Age-related subproteomic analysis of mouse liver and kidney peroxisomes

1477-5956-5-19 1477-5956 Research Age-related subproteomic analysis of mouse liver and kidney peroxisomes Mi Jia Jia.Mi@dbb.su.se Garcia-Arcos Itsaso ofbgaari@ehu.es Alvarez Ruben rubenalva@andaluciajunta.es Cristobal Susana Susana.Cristobal@dbb.su.se

Department of Cell and Molecular Biology, Biomedical Center, Box 596, Uppsala University, SE-751 24 Uppsala, Sweden

Department of Biochemistry and Biophysics, The Arrhenius Laboratories for Natural Sciences, Stockholm University, SE-106 91 Stockholm, Sweden

Proteome Science 1477-5956 2007 5 1 19 http://www.proteomesci.com/content/5/1/19 18042274 10.1186/1477-5956-5-19 25 5 2007 27 11 2007 27 11 2007 2007 Mi et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

Despite major recent advances in the understanding of peroxisomal functions and how peroxisomes arise, only scant information is available regarding this organelle in cellular aging. The aim of this study was to characterize the changes in the protein expression profile of aged versus young liver and kidney peroxisome-enriched fractions from mouse and to suggest possible mechanisms underlying peroxisomal aging. Peroxisome-enriched fractions from 10 weeks, 18 months and 24 months C57bl/6J mice were analyzed by quantitative proteomics.

Results

Peroxisomal proteins were enriched by differential and density gradient centrifugation and proteins were separated by two-dimensional electrophoresis (2-DE), quantified and identified by mass spectrometry (MS). In total, sixty-five proteins were identified in both tissues. Among them, 14 proteins were differentially expressed in liver and 21 proteins in kidney. The eight proteins differentially expressed in both tissues were involved in β-oxidation, α-oxidation, isoprenoid biosynthesis, amino acid metabolism, and stress response. Quantitative proteomics, clustering methods, and prediction of transcription factors, all indicated that there is a decline in protein expression at 18 months and a recovery at 24 months.

Conclusion

These results indicate that some peroxisomal proteins show a tissue-specific functional response to aging. This response is probably dependent on their differential regeneration capacity. The differentially expressed proteins could lead several cellular effects: such as alteration of fatty acid metabolism that could alert membrane protein functions, increase of the oxidative stress and contribute to decline in bile salt synthesis. The ability to detect age-related variations in the peroxisomal proteome can help in the search for reliable and valid aging biomarkers.

Background

Aging is a natural phenomenon that affects the entire physiology of an organism. It is a complex process resulting from changes in the expression and regulations of numerous genes over time. It has become evident from high-throughput studies that the metabolic pathways affected in aging are interconnected 12. Therefore, techniques such as proteomics that allow the simultaneous analysis of thousands of molecular parameters within a single experiment could facilitate to identify candidate proteins for aging biomarkers in animal models.

Comparative proteomics has been used to study the effect of aging on the proteome from rat skeletal muscle 3, in epithelial cells 4, brain mice 56, and on specific organelles such as Golgi apparatus, and endoplasmic reticulum 7 or mitochondrial proteins in mice 8, in rat 9, in bovine heart 10 or rat brain 11. These techniques have been applied to examine the effect of anti-aging agents on human endothelial cells 12. Comparative studies using premature aging Hutchinson-Gilford progeria syndrome fibroblasts revealed differential protein expression and glycosylation of membrane proteins 13. Proteomics studies on aged samples have also disclosed various non-enzymatic modifications such as glycosylation and nitration that progress with age 14. These studies clearly indicate the value of additional comprehensive proteomic analysis to find novel aging biomarkers.

It is widely accepted that the reactive oxygen species (ROS) are one of the mediators of aging in most species 15, either being a direct cause of aging or as a by-product of a genetically programmed process 16. So far, most of the subcellular studies have been focused on the mitochondrion that generates the main proportion of cellular ROS. Likewise, the peroxisomal oxidative metabolism is an additional source of ROS. The peroxisome also responds to oxidative stress and protects against oxidative damage. However, the information about which process initiates the aging cascade in the peroxisome is still scarce.

Studies have reported a general decrease in peroxisomal function with aging 17. Decreases in catalase (CAT) activity has been found in various studies 1718. But in particular, inconsistencies have been reported about the age-related effects in the peroxisomal β-oxidation 19. Recently, peroxisome senescence has been studied in human fibroblasts showing that aging comprises the peroxisomal targeting signal protein import and the key antioxidant enzyme CAT 20. The lack of peroxisomal CAT in the nematode Caenorhabditis elegans has been shown to cause a progeric phenotype 21.

