Uterine Deletion of Trp53 Compromises Antioxidant Responses in the Mouse Decidua

Abstract

Preterm birth is a global health issue impacting millions of mothers and babies. However, the etiology of preterm birth is not clearly understood. Our recent finding that premature decidual senescence with terminal differentiation is a cause of preterm birth in mice with uterine Trp53 deletion, encoding p53 protein, led us to explore other potential factors that are related to preterm birth. Using proteomics approaches, here, we show that 183 candidate proteins show significant changes in deciduae with Trp53 deletion as compared with normal deciduae. Functional categorization of these proteins unveiled new pathways that are influenced by p53. In particular, down-regulation of a cluster of antioxidant enzymes in p53-deficient deciduae suggests that increased oxidative stress could be one cause of preterm birth in mice harboring uterine deletion of Trp53.

Pregnancy is a complex process that encompasses major events such as ovulation, fertilization, preimplantation embryo development, implantation, decidualization, placentation, and parturition. Each event is critical for the success of pregnancy. Preterm birth is a pathological state of pregnancy and a global health concern. Prematurity is a direct cause of nearly 30% of all neonatal deaths, and many babies who survive premature birth suffer serious long-term disabilities (1). Although inflammation/infection, stress, uterine overdistension, cervical aberration, and genetic and environmental factors are potential risk factors, the definitive cause of preterm birth remains elusive. Therefore, intense basic research exploring new approaches to combat preterm birth and prematurity is warranted. Animal models that show spontaneous preterm birth are desired tools for studying the underlying mechanism and for developing prevention and treatment strategies. We have recently generated mice with conditional uterine deletion of p53 that show spontaneous preterm birth (2).

Mutation of the Trp53 gene encoding p53 is present in various types of cancers. However, its functions in normal physiological processes, including female reproduction, remain ill understood (3). To examine the role of p53 during pregnancy events, we generated mice with conditional deletion of uterine Trp53 (p53^d/d) by crossing floxed Trp53 (Trp53^loxP/loxP) mice with progesterone receptor (Pgr)-Cre mice (Pgr^Cre/+) (2, 4, 5). Trp53^loxP/loxPPgr^+/+ (p53^f/f) mice, obtained as littermates of p53^d/d mice, were used as controls. Although p53^d/d females showed normal implantation, more than 50% of them had preterm birth with neonatal death due to premature decidual senescence resulting from terminal differentiation and senescence-associated growth restriction of decidual cells with up-regulation of p21, phosphorylated (p) protein kinase B (Akt), and phosphorylated S6 levels (2, 6). S6 kinase, which phosphorylates S6, is a downstream player in the mammalian target of rapamycin complex 1 (mTORC1) pathway. mTORC1 activity is activated by pAkt (7, 8). Because either inhibiting mTORC1 or deleting p21 attenuated senescence and prevented preterm birth in p53^d/d mice, these molecules are thought to be p53 downstream targets in deciduae (6). In fact, it has been shown in an in vitro model of human decidualization that withdrawal of decidualization stimuli down-regulates p53 but increases the level of pAkt (9, 10). It is also known that p21 participates in decidualization (11). However, the result that deciduae deficient in Trp53 have heightened p21 levels is surprising, because p21 is normally down-regulated in cells with a Trp53 mutation or deletion, such as most cancer cells with the exception of few cases (12, 13). These results prompted us to pursue further studies to explore downstream targets of p53 in deciduae. Because treatment with rapamycin, an inhibitor of mTORC1, attenuates early decidual senescence and preterm labor in p53^d/d mice, we sought to explore whether premature senescence and growth restriction affects other pathways that could be associated with adverse pregnancy outcome in p53^d/d females.

In the present study, we performed proteomic analyses using p53^d/d and p53^f/f deciduae on d 8 of pregnancy to identify additional potential downstream targets of p53. Mass spectrometry (MS) characterization of proteins was performed by both a standard commercial MS platform and a new platform coupling an ion mobility separation (IMS) with MS to increase sensitivity and provide greater proteome coverage (14, 15). Because IMS separates ions based on shape, the IMS-MS analysis effectively provides multidimensional separations, greater sensitivity due to efficient ion utilization, and a different view of proteomes of samples without sacrificing measurement throughput (16). These analyses indicate that although decidual Trp53 deficiency positively regulates DNA replication, this deficiency adversely affects antioxidant status, ATP production, and the proteinase system during decidualization. It is an exciting finding to see that Trp53 deletion down-regulates the expression of 10 antioxidant enzymes, including peroxiredoxin-6 (PRDX6), a unique antioxidant enzyme. We previously observed that Prdx6 is robustly expressed in the deciduum surrounding the embryo and acts against reactive oxygen species (ROS) during implantation (17). We also found that 21 proteins in mitochondrial ATP production were down-regulated in p53^d/d deciduae, whereas three proteins associated with DNA replication were up-regulated. Moreover, Trp53 deficiency attenuated the levels of five proteinase inhibitors, including α2-macroglobulin-P (A2MP), which is known to be expressed predominantly in mouse deciduae and is thought to regulate trophoblast invasion (18, 19). Collectively, our results show that loss of decidual Trp53 is associated with increased DNA replication and suppressed antioxidant status, ATP production, and proteolytic activity, unveiling new pathways that are influenced by p53 in mouse deciduae during early pregnancy.

