In neuroblastoma cells, apoptotic programs can be activated by cytokines and cytostatic drugs. Apoptotic dysfunction confers resistance against therapeutic drugs and is a major complication for achieving optimal therapy response. Deregulated expression of the MYCN gene is a critical determinant in neuroblastoma progression, and one of the pleiotropic functions of the MYCN protein is cellular sensitization to cytokine-induced and drug-induced apoptosis. By using the functional approach of technical knockout (TKO), we have identified five genes that regulate sensitization for IFN-γ-induced cell death. Most efficient among them is the newly identified SOXN (neuroblastoma-derived sulfhydryl oxidase), which comprises 12 exons and maps to 9q34.3. SOXN encodes a putative protein of 698 amino acids that contains a signal sequence, a protein-disulfide-isomerase-type thioredoxin and a yeast ERV1 domain and is highly homologous to members of the sulfhydryl oxidase/Quiescin6 family. The SOXN protein is predominantly located in the plasma and in the nuclear membrane. Antisense SOXN confers resistance to IFN-γ-induced apoptosis. In contrast, ectopic overexpression of sense-SOXN sensitizes the cells to induced cell death. These results identify SOXN as a major player in regulating the sensitization of neuroblastoma cells for IFN-γ-induced apoptosis.

Neuroblastoma, after brain tumors, is the second most common solid tumor in young children. The tumors are biologically heterogeneous, and various clinical and genetic factors have been shown to be predictive of patient outcome and to correlate with therapy response (1). Although age of the patient, stage of disease, tumor cell ploidy, level of TRK-A gene expression, presence or absence of 1p deletion or 17q gain, and MYCN gene copy number are strong prognostic factors, thus far no single genetic change common to neuroblastoma has been identified (2). The spectrum of clinical behavior ranges from rapidly progressing, therapy-resistant tumors to disseminated tumors that regress spontaneously. This clinical and biological heterogeneity indicates the participation of multiple pathways and involved factors.

Normal neuroblast development includes programmed differentiation, as well as programmed cell death (3, 4). Spontaneous regression has been reviewed as the delayed activation of this normal apoptotic pathway. Accordingly, progressively growing tumors would carry a defective apoptotic machinery, resulting in resistance to developmental and therapeutic triggers of cell death (5). The drug-induced and spontaneous remission of neuroblastoma was shown to be associated with programmed cell death (3, 6). While high expression of proapoptotic gene products, including tumor necrosis factor-related apoptosis-inducing ligand, caspase-8 (7), interleukin 1β converting enzyme, CPP32 (8), or CD95 (9), were found to be correlated with improved disease outcome, cells highly expressing BCL-2 show a reduced level of differentiation and apoptosis (10).

Although the general pathways leading to apoptosis are well characterized, little is known about specific genes that sensitize neuroblastoma cells to proapoptotic stimuli. To isolate such genes, we established a functional approach of gene identification in the neuroblastoma Tet21N in vitro system (11). In this cell line, apoptosis can be induced by combining tetracycline-dependent MYCN gene expression with serum withdrawal and cytostatic drug or IFN-γ treatment (12, 13). The functional TKO1 approach is based on the random inactivation of death-promoting genes by transfecting cells with an antisense cDNA library (14). In the presence of the death-inducing low serum and MYCN/IFN-γ environment, only those cells survive in which proapoptotic genes are functionally knocked out by antisense transcripts. Candidate fragments can be isolated from surviving cells because the episomal TKO expression vector does not integrate into the host genome. We have isolated five different candidate genes that augment the susceptibility of neuroblastoma Tet21N cells to the death-inducing stimulus of IFN-γ. Three of them correspond to known genes, whereas the other two represent fragments of as far uncharacterized ESTs. The full-length cDNA of the gene that suppressed apoptosis most efficiently when expressed in antisense was cloned. The predicted protein was shown to be highly homologous to the SOX/Q6 gene family.

SOXs catalyze the oxidation of thiols to disulfides by reducing molecular oxygen to hydrogen peroxide following the reaction 2R-SH + O2→R-S-S-r + H2O2. There are at least three different subclasses of these enzymes that depend on the cofactors iron, copper, or FAD for their catalytic activity (15). FAD-dependent SOXs have been found in lower and higher eukaryotes and were shown to oxidize protein thiols or low molecular weight thiols as glutathione (16, 17). Although the ERV1 SOX was shown to be involved in the yeast cell cycle and in the biogenesis of mitochondria (18), its human homologue ALR1 has been implicated in a range of developmental processes from spermatogenesis to the regeneration of liver tissue (19).

Recently, a new family of FAD-dependent SOX-related genes has been described. All members of the so called SOX/Q6 family contain a PDI-type TRX motif besides a catalytically active ERV1 domain. The human Q6 gene is specifically induced when human fibroblasts enter reversible quiescence (20). Homologous genes were found to be active in the rat male genital tract (21), in the transitional layer of the rat epidermis (22), and in chicken egg white (23). Additional genes with high homology have been described (mouse SOX, Gec3, Q6-related products of Caenorhabditis elegans and Drosophila).

Because of the structural homology of the newly isolated cell death-associated neuroblastoma gene to the SOX/Q6 family, we called the gene SOXN. SOXN is the first described human Q6 homologue containing a PDI-like TRX and the yeast ERV1 domain. The deduced SOXN protein contains, like all other SOX members, a putative signal peptide and a conserved potential N-glycosylation site. The data presented suggest a role of SOXN in sensitization to IFN-γ-induced apoptosis.

Construction of pTKO-based cDNA Expression Library.

A subtractive library, enriched for proapoptotic genes, was constructed from Tet21N neuroblastoma cells by subtracting cDNA of noninduced cells from stimulated cells (treated for 2, 4, 8, 12, and 24 h with MCYN and IFN-γ in 0.5% FCS; Ref. 24). Inserts were isolated by NotI/BamHI restriction, their size ranged from 100 to 800 bp. The fragments were cloned into NotI/BglII-restricted episomal pTKO-CZ (11), resulting in the nondirected, cytomegalovirus promoter-driven cDNA library TKO-SK. The library was introduced into DH10B cells (Stratagene) by electroporation, and 40,000 clones were rescued. Plasmid DNA was isolated and purified by phenol/chloroform extraction. Clones were sequenced with the TKO-specific oligonucleotide TKO-3′ (5′-TATATTTACCTTAGAGCTT).

Rescue of Antisense cDNAs, Transfections, and Cell Culture.

Neuroblastoma cell lines were cultivated in RPMI 1640 as described previously (25). The TKO-SK cDNA library (40 μg of DNA) was transfected into 6 × 105 Tet21N-EBNA-1-3 cells using Effectene (Qiagen; Ref. 26). After 48 h, cultures were split, and selection was initiated with Zeocin (250 μg/ml; Invitrogen) and IFN-γ (500 units/ml; Roche) in RPMI 1640 containing 1% FCS. Selective medium was changed every 3–4 days. After 24 days of selection, the surviving colonies were pooled, expanded for 7 days in the presence of Zeocin and 10% serum, and the pTKO plasmids were prepared (27). Episomal DNA was introduced into DH10B competent Escherichia coli cells (Stratagene) by electroporation. Plasmid DNA was selected from bacterial colonies and sequenced using the pTKO-plasmid-specific primer TKO-3′. Individual episomes were transfected into Tet21N-EBNA-1-3 cells and tested in a second round of Zeocin and IFN-γ selection. For the generation of stable clones, 3 × 105 Tet21N-EBNA-1-3 cells were transfected with individual episomes and subjected 14 days to Zeocin selection (250 μg/ml) in RPMI 1640 containing 10% FCS. Pools of 104 Zeocin-resistant clones were kept as stable polyclonal populations.

Isolation and Sequence Analysis of SOXN Full-Length cDNA.

Homology search with the 130-bp candidate-1 fragment using the BlastN2 algorithm (Husar program package)2 identified an EST with 99% identity (AC: Aw408680), as well as the working draft sequence of the genomic BAC RP11–83N9 clone (AC: AL138781) that spans 155 kbp and consists of eight contigs of unknown order. Hybridization of a 1.2-kb insert of the Aw408680 overlapping IMAGE clone 2118002 (AC: Ai524259) to IMAGE cDNA Clone Collection filters (library 998, RZPD)3 did not identify any positive cDNA clones. Hybridization of the same fragment to the PAC library RPCI1,3-54 identified six positive PAC clones, three of which turned out to be positive in Southern blot analysis (data not shown). By PCR analysis of these three PACs, the eight nonassembled contigs of BAC RP11-83N9 were ordered (data not shown), and the genomic area encompassing the putative gene locus was subjected to the gene prediction program Genscan and the EST-Cluster software (available at HUSAR program package). The expression of putative exons was tested by RT-PCR. The 5′-region of the gene was not included in the genomic AL138781 sequence, but the BLASTN2 search on the Celera publication site5 identified a 124-kbp fragment (GA x54KRCD6M81) encompassing the whole first exon.

Gene-specific primers were designed on the basis of the predicted genomic structure of the new gene. To generate the full-length cDNA of the coding region, two PCR fragments were cloned into Bluescript II SK+ plasmid (Stratagene). As template, first strand cDNA was generated by reverse transcription of 0.1 μg of Tet21N poly(A)+ RNA with 200 units of Superscript II RNaseH Reverse Transcriptase and 0.1 mm oligo(dT) (Life Technologies, Inc.) in a total volume of 25 μl according to the 5′-RACE system (Life Technologies, Inc.). In a first PCR reaction the GC-rich 5′-end was amplified with Pwo polymerase (Roche) using oligonucleotides 5′-GGGTGAAAGTTCAGCGCG-3′ and 5′-AGAACTGCACGAGCCACG-3′ in the presence of 10% DMSO. Cycling conditions were 94°C for 2 min followed by 30 cycles of 93°C for 30 s, 63°C for 30 s, 72°C for 1 min, and a final extension at 72°C for 5 min. The resulting 293-bp fragment was ligated into the SmaI site of Bluescript II SK+, resulting in pB-SOXN5. A second fragment of 4 kb was amplified by the Expand Long Template PCR System (Roche). PCR was carried out in a GeneAmp 2400 PCR system from Perkin-Elmer by 30 cycles of 92°C for 20 s, 58°C for 30 s, 68°C for 5 min, and a final extension at 68°C for 10 min using oligonucleotides 5′-GTGTGGGTGCTGGACAGC-3′ and 5′-AATGTTCACTTTCCTTGCAGG-3′. After digestion with BssSI and XbaI, a 2.1-kb fragment was cloned into BssSI- and XbaI-restricted pB-SOXN5, resulting in the plasmid pB-SOXN-CDS that comprises the entire SOXN coding region.

