Human stanniocalcin-1 interacts with nuclear and cytoplasmic proteins and acts as a SUMO E3 ligase

Abstract

Human stanniocalcin-1 (STC1) is a glycoprotein that has been implicated in different physiological process, including angiogenesis, apoptosis and carcinogenesis. Here we identified STC1 as a putative molecular marker for the leukemic bone marrow microenvironment and identified new interacting protein partners for STC1. Seven selected interactions retrieved from yeast two-hybrid screens were confirmed by GST-pull down assays in vitro. The N-terminal region was mapped to be the region that mediates the interaction with cytoplasmic, mitochondrial and nuclear proteins. STC1 interacts with SUMO-1 and several proteins that have been shown to be SUMOylated and localized to SUMOylation related nuclear bodies. Although STC1 interacts with SUMO-1 and has a high theoretical prediction score for a SUMOylation site, endogenous co-immunoprecipitation and in vitroSUMOylation assays with the purified recombinant protein could not detect STC1 SUMOylation. However, when we tested STC1 for SUMO E3 ligase activity, we found in an in vitro assay, that it significantly increases the SUMOylation of two other proteins. Confocal microscopic subcellular localization studies using both transfected cells and specific antibodies for endogenous STC1 revealed a cytoplasmic and nuclear deposition, the latter in the form of some specific dot-like substructure resembling SUMOylation related nuclear bodies. Together, these findings suggest a new role for STC1 in SUMOylation pathways, in nuclear bodies.

---

Introduction

Triggering of angiogenesis is a fundamental process to tumor–host interaction and essential for cancer progression.¹ Solid tumors, as well as leukemias, benefit from this new vascularization by obtaining oxygen, nutrients, growth factors and survival signals. Some angiogenesis inhibitors showed effective anti-leukemia activity in animal models of human leukemia, indicating leukemia dependence on angiogenesis.² Leukemic cells are able to modulate their bone marrow (BM) microenvironment for their own benefit.³ Microarray analysis, carried out in order to characterize leukemia-modulated genes in BM endothelial cells (BMEC), by stimulation either with primary acute lymphoblastic leukemia (ALL) cells or with the patient's plasma, resulted in the identification of several activated genes. Here, in validating five of these potential BM micro-environmental markers by Quantitative Real-Time PCR (RQ-PCR) we found stanniocalcin-1 (STC1) as a significantly activated gene in BMEC after 6 hours stimulus with plasma of children with ALL.

STC1 is an endocrine hormone originally discovered in bony fishes and later identified in humans by mRNA differential display of genes related to cellular immortalization.⁴ Different from its role in fish, human STC1 seems to play a more complex role in numerous developmental, physiological, and pathological processes, including cancer, pregnancy, lactation, angiogenesis, organogenesis, cerebral ischemia, and hypertonic stress.⁵ The precise role of STC1 in carcinogenesis is still unclear, and in particular, the involvement of cancer cells in the differential STC1 expression remains to be determined. STC1 mRNA was present in the bone marrow and blood of breast cancer patients, whereas no STC1 mRNA was evident in healthy volunteers,⁶ thereby suggesting STC1 as a potential novel marker for human breast cancer. Tumor vasculature may also be responsible for the increased expression of STC1,^7,8 since VEGF induces STC1 expression.^9,10 Further studies suggest that STC1 activates antioxidant pathways in endothelial cells and macrophages and displays cytoprotective and anti-inflammatory actions.¹¹

Albeit STC1 has been implicated in so many processes, neither its receptor nor any other interacting molecular partners have been reported to date. In fact its cellular localization is still dubious, since several reports related it to be localized to the cytoplasm, the mitochondria¹² and even the nucleus,¹³ although it is a secreted protein supposed to enter the secretory pathway of the cell.¹⁴ However secreted protein may be absorbed by other cells, possibly resulting in the uptake in cytoplasmic compartments. In view of the lack of information about molecular partners and specific STC1 function, we performed here yeast two hybrid screens in order to trace potential mechanisms underlying its activities in human cells.

Results

Validation of STC1 as a BM microenvironment molecular marker in ALL

Previous microarray data (Jose A. Yunes and Angelo A. Cardoso, unpublished data) showed around an eight-fold increase of STC1 mRNA in BMEC after treatment with ALL patients' plasma. We were able to confirm STC1 mRNA up-regulation during a co-culture of bone marrow stromal cells with leukemic blasts by Real Time Quantitative PCR (RQ-PCR, Table 1, see Materials and Methods for details) and found a similar increase (∼7 fold) in stromal cells co-cultured with 3 different primary ALL cells after 24 hours of treatment. After only six hours of treatment STC1 mRNA level was however ∼5.4 fold below its initial level (0 h) (Fig. 1).

Fig. 1 Analysis of candidate gene expression by quantitative real time PCR (RQ-PCR). (A) Bone Marrow (BM) stromal cells were cultured until they reached confluency and were then subjected to a co-culture with primary B-precursor acute lymphoblastic leukemic cells obtained from different patients for the indicated periods of time. BM stromal cells were separated and processed for RNA extraction, cDNA synthesis and Real-time Quantitative PCR (RQ-PCR). (B) mRNA expression of different candidate genes in BM stromal cells either cultured in fresh media without fetal calf serum (FCS) during 6 hours and 24 hours (both with or without FCS), or stimulated by co-culture with primary leukemic blasts of three different ALL patient's (samples 1, 2 and 3) in fresh media without bovine fetal serum (BFS). The same samples of BM stromal cells before the co-cultivation (0 h) were used for calibration (set to 1.0). Bars indicate standard errors with 95% of confidence. DKK1 and STC1 presented initially a reduction in the expression levels, after 6 hours of co-culture. STC1 demonstrated then, after 24 hours, a great increase in its expression level, validating the microarray data and suggesting that STC1 may serve as tumor microenvironment marker.

Identification of STC1 interacting proteins by yeast two hybrid screenings

Before the yeast two-hybrid screening the construction of LexA–STC1ΔSPa (without the signal peptide) was tested for autonomous activation of reporter genes (HIS3 and LacZ) and no auto-activation of reporters was found (not shown). This demonstrates that STC1 alone has no capacity to activate the reporters and to generate false positive clones. The identified interacting clones demonstrate a specific interaction with the STC1 protein. The absence of auto-activation of the bait construct is an absolute prerequisite for the successful performance of a yeast two-hybrid system library screen. We then employed the yeast two-hybrid system to screen human fetal brain, bone marrow and leukocyte cDNA libraries. LexA–STC1ΔSPa construct was used as bait, and a total of about 1.5 × 10⁶ co-transformed clones in the three screenings were assayed for both reporters.

An interaction map designed for STC1 interactions exhibits a complex network (Fig. 2). A total of 22 proteins were identified and their retrieved fragments, domain composition as well as their cellular locations according to GeneOnthology¹⁵ are presented in Table 2.

Fig. 2 Protein interaction map of human STC1. Yeast two hybrid screens of bone marrow, leukocyte and fetal brain cDNA libraries were performed. Diagram shows a simplified network of protein interactions found by yeast two hybrid screens (ZBTB16 has been omitted here since it may be a false positive). Lines indicate interactions and each color is indicative of published data about the interaction (the present work new findings are indicated by yellow links of STC1 to its preys). Each protein is represented by a different color code assigned by the Osprey program⁸² according to GeneOnthology.

Table 1 Primer sequences and concentrations used in the RQ-PCR

Primer name	Primer sequence (5′-3′)	Product size (bp)	Concentration in RQ-PCR reaction/nM
AF1Q-F	GGCCTGGGTCTGTCAGATACA	72	400
AF1Q-R	TGCTTGCCCGATCATTTTG
CGI109-F	AGGGAGGAAGAAAATTGCCTTT	95	400
CGI109-R	TCCATAAAACTGAACAGTGCAGAATA
DKK1-F	AGGAAGCGCCGAAAACG	88	400
DKK1-R	TTTGATCAGAAGACACACATATTCCA
STC1-F	TGAGGCGGAGCAGAATGACT	78	100
STC1-R	CAACGAACCACTTCAGCTGAGT
TCTE1L-F	ATGCTGAGGAAGCCCACAAT	148	250
TCTE1L-R	TTATAGGCTTTTCCCAACTTAACCA
GAPDH-F	ATGGAAATCCCATCACCATCTT	68	400
GAPDH-R	CAGCATCGCCCCACTTG

