Stanniocalcin-2 Inhibits Mammalian Growth by Proteolytic Inhibition of the Insulin-like Growth Factor Axis*

Background: The biological function and biochemical activity of mammalian stanniocalcin-2 are unknown. Results: Stanniocalcin-2 inhibits proteolytic release of insulin-like growth factor (IGF), and its ability to cause growth retardation upon transgenic overexpression in mice depends on its proteinase inhibitory function. Conclusion: Stanniocalcin-2 is a novel component of the IGF axis. Significance: Altered stanniocalcin-2 expression may affect IGF signaling under pathological conditions. Mammalian stanniocalcin-2 (STC2) is a secreted polypeptide widely expressed in developing and adult tissues. However, although transgenic expression in mice is known to cause severe dwarfism, and targeted deletion of STC2 causes increased postnatal growth, its precise biological role is still unknown. We found that STC2 potently inhibits the proteolytic activity of the growth-promoting metalloproteinase, pregnancy-associated plasma protein-A (PAPP-A). Proteolytic inhibition requires covalent binding of STC2 to PAPP-A and is mediated by a disulfide bond, which involves Cys-120 of STC2. Binding of STC2 prevents PAPP-A cleavage of insulin-like growth factor-binding protein (IGFBP)-4 and hence release within tissues of bioactive IGF, required for normal growth. Concordantly, we show that STC2 efficiently inhibits PAPP-A-mediated IGF receptor signaling in vitro and that transgenic mice expressing a mutated variant of STC2, STC2(C120A), which is unable to inhibit PAPP-A, grow like wild-type mice. Our work identifies STC2 as a novel proteinase inhibitor and a previously unrecognized extracellular component of the IGF system.

The insulin-like growth factors (IGF-I and -II) are ubiquitous polypeptides, involved in cell proliferation and survival (1,2). IGF signaling is mediated by the type 1 IGF receptor (IGF1R), 3 but six high affinity IGF-binding proteins, IGFBP-1 to 6, are able to sequester the IGFs from the receptor (3,4). Compared with insulin signaling, the presence of IGFBPs thus represents an additional layer of regulation with differential synthesis of the different IGFBPs. However, bioactive IGF can be released by means of proteolytic cleavage of the binding protein, causing the generation of IGFBP fragments with diminished affinity for IGF (5). Specific and limited proteolysis represents the principal mechanism of IGF activation.
One IGF-releasing enzyme is the large metalloproteinase, pregnancy-associated plasma protein-A (PAPP-A, pappalysin-1, EC 3.4.24.79), a member of the metzincin superfamily of metalloproteinases (6), that is able to cleave IGFBP-4 (7), IGFBP-5 (8), and IGFBP-2 (9). PAPP-A functions intimately with IGFBP-4 (8, 10 -15), and highly specific IGF-dependent (8) proteolytic cleavage by surface-tethered PAPP-A releases bioactive IGF from the inactive IGF⅐IGFBP-4 complex in close proximity to the receptor (16). In agreement with this mechanism, PAPP-A null mice (17) and IGF-II null mice (18) both have a 40% reduced body mass compared with wild-type littermates. PAPP-A is not only important in development, it is also thought to be involved in vascular disease in humans (19), and PAPP-A knock-out mice show resistance to the development of atherosclerotic lesions (20), diabetic nephropathy (21), and cancer (22). Similar to several other genes of the IGF axis, the absence of PAPP-A causes a markedly (35%) increased healthy life span in mice (23).
The stanniocalcins (STC1 and STC2) are widely expressed secreted glycoproteins. Numerous published studies have indicated the possible involvement of the two mammalian STCs in diverse biological processes such as cell proliferation, calcium regulation, and cancer development (24 -26). However, their precise biological role remains unknown. Of particular interest, mice overexpressing STC2 are reduced 45% in size (27), whereas STC2 knock-out mice are 15% larger than wild-type littermates (28). Therefore, although the STCs have never been connected with the IGF system, we reasoned that STC2 might influence IGF signaling. In one possible scenario, supported by the severe proportional dwarf phenotype observed upon either overexpression of STC2 or knock-out of PAPP-A, STC2 might negatively affect the strong growth-promoting activity of PAPP-A.