Our recent improvements in a comparative proteomic technique aimed for concentrating the subproteome of interest avoiding a complex cellular fractionation 22. We have recently characterized peroxisome-enriched fractions from two mouse tissues: liver and kidney by 2-DE based proteomics 23, predicted peroxisomal proteomes from sequenced genomes 24, and analyzed peroxisomal proteomes from invertebrate species 25. Those studies on invertebrates provided the largest number of identified proteins from mussel's peroxisomal proteome and the proteins have been utilized to identify novel and poor understood pathways affected by xenobiotics 222627.

Applying comparative proteomics to peroxisomal samples could provide new clues to which molecular events were associated with aging in peroxisomes. In this paper, peroxisome-enriched fractions from two mouse tissues: liver and kidney and three ages: 10 weeks, 18 months, 24 months were analyzed by quantitative proteomics. First, we present tissue-specific protein expression profiles from the different ages and a common protein profile to both tissues; thereafter, differentially expressed proteins were identified by MALDI-TOF MS and the differentially expressed proteins were functionally classified. Finally, Western blotting and analysis of predicted transcription factors are in agreement with the quantitative proteomic data. Our results provide an age-related subproteomic analysis of a mouse peroxisome-enriched fraction from liver and kidney, two tissues with different generation capacity.

Results

Quantitative analysis of the peroxisome-enriched fraction from young and old mouse liver and kidney

The principal aim of this study was to characterize age-dependent changes in the peroxisomal proteome of liver and kidney. The tissues were subjected to differential centrifugation and the peroxisome-enriched fraction was further purified with an iodixanol gradient in a density gradient centrifugation (Fig. 1A). The quality of each isolation procedure was assessed. The purity of the peroxisome-enriched fraction was based on the measurement of the marker enzyme, CAT, and verified by protein gel blot analysis routinely 28. The quality of each isolation procedure was assessed and the possible cross-contamination with other organelles was followed by protein gel blot analysis (Fig. 1B and 1C). The method provided preparations from different tissues and ages with equivalent quality 23. Proteins were subjected to 2-DE followed by colloidal Coomassie staining. Statistical analyses were applied to compare the average spot ratio of expression between the 2-DE maps from tissues of different ages. On average, about 140 spots were identified in both tissues that corresponded to 65 different proteins [see Additional file 1] 22.

Additional file 1

Identification of proteins from mouse peroxisomal enriched fractions. Proteins were identified by MALDI-TOF-MS.

Click here for file

Figure 1

Scheme of the protein purification process in A and immunoblot analysis of the peroxisome-enriched fractions in B and C.

Scheme of the protein purification process in A and immunoblot analysis of the peroxisome-enriched fractions in B and C. The enrichment of specific organelle proteins was followed by loading the same amount of protein (20 μg each) of total homogenate (lane A), light mitochondrial fraction (laneD) and highly purified peroxisomal fraction (lane P) onto a 12.5% T polyacrylamide gel. In B section, the immunoblot of CAT were more intensively stained in the fraction P than A and in C section, MDH immunoreaction is found mainly in fraction D. CAT, catalase; MDH, malate dehygrogenase.

In the liver, 14 spots showing changes in protein expression between young and old mouse tissues was observed. The 18 months samples showed a down-regulation response that was compensated by an up-regulation at 24-month. The differentially expressed proteins are illustrated in Fig. 2A, B and 2C, Fig. 3A and Table 1. In the kidney, the statistically significant difference in protein expression affected to 21 spots. The variations at 18 months were equally distributed between up- and down-regulations whereas at 24 months, down-regulation was the main aging-associated effect. The differentially expressed proteins in the kidney are illustrated in Fig. 2D, E and 2F, Fig 3B and Table 2. Eight proteins were up or down-regulated under all condition studied. These proteins composed an aging protein expression profile common to liver and kidney. The most of these eight proteins were down-regulated in the 18 months liver and 24 months kidney samples. However, in the 24 months old liver samples, these proteins were mainly up-regulated (Fig 3C).

Figure 2

Representatives 2-DE gels from peroxisome-enriched fractions were isolated from liver and kidney of M. musculus and separated by 2-DE with denaturing isoelectric focusing (IEF) on immobilized pH gradients in the first dimension between pH 3–10 non-linear (11 cm) and SDS-PAGE 12.5% in the second dimension. Gels were calibrated for molecular mass (in kDa) and pI (in pH units) by external pH and mass standards and stained by colloidal Coomassie. Differential expressed proteins are marked with numbers in A for liver and D for kidney. A. tissue: liver, age: 10 weeks; B. tissue: liver, age: 18 months; C. tissue: liver, age: 24 months; D. tissue: kidney, age: 10 weeks; E. tissue: kidney, age: 18 months; F. tissue: kidney, age: 24 months.

