- Email: [email protected]
Paraoxonase 1 deficiency and hyperhomocysteinemia alter the expression of mouse kidney proteins involved in renal disease
Molecular Genetics and Metabolism 113 (2014) 200–206
Contents lists available at ScienceDirect
Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme
Paraoxonase 1 deficiency and hyperhomocysteinemia alter the expression of mouse kidney proteins involved in renal disease☆ Joanna Suszyńska-Zajczyk a, Marta Sikora a, Hieronim Jakubowski a,b,c,⁎ a b c
Institute of Bioorganic Chemistry, Poznań, Poland Department of Biochemistry and Biotechnology, University of Life Sciences, Poznań, Poland Department of Microbiology & Molecular Genetics, Rutgers-New Jersey Medical School, International Center for Public Health, Newark, NJ, USA
a r t i c l e
i n f o
Article history: Received 4 June 2014 Received in revised form 9 July 2014 Accepted 10 July 2014 Available online 17 July 2014 Keywords: Paraoxonase 1 Polyuria High-methionine diet Hyperhomocysteinemia Mouse kidney proteome Kidney disease
a b s t r a c t Scope: Hyperhomocysteinemia (HHcy) is associated with kidney disease and leads to atherosclerosis and thrombosis. Paraoxonase 1 (Pon1), a hydrolase that participates in homocysteine (Hcy) metabolism and is carried in the circulation on high-density lipoprotein, has also been linked to kidney disease and atherothrombosis. Pon1-knockout mice are susceptible to atherosclerosis and exhibit a kidney-associated phenotype, polyuria or urine dilution. We hypothesize that HHcy and Pon1 deficiency are toxic to kidney function because they impair metabolic pathways important for normal kidney homeostasis. Methods and results: We examined changes in the mouse kidney proteome induced by Pon1 gene deletion and dietary HHcy, using 2D IEF/SDS-PAGE gel electrophoresis and MALDI-TOF mass spectrometry. We found that the expression of ten mouse kidney proteins was altered by the Pon1−/− genotype or HHcy. Proteins involved in metabolism of lipid (ApoA-I), protein (Hspd1), carbohydrate (Pdhb, Fbp1-isoform2, Eno1), and energy (Ndufs8, Ldhd) were down-regulated. Proteins involved in lipid transport (Pebp1), oxidative stress response (Prdx2), and cellular detoxification (Glo1) were up-regulated. The kidney proteins altered by HHcy or Pon1 are also altered in renal disease. Conclusion: Our findings suggest that excess Hcy is toxic because it deregulates the expression of proteins involved in diverse cellular processes–from lipid, protein, carbohydrate, and energy metabolisms to detoxification and antioxidant defenses–that are essential for normal kidney homeostasis. Dysregulation of these processes can account for the involvement of HHcy and reduced Pon1 in kidney disease. Our findings also show that Pon1 plays an important role in maintaining normal kidney homeostasis. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Paraoxonase 1 (PON1) is expressed in many tissues, including the kidney, liver, lung, brain, and colon (both PON1 mRNA and protein are found in those organs), circulates in the blood attached to highdensity lipoproteins (HDL), and is distributed to all tissues both in humans and rodents [1–3]. PON1 has a cardio-protective function in mice [4] and humans [5], which has been linked to its ability to mediate detoxification of oxidized lipids and homocysteine (Hcy)-thiolactone [6]. Reduced PON1 activity is linked to cardiovascular disease [7,8]. In
☆ This work was supported in part by grants from the American Heart Association (12GRNT9420014), the National Science Center (2011/01/B/NZ1/03417, 2011/02/A/ NZ1/00010, 2012/07/B/NZ7/01178, and 2013/09/B/NZ5/02794) and the MNiSW, Poland (N401 065321504 and N N302 434439). ⁎ Corresponding author at: Department of Microbiology, Biochemistry & Molecular Genetics, Rutgers-New Jersey Medical School, 225 Warren Street, Newark, NJ 07101-1709, USA. Fax: +1 973 972 8982. E-mail address: [email protected] (H. Jakubowski).
http://dx.doi.org/10.1016/j.ymgme.2014.07.011 1096-7192/© 2014 Elsevier Inc. All rights reserved.
mice, Pon1 gene deletion impairs urine concentrating ability, which leads to a polyuria phenotype [9]. Pon1 activity is reduced in patients with chronic kidney disease (CKD) [10] and predicts risk of heart attack, stroke, and death [11]. CKD patients also have elevated plasma Hcy [12], an emerging cardiovascular risk factor. Plasma Hcy is negatively correlated with glomerular filtration rate, positively with mortality risk, and accounts for a significant portion of morality in CKD [13,14]. However, mechanisms underlying the toxicity of elevated Hcy and reduced Pon1 in the kidney are not understood [15, 16]. Pon1 links HDL and Hcy metabolisms, which may account for their role in kidney disease. Hcy is a negative determinant of HDL and PON1 [17] and reduces Apoa1 and Pon1 gene expression [18]. HDL and purified PON1 detoxify Hcy-thiolactone [19] and protect against protein Nhomocysteinylation [9,20]. N-Homocysteinylation causes protein damage and is linked to kidney and cardiovascular diseases (reviewed in [21]). Atherosclerosis-related (N-homocysteinylated ApoA-I or N-Hcy-ApoA-I) and other N-Hcy-proteins [22] increase in hyperhomocysteinemia (HHcy), including CKD patients [12] and kidneys of Pcft−/− mice [21].
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
201
Fig. 1. Representative IEF/SDS-PAGE gel showing the kidney proteome of a wild type C57BL/6J mouse. Left to right: IEF-pH gradient pH 4 → pH 7. Top to bottom: SDS-PAGE; molecular weight markers, 10–260 kDa, shown on the left. Numbers indicate protein spots whose identity was established by mass spectrometry. White rectangles outline areas containing proteins whose expression was affected by the Pon1 genotype and/or high-Met diet. Identities of the numbered protein spots are described in Supplementary Table S1.
The Pon1-null mice are more susceptible to Hcy-thiolactone toxicity than wild type mice and exhibit a polyuria phenotype: produce twice as much urine as their wild-type littermates [9]. Taken together, these findings suggest that Pon1 plays important roles in the kidney.
We hypothesize that HHcy and Pon1 deficiency are toxic because they impair metabolic pathways important for kidney homeostasis. To test this hypothesis, we examined changes in the mouse kidney proteome induced by dietary HHcy in wild type and Pon1-null mice.
Table 1 Differentially expressed kidney proteins regulated by Pon1−/− genotype and/or high-Met diet. Protein description (spot #)
Lipoprotein and lipid metabolism Apolipoprotein A-I (#78) Phosphatidylethanolamine-binding protein 1 (#86) Amino acid and protein metabolism 60 kDa heat shock protein, mitochondrial (#14) Energy metabolism NADH dehydrogenase [ubiquinone] iron–sulfur protein 8, mitochondrial (#72) Probable D-lactate dehydrogenase, mitochondrial (#18) Carbohydrate metabolism Pyruvate dehydrogenase E1 component subunit beta, mitochondrial (#47) Fructose-1,6-bisphosphatase 1 isoform 2 (#62) Fructose-1,6-bisphosphatase 1 isoform 1 (#63) Alpha enolase (#34) Oxidative stress response Peroxiredoxin 2 (#85) Methylglyoxal detoxification Glyoxalase 1 (#87) Spot # refers to the numbering on the IEF/SDS-PAGE gels in Fig. 1 and Fig. 2. Significantly different: aP b 0.001, bP b 0.01, cP b 0.05.
Gene name
Fold change Pon1−/− vs. Pon1+/+
Fold change std vs. 1% Met diet
Std diet
Pon1+/+
Pon1−/−
−2.27a 1.23
−1.37a 1.02
1%-Met diet
Apoa1 Pebp1
−1.47a 1.32b
Hspd1
−1.44a
−1.22b
−1.22b
1.04
Ndufs8 Ldhd
−1.96a −1.27
1.07 1.77a
−2.00a −2.00a
1.05 1.11
Pdhb Fbp1 Fbp1 Eno1
−1.19b −2.00a 1.01 −2.00a
−1.02 1.49b −1.01 1.17
−1.44a −2.27a 1.25c −1.59a
−1.23a 1.29c 1.23 1.48a
Prdx2
1.88a
1.21a
1.78a
1.14a
Glo1
1.17
1.45b
1.08
1.15
1.11 1.09
202
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
2. Methods
were analyzed by ANOVA. Unpaired Student's t-test was used to test differences between two groups. Statistica 8.0 software was used.
2.1. Mice and diets 3. Results Pon1−/− mice [4] and wild type Pon1+/+ littermates (C57BL/6J) were maintained at the Rutgers-New Jersey Medical School Animal Facility. Female mice (n = 8) were fed a standard rodent chow containing 0.66% Met (LabDiet 5010, Purina Mills International, St. Louis, MO). At 4 weeks of age, half of Pon1−/− and Pon1+/+ mice were provided 1%Met in drinking water (high-Met diet) for 8 weeks [9]. Four experimental groups of mice were studied: 1) Pon1−/−, control diet; 2) Pon1+/+, control diet; 3) Pon1−/−, high-Met diet; and 4) Pon1+/+, high-Met diet. Drinking water supplementation with 1%-Met did not affect animals' water intake and body weight. Animal procedures were approved by the Institutional Animal Care and Use Committee.
