To confirm my proteomics analysis for identifying Prdx5 interacting partners, coprecipitation experiments were performed with some representative proteins. As shown in Figure 5, DBT, Rab43, Alb, and Pccb were shown to strongly coprecipitate with Prdx5 in hypoxia, consistent with the proteomics analysis in Table 3. Taken together, these findings suggested that Prdx5 could act as a direct regulator in hypoxia and be involved in maintaining kidney homeostasis.
11
Table 1. Schematic representation of mammalian peroxiredoxin family members.
Name Structurea Localization Electron
donor
Subunit References
Prdx1
Cytosol Thioredoxin GSH
Dimer
3, 4, 8 Prdx2
Cytosol Thioredoxin Dimer Decamer
Prdx3
Mitochondria Thioredoxin Dimer
Prdx4
Plasma Membrane
Thioredoxin GSH
Dimer
Prdx5
Mitochondria Peroxisome
Cytosol
Thioredoxin Monomer
Prdx6
Plasma GSH?
Cyclophillin A?
Monomer
aThe cysteins that relate with peroxidase activity are indicated as Cp (peroxidase cystein) or Cr (resolving cystein). Prdx3 and Prdx5 have mitochondrial import signals at their N-terminal regions, beside that Prdx5 also has a peroxisomal localization signal at its C-terminus. Prdx4 has a signal peptide for secretion at the N-terminus (8). Prdx5 exists in ubiquitous cell line and appears to be multifunctional, in some case it plays a role as a stress-inducible factor under specialized oxidative stress conditions, especially hypoxic stress (42).
Cp
Cp Cr
Cp Cr
Cp Cr
Cp Cr
Cp Cr
12
Table 2. The list of primers used for RT-PCR and Realtime-PCR.
Gene Primer name Sequence Purpose
VEGFa VEGFa-F 5’-ACATCTTCAAGCCGTCCTGTGTGC-3’ RT-PCR VEGFa-R 5’-AAATGGCGAATCCAGTCCCACGAG-3’
β-actin Actin-F 5’-AGCGGGTCGTGCGTG-3’ RT-PCR
Actin-R 5-CAGGGTACATGGTGGTGC-3’
VEGFa
VEGFa-F1 5’-CTCACTTCCAGAAACACGACAAA-3’ Real-time VEGFa-R1 5’GCATCTTTATCTCTTTCTCTGTCATCA-3’ PCR
β-actin Actin-F1 CTGTCCACCTTCCAGCAGATGT Real-time
Actin-R1 5’-ACAGTCCGCCTAGAAGCACTTG-3’ PCR
PPIA
PPIA-F1 5’-CCCCATCTGCTCGCAATG-3’ Real-time PCR PPIA-R1 5’-GAGGAAAATATGGAACCCAAAGAA-3’
13
Table 3. Putative target protein altered interaction by Prdx5 immunoprecipitation under hypoxic stressa.
Accession Description Gene Score MS/MS
spectra
Frequencyb Normoxia Hypoxia
IPI00130467 Ras related protein Rab 43 isoform b Rab43 228.7 6 ND 1/3
IPI00130535 Lipoamide acyltransferase component of branched chain α-keto acid dehydrogenase complex, mitochondrial
Dbt 302.6 8 ND 2/3
IPI00131695 Serum albumin Alb 325.3 3 1/3 3/3
IPI00330523 Propionyl CoA carboxylase alpha chain, mitochondrial
Pcca 474.0 21 2/3 3/3
IPI00606510 Propionyl CoA carboxylase beta chain, mitochondrial
Pccb 266.6 10 2/3 3/3
IPI00346834 Keratin type II cytoskeletal 2, oral Krt76 184.2 7 ND 1/3
IPI00227140 Keratin type I cytoskeletal 14 Krt14 1399.7 16 ND 1/3
IPI00230365 Keratin type I cytoskeletal 17 Krt17 1322.1 16 ND 1/3
IPI00347019 Keratin type II cuticular Hb4 Krt84 330.8 10 ND 1/3
IPI00347096 Keratin type II cytoskeletal 72 Krt72 337.2 8 ND 1/3
IPI00462140 Keratin type II cytoskeletal 1b Krt77 2014.2 7 ND 1/3
IPI00468696 Keratin type I cytoskeletal 42 Krt42 798.6 12 ND 1/3
IPI00420970 Keratin type II cytoskeletal 74 Krt74 1999.7 4 1/3 3/3
IPI00226993 Thioredoxin Txn1 1155.1 3 3/3 2/3
IPI00225123 Non-lysosomal glucosylceramidase Gba2 191.4 9 1/3 ND
IPI00348328 Keratin Kb40 Krt78 580.5 19 1/3 ND
IPI00122281 Keratin type I cuticular Ha2 Krt32 260.3 8 1/3 ND
14
aProteins were affinity-purified from mouse kidneys under both normoxic and hypoxic conditions as bound interactors with Prdx5 immunoprecipitation. The purified immunoprecipitates were applied to acrylamide gel-associated tryptic digestion and subjected to nano-UPLC-MS/MS for protein identification.
bFrequency represents the number of times that the interactors were observed in three independent experiments. ND, not detected.
