Phosphatidylcholine (PC) is a major phospholipid species, the principal structural component of cell membrane bilayers accounting for approximately 40% of all eukaryotic cellular membranes (1). The regulation of PC synthesis is altered in a host of disorders including cardiovascular, neurological, and pulmonary diseases (2,3,4). Current methods for studying PC metabolism involve pulsing cells with radiolabeled choline and then separating and quantifying the PC spot by two consecutive steps, a procedure for lipid extraction followed by a subsequent method for PC separation using thin layer chromatography (TLC) or high-performance liquid chromatography (HPLC) (2,5). The whole process is labor-intensive and time-consuming, taking 24 to 48 h. Since the 1950s, three major lipid extraction methods have been introduced, including Folch (6), Bligh-Dyer (7), and Radin (8). All of these methods extract total lipids and some polar components nondiscriminatively, requiring a follow-up procedure such as TLC (2) to further separate PC from other choline-containing lipids. Based on the different solubility properties between PC and substrate choline, we report here a rapid and simple method that allows selective extraction and quantitation of PC from cultured cells in a single step. This method is efficient, quantitative, and less expensive to perform than other commonly used methods.

The goal of this study was to identify a simple system to cleanly separate choline from the metabolic product, PC. Our strategy was to evaluate the differences in physical properties between PC and the substrate choline and then use these differences for solvent system selection. To do this, we first identified the optimal extraction solvent for our system and compared extraction efficiency with published methods. Among the three major lipid extraction methods, Radin's method has several advantages and has been widely used with cultured cells (8,9). In this method, a ratio of 3:2 hexanes:isopropanol is used, which allows the solvent to penetrate cell membrane bilayers and extract a wide range of components, including phospholipids, fatty acids, steroids, and even some proteins (8,9). Choline is a highly polar salt, not soluble in hexanes but soluble in water and isopropanol. In contrast, PC is highly soluble in hexanes. We hypothesized that an appropriate mixture of hexanes and isopropanol could be capable of penetrating the cell membrane bilayers and efficiently extracting PC out of aqueous cell membranes while preventing any choline extraction.

To test extraction efficiencies, Caco-2 cells, obtained from ATCC (Manassas, VA, USA) and cultured in Dulbecco's modified Eagle's medium (DMEM), were pulsed with either 1.5 µCi ¹⁴C-choline or 1.0 µCi ³H-choline at the last 1 h of inoculation. After washing with cold phosphate-buffered saline (PBS), lipids were extracted using several methods: (i) Bligh-Dyer (2), (ii) with hexanes alone, (iii) hexanes:isopropanol at various ratios, or (iv) hexanes:isopropanol at a ratio of 3:2 (Radin's) (8) for 15 min on ice. Radioactive products were quantified by scintillation counting in the presence of ScintiSafe® Econo 2 (Fisher Scientific, San Jose, CA, USA) (2). Both Bligh-Dyer and Radin extraction methods are known to efficiently extract all phospholipids (8,9). It was observed that pure hexanes or mixtures of hexanes:isopropanol in a ratio of 98:2 or 95:5 failed to extract any radioactive components (Figure 1, left column, and data not shown). Hexanes:isopropanol in a ratio of 8:2, 7:3, 3:2, or 9:1 recovered substantial amounts of radioactive components from all cells (Figure 1, middle and right columns, and data not shown).

Figure 1. Extraction of ¹⁴C-phosphatidylcholine (¹⁴C-PC) from Caco-2 cell cultures with or without oleic acid. (Click to enlarge)

To elucidate the specificity of extracting PC with the different methods, we analyzed the extraction products by TLC as described previously (2), followed by phosphorimaging analysis. As shown in Figure 1, TLC imaging confirmed that no radioactive component was obtained by 100% hexanes, while ¹⁴C-PC, ¹⁴C-choline, and an unidentified ¹⁴C-incorporated lipid, most likely lyso-PC, were extracted using Radin's solvent. However, the 9:1 hexanes:isopropanol gave a very clean TLC profile, with only PC appearing to be extracted (¹⁴C-choline and the presumed ¹⁴C-lyso-PC bands could be recovered using Radin's method in the cells pre-extracted with the 9:1 solvent). Based on these results, we termed the 9:1 hexanes:isopropanol mixture the one-step extraction system (OSES) for separating radioactive PC from its choline precursor.

Unsaturated fatty acids have been reported to increase PC synthesis in various cell lines (10). Therefore, we used 500 or 1000 µM oleic acids to increase the amount of PC synthesized as a measure of extraction efficiency under the different synthesis conditions. As shown in Figure 1, the OSES extracted comparable, if not larger amounts of ¹⁴C-PC compared with Radin's method under conditions of increasing PC synthesis.

To confirm extraction efficiency, cells were extracted by OSES and followed with two subsequent extractions; over 88% of the radioactive component was recovered in the first extraction (data not shown). Next, the OSES was compared with the Bligh-Dyer. After pulsing with ³H-choline, lipids were either extracted from cultured cells by OSES and quantified directly; or cells were lysed, total lipids were extracted by Bligh-Dyer and then resolved by TLC, and the ³H-PC spots were subsequently quantified by scintillation counting (2). Results show virtually identical results with the two methods (Figure 2, open bars versus solid bars), demonstrating that the OSES system, while much simpler than the Bligh-Dyer, provides comparable results.

Figure 2. One-step lipid extraction and separation method compared to the traditional thin layer chromatography (TLC) procedure. (Click to enlarge)

Further evidence of the selectivity of OSES lies in the measurement of cellular protein concentrations. Consistent with previous reports (9), an approximately 44% decrease of protein concentrations was observed in samples extracted by Radin's (4.78 ±1.31 µg/mL, P < 0.05) relative to nonextracted control cells (8.51 ± 1.34 µg/mL), while no change in protein concentration by OSES was seen (7.72 ± 1.32 µg/mL). OSES is sensitive to residual buffer or media in the cultured cells, but less sensitive to temperature or time of exposure (data not shown).

In comparison with the TLC and HPLC procedures, OSES has several advantages. OSES extracts and separates PC directly from cultured cells in a single step. This minimization of handling not only saves labor and expense, but also reduces inevitable errors introduced during multiple steps required for other conventional methods, which might lead to larger standard deviations (2).

Another advantage of OSES is timesaving. The whole process of OSES can be finished in 30 min, which is a significantly shorter time compared to the 24 h required for other conventional methods starting from the same cell culture stage. This allows researchers to handle larger numbers of samples in shorter time using fewer resources, making this procedure amenable to automation for high-throughput screening.