Progesterone Receptor-Mediated Regulation of N-Acetylneuraminate Pyruvate Lyase (NPL) in Mouse Uterine Luminal Epithelium and Nonessential Role of NPL in Uterine Function

N-acetylneuraminate pyruvate lyase (NPL) catalyzes N-acetylneuraminic acid, the predominant sialic acid. Microarray analysis of the periimplantation mouse uterine luminal epithelium (LE) revealed Npl being the most downregulated (35×) gene in the LE upon embryo implantation. In natural pregnant mouse uterus, Npl expression increased 56× from gestation day 0.5 (D0.5) to D2.5. In ovariectomized mouse uterus, Npl was significantly upregulated by progesterone (P4) but downregulated by 17β-estradiol (E2). Progesterone receptor (PR) antagonist RU486 blocked the upregulation of Npl in both preimplantation uterus and P4-treated ovariectomized uterus. Npl was specifically localized in the preimplantation D2.5 and D3.5 uterine LE. Since LE is essential for establishing uterine receptivity, it was hypothesized that NPL might play a critical role in uterine function, especially during embryo implantation. This hypothesis was tested in the Npl (−/−) mice. No significant differences were observed in the numbers of implantation sites on D4.5, gestation periods, litter sizes, and postnatal offspring growth between wild type (WT) and Npl (−/−) females from mating with WT males. Npl (−/−)xNpl (−/−) crosses produced comparable little sizes as that from WTxWT crosses. Comparable mRNA expression levels of several genes involved in sialic acid metabolism were observed in D3.5 uterus and uterine LE between WT and Npl (−/−), indicating no compensatory upregulation in the D3.5 Npl (−/−) uterus and LE. This study demonstrates PR-mediated dynamic expression of Npl in the periimplantation uterus and dispensable role of Npl in uterine function and embryo development.


Introduction
N-acetylneuraminate pyruvate lyase (NPL), also named sialic acid aldolase or N-acetylneuraminate lyase, was originally purified and characterized in human related pathogenic as well as nonpathogenic bacteria that utilize the carbon sources in the mucusrich surfaces of the human body, such as Clostridium perfringens and Escherichia coli [1][2][3][4]. The human NPL consists of 320 amino acids (33 kDa) and has a crystal structure of tetramer [5,6]. Mammalian NPL proteins have 86 highly conserved amino acids, which are slightly different from the bacterial counterpart [6]. A splice variant of human NPL is highly expressed in human liver, kidney, ovary, and peripheral blood leukocyte [7].
NPL catalyzes the breaking of carbon-carbon bonds of Nacetylneuraminic acid, the predominant sialic acid, into Nacetylmannosamine and pyruvate, thus regulating the cellular concentrations of sialic acid and preventing the recycling of sialic acid for further sialiation with glycoconjuates in the Golgi compartment [8,9]. In both bacteria and mammalians, sialic acids such as N-acetylneuraminic acid (Neu5Ac) and N-glycolylneur aminic acid (Neu5Gc) are involved in the sialylation and provide the diversity of sialylated oligosaccharides [10][11][12]. Sialic acids have been associated with intercellular adhesion, protein recognition, and immune-related mechanisms [13,14].
Limited reports suggest that NPL may have functions in female reproduction but the potential role of NPL in female reproduction has not been previously investigated. For example, serum sialic acid level increases with the progression of pregnancy compared with that in non-pregnant women [15][16][17]; uterine sialic acid concentration decreases upon ovariectomy in Indian langur monkeys, but increases upon ovarian hormones E2 or E2+P4 treatments [18]. NPL came to our attention from our microarray analysis of mouse periimplantation uterine luminal epithelium (LE) (GEO number: GSE44451). Npl was the most downregulated (356) gene in the postimplantation gestation day 4.5 (D4.5) LE compared with that in the preimplantation D3.5 LE (Xiao et al, submitted). Further analysis indicated peak expression of Npl in the preimplantation D2.5 and D3.5 uterine LE. Since LE is critical for the receptive sensitivity of the uterus [19,20], we hypothesized that NPL might be involved in uterine function, especially uterine preparation for embryo implantation. This hypothesis was tested in Npl (2/2) mice.