Figure 3

Proteins differentially expressed in the 18 months and 24 months versus 10-weeks old group

Proteins differentially expressed in the 18 months and 24 months versus 10-weeks old group. The ratios were calculated dividing the volume percentage per each spot from the 18 months or 24 months by the volume percentage per spot in the 10 weeks old group. The vertical axis corresponds to the average ratio of expression, above the 0 value for the up-regulated proteins and below the 0 value for the down-regulated ones. According to 18 months old group, in the horizontal axis the down-regulated proteins are organized with the lowest values on the left side and the up-regulated ones show the highest values on the right side. Color code: green for 18 months group and orange for the 24 months old group. A. Liver age-related proteins. B. Kidney age-related proteins. C. Differential spot detection in fold from the 8 proteins composing the common age-related PES to both tissues. The four groups correspond to 18 months liver, 24 months liver, 18 months kidney and 24 months kidney samples.

Table 1

Subset of age-associated proteins from quantitative proteomic analysis of liver peroxisomal proteins. Proteins were identified by MALDI-TOF MS ^a).

10 wks

18 mns

24 mns

A (Fold)

SDV(Fold)

B (Fold)

SDV(Fold)

C (Fold)

SDV(Fold)

Pathway

Location

NCBI nr

Cvage

Score

Mr ob

pI ob

0.29

-2.12

0.03

1.63

0.22

β-oxidation

pero

gi|66793429

24%

0.1

25000

0.31

-1.62

0.47

3.11

0.52

β-oxidation

pero

giI30525893

26%

0.0251189

40000

9.0

0.48

0.00

2.20

0.52

α-oxidation

pero

gi|31560355

30%

104

3.981E-06

65000

6.8

0.15

-1.48

0.34

-3.28

0.06

isoprenoid biosynthesis

pero

gi|20965433

33%

0.0158489

26000

8.7

0.50

-1.31

0.39

2.05

0.69

isoprenoid biosynthesis related

pero

gi|409499

24%

0.01

33000

9.1

0.42

2.78

0.43

1.48

0.23

amino acid metabolism

pero

gi|19388006

24%

0.0501187

42000

7.8

0.09

1.02

0.25

-3.77

0.13

putative peroxisomal prot

pero

gi|6678726

18%

0.1

67000

6.1

0.43

2.14

0.39

-1.09

0.41

putative peroxisomal prot

pero

gi|6753036

40%

117

1.995E-07

60000

6.6

0.18

-1.33

0.10

-2.87

0.13

β-oxidation

mito

gi|21431780

24%

0.3162278

33000

8.8

0.21

0.00

9.36

0.15

β-oxidation

mito

gi|20810027

23%

0.01

46000

7.8

0.14

2.26

0.62

1.09

0.29

glutamate biosynthesis

mito

gi|6680027

22%

0.0398107

60000

7.4–7.5

0.38

-1.20

0.15

-3.31

0.21

stress response

mito

gi|14917005

25%

0.0501187

75000

6.2

0.32

2.52

0.53

1.32

0.38

glutamine metabolism

cyt

gi|15419027

28%

0.1258925

42000

7.5

0.05

7.35

0.46

0.00

protein folding

cyt

gi|129729

28%

0.0398107

61000

^a)Protein numbers correspond to number from the 2-DE gels shown in Figure 2. Data from molecular mass in kDa and pI can assist the localization in the Figure 2. The 10 weeks old group correspond to A, 18 months old group as B and 24 months old group as C. The 10 weeks old group has been taken as a reference. Therefore, the change in protein expression associated with age is presented as the ratio A/A or B/A or C/A for liver. The ratios were calculated dividing the volume percentage per each spot in the 10 weeks, 18 months or 24 months divide by the volume percentage per spot in the 10 weeks. The standard deviations data are included. The biochemical pathways and subcellular localization are also indicated.

Table 2

Subset of age-associated proteins from quantitative proteomic analysis of kidney peroxisomal proteins. Proteins were identified by MALDI-TOF MS ^a).