For subcloning, a 2.4-kb SOXN cDNA fragment was EcoRV/NotI digested from pB-SOXN-CDS, blunt-ended, and ligated to blunted pTKO-CZ or into the HincII site of pSP64poly(A) (Promega), resulting in the expression plasmid pTKO-SOXN-CDS and the vector pSP64-SOXN-CDS.

Cell Growth Assay.

Stably transfected cell populations were seeded in 96-well microtiter plates at an initial number of 5000 cells/well. Protection against cell death was monitored using the colorimetric Sulforhodamine B assay that quantifies cell growth with an accuracy comparable with that of Coulter counting (28). Each treatment group consisted of six replicates. Cells were fixed with trichloroacetic acid (10% final concentration) and stained with 0.4% Sulforhodamine B (Sigma) in 1% acetic acid for 30 min. After washing three times with 1% acetic acid, the bound dye was eluted from the cells with 100 μl of Tris-HCl (pH 10.5), and absorbance was measured at 490 nm using a Micro-Elisa autoreader. The starting cell number was determined by fixing one plate 8 h after seeding and was defined as 1. When expressed as a factor of initially seeded cells, an increased cell number corresponds to values > 1, and a decrease of initial cell number is equivalent to values < 1.

RNA Extraction, Northern Blot Analysis, and RT-PCR.

Total RNA was prepared using Trizol reagent (Life Technologies, Inc.) following the manufacturers instructions. Poly(A)+ RNA was isolated with Oligotex mRNA Kit (Qiagen). Two μg of poly(A)+ RNA were separated through a 1% agarose-formaldehyde gel, blotted onto Hybond N+ filters (Amersham), and hybridized with randomly primed 32P-labeled candidate-1 (SOXN) 1-kb fragment and β-actin probes. Hybridization and washing conditions were used according to the protocol of the manufacturer (Hybond N+; Amersham). Adult and fetal multiple tissue Northern blots were obtained from BD-Clontech. Candidate-1 (SOXN) antisense expression was monitored by reverse transcription of 1 μg of total RNA with 0.1 mm oligo(dT)-Primer (Roche) using the Omniscript RT-PCR system (Qiagen) and subsequent PCR with oligonucleotides TKO2-3′ (5′-AGGTTCCTTCACAAAGATCCC-3′) and SOXN-5′-AAGAATGTGAGCACGTGGC-3′. Cycling parameters were 94°C for 2 min followed by 25 cycles of 94°C for 30 s, 60°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 5 min. Amplification of SOXN 3′-nontranslated region was performed with oligonucleotides SOXN-5′-TCTAAAGCCCCCCCTGTC-3′ and SOXN-5′-CATCTTCAGAGGCAACGGAC-3′ under the same conditions.

Loss of Heterozygosity Analysis.

The D9S158 microsatellite locus was amplified with fluorescence-labeled SOXN-specific primer D9S158a: 5′, TTTGGCTAAAATAGGCTCAG; and D9S159b: 5′, AACATAAAGTAGGCAAAGCG and analyzed by capillary electrophoresis using an ABI 310 (29).

FISH.

Two-color FISH was performed with centromeric chromosome 1 probe D1Z5 and the SOXN-specific PAC RPCI-4 538A22. Probes were labeled by nick translation with biotin-14-dCTP (Life Technologies, Inc.) and digoxigenin-11-dUTP (Roche), prehybridized in the presence of Cot-1 DNA and hybridized to metaphase spreads of normal EBV-transformed lymphocytes (30). For detection of biotin-labeled probes, avidin-FITC and biotinylated anti-avidin (Vector) were used. Sites hybridized to digoxigenin were detected by mouse-anti-digoxigenin (Roche) and Cy3-conjugated antimouse IgG (Dianova). Chromosomes were counterstained with DAPI. Images were taken on a Zeiss Axiophot microscope with a charge-coupled device (CCD) camera (KAF 1400; Photometrics) and processed with IPLab-Spectrum and Adobe PhotoShop software. Ten well spread and well-banded metaphases were analyzed to localize the hybridization signals. Giemsa staining followed routine procedures.

In Vitro Transcription and Translation.

Circular pSP64-SOXN-CDS and pSP64-MYCN plasmid DNA (0.5 μg) were transcribed from the SP6 promoter and translated using the TNT Coupled Reticulocyte Lysate Systems (Promega) in the presence of [35S]methionine (Amersham), following the manufacturer’s instructions. Translation products were separated by 12% SDS-PAGE and visualized by fluorographpic enhancement of the [35S]-signal using Amplify Reagent (Amersham) and Hyperfilm MP (Amersham).

Preparation of SOXN Antibody and Immunoblot Analysis.

Polyclonal SOXN antiserum was raised in rabbits against the keyhole limpet hemocyanin-linked peptide LPLPEKPHKEENS (amino acids 297–310 of SOXN protein) and purified following the method of Chersi et al. (31). Specificity was tested by preincubating the purified antibody (2.4 μg/ml) with the peptide (5 μg/ml) for 1 h on ice in PBS.

Total cell lysates (25), pellet-2 plasma membrane fractions (32), isolated mitochondrial fractions (ApoAlert Cell Fractionation kit; BD-Clontech) or RIPA buffer-insoluble protein fractions (33) were resolved by 10% SDS-page and blotted onto nitrocellulose membranes (0.45 μm; Schleicher and Schuell). Blots were incubated with purified anti-SOXN antibody in a 1:1000 dilution overnight at 4°C, washed four times with PBST, incubated with a goat antirabbit antibody conjugated to horseradish peroxidase (dilution 1:5000; PharMingen) and developed using the BM Chemiluminescence Blotting Substrate (Roche). Molecular weight was determined using the Benchmark Prestained Protein Ladder (Life Technologies, Inc.).

Immunofluorescence.

Nontreated, IFN-γ-stimulated (1000 units/ml in RPMI 1640 containing 1 or 10% FCS) or transiently pTKO-SOXN-CDS-transfected Tet21N-EBNA-1-3 cells were grown on glass slides, fixed for 10 min in 3.7% formaldehyde, and permeabilized for 5 min with 0.05% NP40, 0.1% BSA in PBS. Cells were then incubated for 30 min with primary anti-SOXN antibody (purified polyclonal rabbit anti-SOXN, 1:1000 diluted in PBS containing 0.1% BSA) at room temperature. After washing with PBS, cells were incubated with secondary conjugated antibody [Cy3-labeled goat antirabbit ML (Dianova) in 1:250 dilution] for 1 h at room temperature. Nuclei were stained for 2 min in 50 ng/ml DAPI in 2× SSC. The slides were mounted in 1,4-diazabicyclo(2.2.2)octane (DABCO, Sigma) and examined using the Zeiss confocal laser scanning microscope LSM 510UV or the Zeiss Axiophot microscope using IPLab-Spectrum software.

Structural Analysis.

Multiple alignment was performed using the Malign software and protein structure analysis by ProtSweep Protein Sequence Identification tool, both available at the Husar program package.2

Nucleotide Sequences.

SOXN cDNA was published under European Molecular Biology Laboratory database accession number AJ318051.

Rescue of cDNA Fragments That Confer Resistance of Neuroblastoma Tet21N Cells to Apoptotic Stimuli.

To isolate genes involved in apoptotic processes in neuroblastoma, the functional TKO system was established in cell line Tet21N (11). In this approach, death-associated genes are identified by their inactivation through overexpressed antisense sequences. For the selection, a Tet21N-derived subtractive library enriched for apoptotic genes (24) was subcloned into the pTKO-CZ expression vector. The resulting TKO library was nondirected, and thus, ∼50% of inserts were in antisense orientation. To test the redundancy of this library, 50 randomly picked clones were sequenced with a TKO-specific oligonucleotide. In a total of 32 different inserts, the IFN-γ targets IRF-1 (8 of 50), RALY (5 of 50), and VARS (3 of 50) were mostly abundant. The library was transfected into the stably EBNA-1-expressing cell clone Tet21N-EBNA-1-3 that is highly permissive for transfection with the EBV-derived episomal pTKO-CZ plasmid (11). Two days after transfection, cells were placed under the selective pressure of low serum/MYCN/IFN-γ and Zeocin. Episomes were extracted from surviving clones after 24 days and transformed into DH10B E. coli. Plasmid DNA from 107 colonies was isolated and sequenced, resulting in a total of 13 individual episomes after the first round of selection. Each single episome was transfected into Tet21N-EBNA-1-3 cells and examined for protection against low serum/MYCN/IFN-γ-induced cell death in a second round of selection. Five clones conferred resistance to the death stimulus. Although candidates 1 and 2 were highly represented in the initial 107 sequenced clones (candidate-1: 15/107; candidate 2: 80/107), only single recombinants of candidates 3–5 were rescued. Approximately 0.1% of Tet21N cells transfected with the antisense fragment of candidate-1 survived long-term IFN-γ treatment. Because of this protective effect and its high representation in the TKO selection, candidate-1, which corresponds to an as yet uncharacterized EST, was chosen for additional characterization.

Antisense Expression of Candidate-1 Protects Neuroblastoma Tet21N Cells from Cell Death.