Animals and genotyping
Npl (2/2) mice were generated from the mouse strain B6/129S5-Npl Gt(IRESBetageo)332Lex /Mmucd (identification number 011743-UCD) and purchased from the Mutant Mouse Regional Resource Center (MMRRC) at UC Davis, a NCRR-NIH funded strain repository. Npl (2/2) mice were genotyped using tail genomic DNA and three primers in PCR reactions: Primer 0920-59: GGCATA-TATGTGCAGGCAGAATGC, Primer LTR-rev: ATAAACCC-TCTTGCAGTTGCATC, and Primer 0920-39: TCTAGAAAT-GAGTCTGAACCGGAC. The genotyping PCR cycles were: 10 cycles of 94uC for 15s, 65uC for 30 s (decreased 1uC/cycle), 72uC for 40 s; and 30 cycles of 94uC for 15 s, 55uC for 30 s and 72uC for 40 s. The expected PCR product sizes for wild type (WT) (Primer 0920-59 and Primer 0920-39) and Npl (2/2) (Primer 0920-59 and Primer LTR-rev) were 115 bp and 163 bp, respectively. All mice were housed in polypropylene cages with free access to regular food and water from water sip tubes in a reverse osmosis system. The animal facility is on a 12-hour light/dark cycle (7:00 AM to 7:00 PM) at 2361uC with 30-50% relative humidity. All methods used in this study were approved by the University of Georgia Institutional Animal Care and Use Committee (IACUC) and conform to the National Institutes of Health guidelines and public law. All the animal studies were summarized in Table 1.

Mating and uterine tissue collection
Young virgin females were mated naturally with WT stud males and checked for a vaginal plug the next morning. The day a vaginal plug identified was designated as gestation day 0.5 (D0.5, mating night as D0). Uterine tissues were collected from euthanized females between 11:00 h and 12:00 h on D0.5, D1.5, D2.5, D3.5, D4.5, D5.5, and D7.5, respectively. Uterine horns from D0.5 to D2.5 females were quickly removed and snapfrozen on dry ice. Oviducts from these mice were flushed with 16PBS for the presence of eggs or fertilized embryos to determine the pregnancy status. About 1/3 of a uterine horn from each euthanized D3.5 female was frozen on dry ice for tissue sectioning. The remaining D3.5 uterine horns were flushed with 16PBS (to determine the status of pregnancy and to remove the influence of embryos on uterine gene expression) and frozen on dry ice for RNA isolation. On D4.5, D5.5, or D7.5, mice were anesthetized with isoflurane by inhalation and intravenously (i.v.) injected with Evans blue dye to visualize the implantation sites as previously described [21]. At least three pregnant mice were included in each group.

Hormonal treatment
Progesterone (P4), 17b-estradiol (E2), ICI 182780 (ER antagonist), or RU486 (PR antagonist) treatments on ovariectomized WT mice or early pregnant WT mice were done as previously reported [22][23][24]. Briefly, the ovariectomized 6 weeks old virgin WT females (recovered for 2 weeks after surgery) were s.c. injected with 0.1 ml sesame oil (vehicle group) or 0.1 ml 20 mg/ml P4 three times on 0 h, 24 h and 48 h, respectively. In the E2-treated group, the ovariectomized mice were injected with 0.1 ml oil on 0 h and 24 h, then 0.1 ml 1 mg/ml E2 on 48 h. In the P4+E2treated group, the ovariectomized mice were treated the same as the P4 treated group except an additional injection of 0.1 ml 1 mg/ml E2 on 48 h. All injected mice were dissected 6 hours after

LE isolation
D3.5 uteri from naturally mated WT and Npl (2/2) mice were processed for LE isolation as previously described using 0.5% dispase enzyme and gentle scraping [22]. The pregnancy status was determined by the presence of blastocyst(s). At least five pregnant mice were included in each group.

Statistical analyses
Two-tail unequal variance Student's t-test and one-way ANOVA with Dunnett-t test were used for various comparisons. The significant level was set at p,0.05.