3.1. Dietary hyperhomocysteinemia To induce HHcy, mice were fed with a high-Met diet for 8 weeks [23]. HHcy was confirmed by plasma tHcy and N-Hcy-protein assays. Plasma tHcy levels in Pon1−/− and Pon1+/+ mice fed a standard diet were 8.5 ± 1.9 μM and 7.4 ± 2.2 μM, and increased to 48 ± 16 μM and 77 ± 45 μM, respectively, in animals fed with a high-Met diet. Plasma N-Hcy-protein levels increased from basal levels of 1.4 ± 0.5 μM and 1.2 ± 0.4 μM in mice fed with a standard diet to 3.8 ± 1.8 μM and 5.4 ± 2.9 μM in hyperhomocysteinemic Pon1−/− and Pon1+/+ animals that were fed with a high-Met diet, respectively.
2.2. Genotyping 3.2. Proteins differentially expressed in Pon1−/− kidney Genotyping of the Pon1 locus [4] was carried out as described in the Supplementary Methods [23]. The Pon1−/− genotype was confirmed by the absence of enzymatic activities of Pon1 (Hcy-thiolactonase, paraoxonase, arylesterase) in serum [19,20]. 2.3. Hcy assays Total Hcy and N-Hcy-protein were assayed as previously described [22]. 2.4. Proteomic analyses Protein sample preparation, 2D-IEF/SDS-PAGE analyses, and protein identification by MALDI-TOF mass spectrometry [23–25] were carried out as described in the Supplementary Methods. 2.5. Data treatment For each animal in the four experimental group (4 animals/group), analyses were repeated 2–3 times. Relative abundances of protein spots–% volume–were calculated by dividing their volume by the total volume of all spots. Data are expressed as mean ± SD. Data for each protein spot had a normal distribution. The differences between the groups
Mouse kidney protein separation by IEF/SDS-PAGE yielded several hundred distinct protein spots (Fig. 1), ninety two of which have been identified by MALDI-TOF mass spectroscopy (Supplementary Table S1). Nine of these proteins were found to have significantly altered expression in Pon1−/− mice relative to Pon1+/+ littermates, and nine had significantly changed expression in response to high-Met diet (Table 1). The expression levels of other identified proteins were not affected by the Pon1−/− genotype or high-Met diet. Close-up views of representative IEF/SDSPAGE separations of differentially expressed proteins are shown in Fig. 2. Quantification of the level–% volume–for each of the differentially expressed protein is shown in Fig. 3. 3.3. Kidney proteins regulated by Pon1 genotype In mice fed with a standard diet, seven kidney proteins were downregulated by the Pon1−/− genotype (−1.19 to −2.00-fold, P b 0.01) and three proteins were up-regulated (1.32–1.88-fold, P b 0.01) (Table 1). The proteins down-regulated by the Pon1−/− genotype include those involved in lipoprotein metabolism (ApoA-I), energy metabolism (Ndufs8, Ldhd), carbohydrate metabolism (Pdhb, Fbp1-isoform2, Eno1), and protein metabolism (Hspd1). The three proteins that were upregulated in kidneys of Pon1−/− mice are involved in lipid transport
Fig. 2. Close-up views of representative IEF/SDS-PAGE gels showing mouse kidney proteins whose expression was affected by Pon1 genotype and/or high-Met diet. Proteomic analyses were carried out for the following groups of mice: A, E — Pon1+/+, standard diet; B, F — Pon1+/+, high-Met diet; C, D — Pon1−/−, standard diet; D, H — Pon1−/−, high-Met diet. Arrows indicate the direction of the change dependent on Pon1 genotype (C, D, G, H) and high-Met diet (B, D, F, H).
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
(Pebp1), oxidative stress response (Prdx2), and methylglyoxal detoxification (Glo1) (Table 1). 3.4. Kidney proteins regulated by hyperhomocysteinemia In Pon1+/+ mice, seven kidney proteins were down-regulated (ApoA-I, Hspd1, Ndufs8, Ldhd, Pdhb, Fbp1-isoform2, Eno1; − 1.22 to − 2.27-fold, P b 0.01) and three were up-regulated by HHcy (Pebp1, Fbp-isoform1, Prdx2; 1.23–1.78-fold, P b 0.01). The expression of these proteins was similarly affected by HHcy and the Pon1−/− genotype. However, HHcy did not affect the expression of Glo1 that was affected by the Pon1−/− genotype (Table 1). 3.5. Kidney proteins regulated by Pon1 genotype in the presence of hyperhomocysteinemia The expression of kidney Hspd1 and Prdx2, regulated by the Pon1−/− genotype in mice fed with high-Met diet was similar to their expression in animals fed with a standard diet (Table 1). In contrast, the expression of kidney ApoA-I, Pebp1, Ndufs8, Pdhb, and Eno1, altered by the Pon1−/− genotype in mice fed with a standard diet became Pon1 genotypeindependent when the animals were fed with a high-Met diet. The kidney Fbp1-isoform2 exhibited a more complex expression patterns: down-regulated in mice fed with a standard diet, but up-regulated by the Pon1−/− genotype in mice fed with a high-Met diet (Table 1). Kidney Glo1 that tended to be up-regulated in Pon1−/− mice fed with a control diet became significantly up-regulated in Pon1−/− mice fed with a highMet diet (Table 1). Western blot analysis (Supplementary Methods) was performed for kidney ApoA-I and Prdx2 to validate the IEF/SDS-PAGE results. The expression of ApoA-I was attenuated by high-Met diet both in Pon1−/− and Pon1+/+ mice (Supplementary Fig. S1A), similar to results of IEF/ SDS-PAGE analyses (Supplementary Fig. S1D). Western blots also showed that Prdx2 was elevated by high-Met diet in Pon1−/− and Pon1+/+ mice (Supplementary Fig. S1B), consistent with the changes in Prdx2 that were observed in IEF/SDS-PAGE analyses (Supplementary Fig. S1E). 4. Discussion To explain the renal toxicity of reduced Pon1 and HHcy, we studied kidney proteomes of Pon1-knockout mice and their wild type littermates. We found that several kidney metabolic pathways were regulated by the Pon1−/− genotype and/or dietary HHcy. Specifically, 1) proteins involved in lipid transport (ApoA-I, Pebp1), protein metabolism (Hspd1), carbohydrate metabolism (Pdhb, Fbp1, Eno1), energy metabolism (Ndufs8, Ldhd), oxidative stress response (Prdx2), and metabolic detoxification (Glo1) were differentially expressed in kidneys of Pon1-knockout mice; 2) the Pon1−/− genotype-regulated proteins were similarly affected by HHcy; and 3) responses of renal proteins to Pon1 deficiency or HHcy in mice mimic their responses to human renal disease (discussed below). Previous studies have linked both Pon1 and HHcy to CKD and kidney carcinoma, although the underlying mechanisms are not understood. Our present results provide evidence supporting a mechanistic link between Pon1, HHcy, and kidney diseases. First, the Pon1−/− genotype and dietary HHcy cause similar changes in the expression of kidney proteins (Table 1), which suggests that the changes in these proteins' expression in HHcy can be due to Hcy-induced Pon1-deficiency [18]. Second, the mouse kidney proteins regulated by the Pon1−/− genotype or HHcy (Table 1) are known to play important roles in kidney homeostasis and their expression is altered in human kidney disease (Table 2). Thus, our findings point to a novel role for Pon1 and suggest a mechanistic explanation for the association between HHcy, Pon1, and kidney diseases: Pon1 deficiency alters the expression of proteins involved in lipid transport, and protein, carbohydrate, and energy
203
metabolisms that are crucial for normal kidney function, while HHcy also alters the expression of the same proteins (Table 1) either independently or by reducing Pon1 gene expression [18]. Further, our findings that the Pon1−/− genotype and HHcy up-regulate the expression of the antioxidant defense protein Prdx2 (Table 1) provide an explanation for red-ox changes associated with HHcy [16] and Pon1 deficiency [4–6, 26]. The function of Pon1 in the kidney has not been explored previously. As shown in the present study, in Pon1-knockout mice fed with a control diet, seven kidney proteins were identified with decreased expression (ApoA-I, Hspd1, Ndufs8, Ldhd, Pdhb, Fbp1(2), and Eno1) and three kidney proteins were identified with increased expression (Pepb1, Prdx2, and Glo1). These findings suggest that in the absence of HHcy, Pon1 interacts with proteins involved in lipid transport (ApoA-I, Pebp1), protein (Hspd1), carbohydrate (Pdhb, Fbp1, Eno1), and energy metabolisms (Ndufs8, Ldhd), oxidative stress response (Prdx2), and glyoxal detoxification (Glo1). In kidneys of Pon1-knockout mice fed with a hyperhomocysteinemic high-Met diet, one protein was identified with decreased expression (Hspd1) and four proteins were identified with increased expression (Ldhd, Fbp1(2), Prdx2, and Glo1). Further, effects of the Pon1−/− genotype on protein expression were modified in mice fed with high-Met diet. For some proteins, such as Hspd1 and Prdx2, the regulation by the Pon1−/− genotype did not depend on high-Met diet. However, the regulation of the expression of other proteins, such as ApoA-I, Pepb1, Ndufs8, Ldhd, Pdhb, Fbp1(2), and