The cell death protective effect of candidate-1 in transiently transfected cells was additionally examined by comparing the residual cell viability of stably candidate-1 and control plasmid-transfected neuroblastoma cells after IFN-γ treatment. Polyclonal cell populations were generated from candidate-1 vector (pTKO including the 130-bp candidate-1 antisense fragment) or pTKO-βGal control plasmid-transfected Tet21N-EBNA-1-3 cells and maintained in the presence of Zeocin. Expression of the 130-bp antisense fragment was examined by RT-PCR with an insert-specific and a plasmid-specific primer flanking the BglII site of pTKO-CZ (Fig. 1,A). In parallel, endogenous expression of the candidate-1 corresponding mRNA was determined with oligonucleotides specific for the 3′-noncoding region of the gene. Both populations showed endogenous gene expression, the exogenous antisense fragment was exclusively amplified from candidate-1-transfected cells. In colony formation assays, the antisense-expressing cell population was protected from low serum/MYCN/IFN-γ-induced cell death while only small proportions of the cells transfected with the control plasmid survived long-time IFN-γ treatment (Fig. 1,B). To determine the effect on a quantitative basis, the number of surviving cells was measured by the colorimetric SRB assay (Fig. 1 C). In the absence of IFN-γ, the candidate-1 and the control population behaved similarly, suggesting that the antisense expression had no effects on the normal growth of the cells. In contrast, IFN-γ-treated cells displayed a clear difference in viability from day 3 after IFN-γ treatment and later on. Although the control population showed massive cell death resulting in 18% viability on day 13, the antisense-expressing cells merely stopped to proliferate but were resistant against IFN-γ-induced cell death (cell number on day 13 according to 150% of initially seeded cells). These results show that the expression of candidate-1 antisense RNA protects Tet21N neuroblastoma cells from cell death but not from the growth suppressive action of IFN-γ.

To further analyze if candidate expression was not able to inhibit the IFN-γ-induced suppression of cell growth, we tested the IFN-γ response of both antisense RNA expressing and control populations in medium containing 10% FCS. As cell death in the Tet21N cells can only be induced under low serum conditions (12), high serum was expected to protect both populations from cell death. In fact, 3 days after exposure to IFN-γ, both populations stopped to proliferate but remained viable (Fig. 1 D). Consistent with the observations under 1% serum, the candidate-1 antisense fragment was not capable of inhibiting the IFN-γ-induced growth suppression. This suggests that other candidate-1-independent pathways participate in the IFN-γ-induced proliferation arrest.

Candidate-1 Corresponding mRNA Is Expressed in a Variety of Tissues and Induced in Apoptotic Neuroblastoma Cells.

As the candidate-1 antisense fragment was isolated from a subtractive library that is enriched for proapoptotic genes, we examined the influence of the death-promoting low serum/MYCN/IFN-γ environment on endogenous candidate-1 gene expression in Tet21N cells. MYCN-expressing cells were treated with IFN-γ (1000 units/ml) in medium containing 1% FCS. Poly(A)+ RNA was extracted at various time points after induction, blotted and probed with 32P-labeled candidate-1 cDNA. Three mRNA transcripts of 3.3, 4.6, and 7.5 kb were detected in all lanes (Fig. 2 A). The expression level of the two larger transcripts was increased 9 h after apoptosis induction; this level was maintained until at least 48 h (data not shown).

The distribution of candidate-1 transcripts was examined in normal adult and fetal human tissues by Northern blotting (Fig. 2 B). Although transcripts were expressed in all tested tissues, a tissue-specific variation was observed. Pancreas, brain, placenta, kidney, heart, and all tested fetal tissues expressed higher steady-state levels of candidate-1 corresponding mRNAs relative to lung, liver, and skeletal muscle. The candidate-1 corresponding UniGene clusters Hs.30853 and Hs.3091656 report ESTs also from various other tissues, including colon, skin, breast, uterus, ovary, testis, prostate, and B cells. Virtual Northern blotting7 identified transcripts in normal as well as tumor tissue of brain, breast, colon, and other tissues (data not shown). These data indicate that the candidate-1 mRNA is expressed in a variety of tissues.

Candidate-1 Is Homologous to the SOX/Q6 Gene Family.

BlastN2 search with the protective candidate-1 fragment identified the 155-kb BAC-clone RP11-83N9 (accession no. AL138781) that was sequenced by the Sanger Centre (Hinxton, Cambridgeshire, United Kingdom). Additionally, a genomic sequence (Ga x54KRCD6M81) embedding the entire first exon was found on the Celera publication site. Using the gene prediction program Genscan and the EST cluster software (both available at the Husar software package),8 putative exons were identified and specific primers were designed. Long-range PCR was performed on reverse-transcribed Tet21N cDNA, and from the different PCR products, a cDNA sequence of 4555 bp was assembled which corresponds to the 4.6-kb band seen in the Northern Blot (Fig. 2 A). Sequencing and subsequent BlastN2 search revealed homology of the sequence to members of the SOX/Q6 gene family. On the basis of this similarity, the gene was called SOXN (European Molecular Biology Laboratory database accession no. AJ318051). Q6 has recently been renamed QSOX1 and SOXN, whose existence as the putative human Q6 paralog was deduced from database sequences, was termed QSOX2(34). In this work, the genes are addressed as Q6 and SOXN.

The third candidate-1/SOXN transcript of 7.5 kb could not be amplified. Whether the transcript results from an additional 5′-untranslated exon or includes other yet unidentified coding exons remains to be elucidated.

SOXN Sequence Analysis.

The SOXN 4.6-kb cDNA includes a putative in frame start codon (aacATGg, bold and underlined in Fig. 3) at bp 39 that matches in a strong context the Kozak translation initiation consensus sequence (annATGg; Ref. 35). This putative translation start is embedded in a highly GC-rich region, displaying 77% GC content in the 600 bp surrounding the ATG codon. The open reading frame ends at bp 2135 with a TGA stop codon. Thus, SOXN encodes a predicted protein of 698 amino acids.

To identify potential transcriptional start sites, the 5′-genomic sequence of SOXN was examined using a neural network promoter prediction program9 and the Transcription Element Search System (TESS) program,10 that predicts potential transcription factor binding sites. Five strong promoters were identified ranging from bp −730 to −65 of the predicted start codon, none of which showed a TATA box. Interestingly, TESS identified an E-box consensus sequence (CACGTG) at position −55 of the ATG start codon showing 88.3% homology to the MYCN protein binding site and 88.0% homology to the USF transcription factor binding site (bold and underlined in Fig. 3). Additionally, several Sp1 binding sites were predicted in the GC-rich region from bp −242 to −70 and +138. The question whether SOXN transcription can be modulated by MYCN awaits, as well as the experimental determination of the SOXN transcription start, additional studies.

A search for polyadenylation signals identified two sequences at nucleotide positions 3376 and 4519 (bold in Fig. 3). BlastN2 identified several oligo(dT)-primed ESTs in both regions corresponding to the UniGene clusters Hs.309165 (bp 3376, 18 ESTs) and Hs.30853 (bp 4519, 33 ESTs).6 Whether the 3.3- and the 4.6-kb transcripts seen in Northern blots account to an alternate usage of polyadenylation signals or to alternative splicing remains to be elucidated.

Finally, a dinucleotide (GT) repeat was identified in the 3′-untranslated region of the SOXN cDNA (underlined in Fig. 3). This CA-repeat had been characterized in the Généthon database (ID no. AFM073YB11) and was reported to be polymorphic (D9S158 marker, showing a maximal heterozygosity of 70%).

SOXN Is Organized in 12 Exons and Maps to 9q34.3.

Homology search of SOXN full-length cDNA on the public database and on the Celera publication site identified 12 exons spread over a genomic region of ∼32 kb. The intron and exon sizes and the intron/exon-junctional sequences are listed in Table 1. All splice sites, except for intron 7, are closely related to the consensus splicing sites (5′mGT-AGm3′). Intron 7 has a 5′mGC-AGm3′-splicing signal, a structure that has been identified in other genes (36). Remarkably, the human SOXN homologue Q6 is also organized in 12 exons of comparable size (Table 1, Q6 splice sites not shown), underlining the relationship of the two genes.

The SOXN chromosomal localization was determined on metaphase spreads of human lymphoblastoid cells by FISH using a SOXN PAC as a probe. The combined analysis of fluorescent signals and Giemsa staining identified 9q34.3 as the genomic site (Fig. 4). This location is confirmed by the previously reported mapping of the polymorphic SOXN-specific D9S158 marker to the very distal 163 cM position of the integrated Généthon map on chromosome 9q (37).

SOXN Shares Homology with Members of the SOX/Q6 Family.

The predicted SOXN protein consists of 698 amino acids with a calculated mass of 77.33 kDa. BlastP2 search of the deduced SOXN protein revealed homology to several proteins of the SOX/Q6 family. Besides the known human Q6 protein and its related Gec-3 (guinea pig) and SOx (rat and mouse) homologues, an additional putative mouse protein of 692 amino acids (SOXNmms) could be aligned to SOXN in multiple sequence alignment (Fig. 5,A). This SOXN-related mouse sequence corresponds to a 3.3-kb cDNA that was assembled from several ESTs (using the CAP program, Husar software package, data not shown). Although the cDNA shows 80% homology to the human SOXN gene, its predicted protein SOXNmms shows 76.6% overall homology to SOXN (Fig. 5 A), making it likely that the murine SOXNmms encodes a SOXN homologue.

SOXN displays a number of features that are characteristic for the SOX/Q6 family (Fig. 5,B). Thus, it contains a putative signal peptide (amino acids 1–50), a PDI-like TRX domain (amino acids 76–147, active site: WCGHC), an ERV1 domain (amino acids 408–540, catalytic domain: CXXC), and a conserved N-glycosylation motif (amino acid 266). Amino acid sequence alignment confirmed a high identity of ∼60% in these functional domains and in so-called Q6-like regions (20), whereas other regions were less conserved among the SOXN and Q6 family members (Fig. 5 A). Hydrophobicity analysis identified a COOH-terminal extension in the deduced SOXN and the previously updated Q6 sequence (accession no. NM_002826) that was predicted to span the membrane (SOXN: amino acids 661–683, Q6: amino acids 707–729, as determined by the TMHMM program of the Center for Biological Sequence Analysis at the Technical University of Denmark server, data not shown).11

In vitro translation of SOXN cDNA in reticulocyte lysates produced a protein of ∼78 kDa (Fig. 6 A). A second protein of ∼60 kDa translated at lower efficiency may be the result of alternate translation initiation at codon 123.

SOXN Is a Membrane-Associated Protein.