---

Table 2 Proteins retrieved from yeast two hybrid screenings

cDNA library (no. of clones)	Official symbol (Entrez GeneID)^87	Protein fragment	Domain(s) present	Component/compartment according to GeneOntology^15
BM: bone marrow; FB: fetal brain; L: leukocyte; AAT I PLP: aspartate aminotransferase superfamily (fold type I) of pyridoxal phosphate; BTP: bromodomain transcription factors and PHD domain containing proteins; CC: coiled coil; DUF2359: uncharacterized conserved protein; FBP AIa: fructose-1,6-bisphosphate aldolase Ia; FN3: fibronectin type III; G: globin; HMG: high mobility group; IF tail: intermediate filament tail; NADHdh: NADH dehydrogenase; PQQ: beta-propeller repeat occurring in enzymes with pyrrolo–quinoline quinone; Tryp SPc: trypsin-like serine protease; UBQ: ubiquitin homologues; ZnF RBZ: zinc finger domain in ran-binding proteins.
BM (9)	FUS (2521)	381–526 (526)	ZnF RBZ	- Nucleus - Membrane - Cellular component unknown
BM (8)	HBA1/2 (3039/3040)	1–142 (142)	G	- Hemoglobin complex
BM (5)	JSRP1 (126306)	1–314 (331)		- Undetermined
BM (4)	FLJ20254 (54867)	601–689 (689)	DUF2359	- Undetermined
L (4)	SAP18 (10284)	81–153 (153)		- Histone deacetylase complex
FB (1) & L (1)	MT-ND1 (4535)	9–60 (318)	NADHdh	- Mitochondrial respiratory chain complex I - Integral mitochondrial inner membrane
FB (2)	FNDC4 (64838)	15–234 (234)	FN3	- Integral to plasma membrane
L (2)	ERN1 (2081)	1–59 (977)	PQQ	- Integral to endoplasmic reticulum membrane
L (2)	CMTM3 (123920)	34–182 (182)		- Extracellular space - Integral to plasma membrane
L (2)	MAPK14 (1432)	1–38 (360)		- Cytoplasm - Nucleus - Spindle pole
L (2)	ANPEP (290)	875–967 (967)		- Cytoplasm - Integral to plasma membrane - ER-Golgi intermediate compartment
BM (2)	ELA2 (1991)	190–267 (267)	Tryp SPc	- Cell surface - Extracellular region
L (2)	SP100 (6672)	704–852 (879)	HMG, CC	- Nucleus - PML body - Chromatin/chromosome
BM (1)	LMNA (4000)	439–572 (572)	IF tail	- Nucleus - Nuclear lamina intermediate filament
BM (1)	ALDOA (226)	89–364 (364)	FBP AIa	- Undetermined
BM (1)	FTL (2512)	1–175 (175)		- Ferritin complex
BM (1)	ZBTB16 (7704)	55–176 (673)	BTP	- Nucleus - Nuclear speckles - PML body - Transcriptional repressor complex
BM (1)	ALAS2 (212)	341–460 (587)	AAT I PLP	- Endoplasmic reticulum - Mitochondrial matrix - Integral to mitochondrial inner membrane
L (1)	QRICH1 (54870)	502–667 (776)		- Intracellular
L (1)	SUMO1 (7341)	1–101 (101)	UBQ	- Cytoplasm - Membrane - Nucleus - Nuclear pore - Nuclear membrane - Nuclear speckles
FB (1)	TMEM132A (54972)	808–932 (1024)	CC	- Endoplasmic reticulum - Golgi apparatus - Integral to endoplasmic reticulum membrane
L (1)	DAGLB (221955)	613–668 (672)		- Integral to plasma membrane

---

Proteins were grouped according to their cellular compartment in seven groups, although some are present in more than one group: nuclear (LMNA, MAPK14, FUS, QRICH1, SAP18, SP100 and SUMO1), endoplasmic reticulum (ERN1, JSRP1, TMEM132A and FLJ20254), mitochondria (MTND1 and ALAS2), cytoplasm (ALDOA, FTL, MAPK14 and SUMO1), plasma membrane (ANPEP, CMTM3, DAGLB, FNDC4 and TMEM132A), extracellular/secreted (ELA2) and red blood cells (HBA1/2). Although QRICH1 and FLJ20254 had no GeneOnthology information about their location we used the prediction of PSORT II program¹⁶ to infer their possible localization, and hence be able to group them, too.

Confirmation of interactions by GST-pull down assays

In order to confirm interactions observed in the yeast two-hybrid screens we performed GST-pull down assays, using prey proteins, retrieved from the screenings, fused to a GST-tag and a His-tagged STC1 (STC1-HT) produced in insect cells.¹⁷ Starting with 22 clones, 18 cDNAs had cloning strategy frame, for direct transfer from pACT2 vector to GST-tag expression vector. From this 18, we could sub-clone 11 and, from those, only seven (presented in Fig. 3—SP100, FUS, JSRP1, SUMO1, TMEM132A, FDNC4-FIII and LMNA) expressed as soluble proteins. In the case of FNDC4 protein, primers were designed to amplify only the extracellular portion (mainly comprising its fibronectin type III domain [FNIII]). Even using PMSF to inhibit proteases activity, some of the GST-tag proteins expressed showed some degradation in the expression conditions that we were able to establish for the soluble expression of these proteins. Recombinant STC1-HT have bound to all seven GST-tagged proteins but not to GST alone demonstrating the specificity of observed interactions (Fig. 3).

Fig. 3 In vitro confirmation of STC1 interaction with selected retrieved proteins. GST-pull down assay between STC1-HT produced in the baculovirus system and fragments of selected prey proteins retrieved from the yeast two-hybrid screen. All preys were GST fusion proteins. Asterisks indicate theoretically expected GST fusion recombinant protein sizes. Arrow head indicates GST. INPUT indicates GST input in the top Western-blot (WB: anti-GST) and STC1-HT in the bottom Western-blot (WB: anti-6xHis).

STC1 interacts with partners by its amino-terminus

We further generated two constructs of STC1 fused to LexA DNA binding domain splitting STC1 into approximately two halves (an amino and a carboxyl end construct, named LexA–N STC1ΔPSb and LexA–C STC1 respectively). The cut between these two arbitrary halves was performed in a region where no disulfide bonds showed cross-connection of distant parts of the primary sequence (Fig. 4A), and both constructs were tested against the prey constructs retrieved from yeast two hybrid screenings. The great majority of the proteins interacted with the amino terminus of STC1 but not with its carboxy terminal end. Interestingly, FNDC4 interacted with both the N- and C-terminal constructs of STC1 (Fig. 4B).

Fig. 4 Mapping of the STC1 interaction site with the prey proteins using the yeast two-hybrid system. (A) Schematic representation of full-length human STC1 and different constructs with LexA DNA binding domain used here. A linear representation of STC1-HT amino acid sequence is shown with indication of the different portions (signal peptide [SP] in white, pro-peptide [PP] in dark gray, and mature protein chain in black), its disulfide bonds intra- (brackets) and inter-chain (Dimer box). All constructs were produced as chimeric proteins with LexA fused to the N-terminus. LexA–STC1ΔSPa was used initially as a bait in the yeast two-hybrid screens and all three constructs were used to map the interaction site of STC1. Light gray bar indicates a linker region derived from cloning. (B) Colony lift β-galactosidase assay of the yeast L40 strain co-transformed with pBTM116 constructs expressing LexA fusion protein baits (indicated on the right side) and Gal4 activation domain fusion preys (indicated on the top) as retrieved from the screens.

During the mapping of the interaction between ZBTB16 and STC1 N- and C-terminal constructs this prey construct alone also presented non-specific β-galactosidase activity with a LexA-only construct (data not shown). The confirmation assay of all prey plasmids alone is therefore essential to rule out that the prey has transcriptional activation characteristics per se which do not depend on its productive interaction with the bait protein (STC1). In the case of the transcription regulatory ZBTB16 the potential auto-activation of the reporter genes was almost expected.^18,19 Accordingly, ZBTB16 was regarded as a false positive and discarded further on.

The role of stanniocalcin-1 and SUMO-1 interaction

Since SUMO-1 was retrieved as one of the preys from our yeast two hybrid screen and because we were able to confirm this interaction by GST-pull down (Fig. 3), we decided to analyze if the STC1 has a predicted posttranslational SUMOylation site using the SUMOplot^TM web available tool from Abgent (http://www.abgent.com/doc/SUMOplot). This analysis showed that lysine 83 in motif AK₈₃FD of STC1 is likely a target site of SUMOylation, with a relatively high score of 79%. By comparing this motif with the motifs of other known SUMOylated proteins and their scores, we speculated that STC1 not only interacts with SUMO-1 but may be also a target of SUMOylation. Furthermore, it has been previously shown that around 23% of SUMOylated proteins have a non-canonical SUMOylation motif (i.e. motifs that do not follow the rule: ΨKxD/E—Fig. 5A).²⁰ As SUMOylated proteins acquire novel functions, localization, activity or interactions with other proteins,²¹ we studied the conservation of the predicted SUMOylation motif AK₈₃FD in different STC1 orthologs from different species. We discovered by analyzing the phylogenetic tree, that this SUMOylation site does not appear until the appearance of tetrapodes, but is 100% conserved in this group (Fig. S1, ESI†). This finding matches well with the fact that in bony fish the function of STC1 has been strictly and solely associated to the calcium/phosphate homeostasis.²² However in tetrapodes, especially in mammals,^23,24,25 there seems to have been a gain or change of functions for STC1, since several new functions were related in mammalian species that include lactation, hypoxia, bone size regulation and no significant effects on mineral homeostasis have been related in mammalians.¹¹

Fig. 5 Stanniocalcin-1 is not SUMOylated but acts as an E3 SUMO-1 ligase. (A) Bioinformatics analysis of human STC1 protein sequence with the SUMOplot^TM web tool. The consensus motif ΨKxD/E (where Ψ represents a bulky aliphatic residue and x represents any amino acid)⁸⁵ is used by the program to analyze the sequences. The program retrieved the shown scores for the different shown input sequences. Sequences with a (*) are experimental identified SUMOylation sites: PML and Sin3A,⁸⁶Cdc3 and Fak1.²⁰ The score for STC1 is classified as a high probability motif. (B) In vitroSUMOylation assay. The recombinant STC1-HT from baculovirus was submitted to the SUMOylation reaction provided by the BIOMOL kit. As a positive control the RanGAP-1 protein was used. The negative controls are the same reactions, without ATP. The in vitroSUMOylation reaction could not detect SUMOylation of STC1. (C) HEK293 cells were used for the co-immunoprecipitation assay. The co-immunoprecipitation could also not detect endogenous STC1 SUMOylation, but on the other hand confirmed the interaction between both proteins. The total lysate was loaded onto the gels and analyzed by Western blots, but we could not detect neither STC1 nor SUMO-1 (not shown). This was probably because there was not enough specific protein in the lysate lanes, compared to the IP lanes, where all antibody bound protein was concentrated at only one place. (D) SUMO E3 ligase activity of STC1. The in vitroSUMOylation reactions were done with two different proteins, with or without STC1. The reactions in which STC1 was present had a significant increase of the target protein SUMOylation (WB anti SUMO-1), showing that STC1 displayed a SUMO E3 ligase activity. GST protein was used as a negative control, and the WBs anti-GST and the WB anti-Ki 1/57 showed that the same quantities of proteins were loaded.