Materials and Methods

Mice

Trp53^loxP/loxP mice were obtained from the Mouse Models of Human Cancer Consortium (Frederick, MD), whereas Pgr^Cre/+ mice were provided by John B. Lydon and Francesco J. DeMayo (Baylor College of Medicine, Houston, TX) (4, 5). To generate mice with uterine deletion of Trp53, Trp53^loxP/loxP mice (FVB/129) were crossed with Pgr^Cre/+ mice (C57BL6/129). To collect d 8 implantation sites (d 1, vaginal plug), littermate Trp53^loxP/loxP/Pgr^+/+ (p53^f/f) and Trp53^loxP/loxP/Pgr^Cre/+ (p53^d/d) mice on the mixed background were mated with wild-type fertile males to induce pregnancy. All mice used in this investigation were housed in the Cincinnati Children's Hospital Medical Center Animal Care Facility according to National Institutes of Health and institutional guidelines for the use of laboratory animals. All protocols of the present study were reviewed and approved by Cincinnati Children's Hospital Research Foundation Institutional Animal Care and Use Committee.

Preparation of protein samples for MS

Deciduae were dissected from d 8 mouse implantation sites free of embryos and myometria. Tissues were collected from three p53^d/d and three p53^f/f mice. Multiple deciduae were pooled from each mouse, and they were homogenized in 100 mm ammonium bicarbonate. The initial protein concentration was determined by BCA Protein Assay (Pierce, Rockford, IL). Urea and dithiothreitol (Sigma-Aldrich, St. Louis, MO) were added at a final concentration of 8 m and 10 mm, respectively, and incubated for 1 h at 37 C with shaking at 1000 rpm in Thermomixer R. The samples were diluted 5-fold with 50 mm ammonium bicarbonate (pH 7.8) before adding sequencing grade modified trypsin (Promega, Madison, WI) and 1 mm CaCl₂ (to stabilize the trypsin and reduce autolysis) for protein digestion (1:50 wt/wt trypsin-to-protein). The samples were then placed in a Thermomixer R for incubation overnight at 37 C and cleaned via strong cation exchange solid phase extraction (Supelco, Bellefonte, PA). All samples were concentrated down (∼50 μl) in a Speed-Vac SC 250 Express (Thermo Savant, Holbrook, NY). The final peptide concentration was determined by BCA Protein Assay, and peptide samples were stored at −80 C until MS analysis.

Liquid chromatography (LC)-tandem MS (MS/MS), LC-MS, and LC-IMS-MS analyses

Analysis of digested peptide mixtures of samples was performed on both Thermo Fisher Scientific LTQ Orbitrap Velos MS (Velos) (San Jose, CA) operated in MS/MS mode and an in-house built instrument that couples a 1-m IMS (IMS-MS) with an Agilent 6224 time-of-flight MS (Agilent Technologies, Santa Clara, CA) upgraded to a 1.5-m flight tube providing resolution of approximately 25,000 (14). A fully automated in-house built four-column HPLC system equipped with in-house packed capillary columns was used for both instruments with mobile phase A consisting of 0.1% formic acid in water and phase B comprised of 0.1% formic acid in acetonitrile (20). A 100-min LC gradient was performed on the Velos MS (using 60-cm-long columns with an outer diameter of 360 μm, inner diameter of 75 μm, and 3-μm C₁₈ packing material), whereas only a 60-min gradient with shorter columns (30 cm long with same dimensions and packing) was used with the IMS-MS; this additional IMS separation helps address detector suppression. Both gradients linearly increased mobile phase B from 0 to 60% until the final 2 min of the run when B was purged at 95%. Each sample (5 μl) was injected for both analyses, and the HPLC was operated under a constant flow rate of 0.4 μl/min for the 100-min gradient and 1 μl/min for the 60-min gradient. The Velos MS data were collected from 400-2000 m/z at a resolution of 60,000 (automatic gain control target, 1 × 10⁶) followed by data dependent ion trap MS/MS spectra (automatic gain control target, 1 × 10⁴) of the 10 most abundant ions using a collision energy setting of 35%. A dynamic exclusion time of 60 sec was used to discriminate against previously analyzed ions. IMS-MS data were collected from 100-3200 m/z.

Identification and quantification of the detected peptide peaks were performed using the accurate mass and time (AMT) tag approach (21). Briefly, Velos MS/MS data were searched by SEQUEST (Thermo Scientific, Rockford, IL), and the resulting 13,448 peptides that had a MS-GF score greater than or equal to 1E-9 were used to populate our AMT tag database (22, 23). Multiple in-house developed (publicly available http://ncrr.pnnl.gov/software) informatics tools were used to process the LC-MS data and correlate the resulting LC-MS features to an AMT tag database that contained accurate mass and LC separation elution time information for peptide tags generated from decidual proteins. Among the tools used were algorithms for peak-picking and for determining isotopic distributions and charge states (24). Further downstream data analysis incorporated all possible detected peptides into a visualization program, VIPER, to correlate LC-MS features to the peptide identifications in the AMT tag database (25). VIPER provided an intensity report for all detected features, normalized LC elution times via alignment to the database, and featured identification. In DAnTE software, peptide peak intensity values were converted to a log2 scale, normalized with a central tendency algorithm, and assessed at a protein level using rollup parameters; statistical comparisons were done at the peptide and protein level using ANOVA performed as t tests (p53^d/d vs. p53^f/f) in our analysis (26).