To examine the localization of the SOXN protein, a polyclonal antibody was raised against a synthetic peptide comprising residues 297–310 of SOXN as antigen (Fig. 5,A). In immunoblots, the purified antibody recognized the in vitro-translated SOXN (data not shown), as well as a recombinant protein of the same size that was produced in transiently pTKO-SOXN-CDS-transfected Tet21N cells (Fig. 6,B). In nontransfected cells, an endogenous SOXN protein was exclusively detected in the RIPA buffer-insoluble fraction and in the pellet-2 membrane fraction but not in total cell lysates, the mitochondrial fraction, or the RIPA buffer-soluble fractions of Tet21N cells (Fig. 6, B and C, data for RIPA buffer-soluble fractions not shown). The enlarged size of the endogenous SOXN in the pellet-2 fractions might be attributable to glycosylation.

After apoptosis induction, a down-regulation of SOXN in pellet-2 membrane fractions was observed while the protein was enriched in the RIPA buffer-insoluble fraction of IFN-γ-induced Tet21N cells (Fig. 6 C). Thus, SOXN seems to be translocated from plasma membranes to other, thus far uncharacterized, cellular compartments after apoptosis induction.

The membrane-associated localization of SOXN was confirmed by indirect immunofluorescence of the neuroblastoma cell lines Tet21N, Kelly, GI-ME-N, CHP-134, and NGP. All nonpermeabilized cells exhibited a punctuated SOXN-specific staining of the plasma membrane (data only shown for Tet21N cells, Fig. 7). Some cells displayed also a weak staining of the perinuclear region (Fig. 7,A). The binding was, as in immunoblots, completely blocked by preincubating the antibody with competing peptide (data not shown). Serum withdrawal and subsequent stimulation of proliferation by 10% FCS resulted in strong coloring of nonpermeabilized proliferating cells (Fig. 7,B). When apoptosis was induced by combined low serum/MYCN and IFN-γ treatment, a strong staining of apoptotic bodies was observed (Fig. 7,D). Some cells transiently transfected with pTKO-SOXN-CDS displayed an augmented staining of the outer plasma membrane (Fig. 7,E). Permeabilized, nontreated cells showed a weak background staining of nuclei (Fig. 7,A), whereas IFN-γ-treated cells displayed a compact juxtanuclear (Fig. 7,C) and punctuated cytoplasmic staining (Fig. 7,D). This pattern was even more intensive in transiently pTKO-SOXN-CDS-transfected cells (Fig. 7 E).

Collectively, these results indicate that SOXN is predominantly targeted to the nuclear and outer plasma membrane. Induction of apoptosis resulted in an augmented intracellular SOXN expression and strong staining of apoptotic bodies, suggesting that the protein can also be located to divergent, thus far, uncharacterized compartments.

Overexpression of SOXN Sensitizes Tet21N-EBNA-1-3 Cells to Induced Apoptosis.

To test if overexpression of SOXN causes cell death or sensitizes cells to IFN-γ-induced apoptosis, Tet21N-EBNA-1-3 cells were transfected with pTKO-SOXN-CDS or pTKO control plasmid. Selection for the plasmids was started 48 h after transfection and was performed in the presence of the MYCN protein and 250 μg/ml Zeocin. One group of cells was initially treated with apoptosis inducing low serum/IFN-γ stimulus for 3 days and subsequently cultivated for 7 days in 10% serum without IFN-γ treatment. The second group was selected for 10 days in the presence of 10% FCS without IFN-γ treatment. The number of colonies was scored on day 10 of selection. Ectopic expression of SOXN resulted in a significant reduction in the number of Zeocin-resistant colonies when the cells were selected after the initial apoptotic stimulus (+ pTKO-SOXN-CDS +IFN: 28 colonies mean; + pTKO-NK +IFN: 158 colonies mean, Fig. 8). Cells selected without initial IFN-γ treatment did not display significant differences (+ pTKO-SOXN-CDS-IFN: 390 colonies mean; + pTKO-NK-IFN: 410 colonies mean). This suggests that increased SOXN expression alone is not sufficient to induce significant cell death but that augmented SOXN expression can sensitize neuroblastoma cells to IFN-γ-induced cell death.

The identification of genes that sensitize neuroblastoma cells to apoptotic stimuli has important implications for the development of novel therapeutic strategies. To isolate such genes, we have established a functional antisense approach of gene cloning in a neuroblastoma in vitro cell system. This strategy of cloning death-promoting genes through their functional inactivation by antisense RNA has already led to the identification of several novel genes that function as positive mediators of cell death (38, 39).

The transfection and long-time selection of a subtractive library that was enriched for proapoptotic transcripts has resulted in the isolation of five different candidate genes obviously participating in low serum/MYCN/IFN-γ-induced cell death. Among these, the 130-bp candidate-1 antisense fragment of the novel SOXN gene was highly enriched. As candidate-1 was not overrepresented in the initially transfected library, the high frequency of appearance identifies candidate-1 as an efficient mediator of resistance against induced cell death. Indeed, stable transfection and expression of candidate-1 protects Tet21N neuroblastoma cells efficiently from induced apoptosis in colony formation and SRB assays compared with controls. The fact that candidate-1 expression could not protect cells from IFN-γ-induced proliferation arrest suggests that the cytostatic effect is mediated through other IFN-γ-induced genes.

Cloning and sequencing of the full-length cDNA established SOXN as a new member of the SOX/Q6 family. The human Q6 gene has evolved by the fusion of two ancient genes, TRX and ERV1(20), and was reported to be up-regulated in quiescent fibroblasts. The function of Q6 as a SOX is predicted on the basis of its high homology to both the chicken egg white SOX and the rat seminal vesicle SOX. For both enzymes, flavin-dependent catalytic activity of a redox-active disulfide bridge in the ERV1 domain toward reduced peptides and proteins has been demonstrated (15, 21). Recently, SOX activity was also shown for the yeast ERV1(40), yeast Erv2p(41), the human ARV1 gene (42), and the viral E10R ERV-1 family member (43).

The overall homology of the SOXN and Q6 proteins is ∼40%, the identity of functional regions such as the TRX-like or the ERV1 domain is as high as 68%. Besides the sequence homology of SOXN and Q6, the relationship between the two genes is supported by their common structural features as both genes contain 12 exons of identical lengths and show a high GC content in the first exon. The differences in the SOXN and Q6 expression pattern with two transcripts of Q6 (3.2 and 3.8 kb) in placenta, liver, lung, and heart, but very weak expression levels in pancreas, brain, kidney, and skeletal muscle (44) indicate that the two genes have different functions or take on redundant functions in different tissues. However, the ubiquitous expression of SOXN and its high steady-state level in some tissues suggests that SOXN function as a positive mediator of cell death is not restricted to neuroblastoma and that the protein is likely to have additional functions not directly related to apoptosis.

A putative relationship of SOXN to neuroblastoma had emerged from the observation that the loss of heterozygosity of chromosome 9-related markers in neuroblastoma patients included the SOXN-specific D9S158 marker. Of 39 informative neuroblastoma samples, the SOXN marker was found to be deleted in 8 cases (45). However, our own analysis of 48 informative tumor DNAs from neuroblastoma patients identified only 3 with an allelic deletion of D9S158 (data not shown). Thus, allelic deletion of the SOXN locus is unlikely to contribute to the inactivation of SOXN in a large proportion of neuroblastomas. Still, SOXN could be inactivated by other mechanisms such as promoter methylation.

The observed relative molecular mass of in vitro-translated SOXN is consistent with the predicted mass of the protein and supports the functionality of the cloned cDNA. The higher apparent molecular mass of endogenous SOXN probably reflects glycosylation of the membrane-associated protein. The localization of endogenous SOXN in dot-like patterns on the outer plasma membrane and, in lower concentration, in the nuclear membrane, additionally support that SOXN is restricted to membrane fractions. The augmented intracellular SOXN staining and the strong staining of the membranes of apoptotic bodies after IFN-γ stimulation is consistent with the predicted role of SOXN in programmed cell death. However, at this time, the exact biological function of SOXN remains unclear. Does the protein act as a sensor for the cellular redox state or can its function be modulated by exogenous stimuli? A participation in different processes as proliferation and apoptosis has been demonstrated for several proteins. A prominent example is the MYC protein that can act, depending on the physiological status of the cell, as growth promoter or as a mediator of cell death (46). In a similar way, the redox-active TRX and TRX-reductase genes may have dual activities. Both genes were first described to be growth stimulatory (47, 48, 49) until their functional inactivation by TKO elucidated that the encoded proteins can also participate in IFN-induced cell death (14, 39, 50). Recently, TRX-reductase and TRX were reported to induce the expression of p53 target genes, among these several oxidoreductases (51), after IFN and retinoic acid stimulation (52). SOXN, located at the nuclear membrane, could also participate in the redox regulation of transcription factors as it has been reported for p53 (53), nuclear factor-κB, and AP1 (54) signaling and which seems to be a general regulatory mechanism in signal transduction (55).

The functional importance of the TRX domain has not been established for any SOX/Q6 family member. However, yeast Erv2p and its viral homologue E10R, both displaying SOX activity by their ERV1 domains, were shown to transfer oxidizing equivalents directly to the yeast PDI homologue Pdi1p or the virus-encoded glutaredoxin G4L (43, 56). This cooperation of ERV1-like and TRX-like domains of nonlinked molecules makes it reasonable to assume a similar mechanism of internal disulfide transfer in SOX/Q6 proteins. In this model, TRX could serve as a redox sensor judging about SOX activity in mediating downstream apoptotic processes. An example for TRX redox state-dependent apoptosis is tumor necrosis factor α-induced, ASK1-mediated cell death (57). Under reducing conditions, ASK1 interacts with TRX and is inactivated. ROS-induced oxidation of TRX results in its dissociation from the complex, subsequent ASK1 kinase activation, and execution of apoptosis. If the function of the SOX/Q6 family members was regulated by a similar mechanism, conformational changes and activity of SOX proteins could be easily driven by the cellular redox state.

An alternative possibility for SOXN function is based on the observation that cellular TRX can be translocated to the extracellular matrix or nucleus upon oxidative stress (58). Such translocational processes are well established in cells to ensure a rapid execution of apoptosis. Intracellularly sequestrated death receptors can, for example, be redistributed from the Golgi apparatus to the cell surface after p53 induction (59). The augmented juxtanuclear and cytoplasmic SOXN level in IFN-γ-treated Tet21N neuroblastoma cells as well as its down-regulation in the pellet-2 plasma membrane fraction and the subsequent enrichment in the RIPA buffer-insoluble fraction may reflect such a signal-dependent translocation.