So, we attempted to identify if STC1 is a target of SUMOylation either by performing co-immunoprecipitations of endogenous STC1 from HEK293 cells or by in vitroSUMOylation assay using recombinant STC1-HT as a target. As shown in Fig. 5B and C, we could not detect any SUMO-1 attachment to the STC1 protein. However, since STC1 and SUMO-1 interacted, we also tested whether STC1 could increase the SUMOylation level of two other known SUMOylated proteins. Most interestingly we observed that STC-1 acted as a SUMO E3 ligase capable of promoting a significant enhancement of SUMOylation, in a ATP dependent fashion, of both test proteins Ki-1/57-GST and RanGAP-1-GST, but not of the control protein GST (Fig. 5D).

Cytoplasm and nuclear sub-cellular localizations for STC1

The proteins identified in the yeast two-hybrid screenings, as described here, are localized both in the cytoplasm and nucleus, suggesting that STC1 may also localize to the same compartments. Some nuclear proteins, especially SP100, SUMO-1 and SAP18 further suggest a localization of STC1 to nuclear bodies, possibly PML bodies, in which all of these interacting proteins have been found [SP100;²⁶SUMO-1;²⁷SAP18²⁸]. We checked this by subcellular immunolocalization experiments. Confocal microscopic analysis of both K562 leukemia cell and a monoclonal antibody against STC1 to detect the endogenous protein or HEK293 cells transfected with pEGFPN₂–STC1 plasmid, revealed that STC1 localizes to the cytoplasm and the nucleus, where it appears, as expected, in dot shaped fashion, thereby reinforcing the idea of a localization to specific nuclear bodies (Fig. 6). Moreover, both endogenous and transfected STC1 showed a characteristic perinuclear deposition (Fig. 6).

Fig. 6 Stanniocalcin-1 localizes to the cytoplasm and to dot like structures in the nucleus. Fluorescence microscope analyses suggested that STC1 was localized preferentially in the cytoplasm and also to the nucleus in a dot like fashion. Confocal analyses confirmed the nuclear corpuscles, suggesting, according to the proteins retrieved from yeast two-hybrid screen, that this dots may be nuclear corpuscles. For both cell types we observe the same patterns. (A) HEK293 cells transfected with STC1-GFP. (B) K562 leukemic cell type, marked for endogenous STC1. The arrows (red in A and green in B) points out the nuclear corpuscles that co-localizes in the DAPI stained area.

Discussion

Looking for genes that may be activated at the cross talk between the bone marrow microenvironment and leukemia cells we identified by microarray data and validated by real time PCR a molecular marker, STC1, which has been previously suggested to be involved in angiogenesis⁵ and was described as a minimal residual disease marker in ALL.²⁹

Although STC1 has been implicated in diverse cellular processes, its molecular partners have not been characterized yet. Here we tried for the first time to identify protein interaction partners for STC1 in order to better understand its cellular role. Among all retrieved STC1 interactors, there were proteins from different cellular compartments such as the nucleus, the cytoplasm, mitochondria, endoplasmic reticulum (ER)/Golgi complex, plasma membrane and from the extracellular space. Although STC1 has been described to be a secreted hormone, that is present in all secretory compartments from ER onward, through the Golgi complex to the extracellular space, it has also been shown to have high affinity ‘receptors’ at the plasma and mitochondria membranes¹² and was found in the nuclei of mammary gland alveolar cells.¹³ By confocal microscopy analysis we were able to confirm the cytoplasmic and nuclear localizations of STC. It remains to be determined if, and how the secreted protein is able to be internalized in an auto- or paracrine fashion by cells. STC1 may however in addition also bind to specific, although yet unknown, receptors in the plasma membrane, possibly triggering cell signaling cascades. In the following sections we will address in more detail the possible implications of STC1 interaction with proteins from the different mentioned compartments.

Nuclear proteins

The interphasic nucleus is a highly compartmentalized organelle, in which chromosomes occupy discrete territories and various regulatory proteins are present in specific nuclear bodies (NBs) and/or are diffusely distributed throughout the nucleoplasm.³⁰ The lamins, which belong to the intermediate filament family of proteins, are the major components of a filamentous network underlying the inner nuclear membrane, termed the nuclear lamina. Among roles attributed to lamins are maintenance of nuclear envelope integrity, chromatin anchoring sites supply and determination of interphase nuclear architecture. STC1 interactor Lamin A/C (encoded by LMNA gene) seems to have a role in spatial organization of the speckles, one of the ever-increasing number of NBs, where it seems to organize RNA splicing factors and polymerase II transcription.^31–33

STC1 interactor FUS, also known as TLS, is a member of the FET family of RNA binding proteins, which also includes EWS and TAF15 (TAFII68, TAF2N, RBP56) proteins, that had been found in the nucleus as well as in the cytoplasm and was shown to shuttle between these locations. FET proteins are targeted to stress granules induced by heat shock and oxidative stress and FUS required its RNA binding domain for this translocation.³⁴ Also, the small nuclear ribonucleoprotein (snRNP)—associated protein of 69 kDa (69 KD), a protein that shares structural similarity with members of FET (TLS and EWS, 95% and 65% identity, respectively), has been shown to be present in speckles.³⁵ Interestingly FUS was one of the most abundant proteins retrieved in our STC1 yeast two hybrid screenings. Most interestingly, STCs have been previously implicated in hypoxia³⁶ and oxidative stress,³⁷ so it is likely that it could be found in stress granules together with FUS and LMNA. FUS/TLS protein has been described as acting not only in nucleus–cytoplasm shuttling but also as a SUMO1 E3 ligase for Ebp1, a putative tumor suppressor protein.³⁸ Once interacting with STC1, FUS can be involved in STC1 shuttling and both may act in a concerted fashion to enhance protein SUMOylation as E3 ligases. The fact that STC1 could increase Ki-1/57 SUMOylation maybe linked to the identified complex of OTUB1 protein, which participates in a protein complex composed of FUS/TLS, CHOP and RACK1.³⁹RACK1 and Ki 1/57 have been previously described to interact with each other.⁴⁰ There are no SUMO E3 ligase described that co-localize with nuclear lamins, but two different Ubiquitin E3 ligases, Ret Finger Protein⁴¹ and RNF13⁴² have been described to do so. Since SUMOylation and ubiquitination process are very similar, we suggest that probably the SUMO E3 ligase activity of STC1 occurs at the lamins. All the analyzed confocal slices (endogenous or transfected STC1) show a characteristic perinuclear deposition of STC1.

Other nuclear sub domains are the promyelocytic leukemia nuclear bodies (PML-NBs), which recruit and locally accumulate a large number of proteins, many of which are key regulators of various processes, including splicing and transcription. PML is present both in the nucleoplasm and in NBs, which also are nuclear matrix associated,⁴³ and plays a role in the organization of these PML-NBs, targeting proteins such as STC1 interacting protein SP100, CBP, or Daxx to these domains.⁴⁴SUMOylation was first proposed to target PML toward NBs, behaving as an NB-targeting signal but recently SUMOylation is not anymore seen as an NB-targeting signal, but rather as a consequence of the proteins localization to the NB, although the functional significance of the presence of so many SUMOylated proteins in the NBs remains to be obscure.^43,44 Curiously, STC-1 interactor ZBTB16, although discarded during our mapping experiments as a putative false positive, has also been described as a component of PML-NBs.⁴⁵

Interestingly, most of PML-NB proteins are transient residents and are recruited or released upon different cellular stress signals. An example is p53 protein that during different kinds of stresses is stabilized and activated at the PML-NB and then can induce the transcription of its targets genes leading to cell cycle arrest, senescence or apoptosis.^46,47 Another link connecting STC1 to PML-NB is the observation that during cell death generated by iodoacetamide-induced oxidative stress, p53 is stabilized and acetylated, ultimately resulting in upregulated STC1 gene expression.⁴⁸

Also, exposure of cells to high concentrations of heavy metal affects SP100 and PML proteins of PML-NB altering their number and distribution within the cell.²⁶ Turnover dynamics of PML and SP100 at NBs is modulated by SUMOylation⁴⁹ and some studies hypothesize that PML-NBs are SUMOylation hotspots.⁵⁰

Another STC1 interacting protein retrieved in yeast two-hybrid screen, SAP18 (Sin3-associated polypeptide of 18 kDa), has been shown to play a key role in the gene-specific recruitment of the HDAC complex by a number of transcription factors, acting as a protein–protein adapter module bridging the Sin3-HDAC complex to transcription factors, which is supported by its ubiquitin-like fold.⁵¹ The transcription of STC1 has already been described to occur under HDAC3 regulation in breast cancer cells. When the progesterone receptors (PR), binding at the promoter region of STC1, are SUMOylated, HDAC3 is able to bind to PR and inhibit the transcription. Alternatively, when the PR are phosphorylated, SUMO-1 cannot attach to PR allowing for HDAC3 to promote STC1 transcription.⁵² Recently HDAC7 has been shown to act as a SUMO E3 ligase promoting PML SUMOylation, in a deacetylase independent fashion, and playing an important role in regulation of PML-NB formation.^53,54

The fact that STC1 interacted with its partners predominantly by its amino-terminal region (Fig. 4B) may have interesting structural implications. In previous SAXS experiments¹⁷ we had postulated that STC1 forms an anti-parallel dimer mediated through its carboxy-terminal end, leaving both of its amino-termini exposed, which can possibly give the ability to STC1 to interact with two partners at the same time. This may also be crucial for STC1 capacity to act as a SUMO E3 ligase, which would have to bind at one point in time simultaneously to both SUMO1 and the target protein of SUMOylation.

