Fig 1.
Analysis of the sequencing data derived from the SGDP data set and the 1000GP data set.
(A) Mean allele frequency in the Simons Genome Diversity Project (SGDP) data set of all homozygous (colored) and heterozygous SNPs in the TPC2 gene leading to a missense mutation. (B) Cartoon showing the estimated positions of the polymorphisms in the human TPC2 protein. (C) Map of the worldwide distribution of TPC2L564P in the SGDP data set. Each circle represents one individual from an indigenous population, respectively. White circles represent individuals homozygous for TPC2P564L, red circles represent individuals homozygous for TPC2L564P, and pink circles represent heterozygous individuals. (D) Map of the worldwide distribution of TPC2L564P in the 1000GP data set. Each circle represents one of the 26 populations (indigenous or non-indigenous (other descent) as indicated). Color coding scheme corresponds to the one in (C).
Fig 2.
Geographical distribution of all homozygous TPC2 SNPs found in the SGDP data set.
The 297 samples were grouped according to their geographic origin. (A-F) Mean allele frequency of each SNP (mean ± SEM) grouped by geographic region, each symbol (circle, square, triangle) represents one sample (G) Mean allele frequency of each SNP grouped by continent.
Fig 3.
Geographical distribution of all homozygous TPC2 SNPs found in the 1000GP data set.
The 2054 samples were grouped according to geographic origin. (A-G) Mean allele frequency of each SNP (mean ± SEM) in each population grouped by geographic region, each symbol (circle, square, triangle, rhombus) represents one sample. (H) Mean allele frequency of each SNP grouped by continent.
Fig 4.
TPC2L564P in association with other TPC2 SNPs.
(A, B) Representative PI(3,5)P2 (1 μM) activated current densities in vacuolin-enlarged lysosomal vesicles expressing TPC2L564P or TPC2P564L and the corresponding currents after block with ATP (1 mM). (C-F) Frequency of other homozygous and heterozygous TPC2 SNPs on either L564 or 564P background. (C, D) Results from the SGDP data set. (E, F) Results from the 1000GP data set. (G-H) Frequency of homozygous SNPs on either L564 or 564P background in the SGDP data set or the 1000GP data set.
Fig 5.
Effect of PI(3,5)P2 on different variations of TPC2.
(A-I) Representative PI(3,5)P2 (1 μM) activated current densities in vacuolin-enlarged lysosomal vesicles isolated from HEK293 cells overexpressing different human TPC2 variants and the corresponding currents after block with ATP (1 mM). (J) Statistical summary of data as shown in A-I and in Fig 4A and 4B. Shown are average current densities (mean ± SEM) at -100 mV. Unpaired t-test were applied. *p < 0.05 and **p < 0.01.
Fig 6.
GWAS analysis to assess TPC2 SNP disease association.
(A-G) The Type 2 Diabetes Knowledge Portal was employed to assess phenotypes associated with high-frequency TPC2 SNPs [33]. Odds ratios were converted into effect sizes to render the data format uniform, and standard error of the mean estimated from P values, sample sizes, and effect sizes. Forest plots were subsequently plotted for the top five associated traits, centered around an effect size of 0. Decreasing effect sizes would suggest inverse associations between the trait and the SNP (such as for M484L and type 2 diabetes), while increasing effect sizes suggest direct associations (such as for G734E and bone mineral density). Colors indicate statistical significance, with dark red data-points indicating genome-wide significance (P < 10−8). (H) We further investigated the association between V219I and other SNPs. We found V219I and M484L SNPs to be linked, with less linkage between M484L and other SNPs. Since M484L affects channel function more dramatically than V219I, we assume shared phenotypes associated with V219I to be largely attributable to the linked M484L.