Combined action of albumin and heparin regulates lipoprotein lipase oligomerization, stability, and ligand interactions

Abstract

Lipoprotein lipase (LPL), a crucial enzyme in the intravascular hydrolysis of triglyceride-rich lipoproteins, is a potential drug target for the treatment of hypertriglyceridemia. The activity and stability of LPL are influenced by a complex ligand network. Previous studies performed in dilute solutions suggest that LPL can appear in various oligomeric states. However, it was not known how the physiological environment, that is blood plasma, affects the action of LPL. In the current study, we demonstrate that albumin, the major protein component in blood plasma, has a significant impact on LPL stability, oligomerization, and ligand interactions. The effects induced by albumin could not solely be reproduced by the macromolecular crowding effect. Stabilization, isothermal titration calorimetry, and surface plasmon resonance studies revealed that albumin binds to LPL with affinity sufficient to form a complex in both the interstitial space and the capillaries. Negative stain transmission electron microscopy and raster image correlation spectroscopy showed that albumin, like heparin, induced reversible oligomerization of LPL. However, the albumin induced oligomers were structurally different from heparin-induced filament-like LPL oligomers. An intriguing observation was that no oligomers of either type were formed in the simultaneous presence of albumin and heparin. Our data also suggested that the oligomer formation protected LPL from the inactivation by its physiological regulator angiopoietin-like protein 4. The concentration of LPL and its environment could influence whether LPL follows irreversible inactivation and aggregation or reversible LPL oligomer formation, which might affect interactions with various ligands and drugs. In conclusion, the interplay between albumin and heparin could provide a mechanism for ensuring the dissociation of heparan sulfate-bound LPL oligomers into active LPL upon secretion into the interstitial space.

Fig 1. Catalytic activity of LPL measured with DGGR after incubation in various conditions.

(A) 200 nM LPL was incubated in buffer A (▲), which contained 10 IU/ml heparin (◆), 50 mg/ml BSA (■) or both (●). Alternatively, LPL was incubated in LFHP (▼). LPL activity is expressed relative to the initial activity of the experiment that contained both heparin and BSA. Results at 75 minutes were compared by two-tailed Student’s t-test. *P<0.05. LPL loses its activity quickly in plain buffer but is stabilized by lipoprotein free human plasma (LFHP) or its main constituent albumin (BSA). The most significant stabilization was observed when both BSA and heparin were used together. (B) 200 nM LPL was incubated in macromolecularly crowded buffer A. LPL activity is expressed as relative fluorescence units per second (RFU/s). PEG 6 nor dextran 40 could stabilize LPL like BSA, indicating that macromolecular crowding alone is not sufficient for LPL stabilization. (C) 200 nM LPL was incubated in buffer A with various proteins. Casein and β-lactoglobulin were as efficient as BSA in stabilizing LPL, however lysozyme failed to exert any effect. This suggests that LPL stabilization by proteins depends on their specific characteristics such as charge or hydrophobicity.

Fig 2. Catalytic activity of LPL on triglyceride-rich lipoproteins in the presence of fatty acid acceptors and macromolecular crowders.

Measurements were performed using ITC. LPL activity is expressed as heat rate (μJ/s). The substrate mixture contained CM/VLDL (adjusted to 0.5 mM triglycerides), 50 mg/ml BSA (0.75 mM) or 10 mg/ml β-cyclodextrin (8.81 mM) (CYC) and either 10% PEG 6 or 10% dextran 40 (DEX).

Fig 3. Interaction of LPL with BSA as studied using ITC, stabilization of LPL, and SPR.

(A) An example ITC thermogram for the titration of BSA into a solution with bLPL, obtained after subtraction of the BSA dilution effect. The concentration of bLPL in the cell was 1.695 μM. (B) Fitted isotherm for the binding between BSA and bLPL from the ITC experiment on panel A. (C) Enzymatic stability of LPL in the presence of BSA as determined with DGGR. 200 nM bLPL or hLPL was incubated for 60 minutes in buffer A at various BSA concentrations. The values are calculated relative to the initial activity of LPL in the same conditions. BSA stabilized both bLPL and hLPL in a concentration-dependent manner. (D) SPR sensograms showing binding of BSA to biotinylated bLPL that was attached to pre-immobilized neutravidin. (E) SPR sensorgrams showing BSA binding to bLPL that was attached to pre-immobilized GPIHBP1. In D and E, BSA concentrations are shown on sensograms. Non-specific binding sensorgrams of BSA to streptavidin and GPIHBP1, respectively, have been subtracted. (F) Plateau values of sensorgrams plotted against BSA concentration. ●—Binding of BSA with 0.1 IU/ml heparin to GPIHBP1-bound bLPL. ▲—Binding of BSA to biotinylated bLPL. ◆—Binding of BSA to 5D2-bound bLPL. Dashed line—1:1 ratio of immobilized bLPL to bound BSA in the experiment with GPIHBP1-bound bLPL. The results indicate that BSA can bind to neutravidin-bound LPL or GPIHBP1-bound LPL but not 5D2-bound LPL.