It is now widely accepted that SUMOylation is a process that is required for PML-NB formation. Therefore the SUMO E3 ligase activity of STC1 in these bodies could be relevant for its interactions with several important proteins associated with PML-NB that may be targets of SUMOylation.

Endoplasmic reticulum proteins

Among the four ER retrieved proteins interacting with STC1 two are related to the unfolded protein response (UPR), like another member of the STC family, itself: STC2. The first is the UPR transducer ERN1 and the second is TMEM132a, which has also been renamed to HSPA5 binding protein (HSPA5BP) by its high homology to its rat ortholog.⁵⁵ The UPR is activated by a variety of insults that disrupt protein folding in the ER lumen thus preventing accumulation of unfolded proteins in the ER lumen. Among these insults are changes in intralumenal calcium, altered glycosylation, nutrient deprivation, pathogen infection, expression of folding-defective proteins, and changes in redox status.⁵⁶ The mammalian UPR has been shown to be activated by three interconnected signaling proteins: activating transcription factor 6 (ATF6), IRE1 (first identified in a yeast mutant with inositol-requiring phenotype, and also known as ERN1), and double-stranded RNA-activated kinase (PRK)-like ER kinase PERK. Each of these proteins are localized to the ER membrane bound to glucose-regulated protein 78 (Grp78, also known as BIP and HSPA5), a soluble ER-resident molecular chaperone that upon ER stress conditions is released from the trio of UPR transducers.⁵⁷

Most interestingly the other member of STC family of proteins, STC2, has been shown to be induced by ER and oxidative stress agents, and its knockdown was shown to promote thapsigargin-induced N2a cell death. Thapsigargin is an inhibitor of SERCA (sarco/endoplasmic reticulumCa²⁺ ATPase) viaUPR.³⁵IRE1 has been related to participate in the relationship between ER stress and angiogenesis in the placenta during pregnancy in mammals.⁵⁸ Furthermore the nuclear targeting of stanniocalcin-1 during pregnancy and lactation, together with its role in angiogenesis has been previously reported, emphasizing the functional connections between STCs and IRE1.¹³

The other two ER proteins retrieved with STC1 as bait are the junctional sarcoplasmic reticulum protein 1 (JSRP1), and the FLJ20254 protein (recently named TMEM214) of unknown function. JSRP1 plays a modulatory role in the calcium turnover in muscle, as shown by its interaction to both dihydropyridine receptor (DHPR) voltage sensors and inside the sarcoplasmic reticulum's lumen, where it binds to calsequestrin which in turn is linked to Ryanodine receptors family of calcium release channels of the reticulum.^59,60 In bony fish STC1 has been well characterized to play roles in calcium and phosphate homeostasis.^61–63 Increasing evidences have been showing that the human STC1 may also have a role in stress related to calcium-overload.^64–66 Therefore, it is a mechanistic possibility that STC1 may interact with and therefore inhibit JSRP1.

Mitochondrial proteins

STC1 has been previously found to localize within the mitochondrial matrix, exerting a concentration-dependent stimulatory effect on NADH oxidation.¹² This was shown to be attenuated upon occupancy of the STC1 ATP binding cassette by purine nucleotides such as ATP and GTP.⁶⁷STC1 effect on the electron transport chain may be mediated through its binding to its interacting protein MT-ND1 (Table 2), possibly by enhancing its activity. In fact, high affinity receptors of STC1 have already been observed.¹²

Cytoplasmatic proteins

STC-1 interactor mitogen-activated protein kinase 14 (MAPK14 or p38α^MAPK) is a key kinase that responds to activation by environmental stress and pro-inflammatory cytokines by phosphorylating a number of transcription factors and several downstream kinases. Another branch of the mitogen activated protein kinases are the extracellular signal regulated kinases (ERKs) 1 and 2 (ERK1/2), which are part of the classical mitogen kinase cascade, regulating proliferation, differentiation, and cell cycle progression.⁶⁸ Another study proposed that STC1 expression induced by VEGF-A₁₆₅ would be mediated primarily by PKC, ERK and calcium signaling pathways.⁶⁹ Since MAPK14 interacted with STC1 we speculate that STC1 may act in a modulatory fashion inhibiting the stress-related signaling pathway of MAPK by binding to it.

Red blood cell and secreted proteins

The second more abundant protein retrieved from the STC1 yeast two hybrid screenings were the two hemoglobin beta chain genes (HBA1 and HBA2), which code for the same protein. Circulating STC1 is usually not detected, except in pregnancy,⁵ however James and coworkers⁷⁰ demonstrated a high binding of STC1 to red blood cells postulating that the inability of STC1 detection in serum may be due to its sequestration by red blood cells, which either could represent a mechanism of delivery, storage or removal of STC1 from blood circulation and in any case HBA1/2 may serve as the binding site.

Membrane proteins

STC1 interactor Aminopeptidase N (ANPEP, CD13) is a broad specificity aminopeptidase that plays a role in the final digestion of peptides generated from hydrolysis of proteins by gastric and pancreatic proteases and serves as receptor for human conavirus and cytomegalovirus infections, as well as for tumor-homing peptides, more specifically NGR peptides. In addition ANPEP seems to be an important regulator of endothelial morphogenesis during angiogenesis.^71,72

Chemokine-like factor superfamily 3 (CMTM3: CKLF-like MARVEL transmembrane containing motif protein 3, CKLFSF3) belongs to a novel family of proteins that play important roles in the immune system and participate in tumorigenesis. Retrieved as an STC1 interactor, CMTM3 in particular is highly expressed in the testes along with leukocytes and placenta.^73,74

STC1 interactor fibronectin domain containing protein 4 (FNDC4, FRCP1) is a protein that is strongly expressed in brain and liver, and like fibronectin type III (FNIII) and Arg-Gly-Asp (RGD) containing plasma membrane proteins might serve either as cell adhesion molecules or as receptors with no identified ligand, yet.⁷⁵

Taking into account that STC1 does not have a biochemically characterized receptor on the cell surface it would be interesting to perform further experiments in order to test whether any of these three transmembrane proteins may function as a functional STC1 receptorin vivo.

Conclusions

Once we have validated STC1 as a potential molecular marker of the bone marrow tumor microenvironment we set out to identify its molecular interaction partners in human cells. Our yeast two hybrid screen identified a broad spectrum of candidate protein interactors and emphasized the notion that STC1 is a truly multifunctional protein (Fig. 2, Table 2), capable of interacting with proteins from different cellular compartments, including the nucleus. By focusing on a specific interaction of STC1 with SUMO1 we were able to establish a new specific function for STC1, which we show here, can act as a SUMO E3 ligase. However, many of the other interactions briefly descried in our discussion, open promising new avenues to be explored in future detailed studies, since they all show interesting connections with previous functional studies on STC1.

Author's contribution

DMT conceived and designed, under JAY and JK supervision, the real time PCR and the yeast two hybrid screens. MTS conceived and designed, under JK supervision and with help of DMT, the confirmation of interactions by GST-pull down, the mapping interactions in yeast, all SUMOylations analyses and assays and the sub-cellular localization assays. KAG helped with in vitroSUMOylation of Ki-1/57-GST and GCB expressed and purified Ki-1/57-GST protein. FA established expression protocol for FNDC4 protein. MTS, DMT and JK discussed all the results and wrote the paper. JK supervised the project. All authors read and approved the final version of the manuscript.

Materials and methods

Co-culture assay for the validation of the molecular marker candidate genes differential expression in BM stromal cells

Primary leukemia samples used for in vitro assays were obtained from patients with a newly diagnosed B-lineage acute lymphoblastic leukemia (ALL). Co-culture experiments were done with cryopreserved mononuclear cells from leukemic BM aspirates. Normal BM samples were obtained from healthy transplant donors included as controls. The study was approved by the Research Ethics Committee from Centro Infantil Boldrini. Bone marrow stromal cells (BMSCs) were obtained by culturing donor BM mononucleated cells in Dulbecco's Modified Eagle medium (DMEM) high glucose (Sigma) supplemented with 10% fetal calf serum (FCS), 1 μM hydrocortisone and penicillin–streptomycin, at 37 °C and 5% CO₂. After one week, plates were washed to remove non-adherent cells, and the culture was continued until confluence was reached. Cells were detached by the addition of a solution containing 0.25% trypsin. Cells at passages 1 to 3 were used in the experiments.

In order to validate previous microarray data we performed co-culture of BM stromal cells with leukemic blasts. BM stromal cells were cultured in DMEM media with 10% fetal calf serum (FCS) in six-well plates until they reached around 70% of confluency. At this point the medium was changed either to fresh media (control) or fresh media containing 3 × 10⁶ primary leukemic cells (three patient samples were used in three parallel experiments). Samples were collected from non-treated BM stromal cells (0 h) and from those treated with fresh medium alone, containing or not, the leukemic blasts. Cells were then washed with HBSS 1× in order to wash of most leukemic blasts from the stromal layer that tightly adhere to the plastic surface of the tissue culture flasks. Ultimately the stromal cells were trypsinized, collected and their total RNA was extracted.

