A METABOLOMIC, STRUCTURAL AND FUNCTIONAL STUDY OF POLYCYSTIN1 AND NEPHROCYSTIN1

Abstract Nephronophthisis (NPHP) and Autosomal Dominant Polycystic Kidney Disease (ADPKD) are two genetic renal cystic diseases that lead to End Stage Renal Disease (ESRD) in childhood or adolescence and at the late-middle age, respectively1. At present the mechanisms at the basis of cystogenesis in both diseases are poorly understood and pharmacological therapies are still lacking. On the one hand NPHP is characterized by kidney tubular atrophy and cysts formation occurring primarily at the cortico-medullary border, that leads to a reduction of kidneys volume. It is caused by mutations in 11 different genes. The most common form of NPHP is the juvenile one (NPHP type 1), which is caused, in most of the cases, by homozygous deletion of NPHP1 gene (2q13 chromosome). NPHP1 gene encodes for nephrocystin1 (NPHP1), a 732 amino acids long cytoplasmatic protein. It presents a widespread expression, but the pathogenic consequences due to loss-of-function are mainly confined to the kidney. In particular, NPHP1 is localized to cell-cell junctions, cilia and cell-matrix adhesion sites and it is involved in signaling transduction, cell-cell adhesion and cell polarity2-4. On the other hand the hallmark of ADPKD is the formation of bilateral cysts that leads to an increase of kidneys volume weight5,6. The disease is caused by mutations in two genes: Pkd1 gene (16p13.3 chromosome) which is mutated in 85% of cases and Pkd2 (4q21 chromosome) which is mutated in the remaining part of cases. These genes encode respectively for polycystin1 (PC1), a large plasma membrane receptor 4302 amino acids long and polycystin2 (PC2), a 968 amino acids long calcium channel. The two proteins interact through their C-terminal coiled coil domains and are involved in common signaling pathways, thus explaining the same phenotype5,7,8. In this thesis we focused our attention on two new possible roles of PC1: the regulation of metabolism and its interaction with NPHP1. To disentangle PC1 role as \u201cmetabolic regulator\u201d and its possible interaction with NPHP1 through its coiled coil domain, we have selected nuclear magnetic resonance (NMR) spectroscopy as primary investigation technique. PC1 as metabolic regulator. During a routine culture of mouse embryonic fibroblasts (MEFs) our collaborators (Dr. Boletta\u2019s group in San Raffaele Hospital, Milan, Italy) have observed that the knock out (Pkd1 KO) MEFs for PC1 acidify the medium faster than the wild type (WT) ones. This finding rises the hypothesis that PC1 is involved in the regulation of cellular metabolism. Hence, exploiting nuclear magnetic resonance (NMR) spectroscopy, we have applied a metabolomic approach in order to investigate the metabolic pathways and the signaling cascade deregulated by the loss of PC1. Importantly, the metabolomic analysis was performed both in cells and in mouse kidneys (WT and Pkd1 KO) in order to compare the results both in vitro and in vivo. The analysis of the 1D 1H NMR spectra of the media conditioned by the Pkd1 KO and by the WT MEF cells (exometabolome) has highlighted some metabolic differences: the exometabolome of the Pkd1 KO MEF cells displays an increased content of lactate, glutamate and alanine, presenting features similar to tumor cell metabolism9-11whereas the exometabolome of the WT MEF cells has an increased content of choline, formate, glucose, glycine and pyruvate. The two metabolites that vary mostly between the two cell lines are glucose and lactate, as highlighted by PCA analysis. In particular Pkd1 KO MEF cells display a major consumption of glucose, which is in turn related to a major production of lactate, showing the aerobic glycolysis, also called Warburg effect, a metabolic hallmark usually observed in cancer cells. In order to prove the occurrence of the Warburg effect in vivo, Ksp-Cre: Pkd1flox/- and Pkd1flox/+ mice12 were treated with uniformly labeled 13C glucose13,14. 