Culture and Analysis of Mouse Embryonic Fibroblasts-Primary cultures of mouse embryonic fibroblasts (MEFs) were derived from E13.5 embryos resulting from mating B6;D2-Tg(STC2)N1 female and male littermates. The embryos were washed, minced, and digested with papain, and cell suspensions were plated in DMEM culture medium supplemented with 55 M ␤-mercaptoethanol (Invitrogen). Cells at passages 2-4 were used for experiments. Residual tissue from each embryo was used for genotyping, and the presence of transgene-derived STC2 secreted into the culture media was confirmed by ELISA. Culture media from nontransgenic or transgenic MEFs were assessed for the presence of proteolytic activity toward IGFBP-4, with or without the addition of inhibitory mAb 1/41 (100 nM) (32).
Transgenic STC2 contained in the media of confluent plates was immunoprecipitated from 20 ml of media using 30 l of protein G-Sepharose 4 Fast Flow (GE Healthcare), to which mAb STC221 was immobilized at 2 mg/ml by using dimethyl pimelimidate dihydrochloride (Sigma) cross-linking, essentially as described (33). Eluted protein was assessed for the presence of endogenous murine PAPP-A by Western blotting using mAb D8-mIgG2a.
Proteinase Assays-Purified IGFBP-4 was labeled with 125 I (Amersham Biosciences), and cleavage reactions were carried out as described previously (10,34). In brief, media harvested from cells transfected with human PAPP-A cDNA with or without pSTC2 or pSTC2(C120A) were diluted (1:500) to 50 pM PAPP-A and mixed with preincubated 125 I-IGFBP-4 (10 nM) and IGF-II (100 nM) (GroPep Bioreagents) in 50 mM Tris-HCl, 100 mM NaCl, 1 mM CaCl 2 , pH 7.5. Following 10 -40 min of incubation at 37°C, the reactions were terminated by the addition of hot SDS-PAGE sample buffer supplemented with 25 mM EDTA. Substrate and cleavage products were separated by 12% nonreducing SDS-PAGE and visualized by autoradiography using a storage phosphor screen (GE Healthcare) and a Typhoon imaging system (GE Healthcare). For some reactions, media from cells transfected with human or murine PAPP-A cDNA were incubated for 0 or 16 h at 37°C with culture media containing a molar excess of STC2 prior to dilution and analysis of activity.
Cell Culture and Transfection-Human embryonic kidney 293T cells (293tsA1609neo) were maintained in high glucose DMEM supplemented with 10% fetal bovine serum, 2 mM glutamine, nonessential amino acids, and gentamicin (Invitrogen). For transient transfection, 6.0 ϫ 10 6 cells were plated onto 10-cm dishes and transfected 18 h later by calcium phosphate coprecipitation using 10 g of plasmid DNA prepared by Gen-Elute HP plasmid miniprep kit (Sigma). Culture supernatants were harvested 48 h post-transfection and cleared by centrifugation, or the cells were further cultured in serum-free medium (CD293, Invitrogen) to facilitate purification. Secreted levels of PAPP-A were determined by a commercial ELISA (AL-101, Ansh Labs).
Protein Purification-Purification of His-tagged recombinant proteins was carried out by affinity chromatography on a 1-ml HisTrap HP column (GE Healthcare). Serum-free media were diluted 1:1 in 20 mM NaH 2 PO 4 , 150 mM NaCl, pH 7.4 (PBS), and loaded onto the column with a flow rate of 1 ml/min. The column was washed with 20 column volumes of 50 mM NaH 2 PO 4 , 1 M NaCl, 20 mM imidazole, 0.05% Tween 20, pH 7.4, followed by 5 column volumes of PBS. The proteins were eluted with 50 mM NaH 2 PO 4 , 300 mM imidazole, pH 7.4, and dialyzed against 20 mM HEPES, 150 mM NaCl, pH 7.4. Prior to iodination, IGFBP-4 was further purified by reversed-phase (RP)-HPLC on a Discovery BIO wide pore C5 column (4 ϫ 250 mm, Sigma), as described (16). Protein purity was assessed by SDS-PAGE, and quantification of purified proteins was done by amino acid analysis.