Real-time quantitative PCR (RQ-PCR)

For RQ-PCR, total RNA was extracted from samples of treated or untreated stromal cells using TRIZOL reagent (Invitrogen) according to the manufacturer's protocol and quantified by spectrophotometrical methods. 2 μg of total RNA were treated with DNase (GE Healthcare) and then transcribed into first strand cDNA using First Strand cDNA Synthesis Kit (Amersham Biosciences/GE Healthcare) modified as previously described.⁷⁶ Briefly, RNA was treated with 20 U of DNase I (Amersham Biosciences/GE Healthcare) in Tris–HCl 40 mM pH 7.5, MgCl₂ 6 mM buffer, after 15 minutes DNase was inactivated by heating at 80 °C for 10 minutes. For the cDNA synthesis, we proceeded similar to manufacturer's protocol except that 500 ng of random hexamers were used.

RQ-PCR assays were performed using the Applied Biosystems 7500 Systems (Applied Biosystems) and each sample was run in triplicates. All PCR reactions were carried out in a final volume of 25 μL containing 1X of SYBR Green PCR Master Mix (Applied Biosystems), a previously determined concentration of each gene specific primers (Table 1), 1 μL stromal cell cDNA, and sterile deionized water. The standard cycling condition was 50 °C for 2 min, 90 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Data were analyzed by relative quantification using GADPH as endogenous control and non-treated BM stromal cell cDNA as the calibrator sample.

Plasmid constructs

In order to obtain all STC1 constructs we used pGEMT–STC1 full length (1–247)¹⁷construct as template. To generate pBTM116–STC1ΔSPa (STC1 lacking the first 22 amino acids, including the signal peptide), a construct used in the yeast two-hybrid screen as a bait, which encodes a LexA fused protein, we PCR amplified STC1 (23–247) with STC1ΔSPa F (5′-aaggatccCAGAATGACTCTGTGAGCCC-3′) and STC1 Rev (5′-ccgaattCCTCTCCCTGGTTATGCAC-3′) primers, cloned it into plasmid pUC18 (Fermentas) and then sub-cloned it into yeast bait vector pBTM116 plasmid. For pBTM116-STC1NtermΔPSb (STC1 amino-termini lacking the signal peptide, but coding residues 18 to 134) and pBTM116Cterm (STC1 carboxy end, coding residues 129 to 247), we PCR amplified the inserts with STC1ΔPSb F2 (5′-gaattcACCCATGAGGCGGAGCAG-3′) and NSTC1 + stop R2 (5′-ggatccttaCACATTCAGCTTGCTGTAG-3′) primers for the amino-terminus and with CSTC1 F2 (5′-gaattcTACAGCAAGCTGAATGTGTG-3′) and CSTC1 R2 (5′-ggatccTTATGCACTCTCATGGGATG-3′) primers for the carboxyl end fragment. The PCR products were cloned into pGEM plasmid and then sub-cloned into the pBTM116 yeast plasmid.

In order to obtain constructs used in the GST-pull down assays, all nucleotide sequences encoding the proteins retrieved to interact with the STC1 (except that encoding FNDC4) were sub-cloned from the vector pACT2 (Clontech) to the modified bacterial expression vector pET28a-GST-Tev⁷⁷ resulting in GST fusion protein constructs. The FDNC4 sequence was PCR amplified from the pACT2 vector with FNDC4 F (5′-aaggatccCGGCCTCCCTCTCCTGTG-3′) and FNDC4 R (5′-gggaattcACTCAAACGTCGATGGTGTTG-3′) primers and the obtained construct (GST-FNDC4 [FIII]) lacks the carboxyl terminus portion, which encodes its transmembrane and cytoplasmic domains, and which had been present in the clone retrieved in the yeast two hybrid screen. It had to be removed in order to allow soluble protein expression in E. coli.

Full length Ki-1/57 cDNA was cloned into the bacterial expression vector pGEX 2TK to allow its expression as a GST-tagged fusion protein.

For the pEGFPN₂–STC1 construct, used for the HEK293 cell transfections, we perform a sub-cloning of the STC1 gene from pGEM-STC1, resulting in fluorescent GFP fusion recombinant protein construct.

Yeast two-hybrid screen and DNA sequence analyses

The yeast two-hybrid screens of three human cDNA libraries (Clontech) from fetal brain, bone marrow and leukocytes were performed using the yeast strain L40 (trp1-901, his3del200, leu2-3, ade2 LYS2::(lexAop)4-HIS3 URA3::(lexAop)8-lac GAL4) and STC1ΔPSa as a bait in pBTM116 vector. L40 was transformed according to the protocols supplied by Clontech and HIS3 and LacZ reporters were assayed as previously described.⁷⁸

The screens were performed on Synthetic Dropout minimal medium plates (SD) without tryptophan, leucine, and histidine, according to the library screening protocols supplied by Clontech. Recombinant pACT2 plasmids of positive clones for activation of both reporters genes were isolated. The obtained DNA sequence data were annotated using Basic Local Alignment Search Tool (BLAST) web interface⁷⁹ and clones aligning to genomic contigs or untranslated regions were discarded. All the constructions in this article were verified by DNA sequencing in order to ascertain the correct nucleotide sequences by DNA ABI PRISM 377 Genetic Analyzer (Applied Biosystems).

Mapping of interactions

Both constructs, amino (LexA–N STC1ΔPSb) and carboxy end (LexA–C STC1) were tested against prey constructs retrieved from yeast two hybrid screenings additionally to the full length construction (LexA–STC1ΔPSa) as a positive control and to the empty pACT2 plasmid, as a negative control, assayed in the same way that we described above.

Protein expression and purification

Bacterial expression of yeast two hybrid retrieved proteins and of Ki-1/57 was obtained expressing GST fusion protein constructs in Escherichia coliBL21 (DE3) pRARE cells at the following conditions: (1) 37 °C using 0.5 mM isopropyl 1-thio-B-D-galactopyranoside (IPTG) for 4 h (GST–FNDC4[FIII], GST–JRSP1 and GST–SUMO1), (2) 25 °C using 10 mM lactose for 16 hours (GST–FUS, GST–LMNA, GST–SP100 and GST–TMEM132A) and (3) 37 °C using 0.5 mM of isopropyl-b-D-thiogalactoside (IPTG) for 4 hours (GST control and GST-Ki-1/57). After harvest and lysis the resulting suspension was cleared by centrifugation. The obtained supernatant was loaded onto a GST-Trap (Amersham) and eluted in buffer (50 mM Tris–HCl pH 8.0; 50 mM NaCl; 0.1 mM EDTA; 20 mM reduced glutathione). The obtained GST-affinity purified fractions were pooled and dialyzed against the buffer: 50 mM Tris–HCl pH 8.0; 50 mM NaCl; 0.1 mM EDTA (GST-Ki-1/57). STC1-HT was expressed in insect cells and purified as previously described.¹⁷

GST pull-down assay

For each construct, 25 mL of transformed E. coliBL21(DE3) pRARE culture were harvested by centrifugation at 4500× g for 10 min, and the cell pellet was resuspended and incubated for 1 hour on ice in 1 mL of lysis buffer (PBS [137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.4] containing 0.2 mg mL⁻¹lysozyme, 1 μg mL⁻¹ DNaseI, 1 μg mL⁻¹RNAse, 1mM PMSF and 6 mM MgCl₂) and lysed by sonication. Samples were centrifuged at 20 000× g for 30 min at 4 °C and the supernatant was used as soluble total lysate.

One millilitre of total lysate of all GST tagged constructs or of GST alone was allowed to bind to 25 μL of Glutathione-Uniflow Resin (Clontech) for 1 h at 4 °C. After incubation, the beads containing bound recombinant proteins were washed three times with ice cold PBS. Twenty-five μg of purified STC1-HT protein were added to the resins containing GST fusion proteins or GST alone and further incubated in 0.1 mL of PBS for 4 h at 4 °C to allow protein–protein interactions to occur. Beads were then washed three times with 0.5 mL of PBS, followed by three washes with 0.5 mL of PBS containing 0.2% Triton-X-100, then three washes with 0.5 mL of PBS without Triton X-100. Resin-bound protein complexes were resolved on two separate 12.5% SDS-PAGE polyacrylamide gels and after electrophoresis, the proteins were transferred to PVDF membranes by semi-dry electroblotting. Membranes were blocked with 5% BSA in TTBS (0.15 M NaCl, 20 mM Tris–HCl, 0.05% Tween-20, pH 7.2) for 1 hour, then incubated either with a mouse monoclonal anti-His tag (1 ∶ 5000; QIAgen) or mouse monoclonal anti-GST⁸⁰ in blocking solution for 1 h. After three washes with TTBS membranes were incubated with the HRP-conjugated anti-mouse antibody (1 ∶ 5000; Santa Cruz Biotechnology) for 1 h and washed again three times with TTBS. Blots were developed by Luminol reagent (Santa Cruz).

Co-immunoprecipitation

HEK293 cells from three 175 cm² flasks cultivated under 37 °C and 5% CO₂ (∼70% confluency) were washed 3 times with PBS and lyzed with 500 μL lysis buffer (Triton-X100 0.5%, NaCl 150 mM, Tris 10 mM, EDTA 1 mM, EGTA 1 mM, PMSF 0.5 mM) at 4 °C under agitation for 1 hour. The lysate was sonicated in 6 cycles of 5 seconds pulse/15 seconds paused, under 40% pulse intensity and centrifuged at 14 000 rpm at 4 °C for 15 minutes. For pre-clearing the supernatant was incubated at 4 °C for 30 minutes, first with 30 μL of Protein G Resin (Amersham).

