Mutations in lipoprotein lipase that block binding to the endothelial cell transporter GPIHBP1

Abstract

GPIHBP1, a glycosylphosphatidylinositol-anchored protein of capillary endothelial cells, shuttles lipoprotein lipase (LPL) from subendothelial spaces to the capillary lumen. An absence of GPIHBP1 prevents the entry of LPL into capillaries, blocking LPL-mediated triglyceride hydrolysis and leading to markedly elevated triglyceride levels in the plasma (i.e., chylomicronemia). Earlier studies have established that chylomicronemia can be caused by LPL mutations that interfere with catalytic activity. We hypothesized that some cases of chylomicronemia might be caused by LPL mutations that interfere with LPL's ability to bind to GPIHBP1. Any such mutation would provide insights into LPL sequences required for GPIHBP1 binding. Here, we report that two LPL missense mutations initially identified in patients with chylomicronemia, C418Y and E421K, abolish LPL's ability to bind to GPIHBP1 without interfering with LPL catalytic activity or binding to heparin. Both mutations abolish LPL transport across endothelial cells by GPIHBP1. These findings suggest that sequences downstream from LPL's principal heparin-binding domain (amino acids 403-407) are important for GPIHBP1 binding. In support of this idea, a chicken LPL (cLPL)-specific monoclonal antibody, xCAL 1-11 (epitope, cLPL amino acids 416-435), blocks cLPL binding to GPIHBP1 but not to heparin. Also, changing cLPL residues 421 to 425, 426 to 430, and 431 to 435 to alanine blocks cLPL binding to GPIHBP1 without inhibiting catalytic activity. Together, these data define a mechanism by which LPL mutations could elicit disease and provide insights into LPL sequences required for binding to GPIHBP1.

Fig. 1.

Effect of LPL missense mutations on enzymatic activity and binding to heparin. (A) Specific activity of WT LPL, LPL-C418Y, and LPL-E421K. In control experiments, I194T and S132G mutations inhibited LPL catalytic activity, as previously reported (20, 21). (B) Elution profiles of WT LPL, LPL-C418Y, and LPL-E421K from a heparin–Sepharose column with a linear NaCl gradient. The concentration of LPL in each fraction was determined with an ELISA (15). Monoclonal antibody 5F9 was used as the capture antibody; LPL binding was detected with biotinylated antibody 5D2.

Fig. 2.

Western blot assessing binding of WT human LPL, LPL-C418Y, LPL-C438A, and LPL-E421K to GPIHBP1-transfected CHO-K1 cells in the presence or absence of heparin (500 U/mL). After a 2-h incubation, cells were washed extensively, and cell extracts were prepared for Western blotting with GPIHBP1- and LPL-specific antibodies. β-Actin levels were measured as a loading control. Top: Amount of LPL present in the conditioned medium added to the GPIHBP1-transfected CHO-K1 cells.

Fig. 3.

Immunofluorescence microscopy assay of LPL binding to GPIHBP1. CHO-K1 cells transfected with S-protein–tagged human GPIHBP1 were mixed with CHO-K1 cells transfected with WT human LPL, LPL-C418Y, or LPL-E421K and then plated on a coverslip. After a 2-h incubation at 37 °C, permeabilized (A) and nonpermeabilized (B) cells were stained for GPIHBP1 with an antibody against the S-protein tag (red) and LPL with antibody 5D2 (green). Cell nuclei were stained with DAPI (blue). WT LPL secreted by LPL-transfected cells was captured by neighboring GPIHBP1-expressing cells; thus, LPL and GPIHBP1 colocalize in the merged image (arrows). LPL-C418Y and LPL-E421K did not bind to GPIHBP1; hence, no colocalization was observed.

Fig. 4.

Western blot assessing binding of WT and mutant human LPLs (C418Y, C418S, C418P, C438A, C438Y, E421K, and C418Y/E421K) to GPIHBP1-transfected CHO-K1 cells in the presence or absence of heparin (500 U/mL). Quantification of signals with an Odyssey scanner showed that the binding of LPL-C438A and LPL-C438Y were reduced by 70% to 80% compared with WT LPL (n = 3 independent experiments).

Fig. 5.

WT human LPL, but not LPL-C418Y or LPL-E421K, is transported across GPIHBP1-expressing endothelial cells. (A) Ability of rat heart endothelial cells expressing S-protein–tagged versions of GPIHBP1 or CD59 to transport LPL from the basolateral to the apical surface of cells. Equal amounts of WT LPL, LPL-C418Y, or LPL-E421K were added to the basolateral side (lower side) of endothelial cell monolayers and incubated for 1 h at 37 °C. Transported LPL was released from the apical surface (upper side) with heparin (100 U/mL) and dot-blotted onto nitrocellulose. LPL was detected with an antibody against V5, and the dot blots were scanned and quantified with an Odyssey infrared scanner. (B) Transport of LPL from the basolateral to the apical surface of cells, as judged by immunofluorescence microscopy. LPL was added to the basolateral side of endothelial cell that had been transfected with S-protein–tagged GPIHBP1 and grown as confluent monolayers on filters. After a 1-h incubation at 37 °C, cells were fixed and stained with antibodies against the V5 tag (green) and the S-protein (red). Apical and basolateral membranes could be visualized above and below nuclei, which were stained with Draq5 (blue). When WT LPL was added to the basolateral medium, it was transported across cells and could be detected, along with GPIHBP1, on the apical surface of cells. When LPL-C418Y or LPL-E421K was added to the basolateral medium, no transport was observed. No transport of W LPL was observed at 4 °C, or when the endothelial cells expressed CD59 rather than GPIHBP1.

Fig. 6.

Carboxyl-terminal LPL sequences are required for the binding of LPL to mouse GPIHBP1. (A) Inhibition of cLPL binding to GPIHBP1 with increasing amounts of antibody xCAL 1–11. (B) Inhibition of cLPL binding to heparin by cLPL-specific monoclonal antibody xCAL 3–6a (28), but not by antibodies xCAL 3–7a or xCAL 1–11. (C) Ability of mutant cLPLs (in which residues 421–425, 426–430, and 431–435 were changed to alanine) to bind to GPIHBP1 on the surface of transfected pgsA-745 CHO cells. The amount of LPL binding to cells was quantified with an ELISA. (D) Western blot showing that a cLPL mutant lacking residues 416 to 435 (cLPL-Δ416–435) cannot bind to GPIHBP1. (E and F) Immunofluorescence microscopy assay shows that neither cLPL-Δ416–435 nor a cLPL mutant in which residues 421 to 425 were changed to alanine was able to bind to GPIHBP1. CHO-K1 cells transfected with S-protein–tagged human GPIHBP1 were mixed with CHO-K1 cells transfected with WT or mutant cLPL constructs. After a 2-h incubation at 37 °C, permeabilized (E) and nonpermeabilized (F) cells were stained for GPIHBP1 with an antibody against the S-protein tag (red) and LPL with antibody against the V5 tag (green). Cell nuclei were stained with DAPI (blue).