Purification, cellular levels, and functional domains of lipase maturation factor 1

Abstract

Over a third of the US adult population has hypertriglyceridemia, resulting in an increased risk of atherosclerosis, pancreatitis, and metabolic syndrome. Lipoprotein lipase (LPL), a dimeric enzyme, is the main lipase responsible for TG clearance from the blood after food intake. LPL requires an endoplasmic reticulum (ER)-resident, transmembrane protein known as lipase maturation factor 1 (LMF1) for secretion and enzymatic activity. LMF1 is believed to act as a client specific chaperone for dimeric lipases, but the precise mechanism by which LMF1 functions is not understood. Here, we examine which domains of LMF1 contribute to dimeric lipase maturation by assessing the function of truncation variants. N-terminal truncations of LMF1 show that all the domains are necessary for LPL maturation. Fluorescence microscopy and protease protection assays confirmed that these variants were properly oriented in the ER. We measured cellular levels of LMF1 and found that it is expressed at low levels and each molecule of LMF1 promotes the maturation of 50 or more molecules of LPL. Thus we provide evidence for the critical role of the N-terminus of LMF1 for the maturation of LPL and relevant ratio of chaperone to substrate.

Keywords: Cellular protein levels; Chaperone; Endoplasmic reticulum; Lipase maturation factor 1; Lipid metabolism; Lipoprotein lipase.

Figure 1

The N-terminal domain of LMF1 contributes to LPL maturation. A. A schematic of LMF1’s topology. B. Western blots against the C-terminal His tag show that FL, TM3 and TM5 truncations are associated with the pellet fraction. Loading controls include GAPDH for soluble proteins in the lysate fraction and PR for membrane proteins. C. Western blots of the media fraction show that LPL-V5 is secreted in cld/cld cells co-expressing the FL, but not the TM3 or TM5, LMF1 constructs. Arrows indicate expression of LPL-V5. Although we loaded 1/5 as much of the highly expressed TM5 construct in all panels to allow detection of other constructs, saturated pixels were unavoidable in panel 5.

Figure 2

LMF1 truncation variants localize to the ER. Immunocytochemistry of the LMF1 variants (green) was compared to the ER marker mCh-KDEL (red) in COS-7 cells. All LMF1 constructs show perinuclear staining characteristic of the ER as is confirmed by co-localization with mCh-KDEL. Scale bars represent 20 μm.

Figure 3

Topology of LMF1 truncations. A. CD3δ-YFP and CFP-CD3δ were used as positive and negative controls for trypsin cleavage, respectively. The middle and right panels show the Western blots for expression and the PPA. Lane 1 was mock (M) transfected, lane 2 shows untreated cells (UT), lane 3 has digitonin (D) addition for 3 minutes, and lane 4 has trypsin (T) for 2 minutes (after a 1 minute incubation with digitonin). Arrows indicate the FL version of each construct. B. Expression and PPA for WT LMF1 and the two truncation constructs. Lanes are labeled as in A.

Figure 4

LMF1 purification and cellular levels. A. Gel filtration trace of purified LMF1 with a coomassie stained gel. B. Western blot showing that α-LMF1 antibody can detect as little as 17.5 fmole of purified LMF1. Cells are mock transfected (M) or transfected with a plasmid expressing LMF1. PR is a loading control for the pellet fraction. Arrows indicate both PR isoforms. C. The diagram at the top indicates the oligos (D, E, F) used to test for the insertion of the murine endogenous retrovirus into intron 7 of LMF1. PCR products, below, show that cld/wt and cld/cld cells have the expected genotype. D. Western blot for quantification of LPL-V5 released from cld/wt cells. S1 and S2 indicate two replicate media collections.