Overexpression of SOXN is not sufficient to trigger cell death but seems to sensitize neuroblastoma cells to apoptotic stimuli. This indicates that SOXN acts in concert with additional factors to execute programmed cell death. The goal of additional studies will be to precisely define the function of SOXN during neuroblast development and execution of apoptosis.

Grant support: Cooperation Program in Cancer Research of the Deutsches Krebsforschungszentrum and Israel’s Ministry of Science, by the Deutsche Krebshilfe, and by the Ministry of Research and Technology.

I. W. and R. W. contributed equally to this work.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Manfred Schwab, Department of Tumor Genetics (B030), Deutsches Krebsforschungszentrum, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. Phone: 49-6221-423220; Fax: 49-6221-423281; E-mail: [email protected]

1

The abbreviations used are: TKO, technical knockout; EST, expressed sequence tag; TRX, thioredoxin; SOX/Q6, sulfhydryl oxidase/Quiescin6; SOXN, neuroblastoma-derived sulfhydryl oxidase; PDI, protein-disulfide isomerase; RT-PCR, reverse transcription-PCR; oligo(dT), oligodeoxythymidylic acid; FISH, fluorescence in situ hybridization; DAPI, 4′,6-diamidino-2-phenylindole; RIPA, radioimmunoprecipitation assay; FAD, flavin-adenine dinucleotide; SRB, sulforhodamine B.

2

Internet address: http://husar.dkfz-heidelberg.de/.

3

Internet address: https://www.rzpd.de/cgi-bin/db/entrySequence-new.shtml.

4

Internet address: https://www.rzpd.de/cgi-bin/db/entrySequence-new.shtml.

5

Internet address: http://publication.celera.com.

6

Internet address: http://www.ncbi.nlm.nih.gov/UniGene/.

7

Internet address: http://www.ncbi.nlm.nih.gov/SAGE/sagevn.cgi.

8

Internet address: http://genome.dkfz-heidelberg.de/.

9

Internet address: http://www.fruitfly.org/seq_tools/promoter.html.

10

Internet address: http://www.cbil.upenn.edu/tess/.

11

Internet address: http://www.cbs.dtu.dk/services/TMHMM/.

Fig. 1.

Stably candidate-1-transfected cells are protected from IFN-γ-induced cell death. A, only candidate-1 (Tet21N-candidate-1) but not pTKO-βGal control plasmid transfected (Tet21N-NK) Tet21N cells express candidate-1 antisense fragment (candidate-1). Total RNA of stably transfected cells was submitted to RT-PCR amplification with candidate-1-specific forward and plasmid-specific reverse primer (candidate-1), forward and reverse primers specific for the cDNA corresponding to candidate-1 (endogenous candidate-1), and G3PDH-specific primers as control gene (G3PDH). Candidate-1 plasmid (pTKO-candidate-1) served as positive control for antisense expression, a long-range PCR (LR-PCR) product of 4 kb corresponding to candidate-1 (candidate-1 LR-PCR) as positive control for endogenous candidate-1 expression. The arrow indicates the candidate-1 fragment, the additional band of 300 bp corresponds to a nonspecific amplification product. B, colony formation assay. Stable, polyclonal populations were grown in the presence of 1% FCS, MYCN, 250 μg/ml Zeocin, and 1000 units/ml IFN-γ. After 24 days of selection, cells were expanded for additional 7 days in medium containing 10% FCS and Zeocin. The cells were fixed with methanol and stained with Giemsa. Candidate-1, candidate-1-expressing cells; control, pTKO-βGal-transfected cells. Comparable results were obtained in four independent experiments. C and D, SRB assay. Polyclonal candidate-1-expressing (candidate-1, ————) or pTKO-βGal-transfected cells (control, — — — —) were seeded in 96-well microtiter plates at an initial number of 5000 cells/well. Cells were cultivated in the presence of MYCN in medium containing 1% (C)or 10% (D) FCS and treated with 1000 units/ml IFN-γ (triangles) or left without IFN-γ (quarters). At indicated time points, cells were fixed, stained with SRB, and bound dye was measured at 490 nm. Each point represents an average of six replicates (SD within 15% of the mean), expressed as a factor of initially seeded cells. Values < 1 indicate death of initially seeded cells, values > 1 net cell growth. Absorbance values for initial seeded (d0) and on day 13 fixed (d13) cells are: candidate-1(d0): 0.106; control(d0): 0.120; candidate-1-IFN(d13): 1%FCS: 0.443, 10%FCS: 1.405; control-IFN(d13): 1%FCS: 0.382, 10%FCS: 1.412; candidate-1+IFN(d13): 1%FCS: 0.159, 10%FCS: 0.530; control + IFN(d13): 1%FCS: 0.021, 10%FCS: 0.486. Comparable results were obtained in two independent experiments.

Fig. 1.

Stably candidate-1-transfected cells are protected from IFN-γ-induced cell death. A, only candidate-1 (Tet21N-candidate-1) but not pTKO-βGal control plasmid transfected (Tet21N-NK) Tet21N cells express candidate-1 antisense fragment (candidate-1). Total RNA of stably transfected cells was submitted to RT-PCR amplification with candidate-1-specific forward and plasmid-specific reverse primer (candidate-1), forward and reverse primers specific for the cDNA corresponding to candidate-1 (endogenous candidate-1), and G3PDH-specific primers as control gene (G3PDH). Candidate-1 plasmid (pTKO-candidate-1) served as positive control for antisense expression, a long-range PCR (LR-PCR) product of 4 kb corresponding to candidate-1 (candidate-1 LR-PCR) as positive control for endogenous candidate-1 expression. The arrow indicates the candidate-1 fragment, the additional band of 300 bp corresponds to a nonspecific amplification product. B, colony formation assay. Stable, polyclonal populations were grown in the presence of 1% FCS, MYCN, 250 μg/ml Zeocin, and 1000 units/ml IFN-γ. After 24 days of selection, cells were expanded for additional 7 days in medium containing 10% FCS and Zeocin. The cells were fixed with methanol and stained with Giemsa. Candidate-1, candidate-1-expressing cells; control, pTKO-βGal-transfected cells. Comparable results were obtained in four independent experiments. C and D, SRB assay. Polyclonal candidate-1-expressing (candidate-1, ————) or pTKO-βGal-transfected cells (control, — — — —) were seeded in 96-well microtiter plates at an initial number of 5000 cells/well. Cells were cultivated in the presence of MYCN in medium containing 1% (C)or 10% (D) FCS and treated with 1000 units/ml IFN-γ (triangles) or left without IFN-γ (quarters). At indicated time points, cells were fixed, stained with SRB, and bound dye was measured at 490 nm. Each point represents an average of six replicates (SD within 15% of the mean), expressed as a factor of initially seeded cells. Values < 1 indicate death of initially seeded cells, values > 1 net cell growth. Absorbance values for initial seeded (d0) and on day 13 fixed (d13) cells are: candidate-1(d0): 0.106; control(d0): 0.120; candidate-1-IFN(d13): 1%FCS: 0.443, 10%FCS: 1.405; control-IFN(d13): 1%FCS: 0.382, 10%FCS: 1.412; candidate-1+IFN(d13): 1%FCS: 0.159, 10%FCS: 0.530; control + IFN(d13): 1%FCS: 0.021, 10%FCS: 0.486. Comparable results were obtained in two independent experiments.

Close modal
Fig. 2.

Candidate-1 expression. A, expression of a cDNA corresponding to candidate-1 is induced in apoptotic Tet21N neuroblastoma cells. Poly(A)+ RNA was extracted from noninduced cells (0 h), 9, or 24 h after low serum/MYCN/IFN-γ stimulation (24) of cells. Northern blots were hybridized with a 1-kb candidate-1-specific probe and a β-actin probe to verify equivalent RNA loading. B, tissues expressing candidate-1 mRNA. Multiple tissue northern blots (BD-Clontech) were hybridized with 1-kb candidate-1-specific probe and β-actin probe as loading control.

Fig. 2.

Candidate-1 expression. A, expression of a cDNA corresponding to candidate-1 is induced in apoptotic Tet21N neuroblastoma cells. Poly(A)+ RNA was extracted from noninduced cells (0 h), 9, or 24 h after low serum/MYCN/IFN-γ stimulation (24) of cells. Northern blots were hybridized with a 1-kb candidate-1-specific probe and a β-actin probe to verify equivalent RNA loading. B, tissues expressing candidate-1 mRNA. Multiple tissue northern blots (BD-Clontech) were hybridized with 1-kb candidate-1-specific probe and β-actin probe as loading control.

Close modal
Fig. 3.

Nucleotide and deduced amino acid sequence of the candidate-1-corresponding gene SOXN. Numbering: −250 to −1: GC-rich genomic sequence, including potential transcription start. One to 4555: SOXN cDNA according to accession no. AJ318051. Bold printed and underlined: E-box element (bp −17); ATG initiation site (bp 39); candidate-1 (SOXN) antisense fragment (bp 1604–1731); TGA stop-codon (bp 2133); predicted polyadenylation signals (bp 3376 and 4519); and GT repeat (bp 3486–3522). Boxed: TRX-motiv (WCGHC), ERV1-motiv (CKEC).

Fig. 3.

Nucleotide and deduced amino acid sequence of the candidate-1-corresponding gene SOXN. Numbering: −250 to −1: GC-rich genomic sequence, including potential transcription start. One to 4555: SOXN cDNA according to accession no. AJ318051. Bold printed and underlined: E-box element (bp −17); ATG initiation site (bp 39); candidate-1 (SOXN) antisense fragment (bp 1604–1731); TGA stop-codon (bp 2133); predicted polyadenylation signals (bp 3376 and 4519); and GT repeat (bp 3486–3522). Boxed: TRX-motiv (WCGHC), ERV1-motiv (CKEC).

Close modal
Fig. 4.

FISH mapping of SOXN to 9q34.3. Normal human lymphocyte metaphases were Giemsa stained and subsequently hybridized with SOXN-specific probe (Cy3/red) and chromosome 1-specific D1Z5 probe (FITC/green).

Fig. 4.

FISH mapping of SOXN to 9q34.3. Normal human lymphocyte metaphases were Giemsa stained and subsequently hybridized with SOXN-specific probe (Cy3/red) and chromosome 1-specific D1Z5 probe (FITC/green).