Eno1, by the Pon1−/− genotype was altered by high-Met diet. These findings suggest that there is an interaction between the Pon1−/− genotype and HHcy that modulates kidney protein expression. Three of the differentially expressed kidney proteins, ApoA-I, Pebp1, and Prdx2, identified in the present study, are also differentially expressed in other mouse organs in response to the Pon1−/− genotype and/or high-Met diet. For example, ApoA-I is down-regulated and Prdx2 is up-regulated by the Pon1−/− genotype and a high-Met diet both in the kidney and the liver (Supplementary Table S2). Prdx2 is up-regulated by the Pon1−/− genotype in the brain, similar to the kidney. However, in contrast to the up-regulation of Prdx2 by a high-Met diet in the kidney and liver, the brain Prdx2 is down-regulated by a high-Met diet. These findings suggest that while in the kidney and liver the antioxidant defense is enhanced by a high-Met diet, in the brain the antioxidant defense is compromised. The other proteins that were identified as differentially expressed in the kidney (Table 1) were not regulated by the Pon1−/− genotype and/or high-Met diet in the liver [24] or brain [25]. Pebp1 was up-regulated in the kidney and down-regulated in the brain both by the Pon1−/− genotype and highMet diet (Supplementary Table S2). Taken together, these findings indicate that the regulation of protein expression and antioxidant defense mechanisms by the Pon1−/− genotype and high-Met diet is organ specific. The down-regulation of kidney ApoA-I in mice fed with a high-Met diet is most likely a consequence of down regulation of ApoA-I expression by HHcy in the liver. For example, Hcy is known to inhibit hepatic synthesis of ApoA-I in mice (reviewed in [21]). Plasma ApoA-I is also known to be down-regulated in a dietary HHcy mouse model [27], similar to the one used in the present work. Further, in our previous work we found that hepatic ApoA-I is down-regulated by a high-Met diet in wild type mice [24]. ApoA-I is a major interactor of the PPAR signaling pathway involved in the cardio-renal syndrome [28]. Down-regulation of ApoA-I is associated with chronic kidney disease (CKD) and atherosclerosis in humans. For example, patients with CKD exhibit significant lipoprotein derangements, including ApoA-I [29]. ApoA-I is negatively correlated with plasma Hcy, and low ApoA-I levels are a risk factor for cardiovascular disease [30]. Thus, our findings that renal ApoA-I is down-regulated by the Pon1−/− genotype suggest that Pon1 interacts with ApoA-I and has an anti-atherogenic function in the kidney. Further, our finding
204
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
205
Table 2 Kidney proteins that are regulated by Pon1 genotype and/or high-Met diet in the mouse are also differentially expressed in renal disease (see text for discussion and references). Protein name
Lipoprotein metabolism ApoA-I Protein metabolism Hspd1 Energy metabolism Ndufs8 Carbohydrate metabolism Fbp1 isoform 1 Fbp1 isoform 2 Eno1 Oxidative stress response Prdx2 a b
Change in Pon1−/− vs. Pon1+/+ kidneya
Change in 1%-Met vs. std. diet kidneya
Std. diet
1%-Met diet
Pon1+/+
↓
–
↓
↓ in CKD
↓
↓
↓
↓ in human hyperparathyroidism
↓
–
↓
(↓ in LPS-treated mouse)
– ↓ ↓
– ↑ –
↑ ↓ ↓
↓ in human RCC
↑
↑
↑
↓ in human hyperparathyroidism
Change in renal diseaseb
↑ in human RCC
The up ‘↑’ and down ‘↓’ arrows indicate up-regulated and down-regulated proteins, respectively. The dash ‘–’indicates no significant change. CKD, chronic kidney disease; RCC, renal cell carcinoma; LPS, lipopolysaccharide.
that renal ApoA-I is down-regulated by a high-Met diet suggests that elevated Hcy is pro-atherogenic in the kidney. We identified two proteins involved in the kidney energy metabolism (Ndufs8 and Ldhd) as well as three carbohydrate metabolism proteins (Pdhb, Fbp1(2), and Eno1) that were down-regulated by the Pon1−/− genotype and a high-Met diet. Because diminished expression of these proteins would compromise the ability to generate ATP by the mitochondria, our findings suggest that Pon1 plays an important role in kidney energy metabolism and that HHcy impairs the mitochondrial respiratory chain function in the kidney. Altered expression of proteins important for kidney function in response to the Pon1−/− genotype or HHcy identified in the present study is a common finding in kidney disease in humans or animal models. Specifically, the expression of kidney ApoA-I, Hspd1, Ndufs8, Fbp1, Eno1, and Prdx2 is also altered by CKD, renal cancer, or lipopolysaccharide (LPS) treatment (Table 2). For example, kidney ApoA-I that we found to be down-regulated by the Pon1−/− genotype or HHcy in mice (Table 1) is also down-regulated in human CKD [29]. Human diabetic nephropathy is also accompanied by dyslipidemia and decreased plasma ApoA-I [31]. Kidney Ndufs8, a subunit of the mitochondrial complex 1 required for catalysis of NADH oxidation in the mitochondrial respiratory chain, that we found to be down-regulated by the Pon1−/− genotype or HHcy (Table 1) is also down-regulated by the treatment of mice with LPS, which mimics bacterial infection and inflammation (Table 2) [32]. The down-regulation of renal Ndufs8 observed in our present study suggests that Pon1 deficiency or HHcy induce pro-inflammatory changes in the kidney. Hspd1 that we found to be down-regulated by the Pon1−/− genotype (Table 1) is also down-regulated in human renal secondary hyperparathyroidism, which in its late stages becomes autonomous, so that the excessive parathyroid hormone (PTH) secretion no longer responds to physiologic stimuli or to medical treatment [33]. Two proteins involved in carbohydrate metabolism, Fbp1 and Eno1, that we found to be differentially regulated by the Pon1−/− genotype or HHcy (Table 1) are known to be differentially regulated in human renal cell carcinoma (RCC): Fbp1 is down-regulated while Eno1 is upregulated [34]. The antioxidant defense protein Prdx2 that we found to be up-regulated by the Pon1−/− genotype or HHcy (Table 2) is down-regulated in human RCC [34]. The kidney is responsible for maintenance of water and sodium homeostasis. In our previous work we found that Pon1−/− mice exhibit a polyuria (urine dilution) phenotype and produce twice as much 24-h
urine as their wild type Pon1+/+ littermates [9]. Urine volume depends on aquaporin (Aqp) water channels located in epithelial cells of renal tubules. Aqp2 expression is regulated by the bile acid receptor Fxr, a transcription factor also known to regulate lipid and glucose metabolism [35]. Treatments of mice with chenodeoxycholic acid, an endogenous agonist for Fxr, lead to up-regulation of Aqp2 and also Ndufb8, a subunit of the mitochondrial membrane respiratory chain NADH dehydrogenase (complex I), and an increase in urine concentrating capacity [35]. Inactivation of the Fxr gene led to reduced Aqp2 expression and impaired urine concentrating ability, which led to a polyuria phenotype. Although in the present work we were not able to identify aquaporins among the differentially expressed kidney proteins in Pon1−/− mice, we found that kidney Ndufs8, a subunit of the mitochondrial complex 1 required for catalysis of NADH oxidation, was down-regulated by the Pon1−/− genotype (Table 2), suggesting that the down-regulation of Ndufs8 and mitochondrial respiratory chain function may contribute to polyuria. 5. Conclusion In conclusion, our findings show that excess Hcy and reduced Pon1 deregulate the expression of proteins involved in diverse cellular processes–from lipid, protein, carbohydrate, and energy metabolisms to detoxification and antioxidant defenses–that are essential for normal kidney homeostasis. Dysregulation of these processes can account for the toxicity of HHcy and reduced Pon1 observed in kidney disease. Our findings also show that Pon1 plays an important role in maintaining normal kidney homeostasis. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ymgme.2014.07.011. Conflict of interest Authors declare no interests to disclose. Acknowledgments We thank Diana Shih for kindly providing a breeding pair of Pon1knockout mice. References [1] J. Marsillach, B. Mackness, M. Mackness, F. Riu, R. Beltran, J. Joven, J. Camps, Immunohistochemical analysis of paraoxonases-1, 2, and 3 expression in normal mouse tissues, Free Radic. Biol. Med. 45 (2008) 146–157.
Fig. 3. Mouse kidney protein expression as a function of Pon1 genotype and/or high-Met diet. Each panel shows expression level–% volume–for indicated protein and groups of mice: 1 — Pon1+/+, standard diet; 2 — Pon1−/−, standard diet; 3 — Pon1+/+, high-Met diet; 4 — Pon1−/−, high-Met diet. Symbols ‘#’ and ‘*’indicate significant effects of the Pon1−/− genotype and high-Met diet (P b 0.05), respectively.
