Abstract

Apoptosis is a morphologically distinct form of cell death involved in many physiological and pathological processes. The regulation of Fas/Apo-1 involved in membrane-mediated apoptosis has also been known to play crucial roles in many systems. More and more naturally occurring antisense RNAs are now known to regulate, at least in part, a growing number of eukaryotic genes. In this report, we describe the findings of a novel RNA transcribed from the opposite strand of the intron 1 of the human Fas gene. Using orientation-specific RT–PCR and northern blot analysis, we show that this transcript is 1.5 kb in length and was expressed in several human tissues and cell lines. This transcript was cloned by 5′- and 3′-RACE (rapid amplification of cDNA ends) and the transcription start site was determined by primer extension. This novel gene was named Saf. To assess the functions of Saf, Jurkat cells transfected with human Saf or control vector was prepared. The stable Saf-transfectant was highly resistant to Fas-mediated but not to TNF-α-mediated apoptosis. Although the overall mRNA expression level of Fas was not affected, expression of some novel forms of Fas transcripts was increased in Saf-transfectant, especially the inhibitory soluble forms. These findings collectively suggest that Saf might protect T lymphocytes from Fas-mediated apoptosis by blocking the binding of FasL or its agonistic Fas antibody. Saf might regulate the expression of Fas alternative splice forms through pre-mRNA processing.

INTRODUCTION

Fas (Apo-1/CD95) is a receptor belonging to the tumor necrosis factor (TNF) and nerve growth factor receptor family. The human Fas gene consists of nine exons and is transcribed from chromosome 10q24.1. Once Fas is triggered by agonistic antibodies or its natural ligand (FasL), it induces apoptosis via direct and sequential activation of caspases. Fas is physiologically expressed in various tissues such as spleen, lymph nodes and on mature hematologic cells such as T- and B-cells, natural killer cells, monocytes and granulocytes (1). Thus, in both the central and the peripheral immune compartments, apoptosis plays an essential role in maintenance of tolerance and in the termination of an ongoing immune response (2). Deregulation of apoptosis contributes to many diseases including autoimmunity, cancer, stroke, drug resistance in tumors and the acquired immunodeficiency syndrome (3). Fas-mediated apoptosis can be blocked by several mechanisms: (i) production of secreted form of soluble Fas, (ii) lack of cell-surface Fas expression, (iii) overexpression of inhibitory proteins in signal transduction pathways such as Fas associated phosphatase and FLICE-inhibitory protein and (iv) alterations of the primary structure of Fas due to mutations (4,5). In this study, we identified a novel endogenous antisense transcript of Fas and analyzed its role in the modulation of Fas-mediated apoptosis.

A growing number of endogenous antisense RNAs have been identified in many eukaryotic organisms over the last several years (68). Antisense transcripts often code for proteins involved in diverse biological functions. Non-coding antisense transcripts have also been reported. Their role appears to be mainly regulatory (9) and their transcription is often associated with genomic imprinting (10). Although the effect of antisense RNAs on the corresponding sense RNAs are as yet mostly unclear, growing evidences have indicated an involvement in gene regulation (11). Naturally occurring long antisense RNAs described so far in the literature are cis-encoded, meaning that they are transcribed from the opposite strand and correspond to the complete or partial hnRNA sequence of the gene (12). In a small percentage of cases, the transcript initiates within an intron of its counterpart gene or entirely covers its counterpart, for example, c-erbA (13). Mechanisms proposed for the regulation of gene expression by antisense RNA are numerous, and the discovery of novel sense/antisense RNA interactions would be insightful in understanding this mechanism of gene regulation.

Our unpublished data unexpectedly led to the discovery of an endogenous antisense RNA transcript of Fas in the human Jurkat T cell. This novel gene is named Saf. The aim of our study was to identify and characterize this novel antisense transcript. Our data showed that this transcript was 1.5 kb in length and expressed in the heart, placenta, liver, muscle, kidney and pancreas. In addition, apoptosis induced by the Fas agonistic antibody, CH11, was attenuated by Saf expression. These findings demonstrate that Saf, which is transcribed from the strand opposite to Fas, might play an important role in the regulation of the Fas gene expression and Fas-mediated apoptosis.

RESULTS

Identification of the novel antisense transcript

On the basis of the genomic sequence information corresponding to the human Fas, we found two entries in the database of human expressed sequence tags (ESTs) (GenBank accession nos AA018423 and AA058563) that had a transcription orientation opposite to that of Fas (Fig. 1). Sequence analysis was performed using the BLAST search algorithm through the NCBI WWW site. These two overlapping EST sequences are continuous and located within intron 1 of the Fas gene. This novel antisense gene was designated as Saf. Strand-specific RT–PCR was performed to verify the presence of the Saf transcript. It was found that the expected product of 415 bp was generated only when the reverse transcription was primed with the Saf53 primer (Fig. 2A) but not with the Saf33 primer (Fig. 2B). Taken together, these results confirmed the presence of the Saf transcript that was antisense to Fas.

Characterization of the Saf antisense transcript

RACE technique was used to clone both the 5′ and the 3′ ends of the Saf transcript from human placenta RNA. Gel analysis of the RACE products revealed one single band with 3′-RACE and several bands with 5′-RACE (Fig. 3A, upper panel). Southern blot with specific Dig-labeled probe generated by PCR using the Saf53/Saf33 primer pair revealed a specific band of 500 bp with 3′-RACE and a specific band of 550 bp with 5′-RACE (Fig. 3A, lower panel). Transcription start site of Saf was determined by primer extension. Three extension products from Jurkat cells were observed with primer Saf5Race (Fig. 3B, lane 4). Three possible transcription start sites at positions 350, 450 and 550 bp upstream of the Saf5Race primer were then determined. The longest 550 bp product was compatible with the 5′-RACE–PCR result implicating that the transcription start site was at position 550 bp upstream of the primer Saf5Race site. On the basis of the EST information and the RACE results, this Saf transcript was estimated to be 1.5 kb in length.