10 wks

18 mns

24 mns

D (Fold)

SDV(Fold)

E (Fold)

SDV(Fold)

F (Fold)

SDV(Fold)

Pathway

Location

NCBI nr

Cvage

Score

Mr ob

pI ob

0.22

-3.4

0.04

1.02

0.06

auxiliary β-oxidation

pero

gi|6753272

30%

114

3.981E-07

60000

7.5–7.8

0.18

-2.27

0.02

1.04

0.12

auxiliary β-oxidation

pero

gi|6753272

30%

114

3.981E-07

60000

7.5–7.8

0.31

-4.16

0.07

-1.86

0.10

β-oxidation

pero

gi|66793429

24%

0.1

25000

0.16

-1.18

0.46

-3.53

0.13

β-oxidation

pero

giI30525893

26%

0.0251189

40000

9.0

0.26

0.00

-4.38

0.35

α-oxidation

pero

gi|6754564

20%

0.0398107

36000

7.3

0.12

0.00

-1.44

0.22

α-oxidation

pero

gi|31560355

30%

104

3.981E-06

65000

6.8

0.29

3.04

0.58

-1.73

0.05

isoprenoid biosynthesis

pero

gi|20965433

33%

0.0158489

26000

8.7

0.18

-1.11

0.22

-2.61

0.15

isoprenoid biosynthesis related

pero

gi|409499

24%

0.01

33000

9.1

0.14

0.00

6.80

1.10

amino acid metabolism

pero

gi|19388006

24%

0.0501187

42000

7.8

0.17

4.6

0.36

0.00

amino acid metobolism

pero

giI17390882

25%

0.0316228

33000

6.0

0.10

-1.95

0.16

0.00

amino acid metobolism

pero

gi|477004

32%

104

3.981E-06

65000

6.5

0.27

-2.88

0.14

1.14

0.19

amino acid metobolism

pero

gi|477004

32%

104

3.981E-06

65000

6.7

0.23

1.24

0.05

-1.59

0.15

purine/pyrimidine metabolism

pero

gi|18044669

26%

0.0501187

35000

9.0–9.1

0.27

3.45

0.52

-1.21

0.13

putative peroxisomal protein

pero/cyt

gi|226778

24%

0.0199526

33000

7.5

0.19

2.46

0.45

0.00

putative peroxisomal protein

pero

gi|76779273

21%

0.1258925

62000

5.7–5.9

0.14

1.94

0.22

-1.17

0.37

β-oxidation

cyt

gi|13182962

39%

0.1

30000

9.0

0.14

10.66

0.10

0.00

β-oxidation

mito

gi|21431780

24%

0.3162278

33000

8.8

0.18

4.3

0.17

0.00

β-oxidation

mito

gi|26345684

25%

0.0199526

45000

6.5

0.19

4.26

0.62

3.58

0.63

glycolysis

mito

gi|18043470

24%

0.3981072

37000

6.1

0.24

3.61

0.71

0.00

stress response

mito

gi|14917005

25%

0.0501187

75000

6.0

0.20

3.06

0.22

1.37

0.43

stress response

mito

gi|14917005

25%

0.0501187

75000

6.2

^a)Protein numbers correspond to number from the 2-DE gels shown in Figure 2. Data from molecular mass in kDa and pI can assist the localization in the Figure 2. The 10 weeks old group correspond to D, 18 months old group as E and 24 months old group as F. The 10 weeks old group has been taken as a reference. Therefore, the change in protein expression associated with age is presented as the ratio D/D, E/D and F/D for kidney. The ratios were calculated dividing the volume percentage per each spot in the 10 weeks, 18 months or 24 months divide by the volume percentage per spot in the 10 weeks. The standard deviations data are included. The biochemical pathways and subcellular localization are also indicated.

Principal component analysis (PCA) and hierarchical clustering

Cluster analysis methods were applied to confirm the age-related changes observed in the peroxisome-enriched fractions. Hierarchical clustering was used to blindly classify similar gels into classes. All liver gels were clustered together with a 93% similarity level and the kidney gels showed up to 84% similarity. It is remarkable that independently from the tissue, the similarity between the 10 weeks and 24 months gels were close to 95% and were clearly separated from the 18 months gels. For the kidney samples, the clustering differences between the controls and the 24 months samples was stronger with 13.3% difference than the 2% difference between the control and the 18 months samples (Fig 4A).

Figure 4

Organization of data by multivariate analyses

Organization of data by multivariate analyses. A. Heuristic clustering plot B. PCA performed on correlation matrix.

PCA clearly distinguished 2-DE gels based on tissues and ages (Fig 4B). The first component provided a sharp separation between tissues. In the positive side of the x-axis were situated all kidney gels and in the negative side, the corresponding liver gels. The second component separated the different ages. In the positive side of the y-axis the 18 months old gels from both tissues were situated and on the opposite side, the 24 months old gels also from liver and kidney. The control gels from young tissues were situated in the boundary between positive and negative values in the second component.