206
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
[2] B. Mackness, R. Beltran-Debon, G. Aragones, J. Joven, J. Camps, M. Mackness, Human tissue distribution of paraoxonases 1 and 2 mRNA, IUBMB Life 62 (2010) 480–482. [3] S.L. Primo-Parmo, R.C. Sorenson, J. Teiber, B.N. La Du, The human serum paraoxonase/arylesterase gene (PON1) is one member of a multigene family, Genomics 33 (1996) 498–507. [4] D.M. Shih, L. Gu, Y.R. Xia, M. Navab, W.F. Li, S. Hama, L.W. Castellani, C.E. Furlong, L.G. Costa, A.M. Fogelman, A.J. Lusis, Mice lacking serum paraoxonase are susceptible to organophosphate toxicity and atherosclerosis, Nature 394 (1998) 284–287. [5] T. Bhattacharyya, S.J. Nicholls, E.J. Topol, R.L. Zhang, X. Yang, D. Schmitt, X.M. Fu, M.Y. Shao, D.M. Brennan, S.G. Ellis, M.L. Brennan, H. Allayee, A.J. Lusis, S.L. Hazen, Relationship of paraoxonase 1 (PON1) gene polymorphisms and functional activity with systemic oxidative stress and cardiovascular risk, JAMA 299 (2008) 1265–1276. [6] J. Perla-Kajan, H. Jakubowski, Paraoxonase 1 and homocysteine metabolism, Amino Acids 43 (2012) 1405–1417. [7] T.B. Domagala, M. Lacinski, W.H. Trzeciak, B. Mackness, M.I. Mackness, H. Jakubowski, The correlation of homocysteine-thiolactonase activity of the paraoxonase (PON1) protein with coronary heart disease status, Cell. Mol. Biol. (Noisy-legrand) 52 (2006) 4–10. [8] A. Bayrak, T. Bayrak, S.L. Tokgozoglu, B. Volkan-Salanci, A. Deniz, B. Yavuz, M. Alikasifoglu, E. Demirpence, Serum PON-1 activity but not Q192R polymorphism is related to the extent of atherosclerosis, J. Atheroscler. Thromb. 19 (2012) 376–384. [9] K. Borowczyk, D.M. Shih, H. Jakubowski, Metabolism and neurotoxicity of homocysteine thiolactone in mice: evidence for a protective role of paraoxonase 1, J. Alzheimers Dis. 30 (2012) 225–231. [10] A. Gugliucci, K. Kotani, S. Kimura, Paraoxonase 1 in chronic kidney failure, J. Lipid 2012 (2012) 726048. [11] D.J. Kennedy, W.H. Tang, Y. Fan, Y. Wu, S. Mann, M. Pepoy, S.L. Hazen, Diminished antioxidant activity of high-density lipoprotein-associated proteins in chronic kidney disease, J. Am. Heart Assoc. 2 (2013) e000104. [12] A.F. Perna, E. Satta, F. Acanfora, C. Lombardi, D. Ingrosso, N.G. De Santo, Increased plasma protein homocysteinylation in hemodialysis patients, Kidney Int. 69 (2006) 869–876. [13] M.H. Shishehbor, L.P. Oliveira, M.S. Lauer, D.L. Sprecher, K. Wolski, L. Cho, B.J. Hoogwerf, S.L. Hazen, Emerging cardiovascular risk factors that account for a significant portion of attributable mortality risk in chronic kidney disease, Am. J. Cardiol. 101 (2008) 1741–1746. [14] M.C. Chao, S.L. Hu, H.S. Hsu, L.E. Davidson, C.H. Lin, C.I. Li, C.S. Liu, T.C. Li, C.C. Lin, W.Y. Lin, Serum homocysteine level is positively associated with chronic kidney disease in a Taiwan Chinese population, J. Nephrol. (2014) [Epub ahead of print]. [15] H. Jakubowski, Quality control in tRNA charging — editing of homocysteine, Acta Biochim. Pol. 58 (2011) 149–163. [16] B.A. Maron, J. Loscalzo, The treatment of hyperhomocysteinemia, Annu. Rev. Med. 60 (2009) 39–54. [17] M. Lacinski, W. Skorupski, A. Cieslinski, J. Sokolowska, W.H. Trzeciak, H. Jakubowski, Determinants of homocysteine-thiolactonase activity of the paraoxonase-1 (PON1) protein in humans, Cell. Mol. Biol. (Noisy-le-grand) 50 (2004) 885–893. [18] H. Jiang, S.P. Stabler, R.H. Allen, K.N. Maclean, Altered expression of apoA-I, apoA-IV and PON-1 activity in CBS deficient homocystinuria in the presence and absence of treatment: possible implications for cardiovascular outcomes, Mol. Genet. Metab. 107 (2012) 55–65. [19] H. Jakubowski, Calcium-dependent human serum homocysteine thiolactone hydrolase. A protective mechanism against protein N-homocysteinylation, J. Biol. Chem. 275 (2000) 3957–3962.
[20] J. Perla-Kajan, H. Jakubowski, Paraoxonase 1 protects against protein Nhomocysteinylation in humans, FASEB J. 24 (2010) 931–936. [21] H. Jakubowski, Homocysteine in Protein Structure/Function and Human Disease — Chemical Biology of Homocysteine-containing Proteins, Springer, Wien, New York, 2013. [22] H. Jakubowski, G.H. Boers, K.A. Strauss, Mutations in cystathionine {beta}-synthase or methylenetetrahydrofolate reductase gene increase N-homocysteinylated protein levels in humans, FASEB J. 22 (2008) 4071–4076. [23] J. Suszynska-Zajczyk, O. Utyro, H. Jakubowski, Methionine-induced hyperhomocysteinemia and bleomycin hydrolase deficiency alter the expression of mouse kidney proteins involved in renal disease, Mol. Genet. Metab. 112 (2014) 339–346. [24] J. Suszynska-Zajczyk, J. Wroblewski, O. Utyro, M. Luczak, L. Marczak, H. Jakubowski, Bleomycin hydrolase and hyperhomocysteinemia modulate the expression of mouse proteins involved in liver homeostasis, Amino Acids 46 (2014) 1471–1480. [25] J. Suszynska-Zajczyk, M. Luczak, L. Marczak, H. Jakubowski, Hyperhomocysteinemia and bleomycin hydrolase modulate the expression of mouse brain proteins involved in neurodegeneration, J. Alzheimers Dis. 40 (2014) 713–726. [26] C. Noll, J. Dairou, C. Ripoll, J.L. Paul, J.M. Dupret, J.M. Delabar, F. Rodrigues-Lima, N. Janel, Effect of red wine polyphenol dietary supplementation on two phase II enzymes in liver of hyperhomocysteinemic mice, Food Chem Toxicol. 49 (2011) 1764–1769. [27] J. Julve, J.C. Escola-Gil, E. Rodriguez-Millan, J.M. Martin-Campos, M. Jauhiainen, H. Quesada, I.M. Renteria-Obregon, J. Osada, J.L. Sanchez-Quesada, F. Blanco-Vaca, Methionine-induced hyperhomocysteinemia impairs the antioxidant ability of high-density lipoproteins without reducing in vivo macrophage-specific reverse cholesterol transport, Mol. Nutr. Food Res. 57 (2013) 1814–1824. [28] I. Muhlberger, K. Monks, R. Fechete, G. Mayer, R. Oberbauer, B. Mayer, P. Perco, Molecular pathways and crosstalk characterizing the cardiorenal syndrome, OMICS 16 (2012) 105–112. [29] K. Farbakhsh, B.L. Kasiske, Dyslipidemias in patients who have chronic kidney disease, Med. Clin. North Am. 89 (2005) 689–699. [30] R.M. Gueant-Rodriguez, R. Spada, M. Moreno-Garcia, G. Anello, P. Bosco, L. Lagrost, A. Romano, M. Elia, J.L. Gueant, Homocysteine is a determinant of ApoA-I and both are associated with ankle brachial index, in an ambulatory elderly population, Atherosclerosis 214 (2011) 480–485. [31] P.O. Attman, C. Knight-Gibson, M. Tavella, O. Samuelsson, P. Alaupovic, The compositional abnormalities of lipoproteins in diabetic renal failure, Nephrol. Dial. Transplant. 13 (1998) 2833–2841. [32] K.R. Feingold, Y. Wang, A. Moser, J.K. Shigenaga, C. Grunfeld, LPS decreases fatty acid oxidation and nuclear hormone receptors in the kidney, J. Lipid Res. 49 (2008) 2179–2187. [33] I. Santamaria, D. Alvarez-Hernandez, R. Jofre, J.R. Polo, J. Menarguez, J.B. CannataAndia, Progression of secondary hyperparathyroidism involves deregulation of genes related to DNA and RNA stability, Kidney Int. 67 (2005) 2267–2279. [34] N.M. White, O. Masui, L.V. Desouza, O. Krakovska, S. Metias, A.D. Romaschin, R.J. Honey, R. Stewart, K. Pace, J. Lee, M.A. Jewett, G.A. Bjarnason, K.W. Siu, G.M. Yousef, Quantitative proteomic analysis reveals potential diagnostic markers and pathways involved in pathogenesis of renal cell carcinoma, Oncotarget 5 (2014) 506–518. [35] X. Zhang, S. Huang, M. Gao, J. Liu, X. Jia, Q. Han, S. Zheng, Y. Miao, S. Li, H. Weng, X. Xia, S. Du, W. Wu, J.A. Gustafsson, Y. Guan, Farnesoid X receptor (FXR) gene deficiency impairs urine concentration in mice, Proc. Natl. Acad. Sci. U. S. A. 111 (2014) 2277–2282.