Enzyme-linked Immunosorbent Assays (ELISA)-For measurement of STC2 concentration, 96-well plates (MaxiSorp, Nunc) were coated with 100 l per well of 7.5 g/ml of catching antibody (mAb STC221) in 100 mM Na 2 HCO 3 , pH 9.4, by overnight incubation at 4°C. The wells were then blocked with 200 l per well of 2% bovine serum albumin (BSA) in TBS (30 mM Tris-HCl, 300 mM NaCl, 2 mM CaCl 2 , pH 7.4) for 30 min at 37°C. Samples were diluted in TBS-T (TBS containing 0.05% Tween 20) supplemented with 1% BSA and incubated in the coated wells for 1 h at 37°C. Following washing in TBS-T, the wells were incubated for 1 h at 37°C with biotinylated detecting antibody (STC216) diluted in TBS-T with 1% BSA to 1 g/ml, washed again, and incubated (1 h) with avidin-conjugated horseradish peroxidase (P0347, DAKO) diluted 1:20,000 in TBS-T with 1% BSA. Following a final round of washing, the wells were developed using OPD tablets (S2045, DAKO). Calibrators were based on recombinant STC2 diluted in TBS-T with 1% BSA. Eight calibration points (6.25-800 ng/ml) were used. Absorbance was measured at 490 nm on an EnSpire Multimode Plate Reader (PerkinElmer Life Sciences). Blank values were subtracted, and data were analyzed by using cubic curve fitting. The functional sensitivity of the assay at 20% coefficient of variation was 20 ng/ml.
To assess monoclonal antibody binding to STC2 and STC2(C120A), plates were coated with 100 l per well of 2 g/ml polyclonal anti-(STC2). Detection was carried out with available STC2 mAbs (1 g/ml) followed by anti-mouse IgG-HRP (DAKO, P0260) diluted 1:2000. Signals from dilution series of STC2 and STC2(C120A) were normalized and com-pared for each of the antibodies. Buffers for dilution and washing were as described above.
Immunoblotting and Analysis of Complex Formation-Proteins separated by reducing or nonreducing SDS-PAGE were blotted onto a PVDF membrane (Millipore), blocked with 2% Tween 20, and equilibrated in 50 mM Tris-HCl, 500 mM NaCl, 0.1% Tween 20, pH 9.0 (TST). Primary antibodies were diluted (to 1-5 g/ml) in TST containing 0.5% fetal bovine serum, and blots were incubated overnight at room temperature. The blots were incubated for 1 h at room temperature with secondary antibodies (polyclonal swine anti-rabbit IgG-HRP (DAKO, P0217), polyclonal rabbit anti-goat IgG-HRP (DAKO, P0160), or polyclonal rabbit anti-mouse IgG-HRP (DAKO, P0260)) diluted 1:2000 in TST containing 0.5% fetal bovine serum. All washing between the steps was carried out with TST. The blots were developed using enhanced chemiluminescence (ECL Prime, GE Healthcare), and images were captured and analyzed using an ImageQuant LAS 4000 instrument (GE Healthcare). All gel images and immunoblots shown are representative of at least four independent experiments.
To probe for covalent complex formation between proteins synthesized in separate cells, culture media were incubated 0 -16 h at 37°C, separated by nonreducing SDS-PAGE (2 l per lane), and then analyzed by Western blotting as described above. Media from cotransfected cells were not incubated prior to analysis.
