Background

Lipids are the major constituents of cell membranes that delineate biological compartments, and also play an important role in signaling pathways and energy storage (Baenke et al., 2013). Lipid metabolism is dysregulated in a variety of diseases, and recent studies have suggested that disrupting lipid biosynthesis may be an approach to inhibit tumor growth (Svensson and Shaw, 2016). Consequently, there is interest in studying the metabolic pathways that produce cellular lipid species. Culturing cells in media deprived of specific metabolic components can help provide a better understanding of how metabolic pathways support cell function. Standard culture media for most mammalian cells includes animal serum as a source of protein and growth factors (most commonly fetal bovine serum [FBS]). Animal serum contains a variety of lipid species within lipoprotein complexes and bound to serum albumin. The composition of animal serum is not identical across lots, and the nutrients levels in serum are not well defined in most experiments. Cells are able to obtain lipids from both de novo synthesis and from exogenous sources (Kamphorst et al., 2013; Hosios et al., 2016; Balaban et al., 2017), so the presence of serum lipids can complicate studies of de novo lipid metabolism. For example, although inhibitors of fatty acid synthase can inhibit proliferation, this effect is strengthened in lipid-depleted serum (Svensson et al., 2016). Microenvironments within the body likely vary in their lipid composition (Tourtellotte, 1959; Nanjee et al., 2000), suggesting that there are physiological cases where cells may be deprived of lipids.

In addition to serving as a source of lipids, serum also provides growth factors and other components that support the cell proliferation, and most cells cannot be studied for long periods of time in the absence of serum. Serum-free media formulations and lipoprotein-depleted sera are commercially available, but these are often expensive, are not optimized to support the growth of all cell lines, and lack a corresponding lipid-replete serum to serve as a control. To overcome these challenges, we have employed a bi-phasic extraction to remove lipids from FBS without denaturing serum proteins (Cham and Knowles, 1976). This method has been previously described on LipidomicNet, and we modified this approach to generate lipid-depleted serum and a corresponding lipid-replete serum from the same original FBS lot (Hosios et al., 2016). In our hands, some cells can be cultured in this serum for several passages without a substantial change in their proliferation rate, while other cells are more sensitive to the absence of lipids. We have also demonstrated increased de novo lipid synthesis in cells cultured with this lipid-depleted serum, indicating the expected metabolic response to this condition. This protocol provides an efficient way to remove lipids from a range of volumes of FBS, and generates both lipid-depleted serum and dialyzed, lipid-replete control serum for use in cell culture experiments.