The 3′-RACE–PCR and 5′-RACE–PCR products were cloned and sequenced to obtain the full-length transcript sequences. It turned out that the full-length Saf antisense transcript was 1556 bp long and was entirely originated from intron 1 of the Fas gene. Fas and Saf are transcribed from the same locus but in opposite orientations (Fig. 1). After analysis with program TSSW (http://www.softberry.com), one promoter was predicted upstream of the transcription start site. Six transcription factor binding sites were predicted including one TATA box, two GATA-2 boxes, one MZF1 element, one SRY (testis determining factor binding) element and one c/EBPβ element. The 3′ end of the Saf transcript was located 23 bp downstream of the poly(A) signal (AATAAA) (Fig. 4). The Saf cDNA sequence exactly matched the genomic sequence, indicating that Saf is an intronless gene. In addition, the largest open reading frame predicted in this transcript was only 120 bp long, and we could not detect any signals using the coupled transcription–translation system (data not shown). Therefore, it suggests that Saf was not a protein-coding gene.

Northern blot analysis was used to examine the Saf expression in different tissues. A major transcript of 1.5 kb was detected (Fig. 5A). Saf was expressed in human heart, placenta, liver, muscle and pancreas. It was only slightly expressed in kidney and undetectable in brain and lung. As for cancer cell lines, Saf was expressed in Jurkat (acute T-cell leukemia), MES-SA (uterine carcinoma) and U-937 (Histocytic Lymphoma); moderately expressed in CA46 (Burkitt's lymphoma), Raji (Burkitt's lymphoma) and only slightly expressed in MG-63 (osteosarcoma) (Fig. 5B). The expression of Fas was also examined. It was expressed in the heart, liver and Jurkat cells; moderately expressed in brain, MES-SA cells and MG-63 cells; slightly in placenta, kidney, U-937, A-431 and MDA-MD-453 cells (breast carcinoma).

Effect of Saf overexpression on the expression of Fas

To study the potential regulation of the Fas gene expression by the antisense Saf, overexpression of Saf was performed in Jurkat cells. The RNA expression levels of Fas in parental Jurkat cells, vector-transfectant and Saf-transfectant were determined by quantitative, real-time RT–PCR. The overall Fas expression was not significantly affected in the Saf-transfectant in comparison to the parental Jurkat cells and the vector-transfectant. Saf expression in the Saf-transfectant was 1.9-fold that of the parental Jurkat and the vector-transfectant cells (Fig. 6).

Overexpression of Saf protects cells from Fas-mediated apoptosis

To examine the effect of Saf overexpression on the sensitivity of Jurkat cells to Fas-mediated apoptosis and to check whether Saf could protect cells from apoptosis induced by other ligands, we treated cells with the agonistic antibody CH11 (100 ng/ml), FasL (40 ng/ml) and TNF-α (100 ng/ml). Apoptotic cells were measured by using FITC-conjugated annexin V staining and flow cytometry. In the parental Jurkat cells 24 h after treatment by CH11, although the majority of cells have undergone apoptosis (86%), apoptosis was significantly attenuated in the Saf-transfectant (38%). Without the addition of CH11, both the parental Jurkat and the Saf-transfectant showed only basal levels of apoptosis (Fig. 7A). Although the apoptotic effect exerted by FasL was almost identical to that by CH11, however, TNF-α-induced apoptosis was not protected by Saf (Fig. 7A), indicating that the effects of Saf is Fas-dependent. The CH11-induced apoptosis was further confirmed by the activation of caspase 8 and the cleavage of the poly ADP-ribose polymerase (PARP) protein. Although caspase 8 was fully activated and PARP was completely cleaved in the parental Jurkat and the vector-transfectant cells by CH11, caspase 8 activation and PARP cleavage were significantly attenuated in the Saf-transfectant (Fig. 7B).

Effect of Saf overexpression on expression of the Fas alternative splice forms

After analysis with AceView program (http://www.ncbi.nlm.nih.gov/AceView/), several Fas alternative splice forms retaining the Fas intron 1 sequences were found (Fig. 8A). RT–PCR was designed to verify the presence of these alternative splice forms of Fas (Fig. 8A). As shown in Figure 8B, by using various primer pairs, we confirmed the existence of these forms with expected sizes of 182, 353, 600 and 491 bp (Fas splice forms b, c, d and e, respectively). Upon alignment with the Saf sequence, the four transcript variants of Fas are found to overlap with Saf by 48, 88, 105 and 105 bp, respectively (Fig. 8A). Although the expression of b and c forms of Fas was not affected by Saf, expression of the other two Fas variants d and e forms was significantly elevated (Fig. 8B and C). It is noteworthy that the latter forms of Fas are both inhibitory soluble Fas, and the two former forms of Fas are both membranous Fas. When focusing on the Fas transcripts that contain exon 5 and exon 9, the membrane form (367 bp) and the soluble form (304 bp, lacking exon 6) can also be distinguished with the Fas5/Fas6 primer pair. As a result, we found that these two forms of Fas transcript were expressed equally in the parental Jurkat cells and the vector-transfectant, but in the Saf-transfectant, the membrane form was repressed and the soluble form was increased (Fig. 8B).

DISCUSSION

In this study, we describe the findings of an endogenous transcript that is antisense to the Fas transcript. This antisense gene, named Saf, codes for an RNA that is 1556 bp in length. The Saf gene resides entirely in the 12.1 kb intron 1 of the Fas gene and is transcribed in the opposite direction. The Saf transcript ends at 521 bp downstream of the Fas exon 1. A probable functional TATA box is located upstream of the Saf transcription start site. Furthermore, in contrast to the parental Jurkat and vector-transfectant cells, the stable Saf-transfectant was highly resistant to Fas-mediated apoptosis. In order to check whether Saf could protect cells from apoptosis induced by other ligands, we also treated cells with TNF-α and FasL. For FasL-induced apoptosis, Saf could also exert its protection effect, however, TNF-α-induced apoptosis was not protected (Fig. 7A). These findings strongly suggest that Fas-mediated apoptosis in T lymphocytes was inhibited by Saf, and Saf's protection effect was Fas/FasL-pathway specific.