The focus of our analysis was peptide centric, because our MS technologies made possible measurements at the peptide level. Redundant peptide identifications in the case of a single peptide matching multiple proteins (typically protein isoforms) were removed. Therefore, each reported peptide matches back to a single protein. Differences in significance between p53^d/d and p53^f/f data were determined at the peptide level with a P value less than 0.05. A small subset of these peptides (∼7%) was deemed significant from an “all or none comparison,” where a peptide was detected in more than or equal to five datasets of one sample type and not detected in the other sample type. Tables 1–5 and Supplemental Tables 1 and 2, published on The Endocrine Society's Journals Online web site at http://endo.endojournals.org, depict these significant peptides as rolled up protein values (all peptides, regardless of significance, were rolled up to protein values in Supplemental Tables 3 and 4). Supplemental Table 5 contains the final IMS-MS and Velos-filtered results, their associated quantitative information, and functional categorization of each protein. Tables 1–5 highlight functional categories of interest. Proteins reported in this manuscript were to be identified by at least two peptides. Protein functionality was inferred from UniProtKB (http://www.uniprot.org), and links for each protein can be found in Supplemental Table 5, GO annotation in Supplemental Table 6, and KEGG information in Supplemental Tables 7–9. This section is expanded on in Supplemental methods.

Table 1.

Stress-related proteins

UniProt Gene ID	UniProt accession no.	Protein name	log2 (p53d/d/p53f/f)	P value
Antioxidants
Abat	P61922	4-Aminobutyrate aminotransferase, mitochondriala	−0.759	2.4E-05
Cat	P24270	Catalase	−0.593	7.3E-04
Cryab	P23927	α-Crystallin B chaina	−0.373	9.6E-04
FAM120A	Q6A0A9	Constitutive coactivator of PPAR-γ-like protein 1	−0.553	2.5E-03
G6pdx	Q00612	Glucose-6-phosphate 1-dehydrogenase X	−0.511	9.6E-06
Park7	Q99LX0	Protein DJ-1a	−0.932	2.9E-03
Prdx2	Q61171	Peroxiredoxin-2a	−0.444	6.3E-05
Prdx3	P20108	Thioredoxin-dependent peroxide reductase, mitochondriala	−0.444	1.0E-03
Prdx6	O08709	Peroxiredoxin-6a	−1.174	2.3E-08
Sod1	P08228	Superoxide dismutase (Cu-Zn)	−0.628	1.2E-03
Stress-induced proteins
Clu	Q06890	Clusterin	0.450	3.3E-03
Dnaja2	Q9QYJ0	DnaJ homolog subfamily A member 2a	0.735	1.9E-04
Hsp90aa1	P07901	Heat shock protein HSP 90-α	0.663	1.3E-03
Hsp90ab1	P11499	Heat shock protein HSP 90-βa	0.822	1.5E-03
Hspd1	P63038	60-kDa heat shock protein, mitochondrial	0.502	9.1E-03
Hsph1	Q61699	Heat shock protein 105 kDa	0.533	5.0E-04
Myof	Q8R3B4	Myoferlin	0.232	4.0E-03
Stip1	Q60864	Stress-induced-phosphoprotein 1	0.192	1.5E-03
Trap1	Q9CQN1	Heat shock protein 75 kDa, mitochondrial	4.939	2.6E-04
Hspb1	P14602	Heat shock protein β-1	−0.309	3.9E-04
Rbm3	O89086	Putative RNA-binding protein 3	p53f/f onlyb	—

^aCharacterized by IMS-MS and Velos instruments (average values used for fold change and P values).

^bp53^f/f only, p53^d/d samples do not show any signal.

Table 2.

Protein modification and degradation

UniProt Gene ID	UniProt accession no.	Protein name	log2 (p53d/d/p53f/f)	P value
Posttranslational modification
Calu	O35887	Calumenin	−0.486	7.4E-06
Man2a1	P27046	α-Mannosidase 2	p53f/f onlyb	—
Padi2	Q08642	Protein-arginine deiminase type-2	p53f/f onlyb	—
Pdia6	Q922R8	Protein disulfide-isomerase A6a	−1.284	2.3E-03
Rpn1	Q91YQ5	Dolichyl-diphosphooligosaccharide-protein glycosyltransferase subunit 1	−0.739	1.4E-05
Rpn2	Q9DBG6	Dolichyl-diphosphooligosaccharide-protein glycosyltransferase subunit 2a	−0.702	2.5E-03
Sptlc2	P97363	Serine palmitoyltransferase 2	−0.741	2.9E-03
Tgm2	P21981	Protein-glutamine γ-glutamyltransferase 2a	−0.609	3.9E-04
Fkbp5	Q64378	Peptidyl-prolyl cis-trans isomerase FKBP5	−0.810	3.6E-06
Fkbp1a	P26883	Peptidyl-prolyl cis-trans isomerase FKBP1Aa	0.610	1.0E-03
Fkbp4	P30416	Peptidyl-prolyl cis-trans isomerase FKBP4a	0.757	8.9E-04
Proteasome and related proteins
Cand1	Q6ZQ38	Cullin-associated NEDD8-dissociated protein 1	0.440	5.6E-04
Ecm29	Q6PDI5	Proteasome-associated protein ECM29 homolog	0.493	3.7E-03
Psmd1	Q3TXS7	26S proteasome non-ATPase regulatory subunit 1	p53d/d onlyc	—
Psmd12	Q9D8W5	26S proteasome non-ATPase regulatory subunit 12a	0.283	1.3E-03
Psmd2	Q8VDM4	26S proteasome non-ATPase regulatory subunit 2	0.490	2.8E-03
Psmd3	P14685	26S proteasome non-ATPase regulatory subunit 3	0.474	1.3E-04
Psmd8	Q9CX56	26S proteasome non-ATPase regulatory subunit 8	0.483	1.0E-03
Psme3	P61290	Proteasome activator complex subunit 3	0.390	8.3E-04
Trim28	Q62318	Transcription intermediary factor 1-βa	0.441	1.6E-03
Uba1	Q02053	Ubiquitin-like modifier-activating enzyme 1	0.474	7.6E-04
Ubc	Q9Z0H9	Ubiquitin	0.933	2.6E-03
Psma1	Q9R1P4	Proteasome subunit α type-1	−1.641	4.1E-03
Ubqln2	Q9QZM0	Ubiquilin-2	p53f/f onlyb	—
Usp14	Q9JMA1	Ubiquitin carboxyl-terminal hydrolase 14	p53f/f onlyb	—
Proteinases
Clpp	O88696	Putative ATP-dependent Clp protease proteolytic subunit, mitochondrial	−0.553	4.9E-04
Dpep1	P31428	Dipeptidase 1	−0.939	1.6E-03
Gzmb	P04187	Granzyme B(G,H)a	−0.903	1.7E-03
Thop1	Q8C1A5	Thimet oligopeptidase	−1.111	2.6E-03
Spcs2	Q9CYN2	Signal peptidase complex subunit 2	−0.553	8.9E-04
Xpnpep1	Q6P1B1	Xaa-Pro aminopeptidase 1	−0.443	1.3E-05
Capns1	O88456	Calpain small subunit 1	0.703	2.4E-03
Npepps	Q11011	Puromycin-sensitive aminopeptidase	0.195	2.6E-03
Scpep1	Q920A5	Retinoid-inducible serine carboxypeptidasea	0.592	6.6E-04
Proteinases inhibitors
A2mp	Q6GQT1	α-2-Macroglobulin-Pa	−1.999	7.7E-13
Cstb	Q62426	Cystatin-Ba	−1.000	1.0E-03
Serpina3k	P07759	Serine protease inhibitor A3K	−1.109	4.6E-04
Serpinb6	Q60854	Serpin B6a	−0.391	7.1E-04
Serpinc1	P32261	Antithrombin-III	−0.459	2.0E-05