Contents lists available at ScienceDirect
Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme
Paraoxonase 1 deficiency and hyperhomocysteinemia alter the expression of mouse kidney proteins involved in renal disease☆ Joanna Suszyńska-Zajczyk a, Marta Sikora a, Hieronim Jakubowski a,b,c,⁎ a b c
Institute of Bioorganic Chemistry, Poznań, Poland Department of Biochemistry and Biotechnology, University of Life Sciences, Poznań, Poland Department of Microbiology & Molecular Genetics, Rutgers-New Jersey Medical School, International Center for Public Health, Newark, NJ, USA
a r t i c l e
i n f o
Article history: Received 4 June 2014 Received in revised form 9 July 2014 Accepted 10 July 2014 Available online 17 July 2014 Keywords: Paraoxonase 1 Polyuria High-methionine diet Hyperhomocysteinemia Mouse kidney proteome Kidney disease
a b s t r a c t Scope: Hyperhomocysteinemia (HHcy) is associated with kidney disease and leads to atherosclerosis and thrombosis. Paraoxonase 1 (Pon1), a hydrolase that participates in homocysteine (Hcy) metabolism and is carried in the circulation on high-density lipoprotein, has also been linked to kidney disease and atherothrombosis. Pon1-knockout mice are susceptible to atherosclerosis and exhibit a kidney-associated phenotype, polyuria or urine dilution. We hypothesize that HHcy and Pon1 deficiency are toxic to kidney function because they impair metabolic pathways important for normal kidney homeostasis. Methods and results: We examined changes in the mouse kidney proteome induced by Pon1 gene deletion and dietary HHcy, using 2D IEF/SDS-PAGE gel electrophoresis and MALDI-TOF mass spectrometry. We found that the expression of ten mouse kidney proteins was altered by the Pon1−/− genotype or HHcy. Proteins involved in metabolism of lipid (ApoA-I), protein (Hspd1), carbohydrate (Pdhb, Fbp1-isoform2, Eno1), and energy (Ndufs8, Ldhd) were down-regulated. Proteins involved in lipid transport (Pebp1), oxidative stress response (Prdx2), and cellular detoxification (Glo1) were up-regulated. The kidney proteins altered by HHcy or Pon1 are also altered in renal disease. Conclusion: Our findings suggest that excess Hcy is toxic because it deregulates the expression of proteins involved in diverse cellular processes–from lipid, protein, carbohydrate, and energy metabolisms to detoxification and antioxidant defenses–that are essential for normal kidney homeostasis. Dysregulation of these processes can account for the involvement of HHcy and reduced Pon1 in kidney disease. Our findings also show that Pon1 plays an important role in maintaining normal kidney homeostasis. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Paraoxonase 1 (PON1) is expressed in many tissues, including the kidney, liver, lung, brain, and colon (both PON1 mRNA and protein are found in those organs), circulates in the blood attached to highdensity lipoproteins (HDL), and is distributed to all tissues both in humans and rodents [1–3]. PON1 has a cardio-protective function in mice [4] and humans [5], which has been linked to its ability to mediate detoxification of oxidized lipids and homocysteine (Hcy)-thiolactone [6]. Reduced PON1 activity is linked to cardiovascular disease [7,8]. In
☆ This work was supported in part by grants from the American Heart Association (12GRNT9420014), the National Science Center (2011/01/B/NZ1/03417, 2011/02/A/ NZ1/00010, 2012/07/B/NZ7/01178, and 2013/09/B/NZ5/02794) and the MNiSW, Poland (N401 065321504 and N N302 434439). ⁎ Corresponding author at: Department of Microbiology, Biochemistry & Molecular Genetics, Rutgers-New Jersey Medical School, 225 Warren Street, Newark, NJ 07101-1709, USA. Fax: +1 973 972 8982. E-mail address: [email protected] (H. Jakubowski).
http://dx.doi.org/10.1016/j.ymgme.2014.07.011 1096-7192/© 2014 Elsevier Inc. All rights reserved.
mice, Pon1 gene deletion impairs urine concentrating ability, which leads to a polyuria phenotype [9]. Pon1 activity is reduced in patients with chronic kidney disease (CKD) [10] and predicts risk of heart attack, stroke, and death [11]. CKD patients also have elevated plasma Hcy [12], an emerging cardiovascular risk factor. Plasma Hcy is negatively correlated with glomerular filtration rate, positively with mortality risk, and accounts for a significant portion of morality in CKD [13,14]. However, mechanisms underlying the toxicity of elevated Hcy and reduced Pon1 in the kidney are not understood [15, 16]. Pon1 links HDL and Hcy metabolisms, which may account for their role in kidney disease. Hcy is a negative determinant of HDL and PON1 [17] and reduces Apoa1 and Pon1 gene expression [18]. HDL and purified PON1 detoxify Hcy-thiolactone [19] and protect against protein Nhomocysteinylation [9,20]. N-Homocysteinylation causes protein damage and is linked to kidney and cardiovascular diseases (reviewed in [21]). Atherosclerosis-related (N-homocysteinylated ApoA-I or N-Hcy-ApoA-I) and other N-Hcy-proteins [22] increase in hyperhomocysteinemia (HHcy), including CKD patients [12] and kidneys of Pcft−/− mice [21].
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
201
Fig. 1. Representative IEF/SDS-PAGE gel showing the kidney proteome of a wild type C57BL/6J mouse. Left to right: IEF-pH gradient pH 4 → pH 7. Top to bottom: SDS-PAGE; molecular weight markers, 10–260 kDa, shown on the left. Numbers indicate protein spots whose identity was established by mass spectrometry. White rectangles outline areas containing proteins whose expression was affected by the Pon1 genotype and/or high-Met diet. Identities of the numbered protein spots are described in Supplementary Table S1.
The Pon1-null mice are more susceptible to Hcy-thiolactone toxicity than wild type mice and exhibit a polyuria phenotype: produce twice as much urine as their wild-type littermates [9]. Taken together, these findings suggest that Pon1 plays important roles in the kidney.
We hypothesize that HHcy and Pon1 deficiency are toxic because they impair metabolic pathways important for kidney homeostasis. To test this hypothesis, we examined changes in the mouse kidney proteome induced by dietary HHcy in wild type and Pon1-null mice.
Table 1 Differentially expressed kidney proteins regulated by Pon1−/− genotype and/or high-Met diet. Protein description (spot #)
Lipoprotein and lipid metabolism Apolipoprotein A-I (#78) Phosphatidylethanolamine-binding protein 1 (#86) Amino acid and protein metabolism 60 kDa heat shock protein, mitochondrial (#14) Energy metabolism NADH dehydrogenase [ubiquinone] iron–sulfur protein 8, mitochondrial (#72) Probable D-lactate dehydrogenase, mitochondrial (#18) Carbohydrate metabolism Pyruvate dehydrogenase E1 component subunit beta, mitochondrial (#47) Fructose-1,6-bisphosphatase 1 isoform 2 (#62) Fructose-1,6-bisphosphatase 1 isoform 1 (#63) Alpha enolase (#34) Oxidative stress response Peroxiredoxin 2 (#85) Methylglyoxal detoxification Glyoxalase 1 (#87) Spot # refers to the numbering on the IEF/SDS-PAGE gels in Fig. 1 and Fig. 2. Significantly different: aP b 0.001, bP b 0.01, cP b 0.05.
Gene name
Fold change Pon1−/− vs. Pon1+/+
Fold change std vs. 1% Met diet
Std diet
Pon1+/+
Pon1−/−
−2.27a 1.23
−1.37a 1.02
1%-Met diet
Apoa1 Pebp1
−1.47a 1.32b
Hspd1
−1.44a
−1.22b
−1.22b
1.04
Ndufs8 Ldhd
−1.96a −1.27
1.07 1.77a
−2.00a −2.00a
1.05 1.11
Pdhb Fbp1 Fbp1 Eno1
−1.19b −2.00a 1.01 −2.00a
−1.02 1.49b −1.01 1.17
−1.44a −2.27a 1.25c −1.59a
−1.23a 1.29c 1.23 1.48a
Prdx2
1.88a
1.21a
1.78a
1.14a
Glo1
1.17
1.45b
1.08
1.15
1.11 1.09
202
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
2. Methods
were analyzed by ANOVA. Unpaired Student's t-test was used to test differences between two groups. Statistica 8.0 software was used.
2.1. Mice and diets 3. Results Pon1−/− mice [4] and wild type Pon1+/+ littermates (C57BL/6J) were maintained at the Rutgers-New Jersey Medical School Animal Facility. Female mice (n = 8) were fed a standard rodent chow containing 0.66% Met (LabDiet 5010, Purina Mills International, St. Louis, MO). At 4 weeks of age, half of Pon1−/− and Pon1+/+ mice were provided 1%Met in drinking water (high-Met diet) for 8 weeks [9]. Four experimental groups of mice were studied: 1) Pon1−/−, control diet; 2) Pon1+/+, control diet; 3) Pon1−/−, high-Met diet; and 4) Pon1+/+, high-Met diet. Drinking water supplementation with 1%-Met did not affect animals' water intake and body weight. Animal procedures were approved by the Institutional Animal Care and Use Committee.