Fas-mediated apoptosis is an important mechanism of immune homeostasis (14). Dysregulation of the Fas/FasL system leads to severe diseases (15,16). Our discovery of this antisense RNA may represent another level of regulation of the Fas system (17). However, the Fas's exon/intron structure of human, chimpanzee and mouse are all strikingly similar, especially the presence of a relatively long intron 1, whereas sequences of the murine's intron 1 have no homology with those of the primate's. Interestingly, human and chimpanzee share high degree of similarity in intron 1 sequences (99% nucleotide identity in the whole 11.5 kb overlapping region) including the entire Saf region. Thus, this Saf gene appears to be conserved only among the primates and is absent in the rodents. The appearance of the Saf gene may, therefore, represent a relatively recent event in evolution.

Short endogenous antisense RNAs are generally trans-acting and produced from a locus different from that of their corresponding genes, whereas the transcription of long antisense RNA (generally cis-acting) often occurs in the same locus as the sense RNA (18). Our data showed that the Saf transcript did not code for proteins and the Saf gene is transcribed from Fas intron 1 in the opposite direction. Therefore, we propose that this transcript possibly belongs to the category of long antisense RNA and functional non-coding RNAs (8). Totally intronic antisense transcripts would have to target their complementary intronic sense message (in pre-mRNA) to play a regulatory role in processing protein-encoding RNA messages (19). We speculate that sense/antisense intronic transcript pairs may be part of a general regulatory mechanism of gene expression not yet fully understood in eukaryotes. Cis-acting and trans-acting antisense RNAs overlapping with corresponding sense RNAs have been reported for several eukaryotic genes, and they have been implicated in imprinting, alternative splicing regulation, post-transcriptional and translational regulation (20). As reported here, the possibility exists that the Fas gene may be regulated by an antisense mechanism. At least four alternative splice forms of Fas that contain intron 1 sequences overlapping with Saf were found in Jurkat cells. Among them, the two soluble forms (d and e) were upregulated by Saf and the two membrane forms (b and c) were not affected, resulting in an increased ratio of soluble forms to membrane forms (Fig. 8). As the Saf-overlapping regions of forms b and c are different from those of forms d and e, it is possible that either the length or the nature of the overlapping sequences or both are responsible for the difference. It is also worth to note that the 3′-extremity of the Saf transcript corresponds to the 5′-extremity of the d and e forms of the Fas transcript. To investigate the possible involvement of these overlapping sequences in the regulation of Saf on Fas, next we may plan to construct various expression vectors of Saf that have mutations in these overlapping regions to see how these mutations may affect the effects of Saf. The Saf transcript might modulate the splicing of the Fas pre-mRNA by the formation of a dsRNA duplex that favors or disfavors the production of these alternative splice forms. It is possible that these alternative splice forms might play crucial roles in modulating the functions of Fas, for example, soluble forms of Fas that neutralize FasL and/or dominant negative forms of Fas that lack the death domain (Fig. 8). It is, therefore, apparent that intronic sequences may be of functional significance, as has been noted in the mechanism of antisense transcript at the N-myc locus and others (2123) as well as the antisense Saf reported here.

More and more naturally occurring antisense RNAs are being discovered, and researchers have been engineering and expressing antisense transcripts as a means to control gene expression (24,25). Several research groups have been using libraries of genomic fragments to express large numbers of diverse antisense RNAs (2628). In eukaryotes, there are now several examples where expressed microRNA (miRNA) can repress translation in an RNA interference independent manner (29). Hopefully, with more and more understanding of the antisense RNA structures, functions and their regulations, endogenous antisense RNA genes could potentially serve as a target for therapeutic applications. Splice variants have been associated with a variety of cancers and genetic diseases such as cystic fibrosis and thalassemia. The use of antisense oligonucleotides to modulate alternative splicing has been suggested to be a promising RNA repair-based approach to engender the production of therapeutic gene products (30).

It is apparent from the earlier discussion that although many examples of naturally occurring transcripts have been reported to date, no general rule of antisense RNA in the regulation of gene expression has been firmly established. We propose that the Saf transcript might form duplexes with the Fas pre-mRNA in the intron 1 regions and hence modulate gene expression at the RNA splicing level. We will be focusing on those alternative forms of Fas transcripts that have sequences overlapping with the Saf transcript, especially the first exon of the d and e forms of Fas transcripts.

MATERIALS AND METHODS

Cell line

The human cell lines Jurkat (acute T cell leukemia), Raji (Burkitt's lymphoma), Hela (cervical adenocarcinoma) and U-937 (myeloid leukemia) cells were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (Gibco BRL, Rockville, MD, USA), 2 µM glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin at 37°C in a humidified incubator with 5% CO2.

RNA extraction

Total cellular RNA of cultured cells was isolated with the cytoplasmic RNA purification reagent ConcertTM (Invitrogen, Carlsbad, CA, USA). Total RNA was treated with DNase I (Ambion, Austin, TX, USA) for 30 min at 37°C to remove any contaminated genomic DNA before RT–PCR. The quality and quantity of the RNA preparations were tested by agarose gel electrophoresis and measurement of absorption at 260 and 280 nm. The ratio of optical density at 260 and 280 nm was ≥1.8 in all cases.

Strand-specific RT–PCR

First-strand cDNA was synthesized from 1 µg total RNA at 50°C for 50 min with either sense (Saf33) or antisense (Saf53) primer by Superscript II reverse transcriptase (Gibco BRL), and then amplified with primers Saf52 and Saf32 by using Platinum Taq polymerase (Invitrogen). Conditions for PCR were 1 min each at 94, 55 and 72°C for 35 cycles. The designed primer Saf53 would prime cDNA synthesis only from the Saf transcript but not the Fas transcript.

Identification of Fas transcriptional variants by RT–PCR

cDNAs were synthesized from total RNA of parental Jurkat cells, vector-transfectant and Saf-transfectant, then amplified with specific primer pairs of Fas1/Fas2, Fas3/Fas4, Fas5/Fas6 and GAPDH1/GAPDH2, respectively. Reverse transcriptase negative controls (RTase−) were also included. The RT–PCR products were examined with agarose gel electrophoresis.

