalanine isotopologues. A, B, and C are the standards; D, E, and F are the cell extract samples. A and D are PRM chromatograms of unlabeled Phe (M0-Phe; m/z, 166 → 120.081), B and E are PRM chromatograms of [¹³C₆]Phe (M6-Phe; m/z, 172 → 126.102), and C and F are PRM chromatograms of [²H₈]Phe (M8-Phe; m/z, 174 → 128.132). [¹³C₆]Phe, L-[ring-¹³C₆]phenylalanine; [²H₈]Phe, [U-²H₈]phenylalanine; PRM, parallel reaction monitoring.

Figure 2 shows the standard curves plotted as peak area ratio versus the mole ratio of M0/M8 and M6/M8-Phe. Standard curves were used for calculating slope (k_i), intercept (I_i/I₈)₈, and the limits of detection. The standard curves were then used to calculate the amount of phenylalanine in cells and concentration of phenylalanine in media samples.

Figure 2.

Standard curves for unlabeled Phe (M0-Phe) and [¹³C₆]Phe (M6-Phe) measured against [²H₈]Phe (M8-Phe) as the internal standard. Data are plotted as the PRM integrated area ratio (I₀/I₈ and I₆/I₈) vs. mole ratio (n₀/n₈ and n₆/n₈), generating linear curves. The fitted lines are I₀/I₈ = (0.045 ± 0.032) + (0.758 ± 0.013) (n₀/n₈) for unlabeled vs. [²H₈]Phe and I₆/I₈ = (0.048 ± 0.039) + (1.084 ± 0.027) (n₆/n₈) for [¹³C₆]Phe vs. [²H₈]Phe; values are means ±SE. From these standard curves, the limit of detection was calculated to be 46 pmol for measurement of unlabeled Phe and 31 pmol for [¹³C₆]Phe. [¹³C₆]Phe, L-[ring-¹³C₆]phenylalanine; [²H₈]Phe, [U-²H₈]phenylalanine; PRM, parallel reaction monitoring.

Stability of Phenylalanine Concentration versus Incubation Time

Figure 3A shows the concentration of unlabeled Phe and [¹³C₆]Phe measured in the cell culture media of EL4 cells over a 24-h period. The amount of [¹³C₆]Phe added was set to be 250 µM in the media, ∼3 times greater than the unlabeled Phe contained in the RPMI complete media (≈80 µM). The concentration of unlabeled Phe remained constant throughout the 24-h experiment. The [¹³C₆]Phe starts at a zero value at 0-h time point, when no [¹³C₆]Phe has been added to the media and rises to ≈250 µM by the 1st 1-h time point. [¹³C₆]Phe concentration remains constant for 24 h. A minor decrease in the concentration of [¹³C₆]Phe was observed in 24-h media sample (to 233 µM).

Figure 3.

Measurement of phenylalanine for EL4 cells. A: concentration of unlabeled Phe (M0-Phe) and [¹³C₆]Phe (M6-Phe) in the cell culture media. B: amount (pmol/10⁶ cells) of unlabeled Phe (M0-Phe) and [¹³C₆]Phe (M6-Phe) measured in EL4 cell extracts. [¹³C₆]Phe, L-[ring-¹³C₆]phenylalanine; [²H₈]Phe, [U-²H₈]phenylalanine.

We expected a lag in time after the [¹³C₆]Phe was added to the cell culture media before the intracellular [¹³C₆]Phe concentration reached an equilibrium equal to the concentration of [¹³C₆]Phe in the media. Figure 3B shows the amount of unlabeled and phenylalanine isotope measured in the cells over the 24-h experiment. The amount of unlabeled Phe measured in the extracted EL4 cells from 0 to 24 h of incubation remained constant. The [¹³C₆]Phe amount was also constant over the 24 h. These results show that the [¹³C₆]Phe quickly enters the EL4 cells by/before 1st hour of incubation and remains constant.

Cell Volume of EL4 Cells

To derive the average volume of a cell from the amount of phenylalanine (obtained in moles) in the cell extract, we made use of the concentration of phenylalanine isotopes in the media and the cell number for each time point. The number of cells for each time point was 6 × 10⁶ using a hemocytometer. Average cell volume was calculated from the total measured volume divided by the number of cells. These values for the EL4 cell are shown Table 1. The mean cell volume value calculated was 936 ± 58 µm³ (means ± SD) with a 6% coefficient of variation (CV). These data show that incubations as short as 1 h are acceptable for determining cell volume.

Table 1.

Cell Volume Measurements for EL4 Cells (Set 1)

Incubation Time, h	Total Cell Volume per Cell Extract, µL	Volume per Cell, µm3
1	5.45	908
2	5.63	938
4	5.30	882
24	6.10	1,017
Means ± SD	5.62 ± 0.35	936 ± 58
CV	6%	6%

Concentration of [¹³C₆]Phe in media = 250 µM. CV, coefficient of variation.