Differential expression of Npl in periimplantation mouse uterus
To determine the spatiotemporal uterine expression of Npl during early pregnancy, the mRNA expression of Npl was examined in the periimplantation uterus by realtime PCR and in situ hybridization. Realtime PCR indicated that Npl was expressed at a very low level in the D0.5 uterus; it was increased 46 in the D1.5 uterus and 566 in the D2.5 uterus; its expression level was slightly lower (446) in the D3.5 uterus compared to that in the D2.5 uterus; upon embryo implantation, Npl expression level in the D4.5 uterus returned to a level comparable to that of D1.5 uterus (Fig. 1A). In situ hybridization didn't detect any significant Npl signal in the D0.5 and D1.5 uterus (Figs. 1B, 1C). Npl was exclusively detected in the uterine luminal epithelium (LE) on D2.5 and D3.5 with comparable intensity (Figs. 1D, 1E) but undetectable in the postimplantation uterus from D4.5 to D7.5 ( Fig. 1F and data not shown). One study indicated that Npl expression is .56 lower in the LE at the implantation site than that in the LE of inter-implantation site on D4.5 [27]. However, in situ hybridization on longitudinal sections of D4.5 uterus did not detect any Npl signal in the inter-implantation site (data not shown). It suggests that the Npl expression levels in the D4.5 LE from implantation site and inter-implantation site are too low to be detected by in situ hybridization, similar as that in the D0.5 and D1.5 uterus (Fig. 1A). The upregulation of Npl in the preimplantation LE was consistent with microarray (GEO number: GSE44451) and realtime PCR results (Fig. 1A).

PR-mediated upregulation of Npl in preimplantation mouse uterus
Uterine gene expression is largely controlled by ovarian hormones P4 and E2, whose functions are mediated via their receptors PR and ER, respectively [28,29]. To determine the molecular mechanism for Npl upregulation in the preimplantation uterus, D2.5 WT females were treated with PR antagonist RU486 or ER antagonist ICI 182780 and Npl expression level was analyzed in the D3.5 uterus by realtime PCR. No significant difference in Npl expression was observed between ICI 182780treated and vehicle-treated groups ( Fig. 2A). However, dramatically reduced Npl expression was observed in the RU486-treated group (536 compared to vehicle-treated control) ( Fig. 2A). The expression of the house-keeping gene Hprt1 was not changed upon ICI 182780 or RU486 treatments (Fig. 2A). These results indicated that PR mediated the upregulation of Npl in the preimplantation uterus.

Hormonal regulation of Npl in ovariectomized mouse uterus
In ovariectomized WT mice, Npl was significantly upregulated (456) by P4 treatment and downregulated (66) by E2 treatment in the uterus (Fig. 2B). P4-induced Npl upregulation was greatly reduced by co-administration of E2, although the expression level of Npl was still 3-fold higher than that in the vehicle-treated uterus (Fig. 2B). In situ hybridization revealed that P4-induced upregulation of Npl was also detected in the LE of the ovariectomized uterus (data not shown), similar as that in the preimplantation uterus (Figs. 1D, 1E). To determine the involvement of PR in the regulation of Npl in the ovariectomized uterus, another set of experiment was performed. The results indicated that P4-induced upregulation of Npl in the ovariectomized uterus was completely abolished by co-administration of RU486, whereas RU486 alone did not seem to affect Npl expression (Fig. 2C). The expression of the house-keeping gene Hprt1 was not changed upon different hormonal treatments (Fig. 2). These data demonstrated that P4-PR signaling mediated the upregulation of Npl in the mouse uterus.
The coordinated uterine regulation of Npl by both P4 and E2 could explain the temporal expression of Npl in the early pregnant  (Fig. 1A). After D1.5, P4 secretion from the newly formed corpus luteum increases [30] and correspondingly, Npl expression levels increase in the D2.5 and D3.5 uterus (Fig. 1). On D3.5, superimposed ovarian estrogen secretion, which makes the P4primed uterus receptive for embryo implantation [30,31], may contribute to the downregulation of Npl (Figs. 1A, 1F, 2B). However, since ER antagonist ICI 182780 does not significantly affect the expression of Npl in the preimplantation uterus and PR antagonist RU486 dramatically suppresses the expression of Npl in the preimplantation uterus ( Fig. 2A), it is more likely that the downregulation of Npl in the postimplantation LE (Fig. 1F) is the consequence of the downregulation of PR in the postimplantation LE [26].
Since NPL catalyzes the dominant sialic acid N-acetylneuraminic acid, it is expected that its downregulation could lead to elevated sialic acid in the uterus. The downregulation of Npl expression upon E2 treatment in the ovariectomized mouse uterus (Fig. 2B) seems to agree with the increased uterine sialic acid concentration upon E2 treatment in the ovariectomized Indian langur monkeys [18].
LE is the first cellular layer that an implanting embryo communicates with for implantation. Considering the observations that Npl is the most dramatically differentially expressed gene in the periimplantation LE ( Fig. 1) (Xiao S et al, submitted) and Npl is upregulated in the preimplantation uterus via P4-PR signaling (Fig. 2), we hypothesized that Npl might play a role in uterine function, especially uterine preparation for embryo implantation. This hypothesis was tested in the Npl (2/2) mice and appeared to be supported by the preliminary fertility data on MMRRC website, which indicated that the average litter size from Npl (2/2) females (mated with WT males) was 1.6761.53 (N = 3), significantly smaller than that from WT females (mated with Npl (2/2) males) with an average litter size of 7.6762.08 (N = 3, P,0.05) (http:// www.informatics.jax.org/external/ko/lexicon/2492.html).