Rapid amplification of cDNA ends

RACE was performed by using a SMART RACE cDNA amplification kit (Clontech, Palo Alto, CA, USA) with 1 µg total RNA as the template according to the manufacturer's instructions. Gene-specific primers used in the 5′- and 3′-RACE reactions were Saf5Race and Saf3Race, respectively (Fig. 1 and Table 1). The RACE–PCR products were analyzed by agarose gel electrophoresis and Southern blotting. The products were also cloned into the pcDNA3.1–TOPO vector (Invitrogen).

Southern blot analysis

The DNA probes were synthesized by the incorporation of digoxigenin-11-dUTP during PCR using Saf52 and Saf32 as primers (Fig. 1 and Table 1). The RACE–PCR products were separated by gel electrophoresis and detected by the DIG non-radioactive detection system (Roche, Mannheim, Germany).

Primer extension reaction

Primer extension technique was used to determine the transcription start site of Saf. Reaction was carried out with 50 µg of total RNA from the Hela cells or Jurkat cells with Saf5Race primer that is complimentary to the 5′ portion of Saf mRNA. The primer was 32P end-labeled using T4 polynucleotide kinase and hybridized with total RNA in buffer containing 10 mM PIPES, 200 mM NaCl, 1 mM dithiothreitol pH 6.4 and 1 mM dNTPs. It was extended with 400 U MMLV reverse transcriptase (Gibco BRL) for 120 min at 50°C. After termination of the reaction, the reaction products were size fractionated by denaturing the polyacrylamide gel electrophoresis.

Northern blot analysis

To investigate the tissue distribution of Saf expression, northern blotting was performed. The Saf riboprobe used was generated by amplifying an 868 bp fragment from the human Saf cDNA with Saf33 and Saf53 and subcloning of the PCR product into the pcDNA3.1–TOPO vector. Antisense digoxigenin-labeled Saf RNA probe was transcribed from the T7 promoter with DIG RNA labeling kit (Roche). This strand-specific riboprobe was used to hybridize with a Human Multiple Tissue mRNA Blot 7760-1 (Clontech) and FirstChoiceTM Human Blot 3 (Ambion) at 60°C overnight. Detection of RNA was performed with the DIG non-radioactive detection system (Roche). The blot was then washed twice with 2×SSC/0.1% SDS for 45 min, followed by two washes with 0.1×SSC/0.1% SDS at room temperature for 20 min. The RNA hybrids were detected by incubation with an anti-DIG/alkaline phosphatase conjugate and a subsequent chemiluminescence reaction with CSPD and then exposed to X-ray film. The membranes were stripped and reprobed with human Fas and GAPDH.

Construction of recombinant plasmids and transfection into Jurkat cells

Single-stranded Saf cDNA was synthesized using total RNA from normal placental tissue. Primers Saf54 and Saf34 (Fig. 1 and Table 1) were used in PCR to amplify Saf cDNA. The amplified 1681 bp Saf cDNA fragment was ligated into the pcDNA3.1–TOPO vector. The amplified Saf PCR product was DNA sequence verified. Lipofectamine 2000 (Invitrogen) was used to introduce the Saf expression vector into Jurkat cells. After 16 h of incubation, the cells were selected in complete medium containing 0.5 mg/ml G418 (Gibco BRL). Selected transfectants are cloned and maintained in medium containing 0.5 mg/ml G418 and screened for the expression of Saf mRNA by northern blot analysis. Jurkat cells were also transfected with pcDNA3.1–TOPO vector alone and similarly selected for G418 resistance to serve as an experimental control.

Real-time quantitative PCR

The primers used for analysis of Fas (FasQ1/FasQ2), Saf (Saf51/Saf31) and GAPDH (GAPDHQ1/GAPDHQ2) gene expression were shown in Table 1. The PCR reactions were performed in a LightCycler apparatus using LC-Fast Start Reaction Mix SYBR Green I (Roche). Thermocycling was performed in a final volume of 10 µl containing 1.5 µl of cDNA sample; 3 mM MgCl2; 0.5 µM each of the Fas, Saf and GAPDH primers; 1 µl of LC-Fast Start Reaction Mix SYBR Green I and 1 µl of LC-Fast Start DNA Master SYBR Green I/Enzyme (containing Taq DNA polymerase, reaction buffer and deoxyribonucleotide triphosphate mixture). After 10 min at 95°C to denature the cDNA and to activate the Taq DNA polymerase, the cycling conditions were 40 cycles consisting of denaturation at 95°C for 15 s, annealing at 55°C for 15 s and extension at 72°C for 18 s. After PCR, a melting curve was constructed by increasing the temperature from 65 to 95°C with a temperature transition rate of 0.1°C/s. To ensure that the correct product was amplified in the reaction, all samples were examined by 2% agarose gel electrophoresis. The LightCycler apparatus measured the fluorescence at every cycle at the end of the annealing step. The second derivative maximum method was used to automatically determine the crossing point (Cp). This was achieved by a LightCycler software algorithm (version 3.5.3) that identified the first turning point of the fluorescence curve, corresponding to the first maximum of the second derivative curve, which served as the Cp (31). The average Cp value of the GAPDH gene was subtracted from the average Cp value of the interest genes for each sample/primer combination. Five replicates of each reaction were performed. FasΔCp=(Avg. Fas Cp−Avg. GAPDH Cp), SafΔCp=(Avg. Saf Cp−Avg. GAPDH Cp), FasΔΔCp=(Avg. FasΔCp−Avg. FasΔCpParental Jurkat), SafΔΔCp=(Avg. SafΔCp−Avg. SafΔCpParental Jurkat). The fold change in the expression of the target genes (Fas and Saf) relative to the internal control gene (GAPDH) of various cells was calculated.

Apoptosis induction and measurement

For the induction of apoptosis, Jurkat cells with Saf expression vector or pcDNA 3.1–TOPO vector were cultured in six-well plates. Samples were treated with or without the following reagents: human recombinant TNF-α (100 ng/ml; Sigma, St Louis, MO, USA), soluble recombinant human Fas Ligand (40 ng/ml; Alexis, Carlsbad, CA, USA) and the agonistic anti-Fas antibody CH11 (100 ng/ml; Upstate Biotechnology, Lake Placid, NY, USA), for 24 h in complete medium. Cells were then stained with Annexin V-FITC Apoptosis Detection kit I (BD Pharmingen, San Diego, CA, USA) and immediately analyzed by a FACScan flow cytometer (BD Biosciences) using CELLQUEST software. Data were calculated as % apoptosis (% annexin V-FITC-positive cells) and reported as mean % apoptosis±SEM for a minimum of three independent experiments.