Cross-validation of some MALDI-TOF MS identified proteins by immunochemical analysis

Approximately 140 spots were tryptic-digested and analysed by MALDI-TOF MS and proteins were identified by combined data obtained from servers available for the academic communities 242930313233. Here, we confirmed the identification of 65 proteins from the liver and kidney peroxisome-enriched factions 23. In addition, the protein identification by peptide mass fingerprinting was cross-validated with immunochemical analysis for several proteins (CAT, acyl-CoA oxidase (AOX) and multifunctional protein) (Fig. 5A, B C and 5D). The Western blot data from AOX, CAT and multifunctional protein showed firstly, a recovery of these enzymes in the 24 months versus the 18 months group and secondary, a tissue-specify response. However, the different sensitivity range of these two techniques, 2DE-based analysis and Western blot analysis does recommend a qualitative comparison rather than a quantitative data comparison.

Figure 5

SDS-PAGE, immunoblotting from the different peroxisome-enriched fractions

SDS-PAGE, immunoblotting from the different peroxisome-enriched fractions. A. The increase in protein expression in a one-dimensional gel was followed by loading the same amount of protein (20 μg each) of the peroxisome-enriched fraction from liver and kidney samples of 10 weeks old, 18 months old and 24 months (lane P) onto a 12.5% T polyacrylamide gel and stained by Coomassie blue. B. Immunoblot against AOX antibody.C. Fragment from the immunoblot against CAT antibody. C. Fragment from the immunoblot against PH antibody. E. Plot of the differential intensities from the immunoblot against anti-CAT, F. anti-PH and G. anti-AOX for the total and H. Values from the three subunits that immunoreact. In the x-axis, different group organized by tissue and age are represented and in the y-axis, differential band intensity in fold. The value from the band of the immunoblot was normalized against the 10 weeks old liver sample.

Functional classification of the differentially expressed proteins and prediction of transcriptional factors

Twenty-seven proteins showed significant age-dependent differences in protein expression. These proteins were classified in Table 1 and 2 by subcellular localization and functional pathways. In the subcellular classification, the majority of these differentially expressed proteins were peroxisomal proteins, around 20% were mitochondrial proteins and 14% corresponded to cytosolic proteins (Fig. 6). Proteins were divided into 9 main biochemical pathways: β-oxidation, α-oxidation, isoprenoid biosynthesis, amino acid metabolism, purine and pyrimidine metabolism, stress response, protein folding and glycolysis and protein with unknown function. The β-oxidation was the pathway more affected by the effect of age in both tissues.

Figure 6

Functional classification of the age-related changes in the peroxisomal proteome from liver and kidney

Functional classification of the age-related changes in the peroxisomal proteome from liver and kidney. Distribution of protein profiles according to biochemical pathways in a pie-representation: in liver, kidney and common to both tissues. In the center, a Venn diagram of the number of proteins that composes the aging protein profile for each tissues and the common to both of them.

Each cell type or tissue at specific age or developmental stage has its own characteristic gene expression profile that could be defined, to some extent, by the presence of a combination of transcription factors. In order to find additional evidences to the age-related changes in the peroxisomal proteome of liver and kidney, we searched for predicted transcription factors in each of the 65 identified genes. The transcription factors were classified by different factors such as score, e-value, conservation among different genomes and predicted models [see Additional file 2]. In Table 3, we presented transcription factors from Mus musculus that were present in genes from the differentially expressed proteins. None of the factors were found in all eight genes from the common expression profile. Studying which transcription factors appeared in those genes, the R-ALPHA and SPZ1 transcription factors were found in 5 out of those 8 genes. Six different transcription factors appeared in proteins from the β-oxidation pathway and five in the pathways of the isoprenoid biosynthesis. From the 50 predicted transcription factors included in the list, 6 of them were conserved among different genomes whereas most of them were specific from M. musculus.

Additional file 2

Predicted transcription factors from the genes of the identified proteins in this study.

Click here for file

Table 3

Predicted transcription factors from the genes of the most relevant age-related proteins in this study.

β-oxidation

α-oxidation

isoprenoid bio.

aa met.

nucl.met.

TF vs PROT

CAT

ACOX1

ACCA1A

EHHADH

PHYH

HMGCS2

HMGCL

AGXT

DAO1

MRGPRF

Conserved

Model

RXR-ALPHA

T01331

SPZ1

yes

MA0111

NF-Y

yes

M00775

SP-1

M00931

PAX

M00979

PPARG

M00512

ZIC

M00448

BSAP

MA0014

COEI

T01112

CAC-BINDING PROTEIN

M00720

HOX-1.3

yes

M00023

KROX

M00982

SOX

MA0078

MEF-1,2

T00505

BRACHYURY

MA0009