Circular Dichroism (CD) Analysis-Purified wild-type and mutant proteins were compared by using circular dichroism analysis (37). Prior to analysis, purified proteins were dialyzed against 20 mM NaH 2 PO 4 , 20 mM NaF, pH 7.4. Ten CD spectra were recorded at 25°C for each protein on a Jasco J-810 spectropolarimeter (Jasco Spectroscopic, Japan) using a polypeptide concentration of 0.5 mg/ml and a cuvette of 2 mm path length. CD data were obtained in the range from 260 to 190 nm at a resolution of 0.2 nm using a bandwidth of 1.0 nm. The scan speed was 100 nm/min, and the response time was 1 s. Temperature scans were carried out in the range from 260 to 200 nm. ⌬⑀ (expressed in degrees ϫ cm 2 /dmol) values were calculated on the basis of a mean molar mass of 110 g/mol/residue. IGF1R Stimulation Assay-A cell line stably expressing IGF1R, 293-IGFR (clone H), was used to measure IGF receptor phosphorylation, essentially as described (16). Starved cells were rinsed in PBS containing CaCl 2 (0.1 mg/liter) and MgCl 2 (0.1 mg/liter), pH 7.4, and stimulated for 15 min with combinations of IGF-I (10 nM), IGFBP-4 (50 nM), PAPP-A (2.5 nM), and STC2 (15 nM). Before stimulation, IGF-I and IGFBP-4 were incubated for 20 min at 37°C in 20 mM HEPES, 100 mM NaCl, 1 mM CaCl 2 , pH 7.5, to allow the IGF-I⅐IGFBP-4 complex to form. PAPP-A was then added, and cleavage reactions were carried out in the same buffer for 20 min at 37°C. Prior to use, PAPP-A was incubated with or without STC2 or STC2(C120A) in serum medium for 16 h at 37°C. The stimulated cells were lysed on ice with RIPA buffer (Sigma, R0278) and supplemented with proteinase inhibitor mixture (Sigma, P8340) and phosphatase inhibitor mixture set II (EMD Millipore, 524625) for 10 min. Western blotting of cleared and reduced lysates was used to quantitate ␤-subunit phosphorylation using mAb PY99. Detection of total IGF receptor was done using mAb CT-1. For loading controls, blots were stripped and reprobed with actin mAb AC-74. Quantifica-tion of band intensities was carried out using an ImageQuant LAS 4000 instrument (GE Healthcare). The signals of mAb PY99 were subtracted background signals and plotted.
Statistical Analysis-All statistical analyses were performed using GraphPad Prism version 5.0. Statistical analyses of mouse postnatal growth data were performed using the unpaired Student's t test for each time point. Statistical analysis of IGF-I receptor signaling was performed using one-way analysis of variance followed by Dunnett's test. Statistical analysis of antibody binding was performed using one-way analysis of variance followed by Tukey's test. p Ͻ 0.05 was considered statistically significant.

Culture Medium of Embryonic Fibroblasts Derived from STC2 Transgenic Mice Shows a Marked Reduction in Proteo-
lytic Activity toward IGFBP-4 -To test the hypothesis that PAPP-A and STC2 represent an antagonistic pair whose balance may be important for the regulation of IGF signaling, we first generated mice transgenic for human STC2. We find that medium conditioned by MEFs derived from nontransgenic littermates showed specific cleavage of IGFBP-4, whereas medium conditioned by MEFs from STC2 transgenic mice showed a marked reduction in proteolytic activity toward IGFBP-4 (Fig. 1A). We further find that the proteolytic activity toward IGFBP-4, present in medium conditioned by MEFs from nontransgenic mice, can be completely inhibited by using a novel inhibitory monoclonal antibody specific for PAPP-A (Fig. 1B) (32). The latter experiment demonstrates that the IGFBP-4 proteolytic activity secreted by MEFs can be attributed to PAPP-A, and it is in agreement with the earlier finding that MEFs derived from PAPP-A knock-out mice lack proteolytic activity toward IGFBP-4 (17). Together these experiments indicate that either PAPP-A synthesis, secretion, or activity is compromised upon transgenic expression of STC2.

STC2 Present in Culture Medium of Embryonic Fibroblasts Derived from STC2 Transgenic Mice Is Associated with PAPP-A-To probe for a possible association between STC2
and PAPP-A, coimmunoprecipitation was attempted using an STC2-specific antibody followed by PAPP-A Western blotting. In MEF medium from STC2 transgenic mice, but not from nontransgenic littermates, a single band of high molecular weight, which contained strong PAPP-A immunoreactivity, was observed (Fig. 1C). This experiment allows us to conclude that a physical interaction between the trans-   genic STC2 and the endogenous PAPP-A occurs in the MEF medium, suggesting that STC2 might act to inhibit PAPP-A proteolytic activity.