Fig 4. Effect of BSA, Triton X-100 or LFHP on LPL activity in the presence of heparin.

(A) 200 nM LPL was incubated for the indicated timepoints with 50 mg/ml BSA, 10 IU/ml heparin or both in buffer A. The remaining activity (expressed as heat rate, μJ/s) was determined with ITC after a single 5 nM LPL injection into human plasma that contained 1.31 mM triglycerides. LPL activity is expressed relative to the initial activity of the experiment that contained both heparin and BSA and the data is presented as mean ± SD of three independent measurements. LPL activity was significantly lower when only BSA or heparin was used. (B) 1 μM LPL was incubated with 10 IU/ml heparin in buffer A and diluted 5-fold to buffer A with heparin and 0.5% Triton X-100, 50 mg/ml BSA or LFHP. The remaining LPL activity was determined in the same manner as panel A. LPL activity was restored equally with every additive.

Fig 5. Effect of LPL incubation concentration on its activity as measured with ITC.

LPL activity, expressed as heat rate (μJ/s), after incubation of LPL at various concentrations with 50 mg/ml BSA with (▲) or without (●) 10 IU/ml heparin in buffer A. The changes in residual LPL activity suggest that LPL oligomerization triggered by BSA is dependent on LPL concentration. This dependence disappears with the combined use of heparin and BSA which blocks the formation of LPL oligomers.

Fig 6. Oligomerization of LPL as studied by RICS and TEM.

10 nM LPL-ATTO610 with 190 nM unlabeled LPL was incubated in buffer A with the indicated supplements for up to two hours at room temperature. (A) Relationship between diffusion coefficient (D) and number of particles (N) in confocal volume (proportional to concentration) for image sectors recorded in an experiment with 200 nM LPL and 50 mg/ml BSA. Larger particles have lower D values. Notice that a large fraction of measurements leads to estimates with low D and N values (red circle) (B), (D), (F) Representative nsTEM micrographs of unlabeled LPL in similar conditions as RICS experiments. Scale bars are 100 nm. (C) Interaction of LPL and heparin leads to formation of a fraction of estimates with lower D values at N values between 2–3. This is visible in the plot as a smaller set of points grouped around D = 10 μm²/s (red circle) with the larger set of estimates above it. (E) Relationship between D and N for an experiment with LPL, BSA and heparin. Notice that most of the image sectors lead to D and N estimates in a certain region of the plot (red circle).

Fig 7. Inactivation of oligomeric LPL by ANGPTL4.

20 nM LPL (A) or 200 nM LPL (B) was incubated in the presence of 1 μM ANGPTL4 and 10 IU/ml heparin (●) or 50 mg/ml BSA (■) or both (▼) in buffer A. Remaining LPL activity was determined by ITC using undiluted pooled human plasma (0.89 mM triglycerides, n = 5). LPL activity is expressed relative to initial activity in the same conditions without ANGPTL4. Symbols represent individual measurements and lines correspond to their average values. The results suggest that the rate of LPL inactivation by ANGPTL4 depends on the degree of LPL oligomerization.

Fig 8. Interaction of albumin with nANGPTL4.

(A) Effect of BSA at various concentrations on nANGPTL4-induced inactivation of LPL during lipolysis of triglyceride-rich lipoproteins. Raw ITC thermograms of lipolysis of CM/VLDL (adjusted to 0.75 mM triglycerides) by LPL in the presence of 100 nM nANGPTL4 and various concentrations of BSA (2 mg/l, 10 mg/ml, 50 mg/ml). The lipolysis rate is expressed as heat rate, μJ/s. The control experiment was performed in the absence of nANGPTL4 and presence of 50 mg/ml BSA. k_i represents inhibition rate constant of LPL by nANGPTL4 calculated from the data (n = 2). The results demonstrate that the rate of LPL inactivation by nANGPTL4 is increased at higher BSA concentrations. (B) SPR analysis of binding of various concentrations of HSA to immobilized nANGPTL4. Plateau values of SPR sensorgrams were plotted against HSA concentrations after subtracting non-specific binding. The results indicate that HSA interacts with nANGPTL4.

Fig 9. Summarized results and proposed model for oligomerization of LPL.

The model describes possible states of LPL according to in vitro investigations.