Novel Abetalipoproteinemia Missense Mutation Highlights the Importance of the N-Terminal β-Barrel in Microsomal Triglyceride Transfer Protein Function

Abstract

Background: The use of microsomal triglyceride transfer protein (MTP) inhibitors is limited to severe hyperlipidemias because of associated hepatosteatosis and gastrointestinal adverse effects. Comprehensive knowledge about the structure-function of MTP might help design new molecules that avoid steatosis. Characterization of mutations in MTP causing abetalipoproteinemia has revealed that the central α-helical and C-terminal β-sheet domains are important for protein disulfide isomerase binding and lipid transfer activity. Our aim was to identify and characterize mutations in the N-terminal domain to understand its function.

Methods and results: We identified a novel missense mutation (D169V) in a 4-month-old Turkish male child with severe signs of abetalipoproteinemia. To study the effect of this mutation on MTP function, we created mutants via site-directed mutagenesis. Although D169V was expressed in the endoplasmic reticulum and interacted with apolipoprotein B (apoB) 17, it was unable to bind protein disulfide isomerase, transfer lipids, and support apoB secretion. Computational modeling suggested that D169 could form an internal salt bridge with K187 and K189. Mutagenesis of these lysines to leucines abolished protein disulfide isomerase heterodimerization, lipid transfer, and apoB secretion, without affecting apoB17 binding. Furthermore, mutants with preserved charges (D169E, K187R, and K189R) rescued these activities.

Conclusions: D169V is detrimental because it disrupts an internal salt bridge leading to loss of protein disulfide isomerase binding and lipid transfer activities; however, it does not affect apoB binding. Thus, the N-terminal domain of MTP is also important for its lipid transfer activity.

Keywords: apolipoproteins; apolipoproteins B; cardiovascular diseases; chylomicrons; lipids; lipoproteins; microsomal triglyceride transfer protein.

Figure 1

D169V binds apoB17, but is unable to support apoB48 secretion. Cos-7 cells were transfected with plasmids expressing apoB48 or apoB17 and then with plasmids expressing FLAG tagged MTP or D169V. (A–B) Intracellular (A) and media (B) apoB48 were measured by ELISA. Data representative of three experiments; mean, n = 3. Mann-Whitney Test *P ≤0.05. (C) Expression of recombinant proteins was confirmed in cell lysates. The membranes were cut in half and probed for FLAG and GAPDH. (D) Immunofluorescent staining for FLAG (red) and calnexin (green). Experiment performed in duplicate. Multiple fields were scanned. Bars, 25 µm. (E) In one well (apoB17 -), cells were transfected with plasmids expressing MTP. Other wells were transfected with apoB17 and MTP expression plasmids. ApoB17 was immunoprecipitated. Immunoprecipitates were immunoblotted with anti-FLAG antibodies. Membranes were stripped and probed for apoB. (F) Anti-MTP immunoprecipitates were probed for apoB17 and FLAG. Experiments in panels (E) and (F) were performed in duplicate.

Figure 2

D169V does not bind PDI or transfer lipids. Cos-7 cells were transfected with plasmid expressing MTP or D169V. (A) Purified WT or mutant MTP were used to measure phospholipid (PL) transfer activity. (B) Triglyceride (TG) transfer activity was measured in cell lysates. Data in panels (A) and (B) are representative of two experiments performed in triplicate. (C) Immunofluorescent staining for FLAG (red) and PDI (green) in cells expressing WT or mutant MTP. (D) Purified MTP and D169V were probed for FLAG and PDI. Bands were quantified via Image J. The PDI:MTP ratio was normalized to WT MTP. Data representative of two experiments run in duplicate. Experiments were combined for statistical analysis (n=4). Mann-Whitney Test *P≤0.05. The line represents the mean. (E) Cells were lysed in Triton lysis buffer and centrifuged. Supernatants were centrifuged again at 100,000 g. MTP was detected in both the soluble (S) supernatant and insoluble (I) pellet via western blot. The bands were quantified with Image J. To calculate % solubility, the intensity of the soluble fraction was divided by the sum of the pellet and soluble fraction. The experiment was performed in duplicate.