^aCharacterized by IMS-MS and Velos instruments (average values used for fold change and P values).

^bp53^f/f only, p53^d/d samples do not show any signal.

^cp53^d/d only, p53^f/f samples do not show any signal.

Table 3.

Metabolism

UniProt Gene ID	UniProt accession no.	Protein name	log2 (p53d/d/p53f/f)	P value
Lipid metabolism and related proteins
Acot11	Q8VHQ9	Acyl-coenzyme A thioesterase 11	−0.701	1.8E-04
Ech1	O35459	δ(3,5)-δ(2,4)-dienoyl-CoA isomerase, mitochondrial	−0.654	2.3E-03
Pafah1b2	Q61206	Platelet-activating factor acetylhydrolase IB subunit βa	−0.605	2.7E-04
Psap	Q61207	Sulfated glycoprotein 1a	−0.622	5.7E-05
Sgpl1	Q8R0 × 7	Sphingosine-1-phosphate lyase 1a	−0.533	1.6E-04
Apoa1	Q00623	Apolipoprotein A-Ia	0.672	1.4E-03
Apoe	P08226	Apolipoprotein E	0.952	6.9E-04
Mitochondrial ATP production (fatty acid β-oxidation)
Acaa2	Q8BWT1	3-Ketoacyl-CoA thiolase, mitochondriala	−0.698	1.8E-06
Acadvl	P50544	Very long-chain specific acyl-CoA dehydrogenase, mitochondrial	−0.515	2.9E-03
Dbi	P31786	Acyl-CoA-binding proteina	−0.538	2.5E-04
Hadh	Q61425	Hydroxyacyl-coenzyme A dehydrogenase, mitochondrial	−0.725	4.6E-03
Hadha	Q8BMS1	Trifunctional enzyme subunit α, mitochondrial	p53f/f onlyb	—
Hadhb	Q99JY0	Trifunctional enzyme subunit β, mitochondrial	−0.724	2.3E-03
Mitochondrial ATP production (electron transport chain)
Atp5c1	Q91VR2	ATP synthase subunit γ, mitochondrial	−0.568	2.4E-04
Atp5i	Q06185	ATP synthase subunit e, mitochondrial	−0.417	1.0E-03
Cox5b	P19536	Cytochrome c oxidase subunit 5B, mitochondriala	−1.105	2.6E-03
Cox7a2	P48771	Cytochrome c oxidase subunit 7A2, mitochondrial	−0.609	3.8E-05
Cyb5a	P56395	Cytochrome b5	−0.931	3.9E-05
Etfa	Q99LC5	Electron transfer flavoprotein subunit α, mitochondrial	−0.327	5.1E-04
Mtco2	P00405	Cytochrome c oxidase subunit 2	p53f/f onlyb	—
Mtnd5	P03921	NADH-ubiquinone oxidoreductase chain 5	−0.837	1.9E-03
Cisd1	Q91WS0	CDGSH iron sulfur domain-containing protein 1	0.622	4.7E-03
Mitochondrial ATP production (tricarboxylic acid cycle)
Aco2	Q99KI0	Aconitate hydratase, mitochondrial	−0.597	1.7E-03
Dld	O08749	Dihydrolipoyl dehydrogenase, mitochondrial	−0.857	2.3E-03
Idh1	O88844	Isocitrate dehydrogenase (NADP) cytoplasmica	−0.635	9.4E-04
Idh2	P54071	Isocitrate dehydrogenase (NADP), mitochondriala	−0.716	2.7E-03
Idh3g	P70404	Isocitrate dehydrogenase (NAD) subunit-γ, mitochondrial	−0.809	2.7E-03
Mdh2	P08249	Malate dehydrogenase, mitochondrial	−0.532	4.7E-05
Suclg1	Q9WUM5	Succinyl-CoA ligase (GDP-forming) subunit α, mitochondrial	−1.231	1.6E-03
Other metabolism
Acat1	Q8QZT1	Acetyl-CoA acetyltransferase, mitochondrial	−0.609	2.3E-03
Adh5	P28474	Alcohol dehydrogenase class-3	−0.398	1.3E-04
Akr1b1	P45376	Aldose reductase	−1.016	3.0E-03
Aldh9a1	Q9JLJ2	4-Trimethylaminobutyraldehyde dehydrogenase	−0.661	1.6E-03
Asl	Q91YI0	Argininosuccinate lyasea	−0.693	3.4E-05
Ckb	Q04447	Creatine kinase B-type	−0.713	2.9E-03
Cyb5r3	Q9DCN2	NADH-cytochrome b5 reductase 3	−0.378	2.8E-03
Dck	P43346	Deoxycytidine kinase	−0.836	3.6E-03
Dhcr24	Q8VCH6	24-Dehydrocholesterol reductasea	−0.888	1.9E-05
Es1	P23953	Liver carboxylesterase Na	−0.781	1.0E-04
Gatm	Q9D964	Glycine amidinotransferase, mitochondriala	−0.499	5.1E-04
Gda	Q9R111	Guanine deaminase	−0.961	4.1E-04
Glul	P15105	Glutamine synthetasea	−0.953	2.9E-04
Mpi	Q924M7	Mannose-6-phosphate isomerasea	−0.495	3.3E-04
Nampt	Q99KQ4	Nicotinamide phosphoribosyltransferase	−0.300	2.4E-04
Nudt5	Q9JKX6	ADP-sugar pyrophosphatasea	−0.404	4.4E-04
Oat	P29758	Ornithine aminotransferase, mitochondrial	−1.362	3.3E-03
Pik3c3	Q6PF93	Phosphatidylinositol 3-kinase catalytic subunit type 3	−0.754	2.4E-03
Ahcy	P50247	Adenosylhomocysteinase	0.220	1.8E-04
Ak1	Q9R0Y5	Adenylate kinase isoenzyme 1a	0.687	1.1E-04
Aldh1a2	Q62148	Retinal dehydrogenase 2	0.601	2.0E-05
Fmo1	P50285	Dimethylaniline monooxygenase (N-oxide-forming) 1	1.227	3.8E-03
Gc	P21614	Vitamin D-binding protein	0.872	1.7E-03
Gpd1l	Q3ULJ0	Glycerol-3-phosphate dehydrogenase 1-like protein	0.373	3.6E-03
Hdlbp	Q8VDJ3	Vigilin	0.374	2.5E-03
Prpsap1	Q9D0M1	Phosphoribosyl pyrophosphate synthase-associated protein 1	0.463	1.5E-03
Sqrdl	Q9R112	Sulfide:quinone oxidoreductase, mitochondriala	0.800	5.0E-07
Tdo2	P48776	Tryptophan 2,3-dioxygenasea	0.455	2.6E-03
Umps	P13439	Uridine 5′-monophosphate synthase	0.532	1.3E-04