Normal embryo implantation and postimplantation pregnancy in Npl (2/2) females
Embryo implantation initiates around D4.0 in mice [26]. On D4.5, all the pregnant Npl (2/2) females had implantation sites detected by blue dye injection as seen in the WT females (Figs. 3A,  3B). The intensity and spacing of the blue bands, which indicated the implantation sites, and the average numbers of implantation sites between WT and Npl (2/2) females were comparable (Figs. 3A-3C). These data indicated no obvious defect in embryo implantation, which was confirmed by the comparable uterine expression of Prap1 and Abp1 (Figs. 4D-4G), a uterine LE marker [25] and a decidualization marker [32] upon embryo implantation, respectively. These results also indicated that all the preimplantation events, including oogenesis, ovulation, fertilization, embryo transport, and preimplantation embryo development, were not impaired. Postimplantation decidualization was also well developed in the D5.5 and D7.5 Npl (2/2) uteri, demonstrated by the comparable expression of a decidualization marker Dtprp [33] in D5.5 and D7.5 WT and Npl (2/2) uterus (Figs. 3H-3K). In fact, no obvious defect was detected in the Npl (2/2) females during the entire pregnancy, revealed by the comparable gestation periods, litter sizes, survival rates, and postnatal growth of offspring from females with different genotypes when they were mated with WT males (Figs. 4A, 4B, and data not shown). These results proved our hypothesis wrong and didn't support the preliminary fertility data reported in the MMRRC website (http://www.informatics.jax.org/external/ko/lexicon/ 2492.html).