30 μL of new beads were then washed 3 times with lysis buffer (without Triton-X-100) and then incubated for 1 h at 4 °C with 5 μg of goat polyclonal anti-STC1 antiserum (Santa Cruz Biotechnology) and finally washed 3 times with lysis buffer (without Triton-X-100). After 3 hours of incubation with the pre-cleared lysate from above, the beads were recovered and washed 3 times with the lysis buffer (without Triton-X-100). To these beads, 30 μL of SDS 4X buffer were added (Tris–HCl 250 mM pH 6.8, SDS 0.8%, bromophenol blue 0.2%, glycerol 45.5%, 2-mercaptoetanol 20%). This mix was heated to 95 °C for 15 minutes to recover the antibody–protein complexes. The eluates were analyzed by Western blot using mouse monoclonal anti-STC1 (1 ∶ 5000—Abnova Corporation) or against rabbit polyclonal anti-SUMO-1 (1 ∶ 650—Abgent).

Phylogenetic analyses

The complete protein sequence of human STC1 was submitted to the PSI-BLAST web tool. After 5 subsequent submissions excluding alignments with synthetic, hypothetic and predicted sequences, we obtained alignment with STC1 orthologs from 23 different species. The 24 sequences were aligned using the COBALT platform (http://www.ncbi.nlm.nih.gov/tools/cobalt/) and a phylogenetic tree was built using the Fast Minimum evolution tree method, with a maximum difference among sequences of 0.85 using the Kimura protein distance.

In vitro SUMOylation and SUMO E3 ligase assays

For in vitroSUMOylation assays of STC1 we use a SUMOylation Kit (BIOMOL) according to the manufacturer's instructions. For in vitro SUMO E3 ligase assays, we used 200 nM of RanGAP1–GST protein provided by the kit as well as 200 nM Ki-1/57-GST. GST (200 nM) was used as control. The reactions were performed in the presence or absence of 2 pmol of STC1-HT to evaluate its SUMO E3 ligase activity towards RanGAP1–GST or Ki-1/57-GST, two known SUMO-1 modified proteins. For controls all reactions were carried out in the absence of ATP.

The reactions were analyzed by Western blot using rabbit polyclonal anti-SUMO1 (1 ∶ 1000—BIOMOL), or mouse monoclonal anti-Ki-1/57 (A26),⁸¹ followed by membrane stripping and analysis by mouse monoclonal anti-GST antibody.

Sub-cellular localization of STC1

For the sub-cellular localization assays, HEK293 cells were grown on glass coverslips with DMEM culture + 10% FCS medium and transfections were performed by using the calcium phosphate method, using pEGFPN₂–STC1 construct and empty pEGFPN₂ as control. K562 cells were grown on 75 cm² flasks with the required medium and 10⁵cells were pelleted on the slices by the cytospin technique, at 800× g for 5 minutes. Cells were fixed in a solution containing 2% (w/v) paraformaldehyde, 50 μM Taxol and 50 mM EGTA at room temperature for 20 min, and then permeabilized and blocked in a mixture of 0.3% (v/v) Triton-X-100 and 3% (w/v) glycine solution in PBS at room temperature for 30 min. The K562 cells were incubated at room temperature for 1 h with a specific mouse monoclonal anti-STC1 (1 ∶ 100) (that had been generated by three subsequent immunizations of mice with 1 mg STC1-HT protein), in PBS containing 0.1% BSA (w/v). Subsequently, the K562 cells were incubated at room temperature for 1 h, with rhodamine conjugated secondary antibody (Santa Cruz Biotechnology) goat anti-mouse (1 ∶ 200). As a control we used, a slide with K562 cells incubated only with the secondary antibody. DAPI (Molecular Probes) dye was used to stain the nuclei of cells. Cells were examined on a Nikon fluorescence microscope. Confocal microscopy analysis was performed on a Axioplan Carl Zeiss LSM 510 META microscope.

Protein–protein interaction network construction

The Osprey program⁸² was used for STC1 interaction network construction and complemented by searches on the Biological General Repository for Interaction Datasets (BioGRID).^83,84

Acknowledgements

Financially supported by: FAPESP (www.fapesp.br), CNPq (www.cnpq.br) and CNPEM (www.lnls.br/site/home.aspx). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We thank Maria Eugenia R. Camargo for technical assistance, Zildene D. Correa for DNA sequencing support, Dr Silvia Regina Brandalise for providing patient samples, Angelo B. A. Laranjeira for assistance with the co-culture of ALL cells onto BM stromal layers and with cytospin technique and Prof. Dr Sara T. O. Saad for access to confocal microscopy facilities and Janine S. Schincariol and Pedro Bordeaux-Rego for technical support with the confocal.