The next question to address is whether cell volume measurement is dependent on amount of [¹³C₆]Phe added. Such a dependency would indicate problem with the flooding dose approach. A second set of EL4 cells were grown and incubated in triplicate with a [¹³C₆]Phe media concentration of 530 µM and harvested after 2 h, and a third set of EL4 cells were incubated with a [¹³C₆]Phe media concentration of 1,050 µM and harvested in duplicate at 2 h and 4 h. The results of these measurements are shown in Table 2. Similar values of EL4 cell volumes were determined across the three experiments and different sets of cells. The average cell volume across the three experiments was 1,100 ± 160 µm³ (means ±SD). There was no correlation of [¹³C₆]Phe amount added to the media and measured cell volume.

Table 2.

Cell Volume Measurements for EL4 Cells (Sets 2 and 3)

Incubation Time,	Volume per Cell, µm3
h	530 µM*	1,050 µM*
2	1,242	1,311
	1,274	886
	1,261
4		1,139
		1,032
Means ± SD	1,259 ± 16	1,092 ± 179

*Concentration of [¹³C₆]Phe in media.

Cell Volume of Bone Marrow-Derived Dendritic Cells

The cell volumes of a second type of cells were determined: bone marrow-derived dendritic cells (BMDCs). Eight sets of BMDCs (7 × 10⁶ cells/experiment) were grown. Four were incubated for 1 h and the other four were incubated for 2 h with [¹³C₆]Phe. Figure 4A shows the measured concentration of unlabeled Phe and [¹³C₆]Phe in the cell culture media. Figure 4B shows the amount of phenylalanine in the cell extracts at 1 and 2 h of incubation. BMDCs show a constant amount of [¹³C₆]Phe at both 1 h and 2 h. The average cell volume measured was 1,198 ± 66 µm³ (means ± SE) and 1,190 ± 98 µm³ at 1 h and 2 h, respectively.

Figure 4.

Measurement of phenylalanine for BMDCs. Two time points were obtained at 1 and 2 h of incubation (four sets of cells each). A: concentration of phenylalanine in the cell culture media. B: amount (pmol/10⁶ cells) of phenylalanine in the BMDC cell extracts. Gray bars correspond to measurement of unlabeled Phe (M0-Phe), and white bars correspond to measurement of [¹³C₆]Phe (M6-Phe). Bars are means ±SE. Individual data points (sets of measured cells) are shown. BMDC, bone marrow-derived dendritic cell; [¹³C₆]Phe, L-[ring-¹³C₆]phenylalanine; [²H₈]Phe, [U-²H₈]phenylalanine.

DISCUSSION

Cell volume measurement should be a crucial aspect of cell culture studies in metabolomics. First, cell volume changes are known to affect cellular metabolic pathways (1) and, therefore, should be considered. Second, metabolomics methods measure total metabolite amount in a group of cells, not metabolite concentrations. Because cell volume will differ for different cell lines and may change under different experimental treatments for a single cell line, measurements of total metabolite amount can distort or provide false conclusions unless normalized by cell volume. In this paper, we establish a protocol for measuring cell volume for cell culture samples used in metabolomics studies. This single-shot method to measure cell volume is seamlessly integrated within the normal metabolomics workflow. There is no need to separate additional set of cells to be grown specific for cell volume measurement and independent from the metabolomics measurements. The only change to the metabolomics workflow is addition of [¹³C₆]Phe to the cell media and addition of [²H₈]Phe to the cell extract. The instrumentation required for this method is the same LC-MS/MS used for most metabolomics experiments (14). The technique can be used for any type of cell culture that allows for membrane transport of metabolites from the extracellular matrix into the cells. Cells are lysed and, therefore, do not need to be intact for measuring cell volume. Furthermore, results do not depend on whether the cells are adherent or nonadherent.

A high-resolution mass spectrometer allows for accurate measurements by PRM, minimizing interference from compounds with similar molecular masses. We selected SIL [¹³C₆]Phe for measuring cell volume and [²H₈]Phe as the internal standard for normalization. An important advantage of using the [¹³C₆]Phe isotopologue is that its molecular weight (171.1 Da) is significantly higher than unlabeled Phe (165.1 Da), ensuring no natural abundance stable isotope interference from the unlabeled Phe already present in the media. Furthermore, the [²H₈]Phe (173.1 Da) should not interfere with [¹³C₆]Phe measurement. Although the [²H₈]Phe IS we used was highly labeled (98% ²H), there was a minor amount of ²H₆ in the [²H₈]Phe, a contribution of ≈2% that becomes a constant in the intercept for Eq. 2. As alternatives, more highly labeled, commercially available [¹³C₉]Phe or [¹³C₉,¹⁵N]Phe internal standards could be used instead.

