Intramyocellular lipids versus intramyocellular triglycerides

Abstract

Ren et al. recently reported an elegant study of measuring lipid contents of human muscle in vivo using proton magnetic resonance spectroscopy (MRS) techniques (1). The intramyocellular (IMCL) and extramyocellular lipid proton resonances appear as two almost completely resolved peaks for both the methylene and methyl groups of the acyl chains. The methyl resonance peak is used for the quantification of IMCL, in absolute terms. This allows direct comparisons with the results in the literature that are mostly in the absolute terms. The technical advancements shall facilitate the noninvasive investigation of muscle lipids that has been very active for nearly two decades since the initial application by Schick et al. (2). In the report, the authors equated IMCL to IMCL triglycerides (imcTG). This is worth discussion. As MRS quantifies lipids based on the amounts of methylene or methyl protons, the results should reflect TG and phospholipids as well as other fatty acid-containing lipids (3). Cytosolic lipids also include diglycerides, acylcarnitines, and nonesterified fatty acids at substantial levels, especially the latter, with distinct functions (4,5). Lactate methyl resonance peak also overlaps with the lipid methylene resonance (6,7). To measure TG, it is sometimes assumed that cellular phospholipids produce broad peaks that are lost in the background (8). However, no specific tests seem to have been done in this regard to erase the concern and ensure that the resonances are largely free of interference from non-TG lipids. In fact, in most MRS studies this concern is not explicitly addressed. Therefore, it is uncertain whether these resonances can be attributed to imcTG. For this reason, the term IMCL seems more appropriate. However, this term is confused with imcTG in that IMCL is usually studied either as an energy source during exercise (9,10) or as a factor related to insulin resistance (11,12), to which only is imcTG relevant. After more than a decade of such “acclimation,” IMCL seems to have been construed to be almost synonymous with imcTG. Cellular phospholipids are often more abundant than imcTG, a small lipid pool as local energy source (13). In addition to plasma membrane phospholipids, some intracellular phospholipids are in close proximity to or in direct contact with imcTG droplets such as those of mitochondrial membrane as shown by electron microscope (14). Therefore, it seems logical to speculate that MRS may over-estimate imcTG by including non-TG lipids. For instance, IMCL concentration has been reported to be as high as 1.76–2.2% ± 1.1% (IMCL protons as % of water protons) for soleus muscle of obese and healthy lean humans, respectively (15,16). This translates into a TG concentration of 29.3 mmol/kg wet weight after taking into account the differences in the number of protons (62 vs. 2, Ref. 17) and molecular weight of TG (assuming 750) and water molecules, and assuming a muscle water content of 76% (Ref. 18). This is nearly four times higher than the imcTG concentration measured biochemically at 2.1–2.7 mmol/kg wet weight in vastus lateralis of normal and healthy lean humans, (19,20), even after taking into consideration the difference in imcTG concentration between soleus and vastus lateralis (assuming 3:1). There are higher values of imcTG measured biochemically but they are likely contaminated by adipocytes (21). For this reason, we place more confidence on the lower values achieved by using careful dissection techniques (22). Nonetheless, IMCL of soleus and tibialis anterior reported by Ren et al. (1,23) are in the expected range. However, IMCL reported for quadriceps or gastrocnemius are often overestimates (23,24). Interestingly, Howald et al. reported underestimation of tibialis anterior lipids by MRS (14). Although their MRS result is in the expected range for this less oxidative muscle, the results from biochemical and electron microscope (EM) measurements (6.5 and 5.9 mmol/kg ww, respectively) for the same muscle are unrealistically as high as that for oxidative soleus muscle. Therefore, an underestimation by MRS could not be established. It is not clear what caused these overestimations. In particular, it is not known whether non-TG lipids contributed to the methylene and methyl resonances based on which IMCL is quantified. After nearly two decades of active use of the MRS technique, this fundamental question remains unclear. As MRS increasingly becomes a major research tool in the investigation of tissue lipid metabolism, clarifying this issue is critical to better understand the quantitative relation-ship between IMCL and imcTG so that informed comparisons can be made between them. The significance and usefulness of MRS technology to the studies of tissue lipids as related to insulin resistance and sports medicine will depend on the extent that it can reliably probe the relevant lipid pools.