Western blot analysis

At harvest, cells were lysed in 100 µl of cell lysis buffer (20 mM Tris, pH 7.4, 100 mM NaCl, 1% Triton, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin and 10 µg/ml aprotinin). After 30 min on ice, the lysate was clarified by centrifugation and the supernatant was collected. Protein concentration was determined by the bicinchoninic acid assay (Sigma). Equivalent amounts of protein (40 µg/lane) were electrophoresed on 8% SDS–polyacrylamide gel. The gel was electroblotted onto PVDF membrane. Caspase-8, PARP and β-actin were visualized by probing with respective monoclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA).

ACKNOWLEDGEMENTS

This work was supported by the National Health Research Institutes, Taipei, Taiwan, ROC.

Conflict of Interest statement. None declared.

Figure 1. Genomic organization of human Fas and Saf. Saf is transcribed from the Fas intron 1 region but in the reverse orientation. The boxes represent the exons of Fas with coding regions in black. The thick black line is Saf with the two overlapping ESTs (GenBank accession nos AA018423 and AA058563) indicated underneath. The primers are shown in their respective positions and orientations (arrows).

Figure 1. Genomic organization of human Fas and Saf. Saf is transcribed from the Fas intron 1 region but in the reverse orientation. The boxes represent the exons of Fas with coding regions in black. The thick black line is Saf with the two overlapping ESTs (GenBank accession nos AA018423 and AA058563) indicated underneath. The primers are shown in their respective positions and orientations (arrows).

Figure 2. Strand-specific RT–PCR of Saf. Raji, HeLa, Jurkat and U-937 cells were used for the strand-specific RT–PCR analysis. cDNAs were synthesized from total RNA by using Saf-specific antisense primer Saf53 (A) or the sense primer Saf33 (B). cDNA was then amplified with the primer pair Saf52/Saf32 (RTase+). Negative controls of RT–PCR without reverse transcriptase (RTase−) were also included.

Figure 2. Strand-specific RT–PCR of Saf. Raji, HeLa, Jurkat and U-937 cells were used for the strand-specific RT–PCR analysis. cDNAs were synthesized from total RNA by using Saf-specific antisense primer Saf53 (A) or the sense primer Saf33 (B). cDNA was then amplified with the primer pair Saf52/Saf32 (RTase+). Negative controls of RT–PCR without reverse transcriptase (RTase−) were also included.

Figure 3. RACE–PCR and RNA primer extension. (A) Electrophoresis of the Saf RACE–PCR product (upper panel). RNA from human placenta was used for 5′-RACE and 3′-RACE with the primer Saf5Race and Saf3Race, respectively. Southern blot of the RACE–PCR products was performed with specific Saf probe generated by PCR (lower panel). Lane 1, 3′-RACE–PCR products; lane 2, marker and lane 3, 5′-RACE–PCR products. (B) Determination of the transcription start site by primer extension. The transcription start sites of Saf were mapped with Saf5Race primer using total RNA of Hela and Jurkat cells. Three major primer extension products are indicated by black arrows. Lane 1, end-labeled 100 bp marker ladder; lane 2, yeast tRNA; lane 3, Hela cells and lane 4, Jurkat cells.

Figure 3. RACE–PCR and RNA primer extension. (A) Electrophoresis of the Saf RACE–PCR product (upper panel). RNA from human placenta was used for 5′-RACE and 3′-RACE with the primer Saf5Race and Saf3Race, respectively. Southern blot of the RACE–PCR products was performed with specific Saf probe generated by PCR (lower panel). Lane 1, 3′-RACE–PCR products; lane 2, marker and lane 3, 5′-RACE–PCR products. (B) Determination of the transcription start site by primer extension. The transcription start sites of Saf were mapped with Saf5Race primer using total RNA of Hela and Jurkat cells. Three major primer extension products are indicated by black arrows. Lane 1, end-labeled 100 bp marker ladder; lane 2, yeast tRNA; lane 3, Hela cells and lane 4, Jurkat cells.

Figure 4. Nucleotide sequence of the human Saf. Nucleotide sequence is deduced from the full-length cDNA clone derived from the human placenta RNA. Potential transcription factor binding sites and poly(A) signal (AATAAA) are indicated (open box). TSS (transcription start site) is indicated by arrowhead. The promoter region of Saf contains one c/EBPβ(CCAAT-enhancer binding protein) element, one TATA box, two GATA-2 boxes, one MZF1 (myeloid-specific retinoic acid-responsive) element and one SRY (testis determining factor binding) element. The primers used in 5′-RACE (Saf5RACE) and 3′-RACE (Saf3RACE) are also indicated. The promoter region was predicted with the TSSW Program (http://www.softberry.com), which also successfully predicted the transcription start site of Saf.

Figure 4. Nucleotide sequence of the human Saf. Nucleotide sequence is deduced from the full-length cDNA clone derived from the human placenta RNA. Potential transcription factor binding sites and poly(A) signal (AATAAA) are indicated (open box). TSS (transcription start site) is indicated by arrowhead. The promoter region of Saf contains one c/EBPβ(CCAAT-enhancer binding protein) element, one TATA box, two GATA-2 boxes, one MZF1 (myeloid-specific retinoic acid-responsive) element and one SRY (testis determining factor binding) element. The primers used in 5′-RACE (Saf5RACE) and 3′-RACE (Saf3RACE) are also indicated. The promoter region was predicted with the TSSW Program (http://www.softberry.com), which also successfully predicted the transcription start site of Saf.

Figure 5. RNA expression of Saf and Fas in various tissues and cell lines. (A) Expression of Saf and Fas on multiple tissue northern blot (MTN blots, Clontech) containing 2 µg poly(A)+ RNA per lane. (B) Expression of Saf and Fas on multiple cancer cell northern blot (Human blot 3, Ambion) containing 2 µg poly(A)+ RNA per lane. The probes used were single-stranded Saf specific riboprobe (upper panel) and Fas specific riboprobe (middle panel), respectively. Hybridization with the control GAPDH riboprobe confirmed equal amount of RNA loaded in each lane. The same blot was used consecutively for all three probes.