^aCharacterized by IMS-MS and Velos instruments (average values used for fold change and P values).

^bp53^f/f only, p53^d/d samples do not show any signal.

Table 4.

Nuclear proteins

UniProt Gene ID	UniProt accession no.	Protein name	log2 (p53d/d/p53f/f)	P value
Association to DNA
H2afy	Q9QZQ8	Core histone macro-H2A.1	−1.232	6.3E-04
Hist1h1b	P43276	Histone H1.5a	−0.485	2.3E-03
Hist1h4a	Q5T006	Histone H4	−0.892	1.1E-03
Hmgn2	P09602	Nonhistone chromosomal protein HMG-17	−0.703	1.5E-03
Nap1l4	Q78ZA7	Nucleosome assembly protein 1-like 4	0.343	2.1E-03
DNA replication
Mcm5	P49718	DNA replication licensing factor MCM5	0.357	1.8E-04
Mcm6	P97311	DNA replication licensing factor MCM6	0.376	1.1E-03
Pcna	P17918	Proliferating cell nuclear antigen	0.531	3.7E-03
Nuclear import
Kpna2	P52293	Importin subunit α-2	0.352	7.6E-04
Kpnb1	P70168	Importin subunit β-1	0.476	1.9E-03
Ipo9	Q91YE6	Importin-9	0.630	1.0E-05
Ipo7	Q9EPL8	Importin-7	0.776	4.7E-04
Mvp	Q9EQK5	Major vault protein	0.605	2.7E-04

^aCharacterized by IMS-MS and Velos instruments (average values used for fold change and P values).

Table 5.