Non-essential role of Npl in male fertility and embryo development
The WT and Npl (2/2) males had comparable testis weight, sperm counts from cauda epididymis, and litter sizes when they were mated with WT females (data not shown), indicating normal fertility of Npl (2/2) males.
When Npl (+/2) or Npl (2/2) females were mated with Npl (2/2) males, they produced comparable litter sizes to that from WTxWT crosses (Fig. 4B). These data demonstrated that deletion of Npl did not have an obviously adverse effect on embryo development.
No compensatory mRNA expression of other genes involved sialic acid metabolism in D3.5 Npl (2/2) uterus The following genes are known to play roles in sialic acid metabolism: Gne, Nans, and Nanp, which are important for the sialic acid synthesis from the UDP-N-acetylglucosamine (UDP-GlcNAc) to Neu5Ac in cytosol; Cmas and Cmah, which catalyze Neu5AC to CMP-Neu5Ac and then Neu5Gc; Slc35a1, which transports CMP-Neu5Ac and CMP-Neu5Gc to Golgi compartment for further glycosylation; St3gal1 and St3gal4, which are sialytransferases controlling the glycosylation of sialic acid with carbohydrates, glycoproteins, and glycolipids; Neu1 and Neu3, which are neuraminidases responsible for the removal of sialic acid residues from glycoconjugates in different intracellular compartments, such as lysosome, plasma membrane, and mitochondria; Slc17a5, which transports free sialic acid back to the cytosol; and Npl, which degrades the free sialic acid to N-acetylmannosamine and pyruvate in the cytosol [7,34].
Since Npl expression peaks in the preimplantation uterine LE (Fig. 1), the mRNA expression levels of the above mentioned genes involved in sialic acid metabolism were examined in the preimplantation D3.5 WT and Npl (2/2) uteri by realtime RT-PCR. The results showed comparable mRNA expression levels of all these genes between WT and Npl (2/2) uteri (Fig. 5A). Since Npl is an LE-specific gene (Fig. 1) and LE comprises ,10% of the uterine cells [19,20,35], any compensatory mRNA changes of these genes in the LE could potentially be covered in the whole uterine gene expression analysis. Therefore, LE cells from D3.5 WT and Npl (2/2) uteri were isolated for determining the mRNA expression of these genes. Realtime RT-PCR still failed to detect any significant difference of these genes between D3.5 WT and Npl (2/2) LE (Fig. 5B). Npl was included in both uterine and LE analyses (Figs. 5A, 5B) to indicate the deletion of Npl in the Npl (2/ 2) uterus. Based on the relative expression levels compared to the house-keeping gene Gapdh, Npl was enriched in the LE (Fig. 5), consistent with the in situ data (Fig. 1E). These results indicated no compensatory mRNA expression of these genes involved in sialic acid metabolism in the D3.5 Npl (2/2) whole uterus (Fig. 5A) and LE (Fig. 5B).
It has been reported that disruption of sialic acid metabolism and transport could have adverse effects. In mice, inactivation of Gne, which is important for sialic acid synthesis in the cytosol, leads to early postnatal lethality [36]. In humans, mutations of Slc17a5, which transports the free sialic acid from lysosome to cytosol for further degradation, could lead to the sialic acid storage disease (SASD) caused by sialic acid accumulation in the lysosome, and developmental delays and growth retardation [37]. In bacterial species, mutations of Npl, which degrades sialic acid in the cytosol, could lead to toxic overexpression of sialic acid [38]. These results indicate that balanced sialic acid metabolism and compartmentalization are critical for normal physiological functions.
Normal embryo implantation in the Npl (2/2) females ( Fig. 3) indicates that NPL is not essential for embryo implantation. However, being the most downregulated gene in the postimplantation LE implies that it might have redundant, although nonessential, roles in uterine preparation for embryo implantation. Our unpublished microarray data demonstrated that Npl expression was significantly decreased in the pregnant mouse uterus compared to that in the pseudopregnant mouse uterus at 22:00 h on D3.5 (data not shown), right before embryo attachment to the LE for implantation indicated by blue dye reaction [26], suggesting that downregulation of Npl might contribute to the initiation of embryo attachment. Since NPL degrades sialic acid that is involved in cell adhesion [39], it is possible that Figure 5. Deletion of Npl on the expression of sialic acid metabolism related genes in D3.5 uterus (A) and isolated LE (B), respectively. +/+, wild-type; 2/2, Npl (2/2) ; X-axis indicated the names of the examined genes. Y-axis showed the normalized mRNA expression levels by Gapdh (glyceraldehyde 3-phosphate dehydrogenase) x10 3 . Hprt1 (hypoxanthine phosphoribosyltransferase 1) served as the second housekeeping gene. N = 5-6. Error bars, standard deviation. doi:10.1371/journal.pone.0065607.g005 downregulation of Npl could potentially facilitate embryo attachment to the LE for embryo implantation. On the other hand, sialic acid can block the access of antigenic molecules to the cell surface [40] while NPL from C. perfringens can dramatically increase (256) the capacity of B cell antigen presentation [41]. It is possible that NPL might be involved in modulating the uterine immune response during early stages of embryo implantation [42].
In summary, this study demonstrates PR-mediated spatiotemporal expression of Npl in the periimplantation mouse uterus and the nonessential role of Npl in uterine function and embryo development.