Figure 5. RNA expression of Saf and Fas in various tissues and cell lines. (A) Expression of Saf and Fas on multiple tissue northern blot (MTN blots, Clontech) containing 2 µg poly(A)+ RNA per lane. (B) Expression of Saf and Fas on multiple cancer cell northern blot (Human blot 3, Ambion) containing 2 µg poly(A)+ RNA per lane. The probes used were single-stranded Saf specific riboprobe (upper panel) and Fas specific riboprobe (middle panel), respectively. Hybridization with the control GAPDH riboprobe confirmed equal amount of RNA loaded in each lane. The same blot was used consecutively for all three probes.

Figure 6. Effect of Saf overexpression on Fas expression. Relative levels of Fas (black bar) and Saf (white bar) transcripts were determined using real-time quantitative RT–PCR as described in Materials and Methods. cDNA was synthesized from 1 µg total RNA of parental Jurkat cells, vector-transfectant and Saf-transfectant. Aliquots of cDNA were used as template for quantitative PCR reactions containing primers for Fas (FasQ1/FasQ2), Saf (Saf51/Saf31) and GAPDH (GAPDHQ1/GAPDHQ2). Each reaction contained cDNA derived from 10 ng total RNA. Five replicates of each reaction were performed. The fold change in gene expression (Fas and Saf) of vector-transfectant and Saf-transfectant relative to the parental Jurkat cells was determined after normalization with the internal control gene (GAPDH).

Figure 6. Effect of Saf overexpression on Fas expression. Relative levels of Fas (black bar) and Saf (white bar) transcripts were determined using real-time quantitative RT–PCR as described in Materials and Methods. cDNA was synthesized from 1 µg total RNA of parental Jurkat cells, vector-transfectant and Saf-transfectant. Aliquots of cDNA were used as template for quantitative PCR reactions containing primers for Fas (FasQ1/FasQ2), Saf (Saf51/Saf31) and GAPDH (GAPDHQ1/GAPDHQ2). Each reaction contained cDNA derived from 10 ng total RNA. Five replicates of each reaction were performed. The fold change in gene expression (Fas and Saf) of vector-transfectant and Saf-transfectant relative to the parental Jurkat cells was determined after normalization with the internal control gene (GAPDH).

Figure 7. Protection of Saf-transfectant against Fas-mediated apoptosis. (A) Parental Jurkat cells, vector-transfectant and Saf-transfectant were treated with or without TNF-α (100 ng/ml), FasL (40 ng/ml) and the agonistic anti-Fas Ab (CH11, 100 ng/ml) for 24 h. Apoptotic cells were detected by annexin V-FITC staining and flow cytometry. Data were expressed as mean % apoptosis±SEM. (B) Inhibition of the CH11-induced activation of caspase 8 and cleavage of PARP by Saf. The parental Jurkat, vector-transfectant and Saf-transfectant cells were incubated with or without CH11 (100 ng/ml) for 24 h. At harvest, the cell lysates were analyzed by western blot using monoclonal antibodies against caspase 8, PARP and β-actin.

Figure 7. Protection of Saf-transfectant against Fas-mediated apoptosis. (A) Parental Jurkat cells, vector-transfectant and Saf-transfectant were treated with or without TNF-α (100 ng/ml), FasL (40 ng/ml) and the agonistic anti-Fas Ab (CH11, 100 ng/ml) for 24 h. Apoptotic cells were detected by annexin V-FITC staining and flow cytometry. Data were expressed as mean % apoptosis±SEM. (B) Inhibition of the CH11-induced activation of caspase 8 and cleavage of PARP by Saf. The parental Jurkat, vector-transfectant and Saf-transfectant cells were incubated with or without CH11 (100 ng/ml) for 24 h. At harvest, the cell lysates were analyzed by western blot using monoclonal antibodies against caspase 8, PARP and β-actin.

Figure 8. Effects of Saf overexpression on the alternative splice forms of Fas. (A) Exon–intron structure and alternative splice forms at the Fas/Saf locus. The boxes and thin lines represent Fas exons and introns, respectively. TM, transmembrane domain and DD, intracellular death domain. The thick line represents Saf at the bottom in black. The hatched areas indicate the parts of intron 1 sequences that are retained in the four splice forms (b, c, d and e) of Fas and these four variants overlap with Saf by 48, 88, 105 and 105 bp, respectively. The primers are shown in their respective positions and orientations (arrows). (B) RT–PCR with specific primer pairs was used to identify Fas variant transcripts expressed in parental Jurkat cells, vector-transfectant and Saf-transfectant. Transcripts of a/b forms, c/d/e forms, membrane forms, soluble forms and GAPDH were PCR amplified with primer pairs of Fas1/Fas2, Fas3/Fas4, Fas5/Fas6 and GAPDH1/GAPDH2, respectively. Reverse transcriptase negative controls (RTase−) for each sample were also included. (C) Quantification of Fas b/c/d/e forms by densitometry. The densitometry ratio changes in the expression of the b/c/d/e forms of parental Jurkat, vector-transfectant and Saf-transfectant cells were determined after normalization with GAPDH.

Figure 8. Effects of Saf overexpression on the alternative splice forms of Fas. (A) Exon–intron structure and alternative splice forms at the Fas/Saf locus. The boxes and thin lines represent Fas exons and introns, respectively. TM, transmembrane domain and DD, intracellular death domain. The thick line represents Saf at the bottom in black. The hatched areas indicate the parts of intron 1 sequences that are retained in the four splice forms (b, c, d and e) of Fas and these four variants overlap with Saf by 48, 88, 105 and 105 bp, respectively. The primers are shown in their respective positions and orientations (arrows). (B) RT–PCR with specific primer pairs was used to identify Fas variant transcripts expressed in parental Jurkat cells, vector-transfectant and Saf-transfectant. Transcripts of a/b forms, c/d/e forms, membrane forms, soluble forms and GAPDH were PCR amplified with primer pairs of Fas1/Fas2, Fas3/Fas4, Fas5/Fas6 and GAPDH1/GAPDH2, respectively. Reverse transcriptase negative controls (RTase−) for each sample were also included. (C) Quantification of Fas b/c/d/e forms by densitometry. The densitometry ratio changes in the expression of the b/c/d/e forms of parental Jurkat, vector-transfectant and Saf-transfectant cells were determined after normalization with GAPDH.