Close modal
Fig. 5.

SOXN contains conserved domains of the SOX/Q6 family. A, the amino acid sequences of SOXN (accession no. AJ318051), SOxrat (accession no. AF285078), Gec3 (accession no. U82982), and Q6 (accession no. NM_002826) were aligned with Malign. Dark shaded: identical residues; light shaded: similar residues. The predicted cleavage site of the SOXN signal peptide is indicated by a vertical arrow. The conserved TRX and ERV1 disulfides are marked by asterisks, the conserved potential N-glycosylation site by a cross. A synthetic peptide of SOXN amino acids 297–310 (black bracket) was used as antigen for raising an anti-SOXN specific antiserum. B, schematic diagram of the homologous regions of different SOX/Q6 family members. Signal, predicted signal peptide; TRX, TRX-domain; Q6-like, Quiescin6-like regions; N-glyc, conserved N-glycosylation site; ERV1, ERV1-domain; TM, predicted transmembrane segment. The percentage of identical and similar residues (percentage similarity in brackets) to SOXN are shown.

Fig. 5.

SOXN contains conserved domains of the SOX/Q6 family. A, the amino acid sequences of SOXN (accession no. AJ318051), SOxrat (accession no. AF285078), Gec3 (accession no. U82982), and Q6 (accession no. NM_002826) were aligned with Malign. Dark shaded: identical residues; light shaded: similar residues. The predicted cleavage site of the SOXN signal peptide is indicated by a vertical arrow. The conserved TRX and ERV1 disulfides are marked by asterisks, the conserved potential N-glycosylation site by a cross. A synthetic peptide of SOXN amino acids 297–310 (black bracket) was used as antigen for raising an anti-SOXN specific antiserum. B, schematic diagram of the homologous regions of different SOX/Q6 family members. Signal, predicted signal peptide; TRX, TRX-domain; Q6-like, Quiescin6-like regions; N-glyc, conserved N-glycosylation site; ERV1, ERV1-domain; TM, predicted transmembrane segment. The percentage of identical and similar residues (percentage similarity in brackets) to SOXN are shown.

Close modal
Fig. 6.

Expression of SOXN protein. A, in vitro transcription and translation assay of SOXN-cDNA. MYCN-cDNA served as a positive control. The in vitro translated proteins were separated by SDS-PAGE and visualized by enhanced fluorography. Arrows indicate SOXN translation products of 78 and 60 kDa. B, immunoblot analysis of recombinant and cellular SOXN protein. Tet21N-EBNA1-3 cells were transiently transfected with the pTKO-SOXN-CDS expression vector or empty control plasmid. Protein extracts were fractionated by SDS-PAGE, immunoblotted, and reacted with purified polyclonal anti-SOXN-antibody. Exogenous SOXN was detected in total cell lysates (100 μg/lane) of transiently pTKO-SOXN-CDS-transfected (pTKO-SOXN-CDS) but not in control plasmid (pTKO-NK)-transfected cells. Endogenous SOXN was only detected in purified membrane fractions of Tet21N cells but not in total cell lysates or mitochondrial fractions of cells. C, immunoblot analysis of endogenous SOXN after apoptosis induction. Tet21N cells were cultivated in 10% FCS (1) or stimulated with 500 units of IFN-γ in medium containing 0.5% FCS for 24 h (2) or 48 h (4) or were cultivated for 48 h in low serum without IFN-γ treatment (3). Fifty μg of RIPA buffer-insoluble fraction protein or 15 μg of pellet-2 membrane fraction protein were fractionated by SDS-PAGE and blotted. Arrows indicate specific bands recognized by the anti-SOXN antibody. Equal loading was confirmed by incubating the membranes with second antibody (data not shown).

Fig. 6.

Expression of SOXN protein. A, in vitro transcription and translation assay of SOXN-cDNA. MYCN-cDNA served as a positive control. The in vitro translated proteins were separated by SDS-PAGE and visualized by enhanced fluorography. Arrows indicate SOXN translation products of 78 and 60 kDa. B, immunoblot analysis of recombinant and cellular SOXN protein. Tet21N-EBNA1-3 cells were transiently transfected with the pTKO-SOXN-CDS expression vector or empty control plasmid. Protein extracts were fractionated by SDS-PAGE, immunoblotted, and reacted with purified polyclonal anti-SOXN-antibody. Exogenous SOXN was detected in total cell lysates (100 μg/lane) of transiently pTKO-SOXN-CDS-transfected (pTKO-SOXN-CDS) but not in control plasmid (pTKO-NK)-transfected cells. Endogenous SOXN was only detected in purified membrane fractions of Tet21N cells but not in total cell lysates or mitochondrial fractions of cells. C, immunoblot analysis of endogenous SOXN after apoptosis induction. Tet21N cells were cultivated in 10% FCS (1) or stimulated with 500 units of IFN-γ in medium containing 0.5% FCS for 24 h (2) or 48 h (4) or were cultivated for 48 h in low serum without IFN-γ treatment (3). Fifty μg of RIPA buffer-insoluble fraction protein or 15 μg of pellet-2 membrane fraction protein were fractionated by SDS-PAGE and blotted. Arrows indicate specific bands recognized by the anti-SOXN antibody. Equal loading was confirmed by incubating the membranes with second antibody (data not shown).

Close modal
Fig. 7.

Immunofluorescent detection of endogenous and exogenous SOXN in Tet21N neuroblastoma cells. Nonstimulated (10% FCS, -IFN) or low serum/IFN-γ-stimulated (1% FCS and 1000 units/ml IFN-γ) cells were incubated with SOXN-specific antibody, and SOXN-specific signals were detected with Cy3-coupled antirabbit antibody. Nuclei were counterstained with DAPI. Note the augmented extracellular (nonpermeabilized) and intracellular (permeabilized) staining in IFN-γ-treated (C + D) and transiently pTKO-SOXN-CDS-transfected (E) cells. A, nonstimulated cells grown in 10% FCS. B, cells stimulated by 10% FCS after serum withdrawal. C, cells stimulated for 24 h with 1% FCS +1000 units/ml IFN-γ. D, cells stimulated for 48 h with 1% FCS + 1000 units/ml IFN-γ. E, transiently pTKO-SOXN-CDS-transfected cells expressing exogenous SOXN.

Fig. 7.

Immunofluorescent detection of endogenous and exogenous SOXN in Tet21N neuroblastoma cells. Nonstimulated (10% FCS, -IFN) or low serum/IFN-γ-stimulated (1% FCS and 1000 units/ml IFN-γ) cells were incubated with SOXN-specific antibody, and SOXN-specific signals were detected with Cy3-coupled antirabbit antibody. Nuclei were counterstained with DAPI. Note the augmented extracellular (nonpermeabilized) and intracellular (permeabilized) staining in IFN-γ-treated (C + D) and transiently pTKO-SOXN-CDS-transfected (E) cells. A, nonstimulated cells grown in 10% FCS. B, cells stimulated by 10% FCS after serum withdrawal. C, cells stimulated for 24 h with 1% FCS +1000 units/ml IFN-γ. D, cells stimulated for 48 h with 1% FCS + 1000 units/ml IFN-γ. E, transiently pTKO-SOXN-CDS-transfected cells expressing exogenous SOXN.

Close modal
Fig. 8.

Reduced clonogenic survival of SOXN-overexpressing cells after initial apoptotic stimulus. pTKO-SOXN-CDS (SOXN)-transfected or control plasmid NK-transfected cells were grown for 3 days in medium supplemented with 1% FCS, 500 units/ml IFN-γ, 250 μg/ml Zeocin (MYCN switched on), and for additional 7 days in the presence of 1% FCS, Zeocin, and MYCN but without IFN-γ treatment. On day 10 of selection, colonies were stained with Giemsa and scored. Results were repeated in two independent transfections, each carried out in triplicate. Bars represent the mean of three plates ± SD.

Fig. 8.

Reduced clonogenic survival of SOXN-overexpressing cells after initial apoptotic stimulus. pTKO-SOXN-CDS (SOXN)-transfected or control plasmid NK-transfected cells were grown for 3 days in medium supplemented with 1% FCS, 500 units/ml IFN-γ, 250 μg/ml Zeocin (MYCN switched on), and for additional 7 days in the presence of 1% FCS, Zeocin, and MYCN but without IFN-γ treatment. On day 10 of selection, colonies were stained with Giemsa and scored. Results were repeated in two independent transfections, each carried out in triplicate. Bars represent the mean of three plates ± SD.

Close modal
Table 1

Genomic structure of SOXN

SOXN comprises 12 exons. Exon and intron acceptor and donor splice sites are shown with corresponding segment sizes in nucleotide base pairs. The size of intron 11 was determined by long-range PCR (data not shown). Human Q6 is also organized in 12 exons as determined by aligning the Q6 cDNA to the genomic clone AI390718 (intron/exon boundaries not shown).