15
Table 4. List of proteins interacted with Prdx5 is not altered during hypoxic stressa.
Accession Description Gene Score MS/MS
spectra
Frequencyb Normoxia Hypoxia
IPI00625729 Keratin type II cytoskeletal 1 Krt1 2912.3 7 3/3 3/3
IPI00130391 Protease serine 1 Prss1 1025.4 5 3/3 3/3
IPI00988950 Hemoglobin subunit beta 1 like Hbb-b1 1493.7 6 2/3 2/3
IPI00131674 Trypsinogen 2210010
C04Rik
983.5 1 2/3 2/3
IPI00468956 Keratin type II cytoskeletal 71 Krt71 2022.0 4 1/3 1/3
IPI00622240 Keratin type II cytoskeletal 2 epidermal Krt2 657.2 8 1/3 1/3
IPI00311493 Keratin type I cytoskeletal 18 Krt18 297.0 8 1/3 1/3
aProteins were affinity-purified from mouse kidneys under both normoxic and hypoxic conditions as bound interactors with Prdx5 immunoprecipitation. The purified immunoprecipitates were applied to acrylamide gel-associated tryptic digestion and subjected to nano-UPLC-MS/MS for protein identification.
bFrequency represents the number of times that the interactors were observed in three independent experiments. ND, not detected.
16
H. sapiens MGLAGVCALRRSAGYILVGGAGGQSAAAAARRCSEGEWASGGVRSFSRAAAAMAPIKVGD 60 M. musculus MLQLGLRVLGCKASSVLR--ASTCLAGRAGR—-KEAGWECGGARSFSSSAVTMAPIKVGD 56 D. rerio MISTPLLQKGVRAAHCTF---RQLHCTPITSMPIKVGQ 35 D. melanogester MRVLSCKFLGRVVNSALP---QQIISLRSLSKTSAAMV--KVGD 39 * . *.: :. : ***:
H. sapiens AIPAVEVFEGEPGN--KVNLAELFKGKKGVLFGVPGAFTPGCSKTHLPGFVEQAEALKAK 118 M. musculus AIPSVEVFEGEPGK--KVNLAELFKGKKGVLFGVPGAFTPGCSKTHLPGFVEQAGALKAK 114 D. rerio RLPAVEVQEEDPGNSLSMETAELFSCKRGVLFGVPGAFTPGCSKTHLPGFIQMAGELRAK 95 D. melanogester SLPSVDLFEDSPAN--KINTGDLVNGKKVIIFGVPGAFTPGCSKTHLPGYVSSADELKSK 97 :*:*:: * .*.: .:: .:*.. *: ::******************::. * *::*
H. sapiens -GVQVVACLSVNDAFVTGEWGRAHKAEGKVRLLADPTGAFGKETDLLLDDS-LVSIFGNR 176 M. musculus -GAQVVACLSVNDVFVIEEWGRAHQAEGKVRLLADPTGAFGKATDLLLDDS-LVSLFGNR 172 D. rerio -GVDEVACISVNDVFVMSAWGKQNGADGKVRMLADPTGAFTKAVDLVLNNAQLIPVLGNL 154 D. melanogester QGVDEIVCVSVNDPFVMSAWGKEHGAAGKVRLLADPAGGFTKALDVTIDLP----PLGGV 153 *.: :.*:**** ** **: : * ****:****:*.* * *: :: . :*.
H. sapiens RLKRFSMVVQDGIVKALNVEPDGTGLTCSLAPNIISQL 214 M. musculus RLKRFSMVIDNGIVKALNVEPDGTGLTCSLAPNILSQL 210 D. rerio RSQRYAMLIENGVVTKLSVEPDGTGLTCSLASNFLAEV 192 D. melanogester RSKRYSLVVENGKVTELNVEPDGTGLSCSLANNIGKK- 190 * :*:::::::* *. *.********:**** *: :
Figure 1. Alignment of amino acid sequence of human Prdx5 and orthologues in other species. Human Prdx5 and its orthologues were applied to ClustalW 2 program (11). Amino acid numbering based on cytosolic isoform of human Prdx5 sequence. Mitochrondrial targeting sequences (MTSs) are high-lighted in gray. Peroxidatic (CP) and resolving (CR) cysteins are marked in boxes. Asterisks indicate identities, and dots and double dots indicate conservative and highly conservative substitutions, respectively. GenBank accession numbers of peptide sequences using in alignment analysis are NM_012094 (Homo sapiens), NM_012021 (Mus musculus), CN508467 (Danio rerio), and NP_650679 (Drosophila melanogaster).