3.1. Dietary hyperhomocysteinemia To induce HHcy, mice were fed with a high-Met diet for 8 weeks [23]. HHcy was confirmed by plasma tHcy and N-Hcy-protein assays. Plasma tHcy levels in Pon1−/− and Pon1+/+ mice fed a standard diet were 8.5 ± 1.9 μM and 7.4 ± 2.2 μM, and increased to 48 ± 16 μM and 77 ± 45 μM, respectively, in animals fed with a high-Met diet. Plasma N-Hcy-protein levels increased from basal levels of 1.4 ± 0.5 μM and 1.2 ± 0.4 μM in mice fed with a standard diet to 3.8 ± 1.8 μM and 5.4 ± 2.9 μM in hyperhomocysteinemic Pon1−/− and Pon1+/+ animals that were fed with a high-Met diet, respectively.
2.2. Genotyping 3.2. Proteins differentially expressed in Pon1−/− kidney Genotyping of the Pon1 locus [4] was carried out as described in the Supplementary Methods [23]. The Pon1−/− genotype was confirmed by the absence of enzymatic activities of Pon1 (Hcy-thiolactonase, paraoxonase, arylesterase) in serum [19,20]. 2.3. Hcy assays Total Hcy and N-Hcy-protein were assayed as previously described [22]. 2.4. Proteomic analyses Protein sample preparation, 2D-IEF/SDS-PAGE analyses, and protein identification by MALDI-TOF mass spectrometry [23–25] were carried out as described in the Supplementary Methods. 2.5. Data treatment For each animal in the four experimental group (4 animals/group), analyses were repeated 2–3 times. Relative abundances of protein spots–% volume–were calculated by dividing their volume by the total volume of all spots. Data are expressed as mean ± SD. Data for each protein spot had a normal distribution. The differences between the groups
Mouse kidney protein separation by IEF/SDS-PAGE yielded several hundred distinct protein spots (Fig. 1), ninety two of which have been identified by MALDI-TOF mass spectroscopy (Supplementary Table S1). Nine of these proteins were found to have significantly altered expression in Pon1−/− mice relative to Pon1+/+ littermates, and nine had significantly changed expression in response to high-Met diet (Table 1). The expression levels of other identified proteins were not affected by the Pon1−/− genotype or high-Met diet. Close-up views of representative IEF/SDSPAGE separations of differentially expressed proteins are shown in Fig. 2. Quantification of the level–% volume–for each of the differentially expressed protein is shown in Fig. 3. 3.3. Kidney proteins regulated by Pon1 genotype In mice fed with a standard diet, seven kidney proteins were downregulated by the Pon1−/− genotype (−1.19 to −2.00-fold, P b 0.01) and three proteins were up-regulated (1.32–1.88-fold, P b 0.01) (Table 1). The proteins down-regulated by the Pon1−/− genotype include those involved in lipoprotein metabolism (ApoA-I), energy metabolism (Ndufs8, Ldhd), carbohydrate metabolism (Pdhb, Fbp1-isoform2, Eno1), and protein metabolism (Hspd1). The three proteins that were upregulated in kidneys of Pon1−/− mice are involved in lipid transport
Fig. 2. Close-up views of representative IEF/SDS-PAGE gels showing mouse kidney proteins whose expression was affected by Pon1 genotype and/or high-Met diet. Proteomic analyses were carried out for the following groups of mice: A, E — Pon1+/+, standard diet; B, F — Pon1+/+, high-Met diet; C, D — Pon1−/−, standard diet; D, H — Pon1−/−, high-Met diet. Arrows indicate the direction of the change dependent on Pon1 genotype (C, D, G, H) and high-Met diet (B, D, F, H).
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
(Pebp1), oxidative stress response (Prdx2), and methylglyoxal detoxification (Glo1) (Table 1). 3.4. Kidney proteins regulated by hyperhomocysteinemia In Pon1+/+ mice, seven kidney proteins were down-regulated (ApoA-I, Hspd1, Ndufs8, Ldhd, Pdhb, Fbp1-isoform2, Eno1; − 1.22 to − 2.27-fold, P b 0.01) and three were up-regulated by HHcy (Pebp1, Fbp-isoform1, Prdx2; 1.23–1.78-fold, P b 0.01). The expression of these proteins was similarly affected by HHcy and the Pon1−/− genotype. However, HHcy did not affect the expression of Glo1 that was affected by the Pon1−/− genotype (Table 1). 3.5. Kidney proteins regulated by Pon1 genotype in the presence of hyperhomocysteinemia The expression of kidney Hspd1 and Prdx2, regulated by the Pon1−/− genotype in mice fed with high-Met diet was similar to their expression in animals fed with a standard diet (Table 1). In contrast, the expression of kidney ApoA-I, Pebp1, Ndufs8, Pdhb, and Eno1, altered by the Pon1−/− genotype in mice fed with a standard diet became Pon1 genotypeindependent when the animals were fed with a high-Met diet. The kidney Fbp1-isoform2 exhibited a more complex expression patterns: down-regulated in mice fed with a standard diet, but up-regulated by the Pon1−/− genotype in mice fed with a high-Met diet (Table 1). Kidney Glo1 that tended to be up-regulated in Pon1−/− mice fed with a control diet became significantly up-regulated in Pon1−/− mice fed with a highMet diet (Table 1). Western blot analysis (Supplementary Methods) was performed for kidney ApoA-I and Prdx2 to validate the IEF/SDS-PAGE results. The expression of ApoA-I was attenuated by high-Met diet both in Pon1−/− and Pon1+/+ mice (Supplementary Fig. S1A), similar to results of IEF/ SDS-PAGE analyses (Supplementary Fig. S1D). Western blots also showed that Prdx2 was elevated by high-Met diet in Pon1−/− and Pon1+/+ mice (Supplementary Fig. S1B), consistent with the changes in Prdx2 that were observed in IEF/SDS-PAGE analyses (Supplementary Fig. S1E). 4. Discussion To explain the renal toxicity of reduced Pon1 and HHcy, we studied kidney proteomes of Pon1-knockout mice and their wild type littermates. We found that several kidney metabolic pathways were regulated by the Pon1−/− genotype and/or dietary HHcy. Specifically, 1) proteins involved in lipid transport (ApoA-I, Pebp1), protein metabolism (Hspd1), carbohydrate metabolism (Pdhb, Fbp1, Eno1), energy metabolism (Ndufs8, Ldhd), oxidative stress response (Prdx2), and metabolic detoxification (Glo1) were differentially expressed in kidneys of Pon1-knockout mice; 2) the Pon1−/− genotype-regulated proteins were similarly affected by HHcy; and 3) responses of renal proteins to Pon1 deficiency or HHcy in mice mimic their responses to human renal disease (discussed below). Previous studies have linked both Pon1 and HHcy to CKD and kidney carcinoma, although the underlying mechanisms are not understood. Our present results provide evidence supporting a mechanistic link between Pon1, HHcy, and kidney diseases. First, the Pon1−/− genotype and dietary HHcy cause similar changes in the expression of kidney proteins (Table 1), which suggests that the changes in these proteins' expression in HHcy can be due to Hcy-induced Pon1-deficiency [18]. Second, the mouse kidney proteins regulated by the Pon1−/− genotype or HHcy (Table 1) are known to play important roles in kidney homeostasis and their expression is altered in human kidney disease (Table 2). Thus, our findings point to a novel role for Pon1 and suggest a mechanistic explanation for the association between HHcy, Pon1, and kidney diseases: Pon1 deficiency alters the expression of proteins involved in lipid transport, and protein, carbohydrate, and energy
203
metabolisms that are crucial for normal kidney function, while HHcy also alters the expression of the same proteins (Table 1) either independently or by reducing Pon1 gene expression [18]. Further, our findings that the Pon1−/− genotype and HHcy up-regulate the expression of the antioxidant defense protein Prdx2 (Table 1) provide an explanation for red-ox changes associated with HHcy [16] and Pon1 deficiency [4–6, 26]. The function of Pon1 in the kidney has not been explored previously. As shown in the present study, in Pon1-knockout mice fed with a control diet, seven kidney proteins were identified with decreased expression (ApoA-I, Hspd1, Ndufs8, Ldhd, Pdhb, Fbp1(2), and Eno1) and three kidney proteins were identified with increased expression (Pepb1, Prdx2, and Glo1). These findings suggest that in the absence of HHcy, Pon1 interacts with proteins involved in lipid transport (ApoA-I, Pebp1), protein (Hspd1), carbohydrate (Pdhb, Fbp1, Eno1), and energy metabolisms (Ndufs8, Ldhd), oxidative stress response (Prdx2), and glyoxal detoxification (Glo1). In kidneys of Pon1-knockout mice fed with a hyperhomocysteinemic high-Met diet, one protein was identified with decreased expression (Hspd1) and four proteins were identified with increased expression (Ldhd, Fbp1(2), Prdx2, and Glo1). Further, effects of the Pon1−/− genotype on protein expression were modified in mice fed with high-Met diet. For some proteins, such as Hspd1 and Prdx2, the regulation by the Pon1−/− genotype did not depend on high-Met diet. However, the regulation of the expression of other proteins, such as ApoA-I, Pepb1, Ndufs8, Ldhd, Pdhb, Fbp1(2), and