Table 1.

Sequences of the primers

Primer Sequence 
Saf51 5′-
ATG GGA GAT ATG GGA TTG AAC
-3′ 
Saf52 5′-
TCT CCG CGG ATT CTC ACG
-3′ 
Saf53 5′-
CTG TAA GTC GCT GCC TGA GTG G
-3′ 
Saf54 5′-
CGG AGG GCT CGG AAG AA
-3′ 
Saf5Race 5′-
AGA AAA TTG TCA AAG CTT GGA GC
-3′ 
Saf31 5′-
CAT CTC AGC CTC TTG GTG TAA
-3′ 
Saf32 5′-
GAG ATT TTC AAA GCA GCC AAC AT
-3′ 
Saf33 5′-
CTC TGG TAA AGG TGC TAT TGG TA
-3′ 
Saf34 5′-
ACT AAC ACC AGG AAG GTT GAG
-3′ 
Saf3Race 5′-
TCC ATG GAC TCT CAG AGG CTC
-3′ 
Fas1 5′-
ACG GAG TTG GGG AAG CTC TT
-3′ 
Fas2 5′-
TGT CAG TCA CTT GGG CAT TAA CA
-3′ 
Fas3 5′-
CGG AGG GCT CGG AAG AAC G
-3′ 
Fas4 5′-
TGG TGA GTG TGC ATT CCT TGA T
-3′ 
Fas5 5′-
TCA TAC AGG AAT GCA CAC TC
-3′ 
Fas6 5′-
ACT TTC TGT TCT GCT TGT TC
-3′ 
FasQ1 5′-
ATG CTG GGC ATC TGG ACC CTC CTA
-3′ 
FasQ2 5′-
AGT CTT AAA GAG AGA ATT CCC GGA
-3′ 
GAPDH1 5′-
GCA CCG TCA AGG CTG AGA A
-3′ 
GAPDH2 5′-
AGG TCC ACC ACT GAC ACG TTG
-3′ 
GAPDHQ1 5′-
TGG CGC TGA GTA CGT CGT G
-3′ 
GAPDHQ2 5′-
ATG GCA TGG ACT GTG GTC AT
-3′ 
Primer Sequence 
Saf51 5′-
ATG GGA GAT ATG GGA TTG AAC
-3′ 
Saf52 5′-
TCT CCG CGG ATT CTC ACG
-3′ 
Saf53 5′-
CTG TAA GTC GCT GCC TGA GTG G
-3′ 
Saf54 5′-
CGG AGG GCT CGG AAG AA
-3′ 
Saf5Race 5′-
AGA AAA TTG TCA AAG CTT GGA GC
-3′ 
Saf31 5′-
CAT CTC AGC CTC TTG GTG TAA
-3′ 
Saf32 5′-
GAG ATT TTC AAA GCA GCC AAC AT
-3′ 
Saf33 5′-
CTC TGG TAA AGG TGC TAT TGG TA
-3′ 
Saf34 5′-
ACT AAC ACC AGG AAG GTT GAG
-3′ 
Saf3Race 5′-
TCC ATG GAC TCT CAG AGG CTC
-3′ 
Fas1 5′-
ACG GAG TTG GGG AAG CTC TT
-3′ 
Fas2 5′-
TGT CAG TCA CTT GGG CAT TAA CA
-3′ 
Fas3 5′-
CGG AGG GCT CGG AAG AAC G
-3′ 
Fas4 5′-
TGG TGA GTG TGC ATT CCT TGA T
-3′ 
Fas5 5′-
TCA TAC AGG AAT GCA CAC TC
-3′ 
Fas6 5′-
ACT TTC TGT TCT GCT TGT TC
-3′ 
FasQ1 5′-
ATG CTG GGC ATC TGG ACC CTC CTA
-3′ 
FasQ2 5′-
AGT CTT AAA GAG AGA ATT CCC GGA
-3′ 
GAPDH1 5′-
GCA CCG TCA AGG CTG AGA A
-3′ 
GAPDH2 5′-
AGG TCC ACC ACT GAC ACG TTG
-3′ 
GAPDHQ1 5′-
TGG CGC TGA GTA CGT CGT G
-3′ 
GAPDHQ2 5′-
ATG GCA TGG ACT GTG GTC AT
-3′ 