152
CR
48
1
73 CP
Mitochondrial targeting sequences
17
Figure 2. Immunoprecipitation using mouse anti-Prdx5 antibody. 200 àg of total lysate from mouse kidneys (input) was incubated with 5, 10, 20 àl of mouse Prdx5 antibody at 4 oC in circulator for overnight. The immune complex was pulled down by incubating with protein G agarose (Invitrogen, USA) for 4 hours at 4 °C. Unbound and bound fractions were separated by centrifugation at 3000 rpm, followed by washing several times with 1x PBS. After that, unbound and bound fractions were applied into SDS-PAGE and western blot with commercial Prdx5 antibody. The results from western blotting showed that Prdx5 existed in bound fraction from start point at 5 àl antibody. Additionally, the decrease of Prdx5 amount in unbound fraction got along with the increase of Prdx5 amount in bound fraction when I added more Prdx5 antibody (5, 10, 20 àl antibody). Prdx5 did not exist in unbound fraction when Prdx5 antibody reached maximal concentration (20 àl antibody), that indicated Prdx5 was fully pulled down into bound fraction. Taken together, these results indicated Prdx5 antibody produced in my laboratory could co-purified mouse Prdx5 and the concentration using for immunoprecipitation assay in a range 5-20 àl antibody (IgG concentration of Prdx5 about 5.65 àg/àl).
18
Figure 3. Confirmation of VEGFa expression in hypoxia treated kidney via semi-quantified RT-PCR and realtime-PCR. Mouse kidneys from hypoxic and normoxic groups were homogenized in Qiazol under manufacturer’s instructions. They were divided in two groups, one group (n=3) used for RT-PCR (A) and one group (n=3) used for realtime-PCR (B). The results from RT-PCR showed upregulation of VEGFa during hypoxic treatment, 1.30±0.10 versus 0.97±0.06 (hypoxia versus normoxia group, respectively. Relative expressions were normalized with β-actin and represented in mean ± standard deviation, p<0.05. Additionally, relative expression of VEGFa also increased in hypoxic group, 1.40±0.12 (β-actin) and 1.38±0.07 (PPIA) compared with normoxic group, 1.05±0.15 (β-actin) and 1.01±0.03 (PPIA).
A
B
19
Figure 4. Work-flow to identify putative target protein interacting with Prdx5 in hypoxic kidney. C57BL6/J mice were divided into two groups, one group maintained in normoxia (20.0±0.5% O2 ) whereas other maintained in O2 concentration regulated chamber ( 8.0±0.5% O2, for 4 hours). After indicated time point, the mice were sacrificed and collected the kidneys. I prepared the total lysate from normoxic and hypoxic kidneys in a lysis buffer containing 1%
Triton X-100 in 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate, 1 mM sodium orthovanadate, 25 mM sodium fluoride, 1 àg/ml leupeptin, and 1 mM PMSF. Protein was extracted for over 4 hours at
20
4°C followed by sonication. Then 500 àg of protein was incubated with 10 àl Prdx5 antibody.
The co-purified proteins were pulled down by incubated with 50 àl protein G (50% slurry). The immunoprecipitated complex was eluted with 60 mM Tris-HCl (pH 6.8), 2.5% glycerol, 2%
SDS, and 28.8 mM β-mercaptoethanol and then the eluted complex was freeze-dried before being subjected to gel associated trypsin digestion and nano-UPLC-MS/MS analysis for comparative proteomics. Processed ions were mapped against the IPI mouse database (version 3.87) for identify putative protein.
21
Figure 5. Confirmation putative target proteins interacting with Prdx5 by reverse immunoprecipitation. To confirm my proteomics analysis for identifying Prdx5 interacting partners, I carried out reverse immunoprecipitation assay with target protein [(A) Pccb, (B) Alb, (C) Rab43, (D) DBT]. The data indicated that all target proteins strongly co-precipitated with Prdx5 in hypoxic condition, these results also were consistent with my data collected from nano- UPLC-MS/MS comparative analysis.
A
C D
B