ExonAcceptorDonorExon sizeIntron sizeExon size Q6
 616: CCTGGCTGGGGATGTGCGAG 616 >10 kb >340 
  GT GAGGCGCGCCCCCGCGCGGGGACCC    
617: CACTGAAGGTCATGCTCCTGCTTTTC AG 717: CACTTCTACCCCACCTTCCGG 100 1966 99 
 ACTGGGCCAGTGCCATTCGC GT GAGTATCACGCGGCTGCTTGTCCGTTG    
718: TCTTACTTGTTCTGTGTTTTTCCTCTT AG 766: ACTGGAGAAAATTTTAAAG 48 746 46 
 TATTTTAAAGCATTTACAAAG GT AAGGACATAAGCATTGTATTTTAGTCC    
767: CAGCTCTGACCCTGGTGTGCTTCCGGC AG 872: GCGCCTAGACCCCATTCA 105 153 103 
 GACCTGACCGAGAGCTGC GT GAGTAGCTCCATGAAGAAGGCGGCCG    
873: ACTAAAAGCCTCTTTTTTTCCTCTTGAA AG 963: TCCTACCTTGGACGGGAG 90 1818 89 
 GCCCAGTGATGTTCTTTCC GT ACGGCCTCTGCGTGGGCTCCAGGTGTC    
964: TGGAAGTGTCATTGTGTTCTCATCCCCC AG 1109: CGCATGGATTGATTAACGT 143 2632 147 
 GTGATCTTAGACCTGAT GT GAGTGCCCGGGTGGGGTCTGGGAAC    
1110: TCTCACCTTTGCCCTCCCCTCTCTCAC AG 1244: GTTTGGAGAGAATTTGACAA 136 221 135 
 CGTGAAGCCTCTGCGGGCCT GC AAGTGTTCCTCTTGAACTGAAGCCCC    
1245: TTCCCTGACGCTGCCATATCCCGTCCT AG 1375: TTGTGACTGTCTTGGCCAAG 130 1955 133 
 GTCGAAGCTGTACACGGTGG GT TTGGGACGGTGCTGGTCTGGGCTGCAT    
1376: AGCTCTCAGGTGGCTCCTTCCTCATGC AG 1497: AACAACAAGATGCGG 123 1295 122 
 CTGTTCCCTGGACGGCCG GT GAGCCCCAGAACCCTCGCCAGCCACG    
10 1498: TTGGGGTTTTTTTGTTTGTTTGTTTTTC AG 1648: GATGCACTGGTTGGCACAG 150 4704 145 
 ATTTCTGGAATATTCCTTAC GT AAGCTGTGCCCATACAGGCCAGAGGG    
11 1649: CGTTGTCAGCAGTGTTTCTTCCCCTGC AG 1837: AACGGCCGCCTGGCAG 187 1750 177 
 GCTTTGAAGACGACCCC GT GAGAAGCCCCTGGGCATGGGGGGCTC    
12 1838: TCTGGCTCTCCCGCTTCCTTTTCTTCCT AG  2967 1663  
 GCCATCTGAGTGAGGATC     
ExonAcceptorDonorExon sizeIntron sizeExon size Q6
 616: CCTGGCTGGGGATGTGCGAG 616 >10 kb >340 
  GT GAGGCGCGCCCCCGCGCGGGGACCC    
617: CACTGAAGGTCATGCTCCTGCTTTTC AG 717: CACTTCTACCCCACCTTCCGG 100 1966 99 
 ACTGGGCCAGTGCCATTCGC GT GAGTATCACGCGGCTGCTTGTCCGTTG    
718: TCTTACTTGTTCTGTGTTTTTCCTCTT AG 766: ACTGGAGAAAATTTTAAAG 48 746 46 
 TATTTTAAAGCATTTACAAAG GT AAGGACATAAGCATTGTATTTTAGTCC    
767: CAGCTCTGACCCTGGTGTGCTTCCGGC AG 872: GCGCCTAGACCCCATTCA 105 153 103 
 GACCTGACCGAGAGCTGC GT GAGTAGCTCCATGAAGAAGGCGGCCG    
873: ACTAAAAGCCTCTTTTTTTCCTCTTGAA AG 963: TCCTACCTTGGACGGGAG 90 1818 89 
 GCCCAGTGATGTTCTTTCC GT ACGGCCTCTGCGTGGGCTCCAGGTGTC    
964: TGGAAGTGTCATTGTGTTCTCATCCCCC AG 1109: CGCATGGATTGATTAACGT 143 2632 147 
 GTGATCTTAGACCTGAT GT GAGTGCCCGGGTGGGGTCTGGGAAC    
1110: TCTCACCTTTGCCCTCCCCTCTCTCAC AG 1244: GTTTGGAGAGAATTTGACAA 136 221 135 
 CGTGAAGCCTCTGCGGGCCT GC AAGTGTTCCTCTTGAACTGAAGCCCC    
1245: TTCCCTGACGCTGCCATATCCCGTCCT AG 1375: TTGTGACTGTCTTGGCCAAG 130 1955 133 
 GTCGAAGCTGTACACGGTGG GT TTGGGACGGTGCTGGTCTGGGCTGCAT    
1376: AGCTCTCAGGTGGCTCCTTCCTCATGC AG 1497: AACAACAAGATGCGG 123 1295 122 
 CTGTTCCCTGGACGGCCG GT GAGCCCCAGAACCCTCGCCAGCCACG    
10 1498: TTGGGGTTTTTTTGTTTGTTTGTTTTTC AG 1648: GATGCACTGGTTGGCACAG 150 4704 145 
 ATTTCTGGAATATTCCTTAC GT AAGCTGTGCCCATACAGGCCAGAGGG    
11 1649: CGTTGTCAGCAGTGTTTCTTCCCCTGC AG 1837: AACGGCCGCCTGGCAG 187 1750 177 
 GCTTTGAAGACGACCCC GT GAGAAGCCCCTGGGCATGGGGGGCTC    
12 1838: TCTGGCTCTCCCGCTTCCTTTTCTTCCT AG  2967 1663  
 GCCATCTGAGTGAGGATC     

We thank Andreas Claas and Inge Krebs for helpful discussions, Young-Gyu Park for performing the loss of heterozygosity study, and Britta Maedge for providing plasmid pSP64-MYCN.