References

1. J. Folkman, Angiogenesis in cancer, vascular, rheumatoid and other disease, Nat. Med., 1995, 1, 27–31.
2. P. O. Iversen, D. R. Sorensen and H. B. Benestad, Inhibitors of angiogenesis selectively reduce the malignant cell load in rodent models of human myeloid leukemias, Leukemia, 2002, 16, 376–381.
3. J. P. Veiga, L. F. Costa, S. E. Sallan, L. M. Nadler and A. A. Cardoso, Leukemia-stimulated bone marrow endothelium promotes leukemia cell survival, Exp. Hematol., 2006, 34, 610–621.
4. A. C. Chang, J. Janosi, M. Hulsbeek, D. de Jong, K. J. Jeffrey, J. R. Noble and R. R. Reddel, A novel human cDNA highly homologous to the fish hormone stanniocalcin, Mol. Cell. Endocrinol., 1995, 112, 241–247.
5. A. C. Chang, D. A. Jellinek and R. R. Reddel, Mammalian stanniocalcins and cancer, Endocr. Relat. Cancer, 2003, 10, 359–373.
6. R. A. Wascher, K. T. Huynh, A. E. Giuliano, N. M. Hansen, F. R. Singer, D. Elashoff and D. S. Hoon, Stanniocalcin-1: a novel molecular blood and bone marrow marker for human breast cancer, Clin. Cancer Res., 2006, 9, 1427–1435.
7. M. E. Gerritsen, R. Soriano, S. Yang, G. Ingle, C. Zlot, K. Toy, J. Winer, A. Draksharapu, F. Peale, T. D. Wu and P. M. Williams, In silico data filtering to identify new angiogenesis targets from a large in vitro gene profiling data set, Physiol. Genomics, 2002, 10, 13–20.
8. (a) D. M. Anderson, B. J. Beres, J. Wilson-Rawls and A. Rawls, The homeobox gene Mohawk represses transcription by recruiting the sin3A/HDAC co-repressor complex, Dev. Dyn., 2009, 238, 572–580; (b) J. Kahn, F. Mehraban, G. Ingle, X. Xin, J. E. Bryant, G. Vehar, J. Schoenfeld, C. J. Grimaldi, F. Peale, A. Draksharapu, D. A. Lewin and M. E. Gerritsen, Gene expression profiling in an in vitro model of angiogenesis, Am. J. Pathol., 2000, 156, 1887–1900.
9. D. Liu, H. Jia, D. I. Holmes, A. Stannard and I. Zachary, Vascular endothelial growth factor-regulated gene expression in endothelial cells: KDR-mediated induction of Egr3 and the related nuclear receptors Nur77, Nurr1, and Nor1, Arterioscler., Thromb., Vasc. Biol., 2003, 23, 2002–2007.
10. K. K. Wary, G. D. Thakker, J. O. Humtsoe and J. Yang, Analysis of VEGF-responsive genes involved in the activation of endothelial cells, Mol. Cancer, 2003, 2, 25.
11. D. Sheikh-Hamad, Mammalian stanniocalcin-1 activates mitochondrial antioxidant pathways: new paradigms for regulation of macrophages and endothelium, Am. J. Physiol.: Renal Physiol., 2010, 298, F248–F254.
12. C. R. Mccudden, K. A. James, C. Hasilo and G. F. Wagner, Characterization of mammalian stanniocalcin receptors. Mitochondrial targeting of ligand and receptor for regulation of cellular metabolism, J. Biol. Chem., 2002, 277, 45249–45258.
13. C. P. Hasilo, C. R. Mccudden, J. R. Gillespie, K. A. James, E. R. Hirvi, D. Zaidi and G. F. Wagner, Nuclear targeting of stanniocalcin to mammary gland alveolar cells during pregnancy and lactation, Am. J. Physiol.: Endocrinol. Metab., 2005, 289, E634–E642.
14. D. A. Jellinek, A. C. Chang, M. R. Larsen, X. Wang, P. J. Robinson and R. R. Reddel, Stanniocalcin 1 and 2 are secreted as phosphoproteins from human fibrosarcoma cells, Biochem. J., 2000, 350(Pt 2), 453–461.
15. J. Lomax, Get ready to GO! A biologist's guide to the Gene Ontology, Briefings Bioinf., 2005, 6, 298–304.
16. K. Nakai and P. Horton, PSORT: a program for detecting sorting signals in proteins and predicting their subcellular localization, Trends Biochem. Sci., 1999, 24, 34–36.
17. D. M. Trindade, J. C. Silva, M. S. Navarro, C. Torriani and J. Kobarg, Low-resolution structural studies of human Stanniocalcin-1, BMC Struct. Biol., 2009, 9, 57.
18. K. F. Kelly and J. M. Daniel, POZ for effect—POZ-ZF transcription factors in cancer and development, Trends Cell Biol., 2006, 16, 578–587.
19. J. Raberger, A. Schebesta, S. Sakaguchi, N. Boucheron, K. E. Blomberg, A. Berglof, T. Kolbe, C. I. Smith, T. Rulicke and W. Ellmeier, The transcriptional regulator PLZF induces the development of CD44 high memory phenotype T cells, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 17919–17924.
20. Y. Xue, F. Zhou, C. Fu, Y. Xu and X. Yao, SUMOsp: a web server for SUMOylation site prediction, Nucleic Acids Res., 2006, 34(1), W254–W257.
21. M. B. Kroetz and M. Hochstrasser, SUMO protocols, Human Press, 2009, vol. 1, pp. 107–120.
22. G. F. Wagner and G. E. Dimattia, The stanniocalcin family of proteins, J. Exp. Zool. A Comp. Exp. Biol., 2006, 305(9), 769–780.
23. G. Tremblay, P. Bernier-Dodier, L. Delbecchi, G. F. Wagner, B. G. Talbot and P. Lacasse, Local control of mammary involution: is stanniocalcin-1 involved?, J. Dairy Sci., 2009, 92(5), 1998–2006.
24. G. Basini, L. Baioni, S. Bussolati, S. Grolli, L. H. Kramer, G. F. Wagner and F. Grasselli, Expression and localization of stanniocalcin 1 in swine ovary, Gen. Comp. Endocrinol., 2010, 166(2), 404–408.
25. J. Johnston, Y. Ramos-Valdes, L. A. Stanton, S. Ladhani, F. Beier and G. E. Dimattia, Human stanniocalcin-1 or -2 expressed in mice reduces bone size and severely inhibits cranial intramembranous bone growth, Transgenic Res., 2010 DOI:10.1007/s11248-010-9376-7.
26. A. Zimber, Q. D. Nguyen and C. Gespach, Nuclear bodies and compartments: functional roles and cellular signalling in health and disease, Cell. Signalling, 2004, 16, 1085–1104.
27. T. H. Shen, H. K. Lin, P. P. Scaglioni, T. M. Yung and P. P. Pandolfi, The mechanisms of PML-nuclear body formation, Mol. Cell, 2006, 3, 331–339.
28. Y. Zhang, R. Iratni, H. Erdjument-Bromage, P. Tempst and D. Reinberg, Histone deacetylases and SAP18, a novel polypeptide, are components of a human Sin3 complex, Cell, 1997, 89(3), 357–364.
29. Y. Tohmiya, Y. Koide, S. Fujimaki, H. Harigae, T. Funato, M. Kaku, T. Ishii, Y. Munakata, J. Kameoka and T. Sasaki, Stanniocalcin-1 as a novel marker to detect minimal residual disease of human leukemia, Tohoku J. Exp. Med., 2004, 204(2), 125–133.
30. R. I. Kumaran, R. Thakar and D. L. Spector, Chromatin dynamics and gene positioning, Cell, 2008, 132, 929–934.
31. R. I. Kumaran, B. Muralikrishna and V. K. Parnaik, Lamin A/C speckles mediate spatial organization of splicing factor compartments and RNA polymerase II transcription, J. Cell Biol., 2002, 159, 783–793.
32. N. Sylvius, Z. T. Bilinska, J. P. Veinot, A. Fidzianska, P. M. Bolongo, S. Poon, P. McKeown, R. A. Davies, K. L. Chan, A. S. Tang, S. Dyack, J. Grzybowski, W. Ruzyllo, H. McBride and F. Tesson, In vivo and in vitro examination of the functional significances of novel lamin gene mutations in heart failure patients, J. Med. Genet., 2005, 42, 639–647.
33. N. Sylvius, A. Hathaway, E. Boudreau, P. Gupta, S. Labib, P. M. Bolongo, P. Rippstein, H. Mcbride, Z. T. Bilinska and F. Tesson, Specific contribution of lamin A and lamin C in the development of laminopathies, Exp. Cell Res., 2008, 314, 2362–2375.
34. M. K. Andersson, A. Stahlberg, Y. Arvidsson, A. Olofsson, H. Semb, G. Stenman, O. Nilsson and P. Aman, The multifunctional FUS, EWS and TAF15 proto-oncoproteins show cell type-specific expression patterns and involvement in cell spreading and stress response, BMC Cell Biol., 2008, 9, 37.
35. W. Hackl and R. Luhrmann, Molecular cloning and subcellular localization of the snRNP-associated protein 69 KD, a structural homologue of the proto-oncoproteins TLS and EWS with RNA and DNA-binding properties, J. Mol. Biol., 1996, 264, 843–851.
36. H. Y. Yeung, K. P. Lai, H. Y. Chan, N. K. Mak, G. F. Wagner and C. K. Wong, Hypoxia-inducible factor-1-mediated activation of stanniocalcin-1 in human cancer cells, Endocrinology, 2005, 146, 4951–4960.
37. D. Ito, J. R. Walker, C. S. Thompson, I. Moroz, W. Lin, M. L. Veselits, A. M. Hakim, A. A. Fienberg and G. Thinakaran, Characterization of stanniocalcin 2, a novel target of the mammalian unfolded protein response with cytoprotective properties, Mol. Cell. Biol., 2004, 24, 9456–9469.
38. S. M. Oh, Z. Liu, M. Okada, W. Jang, X. Liu, C. B. Chan, H. Luo and K. Ye, Ebp1 SUMOylation, regulated by TLS/FUS E3 ligase, is required for its anti-proliferative activity, Oncogene, 2010, 29(7), 1017–1030.
39. M. J. Edelmann, A. Iphöfer, M. Akutsu, M. Altun, K. di Gleria, H. B. Kramer, E. Fiebiger, S. Dhe-Paganon and B. M. Kessler, Structural basis and specificity of human otubain 1-mediated deubiquitination, Biochem. J., 2009, 418(2), 379–390.
40. F. C. Nery, D. O. Passos, V. S. Garcia and J. Kobarg, Ki-1/57 interacts with RACK1 and is a substrate for the phosphorylation by phorbol 12-myristate 13-acetate-activated protein kinase C, J. Biol. Chem., 2004, 279(12), 11444–11455.
41. I. Gillot, C. Matthews, D. Puel, F. Vidal and P. Lopez, Ret Finger Protein: An E3 Ubiquitin Ligase Juxtaposed to the XY Body in Meiosis, Int. J. Cell Biol., 2009, 524858.
42. J. P. Bocock, S. Carmicle, E. Madamba and A. H. Erickson, Nuclear Targeting of an Endosomal E3 Ubiquitin Ligase, Traffic, 2010, 11(6), 756–766.
43. Y. Takahashi, V. Lallemand-Breitenbach, J. Zhu and H. de The, PML nuclear bodies and apoptosis, Oncogene, 2004, 23, 2819–2824.
44. V. Lallemand-Breitenbach, J. Zhu, F. Puvion, M. Koken, N. Honore, A. Doubeikovsky, E. Duprez, P. P. Pandolfi, E. Puvion, P. Freemont and H. de The, Role of promyelocytic leukemia (PML) SUMOlation in nuclear body formation, 11S proteasome recruitment, and As₂O₃-induced PML or PML/retinoic acid receptor alpha degradation, J. Exp. Med., 2001, 193, 1361–1371.