Other proteins

UniProt Gene ID	UniProt accession no.	Protein name	log2 (p53d/d/p53f/f)	P value
Expressed exclusively in decidua
Prl3c1	Q9QUN5	Prolactin-3C1a	−1.974	9.2E-09
Transport
Anxa1	P10107	Annexin A1	−1.227	1.2E-03
Ap1b1	O35643	AP-1 complex subunit β-1	−0.459	2.3E-03
Atl3	Q91YH5	Atlastin-3a	−0.984	2.2E-03
Cp	Q61147	Ceruloplasmina	−0.653	1.3E-03
Fabp4	P04117	Fatty acid-binding protein, adipocyte	−1.090	8.3E-07
Lman2	Q9DBH5	Vesicular integral-membrane protein VIP36	−0.362	3.0E-03
Rab8a	P55258	Ras-related protein Rab-8A	−0.789	3.5E-03
Sec23b	Q9D662	Protein transport protein Sec23B	−1.085	1.8E-05
Sec61a1	P61620	Protein transport protein Sec61 subunit α isoform 1	−0.521	3.5E-03
Sec61b	Q9CQS8	Protein transport protein Sec61 subunit β	−0.731	1.1E-04
Tom1	O88746	Target of Myb protein 1	−0.375	3.4E-03
Tspo	P50637	Translocator protein	−0.653	2.8E-03
Vapa	Q9WV55	Vesicle-associated membrane protein-associated protein A	−1.014	1.2E-04
Copb1	Q9JIF7	Coatomer subunit β	0.611	4.4E-03
Timm44	O35857	Mitochondrial import inner membrane translocase subunit TIM44	1.502	1.6E-03
Uso1	Q9Z1Z0	General vesicular transport factor p115	0.291	4.0E-03
Cell adhesion
Emilin1	Q99K41	EMILIN-1	0.748	3.8E-03
Fermt2	Q8CIB5	Fermitin family homolog 2	0.435	9.4E-06
Fn1	P11276	Fibronectina	0.539	1.3E-03
Ilk	O55222	Integrin-linked protein kinase	0.371	1.0E-03
Lama5	Q61001	Laminin subunit α-5a	0.604	5.0E-04
Lamb2	Q61292	Laminin subunit β-2a	0.601	4.4E-04
Tgfbi	P82198	Transforming growth factor-β-induced protein ig-h3	1.006	2.7E-03
Tjp1	P39447	Tight junction protein ZO-1	0.341	3.7E-03
Cdh3	P10287	Cadherin-3	−1.201	3.4E-03
Pecam1	Q08481	Platelet endothelial cell adhesion molecule	−0.517	9.5E-04
Cytoskeleton
Actn1	Q7TPR4	α-Actinin-1	0.580	3.5E-07
Actn4	P57780	α-Actinin-4a	0.480	1.4E-03
Add3	Q9QYB5	γ-Adducin	0.455	2.3E-03
Anxa3	O35639	Annexin A3	0.265	3.3E-03
Cap1	P40124	Adenylyl cyclase-associated protein 1a	0.571	4.9E-05
Col14a1	Q80 × 19	Collagen α-1(XIV) chaina	0.491	1.1E-03
Col1a1	P11087	Collagen α-1(I) chaina	0.787	1.4E-05
Csrp1	P97315	Cysteine and glycine-rich protein 1	0.427	1.2E-04
Dctn2	Q99KJ8	Dynactin subunit 2	0.362	4.1E-03
Dstn	Q9R0P5	Destrin	0.594	1.0E-03
Krt18	P05784	Keratin, type I cytoskeletal 18	0.773	1.1E-03
Lima1	Q9ERG0	LIM domain and actin-binding protein 1	0.492	5.9E-04
Map4	Q8CFP5	Microtubule-associated protein 4	0.396	5.0E-04
Myh10	Q61879	Myosin-10a	0.598	1.3E-04
Myh9	Q8VDD5	Myosin-9a	0.429	4.3E-04
P4ha1	Q60715	Prolyl 4-hydroxylase subunit α-1a	0.971	5.7E-04
Parva	Q9EPC1	α-Parvin	0.702	3.3E-04
Plod2	Q9R0B9	Procollagen-lysine,2-oxoglutarate 5-dioxygenase 2a	0.750	3.1E-04
Serpinh1	P19324	Serpin H1a	0.442	1.1E-04
Tpm4	Q6IRU2	Tropomyosin α-4 chaina	0.579	5.1E-05
Ugdh	O70475	UDP-glucose 6-dehydrogenase	0.791	1.2E-03
Cnn3	Q9DAW9	Calponin-3	−0.291	4.3E-03
Hspg2	Q05793	Basement membrane-specific heparan sulfate proteoglycan core proteina	−0.812	1.5E-03
Pfn1	P62962	Profilin-1a	−0.515	2.2E-03
Thy1	P01831	Thy-1 membrane glycoprotein	−0.929	3.3E-03

^aCharacterized by IMS-MS and Velos instruments (average values used for fold change and P values).

Western blotting

Protein extraction and Western blotting were performed as described (27). Antibodies to PRDX6 (LabFrontier Co., Seoul, Korea) and ACTIN (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were used. ACTIN served as a loading control.

Reverse transcription-polymerase chain reaction

RT-PCR was performed as described using primers 5′-TCATGGGGCATTCTCTTTTC-3′ and 5′-AAGGTCCCTGCCCTTATCAT-3′ for Prdx6 (product size 237 bps), 5′-GACCTTGTTGTTGACAAGGACTTA-3′ and 5′-CCTGAATGTACAGTAGAGGAATCA-3′ for A2mp (product size 277 bps), and 5′-GTGGGCCGCCCTAGGCACCAG-3′ and 5′- CTCTTTGATGTCACGCACGATTTC-3′ for β-Actin (product size 540 bps) (27). The housekeeping gene β-Actin served as an internal control.