1
Kaneko Y., Cohn S. L. Ploidy and cytogenetics of neuroblastoma Brodeur G. M. Sawada T. Tsuchida Y. Voute P. A. eds. .
Neuroblastoma
,
41
-56, Elsevier Amsterdam  
2000
.
2
Ambros P. F., Brodeur G. M. Concept of tumorigenesis and regression Brodeur G. M. Sawada T. Tsuchida Y. Voute P. A. eds. .
Neuroblastoma
,
21
-32, Elsevier Amsterdam  
2000
.
3
Nakagawara A. Molecular basis of spontaneous regression of neuroblastoma: role of neurotrophic signals and genetic abnormalities.
Hum Cell.
,
11
:
115
-124,  
1998
.
4
Hamburger V. History of the discovery of neuronal death in embryos.
J. Neurobiol.
,
23
:
1116
-1123,  
1992
.
5
Ikeda H., Castle V. P. Apoptosis in neuroblastoma and pathways involving Bcl-2 and caspases Brodeur G. M. Sawada T. Tsuchida Y. Voute P. A. eds. .
Neuroblastoma
,
197
-206, Elsevier Amsterdam  
2000
.
6
Ikeda H., Hirato J., Akami M., Suzuki N., Takahashi A., Kuroiwa M., Matsuyama S. Massive apoptosis detected by in situ DNA nick end labeling in neuroblastoma.
Am. J. Surg. Pathol.
,
20
:
649
-655,  
1996
.
7
Hopkins-Donaldson S., Bodmer J. L., Bourloud K. B., Brognara C. B., Tschopp J., Gross N. Loss of caspase-8 expression in highly malignant human neuroblastoma cells correlates with resistance to tumor necrosis factor-related apoptosis-inducing ligand-induced apoptosis.
Cancer Res.
,
60
:
4315
-4319,  
2000
.
8
Nakagawara A., Nakamura Y., Ikeda H., Hiwasa T., Kuida K., Su M. S., Zhao H., Cnaan A., Sakiyama S. High levels of expression and nuclear localization of interleukin-1β converting enzyme (ICE) and CPP32 in favorable human neuroblastomas.
Cancer Res.
,
57
:
4578
-4584,  
1997
.
9
Fulda S., Sieverts H., Friesen C., Herr I., Debatin K. M. The CD95 (APO-1/Fas) system mediates drug-induced apoptosis in neuroblastoma cells.
Cancer Res.
,
57
:
3823
-3829,  
1997
.
10
Behrens M. M., Martinez J. L., Moratilla C., Renart J. Apoptosis induced by protein kinase C inhibition in a neuroblastoma cell line.
Cell Growth Differ.
,
6
:
1375
-1380,  
1995
.
11
Wittke I., Madge B., Wiedemeyer R., Kimchi A., Schwab M. DAP-5 is involved in MycN/IFNγ-induced apoptosis in human neuroblastoma cells.
Cancer Lett.
,
162
:
237
-243,  
2001
.
12
Lutz W., Fulda S., Jeremias I., Debatin K. M., Schwab M. MycN and IFNγ cooperate in apoptosis of human neuroblastoma cells.
Oncogene
,
17
:
339
-346,  
1998
.
13
Fulda S., Lutz W., Schwab M., Debatin K. M. MycN sensitizes neuroblastoma cells for drug-induced apoptosis.
Oncogene
,
18
:
1479
-1486,  
1999
.
14
Deiss L. P., Kimchi A. A genetic tool used to identify thioredoxin as a mediator of a growth inhibitory signal.
Science (Wash. DC)
,
252
:
117
-120,  
1991
.
15
Hoober K. L., Joneja B., White H. B., III, Thorpe C. A sulfhydryl oxidase from chicken egg white.
J. Biol. Chem.
,
271
:
30510
-30516,  
1996
.
16
Ostrowski M. C., Kistler W. S. Properties of a flavoprotein sulfhydryl oxidase from rat seminal vesicle secretion.
Biochemistry
,
19
:
2639
-2645,  
1980
.
17
Hoober K. L., Sheasley S. L., Gilbert H. F., Thorpe C. Sulfhydryl oxidase from egg white. A facile catalyst for disulfide bond formation in proteins and peptides.
J. Biol. Chem.
,
274
:
22147
-22150,  
1999
.
18
Lisowsky T. ERV1 is involved in the cell-division cycle and the maintenance of mitochondrial genomes in Saccharomyces cerevisiae.
Curr. Genet.
,
26
:
15
-20,  
1994
.
19
Giorda R., Hagiya M., Seki T., Shimonishi M., Sakai H., Michaelson J., Francavilla A., Starzl T. E., Trucco M. Analysis of the structure and expression of the augmenter of liver regeneration (ALR) gene.
Mol. Med.
,
2
:
97
-108,  
1996
.
20
Coppock D. L., Cina-Poppe D., Gilleran S. The quiescin Q6 gene (QSCN6) is a fusion of two ancient gene families: thioredoxin and ERV1.
Genomics
,
54
:
460
-468,  
1998
.
21
Benayoun B., Esnard-Feve A., Castella S., Courty Y., Esnard F. Rat seminal vesicle FAD-dependent sulfhydryl oxidase. Biochemical characterization and molecular cloning of a member of the new sulfhydryl oxidase/quiescin Q6 gene family.
J. Biol. Chem.
,
276
:
13830
-13837,  
2001
.
22
Hashimoto Y., Suga Y., Matsuba S., Mizoguchi M., Takamori K., Seitz J., Ogawa H. Inquiry into the role of skin sulfhydryl oxidase in epidermal disulfide bond formation: implications of the localization and regulation of skin SOx as revealed by TPA, retinoic acid, and UVB radiation.
J. Investig. Dermatol.
,
117
:
752
-754,  
2001
.
23
Hoober K. L., Glynn N. M., Burnside J., Coppock D. L., Thorpe C. Homology between egg white sulfhydryl oxidase and quiescin Q6 defines a new class of flavin-linked sulfhydryl oxidases.
J. Biol. Chem.
,
274
:
31759
-31762,  
1999
.
24
Wiedemeyer R., Westermann F., Wittke I., Nowock J., Schwab M. Ataxin-2 promotes apoptosis of human neuroblastoma cells.
Oncogene
,
22
:
401
-411,  
2003
.
25
Lutz W., Stohr M., Schurmann J., Wenzel A., Lohr A., Schwab M. Conditional expression of N-myc in human neuroblastoma cells increases expression of α-prothymosin and ornithine decarboxylase and accelerates progression into S-phase early after mitogenic stimulation of quiescent cells.
Oncogene
,
13
:
803
-812,  
1996
.
26
Bannasch D., Weis I., Schwab M. Nmi protein interacts with regions that differ between MycN and Myc and is localized in the cytoplasm of neuroblastoma cells in contrast to nuclear MycN.
Oncogene
,
18
:
6810
-6817,  
1999
.
27
Hirt B. Selective extraction of polyoma DNA from infected mouse cell cultures.
J. Mol. Biol.
,
26
:
365
-369,  
1967
.
28
Skehan P., Storeng R., Scudiero D., Monks A., McMahon J., Vistica D., Warren J. T., Bokesch H., Kenney S., Boyd M. R. New colorimetric cytotoxicity assay for anticancer-drug screening.
J. Natl. Cancer Inst. (Bethesda)
,
82
:
1107
-1112,  
1990
.
29
An H. X., Claas A., Savelyeva L., Seitz S., Schlag P., Scherneck S., Schwab M. Two regions of deletion in 9p23-24 in sporadic breast cancer.
Cancer Res.
,
59
:
3941
-3943,  
1999
.
30
Savelyeva L., Claas A., Gier S., Schlag P., Finke L., Mangion J., Stratton M. R., Schwab M. An interstitial tandem duplication of 9p23–24 coexists with a mutation in the BRCA2 gene in the germ line of three brothers with breast cancer.
Cancer Res.
,
58
:
863
-866,  
1998
.
31
Chersi A., Ruocco E., Muratti E. Preparation and properties of gel-peptide immunoadsorbents.
J. Immunol. Methods
,
122
:
285
-289,  
1989
.
32
Rehm, H. Der Experimentator: Proteinbiochemie, (Ed. 2), Vol. 2te Auflage, p. 217. Stuttgart, Jena, Lübeck, Ulm: Gustav Fischer Verlag, 1997.
33
Wu Y., Ozaki Y., Inoue K., Satoh K., Ohmori T., Yatomi Y., Owadab K. Differential activation and redistribution of c-Src and Fyn in platelets, assessed by MoAb specific for C-terminal tyrosine-dephosphorylated c-Src and Fyn.
Biochim. Biophys. Acta
,
1497
:
27
-36,  
2000
.
34
Thorpe C., Hoober K. L., Raje S., Glynn N. M., Burnside J., Turi G. K., Coppock D. L. Sulfhydryl oxidases: emerging catalysts of protein disulfide bond formation in eukaryotes.
Arch. Biochem. Biophys.
,
405
:
1
-12,  
2002
.
35
Kozak M. An analysis of 5′-noncoding sequences from 699 vertebrate messenger RNAs.
Nucleic Acids Res.
,
15
:
8125
-8148,  
1987
.
36
Devireddy L. R., Jones C. Alternative splicing of the latency-related transcript of bovine herpesvirus 1 yields RNAs containing unique open reading frames.
J Virol.
,
72
:
7294
-7301,  
1998
.
37
Gyapay G., Morissette J., Vignal A., Dib C., Fizames C., Millasseau P., Marc S., Bernardi G., Lathrop M., Weissenbach J. The 1993–94 Genethon human genetic linkage map.
Nat. Genet.
,
7
:
246
-339,  
1994
.
38
Levy-Strumpf N., Kimchi A. Death associated proteins (DAPs): from gene identification to the analysis of their apoptotic and tumor suppressive functions.
Oncogene
,
17
:
3331
-3340,  
1998
.
39
Hofmann E. R., Boyanapalli M., Lindner D. J., Weihua X., Hassel B. A., Jagus R., Gutierrez P. L., Kalvakolanu D. V., Hofman E. R. Thioredoxin reductase mediates cell death effects of the combination of beta interferon and retinoic acid.
Mol. Cell. Biol.
,
18
:
6493
-6504,  
1998
.
40
Lee J., Hofhaus G., Lisowsky T. Erv1p from Saccharomyces cerevisiae is a FAD-linked sulfhydryl oxidase.
FEBS Lett.
,
477
:
62
-66,  
2000
.
41
Gerber J., Muhlenhoff U., Hofhaus G., Lill R., Lisowsky T. Yeast ERV2p is the first microsomal FAD-linked sulfhydryl oxidase of the Erv1p/Alrp protein family.
J. Biol. Chem.
,
276
:
23486
-23491,  
2001
.
42
Lisowsky T., Lee J. E., Polimeno L., Francavilla A., Hofhaus G. Mammalian augmenter of liver regeneration protein is a sulfhydryl oxidase.
Dig. Liver Dis.
,
33
:
173
-180,  
2001
.
43
Senkevich T. G., White C. L., Koonin E. V., Moss B. A viral member of the ERV1/ALR protein family participates in a cytoplasmic pathway of disulfide bond formation.
Proc. Natl. Acad. Sci. USA
,
97
:
12068
-12073,  
2000
.
44
Coppock D., Kopman C., Gudas J., Cina-Poppe D. A. Regulation of the quiescence-induced genes: quiescin Q6, decorin, and ribosomal protein S29.
Biochem. Biophys. Res. Commun.
,
269
:
604
-610,  
2000
.
45
Takita J., Hayashi Y., Kohno T., Yamaguchi N., Hanada R., Yamamoto K., Yokota J. Deletion map of chromosome 9 and p16 (CDKN2A) gene alterations in neuroblastoma.
Cancer Res.
,
57
:
907
-912,  
1997
.
46
Evan G. I., Wyllie A. H., Gilbert C. S., Littlewood T. D., Land H., Brooks M., Waters C. M., Penn L. Z., Hancock D. C. Induction of apoptosis in fibroblasts by c-myc protein.
Cell
,
69
:
119
-128,  
1992
.
47
Gallegos A., Gasdaska J. R., Taylor C. W., Paine-Murrieta G. D., Goodman D., Gasdaska P. Y., Berggren M., Briehl M. M., Powis G. Transfection with human thioredoxin increases cell proliferation and a dominant-negative mutant thioredoxin reverses the transformed phenotype of human breast cancer cells.
Cancer Res.
,
56
:
5765
-5770,  
1996
.
48
Mau B. L., Powis G. Inhibition of cellular thioredoxin reductase by diaziquone and doxorubicin. Relationship to the inhibition of cell proliferation and decreased ribonucleotide reductase activity.
Biochem. Pharmacol.
,
43
:
1621
-1627,  
1992
.
49
Berggren M., Gallegos A., Gasdaska J. R., Gasdaska P. Y., Warneke J., Powis G. Thioredoxin and thioredoxin reductase gene expression in human tumors and cell lines, and the effects of serum stimulation and hypoxia.
Anticancer Res.
,
16
:
3459
-3466,  
1996
.
50
Ma X., Karra S., Lindner D. J., Hu J., Reddy S. P., Kimchi A., Yodoi J., Kalvakolanu D. D. Thioredoxin participates in a cell death pathway induced by interferon and retinoid combination.
Oncogene
,
20
:
3703
-3715,  
2001
.
51
Polyak K., Xia Y., Zweier J. L., Kinzler K. W., Vogelstein B. A model for p53-induced apoptosis.
Nature (Lond.)
,
389
:
300
-305,  
1997
.
52
Hu J., Ma X., Lindner D. J., Karra S., Hofmann E. R., Reddy S. P., Kalvakolanu D. V. Modulation of p53-dependent gene expression and cell death through thioredoxin-thioredoxin reductase by the interferon-retinoid combination.
Oncogene
,
20
:
4235
-4248,  
2001
.
53
Hainaut P., Milner J. Redox modulation of p53 conformation and sequence-specific DNA binding in vitro.
Cancer Res.
,
53
:
4469
-4473,  
1993
.
54
Schenk H., Klein M., Erdbrugger W., Droge W., Schulze-Osthoff K. Distinct effects of thioredoxin and antioxidants on the activation of transcription factors NF-κB and AP-1.
Proc. Natl. Acad. Sci. USA
,
91
:
1672
-1676,  
1994
.
55
Arner E. S., Holmgren A. Physiological functions of thioredoxin and thioredoxin reductase.
Eur. J. Biochem.
,
267
:
6102
-6109,  
2000
.
56
Sevier C. S., Cuozzo J. W., Vala A., Aslund F., Kaiser C. A. A flavoprotein oxidase defines a new endoplasmic reticulum pathway for biosynthetic disulphide bond formation.
Nat. Cell Biol.
,
3
:
874
-882,  
2001
.
57
Saitoh M., Nishitoh H., Fujii M., Takeda K., Tobiume K., Sawada Y., Kawabata M., Miyazono K., Ichijo H. Mammalian thioredoxin is a direct inhibitor of apoptosis signal-regulating kinase (ASK) 1.
EMBO J.
,
17
:
2596
-2606,  
1998
.
58
Arner E. S., Zhong L., Holmgren A. Preparation and assay of mammalian thioredoxin and thioredoxin reductase.
Methods Enzymol.
,
300
:
226
-239,  
1999
.
59
Bennett M., Macdonald K., Chan S. W., Luzio J. P., Simari R., Weissberg P. Cell surface trafficking of Fas: a rapid mechanism of p53-mediated apoptosis.
Science (Wash. DC)
,
282
:
290
-293,  
1998
.