STC2 Forms a Covalent Complex with PAPP-A and Irreversibly Inhibits Its
Proteolytic Activity-We further analyzed the interaction between STC2 and PAPP-A in vitro in a system based on human recombinant proteins. PAPP-A secreted from transfected mammalian cells rapidly cleaves IGFBP-4 at a single site, resulting in two proteolytic fragments that comigrate in SDS-PAGE (34). However, no proteolytic activity toward IGFBP-4 was detected upon cotransfection with STC2 cDNA ( Fig. 2A), even though PAPP-A was still secreted from the cells (Fig. 2B). Therefore, to directly test the hypothesis that STC2 is an inhibitor of PAPP-A, separately synthesized and purified STC2 was added to the cleavage reaction. Surprisingly, STC2 did not cause an alteration of PAPP-A activity in this experiment (Fig. 2C).
Following separate transfection, PAPP-A and STC2 migrated in SDS-PAGE as dimers with the expected sizes of ϳ400 and 90 kDa, respectively, but upon cotransfection, a high molecular mass band of ϳ500 kDa, which contained both antigens, was formed (Fig. 3, A  and B). Thus, STC2 is capable of forming a complex with PAPP-A, which resists separation in denaturing PAGE and therefore most likely is covalent. However, from this experiment, we cannot distinguish between complex formation before or after protein secretion.
We further analyzed the process of PAPP-A⅐STC2 complex formation by incubating separately synthesized PAPP-A and STC2. A PAPP-A-containing band of 500 kDa gradually appeared during the incubation (Fig. 3C), demonstrating that complex formation in the extracellular environment is possible. PAPP-A was proteolytically active following incubation in the absence of STC2, while after 16 h of incubation with STC2, PAPP-A showed no activity (Fig. 3D). Importantly, this explains the lack of STC2 inhibitory activity observed without prolonged preincubation (Fig. 2C).
Based on the above, we conclude that STC2 is a potent proteinase inhibitor of PAPP-A. Curiously, inhibition requires formation of a covalent complex between PAPP-A and STC2. For- FIGURE 4. Cys-120 of STC2 is required for covalent complex formation and proteolytic inhibition. A, sequence alignment of human STC1 (NM_003155.2) and STC2 (NM_003714.2) using Clustal Omega. Disulfide bonds and dimerization disulfide of STC1, indicated by lines, are according to published data (38). Three cysteine residues of STC2 (Cys-120, Cys-197, and Cys-205), which have no counterpart in STC1, are indicated in blue. Cys-120 is conserved in all known mammalian STC2 sequences, and it has no counterpart in any known STC1 sequence (data not shown). B, STC2 Western blot following reducing SDS-PAGE of medium from HEK293T cells transfected with wild-type or mutated variants of STC2 cDNA as indicated. C, PAPP-A Western blot (WB) following nonreducing SDS-PAGE of media from HEK293T cells transfected with combinations of cDNAs as indicated. Note that STC2(C120A) is unable to form a covalent complex with PAPP-A. D, proteolytic activity toward radiolabeled IGFBP-4 in medium from HEK293T cells transfected with cDNA encoding PAPP-A, preincubated for 16 h with or without STC2 or STC2(C120A).
mation of this complex appears to be a relatively slow process under the given experimental conditions.
Abilities of STC2 to Covalently Bind and Inhibit PAPP-A and to Inhibit IGF Signaling in Vitro Depend on Cys-120 -The homologous STC1 contains 11 cysteine residues. It has previously been shown that 10 of these form five intramolecular disulfide bonds, and that one residue is responsible for dimerization (38). These 11 cysteine residues are conserved in STC2, which contains three additional cysteines, Cys-120, Cys-197, and Cys-205 (Fig. 4A). This uneven number of additional cysteines prompted us to test the possibility that formation of the PAPP-A⅐STC2 complex might be based on disulfide bonding between PAPP-A and STC2.
