A polymorphism in New Zealand inbred mouse strains that inactivates phosphatidylcholine transfer protein

Abstract

New Zealand obese (NZO/HlLt) male mice develop polygenic diabetes and altered phosphatidylcholine metabolism. The gene encoding phosphatidylcholine transfer protein (PC-TP) is sited within the support interval for Nidd3, a recessive NZO-derived locus on Chromosome 11 identified by prior segregation analysis between NZO/HlLt and NON/Lt. Sequence analysis revealed that the NZO-derived PC-TP contained a non-synonymous point mutation that resulted in an Arg120His substitution, which was shared by the related NZB/BlNJ and NZW/LacJ mouse strains. Consistent with the structure-based predictions, functional studies demonstrated that Arg120His PC-TP was inactive, suggesting that this mutation contributes to the deficiencies in phosphatidylcholine metabolism observed in NZO mice.

Figure 1. Localization of Pctp within the Nidd3 locus

The Pctp gene maps within a previously published QTL peak for Nidd3 affecting both serum insulin and body weight [1]. The diabetogenic contribution was from NZO.

Figure 2. Sequence variations of Pctp in NZO compared with other mouse strains

In this schematic diagram of Pctp, the boxes represent exons. The solid lines indicate the 5′ promoter region and 3′ flanking DNA. The dashed lines indicate introns, which are not drawn to scale. Numeric values indicate bp positions upstream (negative) or downstream (positive) with respect to the transcription initiation site (http://www.ensembl.org/Mus_musculus). The positions of the ATG start codon, the TAA stop codon and AATAAA polyadenylation sequence are depicted as reference points. Arrows indicate location and identity of sequence variations identified in NZO Pctp: There was an A to G substitution at bp −86 within in the CCAAT-box of the promoter region and a deletion (represented by dashes) of 12 bp from −80 to −69. A C-to-T substitution at bp 338 within exon 3 was synonymous, leaving Ser at aa 110 unaltered. However, a G-to-A substitution at bp 367 in exon 4, was non-synonymous, resulting the substitution of Arg at aa 120 with His. There were two additional sequence differences outside the coding region in exon 6, each of which consisted of two bp insertions in NZO mice, as well as a 4 bp insertion and a 1 bp deletion in the NON strain.

Figure 3. Predicted influence of the Arg120His substitution on PC-TP structure

The three-dimensional structure of PC-TP (gray) in complex with phosphatidylcholine (green). The interactions of Arg120 with Asp70 and Asp122 are depicted by dotted lines, as are the interactions of the phenolic hydroxyl group of Tyr72 with the ligand. A modeled position of His120 (blue) is also depicted. Residues Leu68, Asp70 and Tyr72 are donated from helix α2 (residues 64 – 74; pink). This helix may mediate the effects of the Arg120His mutation to the lipid binding pocket.

Figure 4. The Arg120His substitution inactivates PC-TP in vitro without altering membrane binding

(A) PC-TP activities were compared among vector, vector containing wild type or Arg120His PC-TP. The indicated amounts of PC-TP in E. coli cytosol were used to measure PC transfer activity in vitro. The assay of vector control consisted of the corresponding amount of cytosolic protein. Data are represented as mean ± SD. (B) Samples of cytosol containing wild type and mutant PC-TP were admixed (50μg total PC-TP) to achieve varied molar ratios of wild type and Arg120His PC-TP. The nearly linear increase in activity as a function of increasing proportion of wild type PC-TP indicated that Arg120His did not inactivate the wild type protein in vitro. (C) Membrane binding of His-tag wild type and Arg120His PC-TPs was determined according to the free fraction of protein as a function of increasing phospholipids/protein ratios. The data are representative of duplicate determinations in two experiments.