1H-13C HSQC NMR spectra on the polar extracts of kidneys and livers (used as control organ) were acquired and analyzed in order to simplify the detection of the metabolic conversion of glucose into lactate. The Pkd1 KO kidneys displayed an increased content of glucose and lactate as compared to the WT ones, whereas livers did not display any difference. These results confirm that the aerobic glycolysis occurs also in vivo and is caused by the loss of functional PC1. In order to restore the metabolic basal condition, mice were treated with 2-deoxy-glucose (2DG), a drug able to block directly the glucose metabolism15,16. For testing the efficacy of the treatment, Ksp-Cre: Pkd1flox/- and Pkd1flox/+ mice were treated at the same time with 13C glucose and 2DG and the polar extracts of kidneys and livers were analyzed through 1H-13C HSQC NMR spectra. The Pkd1 KO kidneys treated with 2DG display a decrease of the lactate content, whereas the amount of glucose does not change, hence the treatment with 2DG affects the glucose metabolism but not its uptake. Moreover, from the histological point of view, the treated Pkd1 KO kidneys show a reduction of the volume of cysts and consequently a reduction of the weight of the kidney, compared to the not treated Pkd1 KO kidneys. Pkd1 KO livers do not present any changes after treatment with 2DG. These data show that 2DG acts specifically on Pkd1 KO kidneys, improving both the altered metabolism and the histology of the Pkd1 KO kidneys, without involving other organs that, instead, express functional PC1 (eg. the liver). In conclusion, our findings indicate that PC1 acts as \u201cmetabolic regulator\u201d, as shown by the analysis of the medium conditioned Pkd1 KO cells. In particular Pkd1 KO cells and Pkd1 KO kidneys are affected by the Warburg effect, a metabolic hallmark of accelerated metabolism observed also in cancer cells. Moreover, the analysis of Pkd1 KO kidneys of Ksp-Cre: Pkd1flox/- and Pkd1flox/+ mice treated with 2DG has highlighted an improvement of the Warburg effect and the renal histology. These preliminary results are very promising for a possible therapy for ADPKD patients and will be object of further in vivo studies. The coiled coil domain of PC1 does not interact with the N-terminus of NPHP1. We have recently demonstrated that one of the poly-proline motives, belonging to cytoplasmic C-terminus of PC1, is a binding partner for SH3 domain of NPHP1 and this complex is involved in the regulation of apoptosis17. The weak affinity of the resulting complex, as indicated by the estimated dissociation constant (Kd=0.3 +/-0.02 mM), suggests that other proteins or other domains of the same proteins could contribute to the interaction. In particular, the presence of coiled coil domains at PC1 C-terminus and at NPHP1 N-terminus, suggests the formation of a more complex macromolecular system involving interactions of both coiled coils. In order to verify and to structurally characterize at atomic level the interactions between PC1 and NPHP1 we have applied NMR spectroscopy. The first 115 amino acids (13kDa) of NPHP1 were predicted to form a coiled domain by the Web Server PCOIL. The coiled coil domain is a \u3b1 helical motif involved in homo and heteromultimerization18. Unexpectedly, recombinant NPHP11-115 behaves as a monomer, as assessed by the elution volume in size exclusion chromatography and by the overall correlation time deduced from NMR relaxation experiments. Circular dichroism experiments confirmed the presence of \u334 45% of \u3b1 helix and indicated a high thermostability of the domain, with a melting temperature of 65\ub0C. The solution structure of the domain shows that the first 115 amino acids of NPHP1 present a three helix bundle fold stabilized by hydrophobic interactions. The supposed interaction between the N-terminus of NPHP1 and the coiled coil domain of PC1 (CC_PC1) was next verified performing an NMR titration with the purified recombinant proteins, recording 1H-15N HSQC NMR spectra of NPHP1 upon addition of unlabelled CC-PC1. CC_PC1 was produced with the tags in order to improve the solubility and to avoid aggregation19. The recorded 1H -15N HSQC spectrum does not show any chemical shift displacement, indicating, that the two proteins do not interact. In conclusion our data show that the N-terminus of NPHP1 is a monomeric and thermostable domain characterized by the presence of three \u3b1 helices that pack together to form a three helix bundle. The N-terminus of NPHP1 is not a coiled coil domain, as erroneously predicted, and does not interact with CC_PC1. Interactions between NPHP1-115 and PC1 should be modulated by other protein domains or might involve additional proteins. Further experiments are needed to clarify this point