Eno1, by the Pon1−/− genotype was altered by high-Met diet. These findings suggest that there is an interaction between the Pon1−/− genotype and HHcy that modulates kidney protein expression. Three of the differentially expressed kidney proteins, ApoA-I, Pebp1, and Prdx2, identified in the present study, are also differentially expressed in other mouse organs in response to the Pon1−/− genotype and/or high-Met diet. For example, ApoA-I is down-regulated and Prdx2 is up-regulated by the Pon1−/− genotype and a high-Met diet both in the kidney and the liver (Supplementary Table S2). Prdx2 is up-regulated by the Pon1−/− genotype in the brain, similar to the kidney. However, in contrast to the up-regulation of Prdx2 by a high-Met diet in the kidney and liver, the brain Prdx2 is down-regulated by a high-Met diet. These findings suggest that while in the kidney and liver the antioxidant defense is enhanced by a high-Met diet, in the brain the antioxidant defense is compromised. The other proteins that were identified as differentially expressed in the kidney (Table 1) were not regulated by the Pon1−/− genotype and/or high-Met diet in the liver [24] or brain [25]. Pebp1 was up-regulated in the kidney and down-regulated in the brain both by the Pon1−/− genotype and highMet diet (Supplementary Table S2). Taken together, these findings indicate that the regulation of protein expression and antioxidant defense mechanisms by the Pon1−/− genotype and high-Met diet is organ specific. The down-regulation of kidney ApoA-I in mice fed with a high-Met diet is most likely a consequence of down regulation of ApoA-I expression by HHcy in the liver. For example, Hcy is known to inhibit hepatic synthesis of ApoA-I in mice (reviewed in [21]). Plasma ApoA-I is also known to be down-regulated in a dietary HHcy mouse model [27], similar to the one used in the present work. Further, in our previous work we found that hepatic ApoA-I is down-regulated by a high-Met diet in wild type mice [24]. ApoA-I is a major interactor of the PPAR signaling pathway involved in the cardio-renal syndrome [28]. Down-regulation of ApoA-I is associated with chronic kidney disease (CKD) and atherosclerosis in humans. For example, patients with CKD exhibit significant lipoprotein derangements, including ApoA-I [29]. ApoA-I is negatively correlated with plasma Hcy, and low ApoA-I levels are a risk factor for cardiovascular disease [30]. Thus, our findings that renal ApoA-I is down-regulated by the Pon1−/− genotype suggest that Pon1 interacts with ApoA-I and has an anti-atherogenic function in the kidney. Further, our finding
204
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
205
Table 2 Kidney proteins that are regulated by Pon1 genotype and/or high-Met diet in the mouse are also differentially expressed in renal disease (see text for discussion and references). Protein name
Lipoprotein metabolism ApoA-I Protein metabolism Hspd1 Energy metabolism Ndufs8 Carbohydrate metabolism Fbp1 isoform 1 Fbp1 isoform 2 Eno1 Oxidative stress response Prdx2 a b
Change in Pon1−/− vs. Pon1+/+ kidneya
Change in 1%-Met vs. std. diet kidneya
Std. diet
1%-Met diet
Pon1+/+
↓
–
↓
↓ in CKD
↓
↓
↓
↓ in human hyperparathyroidism
↓
–
↓
(↓ in LPS-treated mouse)
– ↓ ↓
– ↑ –
↑ ↓ ↓
↓ in human RCC
↑
↑
↑
↓ in human hyperparathyroidism
Change in renal diseaseb
↑ in human RCC
The up ‘↑’ and down ‘↓’ arrows indicate up-regulated and down-regulated proteins, respectively. The dash ‘–’indicates no significant change. CKD, chronic kidney disease; RCC, renal cell carcinoma; LPS, lipopolysaccharide.
that renal ApoA-I is down-regulated by a high-Met diet suggests that elevated Hcy is pro-atherogenic in the kidney. We identified two proteins involved in the kidney energy metabolism (Ndufs8 and Ldhd) as well as three carbohydrate metabolism proteins (Pdhb, Fbp1(2), and Eno1) that were down-regulated by the Pon1−/− genotype and a high-Met diet. Because diminished expression of these proteins would compromise the ability to generate ATP by the mitochondria, our findings suggest that Pon1 plays an important role in kidney energy metabolism and that HHcy impairs the mitochondrial respiratory chain function in the kidney. Altered expression of proteins important for kidney function in response to the Pon1−/− genotype or HHcy identified in the present study is a common finding in kidney disease in humans or animal models. Specifically, the expression of kidney ApoA-I, Hspd1, Ndufs8, Fbp1, Eno1, and Prdx2 is also altered by CKD, renal cancer, or lipopolysaccharide (LPS) treatment (Table 2). For example, kidney ApoA-I that we found to be down-regulated by the Pon1−/− genotype or HHcy in mice (Table 1) is also down-regulated in human CKD [29]. Human diabetic nephropathy is also accompanied by dyslipidemia and decreased plasma ApoA-I [31]. Kidney Ndufs8, a subunit of the mitochondrial complex 1 required for catalysis of NADH oxidation in the mitochondrial respiratory chain, that we found to be down-regulated by the Pon1−/− genotype or HHcy (Table 1) is also down-regulated by the treatment of mice with LPS, which mimics bacterial infection and inflammation (Table 2) [32]. The down-regulation of renal Ndufs8 observed in our present study suggests that Pon1 deficiency or HHcy induce pro-inflammatory changes in the kidney. Hspd1 that we found to be down-regulated by the Pon1−/− genotype (Table 1) is also down-regulated in human renal secondary hyperparathyroidism, which in its late stages becomes autonomous, so that the excessive parathyroid hormone (PTH) secretion no longer responds to physiologic stimuli or to medical treatment [33]. Two proteins involved in carbohydrate metabolism, Fbp1 and Eno1, that we found to be differentially regulated by the Pon1−/− genotype or HHcy (Table 1) are known to be differentially regulated in human renal cell carcinoma (RCC): Fbp1 is down-regulated while Eno1 is upregulated [34]. The antioxidant defense protein Prdx2 that we found to be up-regulated by the Pon1−/− genotype or HHcy (Table 2) is down-regulated in human RCC [34]. The kidney is responsible for maintenance of water and sodium homeostasis. In our previous work we found that Pon1−/− mice exhibit a polyuria (urine dilution) phenotype and produce twice as much 24-h
urine as their wild type Pon1+/+ littermates [9]. Urine volume depends on aquaporin (Aqp) water channels located in epithelial cells of renal tubules. Aqp2 expression is regulated by the bile acid receptor Fxr, a transcription factor also known to regulate lipid and glucose metabolism [35]. Treatments of mice with chenodeoxycholic acid, an endogenous agonist for Fxr, lead to up-regulation of Aqp2 and also Ndufb8, a subunit of the mitochondrial membrane respiratory chain NADH dehydrogenase (complex I), and an increase in urine concentrating capacity [35]. Inactivation of the Fxr gene led to reduced Aqp2 expression and impaired urine concentrating ability, which led to a polyuria phenotype. Although in the present work we were not able to identify aquaporins among the differentially expressed kidney proteins in Pon1−/− mice, we found that kidney Ndufs8, a subunit of the mitochondrial complex 1 required for catalysis of NADH oxidation, was down-regulated by the Pon1−/− genotype (Table 2), suggesting that the down-regulation of Ndufs8 and mitochondrial respiratory chain function may contribute to polyuria. 5. Conclusion In conclusion, our findings show that excess Hcy and reduced Pon1 deregulate the expression of proteins involved in diverse cellular processes–from lipid, protein, carbohydrate, and energy metabolisms to detoxification and antioxidant defenses–that are essential for normal kidney homeostasis. Dysregulation of these processes can account for the toxicity of HHcy and reduced Pon1 observed in kidney disease. Our findings also show that Pon1 plays an important role in maintaining normal kidney homeostasis. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ymgme.2014.07.011. Conflict of interest Authors declare no interests to disclose. Acknowledgments We thank Diana Shih for kindly providing a breeding pair of Pon1knockout mice. References [1] J. Marsillach, B. Mackness, M. Mackness, F. Riu, R. Beltran, J. Joven, J. Camps, Immunohistochemical analysis of paraoxonases-1, 2, and 3 expression in normal mouse tissues, Free Radic. Biol. Med. 45 (2008) 146–157.
Fig. 3. Mouse kidney protein expression as a function of Pon1 genotype and/or high-Met diet. Each panel shows expression level–% volume–for indicated protein and groups of mice: 1 — Pon1+/+, standard diet; 2 — Pon1−/−, standard diet; 3 — Pon1+/+, high-Met diet; 4 — Pon1−/−, high-Met diet. Symbols ‘#’ and ‘*’indicate significant effects of the Pon1−/− genotype and high-Met diet (P b 0.05), respectively.