Results

MS characterizes p53-regulated proteins in deciduae

The heterogeneous population of decidual cell types, presumably with different functions, made quantitation sensitivity very important. Because different cell types are difficult to distinguish, we homogenized the whole decidua and diluted protein expression changes that represented only few cells. For increased protein coverage and quantitative information, MS analysis of decidual tissues was performed with both a LTQ Orbitrap Velos MS (Velos) and new IMS-MS platform. These analyses identified a total of 283 proteins (Supplemental Table 5) that showed different levels between d 8 deciduae of p53^d/d and p53^f/f females; 183 of these proteins are represented in the functional categories of Tables 1–5. A reliable overlap was observed between platforms; 76 proteins were found significant by both Velos and IMS-MS (Supplemental Table 5). As expected, IMS-MS illustrated more significant proteins than the Velos alone (15). Figure 1 shows the intensity changes of IMS-MS spectra for selected significant peptides that are down-regulated (Fig. 1, A and B) and up-regulated (Fig. 1C) in p53^d/d deciduae compared with those in p53^f/f deciduae. Tables 1–5 show 103 down-regulated proteins and 80 up-regulated proteins in the p53^d/d deciduae. The P values shown in these tables were calculated using all instrument runs and biological variability among six mice (three each of p53^d/d and p53^f/f) and are depicted by average protein intensities for the IMS-MS subset of the data (Fig. 2). We analyzed protein functions using UniProtKB, GO annotation, and KEGG information (Supplemental Tables 5–9) and categorized proteins into 18 functional pathways (Tables 1–5 and Supplemental Tables 5–9). In the p53^d/d deciduae, most protein species were down-regulated in the following 12 pathways: antioxidant enzymes, posttranscriptional modifications, proteinases, proteinase inhibitors, lipid metabolism, fatty acid β-oxidation, electron transport chain, tricarboxylic acid cycle, metabolism, association to DNA, transporter, and known molecules previously found to be associated with decidualization. Proteins mostly up-regulated in the p53^d/d deciduae are associated with six pathways, including stress-induced proteins, proteasome and related proteins, DNA replication, nuclear import, cell adhesion, and cytoskeleton. Of note, we found that prolactin-3C1, a known decidual marker (28), and platelet endothelial cell adhesion molecule 1 (29), an endothelial cell marker, are down-regulated in p53^d/d deciduae.

Fig. 1.

IMS-MS spectra of A2MP, PRDX6, and myosin-10 (MYH10) peptides. A, YIFIDESHITQALTWLSQQQK from A2MP. B, LIALSIDSVEDHLAWSK from PRDX6. The signal intensity located at the bottom left of each spectra denotes an increase in peptide intensity in the p53^f/f samples by 6.3- and 2.4-fold change for these A2MP and PRDX6 peptides, respectively. C, IMS-MS spectra for the significant peptide QLLQANPILESFGNAK from MYH10 showed a 2.1-fold intensity increase in the p53^d/d sample. Mouse identifiers WT_8718 and KO_7988 (see Supplemental Tables 1–9) were used as examples for this figure.

Fig. 2.

Disruption of p53 signaling adversely affects many cellular processes. Heat maps depict average protein intensities, calculated from multiple IMS-MS analyses, for each of the total six mice (three of p53^d/d and p53^f/f each).

DNA replication is up-regulated in p53^d/d deciduae

Decidualization is a process by which endometrial stromal cells undergo proliferation and differentiation after embryo implantation. We record weights of implantation sites as an index of the extent of decidualization. During decidualization, there are massive increases in cell numbers with heightened DNA synthesis in many cells leading to endoreduplication and polyploidy (11). In p53^d/d females, the number of decidual cells undergoing polyploidy with terminal differentiation and cellular senescence activity substantially increases. These changes are reflected in growth restriction, smaller decidual size, and preterm birth (2). In the present study, we found three proteins related to DNA replication- DNA replication licensing factor minichromosome maintenance complex components 5 and 6 and proliferating cell nuclear antigen are all up-regulated in p53^d/d deciduae compare with those in p53^f/f dams (Table 4).

Antioxidant enzymes are down-regulated in p53^d/d deciduae

Generation of ROS plays important roles during pregnancy. However, ROS must be balanced with antioxidant enzymes, because excessive ROS causes susceptibility to oxidative stress (OS), which results in cellular damage and senescence (17, 30). We have previously observed that increased polyploidy of stromal cells with and growth restriction in p53^d/d deciduae are associated with increased senescence (2), and OS is associated with preterm birth (31). We found that 10 antioxidant enzymes were down-regulated in p53^d/d deciduae compared with p53^f/f deciduae. Using Western blot and RT-PCR analyses, we confirmed that PRDX6 expression is lower in p53^d/d deciduae than p53^f/f deciduae (Fig. 3, A and B). These results suggest that p53^d/d deciduae with decreased expression of antioxidant enzymes are susceptible to OS, which in turn induces compromised decidualization with restricted decidual growth and premature senescence, ultimately triggering preterm birth in p53^d/d deciduae.

Fig. 3.

PRDX6 and A2MP are down-regulated in d 8 p53^d/d deciduae. A and B, Protein and mRNA levels of PRDX6 in d 8 deciduae of p53^d/d and p53^f/f mice were examined by Western blotting (A) and RT-PCR (B). C, mRNA level of A2mp, in d 8 deciduae in p53^d/d and p53^f/f mice was examined by RT-PCR. ACTIN and β-Actin served as loading controls.

Mitochondrial ATP production is down-regulated in p53^d/d deciduae

It is interesting to note that expression levels of 21 proteins in the mitochondrial ATP generation pathways were compromised in p53^d/d deciduae compared with those in p53^f/f females (Table 3). One major function of mitochondria is to generate ATP, and dysfunction of mitochondria accompanied by reduction in ATP production promotes senescence (32–38). Collectively, these findings suggest that reduced ATP generation resulting from mitochondrial dysfunction in p53^d/d deciduae induces premature senescence.