Figure 3

Disposition of D169V in the predicted secondary and tertiary structure of MTP. (A) A schematic secondary structural representation of the M subunit modified from. Green arrows and red cylinders depict β-strands and α-helices, respectively. The top row labels the N-terminal (β^N), middle α-helical domain (α), and C-terminal domain (β^C, β^A sheets). Biochemically characterized loss-of-function ABL mutations are highlighted below and the position of D169V is highlighted above the secondary structure in yellow. (B) The predicted tertiary structure is color coded by structural domains; β^N, green, N-terminal β-sheet; α cyan, central α-helical domains; β^C, β^A dark blue, C-terminal domain. The sphere’s represent D169’s side chain. (C) Magnified view of the N-terminal region. D169 faces the interior of the β-barrel. (D) Ribbon diagram of the interior of the N-terminal β-barrel. Mutated amino acid side chains are depicted as sticks. Light blue lines portray possible salt bridges.

Figure 4

The ionic interactions between D169-K187-K189 are required for apoB secretion Experiments were performed similarly to Figure 1. (A–B) ApoB48 was measured in lysates (A) or media (B) of transfected cells. (C) Immunofluorescent staining for FLAG (red) and Calnexin (green). Bars, 25 µm. (D) Western blot of cell lysates. Black lines depict different areas of the same gel cut and brought together for proper presentation. Data representative of two experiments performed in triplicate. (E) Anti-apoB immunoprecipitates were blotted for apoB and FLAG. (F) Proteins were immunoprecipitated with anti-FLAG and western blotted sequentially for apoB and FLAG. All WT and mutant MTP were run on the same gel. Each experiment (E–F) was performed in duplicate.

Figure 5

The ionic interactions between D169-K187-K189 are required for lipid transfer and PDI binding. Experiments were performed as in Figure 2. (A) Immunofluorescent staining was performed for PDI (green) or FLAG (red). Bars, 25 µm. (B) FLAG and PDI were detected in purified WT or mutant MTP. Two independent experiments were performed in duplicate and combined for quantification (n=4). (C) MTP was detected in the soluble (S) and insoluble (I) fractions (left) and quantified to calculate % solubility (right). The experiment was performed in duplicate. (D) TG transfer activity was measured in cell lysates. (E) PL transfer activity was measured in purified protein. Data representative of two independent experiments. Mean, n = 3. Significance among the MTP variants was calculated by the Kruskal-Wallis test. The Mann-Whitney test compared differences between MTP and each mutant. *P≤0.05.

Figure 6

Preservation of charges in the D169-K187-K189 salt bridge does not affect subcellular localization or apoB binding. Experiments were performed similarly to Figure 1. (A) Immunofluorescent staining for calnexin (green) and FLAG (red). Bars, 25 µm. (B) Western blot of whole cell lysates. (C–D) Cos-7 cells were cotransfected with apoB17 and MTP. ApoB17 (C) or MTP (D) was immunoprecipitated. Immunoprecipitates were immunoblotted for FLAG and apoB. Each blot is representative of an experiment performed in duplicate.

Figure 7

Preservation of charges at D169-K187-K189 restores PDI binding and lipid transfer activities. Experiments were performed similarly to Figure 2. (A) Immunofluoresence staining for PDI (green) and FLAG (red). Bars, 25 µm. (B) MTP was purified using anti-FLAG antibodies and probed for FLAG and PDI (left). MTP and PDI bands were quantified and normalized to WT MTP (right). (C) Solubility of MTP as in Figure 2. (D) TG transfer was measured in cell lysates. (E) PL transfer was measured in purified protein. (F–G) Intracellular (F) and secreted (G) apoB48 were measured. Data representative of two experiments. Significance calculated by the Kruskal-Wallis test amongst four groups followed by the Mann-Whitney test between MTP and individual mutant. Mean, n = 3. *P≤0.05.

Figure 8

Evolutionary conservation and hypothetical models show extended E167-D169-S171-Y178-K187-K189 salt bridge and apoB17 binding site. (A) Amino acids 163–195 in β-strands 7–9 of MTP orthologs were aligned using Clustal Omega. Arrowheads point to the members of the extended salt bridge (E167, D169, S171, Y178, K187, and K189). Green arrows below represent the secondary structure of this region. (B) Magnified interior of the N-terminal β-barrel. Amino acid side chains that could possibly interact with other members of the extended salt bridge are shown. The blue lines represent possible electrostatic interactions. (C) The two major binding sites for PDI are in yellow. This study identified the importance of the N-terminal β-strands 7–9 in PDI binding. Middle α-helices 13–17 have been previously shown to interact with PDI. (D) Expanded view of the N terminal domain illustrating the two microdomains. β-strands 7–9 (yellow) are necessary for PDI binding. β-strands 1–3 (orange) and helix NH1 (red) may be involved in apoB binding. Helix NH2 (purple) appears to divide these two microdomains.