When the three additional cysteine residues were substituted individually to alanine, the level of protein expression was unchanged (Fig. 4B). However, one mutant, STC2(C120A), did not form a covalent complex with PAPP-A (Fig. 4C), indicating that an intermolecular disulfide bond involving Cys-120 is the basis for the covalent linkage between STC2 and PAPP-A. Furthermore, STC2(C120A) lacked the ability to inhibit PAPP-A under conditions where PAPP-A activity was eliminated by wild-type STC2 (Fig. 4D).
We then used a cell-based assay to assess the ability of STC2 to inhibit IGF receptor signaling in vitro. We find that STC2 virtually eliminated PAPP-A-mediated IGF receptor phosphorylation, whereas the effect of STC2(C120A) was nonsignificant (Fig. 5). Thus, not only the ability of STC2 to covalently bind PAPP-A but also its ability to inhibit PAPP-A-mediated IGF signaling is effectively disabled upon mutation of Cys-120.
STC2 Causes Growth Retardation in Mice by Proteolytic Inhibition-Prior to use for in vivo studies, we compared wildtype STC2 and STC2(C120A) to verify that the single amino acid substitution of residue Cys-120 did not impose a change in properties of the protein, except for the lost ability of STC2(C120A) to inhibit PAPP-A. Comparison by CD analysis showed similar spectra (Fig. 6A) of the purified proteins, and furthermore, the CD spectra of both STC2 and STC2(C120A) appeared to be unaffected by a transient exposure to 90°C (Fig.  6B). These data suggest that the structural integrity of STC2(C120A) is intact. In addition, equal recognition of STC2 and STC2(C120A) by six available monoclonal antibodies also supported this interpretation (Fig. 6C). Finally, we verified that the biochemical properties of both STC2 and STC2(C120A) are conserved in the murine system (Fig. 6, D and E).
To finally test the hypothesis that overexpression of STC2 causes growth retardation by proteolytic inhibition, we generated mice transgenic for wild-type STC2 or STC2(C120A) in parallel. In agreement with earlier findings (27), overexpression of wild-type STC2 caused a severe reduction in postnatal growth rate compared with nontransgenic animals (Fig. 7A). Our data further indicate a correlation between the circulating level of STC2 and the severity of the growth retardation (Fig.  7B). In striking contrast, transgenic expression of STC2-(C120A) did not cause a detectable alteration in growth rate, even at high levels of circulating protein (Fig. 7C). We therefore draw the conclusion that STC2 transgenic mice show growth retardation because the proteolytic activity of PAPP-A is compromised.

DISCUSSION
We have identified STC2 as a potent inhibitor of PAPP-A proteolytic activity. STC2 binds covalently to PAPP-A to completely eliminate its activity toward IGFBP-4 and hence PAPP-A-mediated IGF signaling. From previous studies, it is known that although both transgenic overexpression of STC2 (27) and knock-out of PAPP-A (17) cause severe growth retardation, STC2 knock-out mice (28) are larger than wild-type mice. We here further show that mice transgenic for STC2(C120A), which is unable to inhibit PAPP-A, show normal growth (Fig.  8A).
The substantially increased body weight of STC2 knock-out mice (28) suggests that the absence of STC2 causes an elevated PAPP-A activity. A strong anabolic effect of PAPP-A was also demonstrated by a markedly increased bone thickness upon targeted transgenic PAPP-A expression in osteoblasts (39) and similarly by a large increase in skeletal muscle mass upon muscle-specific transgenic expression of PAPP-A (40). Finally, in humans, the growth-promoting potential of PAPP-A is underscored by the finding of a strong, positive correlation between birth weight and maternal first trimester levels of PAPP-A (41).
Taken together, these findings support a model in which a local dynamic balance between PAPP-A proteolytic activity and STC2 inhibitory activity is a critical determinant of IGF receptor stimulation in vivo (Fig. 8B). In this regard, it is important to stress that IGF-mediated growth ultimately depends on the level of pericellular, bioactive IGF in the tissue microenvironment (2). The source of IGF within tissues may be the circulation or cells of the tissue, first emphasized by the finding that liver-specific knock-out of IGF-I dramatically reduces (by 75%) serum levels of IGF-I but has no effect on growth (42,43). It is also emphasized by the observation that the level of circu-  FEBRUARY 6, 2015 • VOLUME 290 • NUMBER 6 lating IGF-I is not altered in mice that have become dwarfs as a result of STC2 overexpression (27).

