Fig 1: Hyperprolinemic mice have reproductive deficits and developmental delays.A Sanger sequence of S352L knock-in mice. B ALDH4A1 protein levels in knock-in mice. C Plasma proline levels (n = 3-4; error = standard error; 1-way ANOVA performed). D Urine proline levels (n = 5; error = standard error; 1-way ANOVA performed). E Litter sizes of knock-in mice. Independent samples t-test performed. F Ratio of genotypes born to heterozygote knock-in mouse crosses. χ2 test performed. G Survival curve of S352L knock-in mice (n = 25 WT, n = 34 HET, and n = 9 HOM). H Body weights of S352L knock-in mice post-weaning for females (left) and males (right). Each data point is an individual mouse connected by a straight line between measurements. Trend lines for each genotype are included (The number (n) of female mice per group is 30, 23, and 12 for WT, HET, and HOM, respectively. The number (n) of male mice per group is 22, 29, and 7 for WT, HET, and HOM, respectively.). I Representative image of male wildtype and homozygous S352L mice's body length. J Brain mass of S352L knock-in mice (n = 3; error = standard error; independent samples t-test performed). K Brain proline levels (n = 4; error = standard error; independent samples t-test performed). L SMS protein levels in mouse brains (n = 3; error = standard error; independent samples t-test performed). M The ratio of astrocytes and neurons measured by GFAP count compared to NeuN count is lower in hippocampus of S352L homozygous mice (n = 4; error = standard error; see Supplementary Data Fig. 6E for representative image). Independent samples t-test performed. *p < 0.05; ****p < 0.0001.
Fig 2: ALDH4A1 deficiency leads to increased susceptibility to aldehydic load, protein aggregation, transcriptomic changes, and cytotoxicity.Purple residue is the position of the S352L HPII variant. Blue residues are the catalytic residues in the active site. Created in BioRender: https://BioRender.com/psdqc2p licensed under CC BY 4.0.
Fig 3: Knock-in S352L iPSCs, iNeurons, and iAstrocytes recapitulate the human phenotype and indicate cell-type specificity.A Sanger sequencing of S352L knock-in point mutation. B ALDH4A1 protein levels in homozygous wild type and S352L human iPSCs. C Intracellular proline in iPSCs (n = 3 independent cultures from the same cell line; error = standard error; independent samples t-test performed). D Extracellular proline in iPSCs (n = 3 independent cultures from the same cell line; error = standard error; independent samples t-test performed). E ALDH4A1 protein levels in iNeurons and iAstrocytes. F Intracellular proline levels in human iNeurons and iAstrocytes (n = 3 independent cultures from the same cell line, error = standard error; independent samples t-test performed). G Extracellular proline levels in iNeurons and iAstrocytes (n = 8 independent cultures from the same cell line, error = standard error; 2-way ANOVA performed). H Knock-in iPSCs are more sensitive to 4-HNE. Cells were treated with vehicle or 4-HNE for 16 hours. Each point is the average of 3 technical replicates (n = 3 independent experiments; error = standard error). I Levels of 4-HNE protein adducts in WT and S352L ALDH4A1 neural stem cells measured by western blot (left) and quantified (right). Cells were treated with vehicle or 30 µM 4-HNE for 1.5 hours. Each point is an independent culture (n = 3 independent cultures from the same cell line; two-way ANOVA performed). ns = non-significant;*p < 0.05; **p < 0.01; ***p < 0.001.
Fig 4: ALDH4A1 is necessary for 4-HNE clearance and the S352L HPII variant cannot oxidize the non-canonical substrate.A 4-HNE does not inactivate recombinant wild-type ALDH4A1. Each large point is the average of n = 3 technical replicates (smaller points); error = standard deviation. B Recombinant S352L ALDH4A1 is unable to oxidize 4-HNE. Each point is the average of 3 technical replicates (n = 3 independent experiments; error = standard error; independent samples t-test performed). C ALDH4A1 knock-down leads to increased sensitivity to 4-HNE-induced toxicity. BE(2)-M17 neuroblastoma cells were treated with vehicle or 30 µM 4-HNE for 1 hour. Each point is the average of 3 technical replicates (n = 3 independent experiments; error = standard error; 2-way ANOVA performed). D ALDH4A1 knock-down leads to increased protein aggregation. BE(2)-M17 neuroblastoma cells were treated with vehicle or 1 µM 4-HNE for 16 hours followed by treatment with vehicle or 30 µM 4-HNE for 15 minutes. Each point is an individual cell (nCTRL, CTRL = 134, nCTRL,4-HNE = 87, nALDH4A1, CTRL = 97, nALDH4A1,4-HNE = 161). Error = standard error; 2-way ANOVA performed. *p < 0.05; **p < 0.01; ****p < 0.0001.
Fig 5: ALDH4A1 is highly conserved, and the S352L HPII variant is highly unstable and less active.A ALDH4A1 metabolic pathway. B Many understudied ALDH4A1 variants exist within the human population and may increase the risk of disease. Green bars indicate positions with > 2 missense variants and blue bars indicate positions with < 3 missense variants. The active site (black), dimerization domain (blue), and NAD+ binding site (orange) are indicated above the histogram. In silico predicted effects of missense variants are denoted with purple (very damaging), green (slightly damaging), blue (benign), or gray (unknown) dots above the protein domains and separated by commonality (> 0.1% allele frequency). Residue S352 is indicated with a red arrow. C Sequence alignment of ALDH4A1 S352 position across species. Sequences coloured according to conservation across species and S352 position outlined in red. D Activity of ALDH4A1 S352L variant compared to wild type with allysine. Each point is the average of n = 3 technical replicates from 3 independent experiments normalized to WT activity levels. Error = standard error. Independent samples t-test performed. E ALDH4A1 protein stability in NIH3T3 cells at time 0 and 24 hours after cycloheximide treatment (n = 3 independent cultures). F ALDH4A1 protein levels in NIH3T3 cells at time 0 (n = 3 independent cultures; error = standard error; independent samples t-test performed) measured by western blot. G ALDH4A1 protein levels in NIH3T3 cells at times 0, 1, 3, 6, 8, and 24 hours after cycloheximide treatment. Quantification of cycloheximide chase protein levels over 24 hours was normalized to protein levels at time = 0 for each variant (n = 3 independent cultures; error = standard error; 2-way ANOVA performed). Large points are means of independent replicates (smaller points). H The half-life of WT and S352L ALDH4A1 was calculated from G. *p < 0.05; **p < 0.01; ***p < 0.001.
Supplier Page from Sino Biological, Inc. for Human ALDH4A1 Gene ORF cDNA clone expression plasmid