References

1
Maurillo, L., Del Poeta, G., Venditti, A., Buccisano, F., Battaglia, A., Santinelli, S., Caravita, T., Epiceno, A.M., Del Moro, B., Tamburini, A. et al. (
2001
) Quantitative analysis of Fas and bcl-2 expression in hematopoietic precursors.
Haematologica
 ,
86
,
237
–243.
2
Kramer, C., Loros, J.J., Dunlap, J.C. and Crosthwaite, S.K. (
2003
) Role for antisense RNA in regulating circadian clock function in
Neurospora crassa. Nature
 ,
421
,
948
–952.
3
Fadeel, B., Orrenius, S. and Zhivotovsky, B. (
1999
) Apoptosis in human disease: a new skin for the old ceremony?
Biochem. Biophys. Res. Commun.
 ,
266
,
699
–717.
4
Shin, M.S., Park, W.S., Kim, S.Y., Kim, H.S., Kang, S.J., Song, K.Y., Park, J.Y., Dong, S.M., Pi, J.H., Oh, R.R. et al. (
1999
) Alterations of Fas (Apo-1/CD95) gene in cutaneous malignant melanoma.
Am. J. Pathol.
 ,
154
,
1785
–1791.
5
Barnhart, B.C., Alappat, E.C. and Peter, M.E. (
2003
) The CD95 type I/type II model.
Semin. Immunol.
 ,
15
,
185
–193.
6
Yelin, R., Dahary, D., Sorek, R., Levanon, E.Y., Goldstein, O., Shoshan, A., Diber, A., Biton, S., Tamir, Y., Khosravi, R. et al. (
2003
) Widespread occurrence of antisense transcription in the human genome.
Nat. Biotechnol.
 ,
21
,
379
–386.
7
Rosok, O. and Sioud, M. (
2004
) Systematic identification of sense–antisense transcripts in mammalian cells.
Nat. Biotechnol.
 ,
22
,
104
–108.
8
Lehner, B., Williams, G., Campbell, R.D. and Sanderson, C.M. (
2002
) Antisense transcripts in the human genome.
Trends Genet.
 ,
18
,
63
–65.
9
Kelley, R.L. and Kuroda, M.I. (
2000
) Noncoding RNA genes in dosage compensation and imprinting.
Cell
 ,
103
,
9
–12.
10
Sleutels, F., Barlow, D.P. and Lyle, R. (
2000
) The uniqueness of the imprinting mechanism.
Curr. Opin. Genet. Dev.
 ,
10
,
229
–233.
11
Kumar, M. and Carmichael, G.G. (
1998
) Antisense RNA: function and fate of duplex RNA in cells of higher eukaryotes.
Microbiol. Mol. Biol. Rev.
 ,
62
,
1415
–1434.
12
Bartsch, H., Voigtsberger, S., Baumann, G., Morano, I. and Luther, H.P. (
2004
) Detection of a novel sense–antisense RNA-hybrid structure by RACE experiments on endogenous troponin I antisense RNA.
RNA
 ,
10
,
1215
–1224.
13
Suzuki, M. and Hayashizaki, Y. (
2004
) Mouse-centric comparative transcriptomics of protein coding and non-coding RNAs.
Bioessays
 ,
26
,
833
–843.
14
Timmer, T., de Vries, E.G. and de Jong, S. (
2002
) Fas receptor-mediated apoptosis: a clinical application?
J. Pathol.
 ,
196
,
125
–134.
15
Holzelova, E., Vonarbourg, C., Stolzenberg, M.C., Arkwright, P.D., Selz, F., Prieur, A.M., Blanche, S., Bartunkova, J., Vilmer, E., Fischer, A. et al. (
2004
) Autoimmune lymphoproliferative syndrome with somatic Fas mutations.
N. Engl. J. Med.
 ,
351
,
1409
–1418.
16
Houston, A. and O'Connell, J. (
2004
) The Fas signalling pathway and its role in the pathogenesis of cancer.
Curr. Opin. Pharmacol.
 ,
4
,
321
–326.
17
Muschen, M., Warskulat, U. and Beckmann, M.W. (
2000
) Defining CD95 as a tumor suppressor gene.
J. Mol. Med.
 ,
78
,
312
–325.
18
Vanhee-Brossollet, C. and Vaquero, C. (
1998
) Do natural antisense transcripts make sense in eukaryotes?
Gene
 ,
211
,
1
–9.
19
Reis, E.M., Nakaya, H.I., Louro, R., Canavez, F.C., Flatschart, A.V., Almeida, G.T., Egidio, C.M., Paquola, A.C., Machado, A.A., Festa, F. et al. (
2004
) Antisense intronic non-coding RNA levels correlate to the degree of tumor differentiation in prostate cancer.
Oncogene
 ,
23
,
6684
–6692.
20
Haddad, F., Bodell, P.W., Qin, A.X., Giger, J.M. and Baldwin, K.M. (
2003
) Role of antisense RNA in coordinating cardiac myosin heavy chain gene switching.
J. Biol. Chem.
 ,
278
,
37132
–37138.
21
Krystal, G.W., Armstrong, B.C. and Battey, J.F. (
1990
) N-myc mRNA forms an RNA–RNA duplex with endogenous antisense transcripts.
Mol. Cell. Biol.
 ,
10
,
4180
–4191.
22
Lease, R.A. and Belfort, M. (
2000
) A trans-acting RNA as a control switch in Escherichia coli: DsrA modulates function by forming alternative structures.
Proc. Natl Acad. Sci. USA
 ,
97
,
9919
–9924.
23
Tasheva, E.S. and Roufa, D.J. (
1995
) Regulation of human RPS14 transcription by intronic antisense RNAs and ribosomal protein S14.
Genes Dev.
 ,
9
,
304
–316.
24
Munroe, S.H. (
2004
) Diversity of antisense regulation in eukaryotes: multiple mechanisms, emerging patterns.
J. Cell. Biochem.
 ,
93
,
664
–671.
25
Luther, H.P., Podlowski, S., Hetzer, R. and Baumann, G. (
2001
) Analysis of sense and naturally occurring antisense transcripts of myosin heavy chain in the human myocardium.
J. Cell. Biochem.
 ,
80
,
596
–605.
26
Shendure, J. and Church, G.M. (
2002
) Computational discovery of sense–antisense transcription in the human and mouse genomes.
Genome Biol.
 ,
3
, research0044.1–0044.14.
27
Kiyosawa, H., Yamanaka, I., Osato, N., Kondo, S. and Hayashizaki, Y. (
2003
) Antisense transcripts with FANTOM2 clone set and their implications for gene regulation.
Genome Res.
 ,
13
,
1324
–1334.
28
Chen, J., Sun, M., Kent, W.J., Huang, X., Xie, H., Wang, W., Zhou, G., Shi, R.Z. and Rowley, J.D. (
2004
) Over 20% of human transcripts might form sense–antisense pairs.
Nucleic Acids Res.
 ,
32
,
4812
–4820.
29
Kondo, N., Ishii, Y., Son, A., Sakakura-Nishiyama, J., Kwon, Y.W., Tanito, M., Nishinaka, Y., Matsuo, Y., Nakayama, T., Taniguchi, M. et al. (
2004
) Cysteine-dependent immune regulation by TRX and MIF/GIF family proteins.
Immunol. Lett.
 ,
92
,
143
–147.
30
Sazani, P. and Kole, R. (
2003
) Therapeutic potential of antisense oligonucleotides as modulators of alternative splicing.
J. Clin. Invest.
 ,
112
,
481
–486.
31
Livak, K.J. and Schmittgen, T.D. (
2001
) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method.
Methods
 ,
25
,
402
–408.