Stanniocalcin-2 (STC2) Proteolytically Inhibits PAPP-A
In humans, adult height is influenced by genetic variants distributed in a large number of loci (44). Genes of these loci tend to group in known biological pathways, such as the IGF system, which was found to be represented by genes encoding IGF1R and PAPP-A, for example. Although not connected with any biological pathway, the gene encoding STC2 also appeared in this analysis. Similar findings have recently resulted from genetic studies of domestic dogs, which show the greatest variation in body size of any land mammal (45). It was found that the occurrence of derived variants at only six genes is a major explanation for the extreme reduction in body size of dogs caused by selective breeding (46). The authors found STC2 to be among the genes, also including IGF1 and IGF1R. Therefore, at least three of these six genes driving the small size of dogs are now connected with the IGF system.
Our finding that STC2 is a proteinase inhibitor is surprising. The two homologous STCs have unique amino acid sequences (47,48), containing no known modules or motifs by which this biochemical function might have been predicted. Although the sequence of STC1 lacks a cysteine residue corresponding to Cys-120 of STC2, it is obvious to speculate that STC1 has biochemical properties similar to STC2. In this regard, it is interesting to note that transgenic overexpression of STC1 (49) also causes growth retardation, although STC1 knock-out mice are normal (50). STC1 may not be able to bind PAPP-A covalently, but a high affinity noncovalent interaction with PAPP-A might compensate for this. Future studies are required to address these questions. Likewise, further analysis is required to determine whether the STCs possess inhibitory activity toward other proteinases, including PAPP-A2 (51), the only paralog of PAPP-A. Because of the striking similarity in phenotypes between PAPP-A knock-out mice and transgenic STC2 mice, it FIGURE 6. Biochemical properties of STC2 are retained in STC2(C120A). A, CD analysis of purified STC2 and STC2(C120A). Inset shows Coomassie-stained SDS-polyacrylamide gel of purified STC2 and STC2(C120A). B, CD spectra of STC2 and STC2(C120A) recorded at 25 and 90°C and again at 25°C following incubation at 90°C for 2 min. C, relative binding of available monoclonal antibodies to STC2 and STC2(C120A) is plotted as a ratio between the levels measured in ELISAs. A value of 1 indicates equal binding to STC2 and STC2(C120A). Results are means Ϯ S.D. from four independent experiments performed in triplicate. Differences in ratios are not statistically significant. D, proteolytic activity toward radiolabeled IGFBP-4 in medium from HEK293T cells transfected with cDNA encoding murine PAPP-A (mPAPP-A), preincubated with or without STC2 or STC2(C120A). E, Western blot (WB) following nonreducing SDS-PAGE of samples from D demonstrating covalent complex formation between murine PAPP-A and STC2 but not STC2(C120A).

Stanniocalcin-2 (STC2) Proteolytically Inhibits PAPP-A
is tempting to speculate that the STCs do not affect proteinases other than PAPP-A. However, because knock-out of PAPP-A2 also causes a growth retardation phenotype in mice, albeit less severe than caused by knock-out of PAPP-A (52), PAPP-A2 might also be an STC target proteinase.
Dysregulated IGF receptor signaling has been extensively linked to various diseases, in particular human cancer (53). Also, the involvement of PAPP-A in cancer development is increasingly recognized (32), and is underscored by the finding that PAPP-A knock-out mice have a remarkably low incidence of spontaneous age-related tumors (22). It is therefore interesting that multiple reports link both stanniocalcins to human cancers, although both up-and down-regulations have been reported (24,26). For example, loss of BRCA1 tumor suppres-sor function in breast cancer causes STC1 expression to become undetectable (54), and late relapse of breast cancer correlates with high expression of STC1 and STC2 (55).
Overall, we have demonstrated that STC2 is a novel proteinase inhibitor of PAPP-A, and therefore is most likely a part of the complex extracellular network that regulates IGF receptor activation in vivo. Our findings connect two areas of research that previously seemed unrelated. Future work will delineate the details of the concerted actions of these molecules in normal and pathological tissues.