Metabolic Profiling of Multiorgan Samples: Evaluation of MODY5/RCAD Mutant Mice

Elevated level of lysophosphatidic acid among patients with HNF1B mutations and its role in RCAD syndrome: a multiomic study

INTRODUCTION

Patients with hepatocyte nuclear factor-1 beta (HNF1B) mutations present a variable phenotype with two main symptoms: maturity onset diabetes of the young (MODY) and polycystic kidney disease (PKD).

OBJECTIVES

Identification of serum metabolites specific for HNF1Bmut and evaluation of their role in disease pathogenesis.

METHODS

We recruited patients with HNF1Bmut (N = 10), HNF1Amut (N = 10), PKD: non-dialyzed and dialyzed (N = 8 and N = 13); and healthy controls (N = 12). Serum fingerprinting was performed by LC-QTOF-MS. Selected metabolite was validated by ELISA (enzyme-linked immunosorbent assay) measurements and then biologically connected with HNF1B by in silico analysis. HepG2 were stimulated with lysophosphatidic acid (LPA) and HNF1B gene was knocked down (kd) by small interfering RNA. Transcriptomic analysis with microarrays and western blot measurements were performed.

RESULTS

Serum levels of six metabolites including: arachidonic acid, hydroxyeicosatetraenoic acid, linoleamide and three LPA (18:1, 18:2 and 20:4), had AUC (the area under the curve) > 0.9 (HNF1Bmut vs comparative groups). The increased level of LPA was confirmed by ELISA measurements. In HepG2^HNF1Bkd cells LPA stimulation lead to downregulation of many pathways associated with cell cycle, lipid metabolism, and upregulation of steroid hormone metabolism and Wnt signaling. Also, increased intracellular protein level of autotaxin was detected in the cells. GSK-3alpha/beta protein level and its phosphorylated ratio were differentially affected by LPA stimulation in HNF1Bkd and control cells.

CONCLUSIONS

LPA is elevated in sera of patients with HNF1Bmut. LPA contributes to the pathogenesis of HNF1B-MODY by affecting Wnt/GSK-3 signaling.

Strain-Specific Liver Metabolite Profiles in Medaka

The relationship between genetic variation and phenotypic traits is often poorly understood since specific genotypes do not always easily translate into associated phenotypes, especially for complex disorders. The genetic background has been shown to affect metabolic pathways and thus contribute to variations in the metabolome. Here, we tested the suitability of NMR metabolomics for comparative analysis of fish lines as a first step towards phenotype-genotype association studies. The Japanese rice fish, medaka (Oryzias latipes), is a widely used genetic vertebrate model with several isogenic inbred laboratory strains. We used liver extracts of medaka iCab and HO5 strains as a paradigm to test the feasibility of distinguishing the metabolome of two different inbred strains. Fifteen metabolites could be detected in uni- and multivariate analyses that showed strain-specific levels. Differences could be assigned to specific metabolic pathways. Our results show that NMR spectroscopy is a suitable method to detect variance of the metabolome caused by subtle genetic differences. Thus, it has the potential to address genotype–phenotype associations in medaka, providing an additional level of phenotypic analysis.

Hnf1b haploinsufficiency differentially affects developmental target genes in a new renal cysts and diabetes mouse model

Heterozygous mutations in HNF1B cause the complex syndrome renal cysts and diabetes (RCAD), characterized by developmental abnormalities of the kidneys, genital tracts and pancreas, and a variety of renal, pancreas and liver dysfunctions. The pathogenesis underlying this syndrome remains unclear as mice with heterozygous null mutations have no phenotype, while constitutive/conditional Hnf1b ablation leads to more severe phenotypes. We generated a novel mouse model carrying an identified human mutation at the intron-2 splice donor site. Unlike heterozygous mice previously characterized, mice heterozygous for the splicing mutation exhibited decreased HNF1B protein levels and bilateral renal cysts from embryonic day 15, originated from glomeruli, early proximal tubules (PTs) and intermediate nephron segments, concurrently with delayed PT differentiation, hydronephrosis and rare genital tract anomalies. Consistently, mRNA sequencing showed that most downregulated genes in embryonic kidneys were primarily expressed in early PTs and the loop of Henle and involved in ion/drug transport, organic acid and lipid metabolic processes, while the expression of previously identified targets upon Hnf1b ablation, including cystic disease genes, was weakly or not affected. Postnatal analyses revealed renal abnormalities, ranging from glomerular cysts to hydronephrosis and, rarely, multicystic dysplasia. Urinary proteomics uncovered a particular profile predictive of progressive decline in kidney function and fibrosis, and displayed common features with a recently reported urine proteome in an RCAD pediatric cohort. Altogether, our results show that reduced HNF1B levels lead to developmental disease phenotypes associated with the deregulation of a subset of HNF1B targets. They further suggest that this model represents a unique clinical/pathological viable model of the RCAD disease.

Application of Multiblock Analysis on Small Metabolomic Multi-Tissue Dataset

Data integration has been proven to provide valuable information. The information extracted using data integration in the form of multiblock analysis can pinpoint both common and unique trends in the different blocks. When working with small multiblock datasets the number of possible integration methods is drastically reduced. To investigate the application of multiblock analysis in cases where one has a few number of samples and a lack of statistical power, we studied a small metabolomic multiblock dataset containing six blocks (i.e., tissue types), only including common metabolites. We used a single model multiblock analysis method called the joint and unique multiblock analysis (JUMBA) and compared it to a commonly used method, concatenated principal component analysis (PCA). These methods were used to detect trends in the dataset and identify underlying factors responsible for metabolic variations. Using JUMBA, we were able to interpret the extracted components and link them to relevant biological properties. JUMBA shows how the observations are related to one another, the stability of these relationships, and to what extent each of the blocks contribute to the components. These results indicate that multiblock methods can be useful even with a small number of samples.