45. P. Dhordain, O. Albagli, N. Honore, F. Guidez, D. Lantoine, M. Schmid, H. D. The, A. Zelent and M. H. Koken, Colocalization and heteromerization between the two human oncogene POZ/zinc finger proteins, LAZ3 (BCL6) and PLZF, Oncogene, 2000, 19, 6240–6250.
46. E. de Stanchina, E. Querido, M. Narita, R. V. Davuluri, P. P. Pandolfi, G. Ferbeyre and S. W. Lowe, PML is a direct p53 target that modulates p53 effector functions, Mol. Cell, 2004, 13, 523–535.
47. E. Krieghoff-Henning and T. G. Hofmann, Role of nuclear bodies in apoptosis signaling, Biochim. Biophys. Acta, 2008, 1783, 2185–2194.
48. K. P. Lai, A. Y. Law, H. Y. Yeung, L. S. Lee, G. F. Wagner and C. K. Wong, Induction of stanniocalcin-1 expression in apoptotic human nasopharyngeal cancer cells by p53, Biochem. Biophys. Res. Commun., 2007, 356, 968–975.
49. S. Weidtkamp-Peters, T. Lenser, D. Negorev, N. Gerstner, T. G. Hofmann, G. Schwanitz, C. Hoischen, G. Maul, P. Dittrich and P. Hemmerich, Dynamics of component exchange at PML nuclear bodies, J. Cell Sci., 2008, 121(16), 2731–2743.
50. E. van Damme, K. Laukens, T. H. Dang and X. van Ostade, A manually curated network of the PML nuclear body interactome reveals an important role for PML-NBs in SUMOylation dynamics, Int. J. Biol. Sci., 2010, 6(1), 51–67.
51. S. A. McCallum, J. F. Bazan, M. Merchant, J. Yin, B. Pan, F. J. de Sauvage and W. J. Fairbrother, Structure of SAP18: a ubiquitin fold in histone deacetylase complex assembly, Biochemistry, 2006, 45, 11974–11982.
52. A. R. Daniel and C. A. Lange, Protein kinases mediate ligand-independent derepression of sumoylated progesterone receptors in breast cancer cells, Proc. Natl. Acad. Sci. U. S. A., 2009, 106(34), 14287–14292.
53. C. Gao, X. Cheng, M. Lam, Y. Liu, Q. Liu, K. S. Chang and H. Y. Kao, Signal-dependent regulation of transcription by histone deacetylase 7 involves recruitment to promyelocytic leukemia protein nuclear bodies, Mol. Biol. Cell., 2008, 19, 3020–3027.
54. C. Gao, C. C. Ho, E. Reineke, M. Lam, X. Cheng, K. J. Stanya, Y. Liu, S. Chakraborty, H. M. Shih and H. Y. Kao, Histone deacetylase 7 promotes PML SUMOylation and is essential for PML nuclear body formation, Mol. Cell. Biol., 2008, 28, 5658–5667.
55. K. Oh-Hashi, Y. Naruse, F. Amaya, G. Shimosato and M. Tanaka, Cloning and characterization of a novel GRP78-binding protein in the rat brain, J. Biol. Chem., 2003, 278, 10531–10537.
56. J. D. Malhotra and R. J. Kaufman, The endoplasmic reticulum and the unfolded protein response, Semin. Cell Dev. Biol., 2007, 18, 716–731.
57. J. N. Gass, H. Y. Jiang, R. C. Wek and J. W. Brewer, The unfolded protein response of B-lymphocytes: PERK-independent development of antibody-secreting cells, Mol. Immunol., 2008, 45, 1035–1043.
58. T. Iwawaki, R. Akai, S. Yamanaka and K. Kohno, Function of IRE1 alpha in the placenta is essential for placental development and embryonic viability, Proc. Natl. Acad. Sci. U. S. A., 2009, 106(39), 16657–16662.
59. E. Gouadon, R. P. Schuhmeier, D. Ursu, A. A. Anderson, S. Treves, F. Zorzato, F. Lehmann-Horn and W. Melzer, A possible role of the junctional face protein JP-45 in modulating Ca²⁺ release in skeletal muscle, J. Physiol., 2006, 572, 269–280.
60. E. Rios, Calcium signalling in muscle: a milestone for modulation studies, J. Physiol., 2006, 572, 1–2.
61. H. S. Olsen, M. A. Cepeda, Q. Q. Zhang, C. A. Rosen and B. L. Vozzolo, Human stanniocalcin: a possible hormonal regulator of mineral metabolism, Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 1792–1796.
62. G. F. Wagner, B. L. Vozzolo, E. Jaworski, M. Haddad, R. L. Kline, H. S. Olsen, C. A. Rosen, M. B. Davidson and J. L. Renfro, Human stanniocalcin inhibits renal phosphate excretion in the rat, J. Bone Miner. Res., 1997, 12, 165–171.
63. K. L. Madsen, M. M. Tavernini, C. Yachimec, D. L. Mendrick, P. J. Alfonso, M. Buergin, H. S. Olsen, M. J. Antonaccio, A. B. Thomson and R. N. Fedorak, Stanniocalcin: a novel protein regulating calcium and phosphate transport across mammalian intestine, Am. J. Physiol., 1998, 274, G96–G102.
64. D. Sheikh-Hamad, R. Bick, G. Y. Wu, B. M. Christensen, P. Razeghi, B. Poindexter, H. Taegtmeyer, A. Wamsley, R. Padda, M. Entman, S. Nielsen and K. Youker, Stanniocalcin-1 is a naturally occurring L-channel inhibitor in cardiomyocytes: relevance to human heart failure, Am. J. Physiol. Heart Circ. Physiol., 2003, 285, H442–H448.
65. D. Sheikh-Hamad, D. Rouse and Y. Yang, Regulation of stanniocalcin in MDCK cells by hypertonicity and extracellular calcium, Am. J. Physiol. Renal Physiol., 2000, 278, F417–F424.
66. K. Koizumi, M. Hoshiai, H. Ishida, K. Ohyama, H. Sugiyama, A. Naito, T. Toda, H. Nakazawa and S. Nakazawa, Stanniocalcin 1 prevents cytosolic Ca²⁺ overload and cell hypercontracture in cardiomyocytes, Circ. J., 2007, 71, 796–801.
67. J. P. Ellard, C. R. McCudden, C. Tanega, K. A. James, S. Ratkovic, J. F. Staples and G. F. Wagner, The respiratory effects of stanniocalcin-1 (STC1) on intact mitochondria and cells: STC1 uncouples oxidative phosphorylation and its actions are modulated by nucleotide triphosphates, Mol. Cell. Endocrinol., 2007, 264, 90–101.
68. P. P. Roux and J. Blenis, ERK and p38 MAPK-activated protein kinases: a family of protein kinases with diverse biological functions, Microbiol. Mol. Biol. Rev., 2004, 68, 320–344.
69. D. I. Holmes and I. C. Zachary, Vascular endothelial growth factor regulates stanniocalcin-1 expression via neuropilin-1-dependent regulation of KDR and synergism with fibroblast growth factor-2, Cell. Signalling, 2008, 20, 569–579.
70. K. James, M. Seitelbach, C. R. Mccudden and G. F. Wagner, Evidence for stanniocalcin binding activity in mammalian blood and glomerular filtrate, Kidney Int., 2005, 67, 477–482.
71. S. V. Bhagwat, J. Lahdenranta, R. Giordano, W. Arap, R. Pasqualini and L. H. Shapiro, CD13/APN is activated by angiogenic signals and is essential for capillary tube formation, Blood, 2001, 97, 652–659.
72. R. Pasqualini, E. Koivunen, R. Kain, J. Lahdenranta, M. Sakamoto, A. Stryhn, R. A. Ashmun, L. H. Shapiro, W. Arap and E. Ruoslahti, Aminopeptidase N is a receptor for tumor-homing peptides and a target for inhibiting angiogenesis, Cancer Res., 2000, 60, 722–727.
73. J. Zhong, Y. Wang, X. Qiu, X. Mo, Y. Liu, T. Li, Q. Song, D. Ma and W. Han, Characterization and expression profile of CMTM3/CKLFSF3, J. Biochem. Mol. Biol., 2006, 39, 537–545.
74. Y. Wang, T. Li, X. Qiu, X. Mo, Y. Zhang, Q. Song, D. Ma and W. Han, CMTM3 can affect the transcription activity of androgen receptor and inhibit the expression level of PSA in LNCaP cells, Biochem. Biophys. Res. Commun., 2008, 371, 54–58.
75. A. Teufel, N. Malik, M. Mukhopadhyay and H. Westphal, Frcp1 and Frcp2, two novel fibronectin type III repeat containing genes, Gene, 2002, 297(1–2), 79–83.
76. E. Beillard, N. Pallisgaard, d. V. V. Van, W. Bi, R. Dee, S. E. van der, E. Delabesse, E. Macintyre, E. Gottardi, G. Saglio, F. Watzinger, T. Lion, J. J. van Dongen, P. Hokland and J. Gabert, Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using ‘real-time’ quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR)—a Europe against cancer program, Leukemia, 2003, 17, 2474–2486.
77. F. R. Carneiro, T. C. Silva, A. C. Alves, T. Haline-Vaz, F. C. Gozzo and N. I. Zanchin, Spectroscopic characterization of the tumor antigen NY-REN-21 and identification of heterodimer formation with SCAND1, Biochem. Biophys. Res. Commun., 2006, 343, 260–268.
78. E. M. Assmann, M. R. Alborghetti, M. E. Camargo and J. Kobarg, FEZ1 dimerization and interaction with transcription regulatory proteins involves its coiled-coil region, J. Biol. Chem., 2006, 281, 9869–9881.
79. M. Johnson, I. Zaretskaya, Y. Raytselis, Y. Merezhuk, S. McGinnis and T. L. Madden, NCBI BLAST: a better web interface, Nucleic Acids Res., 2008, 36, W5–W9.
80. T. A. Lemos, D. O. Passos, F. C. Nery and J. Kobarg, Characterization of a new family of proteins that interact with the C-terminal region of the chromatin-remodeling factor CHD-3, FEBS Lett., 2003, 533, 14–20.
81. J. Kobarg, S. Schnittger, C. Fonatsch, H. Lemke, M. A. Bowen, F. Buck and H. P. Hansen, Characterization, mapping and partial cDNA sequence of the 57 kDa intracellular Ki-1 antigen, Exp. Clin. Immunogenet., 1997, 14, 273–280.
82. B. J. Breitkreutz, C. Stark and M. Tyers, Osprey: a network visualization system, Genome Biol., 2003, 4, R22.
83. B. J. Breitkreutz, C. Stark, T. Reguly, L. Boucher, A. Breitkreutz, M. Livstone, R. Oughtred, D. H. Lackner, J. Bahler, V. Wood, K. Dolinski and M. Tyers, The BioGRID Interaction Database: 2008, Nucleic Acids Res., 2008, 36, D637–D640.
84. C. Stark, B. J. Breitkreutz, T. Reguly, L. Boucher, A. Breitkreutz and M. Tyers, BioGRID: a general repository for interaction datasets, Nucleic Acids Res., 2006, 34, D535–D539.
85. H. D. Ulrich, The SUMO system: an Overview, SUMO protocols, Human Press, 2009, vol. 1, pp. 3–16.
86. S. H. Yang, A. Galanis, J. Witty and A. D. Sharrocks, An extended consensus motif enhances the specificity of substrate modification by SUMO, EMBO J., 2006, 25, 5083–5093.
87. D. Maglott, J. Ostell, K. D. Pruitt and T. Tatusova, Entrez Gene: gene-centered information at NCBI, Nucleic Acids Res., 2007, 35, D26–D31.

---

Footnotes

† Electronic supplementary information (ESI) available: Phylogenetic analyses of putatives SUMOylation sites of STC1 from 24 different species. See DOI: 10.1039/c0mb00088d

‡ MTS and DMT contributed equally to this article.

---