We also measured the concentration of unlabeled Phe in our experiments along with the [¹³C₆]Phe. These measurements were done for completeness but are unnecessary for applying the method. Only the [¹³C₆]Phe and [²H₈]Phe need be measured. The RPMI complete media has an expected concentration of 80 µM unlabeled Phe, and we fixed the concentration of [¹³C₆]Phe three times higher at ≈250 µM. In addition, in the subsequent sets of EL4 cell cultures, we fixed the concentration of [¹³C₆]Phe at 530 µM and 1,050 µM and measured and confirmed the [¹³C₆]Phe in the media to be ≈510 µM and ≈1,050 µM, respectively. However, with proper quantitative addition of the [¹³C₆]Phe to the media, there is no need to confirm by measurement the [¹³C₆]Phe concentration in the media, eliminating another measurement step for the method. The only measurements that need to occur are the amounts of [¹³C₆]Phe and [²H₈]Phe in the cell extracts.

To measure the volume of cells, there is a minimum threshold of the number of cells needed. This number is dependent on the limit of detection (LOD) for phenylalanine and the concentration of SIL phenylalanine added to cell culture media. For the EL4 (first set) and BMDC cells, we used 6 × 10⁶ and 7 × 10⁶ cells per sample, respectively. The amount of phenylalanine in each set of cell extracts was ≈1,400 pmol and was at 45 times the LOD for [¹³C₆]Phe measurement (≈31 pmol). Thus, if we worked at 10 times the LOD, we would only need ≈1 × 10⁶ cells. If the [¹³C₆]Phe concentration in the media were doubled, we would only need 5 × 10⁵ cells to measure cell volume. Similar calculation can be made for cells with greater or lesser volumes than the cells we measured here.

The method as tested requires that the cells being measured do not metabolize phenylalanine. This condition is readily met because most cells do not express the phenylalanine hydroxylase enzyme (15, 16). However, the presence of the phenylalanine hydroxylase enzyme is readily determined by looking for the presence of [¹³C₆]tyrosine formed in the cell extracts or media from [¹³C₆]Phe. This measurement is also part of the standard metabolomics workflow. For example, we sought and found no [¹³C₆]tyrosine in either the EL4 or BMDCs, confirming no metabolism of phenylalanine by the cells.

The flooding dose method assumes that the concentration of intracellular [¹³C₆]Phe is equal to the extracellular concentration. A crucial aspect of this technique involves adding enough [¹³C₆]Phe to force [¹³C₆]Phe into equilibration between cells and media. We initially used an extracellular [¹³C₆]Phe amount three times (≈250 µM) more than unlabeled Phe in the media (80 µM). Transfer of [¹³C₆]Phe from the media into the cells occurred rapidly by/before 1 h of incubation. Cell volume measurements were consistent over the time points for EL4 cells and BMDCs. Even with the flooding dose, there is presumably a small gradient between intra- and extracellular [¹³C₆]Phe concentrations. As long as the gradient is small, the gradient will only add a small consistent bias to the cell volume measurements. However, to test for an effect of the flooding dose amount on the result, we conducted two additional cell volume measurements in EL4 cells, doubling (to 530 µM) and quadrupling (to 1,050 µM) the [¹³C₆]Phe concentration in the media and measured similar cell volumes for all three experiments. There was no apparent effect of amount of flooding dose upon cell volume measurement using phenylalanine.

The diameters of EL4 cells have been previously measured by two-photon microscopy (17). The reported diameters for two different fluorescently labeled sets of EL4 cells were 11.0 ± 0.3 µm and 12.7 ± 0.2 µm, which assuming a spherical cell correspond to volumes of 750 ± 60 µm³ and 1,120 ± 50 µm³. Our measurements of EL4 cell volumes (1,100 ± 160 µm³) fall within reported range of EL4 cell volumes (17). Our measurements of BMDC volumes (769 ± 50 µm³) also fall within the range (690–1,400 µm³) reported in the literature (18). These comparisons give us some confidence that we are measuring the correct volumes.

The technique presented here establishes a method to measure cell volume and convert metabolomics measurements of metabolite amounts into concentrations. The method uses two SIL phenylalanine tracers that are measured using the same LC-MS/MS instrumentation used in most metabolomics workflows. Thus, the method is readily applied to metabolomics experiments using cell culture. The method is also useful and can be applied in general to measure cell volumes in cell culture.

GRANTS

This work was supported by NIH grants P30 GM118228 and S10 OD018126.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

D.E.M. conceived and designed research; K.S.S., E.A., and R.C.B. performed experiments; K.S.S. and D.E.M. analyzed data; K.S.S. and D.E.M. interpreted results of experiments; K.S.S. and D.E.M. prepared figures; K.S.S. drafted manuscript; K.S.S., E.A., R.C.B., and D.E.M. edited and revised manuscript; K.S.S., E.A., R.C.B., and D.E.M. approved final version of manuscript.