Proteinase inhibitors are down-regulated in p53^d/d deciduae

Some proteinases and proteinase inhibitors are known to play roles in pregnancy, such as trophoblast invasion (39). By proteomic analysis, we found that five proteinase inhibitors were down-regulated in p53^d/d deciduae compared with those in p53^f/f females (Table 2). Using RT-PCR analysis, we confirmed that the level of A2MP, induced in deciduae (18), was lower in p53^d/d deciduae compared with p53^f/f deciduae (Fig. 3C). Mouse A2mp, was cloned from mesometrial deciduae and is thought to be up-regulated by uterine natural killer cells to promote remodeling of the spiral artery to support decidualization (18, 19).

Discussion

Our recent observations have shown that uterine deletion of Trp53 increases the incidence of preterm birth resulting from premature decidual senescence with enhanced terminal differentiation of stromal cells and heightened mTORC1 signaling (2, 6). Mouse models for studying spontaneous preterm labor are very limited, because a drop in progesterone levels, unlike humans, is seen in most mouse models of preterm birth. p53^d/d mice serve as a novel model to study the underlying mechanism of spontaneous preterm birth without progesterone withdrawal, a model seems relevant in the pursuit for the cause of premature birth in humans. In addition, our observations of senescence-associated decidual growth restriction and preterm birth with conditional deletion of uterine Trp53 have discovered a new critical role of p53 in uterine biology and parturition involving uterine cyclooxygenase 2-prostagrandin F_2α signaling pathway, which is well known to be associated with preterm delivery. This is relevant to the observation that women at ages 35 or older are at higher risk for preterm birth (40), and p53 activity diminishes with ageing of mice (41).

Because many factors, including genetic alterations, OS, and inflammation/infection, can give rise to cellular senescence and preterm birth, we were interested in exploring other potential factors that are associated with these events by comparing decidual protein profiles in p53^f/f and p53^d/d females. Our findings of down-regulation of a cluster of antioxidant enzymes, including PRDX6, a unique member in the family, and a group of 21 proteins involved in ATP generation, are provocative and add new information to our repertoire. Because down-regulation of antioxidant enzymes provoking OS and mitochondrial dysfunction with reduced ATP generation are contributors to cellular senescence, we believe that preterm birth in p53^d/d females arise from several deregulated pathways converging to the senescence pathway. Our results also show that proteins involved in the DNA replication pathway are up-regulated in p53^d/d deciduae. This finding corroborates with our previous observation of decidual growth restriction with heightened stromal cell polyploidy and increased DNA content without cytokinesis in the absence of p53 (2).

Antioxidant enzymes help balance ROS generated under normal physiological conditions. Pregnancy is an example of a physiological system in which this balance is tightly regulated. Collapse of the balance due to the excessive generation of ROS imposing OS induces adverse effects on embryo development and pregnancy outcome. However, the mechanism by which a balance between ROS and antioxidant enzymes are achieved during pregnancy is not yet clearly understood (42–45). We have previously shown that FK506-binding protein 4 (FKBP52) mutant females with reduced uterine PRDX6 are susceptible to overt OS and show implantation failure. This failure can be overcome by providing mutant females with excess progesterone and antioxidant supplements (17). We also showed that induced expression of PRDX6 rescues H₂O₂-induced cell death in Fkbp52-deficient mouse embryonic fibroblasts with reduced PRDX6 levels. These results suggest that Fkbp52 deficiency diminishes the threshold against OS by reducing PRDX6 levels. In the current study, we again found that many antioxidant enzymes, including PRDX6, have levels that are lower in p53^d/d deciduae, suggesting higher OS levels. Because it was shown that OS activates mTORC1 signaling, one possibility is that loss of Trp53 induces higher OS by reducing antioxidant levels, thus leading to higher mTORC1 activity (46). Moreover, p53 transcriptionally induces the expression of antioxidant enzymes (46), implying that a decidual Trp53 deficiency might reduce antioxidant activity.

Mitochondria's one major function is to generate ATP, and it is thought that its dysfunction plays a role in senescence (32, 35, 36). It has also been shown that decreased F1-ATPase activity might be responsible for age-dependent mitochondrial dysfunction (35, 36). There is also evidence that ADP/ATP ratio is in parallel with declining mitochondrial functions during cellular senescence of human diploid fibroblasts, which exhibit a limited replicative life span (32). Our findings of down-regulation of mitochondrial proteins related to ATP production are consistent with their role in premature decidual senescence in p53^d/d mice. Notably, p53 localizes in mitochondria and directly implies positive effects on mitochondrial biogenesis and function (47, 48).

The family of A2M genes is evolutionarily conserved from lower to higher organisms and categorized as proteinase inhibitors. They are known to play a role in the immune system (49). Two isoforms of A2M, pregnancy zone proteins and A2MP, are expressed in mouse deciduae and are involved in decidualization (18, 50). We found that A2MP is down-regulated in p53^d/d deciduae compared with p53^f/f deciduae. Although most A2M are produced in the liver and are abundant in the blood, A2MP is instead locally found in the uterus and heart (18). Deciduae of Rag2/γC mutant mice without NK or T cells have reduced levels of A2MP and fail to remodel spiral arteries, but A2M supplementation rescues spiral artery remodeling (18). These data suggest that reduced levels of A2MP caused by Trp53 deletion may affect spiral artery remodeling in deciduae.

In conclusion, proteins related to DNA replication, antioxidation, mitochondrial ATP production, and proteinase inhibitors are potential downstream targets of p53 during decidualization. These findings add to our knowledge of repertoire of factors that could be involved in preterm birth and may open up new avenues to design better approaches to combat preterm birth and prematurity.