206
J. Suszyńska-Zajczyk et al. / Molecular Genetics and Metabolism 113 (2014) 200–206
[2] B. Mackness, R. Beltran-Debon, G. Aragones, J. Joven, J. Camps, M. Mackness, Human tissue distribution of paraoxonases 1 and 2 mRNA, IUBMB Life 62 (2010) 480–482. [3] S.L. Primo-Parmo, R.C. Sorenson, J. Teiber, B.N. La Du, The human serum paraoxonase/arylesterase gene (PON1) is one member of a multigene family, Genomics 33 (1996) 498–507. [4] D.M. Shih, L. Gu, Y.R. Xia, M. Navab, W.F. Li, S. Hama, L.W. Castellani, C.E. Furlong, L.G. Costa, A.M. Fogelman, A.J. Lusis, Mice lacking serum paraoxonase are susceptible to organophosphate toxicity and atherosclerosis, Nature 394 (1998) 284–287. [5] T. Bhattacharyya, S.J. Nicholls, E.J. Topol, R.L. Zhang, X. Yang, D. Schmitt, X.M. Fu, M.Y. Shao, D.M. Brennan, S.G. Ellis, M.L. Brennan, H. Allayee, A.J. Lusis, S.L. Hazen, Relationship of paraoxonase 1 (PON1) gene polymorphisms and functional activity with systemic oxidative stress and cardiovascular risk, JAMA 299 (2008) 1265–1276. [6] J. Perla-Kajan, H. Jakubowski, Paraoxonase 1 and homocysteine metabolism, Amino Acids 43 (2012) 1405–1417. [7] T.B. Domagala, M. Lacinski, W.H. Trzeciak, B. Mackness, M.I. Mackness, H. Jakubowski, The correlation of homocysteine-thiolactonase activity of the paraoxonase (PON1) protein with coronary heart disease status, Cell. Mol. Biol. (Noisy-legrand) 52 (2006) 4–10. [8] A. Bayrak, T. Bayrak, S.L. Tokgozoglu, B. Volkan-Salanci, A. Deniz, B. Yavuz, M. Alikasifoglu, E. Demirpence, Serum PON-1 activity but not Q192R polymorphism is related to the extent of atherosclerosis, J. Atheroscler. Thromb. 19 (2012) 376–384. [9] K. Borowczyk, D.M. Shih, H. Jakubowski, Metabolism and neurotoxicity of homocysteine thiolactone in mice: evidence for a protective role of paraoxonase 1, J. Alzheimers Dis. 30 (2012) 225–231. [10] A. Gugliucci, K. Kotani, S. Kimura, Paraoxonase 1 in chronic kidney failure, J. Lipid 2012 (2012) 726048. [11] D.J. Kennedy, W.H. Tang, Y. Fan, Y. Wu, S. Mann, M. Pepoy, S.L. Hazen, Diminished antioxidant activity of high-density lipoprotein-associated proteins in chronic kidney disease, J. Am. Heart Assoc. 2 (2013) e000104. [12] A.F. Perna, E. Satta, F. Acanfora, C. Lombardi, D. Ingrosso, N.G. De Santo, Increased plasma protein homocysteinylation in hemodialysis patients, Kidney Int. 69 (2006) 869–876. [13] M.H. Shishehbor, L.P. Oliveira, M.S. Lauer, D.L. Sprecher, K. Wolski, L. Cho, B.J. Hoogwerf, S.L. Hazen, Emerging cardiovascular risk factors that account for a significant portion of attributable mortality risk in chronic kidney disease, Am. J. Cardiol. 101 (2008) 1741–1746. [14] M.C. Chao, S.L. Hu, H.S. Hsu, L.E. Davidson, C.H. Lin, C.I. Li, C.S. Liu, T.C. Li, C.C. Lin, W.Y. Lin, Serum homocysteine level is positively associated with chronic kidney disease in a Taiwan Chinese population, J. Nephrol. (2014) [Epub ahead of print]. [15] H. Jakubowski, Quality control in tRNA charging — editing of homocysteine, Acta Biochim. Pol. 58 (2011) 149–163. [16] B.A. Maron, J. Loscalzo, The treatment of hyperhomocysteinemia, Annu. Rev. Med. 60 (2009) 39–54. [17] M. Lacinski, W. Skorupski, A. Cieslinski, J. Sokolowska, W.H. Trzeciak, H. Jakubowski, Determinants of homocysteine-thiolactonase activity of the paraoxonase-1 (PON1) protein in humans, Cell. Mol. Biol. (Noisy-le-grand) 50 (2004) 885–893. [18] H. Jiang, S.P. Stabler, R.H. Allen, K.N. Maclean, Altered expression of apoA-I, apoA-IV and PON-1 activity in CBS deficient homocystinuria in the presence and absence of treatment: possible implications for cardiovascular outcomes, Mol. Genet. Metab. 107 (2012) 55–65. [19] H. Jakubowski, Calcium-dependent human serum homocysteine thiolactone hydrolase. A protective mechanism against protein N-homocysteinylation, J. Biol. Chem. 275 (2000) 3957–3962.
[20] J. Perla-Kajan, H. Jakubowski, Paraoxonase 1 protects against protein Nhomocysteinylation in humans, FASEB J. 24 (2010) 931–936. [21] H. Jakubowski, Homocysteine in Protein Structure/Function and Human Disease — Chemical Biology of Homocysteine-containing Proteins, Springer, Wien, New York, 2013. [22] H. Jakubowski, G.H. Boers, K.A. Strauss, Mutations in cystathionine {beta}-synthase or methylenetetrahydrofolate reductase gene increase N-homocysteinylated protein levels in humans, FASEB J. 22 (2008) 4071–4076. [23] J. Suszynska-Zajczyk, O. Utyro, H. Jakubowski, Methionine-induced hyperhomocysteinemia and bleomycin hydrolase deficiency alter the expression of mouse kidney proteins involved in renal disease, Mol. Genet. Metab. 112 (2014) 339–346. [24] J. Suszynska-Zajczyk, J. Wroblewski, O. Utyro, M. Luczak, L. Marczak, H. Jakubowski, Bleomycin hydrolase and hyperhomocysteinemia modulate the expression of mouse proteins involved in liver homeostasis, Amino Acids 46 (2014) 1471–1480. [25] J. Suszynska-Zajczyk, M. Luczak, L. Marczak, H. Jakubowski, Hyperhomocysteinemia and bleomycin hydrolase modulate the expression of mouse brain proteins involved in neurodegeneration, J. Alzheimers Dis. 40 (2014) 713–726. [26] C. Noll, J. Dairou, C. Ripoll, J.L. Paul, J.M. Dupret, J.M. Delabar, F. Rodrigues-Lima, N. Janel, Effect of red wine polyphenol dietary supplementation on two phase II enzymes in liver of hyperhomocysteinemic mice, Food Chem Toxicol. 49 (2011) 1764–1769. [27] J. Julve, J.C. Escola-Gil, E. Rodriguez-Millan, J.M. Martin-Campos, M. Jauhiainen, H. Quesada, I.M. Renteria-Obregon, J. Osada, J.L. Sanchez-Quesada, F. Blanco-Vaca, Methionine-induced hyperhomocysteinemia impairs the antioxidant ability of high-density lipoproteins without reducing in vivo macrophage-specific reverse cholesterol transport, Mol. Nutr. Food Res. 57 (2013) 1814–1824. [28] I. Muhlberger, K. Monks, R. Fechete, G. Mayer, R. Oberbauer, B. Mayer, P. Perco, Molecular pathways and crosstalk characterizing the cardiorenal syndrome, OMICS 16 (2012) 105–112. [29] K. Farbakhsh, B.L. Kasiske, Dyslipidemias in patients who have chronic kidney disease, Med. Clin. North Am. 89 (2005) 689–699. [30] R.M. Gueant-Rodriguez, R. Spada, M. Moreno-Garcia, G. Anello, P. Bosco, L. Lagrost, A. Romano, M. Elia, J.L. Gueant, Homocysteine is a determinant of ApoA-I and both are associated with ankle brachial index, in an ambulatory elderly population, Atherosclerosis 214 (2011) 480–485. [31] P.O. Attman, C. Knight-Gibson, M. Tavella, O. Samuelsson, P. Alaupovic, The compositional abnormalities of lipoproteins in diabetic renal failure, Nephrol. Dial. Transplant. 13 (1998) 2833–2841. [32] K.R. Feingold, Y. Wang, A. Moser, J.K. Shigenaga, C. Grunfeld, LPS decreases fatty acid oxidation and nuclear hormone receptors in the kidney, J. Lipid Res. 49 (2008) 2179–2187. [33] I. Santamaria, D. Alvarez-Hernandez, R. Jofre, J.R. Polo, J. Menarguez, J.B. CannataAndia, Progression of secondary hyperparathyroidism involves deregulation of genes related to DNA and RNA stability, Kidney Int. 67 (2005) 2267–2279. [34] N.M. White, O. Masui, L.V. Desouza, O. Krakovska, S. Metias, A.D. Romaschin, R.J. Honey, R. Stewart, K. Pace, J. Lee, M.A. Jewett, G.A. Bjarnason, K.W. Siu, G.M. Yousef, Quantitative proteomic analysis reveals potential diagnostic markers and pathways involved in pathogenesis of renal cell carcinoma, Oncotarget 5 (2014) 506–518. [35] X. Zhang, S. Huang, M. Gao, J. Liu, X. Jia, Q. Han, S. Zheng, Y. Miao, S. Li, H. Weng, X. Xia, S. Du, W. Wu, J.A. Gustafsson, Y. Guan, Farnesoid X receptor (FXR) gene deficiency impairs urine concentration in mice, Proc. Natl. Acad. Sci. U. S. A. 111 (2014) 2277–2282.